PRELIMINARY DRAFT                                      ECAO-R-053
DO NOT QUOTE OR CITE                                   JULY  1988
                                                       WORKSHOP REVIEW  DRAFT
              SUMMARY REVIEW OF HEALTH EFFECTS ASSOCIATED WITH
        MANMADE MINERAL FIBERS OTHER THAN ASBESTOS,  SYNTHETIC FIBERS,
               AND SELECTED MINERALS:   HEALTH ISSUE  ASSESSMENT
                                  NOTICE

         This  document  is  a preliminary  draft and .is  intended
         for internal Agency  use  only.   It has not been formally
         released  by the  U.S.   Environmental   Protection  Agency
         and should  not  at this stage be  construed  to represent
         Agency  policy.   It is  being circulated  for  comments  on
         its technical merit and policy implications.
                ENVIRONMENTAL CRITERIA AND ASSESSMENT OFFICE
                OFFICE OF HEALTH AND ENVIRONMENTAL ASSESSMENT
                    U.S. ENVIRONMENTAL PROTECTION AGENCY
                      RESEARCH TRIANGLE PARK, NC  27711

-------
                                  DISCLAIMER

     This document  is  a  preliminary draft and is intended for internal  Agency
use only.   It  has  not  been formally released  by  the U.S.  Environmental  Protec-
tion Agency and should not at this stage be construed to  represent  Agency
policy.   It  is  being circulated for comments  on  its technical merit and policy
implications.  Mention of  trade names or commercial  products does not consti-
tute endorsement or recommendation for use.
                                       ii

-------
                                   CONTENTS

                                                                        Page

LIST OF FIGURES 	       x
LIST OF TABLES	      xi
AUTHORS, CONTRIBUTORS, AND REVIEWERS 	      xv

1.  SUMMARY 	     1-1
    1.1  MANMADE MINERAL FIBERS	     1-2
         1.1.1  Fibrous Glass	     1-2
         1.1.2  Mineral Wool 	     1-9
         1.1.3  Ceramic Fibers 	     1-11
    1.2  SYNTHETIC FIBERS	     1-13
         1.2.1  Aramid Fibers 	     1-13
         1.2.2  Carbon Fibers 	     1-15
         1.2.3  Polyolefin Fibers	     1-17
    1.3  MINERALS 	     1-18
         1.3.1  Attapulgite	     1-19
         1.3.2  Erionite	     1-21
         1.3.3  Talc 	     1-23
         1.3.4  Vermiculite	 1-25
    1.4  REFERENCES 	:	     1-27
2.  BACKGROUND INFORMATION 	     2-1
    2.1  PHYSICAL AND CHEMICAL PROPERTIES 	     2-1
         2.1.1  Manmade Mineral Fibers	'	     2-2
                2.1.1.1  Fibrous Glass 	     2-2
                2.1.1.2  Mineral Wool 	     2-3
                2.1.1.3  Ceramic Fibers	     2-4
         2.1.2  Synthetic Fibers 	:	     2-5
                2.1.2.1  Aramid Fibers	     2-5
                2.1.2.2  Carbon Fibers	     2-6
                2.1.2.3  Polyolefin Fibers 	     2-7
         2.1.3  Minerals	     2-7
                2.1.3.1  Attapulgite	•	     2-7
                2.1.3.2  Zeolites 	     2-8
                         2.1.3.2.1  Erionite 	     2-9
                2.1.3.3  Talc 	     2-9
                2.1.3.4  Vermiculite	     2-10
    2.2  PRODUCTION AND INDUSTRIAL USES 	     2-10
         2.2.1  Manmade Mineral Fibers 	     2-10
                2.2.1.1  Fibrous Glass 	     2-12
                2.2.1.2  Mineral Wool 	     2-12
                2.2.1.3  Ceramic Fibers 	     2-12
         2.2.2  Synthetic Fibers 	     2-14
                2.2.2.1  Aramid Fibers 	     2-14
                2.2.2.2  Carbon Fibers 	     2-16
                2.2.2.3  Polyolefin Fibers	     2-18
         2.2.3  Minerals 	     2-19
                2.2.3.1  Attapulgite 	     2-19
                2.2.3.2  Zeolites	     2-21
                2.2.3.3  Talc 	     2-21
                2.2.3.4  Vermiculite 	     2-22

                                      iii

-------
                         CONTENTS (continued)

                                                                    Page

2. 3  REFERENCES	     2-24
SOURCES OF EMISSIONS, AMBIENT LEVELS,  AND  ENVIRONMENTAL FATE ...     3-1
3.1  SOURCES OF EMISSIONS	 •    3-1
3.2  EXPOSURE 	     3-2
     3.2.1  Manmade Mineral  Fibers	     3-3
     3.2.2  Synthetic Fibers 	     3-8
     3.2.3  Minerals 	     3-9
3.3  AMBIENT LEVELS	     3-13
3.4  ENVIRONMENTAL FATE 	     3-16
3.5  ANALYTICAL METHODS 	     3-17
3.6  REFERENCES 	     3-21
RETENTION AND B10D1SPOSIT10N 	     4-1
4.1  INTRODUCTION 	     4-1
     4.1.1  Host Entry and Retention	     4-1
     4.1.2  Translocation and Clearance  	     4-6
4.2  MANDMADE MINERAL FIBERS 	     4-7
     4.2.1  Glass Fibers 	     4-7
            4.2.1.1  Host Entry and Retention	     4-7
            4.2.1.2  Translocation and Clearance  	     4-11
     4.2.2  Mineral Wool 	     4-16
            4.2.2.1  Rock Wool 	     4-17
     4.2.3  Ceramic Fibers	     4-18
4.3  SYNTHETIC FIBERS 	     4-20
     4.3.1  Aramid Fibers 	     4-20
     4.3.2  Carbon Fibers 	     4-21
     4.3.3  Polyolefin Fibers	     4-23
4.4  MINERALS 	     4-23
     4.4.1  Erionite 	     4-23
     4.4.2  Attapulgite 	     4-24
     4.4.3  Talc	     4-25
     4.4.4  Vermiculite	     4-29
4.5 REFERENCES	     4-29
ACUTE, SUBCHRONIC, AND CHRONIC TOXICITY  IN ANIMALS  	     5-1
5.1  ACUTE TOXICITY	     5-1
     5.1.1  Manmade Mineral  Fibers 	'..     5-1
            5.1.1.1  Fibrous Glass 	     5-1
            5.1.1.2  Mineral Wool 	     5-2
            5.1.1.3  Ceramic Fibers 	     5-2
     5.1.2  Synthetic Fibers 	     5-2
            5.1.2.1  Aramid Fibers 	     5-2
            5.1.2.2  Carbon and Polyolefin Fibers	     5-4
     5.1.3  Minerals 	     5-5
            5.1.3.1  Attapulgite 	     5-5
            5.1.3.2  Erionite 	     5-5
            5.1.3.3  Talc 	     5-5
            5.1.3.4  Vermiculite 	     5-6
5.2  SUBCHRONIC AND CHRONIC TOXICITY 	     5-6
     5.2.1  Mandmade Mineral Fibers 	     5-6

-------
                         CONTENTS (continued)

                                                                    Paoe

            5.2.1.1  Fibrous Glass	     5-6
                     5.2.1.1.1  Oral  	     5-6
                     5.2.1.1.2  Inhalation 	     5-6
                     5.2.1.1.3  Intratracheal  	     5-18
                     5.2.1.1.4  Intrapleural  	     5-20
            5.2.1.2  Mineral Wool 	     5-20
            5.2.1.3  Ceramic Fibers 	     5-22
            5.2.1.4  Other Manmade Mineral Fibers	     5-25
     5.2.2  Synthetic Fibers 	     5-28
            5.2.2.1  Aramid Fibers 	     5-28
            5.2.2.2  Carbon Fibers	     5-29
            5.2.2.3  Polyolefin Fibers	     5-30
     5.2.3  Minerals 	     5-31
            5.2.3.1  Attapulgite	     5-31
            5.2.3.2  Erionite and Other Zeolites	     5-31
            5.2.3.3  Talc 	     5-32
            5.2.3.4  Vermiculite 	     5-35
5.3  REFERENCES 	     5-35
DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 	     6-1
6.1  MANMADE MINERAL FIBERS 	     6-1
6.2  SYNTHETIC FIBERS 	     6-2
6.3  MINERALS	     6-2
6.4  REFERENCES	     6-2
MUTAGENICITY, CYTOTOXICITY, AND CARCINOGENICITY 	     7-1
7.1  MUTAGENICITY 	     7-1
                                                                    7-1
                                                                    7-2
                                                                    7-2
                                                                    7-2
                                                                    7-2
                                                                    7-2
                                                                    7-2
                                                                    7-3
                                                                    7-3
                                                                    7-3
7.2  IN VITRO CYTOTOXICITY	     7-4
     772.1  Introduction	     7-4
            7.2.1.1  Procedures Available To Measure
                     Cytotoxicity 	     7-4
            7.2.1.2  Test Material Composition	     7-5
     7.2.2  Manmade Mineral Fibers	     7-6
            7.2.2.1  Glass Fibers 	     7-6
            7.2.2.2  Mineral Wool/Rock Wool/Slag Wool	     7-13
            7.2.2.3  Ceramic Fibers 	     7-14
     7.2.3  Synthetic Fibers 	     7-14
            7.2.3.1  Aramid Fibers	     7-14
            7.2.3.2  Carbon Fibers	     7-15
            7.2.3.3  Polyolefin Fibers 	     "7-15
7.1.1
7.1.2
7.1.3
7.1.4
7.1.5
IN VIT
Manmade Mineral Fibers 	
7.1.1.1 Fibrous Glass 	
7.1.1.2 Mineral Wool 	
7.1.1.3 Ceramic Fibers 	
Synthetic Fibers 	
Minerals 	 ...... 	
7.1 3.1 Erionite 	
7.1.3.2 Attapulgite 	
Talc 	
Vermiculite 	 	 	
RO CYTOTOXICITY 	

-------
                             CONTENTS  (continued)

                                                                        Pace

         7.2.4  Minerals	     7-16
                7.2.4.1  Attapulgite/Sepiolite  	     7-16
                7.2.4.2  Erionite 	     7-17
         7.2.5  Talc 	     7-1E
         7.2.6  Venrn'culite	     7-1S
        ,7.2.7  Mode of Action	     7-19
                7.2.7.1  Cellular Responses	     7-19
                         7.2.7.1.1  Development of  ferruginous
                                    bodies 	     7-19
                         7.2.7.1.2  Effects  of  mineral  fibers on
                                    cell  membranes  	     7-20
                         7.2.7.1.3  Release  of  superoxide  	     7-21
                         7.2.7.1.4  Development of  proteinosis  	     7-22
                7.2.7.2  Physical Properties of Fibers  	     7-22
    7.3  CARCINDGENICITY 	     7-25
         7.3.1  Manmade Mineral Fibers 	     7-25
                7.3.1.1  Fibrous Glass	     7-25
                         7.3.1.1.1  Inhalation	     7-25
                         7.3.1.1.2  Intratracheal  	     7-2E
                         7.3.1.1.3  Intrapleural  	     7-30
                         7.3.1.1.4  Intraperitoneal  	     7-32
                7.3.1.2  Mineral Wool  	     7-34
                7.3.1.3  Ceramic Fibers 	     7-35
         7.3.2  Synthetic Fibers	     7-41
                7.3.2.1  Aramid and Polyolefin  Fibers 	     7-41
                7.3.2.2  Carbon Fibers	     7-43
         7.3.3  Minerals 	     7-43
                7.3.3.1  Attapulgite	     7-43
                7.3.3.2  Erionite and  Other  Zeolites	     7-44
                7.3.3.3  Talc 	     7-45
                7.3.3.4  Vermiculite 	'.	     7-49
         7.3.4  Other Fibers 	     7-49
         7.3.5  Summary	     7-64
    7.4  REFERENCES 	     7-66
8.   EFFECTS ON HUMANS	     6-1
    8.1  MANMADE MINERAL FIBERS 	:.     6-1
         8.1.1  Glass Fibers and Mineral  Wool 	     8-1
                8.1.1.1  Cross-Sectional  Studies,  Clinical
                         Evaluations,  and Case  Reports	     8-1
                         8.1.1.1.1  Eye and  skin irritation	     8-1
                         8.1.1.1.2  Respiratory effects	     8-4
                8.1.1.2  Epidemiological  Studies	     8-9
                         8.1.1.2.1  Epidemiological  studies in
                                    U.S.  workers 	     8-10
                         8.1.1.2.2  Epidemiological  studies in
                                    Canadian workers 	     6-29
                         8.1.1.2.3  Epidemiological  studies in
                                    European workers 	
                                      VI

-------
                         CONTENTS (continued)

                                                                     Page

                     8.1.1.2.4  Epidemiologies'!  study  combining
                                U.S.,  Canadian,  and European
                                workers	      8* 73
              r       8.1.1.2.5  Integration of'tiuman health
                                effects data by  fiber  and  by
                                carcinogenic and noncarcino-
                                genic  effects	      8-53
     8.1.2  Ceramic Fibers 	      8-102
            8.1.2.1  Cross-Sectional Studies,  Clinical
                     Evaluations, and  Case Reports 	      6-1C2
                     8.1.2.1.1  Respiratory effects	      8-1C2
8.2  SYNTHETIC FIBERS 	      8-102
     8.2.1  Aramid Fibers 	      8-1C2
     8.2.2 ^Carbon Fibers 	      6-102
            8.2.2.1  Cross-Sectional Studies,  Clinical
                     Evaluations, and Case Reports 	      £-102
                     8.2.2.1.1  Respiratory effects	      E-1D3
                     8.2.2.1.2  Effects from surgical
                                introduction 	      £-102
     8.2.3:  Polyolefin Fibers 	      £-iC2
6.3  MINERALS .".		      6-102
     8.3.1  Attapulgite and Sepiolite  	      8-102
            8.3.1.1  Cross-Sectional Studies,  Clinical
          :           Evaluations, and  Case Reports 	      8-103
                     8.3.1.1.1  Respiratory effects	      8-103
          ;, 8.3.1.2  Epidemiological Studies 	      8-105
     8.3.2  Erionite 	      6-106
          I?  8.3.2.1  Cross-Sectional Studi-es,  Clinical
          a           Evaluations, and Case Reports 	      8-107
         f           8.3.2.1.1  Respiratory effects	      8-107
          i.           8.3.2.1.2  Immunological  effects	     6-117
          V8.3.2.2  Epidemiological Studies 	!	      8-118
     8. 3.3%7alc 	      8-116
          ''"8.3.3.1  Cross-Sectional Studies,  Clinical
                     Evaluations, and Case Reports 	      8-116
                     8.3.3.1.1  Respiratory effects	      8-116
                     8.3.3.1.2  Effects from dermal
                                appl ication	      8-121
                     8.3.3.1.3  Effects from intrapleural
                                instillation 	      8-122
                     8.3.3.1.4  Effects from surgical
            :                    introduction	      8-123
                     8.3.3.1.5  Effects from intravenous
                                administration.	      8-123
            8.3.3.2  Epidemiological Studies 	      6-124
     8.3.4  Vermiculite	      8-133
            8.3.4.1  Cross-Sectional Studies,  Clinical
                     Evaluations, and Case Reports  	      6-133
                                  vn

-------
                             CONTENTS (continued)

                                                                        Page

                         8.3.4.1.1  Respiratory  effects	     8-133
                8.3.4.2  Epidemiological  Studies 	     8-13*
    8.4  REFERENCES 	     8-135
9.   QUANTIFICATION OF TOXICOLOGICAL EFFECTS 	     9-1
    9.1  PROCEDURES FOR QUANTIFICATION OF TOXICOLOGICAL  EFFECTS  ....     9-1
         9.1.1  Noncarcinogenic Effects 	     9-1
         9.1.2  Carcinogenic Effects 	     9-3
                9.1.2.1  Studies in Humans	     9-3
                9.1.2.2  Studies in Animals	     9-4
    9.2  MANMADE MINERAL FIBERS	     9-6
         9.2.1  Human Health Effects 	     9-6
                9.2.1.1  Fibrous Glass	     9-7
                         9.2.1.1.1  Qualitative  evaluation  of
                                    noncarcinogenic effects 	     9-7
                         9.2.1.1.2  Characterization of  carcino-
                                    genic effects  	     9-9
                9.2.1.2  Mineral Wool 	     9-13
                         9.2.1.2.1  Qualitative  evaluation  of
                                    noncarcinogenic effects 	     9-13
                         9.2.1.2.2  Characterization of  carcino-
                                    genic effects  	     9-13
                9.2.1.3  Ceramic Fibers	     9-17
                9.2.1.4  Quantification of Carcinogenic  Effects
                         of Manmade Mineral Fibers 	     9-17
         9.2.2  Animal Toxicity	     9-20
                9.2.2.1  Characterization of Noncarcinogenic
                         Effects 	     S-20
                9.2.2.2  Quantification of Noncarcinogenic
                         Effects 	     9-21
                         9.2.2.2.1  Fibrous glass	     9-21
                         9.2.2.2.2  Mineral wool 	     9-24
                         9.2.2.2.3  Ceramic fibers 	     9-26
                9.2.2.3  Characterization of Carcinogenic Effects  ..     9-28
                         9.2.2.3.1  Fibrous glass  	     9-29
                         9.2.2.3.2  Mineral wool 	     9-30
                         9.2.2.3.3  Ceramic fibers 	     9-31
         9.2.3  Characterization of Carcinogenic Potential  Based
                on Human and Animal Data	     9-32
    9.3  SYNTHETIC FIBERS	     9-32
         9.3.1  Human Health Effects 	     9-32
         9.3.2  Animal Toxicity	     9-33
                9.3.2.1  Characterization of Noncarcinogenic
                         Effects	   9-33
                9.3.2.2  Quantification of Noncarcinogenic
                         Effects 	     9-33
                         9.3.2.2.1  Aramid fibers	     5-33
                         9.3.2.2.2  Carbon fibers 	     9-34
                         9.3.2.2.3  Polyolefin fibers	     9-34
                                     V111

-------
                         CONTENTS (continued)
            9.3.2.3  Characterization and Quantification  of
                     Carcinogenic Effects 	     9-34
                     9.3.2.3.1  Aramid fibers  	     9-34
                     9.3.2.3.2  Carbon fibers	     9-35
                     9.3.2.3.3  Polyolefin fibers	     9-35
     9.3.3  Characterization of Carcinogenic Potential  Based
            on Human and Animal Data	     9-36
9.4  MINERALS	     9-36
     9.4.1  Human Health Effects	     9-36
            9.4.1.1  Attapulgite	  .   9-36
                     9.4.1.1.1  Qualitative evaluation  of
                                noncarcinogenic  and  carcino-
                                genic effects  	     9-3£
            9.4.1.2  Erionite 	     9-37
                     9.4.1.2.1  Qualitative evaluation  of
                                noncarcinogenic  and  carcino-
                                genic, effects  	     9-37
            9.4.1.3  Talc 	     9-37
                     9.4.1.3.1  Qualitative evaluation  of
                                noncarcinogenic  and  carcino-
                                genic effects  	     9-37
            9.4.1.4  Vermiculite	     9-36
                     9.4.1.4.1  Qualitative evaluation  of
                                noncarcinogenic  and  carcino-
                                genic effects  	     9-36
     9.4.2  Animal Toxicity	     9-3S
            9.4.2.1  Quantification of Noncarcinogenic  Effects .     9-3B
                     9.4.2.1.1  Attapulgite 	     9-38
                     9.4.2.1.2  Erionite	     9-36
                     9.4.2.1.3  Talc 	     9-38
                     9.4.2.1.4  Vermiculite	'.	     9-39
            9.4.2.2  Characterization and Quantification  of
                     Carcinogenic Effects 	     9-39
                     9.4.2.2.1  Attapulgite 	     9-39
                     9.4.2.2.2  Erionite 	     9-39
                     9.4.2.2.3  Talc 	     9-40
                     9.4.2.2.4  Vermiculite 	     9-40
     9.4.3  Characterization of Carcinogenic Potential  Based
            on Human and Animal Data	     9-41
9.5  SUMMARY OF CARCINOGENIC AND NONCARCINOGENIC EVIDENCE 	     9-41
9.6  REFERENCES 	     9-43
                                  IX

-------
                                   FIGURES
                                                                       Page
        The relationship of glass fiber diameter and length  to
        aerodynamic equivalent diameter (D  ) 	    4-3
4-2     The relationship of alveolar deposition to the  diameter
        and length of glass fibers	    4-4
4-3     Bivariate distribution of fiber deposition in the  alveoli
        following nose-only breathing by humans 	    4-5

-------
                                    TABLES
Number                                                                   Page
 2-1     Mineral and vitreous fibers 	     2-2
 2-2     Major U.S.  producers of textile  glass  fibers  	     2-11
 2-3     U.S.  consumption of textile glass  fibers, 1983  	     2-12
 2*4     Major U.S.  alumina-silica producers  	     2-13
 2-5     U.S.  consumption of alumina-silica fiber products by
         industry,  1983 	     2-13
 2-6     Representative applications for  ceramic, alumina, zirconia
         and silica fiber products 	     2-15
 2-7     U.S.  demand for high-performance carbon fibers  	     2-17
 2-8     Major U.S.  producers of carbon fibers  	     2-17
 2-9     Major U.S.  producers of polyolefin fibers  	     2-20
 2-10    Fuller's earth sold or used by U.S.  producers in 1984,
         by use	     2-22
 2-11    U.S.  production of crude talc and  pyrophyllite  in 1984  	     2-23
 2-12    End uses for ground talc and pyrophyllite	     2-23
 3-1     Releases of vermiculite into the environment	     3-2
 3-2     Airborne concentrations of manmade mineral  fibers in
         manufacturing plants 	     3-5
 3-3     NOHS estimates of exposure to mineral  fibers  	     3-7
 3-4     Occupational exposure to airborne  concentrations of
         asbestos in processing plants 	     3-11
 3-5     Summary of occupational and nonoccupational  inhalation
         exposure to asbestos In vermiculite  	     3-12
 3-6     Asbestiform fiber measurement methods  	     3-15
 4-1     The number and size of nonasbestos fibers  in  the  lungs
         of 20 subjects 	     4-25
                                         £
 5-1     Mass and concentrations of Kevlar   fibers  used in an
         inhalation study in male rats 	     5-3

-------
                              TABLES (continued)


Number                                                                   Page

 5-2     The histologic grading system of pulmonary  changes  in rats
         related to the inhalation of fibers  	     5-10

 5-3     Dimensions and concentrations of glass fibers  used  in a
         chronic inhalation study with F344  rats  and cynomolgus
         monkeys 	,	     5-12

 5-4     Dimensions, aerosol  concentrations,  and  particle  ratios of
         fibers used in 2-year inhalation studies with  Osborne-
         Mendel rats and Syrian hamsters  	     5-15

5-5      Incidences of pulmonary lesions  noted in Osborne-Mendel
         (OK) rats and Syrian hamsters exposed to various  glass
         fibers by inhalation for 24 months  	     5-16

 5-6     Type, diameter, and dose of glass fibers intratracheally
         injected in Syrian hamsters 	     5-18

 7-1     Size distribution of some asbestiform fibers used in
         ID vitro cytotoxicity studies 	     7-9

 7-2     Neoplasms and hyperplasia in the lungs of rats after
         inhalation exposure to various MMMF  	     7-27

 7-3     Incidence of mononuclear cell leukemia in F344 rats
         exposed to fibrous glass insulation  via  inhalation  for
         2 years 	     7-26

 7-4     Tumors in Syrian hamsters after intratracheal-instillation
         of glass fibers or crocidolite 	     7-30

 7-5     Size distribution, tumor incidence,  and  lifespan
         following intraperitoneal administration of manmade and
         naturally occurring fibers to female Wistar rats	     7-25

 7-6     Size distribution and number of manmade  and naturally
         occurring fibers administered intrapen'toneally to  female
         Wistar rats 	     7-36

 7-7     Lifespan and tumor incidences in female  Wistar rats after
         intraperitoneal administration of manmade and naturally
         occurring fibers 	     7-36

 7-8     Incidences of lung mesotheliomas in rats after intrapleural
         instillation of various MMMF 	     "
                                      XI 1

-------
TABLES (continued)
Number
7-9

7-10

7-11

7-12

7-13

8-1

8-2

8-3

8-4

8-5

6-6

8-7
8-8

8-9

8-10
8-11

Peritoneal tumor incidences in mice after intraperitoneal
injection of Karain dust 	
Selected data on carcinogenicity of different fibrous
materials following intrapleural implantation in rats 	
Incidence of pleural tumors induced in F344 rats by
ferroactinolite and amosite 	
Sizes of mineral fibers retained in the lungs of rats
at various intervals after intratracheal instillation 	
Summary of carcinogenicity studies on MMMF, synthetic
fibers, and minerals 	
Significantly elevated SMRs by cause for all MMMF workers
between 1978 and 1982 	
Significantly elevated SMRs by cause for all MMMF workers
between 1946 and 1982 	
Significantly decreased SMRs by cause for all MMMF workers
between 1946 and 1982 	
Cancer incidence analysis by site and duration of
exposure 	
Selected cause-specific mortality for production workers
from the U.S., Canadian, European, and combined cohorts ....
Comparison between U.S. and European SMRs for lung cancer
derived from national and local rates 	
Lung cancer mortality by process 	 	 .
Lung cancer mortality among the three combined cohorts by
process and time since first exposure 	
Respiratory cancer mortality by exposure to small-diameter
glass wool fibers and time since first employment 	
Selected cancer mortality for U.S. and European cohorts 	
Mortality from NMRD by cohort 	
Page

7-4£

7-52

7-54

7-54

7-55

8- IS

e-19

8-20

8-71

8-75

8-76
8-77

8-77

8-78
8-80
8-81
       xm

-------
                              TABLES (continued)


Number                                                                    Page

 8-12    Summary of carcinogenic and noncarcinogenic effects from
         historical prospective morbidity and mortality studies  in
         workers exposed to MMMF	      8-84

 9-1     Summary of the weight of evidence for causal relationships
         between exposure to MMMF and carcinogenic and noncarcino-
         genic effects 	      9-8

 9-2     Summary of long-term inhalation studies considered for
         derivation of the RfD for fibrous glass	      9-22

 9-3     Summary of long-term inhalation studies considered for
         derivation of the RfD for rock wool 	      9-25

 9-4     Summary of long-term inhalation studies considered for
         derivation of the RfD for ceramic fibers 	      £-27

 S-5     Sijfr.-ary of carcinogenic and noncarcinogenic effects from
         animal and human data 	      9-42
                                      xiv

-------
                     AUTHORS, CONTRIBUTORS,  AND REVIEWERS


     The following Dynamac  Corporation personnel were  involved in the prepara-
tion of  this  document:   Nicolas P. Hajjar,  Ph.D.  (Project  Manager/Principal
Author); Barrett N.  Fountos,  Claire Kruger-McOermott,  Ph.D., Patricia Turck,
Mary E. Cerny, Dawn Webb, Brion Cook,  Nancy McCarroll,  William McLellan,  Ph.D.,
Christian Alexander, John  Bruno,  Ph.D.,  Edward Flynn,  Charles Rothwell, Ph.D.,
Janice  Range,  and Sharon  Segal,  Ph.D.   (Authors); Karen  Swetlow (Technical
Editor); and Sanjivani  Diwan, Ph.D. and Gloria Fine (Information  Specialists).

     This document was  prepared by Dynamac Corporation under contract to the
Environmental   Criteria  and  Assessment  Office, Research  Triangle  Park,  NC
(Dennis J.  Kotchmar,  M.D.,  Project Manager).
                                       xv

-------
                                   1.   SUMMARY
      Concern surrounding the possible  inhalation  associated health effects of
 asbestos on the general population stimulated interest on the potential health
 effects of  other  natural  and  manmade  fibers.'  This document discusses the
 health effects of the  manmade  mineral  fibers (MMMFs) (fibrous glass, mineral
 wool,  and  ceramic  fibers),  the  synthetic  fibers  (aramid,  carbon,  and  polyolefin
'fibers), and the naturally  occurring  fibers  attapulgite  and erionite,  as well
 as the  health  effects  related to the  nonfibrous  silicate  minerals  talc and
 vermiculite, which have multiple  consumer  uses  and  have  been  shown to  contain
 various concentrations of  asbestiform  fibers.   Data for other fibers  a-'i 'cr
 minerals are also discussed,  primarily for comparative purposes, in various
 parts  of this  report.
      Chapter 2  of  this  document presents  background  information on physical ar.s
 chemical properties, production volumes, and  uses of each of the 10 fibers.
 The  sources of emissions,  ambient levels,  worker  and consumer exposure, envi-
 ronmental  fate, and analytical procedures  used to identify and quantitate each
 fiber  are  discussed in Chapter 3.  Chapter 4 discusses^the aerodynamic factors
 governing  host  entry,  retention,  and  biodisposition.   Chapter  5 provides acute,
 subchronic,  and chronic toxicity data for laboratory animals as the data relate
 to noncarcinogenic or  systemic  effects.  Two teratogenicity studies  are dis-
 cussed in  Chapter 6.   Mutagenicity,  cytotoxicity, mode of action, and animal
 carcinogenicity studies are presented  in Chapter  7.  Chapter 8 discusses human
 health effects data obtained from cross-sectional, clinical, and retrospective
 mortality  studies  and  case  reports.
      The carcinogenic   and/or noncarcinogenic effects of each fiber are charac-
 terized from both animal toxicity and human  health  effects data  in Chapter 9.
 Quantification  of  noncarcinogenic effects was  conducted  by  calculating  a
 Reference  Dose  (RfD)   based on systemic No-Observed-Adverse-  Effect Levels
 (NOAELs) or Lowest-Observed-Adverse-Effect Levels (LOAELs) from  chronic animal
 toxicity data.   When carcinogenicity  was  demonstrated, estimates  of carcinogenic

 i  i.  ir.rr,                           1.1         nPirr.-nn W.T n".-f- n;  r'~~

-------
risk were calculated.  The carcinogenic potential  of each fiber  was  categorizes
based on  evidence  from human and  animal  studies.  Summaries for each fiber are
presented in the following sections.
1.1  MANMADE MINERAL FIBERS
     MMMFs are  groups  of amorphous silicates that are not  discrete  entities
(such as  minerals)  definable by a fixed chemical  composition.   They  are  manu-
factured  by  various  techniques  from a molten mass of raw material, and their
properties depend on their chemical composition.

1.1.1  Fibrous Glass
     Fibrous  glass  is composed  primarily  of the oxides of silicon  (SiO.),
aluminum  (Al-O^), and  boron (ELO,).   The physical and chemical properties of
glass fibers  depend  on the aspects of their  chemical composition that can be
manipulated  during  production to  promote  specific properties.  Three  ma'ir,
classes of glass  fibers  are available:   textile, wool,  and fine fibers, witr,
typical fiber diameters  of 10 to  15 um, 3  to 10  urn, and <_! um, respectively.
Approximately 1200  and 3000 million pounds of textile fibers  and  insulaticr
glass wool,  respectively,  were  produced in 1983.   Textile  fibers  are produced
in various  forms, and  about 50 percent of  the  amount  produced is used in
reinforced plastics.   Approximately 80  percent  of the  insulation  glass  fibers
are  used  in  structural  insulation.   Fibrous glass wools and  filaments  are
relatively more water-soluble and  less  chemical- resistant than mineral  wool,
although  both are susceptible,  in varying degrees, to acid and base attack.
The  physical  changes  that  affect  mineral  fibers  are not homogeneous, since
considerable  variation  in  composition exists along the  length  of a  fiber and
between fibers.

Environmental Release and Fate
     Environmental release  of MMMFs  occurs during manufacture, handling,  use,
and  disposal.   However,  estimates  are  not available.  The  principal  route of
exposure  to  fibers  is  inhalation.   Measurements of airborne concentrations of
glass fibers  in the workplace vary greatly with  sampling  location within the
same plant  and  among  plants performing similar processing  operations.   The,
range of  average  concentrations  of fibrous glass  found  in  several plants was

-------
reported to  be  0.0024  to 0/78 total fibers/cm3 and  0.002  to  0.08  respirable
fibers/cm  (i.e., with nominal diameters <3 urn).
     The highest  fiber concentrations  were  detected in the  production  and
manufacturing areas.   Estimates  of  respirable  fiber concentrations in  the
European glass  wool  industry during  the early  technological  phase (see
section 3.2.1)  indicate  that they were similar to current concentrations  of
about 0.1 fibers/cm  or less.
     Occupational exposure  also  depends  upon  working habits, the processes
involved, physical  characteristics of  the  fibers, and the availability of
control measures.  The  National  Occupational Hazard  Survey (NOHS) provided the
following estimates  of  the  number of workers potentially exposed to MMMF  for
1972-74:  glass  fibers,  62,811;  glass  wool,  57; fiberglass insulation, 80,629.
The National  Occupational Exposure Survey (NOES)  reported no  information  for
these fibers.   Consumer  exposure to MMMF is expected to be low except during
periods when insulating material  is being handled, but estimates were not found
in the available literature.
     Limited  information  is available  on ambient levels  for glass  fibers,
although their  presence appears  to be ubiquitous.  Concentrations of glass
fibers found in ambient air samples collected in California were reported to  be
0.87 ± 0.48 fibers/L with an average diameter of 3.7 urn (range 0.1 to 17.7 prr).
A survey of several  schools for MMMF indicated concentrations  ranging frorr. 0  to
0.087 fibers/L,  whereas concentrations in office tuildings ranged from 0.380  to
10 fibers/L.  Mineral  fibers  have also been found in aquatic  environments and
drinking water supplies.
     This report  discusses  the  environmental  fate of mineral fibers with a
focus on glass  fibers  and compares the Information presented with data avail*
able on the  environmental  fate of asbestos  fibers.   Glass fibers  can  undergo
both chemical  and physical  changes under various environmental  conditions.
These include changes in fiber size, chemistry, and crystalline structure.   The
major environmental  fate of minerals and mineral fibers is probably deposition,
especially of longer fibers, and eventual burial in soils and sediments.
     Various methods have been used to analyze and quantitate fibers.  Earlier
methods measured mass,  and  x-ray diffraction techniques provided  identifica-
tion.   Counting  methods  are used for samples that are too  small  to be  weighed.
Light microscopy methods  are limited in  resolution.  Both scanning and trans-
mission electron  microscopy can  be used  with x-ray analytical methods  to

July 1988                          1-3        DRAFT--DO NOT QUOTE OR CITE

-------
provide  for better  mineralogic identification of  fibers.   Another recer.t
analytical method  is the LAMMA procedure, which combines a high-energy laser
beam with mass spectroscopic analysis.

Metabolism
     Various  factors influence the  deposition of  inhaled particles in  the
respiratory tract;  for  fibers, these include  the aerodynamic equivalent diame-
ter  (D.D)  and  aspect ratio.   Several studies  indicate that glass fibers with a
      fl C
D    larger than 3.5 urn  have  little  chance  of alveolar  deposition; most are
 OC
deposited  above  the  trachea  and in nasal  hairs.  In contrast, maximum alveolar
deposition  in  rats  and  dogs  occurs for  glass  fibers with an actual diameter of
<0.1 um  and a length of about 10 urn or  a  D    of  approximately 1 to 2 um.
However, only  a  small  percentage of these  fibers are retained  (7  to 16  per-
cent), and  most  are exhaled without being deposited.   Alveolar  deposition  was
reported to follow  a similar pattern in man,  but absolute volumes  are  greater
than those  observed  in  rats.   Alveolar deposition  declines steeply  to  about
3 percent of a D   of 3.4, i.e., a count median diameter of 1.1  pm.  The  rete~-
                o 6
tion of  glass fibers following repeated  or chronic-inhalation exposures  is
similar to  that  observed following single exposures.  Retention decreases  with
increasing  diameter  and  length, and increases with repeated dosing.   In  rats,
most of the retained fibers are found in the right upper and left  lobes  follow-
ing  long-term  exposure.   The  fibers retained  in the  lung were  less  than 5 um
long.  Female  rats  apparently tend to have slightly  higher tissue levels  of
                                                       •
fibers than males following chronic inhalation exposure.
     Several mechanisms  are  Involved in clearing fibrous materials  from  the
lung.  These  include removal  by mucociliary  activity, transport by  alveolar
macrophages to regional  lymph nodes, uptake by epithelial cells,  and  direct
translocation of  fibers  between the epithelial cells to the interstitium  and
the  pleura.  However, transport via the blood circulation has also been demon-
strated.    As  indicated   above,  microfibers  are readily eliminated following
single inhalation exposures.   However,  the  fibers retained following single or
repeated inhalation  exposure,  including those with larger diameters (<3 um),
are  cleared slowly  with a half-life of  about  44 days.  The half-life clearance
value following  chronic  exposure was about 600 days.  The rates of  clearance
following  intratracheal  administration  in rats and guinea pigs decrease with
increased  fiber  length  (^10  pm) and dose.  Short fibers  are phagocytized  by

July 1988                          1-4        DRAFT—DO NOT QUOTE  OR  CITE

-------
alveolar macrophages and  cleared  to the lymph nodes.   Long  fibers  are  usually
partially phagocytized.   Fiber  type and composition,  which  affect solubility
and fragmentation, also influence the rates of fiber retention  in  the lung.   In
this respect, glass  fibers  behave differently from amphibole asbestos  fibers,
which are essentially  insoluble u>  vitro and  split  longitudinally into thinner
fibers and  thus  are  not cleared as  readily from the  lungs.   Following intra-
pleural injection  in  rats,  fibers were found  in various  organs; translocation
apparently occurred via the blood capillaries.

Animal Toxicity
     Limited information  was  available  on  the acute toxicity of fibrous glass.
Fibrous glass causes  dermal  irritation in rabbits  and guinea  pigs  following
repeated application for 1 month.   Mild inflammatory reaction was  noted in male
rats, hamsters,  and guinea pigs following 20 days of exposure to fiberglass.
     Inhalation exposure  to  glass fibers  of varying diameters  and lengths for
up  to  2  years  generally  failed to  show any substantial  pulmonary damage  or
carcinogenic response  in  several  experimental species.   In  contrast,  asbestos
exposure results  in  collagen formation and development  of  pulmonary  tumors.
Pulmonary responses to glass fibers generally  consisted of macrophage  infiltra-
tion,  alveolar  dust-laden macrophage  aggregation,  and  alveolar  proteinosis
resulting in some  alteration of lung architecture  in  rats, guinea pigs,  and
hamsters.    The  tissue response  appeared  to  be  influenced  by the physical
dimensions of the fibers.   In general, exposure to long (>20 urn),  thick (>5 prr.)
fibers resulted  in a  minimal  reaction characterized 'by sparse  dust-laden
macrophage accumulation in  the alveoli.  By  contrast, exposure to long,  thin
(<3 urn) fibers resulted in alveolar proteinosis, development of reticulin  fiber
networks, and  some fibrosis.   Short (<5 urn), thin fibers  caused a severe
inflammatory response resulting in dense dust-laden macrophage  aggregation, and
in some cases, enlargement of adjacent  lymph nodes due to macrophage migration.
However, very little alteration of the  lung architecture was observed:
     When the  results of  these studies are  considered  collectively,  it  is
evident that inhalation of glass microfibers results in mild cellular reactions
at  exposure  levels of  <3.0 mg/m .   Higher concentrations of  microfibers  or
fibrous glass (diameter <3 um) result  in more pronounced pulmonary manifesta-
tions.  Using a grading system to score these  lesions, a NOAEL  of 3.0 mg/ir,  and
a LOAEL of  100  mg/m   were identified in rats  and hamsters for  microfibers  with

July 1988                           1-5         DRAFT-DO NOT QUOTE OR CITE

-------
diameters  <0.5 urn.   The RfD  is  0.03 mg/m3.   For  thicker  fibers (>0.5  to
<3.5 pin),  a  LOAEl  of 5 mg/m3 was identified in rats,  and  the RfD value was
calculated to be 0.005 mg/m .
     Intratracheal instillation of glass fibers  into several  species  results  in
pulmonary  manifestations similar  to  those observed following inhalation expo-
sure.   Very  short (<5 pm),  very  thin (<0.3 pm) fibers apparently cause  no
pulmonary  response.  Short  (<10 pm),  thin (<2 urn)  fibers  elicit simple  macro-
phage response.  Longer  (>10 pm)  or  thicker fibers (<3 ym)  result in a more
pronounced response  and  some fibrosis.   However, results from the same  studies
provide  equivocal  evidence on the  carcinogenic potential of glass  fibers
following  intratracheal  administration.   Results  from studies  with  rats  anci
hamsters  indicate  that microfibers (diameter £0.5 urn, length  >10 ym)  induce
pulmonary  tumors, whereas other studies with the same species exposed to fibers
with  similar dimensions provided negative results.   Unlike inhalation  or
intratracheal  instillation  into  the  lungs,  the direct application of  glass
fibers with varying diameters and lengths (and especially  microfibers)  onto  the
pleura!   surface  of  rats  and hamsters or the peritoneum of rats has  been shown
to result  in  the development of mesothelial  and lung tumors.  The tumorigenic
response was  more  pronounced and was elicited by  lower doses following  intra-
peritoneal administration  than  following administration via the  intrapleural
route.   Results from one study with BALB/c mice  indicate that glass  fibers were
not carcinogenic following  intrapleural  administration.  Similarly,  intraperi-
toneal administration  of glass microfibers to hamsters in another study  was
also negative.   Quantification of cancer risk could not be  determined  since
there are  no  methodologies  for the intraperitoneal or intrapleural  route, and
inhalation exposure was negative.   Fibrous glass is not mutagenic in  Salmonella
strains or Escherichia coVj WP 2 with or without S9 activation.   However,  these
systems  may  not be appropriate for  assaying  the  mutagenicity of fibers  and
other particles.   Long fibers (10 ym) of small diameter (0.13 yrn) cause  chromo-
somal aberrations  in  cultured Chinese hamster ovary cells at 0.01 mg/mL  and
induce transformation in cultured Syrian hamster embryo cells.
     Several   in  vitro  cytotoxicity tests have been developed to determine the
mechanism  of  action of mineral fibers  and to predict the fibrogenic and/or
carcinogenic   potential of  specific  fibers.   These include cultures  of  3T3 or
BHK21 fibroblasts;  guinea  pig,  rat,  and mouse  macrophages;  Chinese hamster
V79-4 lung cells;  Syrian  hamster embryo cells;  phagocytic epitheloid asciteB

July 1988                          1-6        DRAFT--DO NOT QUOTE OR  CITE

-------
tumor cells;  and a human tumor  line  of  type 11 alveolar cell origin.  The
cytotoxicity of  glass  fibers  has been confirmed in several studies.  Equivalent
masses and numbers of various glass and asbestos fibers  induced comparable cell
responses, but the  extent  of toxicity was primarily dependent on fiber size.
Long, thin glass fibers  (diameter  £0.3 urn and length 10 to 40 urn for  83 and
74  percent  of the  fibers,  respectively) and  long,  very thin glass fibers
(diameter <0.2 pm and length 10 to 60 urn for 98.1  and 50 percent, respectively)
were as cytotoxic  to  epitheloid ascites tumor cells, mouse peritoneal macro-
phages,  hamster  embryo cells,  and  guinea  pig,  rat, and  dog lung macrophages as
were similar-sized chrysotile,  crocidolite, amosite, and synthetic  fluoramphi-
boles.   Long,  thin asbestos  and glass fibers  also  reduced the phagocytic
ability of macrophages and  inhibited DNA, RNA, and  protein synthesis  in the
above-mentioned  cells.   These glass fibers (mean diameters between 0.12 and
1.9 pm)  also slowed the  growth of adult and fetal  human fibroblasts and human
bronchial  epithelial cells.   The cytotoxic behavior of the long, thin glass
fibers corresponds to  results of j_n vitro  studies in which rats and hamsters
treated intrapleurally and  rats  treated  intraperitoneally showed an increased
incidence of mesotheliomas.
     The  results of an  oral  teratogenicity study with  a  food-grade  sodiuir.
aluminosilicate  indicate  no clear  compound-related  effects on  reproductive
indexes, maternal and  fetal  toxicity,  or soft and skeletal tissue  abnormali-
ties.   The  fiber  was  administered  to pregnant Wistar rats, CD-I mice,
Dutch-belted rabbits,  and  Syrian Golden hamsters  beginning on day  6 of gesta-
tion for  9 to  13 consecutive days  at  doses up to 1600 mg/kg.   The  available
literature contained no information on reproductive effects.

Human Health Effects
     Skin and eye irritation have been associated with occupational  exposure to
fibrous glass.   Mechanical  irritation  of the skin and eye is  common,  although
immunologic effects  in a  small  number of  exposed workers have also been
observed.   Respiratory manifestations including bronchitis, pharyngitis, rhini-
tis, asthma, laryngitis, and sinusitis  have also been  associated with  exposure
to  glass  fibers.  These effects are apparently reversible when exposure to
fibrous glass  is discontinued.   Although no clear  evidence of pulmonary abnor-
malities among exposed workers was found  in  several  cross-sectional  studies,
small  lung  opacities  and  higher   incidences  of   chronic  and dysplastic

July 19BB                          1-7        DRAFT--DO NOT QUOTE OR CITE

-------
 pharyngolaryngitis  have been noted in  a  few workers exposed to fibrous  glass
 for more than 5 years.
      The weight of evidence  from  several  epidemiclogical studies for a caused
 relationship  between  exposure to  microfibers  having diameters  similar  to
 asbestos and  nonmalignant respiratory  disease  (NMRD) or respiratory cancer in
 U.S.  workers is inadequate.   The absence  of  risk is  limited by the small number
 of deaths and low power to detect  NMRD  in these studies.
      The weight of evidence  is  limited for  a causal relationship,  associated
 with latency  and duration of exposure,  between exposure  to glass wool  fibers
 and NMRD and  lung cancer  in  U.S., Canadian,  and European workers.  The  strong-
 est evidence  for a causal  relationship between exposure  to glass wool  fibers
 and NMRD comes  from a  study  conducted  in the  United States  (Bayliss et a!.,
•1976),  whereas  the  evidence  for  lung  cancer  comes  from  a Canadian study
 (Shannon et a!., 1986,  1987)  comparing  data  to  provincial mortality rates.
      Several studies  provide limited evidence  regarding a causal relationship
 between laryngeal cancer  in  European (especially Italian) workers and exposure
 to glass wool  fibers.   Although  only  nine laryngeal  cancer deaths were reports
 the strength  of  this  association  is  based on  the large  SMRs that  increased
 with increasing latency and  length of exposure, the significant findings after
 comparison with high local rates,  and  the identification of  cases in separate
 studies.   In  France,  a threefold  excess  in the  incidence  of buccal cavity
 cancer  was  noted among  production  workers with 10 to 19 or ^20 years of expo-
 sure when compared to  regional  rates.   However, since the evidence comes from
                                                       *
 only one  study,  the  weight  of  evidence  for a causal  relationship  between
 exposure to glass wool  ffbers and  buccal  cavity cancer is limited.
      The weight  of evidence  for a causal relationship between  NMRD  or  lung
 cancer  and  continuous  filament  (textile  fibers)  among exposed  workers  is
 inadequate.   The abse.rce of a risk is limited by the small number of deaths and
 low power to  detect NMRD  or  lung  cancer  in  several  U.S. and  European studies.
 A quantification of cancer risk  or NMRD based on epidemiological  evidence could
 not be  made because exposure levels  existing during the early technological
 phase were not available.
      Based on the weight  of  evidence from human health effects data and animal
 toxicity data  summarized  in  Chapter 9  (see Table  9-5),  the carcinogenicity
 classification for fibrous glass  is Bl  and for  glass microfibers  is B2.
 Julv  1988                           1-8        DRAFT--DO  NOT  QUOTE OR CITE

-------
1.1.2  Mineral Wool
     Mineral wools  also  consist predominantly of SiD2, A1203, and  B203,  but
they differ  from fibrous glass in that they  contain  greater  amounts of  CaO,
MgO, Fe203,  and  Ti02,  and 1esser amounts of  Ba  and the oxides of the alkali
metals.  Two  types  of  mineral wool fibers are available:  rock wool  and  slag
wool.  Rock  wool  is formed by melting natural  igneous  rocks, whereas  slag wool
is produced from material consisting of premelted slag from iron blast furnaces
or  from  other metal-slagging  processes.  Slag and rock wools are  generally
formed into  fibers  with  diameters of 1 to 6 urn,  which provide optimal thermal
conductivity.  These wools have many of the same uses as fibrous glass, primar-
ily  in  insulation.   Approximately 200,000 metric  tons of mineral wool were
produced in  1983.   Mineral  wool  is  relatively  less water-soluble  and more
chemical-resistant  than  glass  fibers,  but  both are  susceptible  to varying
degrees in acids and bases.

Environmental Release and Fate
     Estimates of the  environmental  release  of mineral  wool are  not available.
Similar to  glass fibers,  airborne concentrations  of  mineral  wool  in the
workplace vary greatly with sampling location within the same plant and  among
plants performing similar  operations.   The  range of average concentrations of
mineral wool found in air samples from several plants were reported to be 0.024
to 0.34 total fibers/cm  and 0.001 to 0.463 respirable fibers/cm .  The highest
fiber concentrations were  detected in the production and manufacturing areas.
Estimates of  respirable  fiber  concentrations in the  European mineral wool
industry during the early technological phase ranged from 1 to 10 fibers/cm  as
compared to  current levels of about 0.03 to  0.12  fibers/cm .   NOHS estimated
that 5079 workers were exposed to mineral wool  in  1972-74.  No information  was
reported by  NOES.   Estimates for consumer exposure were not available.  Simi-
larly, ambient levels  for  mineral wool  were  not  found  in  the  available litera-
ture, except  for those reported above for fibrous glass; the concentration of
MMMF ranged  from 0  to 10 fibers/L.   The environmental  fate  and analytical
methods used  to identify  and quantitate  mineral wool  are  similar  to those
described for fibrous glass.

Metabolism
     Factors  affecting  retention and  clearance  of mineral wool  fibers  are
similar  to  those  discussed  for  fibrous  glass  with  similar dimensions.
July 1988                          1-9        DRAFT--DO NOT QUOTE OR CITE

-------
Retention of  fibers  decreases  with increased  diameter and length and increases
with repeated  dosing.   Retention was  highest  for  fibers with a length of 11 pm
and a  diameter of 0.5 pm.   Clearance  was  faster for fibers with diameters less
than 0.5 pm and decreased with increased fiber diameter.  The  rates  of  clear-
ance of  retained fibers following repeated  and chronic inhalation exposure are
similar to those observed for glass  fibers.  Nearly all the fibers  retained
were  less  than  5 pm  long.   Short  fibers- are phagocytized  by  alveolar and
interstitial macrophages  and  cleared  to  lymph nodes.  Fragmented and partially
dissolved fibers  have also been detected.  In hamsters dosed  intratracheally
with rock wool  (1.1  x 27 pm),  pseudoasbestos  bodies were detected in the lungs
after 2 months.

Animal Toxicity
     No information was  found  on the acute toxicity of mineral wool.   Inhala-
tion exposure  to respirable mineral wool in  hamsters and rats for periods of
12 to  24 months elicits pulmonary responses comparable to  those observed with
glass  fibers  having   similar  dimensions.   Pulmonary responses  elicited were
mainly dust-laden  macrophage aggregation with  only -slight changes  in  lung
architecture.    The results  from two  studies  indicate a NOAEL  of  5  mg/rr  for
rock wool, whereas one study with  slag wool  identified  a  LOAEL of  10 mg/rr, .
The corresponding  RfD values  were 0.05 and  0.01 mg/m  .  Investigators found no
evidence of a  carcinogenic response following inhalation exposure to  slag wool
(diameter 2.7  pm,  length 40 pm) or rock wool (diameter <3 pm, length >5 pm).
Unlike inhalation  studies,  increased  incidences  of mesotheliomas and sarcomas
were noted in  rats following intrapleural or  intraperitoneal administration of
rock wool.  The carcinogenic  evidence for slag wool  administered via either
route  to  rats was weaker  (i.e.,  equivocal).   The tumor incidences were low
(0 to  6  percent).   When compared  to  asbestos, the carcinogenic  potential  of
rock wool is much lower.  Quantification of carcinogenic risk was not possible.
     Genotoxicity  studies  were not found in  the  available  literature.   Rock
wool and  slag  wool  can induce  considerable  cytotoxicity  in  V79-4  Chinese
hamster lung  cells,  A549 human tumor cells, and mouse peritoneal  macrophages.
Limited data  suggest  that the cytotoxic  effects of rock wool and slag wool are
comparable to  those  of similar-sized fibrous glass.   In contrast,  long, rela-
tively thin (diameter 3.3 pm,  length  221 pm)  mineral  wool fibers do  not appear
to be cytotoxic to rat pulmonary alveolar macrophages or erythrocytes.

July 1988                          >10       DRAFT-DO NOT QUOTE OR CITE

-------
     Studies describing the  teratogenicity and reproductive effects of mineral
wool or slag wool and rock wool were not found.

Human Health Effects
     Cross-sectional  studies  with workers exposed  to mineral  wool indicate
skin, eye,  and  respiratory manifestations similar to  those observed in workers
exposed  to  glass fibers.  Several  epidemiologic  studies  did  not detect, or
provided  limited  evidence  for,  an association between exposure to mineral wool
and NMRD, especially  when  the data were analyzed by  latency.  Similar to the
studies with glass  fibers, the  absence  of risk was  limited by the  small  number
of deaths and low power to detect NMRD in multiple cohorts.
     Significant  excess  mortalities from various  carcinogenic  effects  were
noted in  workers  exposed  to  mineral wool.  These included respiratory,  lung,
digestive system,  stomach, and  buccal  cavity and  pharyngeal cancers.   The
weight of evidence  for  a causal  relationship from several studies  was limited.
A quantification  of  cancer risk or NMRD based on epidemiologic  evidence  ccuOd
not be made  because  exposure levels existing during  the  early  technological
phase were not available.
     Based  on  the weight  of evidence from human health  effects  and  animal
toxicity  data  summarized  in  Chapter  9  (see Table 9-5),  the carcinogenicity
classification of mineral wool is 61.

1.1.3  Ceramic Fibers
     Ceramic fibers  are produced  by  combining  kaolin'clay or a  mixture of
alumina and  silica  to form aluminosilicate glasses.  The melted  compound  is
blown to  form  fibers  used  primarily for high-temperature  insulation.  Approxi-
mately 100  million  pounds  were  produced 1n  1983.  Alumina and zirconia are  two
other types of ceramic fibers that possess strong physical and chemical refrac-
tory qualities  used  for even higher temperature applications.   The  fibers
produced  for  industrial applications have a desired  diameter  range of 2 to
3.5 pm and  a length  ranging  from a few micrometers  to several centimeters.
Alumina fibers  are  imported  from the United Kingdom; in  1983, 150,000 pounds
were imported and consumed in the United States.   The annual  production capacity
of zirconia  fibers  in the  United  States  for 1983 was about 10,000 to 12,000
pounds.   Ceramic  fibers  exhibit good chemical  stability  and resist oxidation,
reduction, and attack from most corrosive agents, but they are not resistant to
strong acids and alkalies.
i»i» lose                          1-11       DRAFT—DO NOT QUOTE OR CITE

-------
Environmental Release and Fate
     Estimates  of the  environmental  release  of ceramic  fibers are  not
available.  The  range  of average ceramic fiber concentration in air  samples
from three  plants was  reported to be  0.049  to 2.6 fibers/cm  and 0.21 to
2.6 respirable fibers/cm  .   No  information was  found on ambient concentrations
or NOHS and NOES estimates of worker exposure.   The  specialized  applications  of
ceramic fibers  limit  their  potential  for consumer exposure, but no estimates
were available.   Similarly,  no  information was  found on the environmental fate
of  ceramic  fibers.   The  analytical methods  used to identify and quantitate
ceramic fibers are similar to those described for fibrous  glass.

Metabolism
     Limited  information  is  available  on the pulmonary deposition,  transloce-
tion, and clearance  of ceramic  fibers.   Results from  a few available  studies
indicate that pulmonary  retention  and  clearance of  ceramic fibers  are similar
to the  pulmonary retention  and  clearance observed for glass and mineral wed
fibers.    Following  repeated and chronic inhalation exposure,  most  of  the
retained fibers  were  <_5  urn  in length.   However, slight differences  were noted
in the distribution of fibers deposited within the various pulmonary  lobes  when
compared to glass  fibers.   Retained fibers are cleared slowly  from the  lung.
and clearance rates  decreased  with increased fiber  lengths.   Fragmentation of
ceramic fibers,  as well  as  phagocytosis by alveolar macrophages and  clearance
to lymph  nodes,  have  also been  reported.  Migration of Ceramic  fibers  from the
peritoneum to the lymph  nodes  of  rats  injected  intraperitoneally has been
observed.

Animal  Toxicity
     No information was found on the acute toxicity  of ceramic fibers.   Similar
to other  vitreous  fibers, the chronic  inhalation  of ceramic fibers by  rats has
been shown to result  in little pulmonary manifestation other than dust-laden
macrophage aggregation.   Using  a grading system to  score  the results  of three
chronic inhalation toxicity  studies,  a NOAEL of 2.45 mg/m  was  identified  for
refractory Al.O,  fibers  (Saffil  ),  and  a LOAEL  of 8.4  mg/m3 was  identified for
aluminum silicate ceramic microfibers.
     Intraperitoneal  administration of  ceramic  fibers  resulted  in  fibrosis  in
rats.   However,  the  fibrosis contained  little  or no  collagen  and was  in  no way
comparable to the fibrosis  elicited by asbestos.   The results fror  seve-el
i i   ,^.»                           T.IO       nD'fT__nr. »'~T rvi-Tr r.r

-------
studies suggest that the carcinogenic potential of ceramic fibers  is  similar  to
that observed  for  other vitreous fibers.   Chronic inhalation toxicity  studies
with rats  and hamsters exposed  to  aluminum  silicate  ceramic fibers  and the
refractory  ceramic fibers Fiberfrax® and Saffil*1 were  negative.   Similarly,
inlratracheal  administration of  Fiberfrax®  to rats  and  hamsters produced
negative results.
     In  contrast,  Fiberfrax*1 was  found  to be  very carcinogenic  following
intraperitoneal administration (20  to 80  percent incidence), and the evidence
for  the aluminum  silicate microfibers was  equivocal.   Considerably  lower
incidences of tumors (2 to 4 percent) were noted for large-diameter alumina and
zirconium oxide  fibers  following intrapleural  administration.   Similarly,  low
incidences  of  mesothelial tumors  (10 percent) were  noted  for  the aluminurr.
silicate fibers.  Quantification of cancer risk was not possible.
     Genotoxicity  studies  were not  found  in  the available literature.   Several
cytotoxicity  studies  indicate that ceramic fibers with  various  diameters
(0.2 to 5 ^m) and  lengths are not hemolytic to erythrocytes or cytotoxic to rat
pulmonary alveolar macrophages or to macrophagelike P338D-, cells. Similarly,
      Gi                    GL
Saffil    alumina  and Saffil  iirconia (diameter 2  to  6 ym) were  found  to  be
minimally cytotoxic and nonfibrogenic to rat peritoneal macrophages.
     Teratogenicity and reproductive effects  studies were not found  in the
available literature.
     Based  on  the  weight of evidence from animal  toxicity data summarized in
Chapter 9  (see Table 9-5),  the  carcinogenicity classification  for  ceramic
fibers is C.

Human Health Effects
     No information was found on the health effects of ceramic fibers.   .
1.2  SYNTHETIC FIBERS
     Similar to  MMMF,  the synthetic fibers discussed In this document are not
discrete entities  definable  by a fixed chemical  composition.   Their chemical
and physical properties depend upon their chemical composition.

1.2.1  Aramid Fibers
     Aramid fibers  are aromatic  polyamids synthesized  from  aromatic  diarr.ines
and diadic  chlorides.   Two  categories of fibers  are  available:   heat-  anc!
July 1988                          1-13       DRAFT—DO NOT QUOTE OR CITE

-------
flame-resistant  fibers  (Nomex ) and ultra-high-strength, high-modulus fibers
(Kevlar ).  Most aramid fibers are highly crystalline.   They are found in  a  dry
pulp  form,  which consists of  short  fibers  (0.5 to 8 mm) with  fine  fibrils
attached to the surface.  They have excellent wear resistance,  are nonabrasive,
and  have  multiple  applications, including the following:   Nomex*  fibers  are
used  in industrial  protective  clothing,  carpets,  tents,  and solution  paper for
electrical motors and  transformers;  Kevlar   fibers are used in tires, cables,
                                                                             £
parachutes, and body armor.  In 1975, the annual production capacity for Nomex
          £
and Kevlar  was reported to be 9,100 and 23,000 tons/year,  respectively.

Environmental Release and Fate
    • Estimates of environmental  release  of  aramid fibers were  not available.
Also  unavailable were  information on airborne concentrations in manufacturing
plants and ambient  air, NOHS and NOES estimates of worker exposure, and  esti-
mates of  consumer  exposure.   Aramid fibers are  susceptible to  weathering and
aging.  Their chemical composition and tensile strength are affected by photol-
ysis,  temperature,  and  oxidation but not  hydrolysis.   Analytical  methccs
similar to those described for fibrous glass are used to identify and quar.fi-
tate aramid fibers.

Metabolism
     Information on the pulmonary deposition  and clearance  of  aramid fibers  in
laboratory animals  is  very limited.   However,  it appears that  these  are influ-
enced  by  the same  mechanisms  noted  for other fibers.'   Following  inhalation
exposure  or  intratracheal  administration to rats, fibers with  a Dfle of <5  urn
were  found  in the  alveolar region.  Fiber-laden  macrophages were not immedi-
ately observed after inhalation, whereas fragmented and markedly shorter  fibers
were observed within 3 months.

Animal Toxicity
     Mild  to  moderate  foreign body  reaction was  found in  the  lungs  of rats
exposed to aramid  fibers for 2 weeks  at concentrations of  3 and 18 mg./m  but
not at <0.5 mg/m .   Fibrosis was noted  at  the  highest exposure level and was
still  evident 6 months  postexposure.   Only one  chronic inhalation  toxicity
                               £
study was  available for Kevlar   fibers  (diameter 90  percent <1.5 urn, length^
75 percent <20 urn).   Moderate pulmonary toxicity characterized by hyperplasi^H
obi iterative  bronchiolitis  with  collagen  formation,  emphysema,  and a

-------
dose-dependent increase in squamous cell  carcinoma  was  noted  in rats exposed to
concentrations of  0.63 and  2.23 mg/m3.   The NOAEL and  LOAEL  for systemic
effects  were 0.08  and 0.32 mg/m3, respectively,  and  the RfD volume  was
0.0008 mg/m  .  The estimated cancer risk  was  9.4 x  10~2  (ug/m  ).
     The results from three studies with  Wistar rats  injected intraperitonesny
           £
with Kevlar* fibers  (diameter <1.0 pm) indicate a weak carcinogenic potential.
The incidences of  nesotheliomas  and  sarcomas were  low (6 to 13 percent)  and
were not dose-dependent.
     Genotoxicity studies were not found in  the available literature.  Aram id
fibers with  mean diameters  of  0.14 urn and mean lengths of <5 pm were found  to
be  as  cytotoxic  as  asbestos when  incubated  with cultures of rat  pulmonary
alveolar macrophages.  The fibers induced a dose-dependent release  of cytosomal
and lysosomal marker enzymes with a concomitant decrease  in ATP cell content.
     Teratogenicity  and reproductive  effects  studies  were not  found  in the
available literature.

Human Health Effects
     No information was available on the  health effects of aramid  fibers.
     Based on the  weight  of evidence from animal  toxicity data summarized in
Chapter 9 and  in the  absence  of  human  effects data  (see Table 9-5),  the
carcinogenicity classification for aramid fibers is B2.

1.2.2  Carbon Fibers
     Carbon  fibers  are  produced  by controlled  pyro'lysis  of  rayon,
polyacrylonitrile (PAN), or mesophase pitch.   The fibers  have diameters  of 5 to
15 urn and are  characterized by flexibility,  electrical conductivity, chemical
inertness  except  oxidation, refractoriness,  and  high  tensile  strength.
Rayon-based  carbon fibers  are produced in a  high-modulus or  a low-modulus forrr.
and are  used in  tire cords  and aerospace applications.  The  estimated consump-
tion was  100,000 to 250,000 kg/year  for 1970  to 1976.  Applications  of PAN
fibers include use  in structure  fabrication  for the aerospace  industry, sport-
ing goods, prostheses  and artificial  joints, and  reinforcement for plastics.
Production volumes  were not available.   Mesophase pitch-based  carbon  fibers
have a carbon (graphite) content over 99.5 percent.  They are produced  in  three
forms, differing only in filament modulus.
July 1988                          1-15       DRAFT--DO NOT QUOTE 0* CITE

-------
Environmental Release and Fate
     Estimates  of  environmental  release of carbon fibers are not  available.
Airborne  concentrations  of carbon  fibers  vary with  sampling  location and
process within  the  same  plant and from one plant  to  another.   The ranges  of
concentrations  found  in  air samples from two plants  were 0.024  to  1.2  and  0.9
to 659,000  total  fibers/L,  respectively.   The range  of respirable  fibers in a
third plant  was 0.03 to 0.16 mg/m3.  No information was available on ambient
concentrations, NOHS  and NOES estimates of worker exposure, or estimates  of
consumer  exposure.   Information  on  the  environmental fate of carbon fibers wes
unavailable.  However, carbon fibers are chemically inert except for oxidation.
The  analytical  methods  used  to  identify and  quantitate carbon fibers are
similar to  those  used for fibrous glass.  In  addition,  microwave absorption
and induced charge detectors can be used with larger  fibers.

Metabolism
     Phagocytosis of  carbon  fibers  in exposed guinea pigs occurs  within hours
after inhalation and  continues  for 2 or more  years.   The  nonfibrous  dust  is
primarily cellular  in macrophages,  and  the number of. particles  in each macro-
phage increases with  time after exposure.   Human subjects exposed to  a carbon
dust aerosol for 3 minutes at a concentration of 50 mg/mL (mass  median  diameter
11 urn)  retained approximately half  of  the  inhaled carbon dust.   Additional
information was not found in the available literature.
                                                        •
Animal Toxicity
     Short-term toxicity studies with carbon fibers are limited  to tissue reac-
tions associated with implants  in the peritoneal  cavity or bones  of rats  and
rabbits.  Fibrous  tissue ingrowth  and  mild to acute  foreign body response
was  evident.   The  results  from two  subchronic  inhalation  toxicity studies
indicate no pulmonary damage, although macrophage accumulation was.evident.  In
the first study,  rats were  exposed to submicron  carbon dust for 7  to 100 hours
and observed  for  up to 2 years.   In the second study,  hamsters  were exposed to
20 mg/m   Celion,  a PAN-based carbon  fiber with  a  mean diameter of 7 urn and
length  of 20  to 60 urn, for  16 weeks.  Chronic  inhalation toxicity  studies  were
not  found  in the available  literature.   Consequently, an RfD value was not
calculated.    Similarly,   intratracheal,  intrapleural,  and  intraperitoneal
studies were  not  available.   The results from a chronic dermal  toxicity study

July 1968                          i-16       DRAFT-DO NOT QUOTE OR CITE

-------
with male mice  treated  three times & week for  life  with  one  of four carbon
fibers did not induce a compound-related increase in  tumor incidences.
     Genotoxicity studies were  not found in  the available literature.  Carbon
fibers of various  sizes (diameter 0.2 to 15  urn) were only minimally  cytotoxic
to rat peritoneal and pulmonary macrophages.
     Teratogenicity  and reproductive effects studies were  not  found in  the
available literature.

Human Health Effects
     Results from one cross-sectional study provided  no  evidence of  respiratory
manifestation in 88 workers employed in a carbon plant.
     Based on  the weight  of evidence from human health  effects and animal
toxicity data  summarized in  Chapter 9  (see  Table 9-5), the carcinogenicity
classification of carbon fibers is 0.

1.2.3  Polvolefin Fibers
     Pclyolefin fibers are long-chain synthetic  polymers consisting  of at least
85 percent  by weight ethylene, propylene, or other  olefin units.    They are
fabricated commercially by a melt extrusion technique.   This involves expulsion
of the polymer  through  a spinneret,  solidification,  and winding onto packages.
Polyolefin fibers are extremely hydrophobic  and inert  to  acids and  bases and
organic solvents at  room temperature.   Polyolefin fibers  are primarily used in
home furnishings, such as carpet facing and backing  for  upholstery..   Industrial
uses  include  synthetic   turf,  cordage and twine, agricultural  fabrics,  and
sewing threads.  Production  and consumption  values  for 1963 were 603 and 767
million pounds,  respectively.   Production has increased steadily to  correspond
with increased demand.

Environmental Release and Fate
     Estimates of environmental  release of polyolefin  fibers were  not avail-
able.  Similarly, no information was available on airborne concentrations  in
manufacturing plants  and ambient  air.   NOHS  estimated that 96,784 workers were
exposed to polyolefin fibers during 1972-74; NOES provided no  estimates.   No
information was  available on  the environmental fate of  polyolefin fibers,
although they  tend  to be chemically  inert at  room  temperatures.  Analytical
procedures similar to those  described for fibrous glass  are  used to identify
and quantitate polyolefin fibers.
July 1988                          1-17       DRAFT--DO NOT QUOTE OR CITE

-------
Hetabolism
     No  information  was found on  the  retention and clearance of polyolefin
fibers.

Animal Toxicity
     Short-term  toxicity  studies are limited to tissue  reactions  associated
with  implants  in the  peritoneal  cavity and femoral bone of  rats.   Fibrous
tissue ingrowth  was  noted  in both tissues.  Subchronic and chronic inhalation
toxicity  studies  were  not  found  in the  available literature.  Consequently, an
RfD  value could  not  be calculated.  Male rats dosed intratracheally with
1.25 mg  of  polypropylene  fibers  or dust and observed  for 21  months  exhibited
lung granulomas with foci containing irregularly shaped birefringent particles.
No other  effects  or  pulmonary tumors were noted.   Intraperitoneal  administra-
tion of polypropylene fibers (diameter <2.1 pm, length <23 urn) to female Wistar
rats elicited  a  weak carcinogenic  response.  Two of 51 rats  developed  mescthe-
liomas or sarcomas, but the incidence was not significantly different froir that
of control animals.
     Genotoxicity studies were not found in the available literature.  Polye*J
ylene  (diameter  3 to 75 urn) and polypropylene (diameter 4 to 50 urn) exhibited
low  cytotoxicity  when  incubated with  rat peritoneal  or pulmonary alveolar
macrophages at 0.5 mg.   Less  than  1 and 4  percent  of the peritoneal  and  alveo-
lar  macrophages,  respectively,  were killed following phagocytosis of  the
fibers.
     Teratogenicity  and reproductive effects  studied  were not found  in  the
available literature.

Human Health Effects
     No  information was available on the health effects of polyolefin fibers.
     Based  on  the weight of evidence from animal  toxicity data in the absence
of human health  effects  data (summarized in Chapter  9  and  Table  9-5),  the
carcinogenicity classification for polyolefin fibers is D.
1.3  MINERALS
     All  minerals  and their  asbestiform  varieties are defined by a  crystal
structure and  a  chemical  composition that is restricted to a limited range c|

July 198E                          1*18       DRAFT--DO NOT QUOTE OR CITE

-------
permissible substitution.  Their  physical  and  chemical  properties  are deter-
mined  by  their crystal structure, chemical composition, and  crystallization
habit.

1.3.1  Attapulgite
     Attapulgite  (CAS ho.  12174-11-7)  is  a  trade  name for  the  common
asbestiform,  short-fibered (length  0.8 pn>,  diameter 0.06 pm.)  variety of
palygorskite  mined  at Attapulgus, GA.   Palygorskite is a layered  silicate
structurally  related to  the  clay  minerals and  has the  composition of
Mg5(OH)2(H20)4Sig02Q'4H20.  The mineralogic composition  varies  greatly with the
site  of origin and  the presence of  other  minerals.   Approximately  2  million
short  tons  were  produced or used in 1985.   Attapulgite, a  light  cream-colored
powder, is chemically inert and its properties  are unaffected  by  changes  in pH.
Attapulgite finds a  wide commercial  application in both absorbent  and jelling
or thickening agents due, in part, to its  sorptive and colloidal  properties.

Environmental Release and Fate
     Emissions of  minerals are  associated with  natural weathering,  mining
operations, and  the use  of  these materials in secondary  products, but no
estimates   are available  for  attapulgite.   The range of concentrations of
attapulgite in air  samples in mining and milling  facilities was  reported to  be
0.09  to 23 mg/cm   and 0.05 to 2.7 mg respirable  fibers/cm .   NOHS  estimated
that  134,121  workers  were exposed to attapulgite  in  1972-74.   NOES provided  no
estimates.  Estimates  of  consumer exposure were not  available.   Attapulgite  is
inert  and is  not  expected to undergo  chemical transformation.  Analytical
procedures  used to  identify  and quantitate attapulgite  are similar to those
reported  for  fibrous glass.

Metabolism
      Limited  information  was  available  on  the  metabolism of  attapulgite.   The
results of two case  reports  suggest that  inhaled attapulgite can reach the
alveoli and that  ingested attapulgite is absorbed and eliminated in the urine
when  consumed in  large amounts.   A  total  of 14 different mineral groups were
identified  in the  lungs of 11 men and 9 women  with no history of occupational
exposure  at the time of death.  These included  asbestos, apatite, gypsum,  talc,
silica, and attapulgite.  About 86 percent of all  of these fibers were <5 ur  in
length.
July 1988                          1-19       DRAFT—DO NOT QUOTE OR CITE

-------
Animal Toxicily
     No  information  was available  on  the acute and subchronic toxicity  of
attapulgite.   Similarly,  no information was available on  chronic  inhalation
toxicity.   Consequently,  an  RfD value was  not calculated.   Intratracheal
instillation of  100  mg  attapulgite  (0.02  x 0.8 urn) in sheep resulted in bio-
chemical  changes,  accumulation of alveolar macrophages,  and sparse  areas  of
peribronchiolar  fibrosis.   Intrapleural administration of  attapulgite to two
strains  of rats  was not  tumorigenic.   Intraperitoneal  administration  of
attapulgite  (0.07  x 1.3 urn)  obtained  from Caceres, Spain, was tumorigenic;
40 percent  of  the  rats  developed mesotheliomas or  sarcomas.  However,  three
other samples  from Spain,  the United States, and France  (0.04  x 0.8 urn),  were
not  tumorigenic  following intraperitoneal administration  in the  same  study.
The compositions of these samples were not defined.
     Attapulgite does not  induce unscheduled DNA synthesis in primary cultures
of rat  hepatocytes at  concentrations of 1 and  10 pg/mL.   Limited  information
also  suggests  that  attapulgite  does not   increase  the  frequency of  sister
chromatid exchange.  No other genotoxicity studies  were  available.   Attapulgite
(0.06 x  0.77 urn)  generally  appears  to  exhibit  cytotoxic and hemolytic activity
when incubated with  peritoneal  and  alveolar  macrophages.   Attapulgite was mere
cytotoxic and  hemolytic  than  chrysotile asbestos.   Cytotoxicity increased witn
increased fiber  length  and  decreased diameter as shown by increased  cytctox-
icity in mouse peritoneal macrophages,  A549 type II  alveolar  human tumor cells,
and V79-4 Chinese  hamster lung cells.
     Teratogenicity  and  reproductive effects  studies were not found in the
available literature.

Human Health Effects
     Limited information was  available  on the health effects of attapulgite.
The  results  from  a  cross-sectional  morbidity  study indicated  no consistent
relationship between worker exposure and respiratory symptoms.   In attapulgite-
related occupations, pulmonary function was generally low and  the  prevalence of
pneumoconiosis was elevated in  the older age  categories; however,  neither
effect was  statistically  significant for  workers in occupations that generated
dusts other than  attapulgite.   Despite the low power,  when compared with  U.S.
rates,  a  threefold  excess  in  lung  cancer mortality was  noted in a cohort
mortality study of 2300  miners  and  millers with >5  years exposure duration and
^20 years  latency  who were  employed for  at  least  1 month frorr, 1940 to  1977.
' *  """•                          1-?n       DRAFT—DO NOT QUOTE  OR CITE

-------
However,  the  evidence is  inadequate  for a causal relationship because the
excess was  noted  in  only one study, the number of deaths  was  small, the power
was low, and mortality rates were comparable to national mortality  rates.
     Based  on  the weight  of evidence from human health effects and animal
toxicity  data  summarized  in  Chapter  9  (see Table 9-5),  the carcinogenicity
classification for attapulgite is C.

1.3.2  Erionite
     Erionite (CAS No. 66733-21-9)  is a fibrous 2eolite found in deposits in
the Western Intermountain Region of  the  United States.   Erionite  occurs as
either single needles or in clusters and ranges from 0.25  to 1.5 um in  diameter
and  up to  50 um  in  length.   U.S.  erionite  is  classified as "probably
asbestiform," whereas  the erionite found  in  Turkey  is  a  suspected "truly
asbestiform" material owing  to  exposure-associated adverse health  effects.   A
sample taken from the Cappadocia region of Turkey had the  following chemical
composition:  (Na^K-CaMg)^ ^AlgSi-yOy-^HpO.
     Erionite has  been used  for nitrogen  retention in fertilizers  and  concrete
aggregate and in  road surfacing, but  production and  use  volumes are unavail-
able.   The  annual production of natural  zeolites in the  United States  was
estimated to be 2500 to 500D tons for 1982-84.   However,  erionite  apparently  is
not produced in the United States at the present time.

Environmental Release and Fate
     Estimates of  environmental  release of erionite are  not available.  Simi-
larly, estimates  of  airborne concentrations in mining facilities  and  ambient
air, as well  as  worker and consumer exposure,  were not reported.   However,  in
two Turkish villages with  relatively high incidences of mesothelial  tumors
among the villagers, airborne concentrations of respirable dust ranged  from  0.2
to 2.8 mg/m , whereas fiber concentrations ranged from <0.01 to 1.38 fibers/cm .
No information was found on the environmental fate of erionite.  The analytical
procedures  used  for fibrous glass  are  also used to quantitate and identify
erionite fibers.

Metabolism
     Although the mechanisms  governing  lung  deposition  and  clearance  of
erionite fibers are similar to those governing  lung deposition and clearance  of
other  fibers,  very little  information is  available  on  the  deposition and
July 1986                           1-21        DRAFT—DO NOT QUOTE  OR CITE

-------
clearance of erionite.   Erionite  has  been shown to penetrate the pleura! and
parenchyma! tissues  in  Turkish  villagers  exposed to low levels.  Fibers recov-
ered from  the  lungs  of  these  individuals  were uncoated and had a mean diameter
of 0.3 urn and a maximum length of 9 urn.   Ferruginous  bodies  containing  erionite
fibers were noted in a few individuals.

Animal Toxicity
     No information was found on the acute and subchronic  toxicity  of erionite.
Chronic toxicity data from several  studies indicate that erionite samples fron
Oregon and Turkey are highly carcinogenic in  rats  via  the  inhalation, intratra-
cheal, intrapleural, and  intraperitoneal  routes  of administration.  The  tumor
incidences in treated animals were 48 to 100 percent, with  mesothelial tumors
being the most predominant.  Erionite  has also been shown  to be carcinogenic  in
mice following  intraperitoneal  injection.  The carcinogenic  unit risk basec! on
tumor  incidences  following inhalation exposure is 5.56 (mg/m  )   .  However,
nonfibrous, synthetic  zeolites  do  not  appear to  be  carcinogenic  following
intrapleural  or intraperitoneal  administration.
     Erionite has been  shown  to induce  significant increases in aneuploidy .
polyploidy in Chinese hamster lung cells and  to induce chromosomal  aberrations,
cell transformations, and  unscheduled  DNA synthesis  at concentrations  ranging
from 10 to 100  ug/mL.   Erionite is cytotoxic to rat  pulmonary  alveolar macro-
phages and human polymorphonuclear leukocytes, but it exhibits  a weak hemolytic
activity in rat erythrocytes.
     Teratogem'city  and  reproductive  effects studi'es with  erionite were  not
found  in the  available  literature.   A teratogenicity study  with  a  synthetic
zeolite administered orally to  pregnant Sprague-Oawley rats and  New Zealand
rabbits at 160C mg/mL on days 6 to 15 and 6  to 18 of gestation, respectively,
indicated  no  adverse effects on  the  dams,  embryos,  or  fetuses  in  either
species.

Human Health Effects
     Several  cross-sectional morbidity  and mortality  studies indicate  a  strong
causal relationship  between exposure  to erionite  and increased  incidence  of
malignant mesothelial tumors  and  other pathological  changes in inhabitants of
three  Turkish  villages.   Although causal relationships are  not  discerned  bv_
cress-sectional  studies,   the consistency of these  findings among severe

July 1988                          1-22       DRA'T—DO NDT QUOTE C-R  CITE

-------
studies and the presence of erionite rather than asbestos  indicate  an  etiologi-
cal role  for  erionite.   Quantification of cancer risk was not performed  for
these cross-sectional studies.
     Based  on the weight  of  evidence from human health effects and  animal
toxicity  data summarized  in Chapter  9 (see Table 9-5),  the carcinogenicity
classification for erionite is Bl.

1.3.3  Talc
     Talc (CAS No.  14807-96-6)  is made up  of pulverized, foliated  hydrous mag-
nesium silicates  from  minerals  with low crystalline silica content.  Talc is
a  layered silicate,  and both  talc crystals and  fragments are commonly platey
and nonfibrous.   Its  color can  be white, yellow, green, or gray depending on
the  number  of  impurities present.    The  chemical composition of  talc is
(^(Si^Oe^MgCOH)^.  jt  is insoluble  in water and exhibits  varying resistance
to acids, alkalies,  and temperature.   Approximately 1100 short tons  of  talc
were produced  in  the  United States  in 1984.  The largest  end  uses  of  talc are
in ceramics  and paint; only 5  percent are used in cosmetics.

Environmental  Release and Fate
     Estimates of  environmental  release of talc are not available.   Airborne
concentrations in  mining  and milling facilities have been  found to range
between 0.28  and  2.64 mg/cm  for  total particles.   Estimates  of airborne
concentrations  in  ambient air  are  not  available.   NOHS  reports  that
1,536,754 workers  were  potentially  exposed to talc in'1972-74, whereas  NOES
estimated that 18,872  workers,  including  5,244  females,  were potentially
exposed to  talc  in  the workplace  in  1980.   Consumer exposure to talc and
talc-containing products  is potentially high,  but no  estimates were available.
Certain talc  deposits have been shown to contain various concentrations of
asbestos, but estimates of asbestos  exposure from talc during  production  or  use
were not  available.   Talc is  not expected  to  undergo  chemical transformation
when released into the environment.   Analytical procedures used to identify  and
quantitate talc are similar to those described for fibrous glass.

Metabolism
     Following single  inhalation exposures, talc is retained in the  lungs of
exposed hamsters  and  is cleared slowly with  a biological  half-life of 7 to
10 days.   The  talc content decreased to the detection levels  founi in centre!
July 1988                          1-23       DRAFT--DO NOT QUOTE  OR CITE

-------
animals 4  months after  exposure.   Pulmonary deposition of  talc following
repeated exposures  is  dose-dependent.   Pulmonary  clearance  is facilitated by
mucociliary activity  and eliminated  in  the feces via  the  gastrointestinal
tract.  Little,  if  any,  talc is absorbed from the gastrointestinal  tract  of
rats, guinea pigs,  mice,  and hamsters following  oral  intubation.  Most  of  the
material was  eliminated in the feces within  1 to 2  days after  dosing.   No
measurable transport of  talc was  reported following  intravaginal  instillation
into  cynomolgus  monkeys.   In rats, however, intrauterine  or intravaginal
instillation resulted in migration of talc particles  into the ovaries.

Animal Toxicity
     No information was available  on  the acute toxicity of  talc,  although
intratracheal  instillation  of talc  in hamsters  resulted in  pulmonary toxicity
as shown by biochemical  and cellular changes.   Subchronic  inhalation exposure
(3 to  12  months) in  rats  resulted  in pulmonary fibrosis that  increased  in
severity as the  exposure period increased.   A LOAEL  of 10.8  mg/m   and  an  RfD
value  of  0.011 mg/m   was   identified  from this  study.  Intratracheal  or
intraperitoneal  administration  of talc  resulted  in  a  typical granulomatous
lesion  consisting of  dust-laden multinucleated  foreign  body  giant  cells,  as
well  as some  fibrosis  with collagen formation in  several animal species.  In
addition to the  development of granulomas, intratracheal instillation of  talc
resulted  in  moderate  tissue destruction in  hamsters,  and  intraperitoneal
administration has  been  shown to  cause abdominal  adhesions  in rats  and  swine.
Talc  is not  carcinogenic   following  inhalation  exposure or  intratracheal
instillation in  rats  and hamsters.   Similarly,  no evidence  of carcinogenicity
was noted following intrapleural,  intraperitoneal,  or  oral  administration in
rats.  The results  of  one  study suggest  that talc may  act  as a cocarcinogen
following intratracheal  administration  in combination with  benzo(a)pyrene  once
weekly for life, but the exact mechanism is not  known.
     Limited data suggest that talc does not induce gene mutations in Salmonella
or Saccharomyces  at 200 mg/mL, chromosomal  aberrations in  rat  bone marrow
cells,  or  dominant  lethal  mutations  in  the  germinal  cells  of  male rats.
However, no details of these studies  were presented.  Natural or acid-purified
talc  (particle size 0.2 to 20 urn) exhibits  weak cytotoxicity  in  rat peritoneal
and alveolar macrophages.   Talc particles within  the  respirable  range were net
cytctoxic to  A549 alveolar  human tumor  cells  or  V79-4  Chinese  hamster luic
cells.
Julv 196S                          I-?4       DRAFT-DC  NCT Q'j:TE OR C!TE

-------
     Teratogenicity and reproductive effects studies were  not  available.

Human Health Effects
     Several cross-sectional morbidity  studies  of  miners  and  millers  indicate
increased  respiratory  symptoms,  higher prevalence  of pleura! thickening or
calcification and pneumoconiosis,  and  decreased pulmonary function  in workers
exposed  to talc containing  various  amounts of tremolite, anthophyllite, or
silica  fibers.   These effects increased  with age of workers,  intensity  of
smoking, and duration  of  exposure.   However,  exposure to  talc free  of asbesti-
form fibers was  associated with  less pronounced effects.   Increased incidences
of ovarian  cancer  are  reported in a case-control  study of 215  white  females
regularly using talc as a dusting powder on the perineum or on sanitary  napkins,
or both,  when  compared to  controls.   However,  the purity of talc  used was
unknown.   Similar to  the  findings  described above,  several historical prospec-
tive mortality  studies  indicate  an excess  of respiratory  cancer and NMRO  among
mining  and  milling  workers.   However,  the  talc  was  contaminated with  asbestos,
silicates,  carbonates, or radon.   Consequently,  a  causal  relationship for ta'c
alone cannot be established.
     Based  on  the  weight  of evidence  from human health  effects and  ar.itr.a1
toxicity  data  summarized in  Chapter 9 (see Table 9-5),  the  carcinogenicity
classification for talc is D.

1.3.4  Vermiculite
     Vermiculite (CAS  No.  1318-00-9) is a  micaceous hydrate of  magnesium-iron-
aluminum silicates.  Vermiculite  crystals  are composed of two silicate layers
connected  by  a hydrous  layer.   The chemical composition of  vermiculite is
(Mg,Fe,A1)3(Al,Si)4010(OH)2'4H20.  It  is  unique among minerals  in  its ability
to exfoliate or  expand up to 20 times  its original size at high temperatures.
It has  cation  exchange capacity  and a  very low thermal  conductivity.  About
285,000  short tons  were  produced in the United States  in 1985.  Most of the
vermiculite mined  and beneficiated  is exfoliated  and  used  in  construction
aggregates, insulation, and agricultural applications.

Environmental Fate and Release
     Approximately  802  x 10   kg  of  vermiculite were released  into the  air,
89,900  x  103 kg  were  released into  water,  and 2.490 x 10  kg were  release:!  es
solid waste  in  1979 based on a production  volume of 1 millio-  tens.  Consumer
July 19S£                          1-25       DRA-T--DO NOT Q'JCTE OR CITE

-------
use results  in  the release of 36 x  10   kg  into the terrestrial  compartment.
Airborne concentrations in mining and milling facilities were reported to  range
from <10  to 39 fibers/cm.   No  information was found  on  concentrations  in
ambient air.    A  recent study indicates that airborne concentrations  prior tc
1964 were  high  (<182 fibers/cm ).   However, exposure estimates have decreased
significantly and  have  leveled off  from 1976 to 1982 to >1.0 fiber/cm.   NOHS
reports that 104,456 workers  were potentially  exposed to vermiculite  from 1972
to 1974, whereas NOES estimated that 4,293 workers, including 365 females, were
exposed to  vermiculite  1n 1980.   Certain vermiculite deposits have been shown
to contain  various  concentrations  of asbestos.   Estimates  of occupational
exposure levels for  asbestos  ranged from undetectable to 9.7 fibers/cm  of air
in mills  and exfoliation facilities.  Ambient  air concentration of asbestos
near mills and exfoliation facilities ranged from undetectable to 0.5 fibers/err5
            -5             -2    3
and 5.0 xlO   to  2.5 x 10  mg/m ,  respectively.   Nonoccupational  exposure to
vermiculite is  high.   In  1979, approximately 13 million persons  were  estimated
to have  been exposed  to  vermiculite near exfoliation  plants  in the  United
States.   In addition,  about  106  million persons  were  exposed  to  consumer
products  containing   vermiculite.   Vermiculite is  not  expected to undergo
chemical transformation when  released into  the environment.   Analytical proce-
dures  used  to  identify  vermiculite  are similar to those described for fibrous
glass.

Metabolism
     No information  is  available on  the  metabolism of  vermiculite following
oral  administration or  inhalation exposure.

Animal  Toxicity
     No information  was found on the acute,  subchronic, or chronic toxicity  cf
vermiculite.   Consequently,   an  RfD value  was not  calculated.   Female
Sprague-Dawley  rats  injected  intrapleurally  with 25 mg  vermiculite (<2  x <5  urr,
amorphous particles) developed granulornas in the lungs and viscera.  However  no
tumors  developed after  104  weeks.   Genotoxicity and cytotoxicity studies  were
not found in the available literature.  Similarly,  teratogenicity and reproduc-
tive effects studies were not found.
Julv 1988            •              1-26       DRAFT--DO NOT QUOTE OR CITE

-------
Human Health Effects
     An association between  previous  vertniculite exposure and parenchymal  and
pleura!  radiographic  abnormalities has  been found  in two cross-sectional
studies of  miners  and millers.   Although these changes did not correlate with
pulmonary function  tests,  they  did correlate with age  and fiber-years.   Howev-
er, vermiculite  from  both facilities  contained fibrous tremolite/ actinolite.
A  historical  prospective mortality study  of one of the  mining  and  milling
facilities  indicated  a  significant twofold excess in mortality for lung  cancer
and NMRD,  especially  in the  highest  exposure  category and among those  with
>20 year  latency.   However,  the  presence  of  tremolite-actinolite  asbestos
precludes a causal relationship between exposure to  vermiculite and lung cancer
or NMRD.
     Based  on  the weight  of evidence from human h&alth  effects and aniir,= 1
toxicity data  summarized in  Chapter  9 (see Table 9-5),  the  carcinogenicity
classification for vermiculite is D.
1.4  REFERENCES

Bayliss, D.  L. ;  Dement,  J.  M. ; Wagoner, J. K.; Blejer, H. P. (1976) Mortality
     patterns among  fibrous  glass  production workers. Ann.  N.  Y.  Acad.  Sci.
     271:  324-335.
Shannon, H.  S.;  Jamieson,  E.; Julian, J. A.; Muir, D. C. F.; Walsh, C. (1986)
     Mortality experience of glass fibre workers--extended follow-up. Presented
     at: international  symposium on  man-made mineral fibres in the  working
     environment;  October;  Copenhagen,  Denmark.  Copenhagen, Denmark: World
     Health Organization, Regional Office for Europe.
Shannon, H.  S.;  Jamieson,  E.; Julian, J. A.; Muir, D. C. F.; Walsh, C. (19E7)
     Mortality experience of  glass  fibre workers -•  extended  follow-up.  Ann.
     Occup. Hyg.  31: 657-662.
July 1985                          1-27       DRAFT--DO NOT QUOTE OR CITE

-------
                          2.   BACKGROUND INFORMATION
     Concern  surrounding  the possible  health  effects  of asbestos upon  the
general population  has  stimulated  interest in the potential  health effects of
other natural and  manmade fibers.   The health effects of asbestos have  been
recently  reviewed  by the National Research Council  (1984)  and the  U.S.. EPA
(Nicholson, 1986).   The purpose of this Health Assessment document is  to  review
key toxicologic data  on manmade  fibers  (e.g.,  fibrous glass, mineral wool, and
ceramic fibers), synthetic  fibers  (aramid, carbon, and polyolefin fibers), and
naturally occurring fibers other than asbestos (e.g., attapulgite and  erionite)
to assess whether significant human health effects are  associated with envircr-
mental  exposure.   Data on  talc  and vermiculite  are also reviewed  in this
document because of concerns associated with  the  possible presence of  asbestos
as a  contaminant  in some deposits of  these minerals.   In addition,  data or,
other fibers  and/or minerals are discussed in various parts of  this document
primarily for comparative purposes.

2.1  PHYSICAL AND CHEMICAL PROPERTIES
     Minerals and their asbestiform varieties  are defined by a  crystal  structure
and a chemical composition that is restricted  to a limited range of  permissible
substitutions.  Their  physical  and chemical properties are determined  first by
their crystal structure,  second  by their chemical composition,  and third  by
their crystallization habit  (Zoltai, 1986).
     Unlike these environmental substances, which are discrete  entities defina-
ble by  a fixed  chemical  structure, the  products currently referred  to as
manmade vitreous  fibers  or  more commonly as  manmade mineral  fibers  (MMMF)
comprise a  group of materials that are less easily defined (Table 2-1).   MMMF
are manufactured  by various  techniques from  a molten mass  of raw  material
(Klingholz, 1977).    Table 2-1 also lists  minerals that have  a  fibrous  morphol-
ogy.   This  list could  easily contain  several  hundred  minerals if it were
expaned to  include  rare  fibrous occurrences   of  common  minerals or  fibrous
minerals found only in  small amounts.  The minerals selected for inclusion  are
July 19EB                               2-1       DRAFT—DO NOT QUOTE  On  CITE

-------
                    TABLE..2-1.  MINERAL AND VITREOUS FIBERS
     Manmade Mineral Fibers
             Fibrous glass
             Mineral wool
               Rock wool
               Slag wool
             Ceramic fibers
             Other fibers
     Synthetic Fibers
             Aramid fibers
             Carbon fibers
             Polyolefin fibers
     Naturally Occurring Mineral Fibers
             Asbestos
             Chrysotile
             Crocidotite
             Amosite
             Actino1ite--Tremo1ite asbestos
             Anthophyllite asbestos
             Attapulgite (palygorskite)
             Zeolites (erionite)
relatively common throughout  the  world,  have  commercial  applications, and  see"
likely to  have, or  are  known to have,  at  least  some of the properties of
asbestifonr. fibers (National Research Council, 1984).
     Physical properties  common to  asbestiform fibers, as compared with non-
asbestiform  crystals  of   the  same substances, Include, fiberlike  morphology
and dimensions, enhanced  strength and flexibility, diameter-dependent  strength,
increased  physical  and chemical  durability,  and improved surface  structure
(i.e., relatively free of defects).   A continuum  of these properties  is possi-
ble, so  that "high-quality" fibers  have these properties to  a greater extent
when compared  to other more brittle  fibers that  have these properties to  a
lesser extent (National Research Council, 1984).

2.1.1  Manmade Mineral Fibers
2.1.1.1   Fibrous Glass.   Fibrous  glass  (CAS  No.  not available)  is compose:!
primarily of the oxides of Si, Al, and B (Wagner et al., 1980).   Silica (SiO^),
soda ash,  B20,,  fluorspar, rutile,  and clay provide the raw materials for  its
manufacture  (Weil!  et al., 1983).   The manufacturing  process  allows production
of  relatively  narrow ranges  of  fiber size depending upon  methods  employed

Julv 19SS                               2-2       DRAFT--DO NOT  QUOTE  On  CITE

-------
during production and commercial need.   The physical  properties  of glass fibers
depend  upon their  chemical  composition.   During production,  the  chemical
constituents can be  manipulated,  and specific  properties of the glass fibers
can thereby be  promoted.   Glass fibers that include  glass wools and filaments
have greater concentrations  of  Ba  and  the  oxides of the alkali metals Na and K
and lesser  amounts  of  Fe  and the oxides of the alkaline earths Ca,  Hg, Fe, and
Ti; consequently,  their chemical  resistance is decreased and  they  are more
soluble in  water.   Glass  fibers never  split  longitudinally but, instead, split
across their axis and form fragments (Klingholz,  1977).
     The three  main  classes  of  glass fibers are textile fibers, wool fibers,
and fine fibers, but many subclassifications also exist within these classes
(National  Research  Council,  1984).  Textile fibers,  which are  the  coarsest,
consist of continuous-filament glass fibers that  have flame-resistant  qualities
and are used as reinforcement for plastics and  other  materials.  They  typically
have fiber  diameters between 10 and 15 urn, but the diameters can range from 6
to  20  urn.   Wool fibers are  used for insulation; the diameters are  usually
between 3 and 10 urn but can range from 1 to 25  urn.  Fine fibers are  produced by-
flame  attenuation and  are  used  for specialized purposes.  They have  nominal
diameters of  1 urn  or  less  (National  Research  Council,  1984; Wagner  et a!.,
1980),  account  for  less  than 1 percent of  consumption (Lockey,  1981) and
contain no trace elements  of biological significance  (Hill,  1977).
2.1.1.2  Mineral Wool.   Two  types  of fibers fall  under the  general  heading  of
mineral wool.   The  term "rock wool" (CAS  No.  not available) refers to glass
fibers formed  by melting  natural  igneous  rocks, whereas  slag wool (CAS No.  not
available) is produced from material consisting of premelted slag from ironblast
furnaces or  from other metal-slagging processes  (National  Research Council,
1984).  Rock  and slag wools  also  consist  predominantly of  SiO.,  Al.O^,  and
BjO,,  as is  the case with glass fibers,  but they differ in  that they contain
greater amounts of  Fe  and the oxides of the alkaline earths (CaO,  MgO, Fe-O^,
and TiOO and  lesser amounts of Ba and the oxides of the alkali metals.   They
are more  chemical  resistant  than  fibrous glass  and less  soluble  in  water
(Klingholz,  1977).    Spurny  (1983), using  LAMMA analysis, determined that  rock
wool fibers of  unspecified diameter and composition  leached certain  elements
after  long-term exposure in water, alkali, and acid.   Only Si was  detected when
fibers were placed in 2 N rLSO., whereas traces of Ti and Fe were  also detected
when  the mineral wool  was placed  in  2 N  HC1.   Fibers  immersed in  2  N NaOH
contained Mg,  Ti,  and  Fe  and leached Ca,  Si, S, Mn,  Al, and Co es  corrps^ed  to
July 1988                               2-3       DRAFT--DO  NO" Q'JOTE OR CITE

-------
fibers placed  in water.   Rock wool  fibers  exposed  to both 2 N HC1 and  2  N
H.SO^ leached Al,  S,  Ca,  Ti, Co, Mg, and  Fe  as  compared  to  fibers  placed  in
water.  Fibers of  rock and slag wools tend to cleave  transversally.   Slag  and
rock wools  generally  are  formed into fibers that have diameters  of  1  to  6  urr,
which provide optimal  thermal conductivity (Klingholz, 1977).
     The chemical  composition of rigid  rock wool fibers,  supplied  by  Rockwell
International A/S  Denmark,  were  described by Morgan and Holmes  (1984).  The
major components  were SiO.  (45.5 percent),  CaO  (14.9 percent),  AIJD^ (13.1
percent), MgO  (10.8 percent), FeO  (8.0  percent),  Na20 (2.4  percent),  Ti02
(1.7 percent), and K20 (1.0 percent).  The  fibers had  a count median  length cf
about 27 ± 2.2 pm  (geometric standard deviation)  and a count  median diameter  cf
about 1.1 ±  2.2  pm.   These  sizes of rock wool  fibers  are somewhat thinner  (0.3
to  5  urn  thickness) than  those  of commercial  rock wool, which have length-
weighted nominal  diameters of about 4 pm.
2.1.1.3  Ceramic Fibers.  Ceramic fibers (CAS No.  not available) are  produced
by  combining  kaolin clay or a  mixture  of alumina (Al-CU) and SiD- to  fonr.
alutr.inosilicate  glasses.   The  melted compound is  then blown to  form fibers
used primarily  for high-temperature  insulation  (National  Research Council,
1984).   For even  higher temperature applications, fibers of AKO, and 2^ ere
used because they  possess strong refractory qualities. These fibers  are able
to  retain their  physical  and chemical characteristics in  the presence of high
temperatures.   A1203  fibers  exhibit good chemical  stability, and they resist
oxidation,  reduction,  and attack from most corrosive agents.  They are net
resistant to  hydrofluoric and  phosphoric  acids  and  concentrated  alkalies.
Chemical  analysis  of  one commercially  available  ceramic  fiber  consisting
predominantly of Al.O, revealed the following composition  (Carborundum Company,
n.d.):

                            Major elements	(
                          A1203                    51.
                          Si02                     47.6
                          Na20                     0.3
                          B203                     0.15
                          Fe203                    0.02
                          Trace inorganics         0.2
                          teachable chlorides      <50 ppm

-------
     ZrO- fibers are able to resist aqueous solutions of alkali  metal  hydroxides
and  molten  alkali metal  chlorides  and carbonates at elevated temperatures.
2r02 fibers provide the lowest thermal conductivity of any commercially  availa-
ble refractory material (Zircar Products, Inc., 1978).
     Fibers  of  SiCL are  also  used for  insulative purposes, but they  are
restricted  to low-temperature  applications.   Fibers high  in  SiO. content
exhibit superior mechanical properties (SRI International, 1985).
     Ceramic fibers have an amorphous structure when formed, but some  conversion
to crystalline material  can occur at high temperatures  (International Agency
for  Research on Cancer, 1983).  The  majority  of fibers produced are encap-
sulated or  incorporated  into structures.   The  desired range of  fiber  diameters
produced for  industrial  applications  is  2 to  3.5  pm, but diameters can range
from less than  1 to 12 urn.  Fiber  length  varies  greatly and can range from e
few micrometers to several centimeters (National Research Council,  1984).

2.1.2  Synthetic Fibers
2.1.2.1  Aramid  Fibers.   Aratnid fibers (CAS No.  not  available)  are  aromatic
polyamides synthesized by the reaction of aromatic diamines and aromatic diacid
chloride in an amide solvent via interfacial or solution polymerization.   These
fibers are  of exceptional heat and flame  resistance and have  good to quite
remarkable tensile strength and modulus.   The endothermic peak  in a differential
thermal analysis  (DTA)  thermogram ranges from  >AOO to 550°C, and glass transi-
tions  range  from about  250 to >400°C.   These fibers  do not  melt  because
                                                      V
decomposition occurs simultaneously.  Most aramid fibers are highly crystalline
or are crystallizable, and  their densities range from  1.35 to 1.45 g/cm
(Preston, 1979).  They  are found  in a dry pulp form consisting  of short fibers
(0.5 to 8\mm)  with  fine fibrils attached to the surface.  They have excellent
wear resistance  and  are nonabrasive (Lee et al.,  1983;  Dunnigan et a!., 1984).
Aramid fibers  are divided  into two  categories:   heat- and flame-resistant
fibers (Nomex) and ultra-high-strength, high-modulus fibers  (Kevlar).
     Heat- and flame-resistant  aramid fibers contain a  high portion  of meta-
oriented phenylene  rings.   They can retain useful tensile properties for 1  to
2 weeks at  300°C.   They have high  limited  oxygen index (LOI)  values, volume
resistivities, and  dielectric  strengths, and they  are  hydrolytically stable.
They are resistant to acids, strong bases, and ionizing  radiation,  but they  are
susceptible to degradation by ultraviolet (uv) light (Preston,  1979).

Julv 1986                               2-5        DRAFT—DO  NOT QUOTE OR CITE

-------
     Ultra-high-strength,  high-modulus  aramid  fibers  have very high (>5DO°C)
melting/decomposition points  and are heat-resistant polymers.   However, they
have low  elongations  and therefore become too brittle to be of importance in
heat and  flame  resistance.   Their important  features are  (1) the fibers are of
fine diameter  slightly  less  than 0.013 mm and  (2) they  are produced from
rodlike polymers  exhibiting extremely high intrinsic and inherent  viscosity
values.   On  a specific  basis (i.e., lower specific gravity), these  ararr.id
fibers are stronger  and stiffer than glass or steel  (DuPont,  1976;  Preston,
1979).
2.1.2.2   Carbon Fibers.   Carbon  fibers  are filamentary forms of carbon with
diameters of  5 to  15 jjm and are  characterized  by flexibility;  electrical
conductivity; chemical  inertness except  to  oxidation;  refractoriness; high
performance  varieties;   high  Young's modulus  (coefficient  of elasticity
measuring the softness or stiffness of the material);  and high tensile strength
(Volk,  1S79).  They are corrosion resistant,  lightweight,  have a  high degree  cf
stability to traction forces, high degree of  elasticity,  low thermal  expansion,
low density, high  thermal  conductivity,  and  abrasion  and  wear  resistance  (Holt
and Home, 1978;  Neugebauer et  al.,  1981; Owen  et  a!., 1986;  Wolter,  19E3).
However,   carbon  fibers   are  highly  vulnerable to  shearing  forces  due to a
decrease  in  interlamellar  shear  strength with increasing fiber modulus.  The
Young's modulus,  electrical  and  thermal  conductivity,  and tensile  strength
depend upon,  and correspond  with an increase  in,  the degree of preferred
orientation (Volk, 1979;  Wolter,  1983).   Three types  of carbon fibers  produced
are rayon-based  fibers, polyacrylonitrile (PAN)-based fibers, and  mesophase
pitch-based fibers.
     Rayon-based carbon  fibers  (CAS  No.  not  available) are  used in cloth  form,
are very  stable at  high temperatures,  and contain over  99  percent  carbon.
Low-modulus  fibers  have a  density  of  1.43  to 1.7 g/cm  , filament  tensile
strength  of  345  to  690 MPa  (Megapascal),  and  Young's  moduli of  21 GPa
                                                                             •i
(gigapascal); and  high-modulus  fibers have a density  and  modulus of  1.65  g/cm"
and 345 GPa, respectively, or 1.82 g/cm  and 517 GPa,  respectively.
     PAN-based carbon fibers (CAS No.  not available) have a  ladder  polymer
structure that  is  maintained throughout carbonization and a carbon content  of
92 to 95 percent.  PAN-based  fibers have a density of 1.7 to 1.8  g/cir. , teisi'ie
strength of about 2.4 to 2.75 GPa, and a Young's modulus  of 193 to  241 GPa.
     Mesophase pitch-based  carbon fibers (CAS  No.  not available)  are t*e  onu
true graphite fibers  with  a  carbon content  over 9S.5  percent, an ap:-c»irr.;te
Julv 19BS                               2-6       DRA-7--DO N3T QUC'E C-  CITE

-------
density of 2.0 g/cm  ,  and  a  tensile  strength of 2.0 GPa.  They are produced in
three  forms  differing only  in  filament modulus.   Young's moduli for these
three  types  are  138  GPa  (mat product), 138 to 517 GPa  (fabric),  and 345 GPa
(filament yarn) (Volk, 1979).
2.1.2.3   Polyolefin  Fibers.   Polyolefin fibers are  manufactured fibers in
which  the  fiber-forming  substance  is any long-chain synthetic polymer of  at
least  85  percent  by  weight ethylene, propylene,  or other olefin  units.  Two
important  characteristics  of polyolefin fibers and their nearly  total lack of
water  absorption  and their low specific gravity,  which  ensure that wet prop-
erties  are identical with properties  at standard conditions and lead to  a
higher  covering  power.   Polyolefin fibers are extremely hydrophobic and are
inert  to  a wide  variety  of  inorganic  acids and bases and organic solvents at
roorr; temperature, and to polar solvents at any temperature.   Thermal  degradation
is oxygen-sensitive  and  results  from high temperature formation  of hydroper-
oxides, but occurs at a substantially lower temperature  (290CC) than  with nylon
(360°C) and  polyester.   Ultraviolet  light, degradation is similar to thermal
degradation but occurs preferentially on surfaces, thereby degrading  fibers and
films  more rapidly than  thick objects.   Polyolefin fibers are more combustible
than other common polymers due to  their low limiting oxygen  index (LOI) values
and  are  inherently  difficult to dye.   They exhibit  elastic  recovery (70  to
100%)  that is  the fractional length  of fibers recovered  from an  initial exten-
sion.  The product of elastic recovery and work  recovery  corresponds  to the
resiliency of  the fiber.   Polyolefin fibers also  exhibit creep (time-dependent
deformation) and  stress relaxation (spontaneous  relief of internal stress)  that
are  dependent  upon,   and inversely related  to increases in,  fiber molecular
weight, molecular orientation, and heat crystallization  (Buchanan, 1979).
     Polyolefin  fibers  of  commercial  importance are  polypropylene (CAS
No. 25085-53-4)  and  polyethylene  (CAS  No. 9002-88-4).   Polypropylene  fibers
have  a significantly  higher melting  point,  equal  tenacity and abrasion
resistance,  and  substantially better  resilience  than polyethylene.  Melting
points for polypropylene and polyethylene  are 170 and 130°C, respectively, and
densities are 0.90 and 0.95 g/cm , respectively (Buchanan, 1979).

2.1.3  Minerals
2.1.3.1  Attapulgite.  Attapulgite (CAS No. 12174-11-7)   is a naturally  occurring
nonplastic clay  that belongs to  a  group of commercially defined  clays known as
Fuller's earth.   Attapulgite is  composed of short asbestifonr fibers of the
July 19BE                               2-7       DRAFT—DO  NOT QUOTE OR CITE

-------
mineral palygorskite, which  consists  of  SiCL,  AI-O,,  and  MgO  in  the  form of a
complex metallic aluminosilicate with a composition of Mgt(OH)2(H20)4Sig020'4H-0
(Zoltai, 1986;  Zoltai and Stout,  1984;  Huggins etal.,  1962).   Mineralogic
composition varies greatly with the site of origin, and the  presence  of minerals
associated with attapulgite  formation,  such as calcite, quartz,  and  smectite,
have a  direct influence  on  determining its purity (Bignon et al.,  1980).
Attapulgite is  grayish-white in color  end occurs  as  elongated,  lath-shaped
crystals in  bundles that  comprise  thin sheets of  minute interlaced  fibers
(International Agency for  Research  on Cancer,  1987).   It is  soft  and easily
cleaned, chemically  inert, and  its  properties  are  unaffected  by  changes in pH.
Bignon et al.  (1980)  determined the mean size characteristics of  attapulgite
originating from  Caceres, Spain,  to  be 0.8 urn  in length  and  0.06  urn  in
diameter.   Bulk  chemical  analysis  performed  on calcinated samples  of  the
Spanish attapulgite revealed  the following  chemical composition:
              Major elements (%)           Minor elements (mg/kg)
            Si02             80.1             Mn          750
            A1203             4.6             P           100
            Fe203             4.6             Ti         2800
            MgO               9.7             Zn          100
            CaO               3               Sr          100
            Na20              0.5             Zr          120
            K20               0.9             Nb          100
                                              Ba          730
                                              La          730
                                              Ni           11
                                              Cr           90

2.1.3.2  Zeolites.   Zeolites  are a group of  naturally  occurring  crystalline
sedimentary minerals  consisting of over 40 hydrated aluminum  silicates  that
contain alkali  or  alkali  earth elements.   They are poorly  consolidated  soft
rocks  formed  by the  alteration of large beds  of  volcanic ash, usually in
alkaline lakes  (Leineweber, 1980).   In the United States,  over 300  deposits
are thought to  exist in locations across 25  States.  The basic component  of
the crystal structures  of  zeolites is the SiO^ and AID. tetrahedral  coordina-
tion.   They have  a three-dimens.ional  open network that  allows the addition of
cations such  as  Na,  Ba,  Ca, and K to  balance  the negative charge  incurred whe--
Al is  substituted  for Si.   The centra"1 Si  or Al  atom is surrounded by  four
oxygen atoms, each of which is shared by  the  neighboring tetrahedron, creating

-------
a honeycombed  crystalline  structure  that  possesses  unique  features character-
istic to each zeolite (Rom et al.,  1983).
2.1.3.2.1  Erlonite.  Erionite  (CAS  No.  66733-21-9) is  a fibrous zeolite that
has been  found in deposits in Arizona, California,  Nevada,  Oregon,  and Uta^ in
the form  of  thin, pure  beds within sedimentary tuff sequences or as outcrops
in desert  valleys of the intermountain region (Rom et al., 1983).   Erionite
fibers occur as  white prismatic  crystals  in radiating groups, as either single
needles or in  clusters  and are, on the average, shorter than asbestos fibers
(International  Agency for Research on Cancer,  1987).  They  can  attain  a maximum
length of  50 urn, and widths  generally range  from  0.25 to 1.5 urn, although
fibers exhibiting widths  of 0.01 to 5.0 pm have  been recorded  (Wright et al.,
19E3;  Suzuki,  1982).  Mineralogic  investigation of  erionite samples taken fro-
sites near the town of  Karain in  the  Cappadocia region of Turkey yielded tne
following  chemical   composition:   (Na.l^CaMg)^ 5A1qSi27°72**7  H2°  ^purny,
19S3).  Documented  uses  for erionite  are:  noble metal-impregnated catalyst  in
a hydrocarbon  cracking  process  in a U.S. plant; house building materials;
increase soil  fertility  and control  odors in  livestock  production.   Extent  of
its use for  house building is minor and  unknown  for the latter  two  purposes
(International  Agency for Research on Cancer,  1987).
2.1.3.3  Talc.    Talc (CAS No.  14807-96-6) is  composed of pulverized,  natural,
foliated hydrous  magnesium silicates  originating  from nonasbestiform  minerals
that  are  low in crystalline SiO«  content.  It consists of a layer of Mg(OH)2
that  is held between two sheets of SiO-.   Weak van  der  Waals'  forces  hold  the
sheets together.  Platey  crystals  of talc slide past one  another  to  produce
the smoothness associated with  talc  (Davies  eta!., 1983).   The  chemical
composition of  talc is  Mg2(Si205)2Mg(OH)2, with theoretical contents of Mg and
Si of 19.2 and 29.6 percent, respectively (Hanson et al., 1985).  Its  appearance
can be white, yellowish, greenish, or gray depending on  the number  of  impurities
present.    Impurities  and/or minerals  commonly found in commercial talc include
MgO,  TiO., calcium  silicate,  iron  oxide  (a color determinant), carbonates, and
aluminates.  Tremolite and/or anthophyllite asbestos are also found as contami-
nants of  some  talc  deposits (National Research Council, 1984; Lockey, 1981).
The specific gravity of  talc  is  2.6  to  2.9.   It  has perfect cleavage, is
insoluble in water, and depending-on its purity,  exhibits  varying resistance  to
acids, alkalies,  and heat (General Electric, 1978;  International  Agency  for
Research on Cancer, 1987).

July  ISSc                               2-9       DRAFT--DO N:~ Qj"£  C-:  CITE

-------
2.1.3.4  Verim'culite.   Vermiculite  (CAS  No.  1318-00-9) is a micaceous  hydrate
of magnesium-iron-aluminum  silicates.  The chemical composition of vermiculite
varies, but an acceptable general formula is  (Mg.Fe.Al), (Al,Si)4010  (OH)2«4H20
(Bates and Jackson,  1980).   Vermiculite  crystals are  composed of  two silicate
layers connected  by  a hydrous layer.  The thickness of the unit cell in fully
hydrated materials  is about 14  angstroms (Gruner,  1934).   All of the major
commercial deposits  occur  in ultramafic  and  mafic host  rocks.  The  material
that  is  mined  is  mixed-layer vermiculite-biotite and  vermiculite-phlogopite.
Other  minerals  commonly present  in the deposits include  quartz,  feldspar,
apatite, corundum, chlorite,  asbestos,  talc,  and clays.  The  main commercial
mining deposits  of  Vermiculite  are  located  in Montana, Virginia, and South
Carolina as well  as  South  Africa (Dixon  et a!.,   1985).  Vermiculite is  a soft
mineral and unique among minerals in its ability to exfoliate or  expand up  to
20 times its original size when heated between 300 and 2000°F (Meisinger, 1955;
Lockey, 1981).   Exfoliation can  also be  achieved by chemical  processes such as
soaking in hydrogen peroxide, weak acids, and other electrolytes.
     Crude vermiculite  has  a loose bulk density of 640 to 1000 kg/m  ,  whereas
exfoliated vermiculite  expands to a bulk density of 56 to 192 kg/m .  It also
has a  very  low  thermal  conductivity and a relatively  narrow range of  cation
exchange capacity, primarily associated with  magnesium.

2.2  PRODUCTION AND INDUSTRIAL USES
2.2.1  Hanmade Mineral Fibers
2.2.1.1  Fibrous  Glass.  During  production,  glass fibers can  be  divided into
two basic types:   staple and continuous.   Staple fibers are produced  by melting
glass feedstock in furnaces and pulling the liquid melt through small orifices,
forming a  thin  coat of  fibers that are  then collected  on  a  spinning  drum.
Continuous fibers are mechanically  extruded  and collected on drums  as  yarn
(International  Agency for  Research on  Cancer,  1983).   Textile glass  fibers are
produced and  sold as  continuous strands (drawn fibers),  staple yarns  and
slivers (blown fibers),  roving and cake packages, chopped  and milled  fibers,
and mats.   The  major U.S.  producers of  textile glass  fibers  are listed in
Table 2-2.
     Glass fibers are  used primarily in reinforced plastics including  unsatu-
rated polyester resins and polyamide, polyacetyl, and  polybutylene terepnthalate
resins.  Other uses include paper and tape reinforcements, automotive tire

'•••>•• IOGO                               2-10       DRAFT--DO NOT QJOTE OR CITE

-------
           TABLE 2-2.   MAJOR U.S.  PRODUCERS OF TEXTILE  GLASS  FIBERS
      Company and location
          Product types
Owens-Corning Fiberglass Corp.
     Aiken, SC
     Amarillo, TX "
     Anderson, SC
     Ashton, RI
     Huntingdon, PA
     Jackson, TN

PPG Industries, Inc.
     Lexington, NC
     Shelby, NC

Manville Corp. &
     Denver, CO'
     Elizabethton, TN
     Waterville, OH

Reichhold Chemicals, Inc.
     Bremen, OH
     Irwindale, CA
     Nashville; TN

Lundy Electronics and Systems,  Inc.
     Pompano Beach, FL

Certainteed Corp.
     Wichita Falls, TX

F.G.I.  Fibers,"Inc.
     Amsterdam-, NY

Nicofibers, Inc.
     Shawnee, OH
  Ywrti, roving, spun roving, woven
  roving, mat, chopped, milled,
  staple yarn, and sliver
   Yarn,  roving and woven roving, mat,
   and  chopped strand
  thopped  strand, mat, and  sliver
   Mat,  roving  and  chopped  roving,
   woven roving,  mat  chopped,  and
   milled
   Metallized glass,  roving,  strand,
   and mat

   Roving,  woven  roving, mat,  chopped
   strand,  and combination  product

   Roving and chopped
   Mat
Source:  SRI International (1985).


reinforcements,  and  home furnishings (Table 2-3)  (SRI  International,  1965).

The annual  production  capacity of  textile-glass fibers  in  November  1983 was

1618  million poonds.   Approximately 20 million pounds were  imported  and

37.5 million pounds  were exported  in 1983 (SRI International,  1985).   The

predicted growth  rate  for glass fibers is 17.5 percent per year (PTS Prorr.t,

1988).  Data on the uses of glass fibers in insulation were not available.
July 19SB
2-11
DRAFT—DO NOT QUOTE OR  CITE

-------
           TABLE  2-3.   U.S.  CONSUMPTION OF TEXTILE GLASS FIBERS, 1983
 Industrial  and consumer-type products                      Millions of pounds
 Reinforced plastics                                                632.7
 Paper and tape reinforcements                                       33.5
 Electrical  applications                                           	
 Automotive tire  reinforcements                                      24.2
 Other                                                             499.9
 Home  furnishings                                                   13.7
     (curtains, drapery and
     upholstery,  other)
 Total                                                             	
                                                                  1204.0
 Source:   SRI  International  (1985).

 2.2.1.2   Mineral Wool.  The National  Research  Council  (1984) estimated that
 total mineral  wool production in  the  United States was approximately 200,000
 metric tons.   Mineral wool, which has many of the same applications as fibrous
 glass, is mainly used in  insulation  for  industrial  processes.
 2.2.1.3   Ceramic Fibers.    Ceramic  fibers  can be grouped  into  three  general
 categories  of  production:   (1) fibers that are produced in the form of blanket,
 board, or module  as   insulative and refractory materials (alumina- silica,
 Al-0,, and Zr-O,); (2) fibers prepared  in the  form of cloth  for radiation and
 heat-protective  purposes  (SiO.); and (3) fibers used in miscellaneous applica-
 tions (potassium titanate,  boron nitride, yttria, tantulum, and aluminum-boron-
 silicon  fibers).  These exotic  fibers are  in limited production and are used
 only  for highly  specialized applications (SRI International,  1985).
      Alumina-silica  fibers  are produced either by  blowing or centrifugal melt
.spinning.   Alumina-silica producers  are  listed  in Table 2-4.  The total produc-
 tion  volume of  alumina-silica fibers is believed  to have been approximately
 100 million pounds in 1983,  about  75  percent  of  industry  capacity.   With
 exports   of  approximately  10 million  pounds and  negligible  imports, 19E3
 domestic consumption was approximately 90 million  pounds  (SRI  International,
 1985).   Consumption of alumina-silica fibers  by industry is shown in Table 2-5.
      Fibers  of Al.O,  are produced  from  a  solution of aluminum salts using  a
 proprietary process.   They are  manufactured by Imperial  Chemical  Industries
 Limited  (ICI), Mond  Division, at  Widnes,  United  Kingdom,  with a capacity of
 approximately  1.5 million  pounds  per year.   Bulk fiber  and mat  forms  are
 imported into the United States through ICI Americas,  Inc., Wilmington, DE.
 U.S.  consumption in  1983 was approximately  150,000 pounds  (SR! International,
 1985).

-------
                TABLE 2-4.   MAJOR U.S.  ALUMINA-SILICA  PRODUCERS
                                                     Estimated  annual capacity
                                                       as  of January 1, 1954
         Company                                        (millions  of pounds)
The Carborundum Electro Minerals Co.                            55
     Niagara Falls, NY
     New Carlisle, IN
     Ne* Iberia, LA

Babcock and Wilcox                                             40
     Augusta, GA
     Ponce, PR

Johns-Manville Corp.                                            20
     Waukegan, IL

Combustion Engineering, Inc.                                    10
     Erwin, TN

Source:  SRI International (1985).
             TABLE 2-5.   U.S.  CONSUMPTION OF ALUMINA-SILICA FIBER
                          PRODUCTS BY INDUSTRY, 1983
Industry
Iron/steel
Nonferrous metals
Chemicals and petroleum
Ceramics
Automobile
Other8
Total
Consumption
(millions of pounds)
19
11
23
7
11
19
90
Percent of
total
21
12
25
8
13
21
100
alncludes aerospace (acoustical insulation), shipbuilding (fire protection
 and insulation), laboratory OEM industrial furnaces.

Source:  SRI International (1985).
July i9££                               2-13      DRAFT--DO NOT QUOTE OR CITE

-------
     Fibers of Zr^O^  are  manufactured using a precursor process  that  req-ires
impregnation  of  an organic fabric with zirconium  salt  solution.   Subsequent
pyrolysis  removes  the organic  substrate  and  transforms the salt residue to
Z^Oj.  Zircar Products, Inc., located in Florida,  NY,  is the only manufacture*-
of these  fibers  in the United States; their  facility  is believed to. have an
annual  capacity  of 10,000 pounds for  bulk  fiber and 1,000-2,000 pounds  for
Zr203 cloth (SRI International, 1985).
     Fibers of  SiO- are actually high-purity silicon dioxide materials  that
are used  for  many  insulative,  heat-protective, and reinforcement  applications.
Silica  fibers  are   produced  by two distinct  processes.  The  first  involves
leaching of all other oxide containments  by immersing glass yarn  in hydrochloric
acid.    The  second   involves  drawing  the   fibers  from a  heated quartz  rod.
Leached glass  products are  produced  at  the Materials  Division  of  H1TCO, c
subsidiary of Armco,  in Gardena,  CA,  with  an annual  capacity of 1.1  millic-
pounds.    They  are   also produced  by  the  Haveg Division  of  AMETEK,  Inc.,  in
V»i lrr,i ngton, DE, with  an annual capacity  of 1.0  million pounds  (SRI Interns-
ficric",  19S5).  Overall  ceramic  fiber production totaled 33 million pounds  ir
1987 aid  is predicted to grow at a rate  of 26 percent  per  year (PTS  Pror.t,
1986).   These  fibers have major  commercial applications in a wide range cf
refractory and insulative products (Table 2-6).
     A  laser-based method that produces  stronger  and more  flexible cerarric
fiber faster  than  conventional  techniques has been developed by MRI Ventures
in Kansas  City,  MO.  This  technique involves  the injection  of silicon  chloride
and methane gases  into a proprietary reactor and  exposing  them  to radiation
from an yttrium/aluminum  laser.   The  reactor temperature and pressure can be
varied  to  design the fibers  for specific  applications.   This  technique is also
used to produce other fibers (PTS Promt,  1988).

2.2.2  Synthetic Fibers
2.2.2.1  Aramid Fibers.  Aramid fibers are  usually produced by solution polym-
erization polycondensation.  A diacid chloride and a diamine are  polymerized  in
an inert  amide  solvent.   The polymer  is  then  spun  from  the  solution and dried.
Both types of  fibers --  heat-/flame-resistant (Nomex)  and  ultra-high-strencth,
high-modulus  (Kevlar) -- are  relatively  expensive  in  comparison  to  other
synthetic fibers primarily because of low-volume production.  In 1975,  the £.I|
du Pont de Nemours   and Company, Inc.,  in  Richmond,  VA, the only producer  lisle:

-------
         TABLE 2-6.   REPRESENTATIVE  APPLICATIONS  FOR CERAKIC, ALUMINA
                      ZIRCONIA,  AND  SILICA  FIBER  PRODUCTS

Ceramic Fiber Products
     Furnace, kiln,  reformer,  boiler linings
     Flexible high-temperature pipe  insulation
     Pressure and cryogenic vessel fire  protection
     Thermal reactor insulation
     High-temperature gasketing
     High-temperature filtration
     Nuclear insulation applications
     Replacement for asbestos  blankets
     Fill or packing material  in
          Expansion  joints
          Tube seals
          Furnace base seals
     Manufacture of  secondary  forms
     Thermal and electrical insulation
     Combustion chamber liners
     Ingot mold linings
     High-temperature baffles  and muffles
     Molten metal trough and distribution linings
     Heat shields
     Hct-gas duct linings
     Refractory backup for brick and castables
     Patching of refractory cracks and  fissures
     Coating for metal liners
     Substitute for  wet asbestos fibers
     Expansion joint packing
     Pipe, hose, cable protection
     Jet-engine, fuel-line insulation
     Thermal battery liners
     Burner blocks

Alumina Fiber Products
     Furnace, kiln,  reformer,  boiler linings
     Expansion joints
     Tube seals
     Furnace base seals
     Manufacture of secondary product processes
     Molten metal trough and distribution linings
     Resin reinforcement
     Filtration

Zircom'a Fiber Products
     Thermal insulation in laboratory and industrial  furnaces
     Radiation shielding in vacuum furnaces
     High-temperature chemical and gas  filtration
     Battery and fuel-cell separators
     Reinforcement for ablative materials in nozzles,  nose tips, and
          heat shields
     Insulation in hot pressing, crystal growing, and annealing furnaces

                                   (continued on the following page)
                                                            wr.T rvi'"7F fi-,

-------
          TABLE 2-6.  REPRESfNTATIVE APPLICATION FOR CERAMIC,  ALUMINA
	ZIRCONIA. AND SILICA FIBER PRODUCTS (continued)	
Silica Fiber Products
     Reinforcement of resin molding compounds
     Thermal-insulative packing
     Wrapping for seals and gaskets
     Resin prepregs for aerospace
     High-temperature gasketing
     Protection of workers and equipment during welding and burning
          operations
     Thermal protection of hoses and cables
     Reinforcement and insulation in elastomeric ablative coatings
Source:   SRI International (1985).

in available literature,  announced an annual capacity for Nomex  and  Kevlan  of
9,100 and  23,000 tons/year, respectively  (Preston,  1979;  SRI  International,
1967).   The predicted growth rate for aramid fibers is 14 percent per  year.   By
1992, they will represent the greatest production volume (PTS  Promt,  1985).
     Due to  their  resistance  to heat  and  flame,  Nomex fibers  are used in
industrial  protective clothing, carpets, tents, insulation paper  for  electrical
motors and  transformers,  ironing board  covers,  and cargo and  boat covers.  The
ultra-high-strength,  high-modulus  Kevlar fibers are used in tires,   cables,
parachutes, body armor, fire hoses, V- and conveyor belts, and high-performance
boats and aircraft (Preston, 1979).
2.2.2.2  Carbon Fibers.   Carbon fibers are produced  by subjecting   rayon,
polyacrylonitrile,  or  mesophase  pitch  to controlled  pyrolysis  (Owen
et a!.,  1986).   U.S.  demands  for  all  high-performance  carbon  fibers have
increased  from  about 1  million  pounds  in 1980 to  about  5.2  million  pounds
in 19B7,  with  50  to  75  percent of the  fibers used  in  the  aircraft and
aerospace  industry (Table  2-7).   The  predicted  growth rate for carbon
fibers  is  19.5 percent  (PTS Promt,  1988).   The  U.S.  producers and  their
annual  capacity for all  carbon  fibers  are  shown in Table 2-8.   The  process
chemistry  and  uses for  rayon,  polyacrylonitrile,   and mesophase pitch-based
carbon fibers are discussed below.
     Low-modulus,  rayon-based  carbon fibers  are  produced  by heat-treating
rayon at 200 to 350°C in a reactive atmosphere.  The fibers are then  carbonized
at  1000 to 2000°C  followed  by  an  additional heat  treatment  at  temperatures
approaching 3000°C.  To produce  high-modulus fibers, the yarn must  be stretched

-------
          TABLE 2-7.  U.S. DEMAND FOR HIGH-PERFORMANCE CARBON FIBERS
Industry
Aircraft/aerospace
Sporting goods
Industrial
Total
1980
(thousands
of pounds)
550
300
150
Wo
1984
(thousands
of pounds)
2000
500
400
I5U5
19878
(thousands
of pounds)
3900
600
700
5200
a
 Estimated demand.

Source:  SRI International (1985).
               TABLE 2-8.  MAJOR U.S. PRODUCERS OF CARBON FIBERS
Annual capacity
(thousands
Company of pounds)
Arr.occ Corp.
Greenville, SC
Ashland Oil
Catlettsburg, KY
AVCO Corp.
Lowell, MA
BASF Corp.
Rock Hill, SC
The Dexter Corp.
Sacramento, CA
Great Lakes Carbon Corp.
Eli2abethton, TN
Hercules, Inc.
Bacchus, UT
Owens-Corning Fiberglass Corp.
Gardena, CA
Polycarbon, Inc.
800
500
200
100
600
750
1400
2700
100
100
Type/brand Name
Polyacrylonitrile
Mesophase pile!-:
Petroleum pitch
Polyacrylonitrile
Polyacrylonitrile
Polyacrylonitrile
Polyacrylpnitri le
(Fortafir)
Polyacrylopitrile
(Magnamite )
Polyacrylonitrile
(Hl-TEX*)
Pitch and rayon
   North Hollywood, CA

Stackpole Corp.
   Lowell, MA

   Total
 450


F700
Polyacrylonitrile
Source:  SRI (1967).
July 198S
        2-17      DRAFT--DO NOT QJC-TE  OR CITE

-------
during the last heat treatment.   The carbon content of these  fibers  is approxi-
mately 99 percent.  Production volumes are not available;  however, the estimated
consumption for 1970-76 was 100,000 to 250,000 kg/year (Volk,  1979).
     Rayon-based  carbon fibers  are  used in tire cords and aerospace applica-
tions, e.g., phenolic-impregnated heat shields and carbon-carbon  composites  for
missile parts and aircraft brakes (Owen et al., 1986;  Volk, 1979).
     Polyacrylonitrile-based  carbon fibers are  produced  in  a wanner very
similar to that  used  to produce of rayon-based fibers.   PAN  fibers are  heat
treated, under tension, in an oxidizing atmosphere at 190 to 280°C.  At this
stage, a well-oriented ladder polymer structure is developed  and  is  essentially
maintained throughout  the  production  process.  'The yarn is carbonized in  an
inert  atmosphere,  usually between 1000 and 1300°C.   The  fiber  is then heat
treated again  at  2500 to  3000°C.  The filaments are gathered into tow bu-.Cles
and wound.   The carbon content of PAN fibers is approximately 92  to  95 percent.
Production and consumption statistics  are  not  available (Holt and Horne, 197£;
Jones et al., 19£2; Volk,  1979).
     PAN-based ca-bcr, fibers are used in structure fabrication for the ae-cspa:€
industry,  sporting  goods  (e.g.,  golf  clubs, fishing rods, tennis  rackets, bow;
and arrows,  skis,  sailboat masts and spars),   textile  and computer  machi-ne-y,
automotive and general  transportation,  musical instruments,  reinforcement  fcr
plastics,  and  prostheses  and artificial joints (SRI International,  1985; Holt
and Horne, 1978;  Neugebauer et al., 1981;  Owen et al., 1986;  Volk, 1979).
     Mesophase pitch-based carbon fiber production starts with  a commercial
coal tar or  petroleum pitch which  is converted through heat  treatment into  a
mesophase state.    The  pitch is  spun into a highly oriented  fiber and is  ther-
moset  in an  oxidizing atmosphere.   The  yarn is carbonized and then  graphitized
at  3000°C.   The  high orientation  is  maintained  during  thermosetting  and
enhanced during  carbonization and graphitization.   The carbon content, is over
99.5 percent.  Production  and consumption  statistics  are  not available  (Volk,
1979).
     Mesophase pitch-based  carbon  fibers  are  used for electrical conduction,
resistance to  heat distortion,  improved wear, and stiffness  critical applica-
tions  (Volk, 1979).
2.2.2.3  Polyolefin Fibers.   Polyolefin fibers are fabricated commercially by
a  melt-extrusion  technique.   The fundamental   elements of this technique are
expulsion of the  polymer  through a  spinneret,  solidification,  and winding  onto
packages.
Julv 1988                               2'^8      DRAFT--DO N:T QUC-TE OF,  CI'E

-------
     Polymer pellets  are melted and metered through a  filter  system  to  the
spinneret where they  are extruded under pressure.  Fiber form depends upon the
size and  number of  holes in the spinneret.  Extrusion occurs at 100 to 150°C
above the melting point of the polymer.
     As heat is expelled, the fibers begin to solidify.   Due  to  the crystalline
and glass transition  properties of  polyolefin fibers, solidification occurs by
crystallization between  the  melting and  glass  transition temperatures  and  can
continue after winding.
     Takeup velocities  for winding  the  fiber  onto packages  vary from 1 to
2 meters per second (m/s) to speeds  as great as  117 m/s.   Slow  velocities allow
further refining  steps  to  be  incorporated.   Such steps include drawing to
induce further crystallization, texturizing  to  impart  bulk and/or elasticity,
and annealing for stability of fiber structure  and properties (Buchanan,  1S7S).
     Production of polyolefin fibers has  increased steadily due  to an  increased
demand for continuous filament yarn  and monofilaments.   Production and consump-
tion values for 1983  were 803 and 767 million  pounds,  respectively  (Buchanan,
1S79;  Chemical  Economics Handbook,  1983).
     Over 40 producers  of polyolefin fibers  may  be  found in  the United States,
but greater than  50 percent of  the  film and  filament products are manufactured
by Amoco  Corporation  (Chemical  Economics  Handbook, 19S3).  The U.S.  producers
and their annual  capacities are listed in Table 2-9.
     Polyolefin fiber  usage  is  concentrated  in  home furnishings and industrial
areas.   Home furnishing uses include carpet facing and  backing,  upholstery,
woven bags,  and  wall-covering materials.  Industrial uses include  synthetic
turf,  cordage  and  twine, webbing, agricultural  fabrics,  fishing line and nets,
sewing thread, and  book binders (Buchanan, 1979; Chemical  Economics  Handbook,
19B3).

2.2.3  Minerals
2.2.3.1  Attapulgite.   Attapulgite,  which is produced from  regions including
Attapulgus, GA,  and  Quincy,  FL,  is composed predominantly of lath-shaped,
amphibolelike  palygorskite.   Most of  the Fuller's earth produced  in other
areas  of  the United States contains varieties  of montmorillonite.  Attapulgite
has numerous commercial  applications (Table 2-10) due in part  to its sorptive
and colloidal  properties.   It  finds wide application in both  absorbent and
gelling or thickening areas.  The thixctropic properties of attapulgite day

July 1988                               2-19      DRAFT—DO JOT QUOTE OR CITE

-------
             TABLE 2-9.   MAJOR U.S.  PRODUCERS  OF  POLYOLEFIN FIBERS
          Company
Annual capacity
  (thousands
   of pounds)
              Type
Amoco Corp.
  Amoco Fabrics and Fibers Co.
    Andalusia, AL
    Bainbridge, GA
    Hazlehurst, GA
    Nashville, GA
  Sunburst Yarns, Inc.
    Roanoke, AL
    Afton, VA

Avtex Fibers, Inc.
    Front Royal, VA

Camac Corp.
    Bristol, VA

Exxon Corp.
    Hangman, KS
  Sumn-.erville, SC

Fibres South, Inc.
    Trussville, AL

Hercules, Inc.
    Oxford,  GA

Integrated Resources, Inc.
  Fibron Corp. of Tennessee
    Chattanooga, TN
  Synthetic  Industries, Inc.
    Chickamauga, GA

Phillips Petroleum Co.
    Rocky Mount, NC
    Spartanburg, SC
Polyloom Corp. of America
    Dayton, TN

Wayn-Tex, Inc.
    Waynesboro, VA
        >295
          20


          20
          35
         <20
          20


         200




          15

          30
          30
          70
          22
          26
        Monofilament, film
        filament yarn, and
        staple
        Filament yarn
        Monofilament
        Film
         Film, monofilament,
         and  filament yar-

         Filament yarn and
         staple
         Film

         Film
         Filament yarn
         Filament yarn,
         staple, and  tow

         Film
         Monofi lament,  film,
         filament yarn,  ana
         staple
                                (continued on the following page)
July 1968
         2-20
DRAFT--DO N:T QUOTE OF. CITE

-------
       TABLE 2-9.  MAJOR U.S.  PRODUCERS OF POLYOLEFIN FIBERS  (continued)
          Company
Annual capacity
  (thousands
   of pounds)
Type
Polyloom Corp. of America                22
    Dayton, TN
Wayn-Tex, Inc.                           26
  Waynesboro, VA
Wellington Technical Industries, Inc.
  Wellington Leisure Products, Inc.
    Pilot Mtn, NC                        10
    Wellington Synthetic Fibres, Inc.
    Leesville, SC                        10
Other                                   166
Tctai                                  ~9£5
                            Film
                            Monofilament,  film,
                            filament yarn,  and
                            staple
                            Monofilament and film
                            Monofilament,
                            filament yarn,  ar,d
                            staple
aNct available.
Source:   SRI International (1987).
provide the important thickening and viscosity controls necessary for suspending
solids.  Mineral thickeners  are  used in such diverse  markets as paint, joint
compound cement, and  polishes.   The reported estimate of attapulgite produced
or  used  in the United States  in 1985 was 2,059,281 short tons  (PTS Promt,
1988).
2.2.3.2  Zeolites.  Zeolite  minerals have  many  actual  and potential  commercial
uses  due to  their  reversible dehydration (Wright et al.,  1983), ion-exchange
behavior, thermal  and acid stability, and molecular absorbency.   Erionite has
been  used  for  nitrogen retention in fertilizers;  in concrete aggregate; and  in
road  surfacings  (National Research  Council,  1984).  Other uses of  zeolites
include heat exchanging in solar energy, decontamination of radioactive wastes,
and  stack  scrubbing  in  air  pollution  control  systems  (Rom et  al.,  1983).
Annual  U.S.  production  of natural  zeolites  was  estimated to  be  2500 to
5000 tons for the years 1982-1984 (International Agency for Research on Cancer,
1987).
2.2.3.3  Talc.   The  amount  of crude talc and  pyrophyllite produced in the
United States in 1984 is shown in Table 2-11.   The  end uses for ground talc
'•••••• •"•>«•• *                               ?-?i      DRAFT—nr. N-T QUOTE OR CITE

-------
          TABLE 2-10.   FULLER'S EARTH SOLD OR USED  BY  U.S. PRODUCERS
                         IN 1984, BY USE  (short  tons)
Use
Domestic
Adnesi ves
Animal feed
Drilling mud
Fertilizers
Filtering, clarifying, decolorizing
mineral oils and greases
Medical pharmaceutical, cosmetic
Oil and grease absorbents
Paint
Pesticides and related products
Pet weite absorbents
Other0
Miscellaneous
Total
Exports:
^ * ^ ^ .' K _ ft
ur 1 i i i r/g JTiuC
Oil and grease absorbents
Pesticides and related products
Pet waste absorbents
Miscellaneous
Total
Grand total
Attapulgite

1,798
12,139
105,716
52,291

18,017
105
209,831
6,606
96,200
291,943
25,872
26,710
849,228

1,245
41,020
9,135
27,642
10,090
89,133
938,361
Montmorillonite

...
...
...
10,576

784
...
220,878
--
120,223
506,279
1S.841
30,911
909, 4S2

...
38,138
373
12,295
466
51,292
960,784
Quantity

1,796
12,139
105,716
62,867

18,801
105 .
430,705
6,606
216,423
798,222
5,713
59,621
l,71c,72C'

1.2-5
79,15£
9,53£
39,927
10,5/6
140,425
1,859,145
alncludes paper filling and roofing granules.
 Includes common brick; catalyst oil refining;  chemical  manufacturing  glazes,
 glasses, and enamel; gypsum products; mortar  and cement refractories;
 plastics; pottery; and sanitary ware.
clncludes paint and uses not specified.
Source:  U.S.  Bureau of Mines (1986).

and pyrophyllite are  shown in Table 2-12.   The largest  portion of ground  talc
is used  in ceramics  (37  percent), followed by paint  (19 percent),  paper
(10 percent),  roofing  (9  percent), other (9 percent),  plastics  (7  percent),
cosmetics  (5 percent),  rubber (3 percent),  and insecticides  (1 percent)  (U.S.
Bureau of Mines, 1986).
2.2.3.4  Vermiculite.   The  amount  of  vermiculite produced  in the  United States
         ™ — - — - • "-                                                              I
in 1985  was  285,000 metric  tons (PTS Promt,  1988).   The largest  domestic
producer is W.R. Grace and Co., with mines in  Libby, MT, and  Encree, SC.
                                                   **» r. r T _ _ ^ r»

-------
          TABLE 2-11.   U.S.  PRODUCTION OF CRUDE TALC  AND  PYROPHYLLITE
                         IN 1984 (thousand short tons)
State
California
Georgia (talc)
North Carolina
Oregon (talc)
Texash(ta1c)
Other0

Total
• Quantity
74
15
87
a
283

711
IITo
 Less than one-half unit.

 Includes Arkansas, Montana, New York, Vermont,  Virginia,  and Washington.

Source:   U.S.  Bureau of Mines (1986).
            TABLE 2-12.   END USES FOR GROUND TALC AND PYROPHYLLITE
                             (thousand short tons)
Use
Ceramics
Cosmetics
Insecticides
Paint
Paper
Plastics
Refractories
Roofing
Rubber-
Other0
Total
Talc
358
44
8
189
100
67
4
86
29
85
970
Pyrophyllite
2l
m ^ u
14
1
—
1
21
6
11
"83
Quantity
367
44
22
190
100
68
25
92
29
96
1053
'incomplete data; some cosmetic talc is known to be included with "other."
 Includes art sculpture, asphalt filler and coatings, crayons, floor tile,
 foundry facings, rice polishing, stucco, and uses not specified.
cNo data reported.

Source:  U.S. Bureau of Mines (1986).
July 19S£                               2-23      DRAFT-DO N3T QuCTE OR CITE

-------
Exfoliated vermiculite  is  produced by W.R.  Grace and  Co.  at  29 plants  in 24
States (of a  total  of 43 plants in 29 States).   Vermiculite is also mined and
processed by  Patterson Vermiculite  Co.  near Enoree,  SC,  and  by  Virginia
Vermiculite, Ltd., Louisa County, VA (Meisinger, 1985).
     The  largest  portion of  exfoliated  vermiculite is used in  construction

aggregates  (51 percent)  followed  by  insulation  (26 percent),  agriculture
(22 percent),  and  other end  uses  (1  percent).   Vermiculite is sold in five
grades.
2.3  REFERENCES


Bates, R.  L. ;  Jackson,  J.  A., eds.  (1980)  Glossary  of  geology.  2nd  ed.  Falls
     Church, VA:  American Geological Institute; pp.  210, 373, 685, 710.

Bignon,  J. ;  Sebastien,  P.; Gaudichet,  A.;  Jaurand,  M.  C. (1980)  Biologies!
     effects of  attapulgite.  In:  Wagner, J.  C.;  Davis, W., eds.  Biologies!
     effects of  mineral fibres  = Effects biologiques des fibres  minerales:
     proceedings of  a  symposium,  v.  1;  September 1979; Lyon, France.  Lyon,
     France: World Health  Organization,  International Agency for  Research  or.
     Cancer; pp. 163-181.  (IARC  scientific  publication  no. 30; 1NSERN symposia
     series: v. 92).

Buchanan,  D.  R.  (1979)  Olefin  fibers.  In:  Grayson, M. ;  Eckroth, D.,  eds.
     Kirk-Othmer encyclopedia of  chemical technology: v.  16.  3rd ed.  New York,
     NY:  John Wiley & Sons, Inc.; pp. 357-385.

Davies, R. ; Skidmore, J. W.; Griffiths, D. M.; Moncrieff,  C.  B.  (1983) Cytotox-
     icity of talc for macrophages ui vitro. Food Chem.  Toxicol.  21:  201-207.

Dixon, G.  M.;  Doria,  J.;  Freed,  J.  R.;  Wood, P.;  May, I.;  Chambers,  T.;  Desai,
     P.  (1985) Exposure  assessment for  asbestos-contaminated  vermiculite.
     Washington, DC:  U.  S.  Environmental Protection  Agency, Office  of  Pesti-
     cides  and Toxic Substances; EPA  report no.  EPA-560/5-85-013. Available
     from: NTIS, Springfield, VA; PB85-183080.

Dunnigan, J. ; Nadeau, D. ; Paradis, D. (1984) Cytotoxic effects of a'ram id fibers
     on  rat pulmonary  macrophages:   comparison  with chrysotile asbestos.
     Toxicol. Lett. 20:  277-282.

General  Electric.  (1978)  Fibrous glass  (silicate  base).  Schenectady,  NY:
     Materials Information Services; material safety data sheet no. 46.

Gruner,  J.  W.  (1934) The structures  of  vermiculite and their  collapse  by
     dehydration. Am. Mineral. 19: 557-575.
•i"iw IQPC                               ?-?4      DRAFT—DO NOT OUDTE OR CITE

-------
Hanson, R.  L. ;  Benson,  J.  M.;  Henderson,  T.  R. ;  Carpenter,  R.  L. ;  Pickrell, J.
     A.; Brown,  S.  C.  (1985) Method for determining  the  lung  burden of talc in
     rats  and  mice after  inhalation exposure to  talc aerosols.  J.  Appl.
     Toxicol. 5: 283-287.

Hill, J. W.  (1977) Health aspects of man-made mineral fibres.  A review. Ann.
     Occup. Hyg. 20: 161-173.

Holt, P. F.; Home, M. (1978) Dust from carbon fibre. Environ. Res. 17:  276-262.

Huggins, C.  W.;  Denny,  M.  V.; Shell, H. R. (1962) Properties of palygorskite,
     an asbestiform mineral. Washington,  DC:  U.  S. Department of the  Interior,
     Bureau of Mines; report of investigations no. 6071.

International  Agency  for  Research  on Cancer. (1983)  Biological effects of
     man-made mineral fibres:  proceedings of a WHO/IARC conference; April 1S£2;
     Copenhagen,  Denmark.  Geneva,  Switzerland:  World Health Organization,
     Regional Office for Europe; Euro reports and studies no.  81.

International  Agency  for  Research  on Cancer.  (1987) IARC  Monographs on the
     evaluation  of  the  carcinogenic risk of chemicals to  humans:  silica and
     other silicates, v. 42. Lyon, France: World Health Organization.

Jones,  H.  D. ;  Jones,  T. R. ; Lyle,  W.  H.  (1982) Carbon  fiber:  results  of a
     survey  of  process  workers and their environments in  a factory prod-cine
     continuous  filament. Ann.  Occup. Hyg. 26: 861-867.

Klinghol:, R. (1977) Technology and production of man-made  mineral fibres. Ann.
     Occup. Hyg. 20: 153-159.

Lee, K. P.;  Kelly, D.  P.; Kennedy,  G.  L.,  Jr.  (1983) Pulmonary response  to
     inhaled Kevlar aramid synthetic fibers  in  rats. Toxicol. Appl. Pharmacol.
     71: 242-253.

Leineweber,  J.  P.  (1980)  Dust chemistry and physics: mineral and vitreous
     fibres. In:  Wagner,  J.  C.; Davis, W., eds.  Biological effects of mineral
     fibres =  Effets  tiologiques  des fibres  minerales: proceedings of a  sympo-
     sium,  v.  2;  September  1979;  Lyon,  France.  Lyon, France: World Health
     Organization,  International Agency for  Research on Cancer;  pp.  8S1-900.
     (IARC scientific publication no. 30.; INSERM symposia  series: -v.  92).

Lockey, J. E. (1981) Nonasbestos fibrous minerals. Clin.  Chest Med. 2: 203-216.

Meisinger,  A.   C.  (1985)  Vermiculite.  In:  Mineral   facts  and  problems.
     Washington,  DC:  U. S.  Department of the  Interior, Bureau of  Mines;
     bulletin no. 675.

Morgan, A.;  Holmes,  A.  (1984) Solubility of  rockwool  fibres  uj vivo and the
     formation of pseudo-asbestos bodies.  Ann. Occup. Hyg.  28:  307-314.

National  Research Council. (1984) Asbestiform fibers: nonoccupational  health
     risks. Washington, DC: National Academy Press.
July 1955                               2-25      DRArT--DC  NC* QL::"E C'~ -•''-

-------
Neugebauer, R.;  Helbing,  G.;  Wolter, D.; Mohr, W.;  Gistinger,  G.  (1961) The
     body  reaction  to  carbon  fiber particles implanted in the medullary space
     of rabbits. Biomaterials 2: 182-184.

Nicholson, W.  0. (1986) Asbestos health effects update. Research Triangle Pa'k,
     NC: U. S.  Environmental  Protection Agency, Office of Health and Environ-
     mental Assessment,  Environmental  Criteria  and Assessment  Office;  EPA
     report no.  EPA/600/8-84/003F.  Available  from:  NTIS,  Springfield,  VA;
     PB86-242864/AS.

Owen, P. E.;  Glaister, J.  R.;  Ballantyne, B.;  Clary,  J.  J.  (1986)  Subchronic
     inhalation toxicology of carbon fibers.  JOM J. Occup. Med.  28:  373-376.

Preston, J.  (1979) Aramid fibers.  In:  Grayson,  M.; Eckroth, D.. eds.  Kirk-
     Othmer encyclopedia  of chemical  technology:  v.  3.  3rd ed. New  York,
     NY: John Wiley & Sons, Inc.; pp. 213-242.

PTS Promt.  (1988) Chemical production and consumption database.

Rom, W. N. ; Casey,  K.  R. ; Parry, W. T.; Mjaatvedt, C.  H.; Moatamed, F.  (1SE3)
     Health implications  of natural  fibrous zeolites for  the  intermountair,
     West.  Environ.  Res.  30: 1-8.

Spumy, K. R.  (1963)  Measurement and  analysis  of  chemically changed rr.ine-e"
     fibers after  experiments  j_n  vitro and  jr\  vivo.  EHP  Environ.  Health
     Perspect. 51:  343-355.

SRI  International.  (1963) Chemical  economics handbook.  Menlo  Park, CA:  SRI
     International.

SRI  International.  (1985) Chemical  economics handbook.  Menlo  Park, CA:  SRI
     International; section 543.5811A.

SRI International.  (1987) 1987  Directory of chemical producers.  Menlo Park,  CA:
     Stanford Research Institute.

Suzuki, Y. (1982)  Carcinogenic  and fibrogenic effects  of zeolites:  preliminary
     observations.  Environ. Res. 27: 433-445.

The  Carborundum Company,  (n.d.) Fiberfrax ceramic fiber:  product specifica-
     tions. Niagara Falls, NY:  Refractories and Electronics Division.

U. S.  Bureau  of Mines. (1986)  Minerals yearbook  1984.  Washington,  DC: U.  S.
     Department of the Interior.

Volk, H. F. (1979)  Carbon (carbon  and  artifical  graphite).  In: Grayson, M.;
     Eckroth,  D., eds.  Kirk-Othmer encyclopedia of chemical technology: v.  4.
     3rd ed.  New York, NY: John Wiley & Sons, Inc.; pp. 622-628.

Wagner, J.  C.; Berry,  G.;  Pooley, F.  D.  (1980)  Carcinogenesis and miners!
     fibres.  Br. Med.  Bull. 36:  53-56.

-------
Weill, H. ;  Hughes,  0.  M.;  Hammad, Y.  Y. ;  Glindmeyer,  H.  W. ,  III;  Sharon, G. ;
     Jones,  R.  N.   (1983)  Respiratory  health in workers exposed to  man-made
     vitreous fibers. Am.  Rev. Respir.  Dis. 128: 104-112.

Wolter,  D.  (1983)   Biocompatability of  carbon fibre  and carbon  fibre
     macroparticles. Aktuel. Probl. Orthop. 26: 28-36.

Wright,  W.  E.;  Rom, W. N.;  Moatamed,  F. (1983) Characterization of  zeolite
     fiber  sizes using  scanning electron microscopy. Arch.  Environ.  Health  36:
     99-103.

Zircar Products,  Inc.   (1978)  Zirconia bulk fibers type ZYBF2.  Florida,  NY:
     Zircar Products, Inc.; technical data bulletin no. ZPI-210.

Zoltai,  T.  (1986)   Review  of  summary health assessment document for  mine-si
     fibers other than  asbestos [letter to  Dr.  Nicolas Hajjar]. Minneapolis,
     MN:  University of Minnesota, Department of Geology and Geophysics; May 24.

Zoltai, T. ; Stout,  J. H. (1984) Mineralology: concepts and principles. Minneap-
     olis, MN:  Burgess Publishers.
July 198E                                2-27       DRAFT--DC  NCT  Qu:TE

-------
       3.  SOURCES OF EMISSIONS, AMBIENT LEVELS,  AND ENVIRONMENTAL  FATE
3.1  SOURCES OF EMISSIONS
     Mineral fibers are  released  into the atmosphere as a result of both human
and natural actions,  but no estimates  were  available.   Manmade  mineral  fibers
(MMMF) are  used  extensively in the production of a wide variety of  industrial
and consumer products (Corn et al., 1976).   Release of  fibers  into the environ-
ment can  occur  during the manufacture,  handling, use,   and disposal of fiber-
containing  products  (National Research Council,  1984).   Dust from MMM? and
their  products  is generated  in  sawing, grinding, and  cutting  processes  as
well as in the installation and removal  of these materials (Hill, 1S77).
     Trie  release  of   synthetic fibers  into  ambient air occurs  during  tne-'r
manufacture, handling, incineration, or disposal.   However,  no  estimates  were
available.
     Emissions  of  minerals are associated  with natural  weathering,  mining
operations, and  the   use  of these materials  in secondary products.  These
secondary products include  road-surfacing materials, pesticides, fertilizers,
and cosmetics.  Additionally,  release  of  mineral fibers present as  impurities
in  other  minerals used  commercially  may  occur  (National  Research  Council,
1984).
     Concern has  been focused on  the quantity of  airborne fibers released by
mineral fiber-containing  materials.  However,  the specific quantity of  fibers
generated from a given product depends upon the mechanical operations performed
(Hill, 1977).
     Estimates  of the amounts of vermiculUe released into  the environment
during mining,  processing,  transport,  and use have been  reported  by JR8  (as
cited  in  Oixon  et al., 1985)  (Table 3-1).   Based on  a production  volume of
                          Q
1  million tons  (1.2   x 10  kg) of vermiculUe ore mined and  beneficiated in
1979 to produce  314  x 10   kg of crude vermiculite, approximately 802 x  10  kc
were  released  into the  air,  89,900 x 10  kg were  released  into water,  and
2,490  x 10  kg  were  released as solid  waste.   The water releases were dis

July 1S5£                          3-1       DRAFT—DO NOT QUD7E On C3"E

-------
           TABLE 3-1.   RELEASES OF VERMICULITE INTO THE  ENVIRONMENT
Process
Production and beneficiation3
Exfoliation
Transportation0
Consumer use
Release (x 103 kg)
Air
802.0
3,029.0
64.7
282.5
Water
89,900.0
5,787.0
_.e
0.6
Solic
2,490.0
4,856.6
125.0
36,070.0
 Production volume of 1.2 x 109 kg.
 Exfoliation volume of 1.046 x 10€ kg, includes some from South Africa.
cTransportation volume of 360 x 106 kg and 206.5 x 106 kg of crude and
 exfoliated vermiculite, respectively.
 Includes exfoliated and unexfoliated vermiculite.
eNot reported.
Source:   Dixon et al. (1985).

of in settling  ponds,  and the water was recycled.   Releases into air resulted
from  fugitive  releases  from dust control  equipment,' whereas  particulates
collected in the  dust  control  system constituted the solid wastes  that were
disposed of in  landfills.   For estimates of releases during  exfoliation, JRB
(1982) assumed that  dust control  equipment  was  98 percent efficient.  Releases
during transport  are  negligible,  whereas the largest release is during  direct
application of vermiculite, as a component of agricultural  products, to  soil.
3.2  EXPOSURE
     Three routes  of  exposure  are commonly addressed in exposure assessments.
These are  inhalation,  ingestion, and dermal absorption.  The  available  data
indicate that  the  principal  route of exposure to  fibers  is  inhalation.  This
route of exposure  is  also of toxicologic concern.   The objective of defining,
both qualitatively  and  quantitatively,  levels  of exposure to airborne concen-
trations of mineral fibers has centered around the physical  characterization of
the fibers themselves.   Several  of these physical  properties appear to play a
role in the  assessment  of the potential  risk  associated  with  exposure.   One
such property is respirability,  i.e., the capability of the fibers tc enter the
small  airways  of  the  respiratory tract.  The general definition cf respirarle
    - 19SE
3-2
DRA'T--DC NCI QL':TE C" CITE

-------
 fibers  includes those fibers having a diameter of less than approximately 3  prr,.
 Additionally,  the  length and the aspect ratio (length/diameter) of the fibers
 determine  where they will  be deposited in  the  respiratory  tract  (National
 Research  Council,  1984).   The length of a  respirable  fiber is  less  than  200 ptr
 (Hill,  1977),  and  the aspect ratio must be >3 (Johnson et al.,  1969).   Different
 samples  of the same generic mineral  fiber  may vary  greatly in  physical proper-
 ties, thus  presenting different risks upon exposure (National Research Council,
 1984).

 3.2.1   Manmade Mineral Fibers
     Measurements  of  airborne concentrations of fibers have been conducted to
 provide  a  predictive  method for assessing  mineral  fiber exposure.   Following  a
.study of  three fibrous  glass manufacturing facilities,  Corn  and Sansone  (19/4)
 reported  that  fiber concentrations varied greatly with  the  sampling  location
 within  the plant  and that  the physical  characteristics  of  the fibers varied
 oreatly  even  among samples collected in the  same  area.   The  average fiber
 i'is~6ter,  content  of  fine fibers, and production rates  were  cited  as  being  the
 most  important factors  expected to result in increased fiber concentration in
 the  working environment.   In contrast, the  effective use  of dust-suppressing
 agents  is  expected to cause the opposite effect (Ohberg, 1987).  Ohberg  (1S£7)
 also  reported that furnace operators or transport workers  were potentially
 exposed  to much  lower  fiber concentrations than production workers.   Esmen
 et al.  (1978)  found a difference of  fivefold  or greater in  the average  fiber
 concentration  among plants performing similar mineral wool  processing opera-
 tions.   Considerable  variations  within  and  among plants were  also found in
 ceramic  fiber  production  facilities (Esmen et al., 1979a).
     Attempts  to  correlate airborne fiber  concentrations,  expressed as fibers/
 cm  ,  with the weight of total suspended paniculate matter (TSPK)  expressed as
 mg/cm   in individual  air samples, have  been unsuccessful  (Corn et al.,  1576;
 Corn  and Sansone,  1974).   However,  an excellent correlation  was  found  when
 average  TSPM concentrations for similar work  activities were  compared to the
 average  fiber  concentrations  for  these same  activities (Esmen  et al.,  1979a).
     Fowler et al.  (1971) concluded that the  mean  fiber diameter  of  fibrous
 glass was  not a reliable  predictor  of  the mean diameter of the fibers in the
 aerosol  generated  by  that material.  In later studies,  however, Esmen et al.
 (1978,  1979b)  determined that the size distribution of the airborne fibers  was

 July 19S5                          3-3        DRA-T--DO NOT QUOTE OR CITE

-------
consistent.  They  concluded  that  approximately 50 percent or less  of  airborne
fibers released  by the production of mineral fibers with nominal diameters  of
6 urn or  more  would be <3 urn in diameter.  Approximately  50  to 90  percent of
airborne fibers  released  by  the production of fibers with nominal  diameters
<3 ym would be respirable (Esmen et al., 1979b).
     Selected occupational exposure data for MMMF, as indicated by  the airborne
concentration  measured in various  manufacturing plants,  are  presented in
Table 3-2.   Esmen  et  al.  (1979b) determined  airborne  fiber concentrations
in 11 glass fiber  and 5 mineral  wool  factories  as  part of an environmental
epidemiological  investigation of  worker exposure (Enterline  et al., 1986; see
Chapter  8).  The average  concentrations within the  breathing zones were 0.0024
to 0.78  fiber/cm  for glass fibers and 0.024 to 0.34 fiber/cm  for  mineral wool.
The highest fiber  concentrations  were detected in production, manufacturing,
and maintenance  operations.   In general, the average concentrations of fibers
in the respirable  range  encountered in  fiber  production  varied between about
0.03 and 0.2  fibers/cm  (Hill,  1977).   Variations  in exposure depend on  the
habits of  the  workers, the processes  involved, and  the  availability of  control
measures,  as  well  as  on  the physical  characteristics  of  the  fibers  (Esmen
et al.,   1979a).
     As   part  of several epidemiologic  investigations (Saracci  et  al.,  1954;
Simonato et al., 1986a,b;  see Chapter 8),  a  study was conducted by  Cherrie  and
Dodgson   (1986)  to  determine worker exposure  in  13  European MMMF  factories
(4 manufacturing glass, 7  mineral wool, and  2  continuous  filaments).   Exposure
was divided into three "technological phases."  The early technological phase
represented the  period of highest  airborne levels of  MMMF, where  no dust-
suppressing agents were used.  The intermediate technological phase represented
MMKF production  with and  without  dust suppressors.   During the late.phase,  oil
and resin  binders  (dust  suppressors)  were  used during production.   The  results
of an environmental  survey conducted between 1977 and 1980 indicated  that the
average  combined occupational group concentrations  in the rock and glass  wool
plants were generally low (<0.1 fibers/cm  ), whereas airborne  fiber concentra-
tions in the glass continuous-filament  factories  were much lower (<0.01 fibers/
cm ).   The average plant  median  for fiber length ranged from 10 to 20 urn, and
the corresponding  median  diameters  ranged from 0.7  to  2  urn.   In general, the
glass wool  fibers  were  thinner than the  rock wool  fibers  (Cherrie et al.,
1985).
Julv 19£5                          3-4       DRAFT—DO NOT QUOTE Oh CIT:

-------
                                  tAtiir



nl 1 ilipr
finliK nl
MIS Ifl.ISS
MIS <|l .ISS
MIS l| I.ISS
MIS I|I.1SS
' MIS l| 1 .ISS
' MIS IJ I.ISS
' MIS <| I.ISS
1- MIS l| I.ISS
IIMIS l| I.ISS
IIMIS i|lass
IIMIS l|laSS
i-oiis i|lass
IIMIS (| I.ISS
mills ii l.iss
I*.S Wllltl
!•. s wnnl
ISS Wlllll
.IS S WIMt I
.»ss wool
imral wool
iK-r.il Html
ni-ral wool
iii-r.il wnnl
ii d/slai| wool
II Ir/S la<| WIIOl
II d WIMI I
II V WIMI 1
M It WIMll
i-r.imii
ill IIIIM
.iilMin
.ilium
ILtpii li| HP
1 1 .I|IM li| t l.p
.ill
.ill
.tli
fi-i mil til ill*
MIIU.IMT nl s.impli'S
|T
ll.il .1 mil prP'.pnli*
ronci>nlrali«ins, no. <
1'iross
No of
pistils RaiH/l* nf Rain|p ol
sam|>lpil avrraqps cnncpnlr.il innr.
40* 0.41 <0 07-7.11
I -- . -
3 HAC Ml -4. SO
n -- --
11 0. 0074-0. /fl
j
\ 0.04-1?.;'
S NA <0 01-71.09
1 -- 10°
S -- 0.001- .1.0
4 0.3* 0 1-14 5
1 S0.?
I.O
1 l?-36
444-I33J
6
1 0 ?• --„
3 1.8-7B ^.B-13.1
3 0.049-?.fi O.OOI?-5f,.W
1
--. -- 74-l?OOh ,
i 11
17 -- mm-n.yi * to"
7 -- 0.09- 2?. 9fi*
i
3 -- 0.70-2 M
I j-
5-73
I inl
• I,
•I.
'Mil   Ni.l  ili-li-i Inl

'mil/m1

Mi 11 inns  (Mil it li--./ll 1

'l.i-i,m.-li IT  mi-.lii ilni ini| \'MI  l.o  I'lMII.
                                                                                   0.0117

                                                                                   NA


                                                                                   0.011
                                                                                   •O.I
                                                                                   O.OI-O. 12
                                                                                   II. I'l
                                                                                   II .7" !-?.(•
                                                                                                  O  4H
                                                                                                  »0 117-1.03
                                                                                                   O.OIU-0.073
                                                                                                   »0 7
                                                                                                   0.01-0  1
                                                                                                   0.001-0. 1
                                                                                                  0.01-0.34

                                                                                                  II .1)111-0.463
                                                                                                  lt.Oi-0. I

                                                                                                  0.0.3-0. 16*
                                                                                                   0 OS- 7. /I
                                                                                                    i.n
        (19/f.)
Bprlalfi  el al.  (19Wi)
Corn and  Sansone (19/4)
Inlerline and Marsh (I9B2)
tsuen  el  al.  (!9/9b)
Gardner el al.  (I9IW.)
Mill el al.  (19/3)
.Inhnson el al.  (I9f>9)
Maqqioni  el al.  (1900)
Weil I  el  al.  (I9B3)
Bay I Iss el al.  ( 19/f.)
Shannon el al.  (19B2. 19B4.
 19B6.  19B7)
tnlerttne el al. (!9Hft)
Heriattl  el al.  (I9H4)
Moulin el al.  (I90(>)
Cherrle el al.  (l9Bfi)
Moulin el al.  (I9iir.)
Shannon el al.  (1907)
Wriqht  (19GB)
Saraccl el al.  (1904)
Alsbirk el at.  (1903)
ls«en  el  at.  (I979b)
Olsen  and Jensen (I9B4)
Well! el  al.  (19B3)
Wei 11  el  al.  (19113)
Carpenler and Spolyar (1945)
Oodqson el al.  (1907)
Halmberq  el at.  (I9B4)
Skuric  and SlMhuljak-Rerilic (19114)
Saraccl el al.  (I'lHI)
ts«»n  el  al.  (19/9b)
,lon*i  el  at.  (1907)
Henry  el  at.  (1907)
Gieseke el at.  (1904)
(.aoble el al.  (I90S)
Va.wHIer el al. (I9BS)
Gamble el al.  (1907)
W><|man el al.  (1907)
Kleinfeld el al. ( I'lf./)
lockey ol al.  (19114)

-------
     Dodgson  et  al.  (1987) developed  a mathematical  model  to estimate past
exposure  during  the early  technological  phase  to respirable glass wool and
rock wool  fibers in  the European wool insulation industry.  The  model  WES
derived from  data  on airborne fiber concentrations measured  during  1977-60 a-.s
1984,  from  dust  emission measurements, and from the  technical history of ea:>-
plant.   It  was  found that the mean  respirable  fiber  concentrations for rock
wool plants ranged from 1 to 2 fibers/cm  during the  early technological  phases
with levels  of  about  10 fibers/cm3 for  those  areas  with the highest dust
                                                                             •5
concentrations.   The current  exposure  levels  are  about  0.03  to 0.12 fibers'cr--
(see Table 3-2).  Higher  concentrations were  reported by  Cherrie et al.  (1557)
in an  experimental  simulation of a  rock  wool/slag wool production  process tc
estimate exposure during the early phases.  It was found that the time-weig-.ted
average concentration during the period of the experimental  production wcrk *a =
approximately 1 to 5 fibers/cm  when oil was added and approximately 5  fibers/
err,  in  the  absence  of oil.  The  corresponding  model  estimates  for the glass
wcol plants  during the  early technological  phase were  similar to curre-.t .
reipira^le  fiber concentrations  of about  0.1  fibers/err;   or less  (Ecdcsc"
et al., 1987).
     Esmen et al. (197Sa)  have suggested  that,  under  similar plant  CO"- fens,
employee exposures to  fibers  during ceramic fiber production will   be  approxi-
mately four times the exposures during mineral wool production.
     Exposure of workers to  mineral fibers occurs in  locations  other than
fiber production facilities.  Specifically, workers engaged in the  fabrication,
application,  finishing,  and other operations  involved  in the manufacture  and
installation  of  mineral  fiber-containing  products are also  exposed  to airborne
fibers  (Fowler et al.,  1971).   Corn et al. (1976) reported  an  average fiber
concentration of 0.42  (range,  0.11 to  0.95) fibers/cm  in a  production area  in
which ceramic fibers were used to form ceramic blocks.  Fibers <3 urn in diarete-
constituted an average of 75.2 ± 7.7 percent of the total  fiber concentration.
     No simple method  currently exists for estimating the amount of respiratle
fibers  likely to be released when mineral  fiber-containing  products  are  used
in practical  situations (Schneider,  1979).   Measurements  taken  during the
removal and  replacement  of MMMF insulation in  a  loft showed an 8-hour time-
weighted average  (TWA)  for respirable  fibers of 9 and 0.5 fibers/cir  . respec-
tively.  The  high  concentration during removal was attributed  to  the age  and.
type of  insulation involved.   Marconi et al.  (1987) performed personal  ar.c
                                   3-6       DRAFT--DO

-------
 area measurements of airborne fiber  concentrations during the installation of
 rock wool insulation blankets on  walls  and-ceilings of deckhouses in ships of
 a naval dockyard.  Mean  respirable  fiber  concentrations  ranged  from 0.02 to
 0.19 fibers/cm3 with  a  peak level  of  0.65  fibers/cm3.
      The National  Institute  for Occupational Safety and Health  (NIOSH)  he;
.conducted two  workplace  exposure  surveys.   The National  Occupational Hazard
 Survey (NOHS), conducted from 1972  to  1974,  estimated  the number of workers
 potentially  exposed to chemical agents  in the workplace  in 1970  (NOHS, 1976).
 These estimates were derived from observations of the actual  use of the ager.t,
 the use  of  trade  name  products  known to  contain  the agent,  and the use  of
 generic products  suspected of containing the agent.  The NOHS  estimates of
 exposure to  the manmade mineral  fibers  are presented in Table 3-3.

           TABLE  3-3.   NOHS ESTIMATES OF EXPOSURE  TO MINERAL FIBERS
Chemical
t _C - : .
Glass fioers
Glass wool
Fibercless
insulation
Mineral wool
Polyolefin fibers
Attapulgite
Talc
Vermicul ite
No. of
people exposed
62,611
57
80,629

5,079
96,784
134,121
1,536,754
104,456
Source of exposure estirristej__percer.t
Actual
..a
100
74

-•
--
--
14
4
Trade name
61
--
—

100
4
29
31
21
Gene Me
3s
--
26

--
95
71
56
74
 8No  data  available.
 Source:   National  Occupational  Hazard Survey  (1980).

      The  National  Occupational Exposure  Survey  (NOES),  conducted  from  1980  to
 1983, estimated the number  of  workers potentially exposed to chemical agents
 in the workplace  in 1980  (NOES,  1984).   Unlike NOHS, the NOES estimates were
 based only on observations  by  the surveyor of the actual  use  of  the agent.
 Fibrous glass,  mineral  wool, and  ceramic fibers  were  not found  in  the NOES.
      The  American Conference of Governmental Industrial Hygienists (ACGIK) has
 recommended an 8-hour  TWA  threshold  limit value (TLV)  of 10  mg/m   for  mine -a1
 wool  fibers (American  Conference  of  Governmental  Industrial Hygienists, 19£r).
 In 1977,  National  Institute  for  Occupational  Safety  and Health  (1977)
 recommended that  the  standard for  occupational  exposure  to  fifcrojs gU-s
 July 19 = 5                           3-7       DRAFT — CD NO' QJCTE C

-------
should  not  exceed 3 fibers/cm  for fibers with diameters <3.5  um  and  lengths
>10 urn,  determined  as  a TWA concentration for up to a 10-hour work shift in a
40-hour workweek.  However, airborne concentrations determined as total fitrous
glass should be  limited to a TWA concentration of 5 mg/m  of air (FR, 1988).
     The  available  data suggest that consumer exposure to  fibers  is low.   In
many  instances,  the manufacture  of secondary products from  mineral fibers
includes  the addition of a binder to form a matrix, thereby resulting in little
subsequent  fiber exposure.   Many of the uses of MMMF are primarily industrial
rather  than residential.   Fibrous glass and mineral wool are used mainly  for
thermal  insulation.  The  highest exposures would probably occur shortly after
the installation or disturbance of  insulation materials.  Fibrous glass is else
used in filtration applications. .The escape of fibers during use would cor,r-c-
•ir.ise the  effectiveness  and use of the filters.  The specialized applicatio-s c*
ceramic fibers limit their potential for consumer exposure.

3.2.2  Synthetic Fibers
     Measurements of airborne  concentrations of polyolefin and  aratr.'d fise-s
in  manufacturing plants were  net  available.   In  a study by  Chang (19£2),
aerodynamic size analysis and scanning electron microscopy of aerosols ge^e-atec
during  the  manufacture  of  carbon fibers showed the  presence  of a  significa-.t
quantity  of respirafcle  fiber-like particulates.   The mass  median  aerodynamic
diameter  (MMAD)  was  4.0 urn compared to the  geometric  diameter (7  to 8 urn) of
unprocessed  fibers.   Similar to  MMMF,  fiber  concentrations  vary   with  tne
sampling  location and manufacturing process  involved.
     Gieseke et  al.  (1984)  studied carbon fiber emissions during the manufac-
ture  and disposal of carbon  fiber  products.  The average  fiber lengths  enc
diameters  from 12 manufacturing  operations  were sampled; these ranged  fror.
25  to  2000 urn and  from 5.5 to  7.8 urn,  respectively.   Fiber concentrations
ranged  from a low of 900  fibers/m   in  the  handling operation  to  a high cf
6.59 x 10   fibers/m  In the drilling operation.  Of the manufacturing operations
                                                                    _2
studied,  drilling  and  weaving had  the highest  releases  (1.44 x 10   percer.t
and 7.2  x 10  ng/cm ,  respectively).  The authors  concluded  that  release  anc
distribution were highly dependent  upon  the particular operation.  Selected
occupational exposure data for carbon fibers determined in manufacturing r'e-.ts
are presented  in Table  3-2.
JVv 1S5S                          3-5       DRAFT--D: N"T Qj"E C':.

-------
     NOHS  exposure  estimates for polyolefin fibers are  listed  in  Tatle  3-3.
No exposure  estimates  for aramid or carbon fibers were found in the  available
literature.  Similarly, NOES estimates for synthetic fibers were not  found.

3.2.3  Minerals
     Measurements of airborne  concentrations  of attapulgite, talc, and verrr.i-
culite in  mining and milling activities conducted in environmental surveys  as
part of  epidemiological  studies are listed in Table 3-2.  No data  were avail-
able for erionite.   Vermiculite exposure at a mining and milling operation  in
Libby, MT, was  studied by Amandus et al.  (1987)..   The site was divided into
25 location-operations, and  sampling information prior to 1950 through 19£2  was
collected  and analyzed.   Prior to 1964, the greatest  exposure  came  fror cry
mill jobs, and estimates for working areas, sweepers, skipping,  and the qusVit,.
control  laboratory  were  168, 182, 88, and 13 f/cm  , respectively.   During the
years 1964-77,  exposure  estimates decreased greatly.   For  the same operations,
they were  33,  36,  17, and  3 f/cm , respectively.  Prior  to 1971, expcsu-e
estimates  for mining  jobs including  drilling  and nondrilling operations  were 9
to 23,  and  <2 f/cm  ,  respectively.   At the- river  loading  station,
estimates prior to 1971 for ore loading, the river office, the conveyor tur.ne"! ,
the river  dock,  and the river station bin area ranged from 5 to 82, 10 tc 11.
112 to 113, 5 to 117, and 20 to 21 f/cm , respectively.   Exposure estimates for
all operations  decreased  annually from 1972 to 1976  and  leveled  off  to  1952.
Exposure in most  areas  during 1977-82 were <1.0  f/cm   and ranged from 0.6  to
1.0 f/cm   in  the  mill.   Estimates for an 8-hour TWA fiber exposure in 19E1-S2
ranged from 0.6 to 0.8 f/cm  for mine and mill jobs.  However, based on samples
taken by the  company in 1984, the average TWA exposure was 0.1 f/cm .   Due tc
the presence  of  asbestos  as a contaminant In vermiculite, Dixon et al.  (19=5)
conducted  a study  to determine airborne  concentrations  of asbestos  in process-
ing plants.   Sampling was conducted  in 1980 at  the  W. R. Grace mine  and milling
facility near  Libby,  MT,  from October 21 through October  26,  and at  both  the
Grace and  Patterson mine  and processing facilities  near Enoree,  SC, from
November 3  through November 6.   Both air  samples and  bulk samples were
collected  at  each  location.   The  sampling method  employed attempted to capture
particles  during  various  stages  of production.   The possibility existe:  that,
during exfoliation, asbestos fibers trapped between vermiculite plates co-id be
released;  therefore,  samples  were analyzed  as  they were  received  a".i s'te1-
July 15 = :                          2-8       DSA"--D: K" Ql'C": C: II":

-------
laboratory  exfoliation,  a process  that  allowed the  samples  to retain the
asbestos, to  determine if additional fibers were shed.   Laboratory  analysis
suggested that  a  higher quantity of asbestiform  fibers existed  in smaller size
grades of vermiculite  as compared with larger grades.  Multiple grades of bulk
samples, mined  from the  Grace facility  at Libby and observed  under  trans-
mission electron  microscopy  (TEM),  consisted of the amphibole  group minerals
anthophyllite  and  actinolite-tremolite  and contained  a  range of  1 to
1800 x 10  fibers/kg.   Corresponding concentration  ranges were from  1 to
41,000 ppm (Dixon et al., 1985).
     Table 3-4  summarizes  data obtained  from a  study that monitored  amfie-t
concentrations  of asbestos-containing vermiculite  at two different  processing
facilities.    Personnel  sampling was conducted at various  sites  of  activity,
and the  values  are  expressed  as  fibers  per cubic centimeter.  Dixcn  et e".
(1985) also reported estimates of occupational  exposure  to vermiculite.  Tne
data were compiled  from various sources  and are  listed in Table 3-5.   However,
these numbers may overestimate occupations! exposure because  they  may include
clerical  personnel  or  because  nonvermiculite products may also  be manu*act-"-e_
at the plant.
     The NOHS estimates  of exposure to attapulgite, talc, and vermiculite  ere
presented in  Table  3-3;  however,  information on erionite was  not founc! in  the
NOHS.   NOES  estimated that  18,872  workers, including  5,244  females, were
potentially exposed to  talc and  4,293 workers,  including 365  females, we-e
potentially exposed  to vermiculite  in the workplace in 1980.   Information  for
attapulgite and erionite was not found in the NOES.   The ACGIH  has recommenced
an 8-hour threshold limit value (TLV-TWA) of 2 mg/m  respirable dust for  talc
containing no asbestos fibers.   No workplace exposure limits  have  been recor-
mended for attapulgite or erionite (ACGIH, 1985).
     Exposure to  attapulgite,  erionite,  talc,  and vermiculite  may arise frorr,
the use  of  these  materials in consumer products such as  pet waste  absorbents,
fertilizers,  pesticides,  drugs, and cosmetics (Dixon et al.,  1985;  National
Research Council,  1984).   However, reports of consumer  exposure  levels  were
found only for  vermiculite.
     Due to the various uses of  vermiculite, consumer exposure is relatively
high  (Table 3-5).   Table 3-5 also lists  the ranges of asbestos concentrations
to which consumers may be potentially exposed as a result of vermiculite use.

-------
         TABLE 3-4.  OCCUPATIONAL EXPOSURE TO AIRBORNE CONCENTRATIONS
                       OF ASBESTOS IN PROCESSING PLANTS
           Population
Asbestos fiber
 concentration,
  fibers/cm3
Comments
Miners and millers of vermiculite
0 Grace mine and mill at Libby






0 Grace mine at Enoree

0 Grace mill at Enoree

Exfoliators of vermiculite
0 G"s:e facility at Enoree

0 Patterson facility
(benef iciation, exfoliation
at Enoree)

0.03a
<0.01
1.7
0.8
6.4
0.11
0.15
0.15
<0.01
0.03
0.14

0.08
o.oe
0.02
0.16
0.05

Front loader, mine
Pit haul driver, mine
Mine analyst, mine
Bottom operator, mill
No. 2 Operator, mill dozer-
Operator mine
Shuttle truck between plants
Truck driver
Dragline operator
Mill monitor
Mill lab technician

Bagger (Grade 4)
Bagger (Grade 3)
Payload operator
Plant foreman
Bagger/fork! ift
 Concentrations are an average of values determined by two independent
 laboratories, Ontario Research Foundation and IIT Research Institute.

Source:   Dixon et at.  (19S5).
                                   3-11
            ncz.-7-.rv-.

-------
            TABLE 3-5.   SUMMARY OF OCCUPATIONAL AND NONOCCUPATIONAL
                INHALATION EXPOSURE TO ASBESTOS IN VERM1CULITE
Population
Occupational
Miners and millers
Exfol iators
Users of exfoliated
vermiculite
Aggregates and
insulation producers
Transportation
Truck
Rail
Warehouse
Consumers
Attic insulators
Garden fertilizers
Lawn .
Disposal
Food
Drinking water
Ambient air
Near mi 11s
Near exfoliation plants

Ambient water
Ambient land
Number
of persons
exposed
•^250
1,694-1,979

1,694-1,979

298

129
108
Unknown

18£,000
32,000,000
74,400,000




4,600
13,147,496



Exposure level
fibers/cm0 pg/m-51
NDa-9.7
ND-0.38

ND-0.38
K
-- 6,800

<0.01-0.3
Unknown
0

6,800
20
4.4
Unknown
Unknown
Unknown

ND-0.5
S.OxlC'-5
2.5x10-*
Unknown
Unknown
Duration
(hours)
42.0
41.5

41.5

3t 9

3- c
7C C
39. 9

c
1
^




16 =
15£



a
 ND = None detected.
 No data available.

Source:   Dixon et al. (1985).

-------
3.3  AMBIENT LEVELS
     The  presence  of mineral  fibers  appears to  be  ubiquitous;  therefore,
exposure  of  the  general  population  is  expected to some extent.  The sources c*
most fibers  found  in the ambient environment are not currently identifiable.
The  nominal  diameters of  most MMMF exceed  the  respirable range (National
Research  Council,  1984).   The larger fibers tend to  settle out of  ambient  sir
(Balzer,  1976).
     Manmade mineral  fiber concentrations in ambient air  have been  reported
for  a  limited number  of indoor and  outdoor environments.  Balzer  (1976)
collected 300  ambient  air samples from sites around  the State  of California tc
determine  glass  fiber  concentrations  in the  environment.   In 36  samples
selected  for detailed  examination,  the average concentration of glass fibers
was 2.57 i 2.98 fibers/L.  The average diameter of the glass fibers was  4.3 pr,
with a range of 0.1 to 17.7 pm.
     A majority  of the U.S.  population is probably  exposed to mineral fibers
tc some  extent by  living or  working in buildings  with fibrous  glass  insulgfic"
(Na-.iv.-al  Research Council,   1964).   The  airflow  through  fibrous-gless-1 irec
duct systems  in public  buildings  at  six  locations  was analyzed for fiber
concentration  to evaluate  the possible erosion of glass fibers frorr,  the liners
(Cho.lak and  Schafer,  1971).   In each  case,  no glass fibers were found on  the
filter of the particulate  air sampler.  Balzer  (1976) examined 37  samples
taken from 13  fibrous-glass-lined ventilation systems in  the  San Francisco Bay-
area.  Fiber  concentrations  ranged  from  none seen  to 2.0 fibers/L, with  ar.
average concentration  of 0.87 ± 0.48  fibers/L.    The average  diameter  of  tne
glass fibers was 3.7 pm, with a range  of 0.1 to 17.7 pm.
     Schneider  (1936)  determined the  airborne  concentration  of  MMMF in two
groups of buildings.   Group A consisted of  11 schools with mechanical ventila-
tion.    Group  B consisted  of kindergartens, offices,  and one school.  MMMF
levels in group  A  ranged from 0 to  87  fibers/m   for  nine  schools.  Two schocls
were excluded  fron  the  results because of  high  MMMF concentrations due  tc
uncoated  ventilation  ducts  (120 fibers/m ) and  previous  handling of MMMF
(950 fibers/m  ).   Nonrespirable fibers were found in only one school.   MK'-'.F'
levels in group  B  ranged from  380  to  10,000 fibers/m .   Nonrespirable fibers
were found in  all  buildings.
J_.1v< 19:£                          3-13      DRAFT—DD N:T Q':;-£ D: :'."-

-------
     In Australia, 193  air  samples  collected in 22  asbestos-containing office
and plant buildings were analyzed for their mineral  (nonasbestos)  fiber concen-
trations.   The  samples  were  collected  during working hours  under  dynamic
ambient airflow conditions.   Fibers  <3  urn  in diameter,  >5 ym in  length, anc
having  an  aspect  ratio >3:1  were counted.   The fiber concentrations ranged
from 0  to  13 fibers/L,  with 75 percent  having 0 to  1  fibe.r/L  (Altree-Williams
and Preston, 1985).
     In a-pilot study (Van  der Wai  et al., 1987), the concentration  levels  of
MMMF in eight homes  in the Netherlands were investigated during a  thermal
insulation procedure  in which glass  or  rock fibers  were  blown into the cavity
wall.    The average concentration  in  the homes during  the insulation  procedure
varied  between 5  and  25 light microscopically detectable fibers/L, with  pecks
up  to  65  fibers/L.   The  next day the  concentration  had decreased  to 1 tc
10 fibers/L, which was  equal  to or  less than the preinsulation concentrate cr.
found in the homes.
     Ambient  levels  for synthetic fibers  were not found  in the available
literature.
     Mining  and  natural weathering  of  minerals cou.ld lead  to  significa-t
localized  airborne  fiber concentrations.   However, no   reports  of  airborne
concentrations of  the naturally occurring  mineral  fibers were found  (Nations!
Research Council,   1984).  Ban's et  al.  (1981) determined the  airborne concen-
tration of gross  and  respirable eriom'te fibers in two  villages  (Karain and
Karlik) in  Central Turkey.    The concentrations  of  respirable dust  from the
street  were  0.2 to 2.8  mg/m  in one village and 0.3 to 1.5 mg/m  in  the  other
village.   Total dust  concentrations  from street, work,   and recreational and
indoor  samples were <0.01 fibers/cm   in all areas in one  village, and <0.01,
<0.01 to 0.3, and <0.01 to 1.38 fibers/cm , respectively, in  the other village.
Ambient air  concentrations  of  asbestos  near vermiculite  manufacturing  anc
processing plants  have been reported (Dixon et al.,  1985; see  Table  3-5).
     Exposure to  mineral  fibers could also occur from drinking  water.  Sources
of  fiber  concentrations in drinking water could include  natural processes,
such as leaching  or  rainfall runoff, and  industrial discharges (Nations!
Research Council,  1984).  Attapulgite has  been found in some water  supplies
(Millette et al.,  1979b, as  cited in National Research Council,  1954).   Arcie-t
air concentrations of asbestos  near vermiculite manufacturing  and
p!ar,ts  have been reported by Dixon et al. (1985; see Table 3-6).
     19:£                          3-K      DRA-T—DC NOT QJOTE

-------
                    TABLE 3-6.  ASBESTIFORK FIBER MEASUREMENT METHODS'
Measurement
       Col lection
  Quantification
   Identification
Mass, gross     Filter
                Electrostatic


                Impaction

                Hi-vol/filter
Mass,           Horizontal elutriator/
 respiratle     filter
                Cvclone/filter
Count
Impingement


Impaction


Thermal precipitator


Membrane filter




Nucleopore filter



Nucleopore filter
                           Gravimetric
                           Gravimetric
                           precipitator

                           Beta-absorption

                           Microscopic


                           Gravimetric
Gravimetric




Light microscope


Light microscope


Light microscope
                                           Light microscope/
                                           phase contrast
                                           TEM,b SEM,C
                                           image recognition


                                           Light scattering
Mineral identification
by x-ray; chrysotile
(asbestos) identifi-
cation by Mg analysis

Not applicable
(piezoelectric)

Not applicable

Mineral identification
by x-ray

Mineral identifies*.:^
by x-ray; chrysov'le
identification by Kg
analysis

Mineral ident'ificetic*.
by x-ray; chryscv 'e
identification ty M;
analysis

Identification by
morphology

Identification by
morphology

Identification by
morphology

Identification by
morphology; mineral
identification by-
dispersion staining

Mineral identification
by SAED;  chemical
composition by EDXA

Identification of fibers
by magnetic alignre-t
aAdapted from Burdett et al. (1980, as cited in National Research Council, 19£-).

 Transmission electron microscope.

cScanning electron microscope.

 Selectee" area electron diffraction.

 Energy-c^spersive x-ray analysis.

       J-lv !5££                          3-15      DK-rf--:>: N:* Q-"": -'• ::~-

-------
3.4  ENVIRONMENTAL FATE
     The fate of  mineral  fibers in the environment has  been studied  mainly  in
comparison to  that of asbestos  fibers  (Spurny,  1983;  Spumy et al., 19£3).
These studies have  focused mainly on glass fibers  as representative of MMMF.
The  results  of  these studies have shown  that  MMMF can  undergo chemical and
physical changes,  including  changes  in fiber size, chemistry,  and  crystalline
structure under different environmental conditions (Spurny,  1983).
     Manmade mineral  fibers  are mechanically stable and insoluble in aqueous
solutions.   They  are  not  resistant to  hydroxides;  glass fibers dissolved in 2f«
NaOH by  the  end of a 1-year period (Spurny et al., 1983).   Resistance  tc  acic
solutions is dependent  upon  fiber type.   Glass fibers  showed  little evioen:e
of  surface  changes following  exposure  to acids,  but MMMF  other the--  glass
fibers were  not  acid-resistant.   Analysis  indicated that the fiber surface  w=s
rm'croporous  following acid exposure  (Spurny  et  al., 1963).  In addition tc
surface  corrosion,  acid  exposure leached  some  elements  (Na,  K, Ca, Al,  anc  Zr.)
frorc the fibers  (Spurny,  19£3).   Exposure to  stronger  acids ca~  cause g'ess
and rock wool fibers  to dissolve starting from the fiber  ends,  resulting ir,
the development of  inverted  cones  (Wojnarovits-Hrapka,  1978, 1977, as  cite:  ir
Notional Research  Council,  1984).   These  data suggest that the exte-nal fic-er
structure is more acid-resistant than the  internal structure.
     The chemical  and physical  changes  that  affect mineral fibers  are not
homogeneous.   Considerable variation  exists  in composition  along the length  of
a  fiber  as well  as between fibers.   Thus, the effects of  environmental condi-
tions, such  as  acid leaching of elements,  vary with location along a fiber  and
affect individual  fibers  differently.   Fine  glass fibers  (diameter <3  uir.) are
degraded much more  rapidly than coarse fibers (Spurny  et al., 1983).   Unlike
asbestos fibers, which can split longitudinally to produce thinner  fibers,  MMMF
break transversely, creating shorter fibers (Weil! et al.,  1983).   However,  for
brittle  asbestos,  especially noticeable in low fiber quality,  the coherence
between  the fibers may be extensive so that when their separation is  attempted,
some fibers will break transversely before they separate longitudinally.
     The rate at  which fibers settle out from the air is  determined mainly by
their diameters,  with larger  fibers  settling out rapidly  (Hill,  1977).  As
with asbestos fibers,  the major environmental  fate of mineral  fibers is Drcta-
bly the  deposition  and eventual burial of the fibers  in  soils and sediments
(National Research Council, 1984).

-------
     Ultraviolet  light  and gamma  radiation  have been found  to  affect the
ultimate tensile strength and chemical composition of  Kevlar   49  fibers  (Morgan
et a!., 1983; Harper  and McAlister,  1975; Toy, 1987).   Aging by gamma irradia-
tion was enhanced  in  the presence of oxygen (compared to  uj  vacuo conditions)
and  was  more severe  than ultraviolet light.   Ultraviolet light  (120 watts,
2537A)  lowered  the strength  by >25  percent  in  1 week,  whereas  soft white
fluorescent  lighting  had a  lesser,  but significant, effect  (Harper and
McAlister,  1975).  Toy  (1987) subjected Kevlar®  fibers to photodegradation in
   18
an   CL atmosphere, using a concentrated sulfuric solvent, at temperatures of
25 and 196°C.   Rapid  decarboxylation occurred *t 25°C at  the terminal groups
of the fiber  macromolecules.   At 196°C, the  fiber macromolecules broke  dew.
completely.   Kay and Ross (1972) reported that several physical characteristics
of Kevlar5"  fibers such as fabric elogation,  breaking  strength,  and tear strencf
were reduced  by  29,  179, and 20 percent, respectively, following accelerated
weathering.    The  fibers were exposed  for periods  of  up  to 150  hours  in  ar,
accelerated weathering unit before property  testing began.
     The rates of  hydrolytic  degradation of  Kevlar   4S fibers depend on the
relative humidity, time, temperature, and stress  level.  However, Morgan et a*.
(19£2) reported  that  hydrolytic degradation is not a  significant aging mecha-
nism.  Auerbach  (1985)  also studied  the effects  of temperature and humidity cf
Kevlar  29 yarns.   Initial  degradation rates  were slow due to  an inhibitor but
increased rapidly as the  inhibitor was depleted.   Kevlar   29  tolerated huridity
levels in the range of 90 percent or  less.
     Vermiculite and  talc are not expected  to undergo chemical transformation
when  released into the  environment.   Their refractory nature precludes  the
effects of  melting/boiling point, solubility, vapor  pressure,  octanol/water
partition coefficient,  etc.,  on their transport.  This is also  true for the
MMMF and mineral fibers.  Terrestrial and fluvial transport processes affecting
these minerals are not well characterized.
3.5  ANALYTICAL METHODS
     The principal  measurement  techniques  that have been used to quantify a-c
identify  asbestiforir  fibers,  including asbestos,  MMMF,  and synthetic  anc
naturally occurring mineral fibers, are presented in Table 3-6.   These analyti-
cal methods evolved as measurement technology progressed and as tie
for mere precise information grew.
                                   3-i7      Dn-r>-DQ N:T c-.:r~£ o~

-------
     The earliest methods  measured mass,  providing total airborne dust concen-
trations.  The  use  of x-ray diffraction techniques  provided identification  of
the mineral phases present.  Filtration techniques were often applied to  rencve
the  larger particles, thus  permitting a measurement of respirable  particle
concentrations.  However,  these  methods  were limited by the  inability of the
x-ray  diffraction  technique  to  detect  small  particles  (National Research
Council, 1984)  and  by the inability to derive reliable estimates of  the  fiber-
concentration from the total dust concentration (Corn and Sansone, 1974).
     Counting methods, which  are far more sensitive than mass determinations,
can  usually  be applied to  samples  that  are too small to be  weighed.   They
permit the counting  of fibers separately from other  particles  and  emphasize
smaller  particles,  since  they greatly outnumber  large particles in the  a*'r
(National Research Council, 1984).
     The use  of these analytical methods is  dependent on the sampling te:r-
niques, sample preparation, and the resolution capabilities  of the instrur.e-ta-
f':~.  Sampling techniques must  be selected  for their  efficacy untie"  tre
specific conditions  encountered  and must be effective over  the  entire  range c*
concentrations  anticipated.   Current  methods  generally  use Nucleopore poly-
carbonate  or  membrane  filters  for  sample  collection (Altree-Williams  e^c
Preston,  1985;  Rendall and  Schoeman,  1985).  The  great variability  amc-.g
samples collected within and among plants (Esmen et  a!.,  1978; Corn and Sansone,
1974)  indicates  the  need  to define the number,  location, and  timing  of samples
collected (National  Research Council, 1984).
     Sample preparation  entails  the  proper handling  and treatment  of  the
collected  samples  to  facilitate  analysis by the  chosen method.  These  pro-
cedures  include  ensuring  no (or  insignificant) loss  of  sample.   For counting
methods, the fibers must be evenly deposited with spacing between the fibers tc
permit counting (National Research Council,  1984).
     Walton (1962,  as cited  in  National  Research Council,  1984) has  stated
that the use  of the membrane filter/phase  contrast  optical  microscope is the
only practical method for routine occupational monitoring.   Henry et al.  (19£2)
used an open-face membrane filter (Millipore; 0.45-Mm pore)  collection mode  and
a light  microscopy  method to count carbon fibers  and to  measure fiber  size  and
total  fiber mass  concentration.   Samples were obtained from various  st=t:or,a-y
emission sources and from a pilot-scale inhouse incinerator facility.  Pr-icr t
analysis by light microscopy,  carbon fibers were  separated  fror: ct*er  pe'ti:--
late matter vie acid treatment.   The  carbon  fibers  were tner> depcsite:  cv.c

-------
5.0-um pore-size Milipore  filters.   Low magnifications  (about 10X) were used
for  length measurements; width  and short lengths were measured using higher
magnifications (100 to 400X).   Carbon fibers were easily  recognized and  counted;
airborne  concentrations  ranged  from 24 to 1200 fibers/m , and fiber  lengths
were between  48  and 750  ym.   Liu et al.  (1983) used a dichotomous sampler witn
a  V-groove  impactor surface  and  optical microscopy to  detect fibers with
average diameters of 8 pm.
     However,  the  light  microscopy  methods  are  limited  in resolution  (only
fibers greater than 0.5  pm in length are visible) and generally do not provide.
fiber identification.  Greater sensitivity is required to analyze  fiber  concen-
trations  at  ambient levels and to provide  identification of  source  fibers
(Nations! Research  Council, 1984).   Use of  the electron microscope and  associ-
ated analytical  techniques  allows  morphologic  and structural  characterizefc-
and  elemental  analysis of  the  fibers  at high resolution  (Paoletti  et  al.,
19S4).
     6ot> $£'•' and  TEK can be used  with x-ray  analytical  methods  tc P-CV^'CT
mineralogic  identification of  fibers   (National  Research Council,  I9S-).
Electron probe microanalysis  is the term used when either SEM or  TEM is per-
formed in conjunction  with x-ray analysis.   This procedure produces elect-or
optical images of  surfaces and  generates x  rays  that are  characteristic of the
elements contained  within  the irradiated surface of the  compound  being  tested
(Pooley, 1977).  Chang (1983) also used SEM and  aerodynamic  size  analysis to
measure ambient levels of carbon fibers.  The major limitation of  SEM, however,
is the  resolution;  fibers  less  than 0.05 urn in diameter  are  not  resolved well
(Vallyathan and  Green, 1985), and  no structural data are  provided (National
Research Council,  1984).   TEM can  be used with several  identification methods
(Spumy,  1983),   The  high  resolving power  of  TEM,  equipped with  energy-
dispersive x-ray analysis and selected area electron diffraction,  can elucidate
the  morphology,  chemistry,  and  crystalline  structure of  the fibers (Vallyathc-
and  Green,  1985).   This method has been  used successfully  to analyze the
composition of  submicroscopic particles of  asbestos  dust.   One  of the main
advantages of  this  technique  is that it enables  the user  to  conduct  simulta-
neous analysis of  all  elements  that possess atomic  numbers higher  than  that of
sodium  (i.e.,  11)  (Pooley, 1975).    The  resolution  of TEM is high encug- tc
detect fiber dimensions as small as 0.2 nm (Chatfield and Lewis,  19BCO-
July 19SS                          3-19      C-;A-T--DC fC~ QJ:~£

-------
     Another current  analytical  method  is  the  LAMMA procedure, a mass spectrc-
scopic analysis  of  single fibers (Spurny,  1983).  The instrument combines  a
high-energy  laser  beam with a  mass  spectroscope to  provide a microchemical
analysis of individual fibers.
     Loo et al.  (1984)  reviewed the  various other methodologies that can  be.
used in  measuring  ambient carbon fibers.   None of these  methods  appears  to be
as  sensitive  as those  described above.  Passive collectors,  such as sticky
tape and the pridal  veil,  can  be used in conjunction with automated  image
analyzers or a  high-voltage tape analyzer.  Optical counters, such as a  tre-ss-
mitometer or LADAR  can detect only dense clouds  of fibers;  length distribution
may be determined  by observing  the light  scattering or  diffraction  patterns
exhibited by the fibers using  other photodetectors.   Microwave  absorption  is
usec to  irradiate  the carbon fibers with  microwaves and  to sense the  nested
fiber with  an   infrared  detector.   This method  is very  specific  for carbon
fibers and can  detect fibers 1 mm in size.  Other analytical methods involve
using low-voltage or  high-voltage (spark counter) grid detection  and measuring
the induced charge  acquired by  a conducting fiber in an  electric field.   Lcc
et al.  (1984)  also  developed a  prototype-induced charge  detector capable  cf
measuring fiber  length  distribution  between 0.08 and 9 mrc and fiber  concer,t
-------
3.6  REFERENCES


Alsbirk,  A.  K.  £.; Johansson,  M.;  Petersen, ft.  (1983)  Ojensymptomer 05
     eksponering for mineralfibre fra lyddaempende loftsplader [Ocular symptorrs
     and  exposure  to  mineral  fibres in boards for  sound-insulation  of ceil-
     ings]. Ugeskr. Laeg. 145: 43-47.

A1tree-Williams, S.;  Preston,  J.  S.  (1985) Asbestos and other fibre levels in
     buildings. Ann. Occup. Hyg. 29: 357-363.

Amandus,  H. E.;  Wheeler, R.;  Jankovic,  J.;  Tucker,  J.  (1987)  The  morbidity  a-d
     mortality  of   vermiculite  miners  and  millers  exposed  to tremolite-
     actinolite: part 1. exposure estimates. Am.  J.  Ind. Med.  11:  1-1.4.

American  Conference of  Governmental  Industrial  Hygienists.  (1985) Threshold
     limit values  for  chemical  substances in the work  environment adopted b>
     ACGIH with  intended changes for 1985-86.  Cincinnati,  OK:  America^ Ccr.fer-
     ence of Governmental Industrial Hygienitts;  pp. 34-35.

Auerbach, I.  (1986) Kinetics  for the degradation of nylon  and Kevlar pe^acn^te
     materials.  Albuquerque,  NK: Sandia  National  Laboratories, Aerodynamics
     Depa-tment; report  no. SAN2-B5-0853C.

Bslzer,  J.  L.   (1976)  Environmental  data;  airborne concentrations found in
     various operations.  In:  Occupational exposure to  fibrous glass:  proceed-
     ings of  a symposium;  June 1974; College  Park,  MD. Cincinnati,  OH:  U.  S.
     Department  of Health,  Education and  Welfare,  National   Institute  for
     Occupational  Safety and  Health; pp. 83-89; HEW publication  no.  (NlCSf-1)
     76-151. Available from: NTIS, Springfield, VA; PB-258869.

Baris, Y.  I.;  Saracci,  R.; Simonato, L.; Skidmore, J.  W.; Artvinli, M. (19£1)
     Malignant  mesothelioma and  radiological chest abnormalities  in two
     villages  in$Central Turkey:  an epidemiological and environmental investi-
     gation. Lancet 1: 984-987.

Bayliss,  D. L.;  Dement, J. M.; Wagoner, J.  K.; Blejer, H. P.   (1976) Mortality
     patterns  among fibrous glass production •work-ers.  Ann. N.  Y. Acad. Sci.
     271: 324-335.

Bertazzi, P.  A.; Zocchetti, C.; Pesatori,  A.; Radice,  L.; Riboldi, L. (195-:)
     Mortalita per:tumori  in  una coorte di  produttori di  fibre vetrose [Cancer
     mortality among workers in the  glass fibre production industry].  Med.  Lav.
     75:  339-358.  •

Bertazzi, P. A.; Zocchetti, C.; Riboldi,  L.; Pesatori,  A.; Radice,  L.;  Latccca,
     R.  (1986)  Cancer mortality of  an  Italian cohort of workers  in man-traae
     glass-fiber production.  In:  Saracci,  R.« «d.  Contributions  to the IASC
     study  on  mortality and  cancer  incidence  among man-made rr.ineral  fifcer
     production workers. Scand. J. Work  Environ. Health 12 (suppl.  1):  65-71.

-------
Burden, G. ;  LeGuen,  J.  M. ;  Rood, A.  P.;  Rooker,  S.  J.  (1980) Comprehensive
     methods  for rapid  quantitative analysis of  airborne particulates  by
     optical  microscopy,  SEM and TEM with special  reference  to asbestos.  Stud.
     Environ. Sci. 8: 323-328.

Carpenter,  J.  L.;  Spolyar,  L. W. (1945)  Negative  chest  findings  in  a  mineral
     wool industry. J. Indiana State Med. Assoc.  38: 389-390.

Chang,  R. J.  C.  (1983)  Generation and characterization of  participate  aerosols
     from virgin carbon fibers and composites. Oiss. Abst.  Int. Pt.  B 44: 6.

Chatfield,  E. J.; Lewis, G. M. (1980) Development and application of an ana^yt-
     ical technique  for measurement  of asbestos fibers in  vermiculite.  In:
     Scanning  Electron  Microscopy/1980/I. AMF O'Hare,  IL:  Scanning Electro-
     Microscopy, Inc.; pp. 329-340.

Cherrie, J. ;  Dodgson, J. (1986)  Past exposure to airborne  fibers  and otne-
     potential risk  factors  in the European man-made mineral fiber production
     industry.  In:   Saracci,  R. , ed.  Contributions to  the  IARC study  en
     mortality  and  cancer incidence  among man-made  mineral  fiber production
     workers. Scand.  J.  Work Environ. Health 12 (suppl.  1): 26-33.

Cherrie, J. ;  Dodgson,  J. ;  Groat,  S.;  MacLaren, W.  (1986) Environmental surveys
     in tne  European man-made mineral fiber production industry. In:  Serecc'i,
     R.  , ed.  Contributions to the IARC study on  mortality  and cancer incidence
     among  man-made  mineral  fiber production workers.  Scand. J. Work  Environ.
     Health 12 (suppl. 1): 18-25.

Cherrie, J.  ;  Krantz,  S. ;  Schneider,  T. ;  Ohberg,  I.;  Karrstrup, 0.;  Linande-,  w.
     (1967) An experimental  simulation of an early rock  wool/slag wool prcsuc-
     tion process. Ann.  Occup. Hyg. 31: 583-593.

Cholak,  J. ;  Schafer,  L.  J.  (1971) Erosion of fibers  from installed fibrous-
     glass  ducts. Arch.  Environ.  Health 22: 220-229.

Corn, M.; Sansone,  E.  B.  (1974) Determination of  total  suspended particulate
     matter and  airborne  fiber concentrations at three fibrous glass  manufac-
     turing facilities.  Environ.  Res. 8: 37-52.

Corn, M.;  Hammad, Y.; Whittier,  D.;  Kotsko, N.  (1976)  Employee  exposure to
     airborne fiber  and  total particulate matter  in two mineral  wool  facili-
     ties.   Environ.  Res.  12:  59-74.

Crawford, N.  P.;  Kello,  D.;  Jarvisalo, J. 0. (1986) Monitoring and evaluating
     man-made mineral fibres:  work of a WHO/EURO  reference  scheme.  Presented
     at: international  symposium on  man-made mineral  fibres  in  the  working
     environment;  October;  Copenhagen,  Denmark.  Copenhagen,  Denmark:  World
     Health Organization, Regional Office for Europe.

Dixon,  G. M. ;  Doria, J. ;  Freed,  J.  R.; Wood, P.;  May,  I.;  Chambers,  T.;  Desci
     P.   (19E5)   Exposure  assessment  for  asbestos-contaminated vermiculite.
     Washington, DC:  U.  S.  Environmental Protection Agency,  Office  of Pesti-
     cides  and  Toxic Substances; EPA  report no.  EPA-56C/5-85-013.  Ava-'iar'e
     from:   NTIS, Springfield, VA; PBE5-l£30eO.

-------
Dodgson,  J.;  Cherrie, J.;  Groat,  S.  (1987)  Estimates of past exposure  to
     respirable  man-made mineral  fibres  in  the European  insulation  wool
     industry. Ann. Occup. Hyg. 31: 567-582.

Enterline,  P.  E.;  Marsh, G. M.  (1982)  Mortality of workers in the  man-made
     mineral fibre industry. In: Biological effects of man-made mineral fibres:
     proceedings of  a WHO/IARC conference; April  1982;  Copenhagen,  Denmark.
     Geneva,  Switzerland: World  Health Organization,  Regional  Office  for
     Europe; pp. 965-972.

Enterline,  P.  E. ;  Marsh, G. M.; Henderson, V.  (1986) Mortality update  of  a
     cohort of U. S.  man-made mineral fiber workers. Presented at:  internation-
     al  symposium  on man-made  mineral  fibres  in  the working environment;
     October; Copenhagen, Denmark.  Copenhagen,  Denmark:  World  Health Organiza-
     tion, Regional Office for Europe.

Esmen, N.  A.;  Hammad, Y. Y. ; Corn,  M. ;  Whittier,  D.; Kotsko,  N. ;  Haller, K. ;
     Kann,  R.  A.   (1978)  Exposure  of employees  to  man-made mineral  fibers:
     mineral wool production.  Environ. Res. 15:  262-277.

Esmen, N.  A.;  Corn,  M. ;   Hammad, Y.  Y.;  Whittier,  D.; Kotsko,  N.;  Haller, M. ;
     Kahn,  R.  A.   (1979a) Exposure of employees to man-made mineral  fifce-s:
     ceramic fiber production.  Environ. Res. 19: 265-278.

Esmen, N. ;  Corn, M. ;  Hammad,  Y. ;  Whittier, D. ;  Kotsko, N.  (1979b) Summary of
     measurements  of  employee  exposure to airborne dust  and  fiber in sixteen
     facilities producing man-made  mineral  fibers.  Am.  Ind.  Hyg. Assoc.  J.
     40:  106-117.

Fowler,  D.  P.;  Balzer,  J.  L.;  Cooper, W.  C. (1971)  Exposure  of  insulation
     workers to airborne  fibrous glass. Am. Ind. Hyg.  Assoc. J. 32: 86-S1.

Gamble,  J. ;  Greife,  A.; Hancock,  J.  (1982) An  epidemiological-industrial
     hygiene study of talc  workers. In: Walton, W.,  ed.  Inhaled particles V:
     proceedings of an  international symposium; August  1980;  Cardiff, Wales,
     United Kingdom.  Ann. Occup. Hyg. 26: 841-859.

Gamble,  J.;  Sieber,   W.  K.;  Wheeler,  R.  W.;  Reger,  R.;  Hall, B.  (1985) A
     cross-sectional  study  of  U.  S. attapulgite workers.  Presented  at:  sixth
     international  symposium on inhaled particles; September;  Cambridge,  United
     Kingdom.

Gardner,  M.;  Winter,  P.  D.;  Pannett, 8.; Simpson, M.  J.  C.;  Hamilton,  C. ;
     Acheson, E. D.  (1986)  Mortality study of workers in the  man-made mineral
     fiber  production industry  in the United Kingdom. In: Saracci,  R., ed.
     Contributions to the IARC study on mortality  and cancer incidence among
     man-made mineral fiber  production workers.  Scand. J. Work Environ.  Healtr.
     12 (suppl.  1): 85-93.

Gieseke, J.  A.;  Reif, R. B.; Schmidt, E. W. (1984) Characterization of  cs'tcn
     fiber emissions  from current and projected activities for the manufacture
     and disposal  of  carbon fiber products. Research Triangle Park, NC:  Envi-
     ronmental Sciences  Research Laboratory; EPA report  no. EPA-600/3-8--C21.
                                   3-23      D*.-rT--DD  NC-  C.'"":  Oc-  CI'E

-------
Harper,  W.  L. ; McAlister,  C.  E.  (1975) Effect of  radiation  on Kevlar-4?.
     Compatibility Conference, Kansas City, KS; Report No. Y/DA-6527-AES7.

Henry, W.  M. ;  Melton, C.  M.; Schmidt, E. W.  (1982) Method for measuring carbon
     fiber  emissions  from stationary sources.  Research  Triangle  Park, NC:
     Environmental  Sciences  Research Laboratories,  Environmental  Measurement
     and Characterization  Division;  EPA report no. EPA-600/3-82-080.  Available
     from: NT1S, Springfield, VA;  PB83-118760.

Hill, J.  W.  (1977) Health aspects of man-made  mineral  fibres.  A review.  Ann.
     Occup. Hyg. 20:  161-173.

Hill, J.  W. ;  Vhitehead, W. S.;  Cameron,  J.  D.; Hedgecock, G.  A. (1973) Glass
     fibres: absence  of pulmonary hazard in  production  workers.  Br.  J. Ind.
     Med. 30: 174-179.

Johnson, D.  L. ;  Healey, J. J. ;  Ayer, H.  E. ;  Lynch, J.  R. (1959)  Exposure  tc
     fibers  in  the manufacture  of fibrous  glass.  Am.   Ind.  Hyg.  Assoc.  J.
     30:  545-550.

Jones, H.  D. ;  Jones,  T. R. ;  Lyle, W. H. (1982) Carbon  fiber:  results of  a
     survey of  process  workers  and their environments  in a factory producing
     continuous filament. Ann. Occup. Hyg. 26:  861-867.

JRE  (1S£2)  Level  II  materials  balance  - vermiculite  [interim  draft  final].
     Washington, DC:  U.S. Environmental  Protection Agency.

Kleir.feld, M. ;  Messite,  J. ;  Kooyman, 0.; Zaki,  M.  H.  (1967) Mortality ar.r.rc
     talc  miners  and millers  in Nev,  York  State.  Arch.  Environ. Healtn
     14:  668-687.

Liu, B.  Y.  H. ;  Pui,  D. Y.  H. ;  Wang,  X. Q. ;  Lewis,  C.  W. (1983) Sampling of
     carbon fiber aerosols. Aerosol.  Sci. Techno!. 2: 499-511.

Lockey,  J.  E. ;  Brooks,  S.  M.; Jarabek,  A.  M.;  Khoury,  P.  R.;  McKay, R.  T. ;
     Carson, A.;  Morrison,  J.  A.; Wiot,  J.  F.; Spitz,  H. B. (1984) Pulmonary
     changes after exposure to vermiculite contaminated  with fibrous  tremolite.
     Am.  Rev. Respir.  Dis. 129:   952-958.

Loo, B.  W.;  Adler, T. C.;  Goulding,  F.  S.;  Landis, D.  A.; Madden,  N.  W. (1554)
     Advances  in  the measurement of ambient carbon fibers.  Aerosol  Sci.
     Techno!. 3: 31-38.

Maggioni, A.; Meregalli, G.; Sala, C.;  Riva, M. (1980) Patologia respiratone e
     cutanea negli addetti alia  produzione di fibre  di vetro (filato)  [Respira-
     tory and skin diseases In glass  fibre workers]. Med.  Lav.  3:  216-227.

Malmberg,  P.;  Hedenstrom,  H.; Kolmodin-Hedman,  B.; Krantz,  S. (1984) Pulme-ary
     function  in workers of a mineral  rock  fibre plant.  In:  Guthe,  T.,  ec.
     Biological  effects  of man-made mineral  fibres: proceedings of a  WHD'IARI
     conference  in association  with JEMRE   and  TINA,   v.  1;  April  1SE2:
     Copenhagen,  Denmark.  Copenhagen,  Denmark: World  Health  Organization.
     Regional Office  for Europe; pp.  427-435.

-------
Marconi,  A.;  Corradetti,  E. ;  Manrtozzi, A.  (1987)  Concentrations  of man-made
     vitreous fibres  during  installation  of Insulation materials aboard ships
     at Ancona Naval Dockyards. Ann. Occup. Hyg.  31: 595-599.

May, D.  R. ;  Ross,  J.  H. (1972) Accelerated weathering of polyamides, aromatic
     polyamides, and  polybenzimidazole fabrics.  Dayton, OH:  Wright-Patterson
     Air  Force Base, Air Force Materials Laboratory; report no. AFML-TR-72-137.
     Available from: NT1S, Springfield, VA; AD-906385.

Millette, J.  R. ; Twyman,  J.  D. ; GHansen,  E. C. ;  Clark, P.  J.  ; Pansing, M. F.;
     (1979) Chrysotile, palygorskite,  and halloysite in drinking water. Scan-
     ning Electron. Microsc.  1: 579-586.

Morgan, R. J. ;  Pruneda, C.  0.; Kong,  F.  M.  (1983) Aging studies  of Kevlar 49
     fibers.   In:   Proceedings  of  the  American  Chemical  Society  meeting.
     Berkeley,  California:  Lawrence  Livermore National Laboratory; repc-t
     no.  UCRL-89971.

Moulin, J. J.;  Mur, J.  M. ; Wild, P.; Ferreaux, J.  P.; Pharr,,  Q. T.  (1966) Ore'.
     cavity and  laryngeal  cancers  among  man-made  mineral  fiber production
     workers.  Scand. J.  Work Environ. Health 12:  27-31.

National  Institute for Occupational  Safety and  Health. (1977)  Criteria fcr
     a  recommended  standard. .. .occupational   exposure  to  fibrous  o^ass.
     Cincinnati, OH:  U. S.  Department of  Health,  Education, and  Welfare,
     Public  Health  Service,   Center  for  Disease  Control;   DHEW  (NIOSH)
     publication   no.  77-152.   Available  from:  NTIS,  Springfield,  VA;
     PE-274195.

National  Research  Council.  (1984) Asbestiform fibers:  nonoccupational  health
     risks,  Washington, DC: National Academy Press.

NOES,  National  Occupational  Exposure  Survey  (1980-1983) [data base].  (19S-)
     Cincinnati, OH:  U.  S.  Department of  Health and Human Services, National
     Institute for  Occupational Safety and Health. Disc; ASCII.

NOHS,  National  Occupational  Hazard  Survey (1972-1974)  [data base].  (1976)
     Cincinnati, Ori:  U.  S.  Department of  Health and Human Services, Nctic-.c1
     Institute for  Occupational Safety and Health. Disc; ASCII.

NOHS,  National  Occupational  Hazard Survey [data base].  (1980) Cincinnati,  Cu:
     U.  S.  Department  of  Health  and Human Services,  National  Institute for-
     Occupational  Safety and Health. Disc; ASCII.

Ohberg, J. (1987)  Technological  development of the  mineral  wool industry  in
     Europe.  Ann.  Occup. Hyg.  31: 529-545.

Olsen,  J.  H. ; Jensen,  0. M.  (1984) Cancer incidence  among employees in one
     mineral  wool  production  plant in Denmark. Scand. J. Work Environ.  Heeltr
     10:  17-24.
Paoletti,  L. ;  Caiazza,  S. ;  Donelli,  G. ;  Pocchiari,  F.  (1964) Evaluation _&.>
     electron  microscopy 'techniques of asbestos  contarr.inatic-- ir. ini^st ""•=_-_
     cosmetic, and pharmaceutical talcs. P.egul. Toxicc1.  5-.='r-=;:\  *• 222-2Zt.
                                    3-25

-------
Pooley,  F.  D.  (1975)  The identification of  asbestos  dust  with an electron
     microscope microprobe analyser. Ann. Occup. Hyg. 18: 181-186.

Pooley,  F.  D.  (1977)  The use  of an analytical electron microscope  in  the
     analysis  of  mineral dusts.  Philos.  Trans.  R.  Soc.  London Ser.  A.  266:
     625-638.

Kendall, R. E. G.; Schoeman, J. J.  (1985) A membrane filter technique for glass
     fibres. Ann.  Occup. Hyg. 29: 101-108.

Saracci,  R.;  Simonato,  L.;  Acheson, E. D.;  Andersen, A.;  Bertazzi,  P.  A.;
     Claude,  J.;  Charnay,  N.;  Esteve, J.; Frentzel-Beyme,  R.  R.; Gardner,  M.
     J. ; Jensen,  0.  M. ; Maasing, R. ; Olsen,  J. H.; Teppo,  L.; Westerholm, P.;
     Zocchetti, C.  (1984)  Mortality and incidence of cancer of workers in tie
     man made  vitreous fibres  producing industry: an international investiga-
     tion at 13 European plants.  Br. J.  Ind.  Med. 41: 425-436.

Schneider, T.  (1979) Exposures to man-made mineral fibres in user  industries ir
     Scandinavia.  Ann. Occup. Hyg.  22: 153-162.

Schneider, T.  (1986)  Manmade mineral fibers  and other fibers in the air a-.d in
     settled dust. Environ.  Int.  12: 61-65.

$r,3T,-,D-ri H. ;  Hayes,  M. ; Julian,  J. ; Muir,  D. (1982) Mortality experience c*
     glass fibre  workers.  In:  Biological effects of  man-made  miners", fibres:
     proceedings  of  a  WHO/IARC conference, v.  1; April;  Copenhagen,  Der,- = ^,.
     Geneva,  Switzerland:  World  Health  Organization,  Regional  Office  for
     Europe; pp.  347-349.

Shannon, H.  S. ; Hayes, M. ;  Julian,  J.  A.; Muir, D.  (1984) Mortality experience
     of glass  fibre workers. Br.  J.  Ind. Med. 41: 35-38.

Shannon, H.  S. ; Jarr,ieson,  E. ;  Julian, J. A.; Muir, D. C. F. ; Walsh,  C.  (19S6)
     Mortality experience of glass  fibre workers--extended  follow-up. Preser.tec
     at:  international symposium on man-made mineral fibres  in  the  work-ing
     environment;   October;   Copenhagen,  Denmark.  Copenhagen,  Denmark:  World
     Health Organization, Regional  Office for Europe.

Shannon, H.  S.; Jameson,  E.;  Julian, J. A.; Muir, D. C. F.; Walsh,  C.  (1SE7)
     Mortality  experience  of glass  fibre workers --  extended  follow-up. A".
     Occup. Hyg.  31: 657-662.

Simonato,  L. ;  Fletcher, A.  C.;  Cherrie, J. ;  Andersen, A.;  Bertazzi,  P.  A.;
     Charney,  N. ; Claude, J.;  Dodgson, J.;  Esteve,  J.;  Frentzel-Beyme,  R. ;
     Gardner,  M.  J.; Jensen, 0.;  Olsen,  J.; Saracci, R.; Teppo, L.; Westerholir,,
     P.;  Winklemann,   R.; Winter,  P. 0.; Zocchetti, C.  (1986a)  The man-made
     mineral  fibers  (MMMF)  European historical cohort study: extension  of tne
     follow-up.  Presented at:   international  symposium  on  man-made  minere"
     fibres  in the  working  environment;  October;  Copenhagen,   Denmark.
     Copenhagen,  Denmark:  World  Health  Organization,  Regional  Office  for
     Europe.
Julv 19:5                          3-26       DRAFT—DC  NC~  D-I":  Cn  C!

-------
Simonato,  1.;  Fletcher, A.  C.;  Cherrie,  J. ;  Andersen, A.; Bertazzi,  P.  A.;
      Charney,  N.;  Claude,  J.;  Dodgson, J.; Esteve,  J.;  Frentzel-Beyme,  R. ;
      Gardner,  M. J. ; Jensen, 0.; Olsen, J.; Saracci,  R.; Teppo, L. ; Westerned,
      P.;  Winklemann,  R.;  Winter, P. D.;  Zocchetti,  C.  (1986b) Updating  lu-.g
      cancer  mortality among a  cohort  of  man-made mineral fibre  production
      workers in seven  European countries.  Cancer Lett.  30: 189-200.

Skuric,  2.;  Stahuljak-Beritic,  D.  (1984) Occupational exposure and ventilato->
      function  changes in  rock  wool workers.  In:  Guthe,  T.,  ed.   Biological
      effects of  man-made  mineral fibres:  proceedings of a WHO/1ARC conference
      in  association with JEMRB and TIMA,  v. 1; April  1982; Copenhagen, Denma-k.
      Copenhagen,  Denmark:   World Health  Organization,  Regional  Office  for
      Europe; pp. 436-437.

Spumy,  K.  R.  (1983)  Measurement and  analysis of  chemically changed  mine'c1
      fibers  after  experiments  in  vitro  and  in vivo.  EHP Environ.  Hea-'tr.
      Perspect. 51: 343-355.    ~~

.Spu-ny,  K.  R. ; Pott,  F. ;  Stoeber,  W.;  Opiela, H.; Schoermann, J.; Weiss,  G.
      (19S3)  On the chemical changes of asbestos  fibers and MMMFs in bic:cg;'C
      residence  and  in the environment:  part  1.   Am.  Ind.  Hyg.  Assoc.  J.
      44:  833-645.

Tcy.  M.  £.  (I9E7)  Kevlar  photochemical degradation mechanisms. Sunnyve'e,  C-:
      Science Applications  International,  Aerosystems  Department;  report  r,c.  swC
      TK  6007.  Available from: NTIS,  Springfield, VA;  AD-A183273.

Vailyathan, V.; Green,  F.  H. Y.  (1985)  The role of analytical  techniques  in  t-.e
      diagnosis of  asbestos-associated  disease. CRC Crit.  Rev.  din. Lac.  $::.
      22:  1-42.

van  der  Wai,  J.  F.; Ebens, R.;  Tempelman, J.  (1987) Man-made mineral fibres in
      homes caused  by thermal insulation.  Atmos. Environ. 21:  13-19.

Walton,  W.  H.  (1982)  The  nature, hazards  and  assessment of occupational  expo-
      sure to airborne  asbestos dust: a  review. Ann. Occup. Hyg. 25(2).

Waxweiler,  R.  J. ;  Zumwalde, R.  D.;  Ness,  G. 0.; Brown, D. P.  (1985) Mortality
      among  persons mining and milling  attapulgite clay. Cincinnati, OH:  U.  S.
      Department of Health and Human Services,  National Institute for Occupa-
      tional Safety and Health, Centers  for Disease Control.

Wegrr.ar,,  D.  H.; Peters, J.  M.;  Boundy,  M.  G.; Smith, T. J.  (1982)  Evaluation of
      respiratory effects in miners  and  millers exposed  to  talc  free  of asbestos
      and  silica. Br. J. Ind. Med. 39: 233-238.

Wei 11,  H.;  Hughes, J.  M.;  Hammad,  Y.  Y.; Glindmeyer, H.  W., III; Sharon, G. ;
      Jones,  R. N.   (1983)  Respiratory  health  in workers  exposed  to man-mace
      vitreous  fibers.  Am.  Rev. Respir.  Dis. 128: 104-112.

WHO/EURO  Technical Committee for Monitoring  and   Evaluating MMMF.  (19£5)  Tr.e
      WHO/EURO  man-made mineral  fiber reference scheme. Scan:!.  J.   Worn Environ.
      Health 11: 123-129.
July 1S££
3-27

-------
Wojnarovits-Hrapka,  I.  (1977) Structural  examination of  inorganic  fibrous
     material. Part I. Epitoanyag 29: 498-503.

Wojnarovits-Hrapka,  I.  (1978) Structural  examination of  inorganic  fibrous
     materials. Part II. Epitoanyag 30: 13-18.

Wright, G.  W.  (1968) Airborne fibrous  glass  particles: chest  roentgenograrcs c*
     persons with prolonged exposure. Arch. Environ.  Health 16:  175-181.
                                    3-2£       DRi"--DC  M:T

-------
                        4.   RETENTION AND BIODISPOSITION
 4.1  INTRODUCTION
      The two primary factors  that  influence  the  deposition of fibers in the
 respiratory tract are aerodynamic  equivalent  d-iameter and  aspect ratio (i.e.,
 length/actual  diameter).   In general,  deposition  and retention of  fibers
 decrease with increasing diameter  and  length;  most fibers  recovered froir, trie
 lung are <3.5 urn in  diameter  and <5  urn in  length.  Mechanisms  involved in  tie
 clearing of  fibrous  materials from the  lung  include  removal  by mucocilia-y
 activity,  transport  by  alveolar  macrophages  to regional  lymph nodes, uptake
 b>  epithelial cells,  and direct translocation of  fibers  between epitheVis1
 cells to the interstitium and the  pleura.  Transport via the systemic circula-
 tion has also been  observed.   Short (<10 urn)  fibers are usually phagocytizei bj.
 alveolar macrophages prior  to clearance;  longer fibers  fragment easily aic
•often are  found  only partially  phagocytired.   Fiber types and  composition
 affect solubility and fragmentation.  Natural fibers (e.g., asbestos, attap.'-
 gite, erionite)  essentially  are in  soluble  and  split longitudinally in to
 thinner fibers,  making  clearance difficult.  Manmade (glass, mineral wool,  ar.c1
 ceramic) and synthetic  (aramid, carbon and polyolefin) fibers  break crosswise,
 increasing solubility and the  rate  of clearance.

 4.1.1  Host Entry and Retention
      Fibers enter the body by inhalation or by oral intake.   Various factors
 influence the deposition of inhaled  particles in the  respiratory tract.  Whe~
 nonfibrous compact dust particles  are  inhaled, the particles having a diameter
 greater than about  5 urn are generally trapped in  the  nasal passages before
 reaching the respiratory system  (Walton,  1982,  as  cited in National Research
 Council, 1984).   However, the aerodynamic behavior of fibers deviates sigr-'fi-
 cantly from that of  nonfibrous particles.  Fibers  having a  given diameter will
 behave aerodynamically  in the same manner as spherical particles of the same
 density that have  a significantly larger diameter  (Gross,  19E1;  Ti;r,rre" 1
                                     4-1

-------
et al., 1970).  The equivalent or aerodynamic diameter of a particle  is  defined
as  "the  diameter of  a  sphere with a density of  1  g/cm3 that has the same
falling speed as the particle."
     Timbrel! (1965) found that the equivalent diameter of a fiber is determined
predominantly by the actual diameter with length being of secondary importance,
especially when the fiber has a high aspect ratio.   The aerodynamic diameter  of
glass  and  asbestos  fibers was about three  times  the  actual diameter of the
fibers.  Since  the  largest compact particles normally found  in the  lungs ere
about  10 um  in  diameter,  it was concluded that asbestos fibers with  diameters
greater than about 3 um would be very unlikely to reach the alveoli.
     Gross  (1981)  developed graphs representing aerodynamic  behavior charac-
teristics of glass fibers to illustrate the relationship of fiber  dimensions  t:
the equivalent  diameter  of a unit-density-sphere having the  same  aerodyneT^c
behavior  (Figure 4-1).   The  relationship of these  dimensions to alveola--
deposition  was  also studied.   It  was  found that the  aerodynamic  equivalent
dia«»te- of  glass  fibers  in the  respirable  range  was  approximately three fiires
its actual  diameter  when  the aspect ratio  approximated  20.   The  aerod/narric
equivalent diameter increased to roughly three to four times the actual  d'ar.eter
for aspect  ratios  greater than 40 and  up  to lengths  of 150 urn.   It was sug-
gested that  a  glass  fiber with a unit-density-sphere  larger than  3.5 um  (i.e.,
diameter ^1.5 urn  and  length >10 urn)  has little chance of alveolar deposition.
Using  mathematical models,  Gross (1981) also showed  that  alveolar deposition
increased  as the diameters  of the fibers  decreased  (Figure 4-2).   Maximum
alveolar deposition  was  reported for  inhaled fibers with 0.25 to  0.5 pm  actual
diameter (about  17  to 18 percent by count).  Deposition of fibers of various
diameters drops rapidly with increasing length, being about 16 percent by count
for a  glass fiber with 0.5 um actual diameter  and  10 urn in  length  to  about
3 percent by count for 120-um-long fibers with the same diameter.
     In  a  similar study,  Asgharian  and Yu  (1988)  developed a mathematical
model  that  indicated a  maximum relative alveolar deposition  of  fibers  (type
unspecified) with  a  length(£p) of 7  um  and a diatneter(dp)  of 0.75 um following
nose-only breathing by humans (Figure 4-3).  The authors reported that although
total  deposition  is  high  (about 80 percent) with nose-only exposure, tracnec-
bronchial  and  alveolar deposition  of  fibers are  lower for  nose  brectvnc
compared to  mouth  breathing due to a strong impact deposition in the ncse  a->c
interference by nasal hairs.

-------
o:
ijj

Li

i
O

or
Ul

o
O
o
    6 -
5 -
    4 J
3 -
    2 H
    1 -
    0
      0
          20
40
60
 «

80
100
                             ODER LENGTH, urn
120
140
     Figure  4-1. The relationship of glass fiber diameter and length

     to  aerodynamic equivalent diameter  (Dae).

-------
O


V

O.
Z

O
 -
ui
o
o


I
               20
40
  60      80



FIBER LENGTH, urn
                                                 100
                                 120
                                                                  140
          Figure 4-2. The relationship  of  alveolar deposition to the


          diameter and  length of glas^jpers.

-------
  100
Figure  4-3. Bivariotc distribution of fiber deposition in the
olveoli  following nooc-only breathing by humans.
Source: Asghorian and Yu  (1988).

-------
     The significance  of  fiber dimensions on the deposition of  a  variety  of
inhaled fibers  following  nose-only  exposure of rats (sex and strain not  spec-
ified)  was  also investigated  by Morgan  (1979).   It was demonstrated that
alveolar deposition  increased  from  about 10 percent at an aerodynamic equiva-
lent diameter  (Dfle)  of 1  pm to a maximum of about 15 percent at  2 urn.• Deposi-
tion declined  steeply  to  a value of 3  percent  at  a D    of 3.4 urn  (for glass
fibers  with  a  count  median diameter (CMD) of 1.1 pm).   Alveolar deposition in
humans  was  reported  to follow  a similar pattern in  that  the maximum deposition
occurred at  a  D._  of about 2 urn, but absolute  values  are greater  than those
                oe
observed in rats.
     Both the  configuration and  dimensions of asbestiform fibers  determine
whether these  fibers will  be deposited  in  the  "lungs of  rats after  inhe1ef:c-
(Timbrell,   1965; Morgan,  197S; Evans etal., 1973).   Morgan  (157S) repo'te:
that for thin  fibers (e.g., crocidolite, Da  = 1.2 urn), most of the  fibers
                                            3e
inhaled by rats were exhaled without being deposited.  However, of  the ir,ste~ia'. s
tr.ct were deposited, more  than half  were found  below the level  of the  t*e:-e=.
In  contrast,  thick glass  fibers (D    = 3.4 urn)  were  all deposited  ir.  t-e
respiratory tract  when  inhaled, but  very few were found  below  the level  c* the
trachea.  Timbrel! (1955)  reported  that long fibers that escaped depositic- at
the nasal hairs  were  expected  to deposit where respiratory bronchioles branch
to  smaller  sizes and diameters.  The more  symmetrical a fiber  (e.g., glass
fibers), the greater the probability of  its reaching the alveoli.
     Similar results  were  reported  by  Evans et al. (1973),  who studied the
deposition of  inhaled  crocidolite asbestos  in   the  respiratory  tract  of  rats
using radioactive  tracer  techniques.   Examination  of autoradiographs  of thin
lung sections showed that the largest number of fibers  were located  at bifurca-
tions of the smaller bronchioles just downstream from the carina.

4.1.2  Translocation and Clearance
     Several mechanisms are involved in clearing fibrous  materials from the
lung.    These  include  removal by mucociliary activity,  transport by alveolar
rr.acrophages  to  regional lymph  nodes and distal sites,  uptake by epithelial
cells that  line  the  airways and  alveoli, and  direct translocetion  of fibers
between epithelial cells  to the interstitium and the pleura  (National  Rese=-cn
Council, 1964).

-------
     The physical  properties  (i.e.,  length and cross-sectional  dimensions)  of
fibers appear to determine the mechanisms of cellular interaction and transport.
Larger fibers deposited in the upper respiratory tract have a greater chance of
being removed from the lung by clearance mechanisms (Timbrel!, 1965).   Elimina-
tion from  the  lung via the airways  has  been  demonstrated, using radiotracer
techniques,  by  the presence  of radioactive fibers in the  feces  of exposed
animals  (Morgan,  1979;  Evans  et al., 1973).  However, the  rates  of clearance
may be  affected  by infections or smoking (Sanchis et al., 1971).   Fibers with
smaller diameters that penetrate deeply into the alveoli  before being deposited
are engulfed by  pulmonary macrophages.   However,  for  macrophages  to completely
engulf  fibers, they  should not exceed a critical length of approximately 5 p~
(Morgan, 1979,  1980).
     Sebestien et  al.  (1979)  reported that at  least  three  mechanisms  m5j  te
involved in  the  translocation of small chrysotile fibers from the  lung tc t*,e
pleura:   transport in the lymphatic circulation,  direct  penetration of tr,:n
fibers,   and  transport via  the blood  circulation  (especially when asbestcs
fibers are found in nonlymphoid organs).
4.2  MANMADE MINERAL FIBERS
4.2.1  Glass Fibers
4.2.1.1   Host  Entry  and  Retention.   The retention  of  glass  fibers following
single, repeated,  or chronic inhalation exposures  is  dependent on  fiber  dimen-
sions.  Retention  decreases  with increased diameter and  length  and increases
with  repeated  dosing.   Most of  the  retained  fibers were found  in  the rig1--*.
upper lobe and left  lobes of rats.  Those fibers retained in the lung were less
than 5 urn long.
     Morgan  et al.  (1980) and  Morgan and Holmes  (1984a) used  radioactive
tracers to  investigate  the  regional  deposition of inhaled glass  fibers  (of
discrete  size) in  the respiratory tract of the rat.   Groups  of eight Alderly
Park  (Strain 1)  SPF  rats (5 months  old,  approximately  350  g) were exposed
once, for 2  to 3 hours  and  by  nose only,  to  one of six  glass fiber aerosols.
Fibers with  a  nominal  width of  1.5 urn and lengths of 5, 10, 30, or 6C urn were
administered at  a rate  of  119,  99, 108, and 77 ug/L/min,  respectively:  3  *
10 um fibers were  administered  at 48 ug/L/min;  and nonsized fibers were  acrr.^r.-
istered at 236 mg/L/min.  Prior to  administration,  about 20 mg of esc*  fiber
                            24
sample were  irradiated with   Na.   Following irradiation, fibers were Swsre~cec
V.i.. icic                             t-i      r.r'CT._rv«i I<-,T r.
-------
in water  containing  0.001 percent Dowfax.   The amount of  fiber in the lung, in
airways external  to  the lung, and in the gastrointestinal tract,  skinned  head
(nasal passages),  head skin,  gastrointestinal  tract,  feces, and  remaining  .
carcass were measured  by  x-ray spectrometry.  (Data were  then normalized to an
aerosol concentration of 100 pg/L/min.)
     Immediately  after  exposure, the  total  mass of 1.5-pm  diameter  fibers
deposited  in  the animals' bodies ranged from about 16 \ig for the 5-pm long
fibers to  approximately 22 to 25 pg for the 30-  and  60-pm long  fibers.   The
mean total  deposition  of  3-ptn diameter fibers was 25 pg.  Approximately 2 tc
3 pg of  the sized  glass  fibers  were  found on the pelt;  animals  probably
ingested these  fibers  during preening.   Most (95 to  99  percent)  of the  rac'c-
activity recovered  initially was found in the gastrointestinal tract.  Sr,£""
            24
amounts of   Na were found  in the  skinned  head  (1  to 4.5 percent) and-  lures
(0.5 to 2.5 percent).   At 2  days, radioactivity  appeared  almost exclusively ir
                                                              ? '
the feces,  accounting  for S7 to  99 percent  of  the  recovered   Na.   Trace tc
srcs'i  amounts were found in  the  skinnec! head, gastrointestinal tract,  and  r-e"t.
     Longe^ (10  to  60  urn) fibers tended  to  concentrate in the gastrointestinal
tract.   By  contrast,  the  majority of  fibers  deposited in  the respiratory  tract
were shc't  (<1C prr long).   Most  fibers deposited in the  lungs were found i~ l~e
left azyoous and  right  diaphragmatic  lobes (30 and 27 percent of  lung  racic-
                                                  24
activity,   respectively);  about  20 percent of the   Na in the lungs was  in the
right apical  lobe.   The right cardiac and  right  azygous  lobes  each
approximately 11 percent.
     The data  suggest  that  sized  glass  fibers are readily taken  up  by the
lungs and  pass through the gastrointestinal  tract unabsorbed.  The  results
also indicate  that  deposition of glass fibers in  the  lower respiratory tract
anci  alveolar  region  declines with increasing  fiber length.   When  the authc-s
(Morgan  et  a!.,  1980;  Morgan  and Holmes,  1984a)  compared their  data  with
results  from  similar experiments,  they found  that alveolar deposition pesked
at  a diameter of approximately  2  pm  and  dropped dramatically between  2  and
3 pm.   The  authors  concluded  that despite  the apparent size  limitations  o^
respiration of  glass fiber particles,  fibers  with a diameter  of 3  to  6 prr. also
are  respirable and may  have a small but significant alveolar deposition.
     Griff is et al.  (1981) studied the lung burden of rats exposed to a gless
fiber aerosol  by  measuring the acid-insoluble sodium  content  of lungs.   Male
     rets (number,  age, and weight not reported)  were  exposed tc "Tone1  f^be-s

-------
with  a  nominal diameter  of  0.2 um for 6 hours/day for 2  or  5 days.   Lunc
burdens  of  glass  were  determined  1, 2, 4, or  5  days  after exposure.   The
average  mass  concentrations  for these exposures ranged from 35 to  50 mg/m  ;
approximately  two-thirds  of  this was "respirable."  The lung  burden after one
6-hour exposure to  the  aerosol  was 53 ± 11 ug.   Lung burdens  determined 2,  4,
and 5 days  after  exposure indicated that about 41 to  48 percent was cleared
from the upper airways between exposures.
     A study  by Hammad  et al.  (1982) was designed to measure  the retention  of
inhaled  glass fibers present primarily  in the alveolar compartment of the
lungs of male albino rats (strain and number not specified).   The  rats were
exposed, by nose only, 6 hours/day for 6 consecutive days.   They were sacrifice:
5 days after  the  last exposure.  The airborne fibers had a count median le-c*.~
(CM..) of 13  i 2.6 (± geometric  standard deviation, GSD) and  a CKD of 1.2  =
1.8 urn.   The  lungs were ashed and the fibers sized and  counted by optical  phase
microscopy.    The  fiber  content  of  lung tissue was compared with that  of  the
airborne djst  cloud.  The following relationship between  fiber retention a-c
fiber length  for  three  fiber diameter categories was  noted:   the retention of
fibers less than  0.5 urn in diameter was significantly higher  then  for large'
diameters,  with a peak of about  7.6 percent at 21-um  fi'ber  length.   In  contrast.
less than 1  percent  of  the fibers  having  diameters of 0.5 to 0.7  and  C.7 t;
1.0 urn were retained, with peak lengths of about 21 and 5 um, respective~\.
When the effect of fiber aerodynamic equivalent diameter on fiber retention  was
examined, maximum retention  of  about 7 percent was observed  at a diameter  cf
1 urn.   Retention decreased rapidly to about 0.5 percent at  equivalent diameters
of  1.5  to  2.5 um and about 0.01 percent at  a  diameter of about  4 um.    In
addition, retention  of  fibers  in the lung  decreased sharply  as the length  of
the fibers,  irrespective  of  diameter, increased from  2 to  10  um; thereafter,
the decrease  in retention was gradual.
     The distribution and physical  characteristics of  inhaled  resin-free  glass
wool in  the  lung  were examined  by  Le Bouffant et  al. (1984, 1987).   A group of
24 male and 24 female Wistar rats (8 to 9 weeks old)  inhaled an aerosol  of £ mg
glass wool fibers/ml, 5 hours/day, 5 days/week for 1  year.   About  50 perce-.t
of the airborne fibers  had lengths of 10 um or  less.   Approximately 65 percent
were less than 1  um  wide; an additional 15 percent were  between  1 and 2 urr.
wide.   After 1 year, half of the animals were removed frorr,  the dusting  chafers
and exposed to fresh air only.   The other  half were  exposed tc the fibers fcr
an additional  year.   Rets were   killed either immediately  e'ter •irne1s*^:r
     19£5                           4-9      DRAFT — DC- N"  Q."£ CS CTE

-------
12 or 24 months) or at various times after exposure was  terminated:   7,  12,  anc
16 months  for  animals exposed for 1 year  and  4  months  for  those  exposed  for
2 years.   An  unexposed group of 24 male  and 24  female  Wistar  rats  served as
controls.
     At  the end  of 12 months' exposure, the right upper lobe and  left  lobe of
the lung contained the highest concentrations  of glass wool (between 2.1  and
2.8 mg dust/g dry tissue).  The trachea, tracheobronchial  glands,  and diaphragm
contained  the  lowest  levels  of dust  (from below levels of  detection to
0.7 mg/g). , Females tended  to have slightly higher tissue  levels than males.
The authors concluded that  migration  of glass  fibers  (e.g., to the  diaphrag'")
was minimal.  Approximately  83 to 94  percent of  the  fibers deposited  in  the
left  lung  lobe,  airways, and diaphragm of  exposed  rats were  less than 5  y
long.   Within  7 months  after the 12-month exposure period, between 65  &~:
75 percent of the  inhaled fibers were cleared from the rats'  lungs.   There-
after, the rate of clearance declined  more slowly (Le  Bouffant  et  a!.,  195-}.
     At  24  and 28  months after the  inhalation  experiments began,  moderate
sorts': "et^CL'1 in  fibres is was observed in  the  lungs.   No other organs we-e
affected.  Fibers were found either free in the alveolus or  engulfed in a"!ve:"='
or septal  macrophages.   One  pulmonary  and four  digestive tract tumors were
observed (control  animals had none and  one of  these turners, respectively);  t~e
high number of  visceral  tumors suggests that some of  the dust may  have bee-
ingested with  the food  consumed by exposed rats.  During  the postexposure
periods,  the glass  wool  fibers took on an eroded and  coarse appearance.   Sore
fibers also showed a  loss of opacity,  and most  exhibited a reduction in soc'iur,
calcium,  and barium  content.   Small  accumulations of  dust macrophages,  prima-
rily at  the bronchoalveolar junctions,  were found after  12 and 24  months of
exposure.
     Glass  fibers  were analyzed by Johnson et al.  (1986) following recovery
from  rats  and hamsters  after nose-only exposure durations of 6  hours/day,
5 days/week, for  up  to 24 months.  The animals inhaled glass  fibers with  mea-
diameters  of 0.45,  3.1,  5.4,  and 6.1  urn in monodispersed  aerosols ranging  fror;
3 mg/m   for the  0.45-um mean diameter glass fibers to 15 mg/m  for  the 6.1-ur
mean  diameter  glass  fibers.   Fibers  were  recovered from tissues by sodiur
hypochlorite digestion and  examined by electron microscopy.  Animal variation
was noted,  but  the finer the dust, the  greater  the amount  of dust  reccve-e:
from the lung.   The number of fibers found  in lungs of the coarse fiber-exrose"
                                    4*10          DRAFT—D: N:T QU:~E c; ::~E

-------
groups were  larger  in comparison  to  finer dusts when the aerosol fiber concen-
trations were normalized, indicating differential  dust  retention.
4.2.1.2   Translocation and Clearance.   Microfibers  (diameter 0.15 urn, length
5.4 urn) are  readily eliminated  following single inhalation exposures.  Glass
fibers  (diameter  <3 pm for 50  percent  of  the fibers)  retained in the lung
following  single  or  repeated inhalation exposure are  cleared slowly with a
half-life  of  about  44 days.   The half-life clearance  value following chronic
exposure  is  about  600 days.   The rates of  clearance following  intratrachesl
administration decrease with  increased  fiber  length (>10 urn)  and  dose.   Short
fibers are phagocytized by alveolar macrophages and  cleared to the lymph  nodes.
Long fibers  are  usually partially phagocytized.   Fiber type  and compos •if: on.
which affect  solubility and  fragmentation,  also influence the  rates  of  fiber
retention  in  the  lung.   Following intrapleural injection in rats,  fibers we'e
found in  various organs;  translocation apparently  occurred  via  the blood
capillaries.
     Clearance  of   fiber-containing  macrophages  and particles  on alver'ic-
surfaces  from 8-  to 10-week-old male CD(SD)8R rats  was complete within  £ ds/s
                                                                              •3
after a single 5- or 10-hour inhalation to fiberglass particles  (10 tc 2G mg-'rr";
dimensions not  reported)  (Warheit et a!.,  1988).   A companion  j_n  vitro  study
indicated that complement activation may play a critical  role in the  ability  of
pulmonary  phagocytic  cells to selectively detect  inhaled particles on alveolar
surfaces soon after exposure to fibers (i.e.,  within 3  hours).
     Friedberg and  Ullmer  (1984)  also studied the elimination  of  glass  fiber
dust from rat lungs.  Approximately 75 percent of the inhaled airborne pa-ticles
had an aerodynamic  diameter  less than 5.0 pm, 50 percent  had a diameter less
than 3.0  pm,  and  32 and 21 percent  had  diameters of 2.0 and 1.5 pm, respec-
tively.    Each of  88 male Sprague-Dawley rats  (200 g,  age  not specified) was
placed  in an individual  cage and exposed  to 86 mg/m  "TEL" fibers  (which
contain 64.4  percent  Si05) for  10 hours.  Another set  of 40 rats  inhaled 86 rr,=
        3
fibers/in  , 10  hours/day,  for 3 consecutive  days.   Animals  were  killed  at
various times (between 1  and 90 days)  after  the  final exposure period.   The
lungs of  rats  receiving a  single  dose initially  contained  approximately  150 to
200 pg dust; nonexposed control  lungs contained 8 pg/lung.   Glass fiber  content
of exposed rat  lungs  dropped  slowly  but steadily, measuring  about  100, 75, and
50 pg/lung,  respectively, at  30,  60, and 90 days  postinhalation.   The lungs of
multiple-dose rats contained an average of 900 pg fibers at 1 daj  poste>r:sj^e;

-------
within 30 days, the average glass fiber content had dropped  to  500  ug/lung.  At
90 days, the fiber concentration was approximately 250 yg/lung.   Despite  a much
higher glass fiber  body  burden,  rats receiving multiple doses eliminated the
fibers at  nearly  the  same  rate as  the  single-dose  rats, with respective elimi-
nation half-lives of 46 and 40 days.  When data from all  animals  were  combined,
a mean elimination  half-time  of  44.3 days was calculated.  Thus, at 280 days
after exposure, it  was estimated that  the  body would  contain only 1 percent of
the glass  fibers  inhaled,  and after 440 days, only 0.1 percent  of the fibers
would remain in the alveoli.
     Long, thin JM  104/475 glass fibers persisted for an extensive period cf
time  in  the lungs  of female  Wistar rats  following inhalation exposures  of
100 ppm fibers 5  hours/day,  5 days/week for 1 year  (Muhle  et  a!.,  19S7).  A
clearance  half-life of approximately 600  days  was  reported; most of the  fire-s
retained by the  lungs had  a diameter between  O.OS and 0.2  pm  and were  mere
than 1 pm long.
     Griff is et al.  (1983) examined the deposition and elimination of  c'ass
fibers in  docs.   A  group of two male  and  two  female  beagle dogs were expose:
by nose only for  60 minutes to neutron-activated fibers  with  a CMD and  O'.  cf
C.15 and  5.4 urn,  respectively, and an activity median aerodynamic  diamete"  c*
2.5 urn.   Female dogs  were  approximately 2 years old  and  weighed between 6.S
and 9.0 kg;  the  two  3-year-old  male dogs weighed between  9.6 and 10.4  kg.
Following  inhalation  exposure (39  to  88 yg fibers/L), dogs were  placed in
metabolism cages, and urine and  feces  were collected  for 4  days, at which time
the animals were  killed.   Radioactivity was  measured  in excreta  samples  and in
the gastrointestinal  tract, head,  trachea,  and  individual  lung lobes.   The
dogs'  bodies contained approximately 2 to 6 mg (40 to 67  percent, respectively)
of the 5  to 9  mg  of  inhaled glass  fibers.   About  5 to 17 percent -of  this was
deposited  in the  deep lung.  Most  of  the  radioactive dose  (77  percent) was
eliminated within 4 days after exposure, with 96 percent  being  recovered  in the
feces; the  clearance  half-life  for glass  fibers was  calculated  to be 2  days.
Approximately  88  percent of the  radioactivity  that remained in the body  at the
time of sacrifice  was in the lungs, where fibers tended  to  concentrate  in the
right apical lobe.  The authors reported that the deposition patterns  exhibited
by the dogs in this study were in good agreement with previous  predictions  mace
for humans.  They  added  that  the aerodynamic diameter of  glass fibers appea-e:
to be a good predictor of respiratory tract deposition in dogs.
     19££                           4-12

-------
     Bellmann et al.  (1987)  reported  that short fibers  (<5  urn)  generally were
cleared from  rat  lungs  more rapidly than long fibers (>5 urn)  following  intra-
tracheal  instillation.   Fiber  type and composition also  seemed to  influence
retention by the lungs.   Long fibers with a low alkaline earth content (ceramic
wool and  104/475  glass  nticrofibers) and crocidolite had  comparable clearance
half-lives  (780 to  3500 days); long glass microfibers  with a high concentra-
tions of  alkaline  earth metals (104/E-glass) had a  low half-life  (55 days).
Glass and rock wools had intermediate half-lives (about  280  days).
     The  clearance  and  jn  vivo solubility of glass  fibers  in rat  lungs were
examined  by Morgan  et al.  (1980,  1982) and Morgan and Holmes  (1984a).   Fibers
of different lengths, with a nominal diameter of 1.5 or  3 urn,  were  administered
by intratracheal instillation to 5-month-old HMT Alderly Park  SPF  rats  (sex  r.ct
specified).    For  the 1.5-ym-diameter samples, groups of  12 animals  receiver
0.5 mg of 5-,  10-,  or 30-pm-long  fibers  or 1 mg  of  60-pm-long fibers in 0.5 ir._
saline.    Additional  groups  of  rats received 1 mg  doses of  10-  or 60-urr-lcnc,
3-ptr-wide  fibers  in  0.5 ml saline.   Fibers with  a  1.5-um diameter we-e
                24
irradiated with   Na prior to administration.
     After 1 year,  about 90 percent of  the  5 x 1.5 urn and  80 percent  of  the
10-pm-long  (both  diameter types)  fibers  were  cleared  from  the  body, with  s
clearance half-time  of  about 60 days.   No significant clearance of  the  30-  and
60-)jm-long  fibers  was observed during the  same  period  of time.  All fibers
were leached  significantly  during the investigation.  The median diameter  of
the 5 x  1.5 urn fibers was reduced by about 11 percent over  18 months;  for  the
10 x 1.5  urn and the 30  and  60  x 1.5 urn fibers,  diameters decreased  by approxi-
mately  28 and  55  percent,   respectively.   Many  of  the  longer  fibers  had
fragmented and  dissolved over  the course of  the  experiment.  At 19.5 months,
only 1  percent of  the  recovered  alveolar macrophages contained fibers.  No
abnormal  pathology  or change in lung weight  was  observed.   It  was concludec
that fibers with  a  diameter equal to or less than 3 urn and  a  length less  ths-
or equal  to 10 pm  are  efficiently cleared  from  the  lungs  (via macrophage-
mediated processes), while long fibers (>30 pm) are not cleared  to a significant
extent for  at  least 1 year  after  exposure.   The rate of dissolution of glass
fibers also was dependent upon fiber length.  Long fibers dissolved rapid'y  &*c
nonuniformly,  often  resulting  in  fragmentation.   Short  fibers dissolved slo-ly
and evenly.   The results suggest that glass fibers behave differently frcr
                                    «-i3          DRA-V-D: N:~

-------
amphibole asbestos fibers,  which  essentially  are insoluble JD v'tro and  not
readily cleared from the lungs (see Section 7.2.7.2).
     In an attempt to  better characterize the health  effects  of  glass  fibers,
Bernstein et al.  (1980)  exposed male F344 rats to fibers of defined sizes by
intratracheal instillation.  The  neutron-activated fibers had a geometric mean
diameter of 1.5 pm and a length of either 5 or 60 u.m.   The  deposition,  translc-
cation, and ultimate fate of these fibers were then  determined by radioassay  cf
  Zn.  Groups of  four  rats were instilled with 20 ug  of 60-um fibers,  2  ug of
5-um fibers  (equal number  of fibers), or 20 ug  of 5-um  fibers (equal mass of
fibers) and  sacrificed  18  to 20 weeks  after administration.  Additional groups
were instilled with similar doses  of fibers,  and four  rats  from each group were
sacrificed at various  intervals between at 6 months  and 2 years.   The "lungs.
lymph nodes, and  various other tissues were collected for  histologic examina-
tion and autoradiography to determine the presence of  fibers  in these tissues.
     The clearance half-times for  each group (as estimated  by  a one-compa-t<-e-t
expc-£-.tia1 rrocel and  the  percent of the initial dose that remained at sa:"i-
fice) were 35 and 38.5 days, respectively, for rats receiving 20 me of e-:t-;e-
lono or short fibers  and 22.5  days  for rats receiving 2  mg of the  short  fibers
(Bernstein et al., 1980).  The whole-rat clearance rates, however, were sir-ler
among  the  groups.  Histopathologic  examination  showed  differences in the
responses of rats exposed  to the  long and short fibers.  The  major  response  tc
the  short  fibers  was  the production of macrophage aggregations  in the  alveoli
and  lymph nodes, indicating that short fibers  were successfully phagocytizei  ty
alveolar macrophages and were  cleared to the  lymph nodes.  The response tc the
long fibers was notably different.  No long fibers were totally phagocytizei  by
macrophages, although  a  few were observed to be partially  phagocytized.
Moreover, only  a  few  smaller fibers (<60 urn  in  length)  were  observed  in the
lymph  nodes,  suggesting limited  breakage of  long fibers.   The  long fibers
produced a granulomatous response with giant  cells,  a response characteristic
of many types of foreign bodies.   In  addition  to the above findings,  few, if
any, fibers were found by light microscopy outside the respiratory tissues.
     In contrast,  glass fibers were  recovered  from the liver and  spleen'of
female F344  rats  and  rabbits (strain and sex  not reported) 2  to  5 years  e'ter
intratracheal instillation  of  2  to 10 mg of the test material (Spumy et el.,
1983).   Most fibers retained their original size, although it appeared  that
leaching and some dissolution had occurred after fiber migration.   IP  raP-^ts.

-------
preferential  leaching  of elements was  observed,  but the  leaching  was  net
homogeneous at different locations (Spumy et a!.,  1983).
     Wright and  Kuschner  (1977)  examined  the migration of short and long glass
fibers  following intratracheal  instillation Into  guinea  pigs.   Each of 30
guinea pigs (strain, sex, age, and weight not specified) was  administered 25 mg
of short  glass  fibers  (i.e.,  93 percent  of the fibers were  less  than 10 pm
long).  Thirty  animals  in a second group  of guinea pigs each received 12 mg of
a sample  in which 92 percent of the  fibers  were  longer than 10 urn.   Animals
were  sacrificed 2 years  after  dosing.    Fibrotic  lesions  developed at  the
terminal bronchioles of animals in the second group,  and long glass  fibers  we^e
found  in  the  areas  of  fibrosis.  Short  fibers induced alveolar  macrophage
aggregation but  not  fibrosis.   In both the short-  and long-fiber  groups,  hicK-
numbers of short-fiber-filled macrophages were observed  in  the lymph nodes,  tut
fibrosis was  absent.   It was  suggested that many  of  the long fibers  probably
had  fragmented,  resulting in a  short-fiber-like response.   Fragmentation  of
"0-5 fibers was  confirmee! by  electron micrographic studies  of ashed  lung  and
lyrr;ph nodes of  animals  administered  long fibers.   The authors conclude: the*
short  fibers  and fragmented  long  fibers acted similarly  in  that both  were
capable of migrating  from the lung to the lymph nodes and that neither  indeed
fibrosis.
     The translocation  of  JK  104 glass fibers through the respiratory  systerr
after injection  into  the pleura!  cavity  of  rats was  studied  by Monchaux et el.
(19S2).  Groups  of  15  Sprague-Dawley SPF male rats  (8 to  12 weeks  old) were
injected with 20 mg  of  fibers suspended  in 2 ml of  saline.   A total  of 1.8 x
10   glass  fibers with  a mean diameter of 0.229 urn and  mean  length  of 5.89 urr,
was  administered.  Two  rats  were killed  at various  intervals up  to 380 days,
and the fiber content in several lymphoid and nonlymphoid  organs was determined
by TEM.   Ninety days after intrapleural   injection, fibers were found in all
organs  analyzed:   mediastinal lymph  nodes,  lung parenchyma, spleen, liver,
kidneys, and  brain.   The number of all  fibers/g tissue was  approximately  the
same in all organs (about 10  to 10 ), except in the thoracic lymph  nodes where
concentrations  were  10  to 100 times  higher.   The  number  and mass of JM 10-5
glass  fibers  recovered/g mediastinal  lymph node tissue remained relatively
                                       p
constant up to  380 days (about 5 x  10  fibers/g and 0.5 mg/g tissue, respec-
tively).  Less than 1 percent of the fibers migrated from  the pleura!  cavity tc
the  lung.    In addition, the  mean length  of  the  fibers  remaining  ir tne lung
                                    4-15

-------
parenchyma increased with  time.   It was suggested that the  fibers  leave  the
pleura!  cavity  mostly through  the  stomata, reaching the  lymphatics  of the
parietal pleura  and,  by  retrogression,  the mediastinal  lymph nodes.   Alterna-
tively,  the  fibers may migrate across  the  visceral  pleura by a persorption
mechanism and then be  carried by  the  fluid  flow within the lung tissue and the
pulmonary lymphatic channels.   However,  the route that seems to be  the most
important is the blood capillaries in the'lung, since all  fibers  were recovered
from nonlymphoid organs 90 days after injection.
     No  information was  found in  the  available literature  on the translocatior.
and clearance of glass fibers following  oral administration  to animal species.
In humans, Cook  and Olson (1979) found that amphibole mineral fibers ingested
with drinking water are eliminated in the urine.   This finding strongly  suggests
gastrointestinal absorption  and passage  of  fibers  through  the kidney, although
the exact mechanisms and rates could not be clearly defined.   In  another  study,
silica fibers were recovered from the lungs of 20 individuals with no history
of occupational  dust  exposure  (Churg,  1983;  Churg  and Warnock,  195C1)  (see
section 4.4.2 for  details).

4.2.2  Mineral Wool
     Factors affecting retention  and  clearance of mineral fibers  are similar
to those discussed for fibrous glass with  similar dimensions.   Hammad (1954)
examined 49  male rats  (age,  weight,  and strain not reported) that  were exposed
by nose only to an aerosol of mineral wool  fibers  (300 fibers/cm ), 6 hours/cay
for 5  or 6  days (number of  days' exposure  unclear).  The CML and CMD of  the
fibers were 13 and 1.2 pm, respectively.  The respective CML and CMD of  mineral
wool fibers, recovered from  lung tissue of all  animals were 7.0 and 0.5  urn.
     Most of the  fibers recovered  5  days  postexposure  were  between  0.5  anc
1.0 pm wide  and 2 and 30  urn long.   In  tissue  examined at 5 days,  the  lupg
relative percent  deposition  was highest for fibers with a length af 11  pm aic
diameter of  0.5 pm.   Concentrations  were  much lower for fibers with large'
diameters or shorter  lengths.  Fiber clearance  was dependent on  length  anc!
diameter.  Fibers  >1 pm  in diameter and <5 pm in  length  were associated  witr.
the highest  retention  and, at 270 days  postexposure,  accounted  for app-cxi-
mstely 10 percent  of  the fibers initially  retained.   Retention  times dec'-essec'
sequentially for the following fiber  length categories:   5 to 7, 7 to 11.  21 to
30, 15  to  21,  11  to 15,  and greater than 30  pm.  Fibers with c'iar.ete-s   "ess

July 1S££                           4-15           DRtFT--C:  N-"~ C-": C-   CI'E

-------
than 0.5 urn  were  cleared first, followed by fibers with  diameters  of 0.5  to
0.7, 0.7 to 1.0, and 1.0 to 1.3 ptn.  Thus, fiber retention appeared  to increase
as  fiber diameter  increased.   Only 6 percent of the mineral wool fibers tr-.et
had been in  the lungs  at 5 days postdosing  remained in that tissue at 270 days
after exposure.  An overall clearance half-life (for fibers of all sizes) C*
175 days was reported.   The authors suggested that fiber clearance results  frorr.
two mechanisms:  normal biological clearance and elimination by dissolution anc
subsequent breakage.
4.2.2.1  Rock  Wool.   The retention of rock  wool  in  F344 rats (number, sex,
age, and  weight not  reported) exposed to a dust  cloud  of 10 mg rock wool
fibers/ml,  7 hours/day,  5 days/week  for  1 year  was determined by Johnson-et  a1.
(19S4).   Each  microgram  of dust contained approximately 9.2 x 10  fibrous  a-;
54.3 x 10  nonfibrous respirable particles.   Nearly 30 percent of the  resrira:"£
fibers were  between 10 and 20 yro  long and  had  a diameter greater than 1  urr;
about 70 percent were  narrower, 52 percent  were  shorter,  and  17  percent were
longer.   At  the  end of the 1-year exposure period, animals were either kil'e:
or  placed  in  a  nondusty  environment  for various time intervals.   At 6  and
13 months after  inhalation,  lungs  contained about 5 x 10   fibers/mg  tissue.
Retention was  closely  related to fiber diameter:  nearly  all  (93 percent)  cf
the recovered  rock wool fibers were less than 0.9 urn wide, whereas distribution
of  fibers by  length was comparable to that  of  the inhaled glass.  Rock wool
fibers were seen throughout the parenchyma of the left lung, in the airspace of
alveolar lining cells, and in the interstitium.   Although some free fibers  were
recovered,  the majority of  fibers were  found  in  alveolar and interstitial
macrophages.    A  6-ray  microanalysis  showed  no apparent change  in  the  elemental
composition of rock wool  fibers through  13  months after termination  of expo-
sures.
     Le Bouffant et al.  (1984, 1987) examined the clearance rate of rock  woo"
fibers in groups  of 24 male  and 24  female  Wistar rats exposed to  fibers  cf
respirable mass  concentrations (5  mg/m )  and size  (<10 \tm long).  After 12  and
20 months of exposure,  most of the inhaled rock wool  fibers were found in  the
lungs and lymph  nodes.   After 24 months, relatively high fiber concentrations
were also  reported for the trachea,  possibly reflecting increased bronchial
clearance.   The  authors  reported that clearance of rock  wool  fibers was fairly
rapid and  noted that 76  percent of  the  fibers  were removed  from the lungs
within 7 months a*ter exposure  was  terminated.   Nearly all the  fibers t-et
pe-sistei were  less than 5 urr.  long;   the  shortest  fibers were DC-serve:  ir-  *"£

-------
lymph  nodes.   Fibers that were  retained  had  lost  some opacity and appeared
eroded.  In addition, many of the fragmented and partially dissolved  fibers  had
a reduced sodium, calcium, and magnesium content.
     Morgan and Holmes (1984b) studied the solubility of sized rock wool  fibers
in the lungs of rats and the formation of pseudoasbestos bodies in the  lungs  of
Syrian Golden  hamsters  following intratracheal  instillation (see also  Section
7.2.7).  The rock wool used had a CML (± GSD) of about 27 ± 2.2 urn and  a  CMD  of
about 1.1 ± 2.2 urn.  The amounts of rock wool administered to rats and  hamsters
were 0.5 and  0.4 mg, respectively.   The rats were  of the  HKT  strain,  Alderly
Park, SPF derived,  and  the hamsters were of  the  outbred DB and DSNO strain.
Rats were killed at various intervals up to  18 months after dosing,  and the
CKDs of rock wool fibers recovered from the lung were determined by a TEK a-ic a
light microscope.   Although  no significant  change was  observed  in the OKI of
fibers during  the  18-month period,  the  fibers were  apparently  becoming thinner
at tneir ends  than in the middle,  indicating slow  dissolution of the  fibers.
Nc effects  on  the  body  weight  and relative  lung weights of the rats were  nciec
during the  observation  period, and  no  histopathologic changes  were observe:  in
stained sections  of the  lungs  of  rats  killed 18 months after instillatior.
However,  only  a  small  fraction of the dose  was accounted for by long fibers
with diameters  less  than 0.5 urn, which are believed  to be the most  likely tc
elicit a pathologic  response.   In hamsters,  the first  signs  of formation of
pseudoasbestos bodies in the lungs  were detected after about  2 months.   0"".y
fibers with less than 2-um  diameters  were  coated,  although the  coating  was
often discontinuous  on  the longer fibers, suggesting  that  formation of pseudc-
asbestos bodies  does not present the ultimate dissolution of rock wool fibers
(Morgan and Holmes,  1984a).   The longer a  fiber, the greater  the probability
that body  formation will  occur  with  time  (Morgan,  1980; Morgan  and Holmes,
1984a) (see also section 7.2.1).

4.2.3  Ceramic Fibers
     There  is  limited  information on the pulmonary deposition, translocstio-.
and clearance of ceramic fibers.  Results of the few available studies indicate
that pulmonary  retention and clearance of ceramic  fibers  is  similar  tc  U.cse
observed for glass  and  mineral wool fibers.  However,  slight differences we-e
noted  in  the  distribution  of  fibers  deposited  within  the  various p-'rc-c-.,
lobes.   In  addition,  the clearance half-life values  for ceraT.ic fiDe-s  »e-e
loncer then for other fibers.

-------
     Nineteen male  rats  (weight,  age,  and strain not specified) were exposed
by  nose  only to  a  dust cloud,  which  contained an average of 709  ceramic
fibers/ml, 6  hours/day  for 5 days  (Rowhani and  Hammad, 1984).  The CMD and CK.
of  the airborne  fibers  were 0.53 and 3.7 pm, respectively.  The animals were
sacrificed 5  days  after final  exposure  to allow for clearance of about 95 per-
cent  of  the  fibers  deposited in ciliated airways.   (During the same time
period, pulmonary clearance  is less than 5 percent.)   Lungs  were removed, and
lobes were separated, weighed, and digested.   Fibers  deposited  in  the  lungs
had a CML  of  approximately 2.0 ym and  a CMD of 0.35 urn; size distribution o*
ceramic fibers was  similar for all lobes.  The  lobar  fiber burden was weight-
dependent, with the  combined left lung lobes and the  right diaphragmatic  lobe
carrying the greatest relative percent  of the total  lung fiber content (apprc^--
mately 32  to  45  and 22 to 29  percent,  respectively).  The left  lobes  anc tr,e
right apical  lobe appeared to concentrate more of the  longer, wider fibers  tna~.
did the other lobes.  However, these differences were  not statistically signif-
icant (p >0.05).
     In a  study  conducted  by Hammad (1984),  a  group of  49 male  rats  (strain,
age, and weight  not reported) were exposed by nose only to ceramic fiber  dust
clouds (with  an  average  concentration  of 300  fibers/ml)  for  5  consecutive
days.   The mean diameter and length of  the airborne fibers  were 0.7 and 9.C pr,
respectively.   Animals  were  killed  in  groups of seven at  5,  30,  90,  180, and
270 days after the  last day of exposure.  Ceramic fibers were cleared  slowly
from the lungs, with a clearance half-life of 196 days.   Fibers with a diameter
between 0.5 and  0.7 urn  and a  length of <5 urn were  retained  the  longest.   At
270 days postexposure,  approximately 25 percent of the inhaled ceramic fibers
were still present in the lungs.
     The results of  the Hammad (1984)   study suggest that  pulmonary clearance
of ceramic fibers is dependent on fiber length.   Ceramic fibers of all diameters
were cleared at similar rates, but the  clearance rate  increased as fiber length
increased to  approximately 11 to 15 pm.  Short  fibers  (i.e.,  those less then
5 urn  in  length)  showed the  highest  retention  in the  rat lung for  all  time
periods.
     Similar  results  were reported  by  Davis et al.  (1984).    Very  few lone
ceramic fibers were  retained by lung tissue  2 years after  Wistar rats (nj-^er.
sex,  and  age  not  reported)  had  inhaled  8.4 mg fibers/ir. ,   7 hours/day.
5 days/week for 224 days, despite a high concentration of alurinur silicate.

-------
     Pigott et al.  (1981)  and Pigott and Ishmael  (1981) also investigated  the
migration of  ceramic  fibers  following  inhalation or  intraperitoneal injecticr,.
In the  study  conducted by Pigott et al. (1981), groups  of  40 adult Aider!ey
Park, Wistar-derived  rats (20/sex)  inhaled  ceramic  fiber  dust clouds for a
minimum of  6  hours/day,  5 days/week, for 86  weeks.   The  exact  fiber concentra-
tion of the dust clouds was  not  noted,  but  an estimated average exposure  of
2.18 mg/m   and  a total cumulative exposure  of  approximately  7000 mg/m   were
reported.    Two animals of each sex were killed  at weeks  14  and 27, one of eac*--
sex was killed at week 53, and the remaining rats  were allowed to survive u-t: >
"natural" death  occurred  or  until 85 percent mortality  was  reached within a
group.   Fragmented fibers were found in the lungs  of most rats, and the  majesty
of the  ceramic  fibers recovered  from  rat  lung tissue were relatively s~crt
compared to those  inhaled.   No other details on the size  of retained fibers
were given.   At  all  times after  inhalation, fibers  were detected in  alveolar
macrophages and in the mediastinal lymph nodes.   The authors suggested that the
fibers may have been transported via macrophages.   The second study (Picctt en:
Ishmael, 1951) showed  that at 3, 6,  and 12 months  after groups of 12  male end
12 female Wistar rats  (140 to 160 g) were given an intraperitoneal iniecfic- c*
20 mg (0.2  ml of  a 10 percent suspension) of  ceramic fibers,  fibers  co^lc  be
recovered in the sternal and mesenteric lymph nodes as well  as in the  peritoneal
cavity.   The  data  indicate that  ceramic fibers  can migrate  from  the peritoneum
to the lymph nodes.
4.3  SYNTHETIC FIBERS
4.3.1  Aramid Fibers
     Information on  the  pulmonary deposition  and  clearance  of  aramid  fibers  is
very limited.   In  one study, rats (strain, number,  sex,  age,  and weight not
given) were given a single intratracheal instillation of 25 mg Kevlar  (aramid)
polymer dust  in  physiologic  saline (Reinhardt,  1980).   Animals were  sacrificed
at 2,  7,  and 49 days and  3,  6,  12,  and 21 months after treatment.   Particles
were detected  in lung tissue at  all  sacrifice  times.   Large  fibers  (diameter
100 to 150 um)  concentrated  primarily in  the terminal  bronchioles,  and  small
particles (diameter 5 um) were recovered in alveolar ducts.
     In another  study,   fiber  deposition and macrophage  response were cose-
related following inhalation by groups of  five male Spragje-Dawle\ rats e>.;cse:
tc 0.1, 0.5,  3.0, or 16 me ultrafine Kevlar'1" fiDers/ir.l, 6 news'ca..

-------
of 10  days  (Lee  et al.,  1983).   Approximately 85, 87, 24, and 45 percent of
these  fibers,  respectively had  an  aerodynamic diameter <5 pm.   In animals
inhaling  3.0  or 18 mg/mL,  fiber-containing  macrophages  accumulated in the
respiratory bronchioles  and alveolar region  immediately  after  exposure  anc
were still present at 2 weeks postinhalation.   Fibers  appeared to be fragmer.te:
and markedly  shorter  within 3 months'  time.   Some dust  cells containing short
fibers (i.e.,  less  than  2 \im long)  had  migrated  to the  peribronchial lymphoid
tissue or tracheobronchial lymph nodes  by 6 months after exposure.  The results
suggest that following phagocytosis, short aramid fibers are transported to  the
lymph system.   In addition, long Kevlar   fibers appear to  be easily and rea:v>y
dissolved in  the  lung, a process which  may facilitate their clearance from tne
body.

4.3.2  Carbon Fibers
     Phagocytosis  occurs  within  hours  after inhalation  of carbon fibers  ar.d
continues for  2  or more  years.   The nonfibrous  dust  is primarily cellule"  •>-•
rr.a:rcpr,ages,  and  the  number  of  particles  in each macrophage  increases wif;
time after exposure.   Human subjects  exposed to   carbon fibers by  inhalatic*
for about 3 minutes retained approximately half of the inhaled carbon dust.
     Holt and  Home  (1978) and  Holt (1982) examined the phagocytosis of carbon
fibers by macrophages at various periods following inhalation exposure.  Groups
of guinea pigs  (strain,  sex,  weight, and age  not specified) were exposed to an
aerosol dust  cloud  containing 370  carbon fibers/mL,  2.9  black fibers/ml, and
1.6 transparent  fibers/ml.   The  transparent  fibers were believed to be glass
fibers.  Carbon fibers had diameters ranging from submicron to several  microns,
black  fibers  were  approximately 1.0 to 2.5 pm wide and no longer than 15 pm,
and transparent  fibers typically had a diameter  of 1.5 pm and  lengths  up  to
30 urn.   The dust  also contained fibers  that  were above respirable range (i.e.,
10 urn wide and greater than 100 urn  long).  Twelve guinea  pigs inhaled the dust
for a total of 100 to 104 hours and were then sacrificed between 6  and  144  days
postexposure.   Three  more  animals,  exposed for 7, 23, or  62 hours, were killed
at 1,  2, and 12 days after inhalation,  respectively.
     Tissue sections  from  the guinea pig killed  1 day after  a  7-hour expos-re
to the  fiber  aerosol  contained  dust-filled macrophages, suggesting that phago-
cytosis had occurred within hours after inhalation (Holt and  Home,  1S7£).   The
highest concentration  of dust-containing cells generally was  foun: ir.  tissues

-------
obtained later  in  the  study,  with cell numbers  peaking  IDS  and  144 days post-
inhalation.   Phagocytosis  therefore appeared to continue for several months
following exposure.  In  a  similar study (Holt,  1982), the presence of fiber-
containing macrophages at 2 years after inhalation  indicated that engulfme-t  of
carbon dust particles  may  proceed for  an extensive period of time.   Nonfibrous
dust was primarily intracellular in macrophages;  these  cells  were frequently
found  attached  to  the alveolar wall or embedded in  tissue  (Holt  and Home,
1978).   All fibrous  particles  longer than  about 5  or 6  urn were  extracellular,
even at 140 days  postinhalation.   Lung tissue from  all  animals contained  an
                                                            2
average of  two to  four  short  (4 to 10 urn) carbon fibers/cm  section;  long
fibers (those  greater than 10 urn) were recovered  only  in  four guinea  pigs
exposed tc the  dust cloud  for 104  hours.  However,  the authors reported that
through tissue  processing, extracellular  particles  were  washed out  of lur.c
sections  of animals  killed shortly after exposure.   Thus, the degree of phag:-
cytosis,  retention,  and number  of carbon  fibers  could not be accurately
estirsted.
     No unccated transparent  fibers were found  in  any of  the guinea pig lun;s.
although  one or two of these  presumed glass  fibers -were  covered with a thick  •
ferruginous coating  (Hclt  and Horn*,  1978).   Carbon fibers did not  appear
coated.  In lung tissue sections  obtained toward the  end of  the  study,  a slight
thickening of  the  interalveolar  tissue was  observed.   However,  at 27 weeks
postexposure,  sections  showed no  abnormal reticulin or collagen,  and there
appeared  to be no fibrcsis.
     In a followup study (Holt,  1982), 27 guinea  pigs  (age, sex, weight,   and
strain not reported) were  exposed to 370 carbon fibers/mL  for  100 hours and
were sacrificed at various times between 1 and 270  days  after termination of
exposure.   Minor  contaminants,  comprising a  total of  1.2  percent of cell
fibers, included black, and transparent fibers.   Two guinea pigs inhaled the
carbon fibers for  7 and 12 hours, respectively, and  were  killed on the follow-
ing day.   Three additional animals, exposed  for 10  hours each, were killed
after 420, 440, and 470 days.   Lungs were sectioned and  stained, and  macrophages
containing fibers and/or dust  were counted.  Very few extracellular fibers  were
seen,   and the  macrophages  that were recovered 1 day  after inhalation contained
only a few dust particles.   The number of particles in each  macrophage  increased
with time; after 1 year, macrophages containing high concentrations  of  canton
dust were common,  even in  animals exposed to the fiber  aerosol  cloud fc- cr-_.

J_V 195£                            4-22          DRA"7--DD  N""  CJI":  C?. CI'E

-------
                                                           2
10  hours.   There were  about 20 dust-filled macrophages/mm  lung tissue at
440 days  postinhalation;  that number doubled and  then  dropped  to less than
20/mm   within  an additional  120  days.   Engulfment of erythrocytes  was else
evident throughout  the  study.   Dust-laden cells accumulated near bronchi and
blood vessels,  and  some cells that  penetrated the  bronchiolar walls were founc
in  the  sputum.   Inhaled submicron-sized carbon  dust also reached the  alveoli.
No  pathology or fibrosis attributable to  carbon dust  or  fibers was observed at
any time.
     The  tracheobronchial  clearance of  carbon dust particles in  eight  hurr.a-'
subjects  was  investigated by  Camner et al.  (1973).   The subjects inhaled a
test aerosol of 6-um Teflon particles   (density 2  mg/mL,  aerodynamic  diameter
approximately  8 to  9 urn)  tagged  with technetium  Tc 99m, and then  a  carbo-
dust aerosol (concentration  50 mg/mL; mass median diameter 11 urn).    Sutje:ts
"inhaled each of the test materials for approximately 3  minutes;  one  g-oup cf
four inhaled the  carbon dust immediately  after  inhalation of the  test aeroscl,
and a second group  waited 1  hour before  inhaling  the carbon fibers.   Recic-
ac*.••vity  was  monitored  for  2 hours after  exposure  to  Teflon.   Two  cf  the
individuals smoked 10 to 15 cigarettes/day; the  others were nonsmokers.   Slight
irritation in  the  airways  was noted following exposure  to carbon fibers cr,\v.
Subjects  retained an average of 67.2 percent of  the  inhaled Teflon pa-tides
and an  average of  52.0 percent of the  inhaled carbon  dust  (p  <0.05).   The
authors concluded that  since inhalation  of carbon  dust followed  that of Teflon,
the faster clearance rate for carbon indicates a faster mucociliary  transport.

4.3.3  Pol^olefin Fibers
     There was  no  information  in  the available literature on the pharme:o-
kinetics and/or translocation of polyolefin fibers.
4.4  MINERALS
4.4.1  Erionite
     Very little  information  is available on the deposition  and  clearance  cf
erionite fibers,  although  data from several epidemiological  studies  indicate
that  erionite penetrated  the pleura!  and  parenchyma! tissues  in TU-K-IST.
villagers exposed  to  low levels (Ban's et al.,  1976;  Sebastien et al.,  15:1;
Rohl et al.,  1982.   Fibers  recovered from the lungs of these indix:S duals *e-e
                                    4-23

-------
uncoated  and  had a  mean diameter of 0.3 pm and  a  maximum length of 9 pm.
Sebastien  et  al.  (1981)  also detected in these individuals  some  ferruginous
bodies that contained  erionite  fibers (zeolite bodies).   In a study  conducted
by  Moatamed  et al.   (1981),  erionite  fibers  were  distributed throughout rat
lungs following  intratracheal  instillation.  Macrophages,  giant cell-granulo-
matous reactions,  and ferruginous bodies  were present within  1  week  after
instillation.   No other  information  on  the pharmacokinetics or biodispositicn
of erionite fibers was found.

4.4.2  Attapulgite
     Little information  is  available on  the deposition, translocation, anc
clearance  cf  attapulgite.   Churg (1983)  determined  the total nonasbestes
mineral  content  in  the lungs of 20 subjects  who had no history  of occupational
dust exposure.   All  subjects (11 men  and 9 women) were selected from a  general
autopsy service.  All  were aged 40 years or older at the  time  of death.  The
average number of nonasbestos fibers  was 106 x  10   ± 87 x 10  fibers/g  wet  IL~C.
with nc significant differences noted between men  (96 x 10  fibers/g) and wore-:
(115 x 10  fibers/g).  Similarly, no significant differences were noted between
the 15 smokers  (S3  x 10   fibers/g) and 5  nonsmokers  (140  x 10' fibers/g).  A
total  of  13 different  mineral groups other  than  asbestos  were  identified,  of
which apatite,  gypsum,  talc, silica,  and attapulgite accounted for  more than
half the  fibers  (Table 4-1).  Silica  was found in the  lungs of  all 20  subjects
and talc  in IS  of 20.   About 86 percent of all fibers were shorter than 5  prr,
and most  had  aspect ratios  less  than 15.   In addition, the  mean number cf
nonasbestos fibers  was  similar  to the number  of  asbestos  fibers  (1C2  x  10  /g
wet lung)  found  in  these subjects, with chrysotile accounting  for  more than
80 percent of the asbestos fibers (Churg,  1983; Churg and Warnock, 1980).
     The  results  of two case studies suggest  that  inhaled attapulgite  fibers
can penetrate  into  the alveolar spaces  and  that ingested attapulgite is  trans-
ported to the kidney and eliminated in the urine.   In the first case, a 41-year-
old miner with lung  fibrosis had been exposed to attapulgite dusts for 3 years.
Approximately 42,000 fibers/ml  were  found  in lung washing  fluid  recovered  by.
bronchoalveolar  lavage  (Bignon  et al.,  1980).   The average  attapulgite  fine*-
was 1.5 pm long  and 0.11 urn wide.   An  average of  300,OOC fibers/ml  were
recovered from the urine of  a 60-year-old woman who had been treated creTU  for
6 months  with  a drug containing  attapulgite (Eignon  et  al., 19EI).  Nc  ctne-
details were given.

-------
             TABLE 4-1.   THE NUMBER AND SIZE  OF  NONASBESTOS FIBERS
                          IN THE LUNGS OF 20  SUBJECTS
Mineral
Apatite
Talc
Attapulgite
Gypsum
Silica
Ctherb
Not identified
Total fibers8
(x 10s)
19.0 t 47
17.0 1 25
8.4 ± 13
7.0. t 23
6.0 1 6.4
18.4
31.0 1 58
Percent
of total
18.0
16.0
7.9
6.6
5.7
17.0
29.0
Length,
1-4.9
89.0
71.0
100.0
54.0
78.0
88.6
87.0
um, expressed
5-9.9
4.9
23.0
0
33.0
14.0
10.1
7.1
as %
>10
6.1
6.0
0
13.0
e. o
1.3
5.9
 Mean value ± standard deviation.
 Includes rutile, kaolinite, mullite, illite,  pyroxene,  pyrophyllite,
 feldspar, vermiculite.
Source:   Churg (1953).

     Preliminary results on  the  absorption of attapulgite from the gastroin-
testinal tract of  rats  (sex and strain not specified)  were  reported by Masse
et al.   (1980).   Absorption was determined by collecting the lymph fluid at
various times after  oral  administration of 10 or 100 mg attapulgite  to croups
of 20 rats  and  counting the fibers  by  TEM.   No fibers  were found at the  low
dose.   Systemic  absorption of  attapulgite (100 mg) was  much less  than that  cf
asbestos.  No other data on the pharmacokinetics of attapulgite were  availstle.
4.4.3  Talc
     The  pulmonary  deposition of  talc following repeated exposure is dose-
dependent.   Following single  inhalation  exposures,  -pulmonary  clearance  is
slow with  a  biological  half-life of 7 to  10 days.  Talc  is  cleared  by mucoci1,-
iary action from the lungs and eliminated in the feces via the gastrointestinal
tract.
     The pulmonary  deposition,  translocation,  and clearance of  inhaled tele  in
Syrian Golden  hamsters  were studied by Wehner  et al.  (1977b).   A group  of  L-.
10-week-old  female  hamsters were  exposed by nose only  to  nejtrcr.-acf'vstec

-------
talc  aerosols  for 2 hours.  The  mean  aerosol  concentrations  during the t*c
sampling periods (from 15 to 30 and 60 to 90 minutes during the  2-hpur  exposure
period)  were 40  and  75 pg/L.   After  exposure,  the urine  and  feces were
collected.  Groups of four animals were killed at various times,  and the lungs,
liver, kidneys, ovaries, and skinned carcass were collected and  radioassayed  by
6-ray spectrometry.  The  counts  were compared with those  obtained  from nine
unexposed  control  hamsters.   Considerable variations in pulmonary  deposition
among animals  killed at the same  period  were  attributable to variations in
aerosol   concentration at  the  different tier levels for animal exposure and  to
differences  in  individual  breathing  patterns.   It was estimated that  animals
placed on  tier  levels  2 to 4  of  the exposure  chamber inhaled an average of
360 pg talc, whereas  those on tier  1  inhaled  an  average  of 570  pg.  Nonethe-
less, the  data  indicated  high pulmonary talc  levels during the  first  4 d=.. s
after exposure, with  mean total  measurements (±  SD)  of 33.1 ± 31 pg and 21. C :
15.3 pg  at  15  minutes  and 4 days  after exposure,  respectively.   Thereafte-.
t=": cc-'tent in  the  lungs dropped gradually with  mean  values  (3.7 ± l.£ pc)
beinc similar  to  those  found  in controls  (2.3 i  0.4 pg)  after 4 mcr.ths.  T^e
biologies1  half-life of  the talc  deposited in  the  lung was reported tc  be 7  to
10 days,  and pulmonary  clearance was complete within  approximately 4  mr^t-.s
after exposure.
     No  significant  difference was  observed among  mean  radioactivity  levels
found in the  liver,  kidneys,  and ovaries  of  exposed and control  animals.
Radioactive  residues  in  carcasses  were  highest  4  hours  after exposure
(455 ± 338 ug)  and decreased  gradually to control  levels  (about  3.5 pg)  by day
8.  This was attributed to talc deposited on or about the nose and nares d'j-inc
exposure  as  well  as  to talc passing through the  gastrointestinal tract  after
mucociliary clearance.  Fecal  samples  contained  relatively high quantities  of
talc.  The mean talc equivalent 21 hours after exposure was 106  ± 53 pg; by  day
4, this  value  had increased by 3.5-fold (350 ± 292 pg).   The  presence  of talc
in the feces was  attributed to mucociliary clearance  from the  lungs probably
together  with  a  fraction  from licking the nose, lips, etc.  Low  levels of
radioactivity found in the urine were attributed to leached   Cc being  absorbed
in the gastrointestinal tract.
     Hanson et al.  (1985)  developed  a method to  determine the  lung burden of
talc  in  rats and mice  after  repeated  inhalation exposure to te'c  eerescls.
This method was based  on  the  analysis of acid-insoluble  Kg  cc-te-.t ip *h£
lungs by  flame  adsorption spectrometry.   Me was  selectee  be:e_se  'it e::c--ts

-------
for 19.2 percent  of  the composition of talc, and most  of  the  endogenous  Mg
present in  the  lung  can be removed by dilute acids.  Whole-body exposures of
male and female  F344 rats  and  BBCSF^ mice were performed 6 hours/day, 5 days a
week for a  total  of  20 exposures.   The mean  lung burdens in rats were 77,  1£7,
and 806 pg talc/g lung tissue at exposure levels  of  2.3, 4.3, and 17  mg talc/rr  ,
respectively.  In mice,  mean  lung burdens were 114, 325, and  1150 ug talc/g
lung tissue  at  exposure levels of 2.2, 6.3,  and 20.6 mg talc/m , respectively.
The higher  contents  of talc  in mouse lungs as compared to rats were related to
a difference in initial deposition.
     In two  studies  with humans (Churg,  1983; Churg and Warnock, 1980), tei:
comprised approximately  16.0 percent of  the  nonasbestos fibers recovered  frcr
the lungs of  20  individuals  with  no history  of occupational dust exposure (see
section 4.4.2 for more details).
     Several studies  have  been conducted to  determine the metabolism of te^c
powder following  oral  administration.   Little,  if  any,  is absorbed  from  the
gastrointestinal tract  following  oral  intubation  of rats,  guinea pigs, mice.
enfi hemjte^s;  most  of the material is  eliminated  in the feces within  1  or
2 days after dosing.
     Studies on  the  absorption and disposition of   H-labeled  talc  folio*: re
single and  repeated  oral  administration  to rats, mice, and guinea  pigs were
conducted by Phillips  et al.  (1978).   [3H]Talc (specific activity 100 pCi/g)
was administered  by  oral  intubation as a suspension in  aqueous glycerol jelly
solution (10 mg/mL,  1 uCi/mL).  Groups of three male Wistar  rats  and three
female Duncan  Hartley guinea  pigs were  administered a single oral  dose  of
[ H]talc at levels  of 50  and  25  mg/kg,  respectively.   Urine  and feces were
collected at 24-hour intervals for 4  days and on day  10.   The animals were
then killed, and liver, kidney, and gastrointestinal tract  tissues  were  removed
for radioassay.   A similar experiment  was conducted  with three  male Wistar rets
given daily oral doses of 50 mg [ H]ta1cAg for 6 days;  these  animals also were
killed 10 days after the last dose.  Approximately 95.8  and 94.4 percent of  the
administered radioactivity was excreted  in the feces of single-dose rats and
guinea pigs, respectively, within 96 hours after compound  ingestion; less than
2 percent was found in the urine, and only 0.2 percent was  found in the  gastro-
intestinal   tract.   The presence  of radiolabeled material  in  the  urine was
attributed  to labile  H associated with  tritium-labeled talc and to  contar.-:na-
tion of the urine.   No radioactivity was  found  in  the  liver or  kidneys e*te'
                                    4-27

-------
 10 days.  Similarly,  no radioactivity was  eliminated  in  the  feces  by  day 10.  No
  H was found  in the  livers  of  rats  given six daily doses of [ H]talc, although
 less than 0.02 percent of the total  dose  was  detected in  the  kidneys.
      Phillips et al.   (1978)  also administered a single oral dose (40 mg/kc) of
 radiolabeled talc to  four female LACA mice; two animals were killed at 6 hours
 and two at 24 hours  postdosing.  At  both  sacrifice times, all of  the radioacti-
 vity was  recovered  in the  gastrointestinal  tract and  feces,  and  none was
 detected in the carcass.
      Similar results were reported  on  the  absorption of  talc by  female Syria-
 Golden hamsters (Wehner et. al., 1977a).   The  talc  used  was  a  neutron-activate:.
 commercially available  baby powder.  The  talc  was  suspended in physiologic
.saline, and  1-mL  aliquots  of  the   suspension were  administered by gastric
 intubation to  six 10-week-old  hamsters.   An  additional  group of  four hairs ters
 received an  identical  suspension of nonirradiated talc  to determine if the
 ta"lc had a  cathartic  effect.   The urine and feces were collected for 24 hou^s.
 The hamsters were  then killed,  and  the  skinned  carcass,  gastrointestinal  tract.
 lungs,  liver,  and kidneys were collected and radioassayed  for    Co  and  '"r- "•••
                                                                           - W  k.
 6-ray spectrometry.   About  74.5  percent  of the talc was -found  in the feces,
 23.5 percent in the  gut,  and  1.9 percent  in the carcass.  No  radioactivity was
 found in the lungs.   Detectable  amounts  were found  in  the  liver  and  kidneys,
 but these were not statistically different from control values.   Approximately
 0.07 percent of the  total  talc was  found  in the urine,  but  that was attributed
 to leaching of    Co from talc.
      Wehner and Weller  (1986) examined  in monkeys the  migration  of talc frorr,
 the vagina.  Six  female exbreeder  cynomolgus  monkeys each  received 30 intra-
 vaginal  talc applications,  consisting of  125  mg neutron-activated powder per
 application, over  a  45-day period.   Doses were  equivalent to 29 to 52 rr.g
 talc/kg/day, depending  on  weight.   Six  untreated  females served  as controls.
 Animals  were between  4  and 12 years old  and  weighed between 2.4  to 4.35  kg.
 The  radioisotopes    Sc,   Co,   Fe, and   Cr were used as  tracers  for the
 administered talc.   Two days  after the  final application,  abdominal lavaoe
 fluid, ovaries, oviducts,  uterus,  and  vagina  with cervix were  collected  for
 y-ray analysis.   Only  the  vagina  with  cervix  samples  from  dosed  animals
 contained talc tracers.   The  authors concluded that despite repeated applica-
 tions to  the  fornix of cynomolgus  monkeys,  no measurable  quantities (i.e.,
 greater  than 0.5  ug)  of talc translocated to  the uterus  or beycic1.

-------
     In contrast, Henderson .et  al.  (1986)  reported that  talc  particles  intro-
duced  into the  uterus  and vagina of the rat  migrated  to the  ovaries.   Eight
female exbreeder  (7.5  months  old) Sprague-Dawley rats  each received an  intra-
vaginal dose  of  250 uL of a talc-saline suspension (100  mg talc/ml).  Four  cf
the rats received additional uterine instillations on days 6 and 15, and twc  of
these rats were treated again on days 22 and 30.   Ovaries were removed from  all
animals at the  time of sacrifice (days 50, 20,  and 49,  respectively).   In  a
separate experiment,  six  Sprague-Dawley exbreeders were given  250  uL of the
same talc suspension,  and an additional six  rats  received the.saline vehicle
only.   Two animals  from each of  these  groups were sacrificed at 24, 49, and
96 hours after  treatment;  ovaries of were removed  as  before.   Talc particles
were recovered  from the ovaries of  all  animals given intrauterine doses  cf  t~e
test material and frorr; the two  rats that were killed 96  hours after intre.a;m'-
nal instillation.   No  talc was  found in the  ovaries of  the other treate: or
nontreated animals.

<.%.£  Venr.iculite
     There was  no  available  information  on  the  metabolism  of  vermiculite
following oral  administration or inhalation exposure.  However, vermiculite  is
not expected  to  be  absorbed frorr; the gastrointestinal  tract to any  appreciable
extent.
4.5  REFERENCES

Asgharian, B. ;  Yu,  C.  P.  (1988) Deposition of inhaled fibrous  particles  in  the
     human lung. J. Aerosol Med. 1: 37-50.
Baris,  Y.  I.; Sahin, A.  A.;  Ozesmi,  M.; Kerse, I.;  Ozen,  E. ; Kolacan,  E. ;
     Altinors, M. ; Goktepeli, A. (1978) An outbreak of pleura! mesotheliorrs anc
     chronic  fibrosing  pleurisy in the village of Karain/Uerguep in Anatolia.
     Thorax 33: 181-192.
Bellmann, B.;  Muhle,  H.;  Pott, F.; Konig,  H.;  Kloppel,  H.;  Spumy,  K.  (1957)
     Persistence of man-made  mineral  fibres (MMMF) and asbestos in rat lungs.
     Ann. Occup. Hyg.  31:  693-709.
Bernstein, D.  M.;  Drew, R.  T.; Kuschner, M.  (1980) Experimental  a?proa:^eE  for
     exposure to sized  glass  fibers. EHP  Environ. Health Perspect. 34: 47-5".
     1985                           *'29          D*ACT--DO  NT  Q.C": -': -•

-------
Bignon,  J.; Sebastien,  P.; Gaudichet, A.; Jaurand, M.  C.  (1980) Biological
      effects  of attapulgite.  In:  Wagner, J. C.;  Davis,  W.,  eds. Biological
      effects  of mineral  fibres « Effects biologiques  des  fibres minerales:
      proceedings  of a  symposium,  v. 1;  September  1979;  Lyon,  France.  Lyon,
      France:  World Health Organization,  International  Agency  for Research on
      Cancer;  pp.  163-181. (IARC scientific publication no.  30; INSERM symposia
      series:  v. 92).

.Camner,  P.;  Helstrom,  P. A.;  Philipson,  K.  (1973)  Carbon dust  and tnucociliary
      transport. Arch. Environ.  Health  26:  294-296.

Churg,  A.  (1983)  Nonasbestos pulmonary mineral  fibers  in the general popula-
      tion.  Environ.  Res.  31:  189-200.

Churg, A.; Warnock,  M.  L.  (1980) Asbestos fibers in the general  population.  An.
      Rev. Respir.  Dis.  122:  669-678.

Cook,  P.  K. ;  Olson, G.  F. (1979) Ingested mineral  fibers: elimination in  t«- = -
      urine. Science  (Washington, DC) 204:  195-198.

Davis,  J.  M.  G. ;  Addison,  J. ; Bolton, R. E.;  Donaldson, K. ; Jones, A.  D. ;
      Wright,  A.  (1984)  The pathogenic effects  of  fibrous ceramic  aluminum
      silicate.glass  administered to  rats  by  inhalation  or peritoneal, injection.
      Jn: Guthe, T.,  ed.  Biological effects of man-made  mineral  fibres:  prcceec-
      ings c^  a  WHO/IARC  conference  in association with  JEMRE and TIMA,  v.  2:
      Ap^il  1962;  Copenhagen, Denmark.  Copenhagen,  Denmark:  World Health  Orcc--
      ization, Regional  office for  Europe; pp. 3G3-322.

Evans,  J.  C. ;  Evans, R.  J. ;  Holmes, A.;  Hounam, R. F.;  Jones, D. M.; Morgan,
      A.; Walsh, M.  (1973) Studies  on the  deposition of  inhaled fibrous  materiel
      in  the  respiratory tract  of  the  rat and its  subsequent  clearance  using
      radioactive tracer techniques:  1.  UICC  crocidolite asbestos.  Environ.  Res.
      6:  180-201.

Friedberg,  K.  D. ;  Ullmer, S. (1984) Studies on  the  elimination of dust of MM!-'?
      from  the rat  lung.  In: Biological  effects of man-made  mineral fibers:
      proceedings of a  WHO/IARC conference,  v.  2; April; Copenhagen, Denma-k.
      Geneva,  Switzerland:  World  Health  Organization,   Regional  Office  for
      Europe; pp. 18-26.

Griffis, L. C. ; Henderson,  T.  R.;  Pickrell,  J.  A.  (1981) A  method for determin-
      ing glass  in  rat  lung  after  exposure to a glass  fiber aerosol. Am.  IncV
      Hyg. Assoc. J.  42:  566-569.

Griffis, L. C. ; Pickrell, J.  A.; Carpenter,  R.  L.;  Wolf, R.  K.;  McAllen,  S.  J. ;
      Yerkes,  K.  L.  (1983)  Deposition  of  crocidolite asbestos  and glass  micrc-
      fibers inhaled  by  the  beagle  dog.  Am. Ind.  Hyg. Assoc.  J.  44:  216-222.

Gross,  P.  (1981)  Consideration of  the aerodynamic equivalent  diameter  of
      respirable mineral  fibers. Am.  Ind.  Hyg. Assoc. J. 42:  449-452.

Hammac1,  Y.  Y. (1984) Deposition and  elimination of MMMF.  In:  Biologies!  e*?'e:ti
      of  man-made  mineral fibers:   proceedings of a  WHO/IARC conference,  v.  2:
      April;  Copenhagen,  Denmark.  Geneva,  Switzerland:  World Heeltr.  0'-c=";:c-
      tion,  Regional  Office  for  Europe; pp. 126-142.

-------
Hammad, Y. ;  Diem,  0.;  Craighead, J.; Weil!,. H.  (19B2)  Deposition of inhaled
     man-made mineral  fibres  in the lungs  of rats.  In:  Walton,  W.,  ed.  Inhalec1
     particles  V:  proceedings  of  an  international symposium;  August 19SD;
     Cardiff, Wales, United Kingdom. Ann. Occup. Hyg.  26:  179-187.

Hanson, R.  L.;  Benson, J.  M. ; Henderson, T. R.; Carpenter,  R.  L.;  Pickrell,  J.
     A.;  Brown,  S.  C.  (1985)  Method for determining the lung  burden of  talc  in
     rats  and mice after  inhalation  exposure  to  talc  aerosols.  J.  Appl.
     Toxicol. 5: 283-287.

Henderson, W. J.;  Hamilton,  T. C.; Baylis, M.  S.;  Pierrepoint, C. G. ; Grif-
     fiths,  K.  (1986) The  demonstration of the migration  of  talc from the
     vagina  and posterior uterus  to  the  ovary in the rat.  Environ. Res.
     40: 247-250.

Holt,  P.  F.  (1982) Submicron  carbon dust in the guinea  pig  lung.  Environ.  Res.
     28: 434-442.

Holt, P. F. ; Home, M. (1978)  Dust  from  carbon  fibre.  Environ. Res. 17: 27t-2£2,

Johnson, N.  F. ;  Griffiths,  D.  M. ;  Hill, F.  J.  (1984)  Size distribution follow-
     ing  long-term inhalation  of  MMMF. In: Biological effects of men-made
     mineral  fibers:   proceedings  of  a  WHO/1ARC conference,   v.  2;  April;
     Copenhagen,  Denmark.  Geneva,  Switzerland: World  Health Organization,
     Regional Office for Europe; pp. 102-125.

Johnson, N.  F. ;  Ortiz, L.  W.  ;  Yerwood,  A.; Smith,  D. M.  (1985)  Analysis  c'
     recovered man-made  fibers from long-term  inhalation  studies.  Toxicoloc-ist
     6: 63.

Le  Bouffant,  L.;  Henin,  J.  P.;  Martin, J. C.; Normand,  C.; Tichoux, G. ;
     Trolard, E.  (1984)  Distribution of inhaled MMMF in  the  rat  lung - long
     term  effects.  In:  Biological  effects  of  man-made mineral fibers: proceed-
     ings  of  a  WHO/IARC conference, v.  2; April; Copenhagen, Denmark. Geneva,
     Switzerland:  World  Health Organization,  Regional Office  for  Europe;  pp.
     143-169.

Le Bouffant, L.; Daniel, H.; Henin, J.  P.; Martin, J.  C.; Normand,  C.; Tichoj*,
     G.; Trolard,  F.  (1987)  Experimental study on  long-term effects, of inhaled
     MMMF  on the lungs of rats. Ann. Occup. Hyg. 31: 765-790.

Lee, K. P.;  Kelly, D.  P.; Kennedy, G.  L., Jr.  (1983)  Pulmonary  response  tc
     inhaled  Kevlar aramid  synthetic  fibers in rats.  Toxicol.  Appl. Pharmaccl.
     71: 242-253.

Masse,  R.;  Sebastien,  P.;  Monchaux, G.; Bignon, J. (1980)  Experimental demon-
     stration of  the  penetration of asbestos  fibres  into the gastrointestinal
     tract.  In:  Wagner,  J.  C.; Davis,  W.,  eds. Biological  effects of mineral
     fibres = Effets  biologiques des  fibres minerales:  proceedings of a sympc-
     sium,  v.  1;  September  1979;  Lyon, France.  Lyon,  France: World Health
     Organization,  International  Agency for Research on  Cancer;  pp.  321-32E.
     (IARC scientific publication  no. 30.;  INSERM symposia  series:  v. 92).

-------
Moatamed, F. ;  Rom,  W.;  Casey,  K.;  Archer,  V.  (1981)  Early  response to  injected
     fibrous  eriom'te  in the  mouse  peritoneum and  rat  lung.  Respir. Dis.
     123: 26.

Monchaux, G. ;  Bignon,  J.;  Hirsch.  A.; Sebastien, P.  (1982)  Translation of
     mineral  fibres  through the respiratory system  after  injection  into  the
     pleura!  cavity  of  rats.  In: Walton, W.,  ed.  Inhaled particles V:  proceed-
     ings of  an  international  symposium;  August 1980; Cardiff, Wales, United
     Kingdom. Ann. Occup. Hyg.  26:  309-318.

Morgan, A.  (1979) Fiber dimensions:  their significance in the deposition and
     clearance of inhaled  fibrous  dusts.  In:  Lenten,  R.; Dement,  I.  M. ,  eds.
     Dusts and disease:  proceedings of the conference on  occupational expcsu-es
     to fibrous  and  particulate dust  and their  extension into the environment;
     1977; Washington,  DC.  Park Forest South,  IL: Pathotox  Publishers, Inc.;
     pp. 87-96.

Morgan, A. (1960) Effect of length on the clearance of fibres from the lunc a-d
     on body  formation.  In:  Wagner,  J. C.  ; Davis, W., eds. Biological effects
     of mineral  fibres  = Effets biologipues des fibres minerales:  proceedings
     of a symposium,  v.  1;  September  1979; Lyon,  France.  Lyon,  France: Wcrlc
     Health  Organization,   International  Agency  for  Research  on Cancer;
     pp. 32S-335.  (1ARC scientific  publication  no.  30.;   INSERK  syr.pcsia
     series:  v. 92).

Morgan, A.;  Holmes,  A.  (1984a) The deposition of  MMMF in the respiratory  tract
     of the  rat,  their  subsequent  clearance,  solubility in  vivo  and prcte'r
     coating. In: Biological effects of man-made mineral  fibers:  proceedings c*
     a  WHO/1ARC  conference,   v. 2,  April; .Copenhagen,  Denmark.  Geneva,
     Switzerland:  World Health  Organization,  Regional  Office  for  Europe:
     pp. 1-17.

Morgan, A.;  Holmes,  A.  (1984b) Solubility of  rockwool fibres  in  vivo and the
     formation of pseudo-asbestos bodies.  Ann. Occup. Hyg.  28:  307-314.

Morgan, A.; Black, A.; Evans,  N.; Holmes, A.;  Pitchard, J.  N. (1980) Deposition
     of sized glass  fibers  in the respiratory  tract of  the rat.  Ann. Occur.
     Hyg. 23: 353-356.

Morgan, A.;  Holmes,  A.; Davidson,  W.   (1982)  Clearance of  sized glass fibe-s
     from the rat lung and solubility  in vivo. Ann. Occup.  Hyg.  25:   317-321.

Muhle, H. ; Pott,  F.; Bellmann, B.; Takenaka, S.; Ziem, U.  (1987) Inhalation  and
     injection experiments  in  rats to test the carcinogenicity  of MMMF.  Anr.
     Occup.  Hyg.  31: 755-764.

National  Research Council.  (1984)  Asbestiform  fibers: nonoccupational  hea^tf
     risks.  Washington, DC: National Academy Press.

Phillips, J.  C. ;  Young, P. J.; Hardy,  K.; Gangolli,  S. D.   (1978) Studies  c~  t^e
     absorption  and  disposition of 3H-labelled talc in the rat,  mcL'se, guinea-
     pig and  rabbit. Food Cosmet. Toxicol. 16:  161-163.

-------
Pigott, G.  H.;  Ishmael,  J.  (1981) An assessment of  the  fibrogenic  poter.tia1  cf
     two refractory fibers by intraperitoneal injection in rats.  Toxicol.  Lett.
     8: 153-163.

Pigott, G.  H. ;  Gaskell,  B.  A.;  Ishmael,  J.  (1981)  Effects of  long  term inr.eia-
     tion of alumina fibres in rats. Br.  J. Exp. Pathol. 62:  323-331.

Reinhardt,  C.  F.  (1980)  Toxicology of aramid fibers.  In:  Proceedings of the
     National  Workshop  on  Substitutes for  Asbestos;  November.  EPA  report
     no. EPA-560/13-BO-001.

Rohl,  A.  N. ;  Langer,  A.  M.;  Moncure,  G.;  Selikoff,  I.  J.; Fischbein, A.  (15£2)
     Endemic pleura!  disease  associated  with exposure  to  mixed fibrous dust  in
     Turkey. Science (Washington, DC) 216: 518-520.

Rowhani,  F. ; Hammad,  Y.  Y.  (198*)  Lobar  deposition of fiber  in the rst.  A-.
     Ind.  Hyg.  Assoc.  J.  45: 436-439.

Sanchis,  J. ;  Doiovich,  M. ;  Chalmers,  R. ;  Newhouse,  K.  T.  (1971) Reg-ions"
     distribution and lung clearance mechanisms in  smokers and non-smokers.  IT.:
     Walter,, W.  H. , ed.  Inhaled particles III:  proceedings of an interne Vic-.a1
     symposium,  v.  1; September  1970; London,  Umited  Kingdom.  Old Wokir.c.
     Surrey, United Kingdom: Unwin Brothers  Ltd.; pp. 183-191.

Setestien,  P.;  Janson,  X.;  Bonnand, G. ;  Riba,  G. ;  Masse,  R.;  Eignor,, J.  (Ir'rj
     Translocation of  asbestos  fibers through  respiratory tract and gastroin-
     testinal tract according to  fiber type  and size. In:   Lemen, R.; Dement. J.
     K. , eds.  Dusts and  disease:  proceedings of the conference on  occupaticr.e".
     exposures  to  fibrous  and particulate dust and their extension into the
     environment; 1977;  Washington,  DC.  Park Forest South,  IL:  Pathctcx  Pub-
     lishers, Inc.; pp.  65-85.

Sebastier., P.; Gaudichet, A.; Bignon, J.  ; Baris, Y. I.  (1981)  Zeolite  boiies in
     human  lungs from Turkey. Lab.  Invest. 44:  420-425.

Spumy, K.  R.  (1983)  Measurement and  analysis  of  chemically changed  mineral
     fibers  after  experiments  rn vitro   and  u> vivo.   EHP Environ. Health
     Perspect.  51: 343-355.

Spumy, K. R.; Opiela, H.; Schoerman, J.; Weiss, G. (1983) Methods .and investi-
     gations relative  to the stability of mineral  fibers  in vitro  and in vivo.
     Staub Reinhalt.  Luft 44: 169-180.

Timbrel!,  V.  (1965)  The inhalation  of fibrous  dusts.  Ann. N. Y.  Acad.  Sci.
     132:  255-273.

Timbrel!, V.; Bevan, N.  E.; Davies, A. S.; Munday,  0. E.  (1970)  Hollow casts cf
     lungs for experimental purposes.  Nature (London) 225: 97-98.

Wei ton, W.  H.  (1982)  The nature, hazards and assessment  of occupational expo-
     sure to airborne asbestos dust: a review.  Ann. Occup. Hyg.  25(2).

Wehner, A.  P.;  Weller,   R.  E. (1986)  On  talc translation fro- the vecir.e tc
     tne o»'iO'jcts and beyond. Food  Chetn.  Toxicol. 24:  329-I3£.


                                    4-23           DP.APT—::  N"  C>":  ;: ::"E

-------
Wehner, A.  P.;  Tanner, T. M. ; Buschbom,  R.  L.  (1977a) Absorption of ingested
     talc by hamsters. Food Cosmet. Toxicol. 15: 453-455.

Wehner, A.  P.;  Wilkerson, C.  L.;  Cannon,  W.  C.; Buschbom, R.  L.;  Tanner,  T.  K.
     (1977b)  Pulmonary deposition,  translocation  and  clearance  of inhaled
     neutron-activated talc in hamsters. Food Cosmet. Toxicol. 15: 213-224.

Wehner, A. P.; Hall, A. S.; Weller, R. E.; Lepel, E. A.;  Schirmer, R. 1.  (19£r)
     Do particles  translocate from the vagina to the oviducts and beyond? Fooc
     Chem. Toxicol. 23: 367-372.

Wright, G.  W. ;  Kuschner,  M.   (1977) The  influence  of varying lengths of glass
     and  asbestos  fibers  on  tissue response in guinea pigs.  In:  Walton, W. H. .
     ed.  Inhaled Particles IV: v. 2. New York,. NY: Pergamon Press; pp.  4£5-4~i.

-------
             5. ACUTE, SUBCHRONIC, AND CHRONIC TOXICITY IN  ANIMALS
     In this  chapter,  the  available short- and long-term animal  toxicity  data
on manmade mineral  fibers,  synthetic fibers,  and minerals are presented, a—
the  noncarcinogem'c  and systemic  endpoints of toxicity are  discussed.   The
noncarcinogenic effects  detailed  in  this  chapter are  then  summarized  anc
quantified in Chapter 9.
5.1  ACUTE TOXICITY
     Limited information was available on the acute toxicity of  MMi-'r,  synthetic
fioers, and minerals.

5.1.1  Kanmade Mineral Fibers
5.1.1.1   Fibrous Glass.   The ocular,  dermal, and pulmonary effects of  glass
fibers have been  studied  in rabbits,  guinea pigs,  hamsters, and rats.  Gross
et al. (1960) reported observing no ocular irritation in rabbits after exposure
tc finely  milled  glass  flakes.   The flakes, which consisted of  some particles
measuring  400 urn  in diameter  (the  majority being <50 gin), were "thoroughly
dusted" on the eyes of six albino rabbits.  The lids were then either  held  shut
or open  for  1  minute.   No effect was  noted on either the  corneas or  the con-
junctivae.
     The  results of a  study cited by NIOSH (1977) Indicate that fibrous glass
causes dermal irritation  in rabbits and  guinea pigs.  Two  types of unspecified
fibrous  glass,  two types of unspecified competing material,  and two cotton
controls were rubbed  on the skin of 16 rabbits (strain and sex  not  specified)
for 2  minutes each, 5 days/week for 1 month.  The  treated areas consisted of
six depilated patches 4.5 cm in diameter.  Twenty-six guinea pigs were rubcei
at eight  similar  sites  with two fibrous  glass textile materials.  The skin cf
the rabbits exhibited a faint redness, scaling, and superficial  yello* crusting
as a  result  of  the mechanical rubbing of  the  fibers on the skin.   In c^nes

-------
pigs, the dermal  reaction was  similar  reacted  similarly, but with more intense
erythema and crusting.  In another study, blood histamine levels  in  guinea pigs
increased when the skin was rubbed with fibrous glass  (N10SH,  1977).
     The inhalation  of  fiberglass  (diameter 1.2 urn, length <3 urn) was show:; to
result  in  a mild  inflammatory  reaction  in male Sprague-Dawley rats, Syrian
hamsters, and  albino guinea  pigs after 20 days of exposure  (Lee  et  a!.,  197S,
1981).  This  reaction was characterized  by "foamy" dust-laden macrophage and
neutrophil aggregation  in the  bronchiolar and alveolar regions of the lungs.
Also  noted  in the  affected  areas were  proliferation  of cuboidal granular
pneumocytes and the thickening of alveolar walls.   Exposure  was for  6 hours/o=.v.
5 days/week, at a concentration of 420 mg/m  (see  section 5.2.1.1.2).
     No pulmonary  response other.than  minimal  foreign body  reaction was  nctec
a'ter  intratracheal   administration  of glass   fibers  (Gross  et el.,  IScLj.
Twenty-five mg of fine  (1.1 urn  diameter)  or  coarse (<50 urn diameter) glass
fibers were suspended in corn oil and administered to  groups of 15 rats  (strg-'n
and  sex not  specified).   Control animals received either kaolin,  quartz  c'jst,
or  corn  oil  alone.   There was  no  evidence of gross changes  in  the  animals
injected intratracheally  with  the  coarse glass dust.  These particles appea-ecf
to behave like inert foreign material  and  were associated with minimal foreign
body  reaction  without significant collagen  formation.   Similarly,  no gross
changes were seen in  the animals that received the fine dust.
5.1.1.2  Mineral WooJ.   No information on the acute toxicity  of  mineral wool
was found in the available literature.
5.1.1.3  Ceramic  fibers.   No information  on  the  acute  toxicity  of  ceramic
fibers was found in the available literature.

5.1.2  Synthetic Fibers
5.1.2.1  Aramid  Fibers.   The inhalation of aramid  fibers  has been shown to
result in a  mild to moderate foreign body reaction in the lungs  of  rats  after
short-term inhalation exposure.   In  a  group of studies reported  by  Lee  et  a'.
(1983), groups of  20 or 25 male Cr1:CD  rats  were exposed via inhalation to
ultrafine KevlarS  fibers  at concentrations  of 0.1, 0.5, 3.0, or 18 mg./V.
Control animals were exposed to air.   Exposures were conducted for 6 hours/cay.
5  days/week  for 2  weeks.   Mass and fiber concentrations are presented ir,
Table 5-1.   Groups of five rats exposed to 0.1, 0.5,  or 3.0 mg'm  were  k-'He:
after 2-week exposures,  and  the remaining animals  were  observed  for  recede-}
periods cf 2 weeks, or 3 or 6 months postexposure.  Groups of five rets  e^se:
.JL-IV 1S£=                            5-2      DRA'T — DD NC~

-------
         TABLE 5-1.  MASS AND CONCENTRATIONS OF KEVLAR^ FIBERS  USED  IN
                       AN INHALATION STUDY IN MALE RATS
Mean cone.
(mg/m3 ± SD)
0.10 4 0.06
0.52 i 0.14
3.00 ± 0.4
18.20 1 2.8K
17.60 ± 4.4b
Fiber count
(fibers/cm)
1.3
26
280
..a-
...a
% Respirable mass
(diameter <5 ur:,)
65
87
94 .
49
13
 Not determined.
 Commercial Kevlar  fibers.
Source:  Lee et al. (1983).
to IS mg/m   were  killed after 4  days  or 2  weeks  of  exposure, and  the
animals  in  the  group were observed for  recovery  periods  of  1,  3,  or  6  montr;
postexposure.   A  separate group  of rats was  similarly exposed to 1£ mg'r.
co!T;merciaT  Kevlar"  fibers  (2.5 mg/m  of respirable  fibers),  and five  rats  we-e
killed after  the  2-week exposure.   The remaining rats  were subsequently kr-lei
after 3 or 6 months.  All rats were examined grossly; and individual luncs were
examined histologically.
     No  clinical  signs  of toxicity or changes  in body  weight were noted at any
exposure  level  during  the study.   Few  effects were  noted in the lungs after
exposure  to 0.1  or  0.5 mg/m  for  2  weeks.   The architecture   of  the  lungs
appeared  normal after  2 weeks; however, sparse  accumulation of "dust cells,"
described as  alveolar  macrophages containing phagocytized Kevlar  fibers, was
observed.   The  amount  of dust-laden macrophage aggregation appeared to  be  dose
related, but most of the dust cells disappeared by 3 months postexposure.
     Pulmonary  effects  and  dust  cell  accumulation were similar  in rats  expcsec
to 3.0 mg/m  ultrafi'ne  Kevlar or 18 mg/m   commercial  Kevlar fibers,  probsrly
owing to  comparable  levels  of resplrable fibers.  Accumulation of dust-lacer.
macrophages  occurred mainly  at  alveolar duct bifurcations  and resulted in
enlargement  of  the  ridges,  proliferation of fibroblasts,  and inflammatory  cell
infiltration  after  a 2-week exposure.  Some thickening of alveolar ducts, d--e
to proliferation of dust-laden foreign body giant cells and chronic inflammatory
cells, was  also  observed.   Three months after exposure, the aggregation  of
dust-laden  macrophages  in alveolar ducts was  increased.  At this point, the
dust-laden mscrophages appeared to be  filled with shortened  c-  C!-s-.w-1a' f^te-s.
July !=££
5-3
DRA-T--DO fCT

-------
Increased cellularity  of  granulomatous  lesions was also evident.  Six me-ins
after exposure,  slight proliferation of dust-laden macrophages and fibroblasts
was noted in alveolar  ducts but there was no deposition of collagen fiber.
     Pulmonary effects in rats exposed to IB mg/rn  ultrafine Kevlar" fibrils
for  2  weeks were  most severe.  Accumulation  of dust-laden macrophages in
terminal  bronchioles  and  alveolar  ducts of adjacent  alveoli was  observed.
Accumulation was particularly  prominent in  alveolar duct bifurcations, result-
ing in partial blockage of the alveolar ducts.   Fibrotic thickening of alveolar
ducts and proliferation  of  chronic inflammatory  cells  were evident.  The
dust-laden  macrophages appeared to  be  incorporated into  fibroblast  cells.
The thickening of  alveolar ducts decreased progressively  1  to 3 months'afie'
exposure.  However, dense  reticulin fiber networks remained,  and  slight  collage".
fiber deposition was  noted in  the  alveolar  duct region.   Dense accumulator-  cf
dust-laden macrophages was still  evident at alveolar duct bifurcations,  and e
few  adjacent  alveoli   were completely blocked.  As noted  in other exposure
groups,   the Kevlar6  fibers found in dust cells  appeared  to  be fragmented  a-:
shortened.  Lung architecture  appeared  almost  normal at  6  months postexpcsu^e.
although fibrosis was  still apparent in alveolar duct regions.   Some dust-lade-.
macrcphages had  migrated  to  adjacent lymphoid  tissue,  but  no   inflammetcry-
reaction was observed.
     Henderson et al.  (1987)  found that aramid (Kevlar")  is  biocompatible  witn
skeletal  muscle  tissue following  implantation in female  New  Zealand White
rabbits.  Cylindrical  pieces of embedded (methylmethacrylate) or  exposed aran-.id
(12 mm  in diameter  and 4 mm in length) were surgically implanted in the para-
vertebral muscle of  six rabbits.   Silicone was implanted as a control.   After
14 or 28 days,  the  rabbits were killed,  and the tissue reaction  was evaluate:.
No adverse  tissue  response  was elicited in rabbits implanted with  aramid.
Formation of a  thin,  fibrous capsule surrounding  the  implanted  material  was
observed  at  both aramid  and silicone implantation sites, but there  was  nc
evidence of inflammation or necrosis.
5.1.2.2   Carbon  and  Polyolefin Fibers.   Since  carbon  fibers and  polyolefin
fibers  are  used  in  prosthetic devices, several studies have been conductec to
evaluate  tissue  reaction  to  these fiber implants.  In a  study  reportec  b;.
Neugebauer  et al.  (1961), minimal  fibrosis  was  evident  2 to 12 weeks  e'te-
injection of carbon  fiber fragments (diameter 7 urn; length,  20  to IOC J.T)  iru
saline into the medullary  canals of the femur bone of C4EP:CH rassits.  Ca-bc
containing macrophages  were  present,  and a small amount of ne*  bone fc--=fe-
•'.-•>" 1=::                           5-4      DRAFT--DO N:T QjC"  C-, CT:

-------
was evident around  and  including  the carbon fibers.   Carbon fiber particles
were  found  in the  lungs,  liver,  spleen, and  kidneys, but  no  foreign body
reaction was noted.
     The consequences of  release  of  carbon  fiber debris  within the synovia!
joint was reported  by Parsons et  al.  (1985).  One hundred milligrams  of carbon
fibers (mean particle size  was  5  ym) suspended in saline were  injected into
the synovial joints of male New Zealand White rabbits.  The  animals were killed
4 days,  and 2,  4, 8, and 16 weeks after injection.   No  cartilage damage was
apparent, and no  debris were found in the  synovial fluid.  Most of the fiber
material  was  located either in  the suprapatellal region or in  the posterior
aspects  of the  knee.   An acute foreign body response  was evident after  uptake
of carbon debris  into  the synovial lining.   After 8 weeks, the carbon fice'S
migrated to  the deep subsynovial  fatty layers;  the  fibers remained there
16 weeks  after  treatment.   There  were no abscesses in animals  injected with
ca-bon fibers when  compared to  positive  control  animals  treated similarly with
talc.   The authors  reported a more severe foreign body response with  syno»itis
and subsequent  development  of  small  abscesses in subsynovial  tissue  in talc-
injected rabbits.    In addition, selective staining revealed loss of proteogly
can from the  cartilage  after 2  weeks  in  rabbits  injected with talc.   Cartilage
from carbon-injected rabbits retained normal staining  patterns,  suggesting  that
no significant loss of proteoglycan occurred.
     Makisato et al. (1984)  compared  tissue reaction  after implantation of a
1.5- by 1.5-cm strip of carbon fiber  paper or polypropylene  into the  peritoneal
cavity,  under the  skin, and in  the medullary canal of the femoral bone of male
rats  (strain  not  specified).   Animals were   killed and examined 1, 3, 6, and
12 weeks after insertion.  The authors reported that carbon  fibers  induced  more
fibrous  tissue  ingrowth in  both the tissue  and bone than polypropylene, but no
bone  ingrowth was  induced by either fiber.

5.1.3  Minerals
5.1.3.1  Attapulgite.   No  information on the acute toxicity of  attapulgite wss
found in the available  literature.
5.1.3.2  Erionite-  No  information on the acute  toxicity of erionite  was found
in the available literature.
5.1.3.3  Talc.  Intratracheal  instillaton of talc has been shown to  result in
pulmonary toxicity  in  hamsters  (Beck et al., 1987).   Groups of six  har.ste'S
(sex  and strain not specified)  were  given  single  intratracheal  injecfic-s of
July 19££                           5-5      DRAFT—DC N3T QlTE 0* CITE

-------
talc  (mean diameter,  0.8 urn)  suspended in  0.9  percent saline  containing
13.3 (jg/mL  rabbit  surface active  material  (SAM)  at concentrations of 0.15,
0.75, or  3.75 mg/100  g body  weight.   Controls received saline plus SAM.  The
pulmonary  responses  were quantified  by measuring biochemical and cellular
indicators  of injury  in the  bronchoalveolar  lavage  (BAL) fluid.  One day after
talc  instillation, elevated  enzyme levels, pulmonary edema, and  increased cell
numbers in  the  BAL fluid (RBCs, macrophage, and  polymorphonuclear neutrophil
cells) were noted.   Persistent elevation of enzyme levels and inhibition  of
phagocytosis were also noted in a group of four hamsters receiving  3.75  mg/lGT  g
body weight that were  killed 1, 4, 7, and 14 days  after administration.
5.1.3.4 Vermiculite.   No  information on the acute toxicity of vermiculite  »ss
found in the available literature.
5.2  SUSCHRONIC AND CHRONIC TOXICITY
5.2.1  Manrriade Mineral Fibers
5.2.1.1  Fibrous Glass
5.2.1.1.1  Oral.   Only  one study on the  oral  toxicity  of fibrous glass was
found  in  the available literature.  Gross  et  al.  (1960)  fed four groups  cf
10 male white  rats  (strain not specified) a diet  that  contained  either  1C  cr
50 percent of  finely milled glass  flakes.  Two  groups  of animals served  as
pair-fed controls.   The  animals  were maintained on their respective  diets  fcr
1 year.  No significant differences were found in mortality,  weight gain, orga">
weight, or histopathology among the groups.
5.2.1.1.2  Inhalation.   Inhalation studies of up to 2 years  duration  with
glass  fibers  of varying diameters  and  lengths generally  failed to show any
substantial pulmonary damage  in  several experimental species (e.g., Mitchell
et al.,  1986;  Smith et al., 1984;  Pickrell et al.,  1983; Lee et al., 19£1.
1979;  Gross  et al., 1970,  1960;  Schepers and Delahant,  1955).  Pulmonary
responses  generally  consisting of macrophage  infiltration, alveolar dust-lacer.
macrophage aggregation, and alveolar proteinosis resulting in some alteration
of the lung  architecture  occurred in rats, guinea pigs,  and  hamsters  (Lee
et al., 1981,  1979;  Gross  et al., 1970, 1960).  The physical  characteristics
(diameter  and  length)  of  the  fibers appeared  to  have some effect  on the  degree
of tissue  response.   In  general, exposure to long  (>20  urr),  thick  (>5 ur.)
fibers  resulted  in  a  minimal  reaction cnaracterized  by slight  dust-lace-
                                              ^rT--DD N- CJ": 0=.

-------
macrophage accumulation  in  the  alveoli.   By contrast exposure to  long,  thin
(<3 pm)  fibers  resulted in  alveolar  proteinosis, development of  reticulin
fiber networks,  and  some fibrosis.  Short  (<5 pm), thin fibers caused a  severe
inflammatory  response  resulting in dense  dust-laden  macrophage  aggregation,
and  in  some  cases,  enlargement of adjacent  lymph nodes  due to macrophage
migration.   However,  very  little  alteration  of  the  lung  architecture  was
observed.  Chronic inhalation toxicity studies indicated mild cellular reactions
for glass  microfibers  at exposure  levels of <3.0  mg/m .  Higher concentrations
of microfibers  or  fibrous  glass (diameter <3 pm)  resulted  in more pronounces
pulmonary manifestations.   Unlike  asbestos exposure,  the  inhalation of  glass
fibers of  similar  dimensions did not result in collagen formation or develop-
ment of  pulmonary  tumors.   Technical  problems associated with the  design of
inhalation studies using fibers probably contributed to  the lack of a pulr.c-sry
response; only  a small  percentage  of fibers in the  dust  cloud may have beer.
inhaled.
     Pickrell et al. (1982)  suggested that a possible explanation  for  the U:<
of response following inhalation exposure,  as compared to  intratracheal  ir.sv":-
lation,  was that inhalation  introduced a burden that was only 3 to 5 percer*, c*
the lowest  concentration of glass eliciting a response  upon intratraches1
instillation.   Several  other investigators (Gross et al.,  1970;  Lee et al.,
1979,  1981)  have shown that high concentrations  of glass  fibers  in  the  lur.cs
could be  achieved,  and these high  concentrations  increased  the  severity  of the
pulmonary  response  but did  not  result  in  the  collagen development  that was
observed after intratracheal instillation.
     Lee  et  al.  (1979,  1981) conducted two subchronic  inhalation  studies to
determine the effects  of glass fibers, Fybex , and potassium titanate  (PKT)
in rats,  hamsters,  and guinea pigs.  In the  first study  (Lee et al., 197S),
groups of  male  Sprague-Oawley rats, albino guinea pigs, and Syrian hamste-s
were  exposed to ball-milled glass fibers at  a  concentration of  420  mg/rr.'3
(0.42 mg/L) for 6 hours/day, 5 days/week for 90 days.   Controls were expose:! to
air.   The  glass  fibers had  an average diameter of 1.2 pm and length of  <3 pm.
Only 15  percent  of the glass particles were considered to be fibrous (lenct--'
diameter ratio >3).  Animals were killed and examined after 20, 50, and  SO ceys
of exposure  and  6,  12, and  24 months after the last  exposure.  Gross exar-'-.e-
tion revealed numerous  grayish-yellow foci throughout the  visceral  ple-jre c*
glass-exposed animals killed during the exposure  period.  Tne majority, o*
foci   disappeared within 6  mcr.ths  postexposu^e.   The tvila-  lytr.p-  r.ossi c'
.!:,•".• 19t:                           5-7      D2A"--DC NC" Q'jCTE C:.  CITE

-------
animals  exposed  to glass  dust  were  enlarged and appeared mottled (grayisn-
yellow foci).  The pulmonary response.was similar in the three species,  although
the response  in  guinea pigs was generally milder.   Macrophage aggregation  was
noted  in bronchioles  and  alveoli  after 20  days of exposure.   There was a
thickening  of adjacent alveolar walls  that were  lined with proliferating
cuboidal granular  pneumocytes.   By day 50 of exposure, these  responses  were
more prominent and had extended to peripheral alveoli.  At day 90 of exposure,
there  was  evidence of macrophage  disintegration with  release  of granular
material that preceded alveolar proteinosis.   The  granular material  consistec
of  myelin   figures,  lamellar structures, dust  particles,  cellular debris,
degenerating  macrophages,  and  polymorphonuclear  leukocytes.   By  6 months
postexposure, the hyperplastic granular pneumocytes were replaced by mercbra-ic-.s
pneumocytes.  The normal architecture of the alveoli was restored after  rerrcva]
of dust-laden macrophages.   After  2  years,  alveoli  still  containing dust-lase-
macrophages were lined by granular pneumocytes.   However,  there was  nc  signifi-
cant collagen formation.
     Similar results were obtained in a second 90-day inhalation study  re?c"te:
by Lee et  al.  (1981).   In this study,  the fibrogenic potential  of ball-milled
glass fibers was compared to asbestos (positive control),  Fybex'5, and PKT.   The
results  of  Fybex   and  PKT  exposure are  reported in section 5.2.1.4.  Grours of
46 male  Sprague-Dawley rats, 34 hamsters (sex  and  strain  not specified),  anc
32 male  albino guinea  pigs were exposed to  approximately 420 mg/m   of  glass
fibers (average  diameter  1.2 urn and  length <2 urn;   only 7 percent of the dust
particles  were  considered to  be fibrous) or to 300 mg/m   amosite  asbestos
(average dimensions  of 5.0 by 0.4  urn).   The  pulmonary  response observed  in  all
animals  exposed  to glass  dust Included dust-laden  macrophage aggregation with
hyperplastic granular pneumocytes and slight alveolar proteinosis that  appeared
on day 90  of exposure  but disappeared  by 1 year  postexposure.  Ferruginous
bodies were observed in the  lungs of hamsters and guinea pigs but not rats.   By
18 to  24 months  postexposure, 2 of  19  rats  and 2  of 7  guinea pigs exposed  to
glass  dust developed bronchoalveolar adenomas.   There were  no  incidences  of
Tung tumors  in  control rats or  guinea  pigs,  and no hamsters  exposed to  glass
fibers developed  lung  tumors.   In animals exposed  to  asbestos,  marked  brcn-
chiolization and collagenous fibrosis were noted after 90 days of exposure.   Cf
the 16 rats exposed to asbestos, two developed bronchoalveolar adenomas  anc one
an epidermoid carcinoma.   It is difficult, however, to assess  the  p^lmo-e-y
     i9£=                           5-e      DKAFT-DD N:T QL-C-'E o- ci"

-------
effects  obtained  from the  two  studies  reported by Lee et al.  (1979,  19cl)
because  the  total  respirable  concentrations  were not reported  and only  7  to
15 percent of the particles tested were  fibrous.
     Wagner et al. (1984) reported the results of an inhalation  study  compering
the fibrogenic properties  of  glass  fibers, mineral wool (see section 5.2.1.2),
and chrysotile  asbestos fibers.  Three-types of glass fibers  were tested:
glass wool  with and  without  a resin coating and glass microfibers (Johns
Manville Code 100).   Fiber dimensions for all test clouds were approximately
equivalent with  diameters  <3.0 pm and lengths >5 um.  However, fiber  counts
varied considerably as  follows:   240, 323,  1436, and 3832 fibers/cm   for  t-e
glass  wool with  resin,  uncoated glass wool,  glass microfibers, and asbestos.
respectively.  Groups of  56 Fischer 344 rats  (28 males and  2E  females)  were
exposed to 10 mg/m  (respirable fiber concentration) of each  fiber for 7  hcu-s/
day, 5 days/week  for  12 months, and  observed for  life.   A group of  control
animals was included in the study, but no information on exposure  was  repo-tec.
Interim kills were performed at 3 and 12 months during the  exposure pe'io:. a-;-
et 12 months  postexposure  (24 months  on  study),  to  assess  the extent  of  tissje
reaction to the dusts.
     No  fibrosis  was  noted in  animals  inhaling  the  glass  wool  (unccete:  cr
coated)  using  the grading  system of McConnell  et  al.  (1984)  presentee in
Table 5-2.   Mean  scores  at 3  months of  exposure  and  12 months  after  exposure
were very similar, ranging from 2.5 to 2,7.   These results  suggested that tnere
was little  increase in  the severity of  the  cellular  reaction with  increasing
length of exposure.  Mean scores were slightly higher for  the glass microfibers,
ranging  from  2.8  to  3.3, but again there was no increase  in the severity  of
cellular  reaction with  increasing  exposure  length.   Unlike glass fibers.
asbestos exposure  resulted in increased  severity of  cellular reaction  with tne
development  of fibrosis  as the length of exposure increased.  Mean scores  for
asbestos were 3.0 at  3 months  of exposure  and 4.0  at 12  months of exposure.
Animals  dying  during  the  study exhibited a  slightly more  severe  cellule"
reaction,  and  a  few  rats  exposed  to the glass nricrofibers developed some
fibrosis.  Mean scores for animals dying during the study from the glass  ificrc-
fiber group  ranged from 3.4 to 3.7.   Incidenr.es  of  bronchoalveolar hyperplasia
were 3/4£,  3/48,  1/47,  and 5/46 for animals  exposed to the glass micrcfine-s.
glass  wool with  resin,  uncoated glass wool,  and asbestos,  respectively.   3r>
addition, 1/4S animals each from the glass microfiber and coste: g'asi *-c'

-------
        TABLE 5-2.  THE H1STOLOGIC GRADING SYSTEM OF PULMONARY CHANGES
                  IN RATS RELATED TO THE INHALATION OF FIBERS
Histologic change
                              Score
Description
Cellular
Fibrcsis
                     Normal      1  No lesion observed.

                     Minimal     2  A few macrophages in the lumen of the
                                    terminal bronchioles and alveoli.

                     Mild        3  Presence of cuboidal epithelium lining
                                    the proximal -alveoli (bronchiolization).
                                    No collagen, but reticulin fibers may be
                                    present in interstitium at the junction
                                    of the terminal bronchiole and alveolus.

                                    Luminal macrophages are more conszncjc-s .
                                    and mononudear cells may be fojr.c ir
                                    the interstitium.

                     Minimal     4  Minimal rollagen deposition at level c*
                                    terminal bronchiole and alveolus.
                                    Increased bronchiolization with
                                    associated mucoid debris suggesting
                                    glandular pattern.

                     Kile        5  Interlobular linking of lesion des^'oe:
                                    in grade 4 and increased severity cf
                                    fibrosis.

                     Moderate    6  Early consolidation.  Parenchyma!
                                    decrease is apparent.

                     Severe      7  Marked fibrosis and consolidation.

                                 8  Complete obstruction of most airways.
__
 Nonfibrotic lesions.

Source:  McConnell et al. (1984).
groups developed  an  adenocarcinoma,  and 1/47 animals  from  the uncoated  glass
wool  group  developed "a  benign  adenoma during the study.   A  total  of 12/48
animals  exposed  to  asbestos  developed lung  tumors  (11 adenocarcinomas  ar.c
1 adenoma).   One  control  animal exhibited bronchoalveolar hyperplasia, but nc

tumors were observed.
     Le  Bouffant  et al.   (1986,  1967)  also studied the pulmonary effects  c*

glass  wool,  rock wool  (see section 5.1.2.2),  and glass  microfibe's (J:">~s
Manville IOC)  in  Wistar  rats.   Grcuos of 24 males and 24 females we^e expose:

-------
to each fiber at a respirable concentration of 5 mg/m ,  5  hours/day, 5 days/week
for 12  or  24 months.   Control animals were exposed  to air.  Rats exposed for
12 months were  observed  for  up to 16 months, and  rats  exposed  for 24 months
were  observed  for up to 4 months  after  exposure was halted.   Approximately
31.5 percent  of the  glass wool  fibers were  10 to  20 urn in  length, and
4E.4 percent of  the microfibers  were  5 to  10  urn  in length.  The diameters were
as  follows:   glass wool <1 urn (68.7  percent) and  microfibers  0.2 to 0.5 urn
(51.3 percent).   The  quantity of dust deposited in  the lungs  varied  greatly
according to the fiber  type; the shorter  fibers migrated to the  lymph nodes.
Chronic inflammation  with  hypersecretion  of mucus  and epithelial  desquamation
but no  metaplasia  were  observed  in the nasal fossae of  rats  exposed  to  both
glass fiber  samples.  The  trachea,  larynx, and other organs except  lungs we~e
normal, however.  In the lungs, small  aggregations  of macrophages  were observe:
mainly at bronchoalveolar bifurcations.   Mild, localized hyperplasia  of  type II
pneumocytes  was noted  in  areas  adjacent  to  macrophage  aggregation  after
12 months of exposure.  After 24  months of exposure,  the lungs exhibited  sir.e1.1.
sparse  foci  of  macrophagic  reaction with a small  amount  of  septa!  reticclin
fibrcsis.   One  of 45 rats exposed  to the glass wool developed a pulmonary
carcinoma.   Progressive  chronic  nephrosis  was observed in glass-exposed  rais.
The lesions were also noted  in controls but  were  less  severe than in gless-
exposed rats.  No pulmonary tumors were observed in microfiber-treated animals.
     Mitchell et el.  (1986)  and  Moorman et al.   (1985)  reported that  chronic
inhalation of  commercially available glass wool insulation induced  pleursl,
plaque!ike lesions  in rats,  but not in  monkeys.   Groups of  50  male and
50 female F344  rats were exposed to 5 or  15  mg/m   of four fiber samples of
varying lengths  and diameters for 7 hours/day,  5 days/week for 86 weeks,  The
specific fiber dimensions and concentrations tested are presented  in  Table 5-3.
The plaquelike  lesions, described  as multifocal,  tan or  gray, raised,  firm
areas 1 to  6 mm or more in diameter,  were  located  on the  surface  of the  lungs.
These  pleura!  plaques were  found  in  all  glass-exposed animals except these
exposed to the  long,  thick fibers  (4 to 6 uni in diameter).   In addition, the
lesions were most  severe in  rats exposed  to  the long,  thin fibers and  we"e
relatively mild in rats exposed to  the  short, thin  fibers.   The  authors
concluded that  the  plaques resulted from  accumulation of  granulomatous foci in
pleura! and  subpleural  areas of  the lungs.  The only other pulmonary response
noted was the aggregation of macrophages and the presence of granuloires
                                    5-11          DRAFT--D: N:~ QJ

-------
        TABLE 5-3.  DIMENSIONS AND CONCENTRATIONS OF GLASS  FIBERS  USED
      IN A CHRONIC INHALATION STUDY WITH F344 RATS AND CYNOMOLGUS  MONKEYS
Fiber
description
Glass-red binder
Glass-yellow binder
Glass-uncoated
Glass-uncoated
Centre!
Fiber dimensions3
d = 4-6 urn
1 = >20 urn
d = 0.5-3.5 urn
1 = >10 pm
d = <3.5 pm
1 = >10 pm
d = <3.5 pm
1 = <10 pm
--
Target concentration
(mg/ms)
15
15
5
5
0
ad = diameter, 1 = length.
Source:   Mitchell et al. (1986).

containing "fibrous" glass  fibers.   As noted earlier, the  long, thick fibers
did net  elicit any gross pulmonary lesion including macrophage aggregation.   It
is interesting  to  note that macrophage aggregation  and granulomatous inflar-.s-
tion were more severe in animals exposed to the short, thin fibers,  whereas  the
severity of the  plaquelike  lesions  was greater  in  animals  exposed to the  long,
thin fibers.  Macrophage  aggregation and granulomatous inflammation were also
noted in the  thymic  and tracheobronchial lymph nodes of  a few glass-exposed
rats.  Glass  fibers  were  not found  in  any  other  tissues.   Besides  pulmonary
effects, the only other significant finding noted was an  increased  (significant
at p <0.05)  incidence   of mononuclear cell  leukemia in rats  from all  glass
fiber-exposed groups  as compared  to the control  incidence.   However,  the
biological significance of this finding is unknown.
     A  similar  72 weeks  exposure  of  groups  containing  15 male cynomolgu's
monkeys did not induce the  plaquelike lesions  or  an increased incidence  of  MCt
noted in  rats  (Mitchell et  al.,  1986; Moorman  et  al.,  1985).   Scattered  macrc-
phage aggregation  occurred  throughout  the  lungs,  and dust-laden macrophsoes
were found  in  tracheobronchial  lymph nodes as  well  as in  the pleura anc  sut-
pleura in all  glass-exposed animals.  No evidence  of  translocation tc cf-er
organs was  found.   Increased numbers of  lymphoid  nodules  or  aggregates  we-e
                                    5-12          DR;"T--DC N:~

-------
observed in monkeys  exposed  to the long, thick fibers.  However, the autho-s
theorized that this resulted from an immunological  response  to  the binder.  The
results of the monkey study may have been compromised because of  an  intercurrer.t
lung mite  infection  that  resulted in pulmonary lesions in the  majority  of  the
monkeys (59/60) on study.
     Gross et al.  (1970)  and Gross (1976) classified glass  fibers as  a  nonfi-
brogenic or  nuisance-type dust  based  on the tissue reactions  seen in  rats
and  hamsters  after long-term  inhalation exposure.   Groups  of  30  rats and
30 hamsters (strains  and  sex  not  specified) were  exposed to uncoated glass
fibers  or  glass fibers coated with phenol formaldehyde  or  a starch binder
(diameter 0.5 urn,  length  5  to 20 urn) at approximately 100 mg/m  for 6 hoys/
day, 5  days/week for  2  years.   The fiber samples were specially manufactured
tc simulate the  glass fibers found in insulation and  textiles by addition of
coating materials, but  these fibers  had  diameters within the respirable ranee.
An additional group  of  20 rats and 20 hamsters served as controls.   Results
revealed no differences in the pulmonary tissue reaction for the coated versus
the  uncoated glass fibers.   Dust-laden macrophage  aggregation  occurred ir.  the
alveolar region  at the  alveolar duct bifurcations.   No atelectasis was present
anc1 generally,  alveolar architecture was  unchanged.   A few rats exhibited  focal
collagenous fibrosis  typically seen  after inhalation of  large  amounts of  non-
fibrous dust.   The migration of dust-laden macrophages to the  lymph nodes  was
observed, and the  lymph nodes  of a few animals were  enlarged due to macrophage
accumulation.   Consequently,  fibrosis developed in  the lymph nodes.   The tissue
reaction in hamsters  was  similar to  that in  rats except  that macrophage aggre-
gation  was more  dense,  and numerous ferruginous bodies  were observed in  the
alveoli of hamsters.   Diffuse, focal,  bland, acellular, collagenous  fibrosis
of the  pleura was  noted in a few glass-exposed animals.   Calcified., laminated
microliths were  also  noted  in the alveoli  of a  few  hamsters.   In addition, tne
authors stated  that  several  hamsters died of pneumonia  during the  study (the
number  of mortalities was  not reported).   This  early mortality  and  the presence
of alveolar microliths  were  attributed to alveolar  clearance failure,  suggest-
ing that the fiber concentration in the  lungs may have been  excessive.
     In a  comprehensive chronic inhalation study by Smith et al.  (1984, 15=6,
1967),  the effects of several  manmade  vitreous  fibers  were  studied  in  rets and
hamsters.  The  fibers tested  included  an uncoated  glass microfiber  (Jonns
Kanville code  100);  glass  wool  with  a   silicone  lubricant (Certainteec"\

Julv 15S£                           5-12          PRA?T--DC M' C-'-T =  c^ CI~-

-------
Insulsafe  II  ,  "blowing wool" building insulation); two glass  wool  build-ing
insulations coated with phenol formaldehyde  binders; a  refractory ceramic fiber
(see section 5.2.1.3);  slag wool (see section 5.2.1.2); and crocidolite  asbestos
as  a  positive control.   The  dimensions, concentrations, and  particle ratios
(nonfibrous-to-fibrous)  of  the  glass fibers used are presented in Table 5-4.
The particle ratios indicated that the aerosol clouds  contained more  nonfibrous
than fibrous  particles.   Groups of female Osborne-Mendel  rats and  male  Syrian
hamsters were exposed nose-only to the  various  fibers  for  6 hours/day, 5 days/
week for 24  months,  and then observed  for  life (see Table 5-5  for numbers of
animals used).
     The lifespan of  the hamsters  was  not adversely affected  by the  inhalaticr.
of  glass  fibers.   In fact, hamsters from all  the glass fiber  groups exce;t
those exposed to  the  3.1-um mean  diameter  (MD)  glass  wool had significantly
longer lifespans  than both control groups (p value  not reported).  Microscopic
examination revealed  fibers in  every section of  the  lungs.   The  fibers were
generally phagocytized,  and numerous ferruginous bodies were present.   Tnere
was a dose-related  increase in the  degree  of  macrophage aggregation, but tie
macrophage foci did  not progress to fibrosis.  After exposure was  halted,  tr.e
macrophage  aggregation   either  regressed  or  remained  the  same.   Pulmonary
lesions were  scored  using the grading  system  of  McConnell  et al.  (1954) anc
are presented in Table  5-5.
     Similar findings were  observed in rats and again, the major lesion noted
was macrophage  aggregation.   The ferruginous bodies observed in hamsters were
not found  in rats, however.   The  grades  obtained for  lung  lesions  did net
exceed 4 in  any of the  glass-exposed animals, indicating that no collagen had
developed.   The smaller diameter  glass fibers appeared to cause  a  greater
degree of cellular change in the lungs, but there was  no apparent dose relation-
ship.   The  highest  scores were obtained in  animals exposed  to asbestos, but
even these  scores  were   less than 4, suggesting that asbestos exposure did net
result in  collagen  formation  although fibrosis was present.   Slight increases
in  the  incidence of  bronchoalveolar metaplasia  were  observed in hamsters
exposed to  3  mg/m   of 0.45-um MD glass  microfibers  and 1.2 mg/m of  the  5.4-pr
MD glass wool when compared to controls, but the differences were not statisti-
cally significant.  The incidences of fibrosis were comparable between  glass-
exposed and control  animals;  no tumors were found in any of the glass-exposed
animals.   Hamsters exposed to asbestos had significantly higher incidences
Ju'y 19:;                           5-14          DRArT--D: MT QJC'E

-------
               TABLE 5-4.   DIMENSIONS, AEROSOL CONCENTRATIONS,  AND PARTICLE  RATIOS
                 OF FIBERS USED IN 2-YEAR INHALATION STUDIES WITH OSBORNE-MENDEL
                                    RATS AND SYRIAN HAMSTERS
Fiber dimension
Fiber
description
"Glass microfiber
(no binder)
High level
Low level
Glass wool
(silicone
lubricant)
•Glass wool
(binder coatee)
Hig". level
Lov. level
Glass wool
(binder coated)
Refractory
ceramic glass
Slag wool
UICC crocidolite
asbestos
Mean
diameter
(urn)
0.45

3.1
5.4


6.1
1.8
2.7
"d
Mean
length
(um)a
7.5 1 10

37 ± 48
31 ± 33


114 ± 94
35 ± 34
40 ± 63
"d
Fiber
concentration
(fibers/cm3)

3000
300
100
100
10
25
200
200
3000
Mass
concentration
(nig/m3)

3.0
0.3
10.0
12
1.2
s.o
12.0
10.0
7.0
Respirable
mass b
fraction Part'
(% w) rat
4:
61
_b
30 6:
3£:
13
-a
31:
19
35 33:
15 28:
_a
*Mean ± SD.
 Characterized using cyclone sampling data.  Represents the total aerosol mass, not just
 fibrous mass.
cRatio represents nonfibrous*to-fibrous particles.
 Not reported; length ranged from 0.2 to 25 urn.

Source:  Smith et al. (1986, 1987).
        Julv 1S=.£


-------
TABLE 5-5.  INCIDENCES OF PULMONARY LESIONS NOTED IN OSBORNE-MENDEL
   (OK) RATS AND SYRIAN HAMSTERS EXPOSED TO VARIOUS GLASS FIBERS
                    BY INHALATION FOR 24 MONTHS

Exposure
level
Exposure group (mg/m3)
Glass tnicrofiber (MD 0.45 urn)
Syrian hamsters 3
OM rats
Syrian hamsters 0.3
OM rats
Glass wool (MD 3.1 urn)
Syrian hamsters 10
OK rats
Glass wocl (M: 5.4 urr.)
Syriar, hamsters 12
OK rats
.Syrian harsters 1.2
OM rats
Glass wool (K~ 6. 1 urr.)
Syr-; s- f.s-i-.ers 9
1st group
2nc group
0'-', rats
Refractory ceraxic glass
Syrian ha.rsters 12
OK rats
Slag wool
Syrian hamsters 10
OK rats
UICC crocidolite asbestos
Syrian hamsters 7
OM rats
.Chamber controls (clean air)
Syrian hamsters
OK rats
Unmanipulated cage controls
Syrian hamsters
OK rats
One mesothel ioma.
Eronchcal veolar tumor.
'"Significantly different from controls (p
So^-ce: Sir.it h et el (!<<£- 1937).


Lesion
grade
2.3 ± 0.1
2.3 ± 0.1
2.1 ± 0.1
2.3 ± 0.1
2.1 ± 0.1
2.3 ± 0.1
1.6 ± 0.1
2.1 ± 0.1
1.8 ± 0.1
2.4 * 0.1
1.9 ± 0.1
1.3 ± 0.1
2.3 ± 0.1
2.1 ± 0.1
3.0 ± 0.1
2.4 ± 0.1
2.8 ± 0.1
2.8 ± 0.1
3-7 ±0.2
1.8 ± 0.1
2.2 ± 0.1
1.7 ± 0.1
2.4 ± 0.1
value not reporte

Incidence
Broncho-
alveolar
metaplasia
6/69
4/57
2/70
4/57
3/60
1/52
1/65
1/57
7/65
0/61
4/61
0/36
1/56
2/69
1/55
2/69
0/55
11/58*
4/57
3/57
5/59
4/112
3/125
o.

o* arirr.a":
Fibrosis
4/69
4/57
3/70
3/57
2/60
7/52
0'66
4/57
1/65
5/61
C/6I
0/3£
3/5£
1/70
12/55
1/69
9/55
14/58*
30/57*
4/58
6/59
3/112
17/125


s w't1-,:
Primar
tumor:
C/6:
c/5:
0/7(
0/5'
0/c1
r. 't
w / -•
C '£
C/5
0/6
0/c
1
C ~
1/7
0/5
C/6
0/5
0/5
3/5
1/5
L' / 5
C-'l
C'l



                                              L'"..--

-------
(p value not reported) of bronchos! veolar metaplasia  and  fibrosis  than  controls,
but no tumors were Induced.  Rats exposed to asbestos had a  significant increase
in the  incidence of fibrosis when compared  to  controls, and 3/57 developed
pulmonary tumors.  No  tumors  were found in the unmanipulated control  rats or
hamsters.  One  of  58 sham control hamsters developed a bronchoalveolar turner.
No sham control rats developed tumors.
     Although bronchoalveolar metaplasia nay be a preneoplastic event in the
development  of  epithelial tumors, none  of the groups exhibiting  increasec
incidences of  bronchoalveolar metaplasia developed  pulmonary tumors.   There
are several  possible explanations.  The  asbestos  sample  used had a mean lengtr
of 5 urn, which  may have been too  short  to  induce lung tumors.   Although  the
authors stated  that  fiber particles were observed in the lungs, the conce~trs-
tion of  respirable  fibers actually inhaled may have  been too small tc inc'jce
fibrosis or development of tumors.  Also, the generated aerosol  clouds  prots:1y
did net contain enough fibrous particles to adequately assess possible  biclcc/-
csl effects of fibers on pulmonary physiology.
     Goldstein et al.  (19£3)  demonstrated  that  focal peribronchiolar fitros'5
can develop  in  baboons following the  inhalation of a fibrous glass dust clt-jc.
A total of 10 male baboons (age not specified)  were dusted 7 hours/day, 5  c=ys '
week for up  to  35 months, with  either  fibrous  glass (Johns  Manville C102 er.c
C104 glass  wool) or  crocidolite asbestos (positive  control).   No negative
control group was  included.   The glass fibers measured  5 to 100 urn  in length
and 0.1  to 4 urn in diameter  (mean diameter was  approximately 0.75  pm),  anc tne
total respirable  concentration  was 5.8 mg/m  .   The asbestos  fibers ranged from
0.1  to  1 urn in  diameter and 0.4 to  25 pm  in length; the total
concentration was  13.5 mg/m .   Fiber  counts  were similar  for both dusts.
Minimal macroscopic  changes  were  seen at  all  stages  of exposure,  the earliest
microscopic changes  observed (after 8 months of exposure) in lung biopsies of
the  glass-exposed  animals included  bronchiolar  thickening,  the presence  c*
scanty  ferruginous  bodies,  and moderately  compact reticulin  formation.   An
organizing  desquamative  alveolitis  usually preceded  the  appearance  of  peri-
bronchiolar fibrosis.   The  fibrosis  was  slightly more marked after 18 ar.c
30 months of exposure  and did not appear to regress up to 7 weeks after expo-
sure was  halted,  although the animals were to be observed for life anc  tne
study was not yet  complete.   The  fibrotic lesions were morphologically  sirr-'ia"
to those  produced  by asbestos but were less extensive.   The a-tno^s sucrsite:

-------
that  the difference  in  severity  of response  between  fibrous glass  and
crocidolite may be due to differences in fiber concentrations  within  the  lungs,
which were  ultimately attributed  to differences in fiber dimensions.   The
authors  attributed the  appearance  of peribronchiolar fibrosis in baboons but
not in  other  animals  (see above studies) to  differences  in  fiber  dimensions.
and/or  species  used.   However,  an  assessment of this finding  is difficult in
the absence of other supporting evidence and untreated  control  animals.
5.2.1.1.3  Intratracheal.  Glass fibers  of  varying  lengths and diameters were
administered  by  intratracheal  instillation to  several  animal species in  a
number  of studies.  The  results  indicate that the physical characteristics  of
the fibers, as  well as  dose, determine the  nature and extent of tissue injuTj.
Very short (<5 urn), very thin  (<0.3 urn) fibers  apparently cause no pulmonary
responses.   Short  (<10 um), thin  (<2 urn)  fibers elicit  simple macrophage
response.  Longer (>10 urn) or thicker fibers (<3 urn) result in a more pronc-j-.ce:
response and fibrosis.
     Tne relationship between the  physical  dimensions  of  the fiber and tissue
injury was demonstrated  by  Pickrell  et al.  (1963).   Groups of 20 male Syria-
hamsters received two  biweekly  intratracheal  instillations  of uncoated  glass
micro-fibers or household insulation fibers in saline suspensions.   The die-eter
and total dose  administered  for  each fiber sample  is  listed  in Table  5-6.   A
group of 30  controls  was instilled with  saline.   Animals were killed  and
examined 1, 3.5,  and  11 months after the initial instillation.  The  smallest
diameter microfiber (0.1  urn) caused  deaths  from pulmonary edema 2 to  30 days
after the initial instillation.   A slightly larger microfiber (0.2  urn diameter)
and one  type  of  household insulation fiber  (2.3 urn  diameter)  induced  increases
in total collagen and mild pulmonary fibrosis at 11  months, with the  microfiber
producing a greater effect than the insulation fiber.
     Kuschner and Wright  (1976)  demonstrated  that the physical characteristics
of  the  fiber  determine  the  nature and extent of tissue  damage.   Groups o*
30 guinea pigs  (strain  and  sex not specified) received  2  x 12.5 mg or  3  x 4  rr.c
of a short or long glass fiber, respectively.   Animals  were sacrificed after  6.
12, or 24 months.  Short, thin fibers (0.05 to 1.99  urn  in  diameter, only  7  per-
cent >10 urn in  length)  induced a macrophage reaction but  not fibrosis.  Lone.
thin  fibers (0.05  to  1.49 um in diameter,  only 7  percent <10 urn  in  ler:*.-)
induced  peribronchiolar  interstitial fibrosis  after 6 months.  Very  short.
very thin fibers (0.03 to 0.29 umm in diameter, <5  um in length) cause: ~:

-------
             TABLE 5-6.   TYPE, DIAMETER,  AND DOSE  OF  GLASS  FIBERS
                  INTRATRACHEALLY INJECTED IN SYRIAN  HAMSTERS
     Fiber                       Fiber count median                   Total
   description                     diameter (pm)                    dose (mg)
Type 1 glass
Type 2 glass
Type A insulation
Type B insulation
Type C insulation
0.1
0.2
2.3
3.0
4.1
7
2
21
18
17
Source:  Pickrell et al.  (1983).

pulmonary alteration up  to  2 years after injection.  Long,  very  thin  fibers
(0.03  to 0.59 pm in diameter, 50 percent >10 pm  in length)  cat-see1 coarser.ir.c
of the  lung  texture with peribronchiolar interstitial fibrosis involving the
proximal alveoli.   Some  interstitial  fibrosis  was seen  2  years  after trie
injection of  short, thick  fibers  (2 pm in  diameter,  88 percent  <10 pm in
length).  Finally,  animals  that received  the  long,  thick  fibers  exhibited
focal  areas  of  interstitial  fibrosis  at  6 months  postinjection, and a  ccs'se--
ing of lung texture with ferruginous bodies  was noted  at  2 years.
     Drew et al.  (1986,  1987) conducted intratracheal   instillation stucies
with sized glass fibers  in  male  Fischer 344 rats.  The fibers were either  1.5
by 5 pm  or 1.5  by 60 pm (short fibers having a mean length of 5.1 pm  and  lone
fibers  having a  mean length of 54 pm).  Animals  were  killed 6,  12, 18, and
24 months after  administration.  Instillation of  both  short  and long fibers as
a  single 20-mg  dose produced a granulomatous  foreign body  response.   In
contrast, low concentrations  of  both short and long  fibers  in multiple dcses
(0.5 mg  once each week for 10 weeks for a total dose of 5 mg) elicited  a sircle
macrophage response up  to  11 months after the last  instillation.  No lung
tumors were seen.
     The pulmonary  effects  of glass  fibers  with equivalent dimensions  but
different surface properties  were  reported  by Gross et  al.  (1970)  and Gross
(1976).  The tissue reaction of  uncoated glass fibers was  compared with that
of  a phenol formaldehyde-coated glass wool  and  a  textile glass fiber with a
starch  binder  (0.5  pm  in diameter, 5 to 20 pm  in length).   Groups of 15  c<-
30 rats  and  12  hamsters (strains  and sex  not  specified)  were  given  intra-
tracheal injections of 3.5 mg (in 1 ml water) of  one of the  three  samples.   The
glass  fibers were  injected  once  weekly for  1  to 10 weeks, and the ar,ir,= !s  were

-------
then  observed  for life.   No differences in tissue reaction were noted for the
uncoated  versus  coated glass fibers.   Dust-laden macrophage  aggregation  was
observed  in  the  alveolar and bronchiolar regions.   Numerous ferruginous bodies
were  found  in the  airways of  hamsters.   A  diffuse,  bland,  acellular,
collagenous  pleura1  fibrosis was  present in  both  rats and  hamsters.  The
authors classified glass fibers as nonfibrogenic or nuisancelike dusts based on
the tissue reaction  observed in this study.
      Schepers  and Delahant (1955)  studied the  effects of  fiber  diameter on  tr,e
lungs  of  guinea  pigs.   Groups of five  to eight guinea pigs  (strain  and sex  not
specified) were  injected intratracheally with  three types of  glass  wool fibers
(mean  diameters  of 6, 3,  and £3 pm  and average lengths of 20  to  50 urn)  c*.
'concentrations of 0.03 and 0.003 mg in a saline suspension once weekly for 52
to 76 weeks.   The investigators found that the largest fibers were relatively-
inert , the medium-sized  fibers  caused foci of  atelectasis  early in the study
with  major  pulmonary  damage  seen  after  75  weeks,  and the smallest  fibers
appeared  to cause mainly focal atelectasis.  No fibrcsis was  seen in any g"0i.-r,
and the  severity  of the  reaction appeared  to be inversely  related  to the
caliber (i.e., diameter) of the fiber.
5.2.1.1.4  Intrapleural.   The  extent of the  tissue  response after  intrapleui-al
administration has  also  been shown to  depend  on  the physical characteristics
of glass  fibers.   Davis  (1976) injected  a single  dose  of 10  mg (in distilled
water)  of four  different  samples  of boron silicate glass  into the pleura'
cavities  of  groups of 25  BALB/c mice  (sex  not specified).   The glass sarr.ples
consisted  of fibers with  mean  diameters of either  0.05  or  3.5 um  and were
either  long  (over several  hundred microns) or  short (about 20  |jm).   The  long
fibers produced  massive  fibrosis,  whereas the  short fibers  induced  only smell,
discrete  granulomas  with minimal fibrosis.
5.2.1.2   Mineral Wool.    Several  investigators have studied the  pulmona-y
effects  of  mineral  wool  in rats  and  hamsters after  inhalation exposure.
Mineral  wool  having dimensions  within the respirable  range  elicited mainly
dust-laden macrophage  aggregation with only slight changes in lung architecture.
The results  were comparable to those obtained  for glass fibers  having similar
dimensions.   In  the  study by Wagner et al. (1984) and McConnell et al. (15S-)
described earlier, groups  of 56 Fischer 344 rats were exposed via inhalation to
commence! grades  of Swedish rock  wool, as well as  several  glass fiber sables
and chrysotile asbestos  as a positive  control.   The rock woe"1  sample r,e:  a

-------
length of >5 pm,  and  87 percent of the fibers had diameters ranging from <0.02
to 2.0 pm.  The gravimetric concentration was 10 mg/m ,  and the exposure period
was 7 hours/day, 5 days/week for 12 months.   The animals were  then  observed for
life.  Results  indicated  that  the pulmonary reaction to all  tested MMMF was
mi Id  compared  to that for asbestos.   There was no apparent  progression of
severity  as the  exposure  length increased in the MMMF-exposed rats.   Lesion
grades ranged  from  2.7 to 2.9 in rock wool-exposed  rats that  survived  2 years
(see  Table  5-2).   Animals that died  during  the study,  including those that
survived  longer than  24 months, had slightly higher scores ranging from 3.C to
3.3.   The  grades  for asbestos-exposed  animals  ranged  from  3.C  to 4.0  in
animals surviving  24  months and  from 4.3  to 4.7 for rats dying during  the
study.  Two  of  48 'rats exposed to  rock  wool  developed  lung  tumors  (both
benign adenomas),  and one  animal  exhibited bronchoalveolar hyperplasia.  A
total of  12/48  rats'' exposed to asbestos developed  lung tumors  (11 adenocer-
cinomas  and  1  adenoma).    Bronchoalveolar  hyperplasia  was  noted  in 5  of
45 astestos-exposed rats.
                    •,
     Le Bouffant  et'al.  (1986,  19S7) exposed groups of Wistar  rats  (2-*  rr.s'es
and 24 females)  to  rock wool  and  two glass  fiber  samples at  respirable cc.i:e~-
trations  of  5 mg/m   for  5  hours/day, 5  days/week  for  12 or 24 months.   Tne
glass  fibers were discussed  above  (see section  5.2.1.1).   The  length  of
95 percent of  the  rock  wool  fibers was  <50 pm,  and  the  diameter  of
82.7 percent of  the  fibers ranged  from <1  to 4 pm.   Rock  wool  inhalation
elicited  mild  macrophage  aggregation in the  lungs  after 12 or 24 months of
exposure.  Chronic  inflammation was observed in the nasal  fossae,  and kidney
lesions,  described as progressive chronic nephrosis, were noted after 18  mcr.tns
of exposure  to  rock wool.  Kidney lesions were also noted in controls  but were
more  extensive  in rock wool-exposed animals.  One  of  47 rats exposed  to  rock
wool developed a peritoneal carcinoma, but no pulmonary tumors were found.
     In  a chronic  inhalation  study by  Smith  et al.  (1986,  1987) describee
previously (see  Section  5.2.1.1.2), groups  of female Osborne-Mendel rats ar.c
male  Syrian  hamsters  were exposed to several  manmade vitreous fibers inducing
slag wool.   The  slag wool  was described as having a CML of 40 pm and an  KZ-  of
2.7 pm.   Approximately 75 percent of these fibers  were >10 pm  in length, a —
95 percent had  diameters  <2.5 pm.   Animals  were exposed nose-only to sis; *cc'
at a  concentration  of 10  mg/m  for  6 hours/day,  5 days/week for 24 mor.tns. e-c
observed  for life.  Aerosol characterization is presented in Te     -
                                                                    -.

-------
     Mean lifespans were  significantly  shorter  (p value not reported) for rats
exposed to  slag  wool  when compared to  controls  (677,  754,  and 724 days for
rats  exposed  to slag  wool,  chamber controls,  and  unmanipulated controls,
respectively).   No  corresponding  decrease  in mean lifespan was noted in slac
wool-exposed hamsters.  The mean lesion scores based on McConnell  et  al.  (1S6-)
for rats  and  hamsters exposed to  slag wool were 2.8  and 2.4, respectively (see
Table 5-2).   Bronchoalveolar metaplasia was noted in 2 of  69 hamsters but in no
rats  exposed  to  slag  wool.   There was  a slight increase  in fibrosis in slag
wool-exposed rats when compared to controls, but there was  no corresponding
increase  in hamsters.   A  summary  of  the pulmonary effects of slag wool  as we""
as the other manmade vitreous fibers tested in this  study  is presented in Tei'e
5-5.   Nc  pulmonary  tumors were induced in rats or  hamsters exposed to  sis;
wool.
5.2.1.3   Ceramic Fibers.  As  observed  with other vitreous fibers, the inhe'a-
ficr.  cf cerarric  glass fibers by rats has  been  shown to result in  little p_l-
mo-a-y  response  other  than  dust-laden  macrophage aggregation.  Intraperii:-
neal   administration of ceramic  fibers  resulted in fibrosis  in  rats.   Ho*e.-er,
the fibrosis  contained little cr  nc collagen  and was, therefore, in nc w=_.
comparable  to  the  fibrosis  elicited by asbestos.   Several  other long-te-rr
studies using  the  intraperitoneal  and  intrapleural   routes  were conducted  in
hamsters  and  two strains  of rats.   However, only carcinogenic endpoints were
discussed,  and  therefore,  these   studies  are  presented in  Chapter  7  (see
section 7.3.1.3).
      In a comprehensive study by  Smith et al.  (1986,  1987), female  Osborne-
Mendel  rats  and male  Syrian hamsters were  treated  with  refractory  ceramic
                        £
glass fibers  (Fiberfrax )  in addition  to several other MMMF via inhalation or
intratracheal  and intraperi.toneal  injection (see section 5.2.1.1).  The cerarri:
fibers  had  a  mean  diameter of 1.8 ym  and a  mean  length  of 35 urn.   In  the
inhalation study, rats  and hamsters were exposed to a concentration of 12 mg/rr"
(200  fibers/cm )  for  6 hours/day, 5 days/week for 24 months, and observed for
life.    The  aerosol  cloud  reportedly contained more  nonfibrous than  fibrous
particles; the particle ratio was  33:1 (see Table 5-4).  A summary of pulmc-a-\
effects is  presented  in Table 5-5.  Results .indicated that the incidence o*
fibrosis was slightly  higher in rats exposed to ceramic fibers   than in cc-.t-c's;
12/55 ceramic  fiber-exposed rats  developed  fibrosis  compared tc 6/5?  a"
17/125 chamber and unmanipulated controls,  respectively.  Ho*eve~. tne i-:'e=se

-------
 was not statistically significant, and  fibrosis  was  not evident in hamsters
 similarly dosed.  The incidence of bronchoalveolar hyperplasia was comparable
 between controls  and  both rats and hamsters exposed to  ceramic  fibers.   No
 tumors were observed  in  rats.   In hamsters,  however, one animal developed 3
 pulmonary spindle cell mesothelioma.
      In the intratracheal study,  the  lifespan  of hamsters dosed with ceramic
 fibers was slightly shorter  than  that of controls.   However, the lifespan cf
 rats was  unaffected  by  the  instillation of  ceramic  fibers.  A  significant
.increase (p value not reported) in the  incidence of bronchoalveolar metaplasia
 was observed in  rats  but not in hamsters dosed with  ceramic  fibers (6/22  rat;
 dosed with  ceramic  fibers versus 0/25  and  3/125 saline and cage controls,
 respectively).   No primary tumors were  observed, however.  In this experiment.
 rats and  hamsters received  2 mg of ceramic fibers in  0.2 ml of saline once
 weekly for 5 weeks and  were  then observed for life.   Cage  controls  received
 saline.
      In the intraperitoneal  study, 21/36 hamsters died within  30 days  efte-  £
 single  injection  of  25 mg of  ceramic  fibers.   Consequently, an additional
 g-OL'p cf 36 was  dosed,  and within 30 days, 15 hamsters  died.   The cause cf
 death was  reportedly  acute  hemorrhagic  peritonitis  resulting from vascular
 collapse and the  accumulation  of  fluid  in the  abdominal  cavity.   Early deatr.s
 were not observed in  rats.   The lifespans of rats and  hamsters  (surviving the
 initial  30 days  of the  study) injected  with ceramic  fibers  were  significantly
 reduced (p value  not  reported) when  compared  to controls.   Eight of 25 rats
 and 7 of 36 hamsters  dosed with ceramic  fibers  developed mesotheliomas.
      Davis et al. (1984) exposed 48  rats (Wistar AF  Han)  to  a dust clous
 containing ceramic aluminum  silicate  glass insulation  for 7  hours/day, 5 days/
 week for  12  months.   The  mean total  respirable concentration was 10  m'g/rr,
 (95 fibers/cm).  The majority of the fibers within  the dust cloud were <3  un.
 in  length  with  a diameter of  0.3 \tn.   Alveolar  proteinosis and ferruginous
 bodies were found in  the lungs of the animals exposed  to ceramic insulation  at
 the end of the 12-month  exposure period, but little or  no fibrosis was  present.
 Thirty-two months after  study  initiation, slight proteinosis was still evident
 but the majority of  the  particulate dust observed in  the lungs  at 12  months  had
 been phagocytized and cleared.   It is  interesting  to  note  that ferruginous
 bodies are generally  observed  in  hamsters, guinea pigs,  and  monkeys exposed  to
 vitreous fibers  or asbestos but are extremely  unusual in rats.

-------
     Pigott et aT.  (1981)  investigated  the potential  toxicity resulting from
the  inhalation of Saffil   alumina fibers  in rats.  These  fibers are composed
of aluminum oxide and contain about  4 percent  silica.  Three  different batches
of Saffil  alumina fibers with mean diameters of 3.0 to 3.2 pm and  mean  lengths
of 35,  62,  and 10.5 urn, respectively,  were tested.  Three batches  of "aged"
      jpi
Saffil   alumina  (i.e.,  fibers that had undergone periods of intense heat such
as occurs during actual use) were also studied.  These fibers had mean diameters
of 3.0  to  3.3 pm and mean  lengths of 53 to 57 urn.  Groups of 40 Wistar rats
(sex not specified)  were  exposed to  2.18 mg/m   (Saffil6) or 2.45 mg/m   ("aged"
Saffil  ) for  6  hours/day,  5 days/week for 86 weeks.  The pulmonary  response to
the Saffil15 fibers was generally minimal at week 27 of exposure,  and no  fibres-;s
was evident.  At study termination,  there appeared to be a slight  increase ir
                                                                        ,j
alveolar epithelialization  in the animals exposed to the  "aged" Saffil'',  a-d
this appeared to be  more  prevalent in  females.   The  epithelializatior, was
generally confined to the appearance of small foci of pigmented macrophages and
was regarded  as  a typical  response to a nuisance dust.   The authors ccncluced
           ,7
that Saffil" alumina was biologically inert.
     Styles and  Wilson  (1976) administered a single intraperitones"1 injection
of 0.2  ml of  a  10 percent suspension of Saffil" alumina fibers to  20 male a~.d
20 female  SPF Wistar  rats.   An additional group  of rats  received  a single
injection of  0.2 ml  of a  5 percent suspension  of chrysotile asbestos (positive
                      ff
control).  The Saffil   fibers had mean  diameters of 2  to 6 pm and  mean  lengths
of 6  to 100  pm.  The animals were killed  after  6 months.  Saffil   alur.ine
caused  deposition of  connective tissue  containing some collagen but did not
provoke  a  fibrosis that  was in  any  way comparable histologically  to  that
produced by chrysotile asbestos.
     Similar  results  were  obtained in a study by Pigott and Ishmael (1981) in
which groups  of  12  male and  12  female  Wistar  rats were  injected intraperitc-
                                                                  g
neally  with a single dose of 20  mg  of  one of  two types of Saffil   fibers in
2 ml saline.  Animals were then observed for 3, 6, and 12 months.   The  type  A
fibers  had  a  mean,diameter of 2.75 urn and length of 15.5 ym; the type B fibers
had  a mean  diameter of 3.7 urn  and  length of 17 urn.   Survival and  growth we'-e
                                  «
unaffected.    Both types  of Saffil  fibers evoked a mild inflammatory reaction
resulting in the formation of nodules consisting of fibers,  inflammatory cells,
and  a small amount  of collagen.   No progressive fibrosis  was evident 12 mont
after exposure.  It was interesting to note that although  the  surface p'-o
                                    c.
is,
I

-------
 and porosity  of  the two types of Saffil   fibers  differed  considerably,  the
 tissue reaction was identical, suggesting  that surface chemistry of insoluble
 fibers may not influence their reactivity.
      Styles and Wilson  (1976)  also  tested  the fibrogenic potential of Saffil'
 zirconia after administering  a single intraperitoneal injection of 0.2 ml in  a
 10 percent suspension to male and female  rats.  They  found  that Saffil  zirccnia
 also caused deposition  of  connective  tissue  that  contained some collagen, but
 to a  lesser extent  than that seen with Saffil6 alumina fibers.   The fibrosis
 that developed was not comparable histologically to that produced by chrysotile
 asbestos.   The  fibers had mean  diameters  ranging from 2  to  5  pm  and mean
 lengths of 5 to 200 urn.
 5.2.1.4  Other Manmade Mineral Fibers.  The toxicity  of several  other  synthetic
 irineral fibers has  been studied.   Titanium-containing  fibers  are  structurally
 similar to asbestos in that they  are composed of bundles of submicronic neecle-
 shaped fibers and may replace asbestos as  reinforcing  agents.  Titanates  have
 beer, shown to cause fibrosis  after inhalation exposure, intratrachesl  instilla-
 tion,  and  intraperitoneal  injection,  although the fibrosis observed was rrjcr
 less  extensive  that  observed for asbestos.   Calcium  silicate  fibers and
 wollastonite  (a  naturally  occurring  calcium  silicate  fiber)  have  also been
 used as substitutes for asbestos in insulation  materials  and as  reinforcing
 agents in  ceramics  and  other  products.  The pulmonary response was found to be
 similar to other MMMF and consisted primarily of dust-laden macrophage aggrega-
 tion with  little evidence of fibrosis.  Wollastonite  did not appear to  be
 particularly durable in the body  (Pott et al., 1987).
      Lee et al.  (1981)  conducted an inhalation  study to determine the toxic
 effects in  rats,  hamsters,  and guinea pigs of potassium octatitanate  fibers
'(also known as  Fybex  ,  which  is  a registered trademark of  Dupont, and consists
 of inorganic  titanate-reinforcing fibers).   The test material consisted prima-
 rily  of  fibers that  measured <15 \im in length.   The  investigators exposed
 46 male  Sprague-Dawley rats,  35 male albino guinea pigs, and 34 hamsters
 (strain and sex  not specified) to an atmosphere  containing either 2.9 x 10  ,
 13.5 x 106, or  41.8 x  106 Fybex6' fibers/L for 6  hours/day,  5 days/week for
 3 months.   One group  of each species exposed to  air  alone  served as  centre's.
 Dose-related decreases were  observed  in body weight gain  in  rats,
 and guinea  pigs  exposed  to  Fybex .   Dose-related  increases  in  fibroger.i:
 activity were also observed in the Fybex'-exposed animals;  howev-e-,  the  f^-
                                     5-25

-------
 resulting  from  Fybex  exposure was much  less extensive than that of asbestos at
 similar  exposure concentrations.   The fibrosis was accompanied by  dust-laden
 macrophage aggregation,  granular pneumocytic proliferation, bronchiolarization,
 and  ferruginous body production.   Rats  tended to exhibit a more severe fibro-
 genic  pulmonary reaction than either  guinea pigs or hamsters.
      Lee et al.  (1981)  also studied the  potential  toxicity of  pigmentary
 potassium^titanate (PKT)  fibers  by  exposing  46  male Sprague-Dawley  rats,
 35 male  guinea  pigs, and 34 hamsters (strain and sex not  specified)  to an
 atmosphere containing 2.0 x 10  PKT  fibers/L for 6 hours/day, 5 days/week  for
 3 months.   The  test material consisted primarily of  fibers  that were <10 urr, in
 length.    PKT  was  less  fibrogenic  than asbestos  or  Fybex .   Very  minute
 pulmonary  fibrosis occurred in rats  and hamsters  exposed to PKT, but not  in
 guinea pigs.  Other pulmonary changes were also very slight.
     The effects  of  titanium phosphate  fibers administered by intratrachesl
 and  intraperitoneal injection to rats and hamsters were  investigated by Gross
 et al. (1S77).   These fibers measured 0.2 to 0.3 urn in diameter and 10 tc 20 ur.
 in  length, and we-e given  in doses  of  2,  10, or  50  mg  of fibers suspended
 in  5 mL  of  saline.   A  slight  dose-related  fibrogenic  response to titaniur
 phespnate  was  observed  following intratracheal  instillation in  rats.   In
 contrast,  no  effects were  noted in hamsters.  Intraperitoneal injection  did not
 induce any treatment-related  lesions.
     The effects of the  inhalation of commercial  calcium silicate insulation
 materials  (mineral tobermorite) were studied  in  rats  (Bolton et al.,  1SE6).
 Groups  of 48  male Wistar  rats  were  exposed  7  hours/day,  5 days/week  for
 12 months,  and  observed for life.    Three  samples of  a  mineral tobermorite
 (Cat(OH),,$i'  0,,-hLO)  were  tested.  Fiber  diameters  ranged  from  1  to  10 prr,
   5     &   b  ID  t                                       *
.although chamber concentrations were adjusted to  10  mg/m  respirable  fibers.
 Control  animals were  exposed to filtered air.
     Dust-laden macrophage accumulation, particularly adjacent to bronchioles
 and  in alveolar spaces, was noted  in all  animals exposed to calcium silicate
 immediately  after completion of exposure (12  months).   Aggregation decreased
 with  time  during the recovery period.   Fibrosis was noted  in both control  anc
 calcium  silicate-exposed rats that survived more than 10  months after cessation
 of  exposure.   Incidences of fibrosis were  11  percent (5/45) for control anc
 9 percent  (13/139) for calcium  silicate-exposed rats.  Small  amounts of  pe-ticu-
 late  matter were observed  in mediastinal lymph nodes of  all animals exposed tc

-------
calcium  silicate.   However,  more  extensive  changes  were  noted  in animals
exposed to one calcium silicate sample denoted "A."   Rats exposed to  sample  "A"
also exhibited  a  lesion  exclusive  to  this group.  The lesion was characterized
by small  granulomatous nodules  consisting  of oiacrophage and fibroblast cells
within a  network  of reticulin and  located in  the peribronchiolar region of the
lungs.  This sample, however, was contaminated with  quartz (1  percent),  and  the
lesion may have  been quartz-induced.  Significant  increases  (p <0.01) in
lymphocyte counts were noted in rats  exposed  to calcium  silicate when compered
to controls.
     Pulmonary neoplasms were found in  2/139  rats exposed  to calcium silicate.
These were described  as  a  focal squamous carcinoma of low-grade  malignancy  and
a benign,  adenomatous  lesion.   One rat  exposed to calcium  silicate developec  a
peritoneal mesotheliome, but this  was considered to be  spontaneous because  cf
the negative response  obtained  in  a  concurrent intraperitoneal  study in whicr.
neither  fibrosis  nor tumors were  obtained after a  single intraperitonecl
injection of 25 mg calcium silicate in rats (Bolton  et a!., 1986).
     The   fibrogenic potential of intratracheal instillation of wollastonite (e
naturally  occurring  calcium  silicate)  was  compared to  asbestos  in  a  stucy
reported   by  Wozniak et al.  (1986).   Groups  of male Wistar rats receive:  a
single intratracheal instillation of  50 mg of Finnish wollastonite, crocido'iite.
or chrysotile  suspended  in 0.6 ml  saline.   Control  animals  received saline.
Wollastonite fibers  ranged from 10 to 20 urn  in  length  (the  diameter was net
reported).  Rats  receiving wollastonite were killed 3,   6, and 9 months  post-
dose, and those  receiving the  asbestos samples  were killed 3 and 6 months
postdose.
     The  wet  lung weight  of animals receiving  wollastonite  was  increased
compared  to control animals,  accounting for approximately a 21 percent increase
after 6  months and  a 19  percent  increase after 9  months.   Hydroxyproline
content  was  also  increased over controls in  animals receiving  wollastonite.
However,   lung  weights  and  hydroxyproline content were much higher in  aniir-sls
receiving  asbestos,  suggesting that  fibrotic changes  were  not  comparable
between wollastonite  and  asbestos.   Histological  changes  in the lungs of rets
receiving wollastonite were  characterized  by granulomas  consisting of  agc'eje-
tion of dust-laden macrophages  with a few fibroblast and lymphoid cells.  Tnere
was some  accumulation of dust-laden macrophages in mediastinal lymph nodes,  but
wollastonite fibers were  not found in other  tissues. There was  nc evidence c*

-------
cell  necrosis  or fibrosis  in the lungs or  lymph  nodes  of rats exposed to
wollastonite.  Conversely, Instillation of asbestos resulted in the development
of  fibrosis  in  the  lungs characterized by significant increases  in  wet lung
weight  and hydroxyproline content,  as well  as proliferation  of  connective
tissue.

5.2.2   Synthetic Fibers
5.2.2.1  Araitn'd  Fibers.   Only one chronic  inhalation toxicity study was found
in  the  available literature  for Kevlar , a commercially  produced  aramid fiber
(Stadler,  1986).  Specially  refined  Kevlar®  pulp fibers  (90 percent  <1.5 urr, in
diameter and 75  percent <20 urn in length) were used to  determine  the effects
of  chronic exposure to respirable fibers  in the  deep portions of the  lung.
The results indicated moderate pulmonary toxicity characterized by hyperplesie.
obliterative bronchiolitis with  collagen formation,  emphysema,  and development
of neoplastic lesions in the exposed rats.   The study was conducted with.groups
of  ICC  male  and  100 female  Crl:CD BR  rats exposed to the  refined fibers  at
chamber concentrations  of 0.0£,  0.32,  0.63,  or 2.23 mg/m  (equivalent to 2.4,
25.4, 111.2, or  435.5  fibrils/cm ,  respectively).   Exposures  were  for 6 hours/
day, 5  days/week for 1 (2.23 mg/m ) or  2  years  (O.DS, 0.32,  and  C.63 mg/r*).
After 1 year of  exposure to  2.23 mg/m  , animals were observed  for a recove-y
period  of  1  year.   This was  instituted after study initiation because of high
mortality  at this exposure  level.  Interim sacrifices were conducted after 3,
6, or 12 months  of exposure.
     Mortality was  significantly  higher (p <0.05)  at the 2.23-mg/m  level  than
in  controls  after  1 year of  exposure.  The majority of these deaths  were
.attributed to obliterative bronchiolitis characterized by granulomas containing
collagen fibers  that partially blocked airways in the alveolar duct region.
                                                                     g
The mortality rate  during the second year with no exposure to  Kevlar   fibers
was  comparable   to  controls.   Body weight gains  were significantly lower
                                       c
(p <0.05)  than controls for  all  Kevlar -exposed groups  after  1 year,  but  no
differences were noted  after 2 years of exposure.   Absolute and relative lung
weights were  significantly  higher (p  <0.05) for  animals exposed  to 0.63  or
2.23 mg/m  than  for controls.  Dust-laden macrophages were evident in the lungs
                                 g
of  all  animals  exposed  to Kevlar  fibrils,.although the lung architecture  was
unaffected at  the  lowest exposure  level  after 2 years.    After 3  months  of
exposure,  lungs  from the 0.63-  or 2.23-mg/m  groups exhibited type II  ce'l
     19==                           5-2-          DRA-T--D: N:~ c.:~: •-•. ::

-------
hyperplasia in the alveolar region.   This was also noted  in  rats exposed to the
0.32-mg/m   level  for 6  months.   Other lung effects noted at 0.32 mg/m  or
higher doses  were  nodular thickening of ridges at alveolar duct bifurcations
and formation of  ciliated columnar  cells in the upper  regions of the alveolar
ducts.  Rats  exposed to  2.23 mg/m  exhibited emphysema characterized by dila-
tion  of  alveolar ducts  and adjoining alveolar  spaces adjacent to blocked
airways.   This was  evident after 1  year of  exposure and  persisted after the
recovery period of  1 year.  In addition, Kevlar® fibrils  migrated  to adjacent
tracheobronchiolar lymph nodes in rats exposed to 0.32  mg/m   or  higher, altnoug-
fibrils were  not  found  in other tissues.  Neoplastic  lesions were  seen in  the
lungs of rats exposed to Kevlar  fibrils at  levels of 2.23 mg/m   (1/36 male:
and 7/56 females)  after  1 year or O.F3 «g/m  (10/69 female?) after 2 yea-s  c*
exposure.   No tumors were observed in controls.
5.2.2.2  Carbon Fitters.    Very  little information  on the subchronic and chro-ic
toxicity of carbon fibers was found  in the available literature.  Two subchronic
inna'aiion studies with  rats and guinea pigs  indicated no  pulmonary damage.
although macrophage  accumulation  was evident.   Implantation of  ca^bor.' fibers
did not elicit any unusual reaction  in bone and muscle  tissue.
     Holt (1982) reported that exposure to submicron carbon  dust via  inha iatio->
for 7  to  100 hours  did  net  induce  pulmonary fibrosis  or other pathological
effects in guinea  pigs.   Dust-laden macrophage aggregation  was noted up to 2
years after  exposure, but Perls-positive macrophages (ferruginous  bodies) *ere
not observed  before 5 months postexposure.  Similarly,  no evidence  of pulmonary
fibrosis was  noted  in rats following subchronic  inhalation  of  Celion, a PAN-
based carbon  fiber  (Owen et al., 1986).   The study was conducted with  a group
of 60 male  Sprague-Dawley rats exposed  to 20  mg/m  , 6 hours/day,  5 days/week
for 16 weeks.   The fibers had a  mean  diameter of 7 pm and   length of 20  to
60 pm.  Rats  were  killed and examined after  4, 8, 12,  or  16  weeks  of exposure;
the remaining animals were observed for a recovery period of 32  weeks.  -Controls
(60 male  rats) were  exposed to  air.  There Were no significant changes  in
pulmonary function.  Histologically, the only treatment-related  observation was
the presence  of  carbon  fiber-containing macrophages in the  alveoli.  No pul-
monary inflammation  or  fibrosis  was evident.  The authors  did  not report f:€
concentration  of  respirable fibers  or quantitate fiber  concentration  in  tne
lungs.  However, most of the carbon fibers  were  beyond  the respiratle  range
since the mean diameter  was 7 pm.   Consequently, fiber  concentrations ^f^
                                    L.

-------
the  lungs  may  not have been adequate  to  elicit  any pulmonary response  other
than macrophage accumulation.
     To evaluate  the  chronic tissue reaction to carbon fiber implants,  strand
and  powdered forms  of AS3000 nonbraided filament carbon  fibers were  inserted
or  injected  (2.5  mL in a saline suspension) intramuscularly  into the glutee.l
muscle or  wrapped around the femurs of Wistar  rats  (Tayton et a!., 1982).  The
animals were killed and examined an average of  17  months after insertion or
injection  of  the  carbon fibers.   No  inflammatory  reaction  was  observed.
Generally,  the implants  or  the powdered  fibers  were surrounded by a  thin
fibrous capsule.  A few phagocytic  cells  were  present,  and an occasional  giant
cell was  noted.   In rats receiving an implant, considerable  fiber debris we-e
found  in  the  interstitial spaces,  and smaller quantities were also present ir
adjacent  lymphatic  capillaries.  The muscle  tissue  exhibited  a slight tnicke--
ing of the endomysium, but no neoplastic changes were evident.  The wrapping cf
carbor, fibers  around  bone resulted in development  of  a thick fibrous caps-jle
around the  fibers  but did not  cause  unusual  reaction or neoplastic  chance  in
bone tissue.                                                                    I
5.2.2.3  Pclyolefrn Fibers.  Polyolefin fibers did not cause any tissue response
ether  than macrcphage  accumulation after intratracheal  or   intraperitcnea"
administration.  Styles and Wilson  (1973) conducted a study designed to investi-
gate the  relationship  between  cytotoxicity  to macrophages  i_n vitro  and  tne
j_n  vivo fibrogenesis  of several natural  and  manmade fibers.   It was  found that
polyolefin  fibers cause low i_n vitro cytotoxicity (see section 7.2.3.3) ano dc
not produce fibrosis after intraperitoneal injection into Wistar rats.  For the
intraperitoneal study,  groups  of 12 male and 12 female rats were given single
injections  of  50 mg/kg of  either  polyethylene  ("Alkathene,"  3  to. 75 urr. in
diameter)  or polypropylene  (4 to 50 urn in diameter) suspended in  physiologica"
saline.  The animals were observed  for 1 or 3 months; no lesions were evident.
     The  chronic  effects  of polypropylene fibers on rats after a  single intre-
tracheal  instillation of  1.25 mg were  investigated by Moreno and Moreno (1981).
Groups of  40  male Long Evans rats were dosed with 0.5-mL suspensions of poly-
propylene  fibers  ("spurted") or polypropylene dust (obtained  from a splitting
operation)  at  concentrations of 0.25  percent  in  a  0.1 percent Tween 60/wate-
solution.    Controls received 0.5 ml of the  Tween 60/water  vehicle.   Anircls
were observed  for 21 months.  No clinical signs  of  toxicity  or changes  ir. bcc\
weight or  organ  weight were observed  in  the polypropylene-dosed c-.-'r:a's  *ne~
                                    t.-'3~          nc :,"--<

-------
compared to  controls.   Histologically,  the  only observation considered to be
treatment  related  was the presence  of  lung granulomas  with foci containing
irregularly  shaped  birefringent  particles.  The authors considered these to De
residual components  of  the  Instilled material.   However,  it should be  noted
that 48  percent  of the animals from the polypropylene  group died during  the
first week  after instillation  presumably because of dosing errors, thus effec-
tively reducing the group size to half that of controls.

5.2.3  Minerals
5.2.3.1.  Attapulgite.  No  information  on  the subchronic  or chronic toxiciu
of  attapulgite  following  inhalation exposure  or  intrapleural   and  intra-
peritoneal  administration was found in the  available  literature.   Intratre:^6:'
instillation of 100 mg (in 100 ml saline) of attapulgite into the tracnesl Icte
cf the  lungs of  sheep demonstrated  that  this  mineral  is biologically active  ir.
broncnoalveolar  spaces  and  induces  lung lesions that are comparable to these
produced by  short  chrysotile  fibers (Begin et al.,  1927).   The attspjlgne
sarnie consisted  of fibers  having  a mean diameter  of  C.02 urr< and a mean le-.cl-
of 0.6 um  (0.25  to 4  urn),  whereas  the  mean diameter  and length  of the  chrysc-
tile fibers  were 0.08 and 0.92 urn,  respectively.   Results  of bronchos!vec'e'
lavages (BAl), collected on study days 0, 12, 24,  40,  and  60,  indicated that in
attapulgite-exposed sheep, total BAL cells/ml increased significantly (p 
-------
      Suzuki  (1982)  also  investigated the effects  of roordenite, a  zeolite
 mineral  using a mixture  of  fibrous  and granular fragments.   Five male Swiss
 albino mice were given single intraperitoneal injections of either  10 or  30  mg
 of mordenite suspended in 1 ml  of saline  and observed for life.  The granular
 mordenite was 0.33  to 5.7 urn on the long axis and 0.27 to 1.67 urn on the  short
 axis.   The fibrous  mordenite was  0.4  to 6 urn in length  and 0.05 to 0.67 urn in
 width.   Five untreated nice served  as controls.  Peritoneal  fibrotic lesions
 were observed in the mordenite-treated mice, but the effect was not as pro-
 nounced  as that  observed  in  the  erionite-treated animals.
 5.2.3.3   Talc.   Talc is  considered  to be  a  particle rather than a  fiber  and.
 therefore, the tissue response resulting from talc exposure cannot  be directl\
•compared to  that  of fibers.   Talc exposure  generally resulted  in  a typical
 granulomatous lesion  consisting of dust-laden  multinucleated foreign body
 giant cells,  as  well as  some  fibrosis  with collagen  formation  in  several  animal
 species.   In  addition to  the  development of  granulomas,  intratracheal  instilla-
 tion of  talc resulted in moderate tissue  destruction in hamsters,  and int-e-
 pe-itoneal administration has been  shown  to cause  abdominal adhesions  in  rets
 and swine.
      Wagner  et al.   (1977) exposed a group of 24 male and 24 female Wistar rats
 to a dust cloud containing 10.8 mg/m  talc  (mean respirable concentration).   A
 total  of 24  animals inhaled  the  dust for 3 months;  12 were exposed  for  6  months,
 and the  remaining  12 were exposed for 12 months.   An  equal  number o'f rats
 remained unexposed as cage controls,  and a group of  48 rats serving  as  positive
 control  animals  was  exposed to 10 mg/m  chrysotile asbestos.   Exposure was for
 7.5 hours/day, 5 days/week.  The  test material  (Italian 00000 grade) contained
 approximately 92  percent talc,  3  percent  chlorite, 1 percent carbonate
 minerals, and 0.5  to 1.0  percent quartz by weight.  The  remaining 3.0  to
 3.5 percent was  not  defined.  Particle  size ranged up  to 70 urn  but  averaged
 25 urn.   Groups of  rats were  sacrificed at 10 days  and 1 year  after  the end  of
 each exposure period, and the  remaining rats were observed  for life.
      The severity  of fibrosis produced was  similar for  the talc- and asbestos-
 exposed   groups,  ranging  from  minimal to slight.   There was  evidence of
 progression  in  severity  as the  exposure  length increased.   One of 4£ rat;
 exposed  to talc developed a  pulmonary adenoma, and  7 of 4£  asbestos-expcsea
 rats developed pulmonary  tumors.

-------
     Trautwein  and Helmboldt  (1967)  studied  the pulmonary effects  of  talc
after  intratracheal  instillation  into chinchillas.   Groups of male and  ferr.ale
chinchillas received intratracheal instillations of 2 ml of a  2 percent  suspen-
sion of  talcum powder (purified Mg^i^.'H^O;  particle  size  not specified) in
saline.   Controls  were treated similarly to test animals, and  received 2 mL  of
saline per injection.  The study consisted of three experiments.   In the first,
14  chinchillas  received injections on study days 1,  20,  50,  70,  and 90, and
were killed  11 months after the last injection.  A second group of 24 animals
received  a  single  intratracheal injection,  and  two  animals were killed  24, 4£.
72, and  120  hours, and 1, 2,  3,  4,  6, 17, and 28 weeks, postdose.  The lest
group  of  18  animals received  11 weekly intratracheal injections for 10 weeks,
and chinchillas were killed 1,  2,  and 3.5 months after the last injection.
     Several  animals died during  the study from the acute effects  of  taU
exposure.  All  of these  animals  exhibited  some form of pneumonia caused by
talc accumulation  in the lungs.   Two animals from  the  first  experiment died
3 tie\ i after  the  initial  injection  from acute suppurative pneumonia.   Five
animals died  from  the third experiment:  two from proliferative and exucatne
pneumonia after 12 and 21 days  (these animals received two and three injections
respectively),  and three  from extensive granulomatous pneumonia between 41  a-:
46 days after the  initial talc  injection.
     Results  indicated that talc  caused exudative pneumonia  that developer
into  pro!iferative, granulomatous pneumonia.   The granulomas  consisted cf
multinudeated  foreign body  giant cells that surrounded  or engulfed the talc
particles.   Several  stages of transformation were noted  in  alveolar  lining
cells  located only  in areas   of  talc deposition or  in  close proximity to
.granulomas.   Initially,  swelling of  alveolar epithelial  cells was  .noted.  IMS
was  followed by hyperplasia  and  progressed to epithelialiiation.  -Finally,
columnar  cells  appeared,  lining the affected alveoli and imparting an "adenor.a-
tous appearance."   No fibrosis was noted in talc-injected or control animals.
Furthermore,  single  or multiple injections resulted  in  the same  tissue reac-
tions.
     In a chronic  study by Stenback et al.  (1986), groups  of 48 Syrian hamster*
received  weekly intratracheal   instillations of 3 mg talc  alone or in co-iira-
tion with 3  mg benzo(a)pyrene  (BaP)  in saline  for life.   The talc sample cc--.-
sisted of 64  to 66  percent SiOp  and  34 to 36 percent McD..  and 5- percent  c^
the  talc  granules  had a particle size of <10 urn.   Dust-laden mscrcpr.eos

-------
aggregation and  accumulation  of interstitial  cells and histiocytes  were  noted
in hamsters exposed  to talc alone.  There were  also  increases  in fibrillary
material  (elastic  fibers) surrounding alveolar  and interstitial  spaces,  and
some  accumulation  of proteinaceous exudate within the alveoli.   However,  no
fibrosis  or  granulomas were  observed  in talc-exposed hamsters.   Talc alone
caused moderate  tissue destruction,  slight metaplasia, and no dysplasia  in the
tracheobronchial 2one;  woderate hyperplasia was seen in the alveolar zone.
Talc plus BaP resulted in moderate destruction, moderate metaplasia,  and  severe
dysplasia in the tracheobronchial zone, and severe hyperplasia  and dysplasia  ir.
the  alveolar  zone.   Talc alone did  not  induce tumors in hamsters but,  when
given  in  combination  with  BaP, was  found to  be an effective  cocarcinogen,
inducing  an increased  incidence of tumors  in both the upper  and  lower  respira-
tory  tract when  compared to talc or BaP alone.  There was a  high incidence of
peripheral (alveolar)  lung tumors, and  adenocarcinomas  were the predominant
turner type.
     In a study  by Styles and  Wilson  (1973) that was designed to compa-e  tne
_i_n vitro  cytotoxicity with  i_n vivo fibrogenesis  of  several synthetic  arc
mineral fibers,  intraperitoneal injection of  50 mg talc/kg  talc into Wistar
rats  (sex not  specified) resulted in the development  of a granulomatous  res:-
tion  that consisted  of the aggregation of foreign body giant cells containing
talc  particles 1 month after administration.   The lesions were  still  evident
3 months  after injection,  but no fibrosis was evident.   In  this study,  talc
was  found to  have  low j_n vitro cytotoxicity resulting in death of less  than
2 percent  of   peritoneal macrophages  following  phagocytosis   (see  also
section 7.2.5).
     Kaiser et al.  (1982)  reported the development of foreign  body granulorr.as
with  adhesions  in  male and female Wistar rats 12 weeks  after  intraperitonea1
injection of  talcum  (dose and  composition not specified).  The development of
foreign body  granulomas  has also been reported  in swine (Migaki and Gardner.
1969).  The lesions  found in the  inguinal canal  of four castrated  male  swine
consisted of  talc-containing epithelioid  and  giant  cells,  lymphocytes,  and
fibrous tissue.   The pigs were  thought  to  have  been  exposed to  talc  during
castration, but  the  exposure period was not reported.   The  peritoneal tissue
response  to talc was also studied in  rats, by  Pelling and Evans  (1965).   Tel:
(50 mg) was implanted  surgically into the peritoneal  cavity  of 60 male Sp-acje
Dawley rats.   Sham-operated rats served as controls.   Animals were ki'lec 2, 4

JL'V 19=:                           5-3-           D^rT--D:  Nl"  QJ-'E C:  C":

-------
8, 13,  26,  and 52 weeks  after implantation  and  examined.  Talc produced exten-
sive granulomatous  peritonitis  within 2 weeks that persisted until  week  13.
Numerous focal  lesions described as  discrete  pale  nodules were observed in all

animals  receiving talc  (100  percent incidence).   These nodules  were  still
apparent 52 weeks  after exposure.   Both control  and talc-exposed-animals
developed adhesions.  The animals receiving talc,  however,  had a  significantly
increased incidence (p <0.001)  of adhesions when compared to  control rats over
the entire  time period.   Histologically, the lesions consisted of  granulorr.es
containing  dust-laden  macrophages and  dust-laden  multinucleated  giant cells
with some collagen  formation.
5.2.3.4.  Vermiculite.    No information on the subchronic or chronic toxicity o*
vermiculite was found in  the available literature.
5.3  REFERENCES
Ee:v E.  D. ;  Feldrr.an,  H.  A.; Brain, J. D. ; Smith, T. J. ; Hallock, K. ; Ge~sc",
     E.  (15c7) The pulmonary toxicity of talc and granite dust as estimate fro-
     an jj2 vivo hamster bioassay. Toxicol. Appl. Pharmacol.  87:' 222-234.

Begin,  R. ;  Kasse,  S. ;  Ro1a-P1es2C2ynski,  M. ;  Geoffrey,  M.;  Martel,  K. ;
     Desmarais,  Y. ;  Sebastien,  P.  (1967) The  lung  biological activity  of
     American attapulgite. Environ. Res. 42: 328-339.

Belter,, R.  E.; Addison,  J. ;  Davis,  J.  M. G.;  Donaldson, K. ;  Jones, A. D. ;
     Killer,  B.  G.; Wright,  A.  (1986) Effects  of inhalation  of  dusts fro*
     calcium  silicate insulation materials  in  laboratory rats. Environ.  Res.
     3S: 26-43.

Davis,  J. M.  G.  (1976)  Pathological aspects  of the  injection of glass fiber
     into the  pleura!  and peritoneal cavities  of  rats  and  mice.  In: Occupa-
     tional exposure to fibrous  glass: proceedings of a symposium; June 1S74;
     College  Park,  MD.  Cincinnati, OH: U. S. Department of Health, Education,
     and  Welfare,  National  Institute  for  Occupational  Safety and  Health;
     pp. 141-149;  HEW  publication no.   (NIOSH)  76-151.  Available  from:  NTIS,
     Springfield, VA: PB-258869.

Davis,  J.  M.  G.;  Addison,  J.; Bolton, R. £.;  Donaldson, K.;  Jones, A. D. ;
     Wright,  A.  (1984) The  pathogenic effects  of fibrous  ceramic aluminu-
     silicate glass administered to rats by  inhalation or peritoneal  injection.
     In: Guthe, T., ed.  Biological effects of man-made mineral fibres: proceed-
     ings of  a  WHO/IARC conference in  association with JEMRB and TIKA, v.  2;
     April 1982; Copenhagen,  Denmark.  Copenhagen, Denmark:  World Health Organ-
     ization, Regional Office for  Europe; pp. 303-322.
                                    5-35

-------
Drew,  R.  t.; Kuschner,  M.;  Bernstein, D. M.  (1986)  The  chronic effects  of
     exposure of  rats to  sized glass  fibers.  Presented at:  international
     symposium on  man-made  mineral fibres in the working  environment;  October;
     Copenhagen,  Denmark.  Copenhagen,  Denmark: World  Health Organization,
     Regional Office for Europe.

Drew,  R.  T.; Kuschner,  M.;  Bernstein, D. M.  (1987)  The  chronic effects  of
     exposure of rats to sized glass fibers. Ann. Occup. Hyg. 31: 711-729.

Goldstein, B.;  Rendall, R.  E.  G.;  Webster,  I. (1983) A comparison of the
     effects  of  exposure of baboons to  crocidolite  and fibrous-glass  dusts.
     Environ. Res. 32: 344-359.

Gross, P. (1976) The effects of fibrous glass dust on the lungs  of animals.  In:
     Occupational  exposure   to  fibrous glass:  proceedings  of a symposium.
     Cincinnati, OH:  Dept.  of Health,  Education and  Welfare;  NIOSH Publ. No.
     76-151;  Available from: NTIS, Springfield, VA; PB-258869; pp. 166-179.

Gross, P.; Westrick,  M.  L. ; McNerney, J. M. (1960) Glass dust:  a study of  Us
     biologic effects. AMA Arch. Ind. Health 21: 22-35.

Gross, P.; Kaschak,  M.;  Tolker, E.  B.;  Babyak, M.  A.; de Treville, R. T.  P.
     (1970)  The  pulmonary reaction  to high concentrations  of  fibrous glass
     dust: a .preliminary report. Arch.  Environ. Health  20: 696-704.            I

Gross, P.; Kociba,  R.  J. ;  Sparschu, G.  L.;  Norris,  J.  M.  (1977) The bioic:ic
     response to  titanium phosphate:  a  new  synthetic  mineral  fiber.  A^c!-..
     Pathol.  Lab.  Med. 101:   550-554.

Henderson, J. D. ;  Mullarky,  R.  H. ;  Ryan, D. E. (1987)  Tissue biocompetibiiity
     of Kevlar aramid  fibers and polymethylmethacrylate composites in  rabbits.
     J. BiomecJ.  Mat. Res. 21: 59-64.

Holt,  P.   F.  (1982) Submicron carbon dust in the guinea pig lung. Environ. Res.
     28:  434-442.

Kaiser,  V.  W.;  Otten, G.;  Kohnlein, H.-E.; Kitzinger, H.  (1962) Unspecific
     reactions  of web  excited  by  surgical  glove  powder.   Fortschr.  Med.
     25:  1213-1216.

Kuschner, M.; Wright,  G.  (1976) The effects  of intratracheal instillation of
     glass fiber  of  varying sizes in guinea pigs.  In: Occupational exposure to
     fibrous glass:  proceedings  of a symposium; June 1974;  College Park, K2.
     Cincinnati,  OH:  U.  S.  Department of  Health,  Education, and  Welfare,
     National Institute  for  Occupational Safety and  Health;  pp. 151-168; HEw
     publication  no.  (NIOSH) 76-151.  Available from:  NTIS, Springfield,  VA;
     PB-258869.

Le Bouffant,  L.; Daniel, H.; Henin,  J.  P.; Martin, J. C.; Normand, C.; Tichcux,
     G.;  Trolard,  F.  (1986)  Experimental study on long-term effects of M'-V-IF on
     the   lung  of  rats.  Presented at:   international  symposium  on  man-rr.aoe
     mineral  fibres  in  the  working  environment; October; Copenhagen,  De"irr;5"k.
     Copenhagen,  Denmark:  World  Health  Organization,  Regional  Office  fcr
     Europe.

-------
Le Bouffant, L. ; Daniel, H. ; Henin, J. P.; Martin, J.  C. ;  Normand, C. ;  Tichoux,
     G. ; Trolard,  F.  (1987)  Experimental  study  on long-term  effects of  inhaled
     MMMF on the lungs of rats. Ann. Occup. Hyg. 31:  765-790.

Lee, K.  P.;  Barras,  C.  E. ; Griffith,  F.  D. ;  Waritz,  R.  S.  (1979) Pulmonary
     response to glass fiber by inhalation exposure.  Lab.  Invest.  40:  123-133.

Lee, K.  P.;  Barras,  C.  E. ; Griffith,  F.  D. ; Waritz, R. S. ; Lapin,  C. A.  (195:)
     Comparative pulmonary  responses  to  inhaled inorganic  fibers  with  asbestos
     and fiberglass.  Environ. Res. 24: 167-191.

Lee, K.  P.;  Kelly, D.  P.; Kennedy, G.  L. ,  Jr.  (1983) Pulmonary  response  to
     inhaled Kevlar  aramid synthetic  fibers in  rats.  Toxicol.  Appl. Pharmacc"1. .
     71:  242-253.
Makisato, S. ; Paavoleinen, P.; Holmstrom, T. ; Slatis, P. (1984) Tissue
     after  implantation  of carbon  fiber and polypropylene. An
            in rats. Acta Orthop. Scand. 55: 675-676.
McDonnell, E. E.; Wagner, J. C. ; Skidmore, J. W. ; Moore, J. A. (1984) A ccns-
     ative study  of  the fibrogenic and carcinogenic effects  of  U1CC Ca-ac-'c-
     chrysctile E asbestos and glass microfibre (JM100). In: Biological Effects
     of Ksn-made  Mineral  Fibres:  proceedings of a WHO/IARC  conference;  v.  2:
     April  1962.  Copenhagen,  Denmark:  World Health Organization,   Reric-e"
     Office for Europe; pp. 234-250.

Mic-ski, G. ;  Garner,  F.  M.  (196S)  Talc-induced granulomas in swine.  J.  Ar.  vet.
     Mec Assoc.  155:  1595-1596.

Mitchell,  R.  I.; Donafrio, D. J. ;  Moorman,  W.  J.  (1986)  Chronic  inhalation
     toxicity of  fibrous  glass  in rats and  monkeys.  J.  Am.  Coll.   Toxicol.
     5: 545-575.

Moorman, W. J. ; Mitchell, R. I.; Mosberg, A.  J. ; Donafrio, D. J.  (1SE5) Chrcr.ic
     inhalation toxicology  of  fibrous glass  in rats  and monkeys.  Presentee at:
     6th  International  Symposium  on inhaled particles; September; Morgantown,
     VA. Cincinnati, OH: National Institute for Occupational Safety  and HeaHn,
     Division of  Respiratory Disease Studies.

Moreno, 0. M. ;  Moreno,  M.  T.  (1980) Twenty-one month observation after single
     intratracheal  insufflation [unpublished  study].  Spinnertown,   PA:  M.e.
     Research Laboratories, Inc.; Project no. 78-2806.

National  Institute  for Occupational  Safety  and Health.  (1977)  Criteria  for
     a  recommended  standard ____ occupational  exposure  to fibrous  glass.
     Cincinnati,  OH:  U. S. Department of Health, Education, and Welfare, Public
     Health  Service,  Center for  Disease Control; DHEW (NIOSH) publication
     no. 77-152.  Available from: NTIS, Springfield, VA;  PB-274195.

Neugebauer,  R. ;  Helbing,  G. ;  Wolter, D. ; Mohr,  W. ;  Gistinger, G.  (IBS'.) 7r>e
     body  reaction to  carbon fiber particles implanted  in the med'jllarv s;c:e
     of rabbits.  Biomaterials  2: 182-184.

-------
Owen, P.  E.;  Glaister, J. R. ; Ballantyne,  B. ;  Clary,  J.  J.  (1986)  Subchronic
     inhalation toxicology of carbon fibers. JOM J. Occup. Med.  28:  373-376.

Parsons,  J.  R. ;  Bhayani,  S. ; Alexander, H. ;  Weiss,  A.  B.  (1985)  Carbon  fiber
     debris within the  synovia! joint. Clin. Orthop.  196:  69-76.

Pelling,  D.;  Evans,  J. G. (1986) Long-term peritoneal tissue response in rets
     to  mould-release agents and lubricant powder used on surgeons'  gloves.
     Food Chem. Toxicol. 24: 425-430.

Pickrell, -J.  A.;  Hill, J. 0.; Carpenter,  R.  L.;  Hahn, F.  F.; Rebar,  A.  H.
     (1983)  In  vitro and in vivo response  after  exposure to man-made  mineral
     and asbestos insulation fibers. Am. Ind.  Hyg. Assoc.  J. 44:  557-561.

Pigctt,   G.  H. ;  Ishmael, J.  (1981) An assessment of the fibrogenic potential  c*
     two refractory  fibers by intraperitoneal  injection in rats.  Toxicol. Lett.
     8:  153-163.

Pigott,   G.  H. ;  Gaskell, B.  A.;  Ishmael,  J.  (19£1) Effects of long term ir.-.ra-
     tion of alumina  fibres in rats. Br. J. Exp. Pathol. 62: 323-331.

Pott, F. ; Roller,  M. ; Ziem,  U. ; Reiffer, F. J.; Bellmann, B.; Rosenbruc*. V. ;
     Huth,  F.  (1987) Carcinogenicity studies on  natural  and man-made fioe-s
     w;th the  intraperitoneal test  in rats.  Proceedings  of a  Synpcsiir c"
     rrineral  fibers  in  the  non-occupational  environment;  Lyon,  Frar.ce;
     September 8-10.

Schepers, G. W. H. ;  Delahant, A.  B.  (1955) An experimental study of the ejects
     of  glass wool  on  animal lungs. AMA Arch.  Ind. Health 12: 276-27S.

Smith,  D. M. ;  Ortiz, L. W.; Archuleta, R.  (1984) Long-term  exposure of Syria-
     hamsters  and  Osborne-Mendel rats  to  aerosolized 0.45  urn mean diameter
     fibrous glass.  In:  Guthe,  T. , ed. Biological effects of man-made mine re"
     fibres: proceedings of a WHO/IARC conference in association with JEMRE ar.d
     TIMA,  v.  2; April 1982; Copenhagen, Denmark. Copenhagen, Denmark:  World
     Health Organization, Regional Office for Europe; pp.  253-272.

Smith,  D. M. ;  Ortiz, L. W. ; Archuleta, R.  F.; Johnson, N. F. (1986) Long-terr
     health  effects  in hamsters  and rats  exposed chronically to  .man-made
     vitreous fibers.  Presented at:  international  symposium  on man-made mineral
     fibres  in  the  working  environment;  October;   Copenhagen,  Denmark.
     Copenhagen, Denmark: World Health Organization,  Regional Office for Europe.

Smith,  D. M. ;  Ortiz, L. W.; Archuleta, R.  F.; Johnson, N. F. (1967) Long-te--
     health  effects  in hamsters  and rats  exposed chronically to  man-made
     vitreous fibers.  Ann. Occup. Hyg. 31:  731-754.

Stadler, J.  C.  (1986) Two-year inhalation  study  in  rats  with Kevlar ficn'ls
     [unpublished study].  Michigan:  E.I.  du Pont  de  Nemours and  Company, Inc.;
     Report no. 290-86; MR no. 4442-001.

Stenback, F. ;  Wasenius, V.-M. ; Rowland, J. (1985) Alveolar anc1  interstitial
     chanoes in silicate-associated  lung tumors in Syriar,  hamsters. Caice- Res.
     Mono? 2: 195-213.

-------
Styles,  J.  A.; Wilson,  J.  (1973)  Comparison  between j_n vitro  toxicity  of
     polymer  and  mineral dusts  and their  fibrogenicity.  Ann.   Occup.  Hyg.
     16: 241-250.

Styles,  J.  A.;  Wilson, J. (1976) Comparison between  in vitro toxicity of twc
     novel  fibrous  mineral dusts  and their tissue reactions i_n vivo. Ann.
     Occup. Hyg. 19: 63-68.

Suzuki,  Y.  (1982)  Carcinogenic and fibrogenic  effects of zeolites:  preliminary
     observations. Environ. Res. 27: 433-445.

Tayton,  K.; Phillips, G.; Ralis, 2. (1982) Long-term effects  of  carbon fiber  on
     soft tissue.  J. Bone Joint Surg. 64: 112-114.

Trautwein,  G. ;  Helmboldt, C.  F.  (1967) Experimental pulmonary talcum granule's
     and epithelial hyperplasia in  the chinchi.Ha. Path. Vet. 4: 254-267.

Wagner,  J.  C. ;  Berry,  G. ; Cooke, T. J. ; Hill, R. J. ;  Pooley, F. D. ;  Skic-r^e.
     J.  W.  (1S77)  Animal experiments with talc.  In: Walton,  W.  H., ed. Inre'e:
     Particles  IV: v. 2.  Oxford, England: Pergamon Press; pp. 647-654.

Wagner, J.  C.;  Berry, G.  B.; Hill,  R. J.; Munday, D.   E.; Skidmore,  J.  W.  (19:-;
     Animal  experiments with MMM(V)F--effects of inhalation and intra^le^'a"
     inoculation  in  rats.  In:  Guthe, T. , ed.  Biological effects of  n-.£--rr£ce
     ir.inei-ai  fibres: proceedings  of a WHO/IARC  conference in association *~'~
     JEKnB  and TINA,  v. 2;  April  1962; Copenhagen,  Denmark.   Copenr.ace-,
     Denmark:  World  Health   Organization,   Regional  Office for  Europe;
     pp. 209-233.

Woznisk,  H. ;  Wiecek,  E. ;  Tossavainen,  A.;  Lao, I.;   Kolakowski,  J.   (1985;
     Comparative  studies of  fibrogenic properties of wollastonite, chrysctile
     and crocidolite. Med. Pracy 5: 288-296.
                                      -

-------
                  6.  DEVELOPMENTAL AND REPRODUCTIVE TOXICITY
     A review  of  the literature revealed that very  little  research  has  bee-
conducted  on  the  developmental and reproductive toxicity of manmade  fibers,
synthetic fibers,  and minerals.
6.1  MANMADE MINERAL FIBERS
     The only  available  information  on  the  teretogenicity  of mineral fibers  is
a  letter  by Nolen  and Dierckman (1963) and a  report  by the Food anc! Drug
Research Laboratories,  Inc.  (1973).   The report presented the  results  o' fc-"
te-strgevlcity  studies  conducted with  a fine  white powdered sodium silicc-
aluminate material.   The composition and dimensions of the fibers used  were net
reported (Food and Drug Research Laboratories,  Inc., 1973).  However,  Ncler' a re
Dierckman (19S3) stated that the above studies  were conducted with a food-c-ase
sodium  aluminosilicate  with an  Na:Al:Si  ratio  of 1:1:13.   The  studies  were
conducted with CD-I mice,  Wistar  rats, golden  hamsters,  and  Dutch-belted
rabbits.  Adequate  numbers  (23  to  32  mice, rats, and  hamsters  and 11  to
18 rabbits) of pregnant females were dosed,  by  oral intubation,  daily  beginning
with day 6  of  gestation for 9 to 13 consecutive days.  All  animals received
oral doses  of  16.0,  74.3,  345.0, or  1600 mg/kg  body weight.  Dams were weighed
at  various  intervals during  gestatip.n  and wne  then  subjected to  cesarear.
section upon completion  of pregnancy.   The body wights of the live pups were
recorded, and  the urogenital tract of each dam  was examined.   Fetuses  were also
examined for external  congenital abnormalities.   One-third of the fetuses  frorr
each  litter underwent detailed  visceral examinations,  whereas  the  remaining
two-thirds  were  cleared,  stained, and  examined  for  skeletal  defects.   The
results indicated no clear compound-related effects on  reproductive indexes.
No  maternal  or fetal toxicity was noted,  and   the number of variations  anc
abnormalities  seen  in  either soft or skeletal  tissues  of test  fetuses  were net
different frorr: those observed in sham controls.

July 19EE                          €-1          DRA-T —SO  N:T C-1":  ^  ::"E

-------
6.2  SYNTHETIC FIBERS
     No information  on  the  developmental or  reproductive toxicity of synthetic
fibers was found in the available literature.
6.3  MINERALS
     No information  on  the  developmental  or reproductive toxicity of attapul-
gites, talc,  or  vermiculite was found in the available literature.   Nolen enc
Dierckman  (1983)  reported  results  of  a  teratogenicity study with synthetic
zeolite (Type A)  tested in Sprague-Dawley rats and New  Zealand rabbits.  The
zeolite contained  15.8,  19.0,  and 20.1 percent Na, Si,  and  Al, respectively.
It was adrinistered  in distilled water by gavage to rats  at doses  of  74 or Ifl!
mg/kg body weight  on days 6 to 15 and to rabbits at 74,  345, or 1600 mg/kg o~
days 6 to  18.  Zeolite  A produced no  adverse  effects on  the  dams,  the  errc-ycs,
or the fetuses in either species at any of the doses tested.
6.4  REFERENCES

Food .and Drug  Research Laboratories,  Inc.  (1S73)  Teratologic  evaluation  cr"  rD-
     71-45 (sodium silicoaluminate).  Rockville, MD:  U.  S.  Department of Hea":tn.
     Education,  and  Welfare,  Food and Drug Administration; contract  nc.  FDA
     71-260.  Available from-. NTIS, Springfield, VA;  PB-223S10.
Nolen, G. A.;  Dierckman,  T. A. (1983) Test for aluminosilicate teratogenicity
     in rats. Food Chem. Toxicol.  21:  697.

-------
              7.  MUTAGENICITY, CYTOTOXICITY,  AND CARCINOGENICITY
7.1  MUTAGENICITY
7.1.1  Manmade Mineral Fibers
7.1.1.1  Fibrous Glass.  Glass  fibers  (0.12  and 1.9 pm in diameter) sho*  nc
evidence of mutagenicity  in bacterial  assays.   Studies with eukaryotic ce^s
sho* that rm'crofibers with average diameters  of 0.1 to 0.3 pm induce chrcmcsc-e
aberrations in  two hamster  cell  lines, whereas  thicker  glass fibers wit1-.
average diameters of 0.8 pm are much less potent.   Chamberlain and Tarmy  (1977)
tested glass fibers in bacterial strains and  found no evidence of  mutagenic'ty.
Fibers were Johns Manville codes 100 and 110  with mean particle sizes (dianete-
x length) of 0.12 x 2.7 uir, and 1.9 x 26.0 urn, respectively.   No reverts--..;  *5~e
found in Escherichia  col i  W-2uvrA or Salmonella typhimurium strains TA152; cr
TA153E with or  without S9  activation at levels  of  0.1  or  1 mg/plate.  However,
as noted by the authors,  bacterial  systems may  not be  appropriate  for assaying
the mutagenicity of  particulates,  since the  fibers do  not enter  the bacte'ia
as  they do  eukaryotic  cells.   Crocidolite,   chrysotile,  amosite,  and
anthophyllite were inactive in the same systems at the same dose levels.
     Brown  et   al.  (1979a) tested  glass  fibers (codes  110  and  100)  for
genotoxicity in  Chinese  hamster V79-4 cells.  The  respirable  fraction of  the
coarser material (110R)  was more cytotoxic than the  total fraction (HOT); a
50 percent  reduction  in cloning  efficiency  occurred  at  20 pg/mL  110R when
compared to 50  g/mL  HOT.   This concentration  also caused  a significant
(p <0.01)  increase  in chromosomal  fragments  and breaks,  and  a nonsignificant
increase in gaps.   The distribution of fibers  with a  diameter  of  0.1 to 0.3 prr
and lengths of  2 to 5 pm or 5 to 10 pm was 33  and  19  percent  for  110R, respec-
tively.   Results  on  chromosomal  aberration  with  code  100 fibers  were not
presented.   Sincock  et al.  (1982) exposed Chinese  hamster ovary  Kl cells  for
48 hours to 0.01 mg/mL of  code 100  and  110  glass  fibers and found that  code
100 induced chromosome aberrations  (all chromosome breaks)  in 14 percent  cf
the scored metaphases as  compared to a 2 percent aberrant all  frequency  ir. the
control  group.   The code 110 fibers  produced negative results.
•>• •>•• lest                            7-1         DRArT--DD N" C.'I'H t~- --"-

-------
     Hesterberg  and Barrett (1984) studied glass fiber codes  100 and 110 for
 their  ability to transform Syrian hamster embryo cells  in  culture.  Code 100
 glass  fibers  had an average diameter  of  0.13 ±  0.11 urn,  ar>d  code  HO  had an
 average  diameter of 0.8 ± 0.06 urn.  The average length of  the thinner  fibers
 was  9.5 ± 0.7 urn.   The  thin  code 100 glass fibers were  about as  active  as
 chrysotile, on a weight  basis,  in  inducing morphological transformation.  There
 was  a  good linear  correlation  between  dose  and  transformation frequency, and
 cytotoxicity  and cell  transformation frequency were also correlated.   Transfo'r-
                                                   2
 mation  frequency was about 1.2 percent at 0.5 g/cm  of  code  100 glass  fibers
                             2
 and  1.6 percent  at 0.5 ug/cm   chrysotile.  Code 110 thick  glass  fibers we<-e
 20 times  less potent on a basis of weight in their  cytotoxic  effect  and induc-
 tion  of transformation.  Milling  of  code 100 glass beads  (10 minutes  in  5-
 electric  blender)  reduced the  average  fiber length from 9-5 to 1.7 pm but hac
 little  effect on diameter (0.11 versus 0.13  urn); milled fibers had  decreased
 cytotoxicity  and a  10-fold  decrease in transformation ability.  Mining  of
 chrysotile  also  destroyed  transforming ability, but these fibers tended  to
 clump after size reduction.
 7.1.1.2   Mineral Wool.   No information on the mutagenicity  of mineral  woo"  was
 found in the  available  literature.
 7.1.1.3   Ceramic Fibers.   No  information on the  mutagenicity  of ceramic fibers
,was  found  in  the available  literature.

 7.1.2   Synthetic Fibers
     No  information on  the mutagenicity of aramid, carbon, or  polyolefin fibers
 was  found  in  the available  literature.

 7.1.3   Minerals
 7.1.3.1   Erionite.   In vitro cytogenetic studies conducted by Palekar  et el.
 (1987)  indicated that  erionite fibers, over a  mass dose  range  of 10 to
 100  ug/ml,  induced  significant but not  clearly dose-related  increases  ir
 aneuploidy  and  polyploidy in V79 Chinese hamster lung  cells.   Although  the  per
 mass dose  response was  comparable to that elicited by the two asbestos fibe-s
 included  in the  study  (UICC chrysotile  and U1CC  crocidolite),  the findings or a
 per  fiber basis  suggested  that erionite  was more reactive  than  the  otne-
 mineral  fibers.   The same investigators found that  100 pg erionite/mL  caused a
 significant increase in  chromosome aberrations (chromatid breaks and frac^e'ts)

-------
however,  the  erionite  was less effective in inducing  structural  aberrations
than the asbestos fibers.
     Poole  et  al.  (1983) found that Oregon  erionite  of  respirafcle  size was
genotoxic  in  cell  transformation  and unscheduled DNA  synthesis (UDS) tests.
The median  length and diameter of  fibers was  1.7 x 0.2 urn;  all  fibers were  less
than 1.0 pm in  diameter, and 80 percent were <4 \im in  length.  Type III trans-
formed  foci of C3H10T  1/2  cells  were  -found in 4/20  flasks at 20 pg/mL*1
erionite;  at  this  level  46  percent of the cells  survived, and the calculatec
transformation  frequency was  0.6  foci/flask  as compared to 0/20  foci  in  the
control  group.   In a second experiment,  levels of 15, 20, 25, and 30 pg/rrL
erionite were tested.   There were Type III  foci  in 6/12  (50 percent)  flasks
(0.83/flask) at 20 pg/mL, but increased cytotoxicity  and decreased transfo-~a-
tion were noted at higher levels.   Erionite was, therefore, considered  positive
in this  test  system.   Stimulation of UDS  in  C3H1DT 1/2 cells and  in A54S  hi.r=->
cell cultures by  erionite was also found at levels between  100 and 200 ug/rcL
(Poole et al., 1983).
7.1.3.2   Attapulcjite.    Dem'zeau  et al.  (1985a) reported  the absence cf
genotoxic  effects  when  samples   of  attapulgite, xonolite,  and  synthetic
sepiolite were  tested  for UDS in  primary  cultures of  rat hepatocytes at le.e's
of 1 and  10 pg/mL.   Xonolite, however, was  found to  reduce  UDS stimulate:: by
2-acetylaminofluorene (0.05 and 0.25 g/mL) in the rat  hepatocyte system.

7.1.4  Talc
     Lord  (1978)  reported that Litton Bionetics, Inc., evaluated talc  for
mutagenicity in 1973.   There were no increases in gene mutation in Salmonella
or Saccharomyces  at  200 mg/mL, and no  increase was reported for  chrome-seme"!
aberrations in  rat bone marrow cells or  for dominant lethal mutation  in  the
germinal  cells  of  male  rats exposed to 30, 300, or 3000 mg/kg.  Study  details
were not available.

7.1.5  Vermiculite
     No  information  on the  mutagenicity of vermiculite was found  in the avail-
able literature.

-------
7.2  IN VITRO CYTOTOX1CITY
7.2.1  Introduction
     In vitro cytotoxicity screening tests are rapid,  cost-effective,  reproc'jc-
ible, and  sensitive  methods  of determining the potential  of mineral  fibers to
induce cellular  damage in cultured mammalian cells.  The  available  assays
employ a  variety of  cell  systems  (erythrocytes,  V79-4 Chinese hamster  lung
cells, A549  type II  alveolar  human tumor cells,  and  P388D1 macrophage-type
cells) to  measure critical endpoints of  cytotoxicity.  The  assays  correlate
well with  _in  vivo studies, provide valuable  information regarding the physics"
and/or chemical  modes of  cytotoxicity,  and are considered sensitive predictors
of  potential   mineral-fiber-induced  fibrogenicity  and/or  carcinogenicity
(Mossman et al., 1985).
7.2.1.1  Procedures  Available  To Measure  Cytotoxicity.  Much of the vn  vitrs-
cytotoxicity research examines fiber-induced membrane disruption.   Biologically
active fibers that lyse erythrocyte membranes, causing a release of hemcclcbi r,,
are said to be hemolytic.   Biologically active fibers that disrupt meir.bra-es i-
other mammalian  cells,  such  as fibroblasts and macrophages, are classifies' as
cytotoxic.   For  these cells,  the effects  of  the  test agent on intracellulc-
menbranes  can  be assessed.   For  example, lactic  dehydrogenase release  is
indicative of  adverse effects  on plasma membranes.   Lysosomal  membrane  disrup-
tion  is  indicated  by  the release  of  B-glucurcnidase,  acid  phosphatase,
p-N-acetylglucosaminidase, or  B-galactosidase.  Lysis of plasma  and  nuclear
membranes  or  a  breakdown  of  DNA polymer  can  be  determined  by the release of
tritiated  thymidine-labeled  DNA into the cell culture medium  (Richards  and
Hunt, 1983).
     Compound effects  on  cell  viability are measured by trypan blue exclusion
and  fluorescein  d'iacetate uptake.    The inhibition of colonial growth  assay
developed  by Chamberlain and Brown (1978) measures the ability of cells  treated
with fibers to  form colonies  (a colony  is an aggregate  of 50  or  more  cells
arising from  a  single progenitor).  Cytotoxicity can  also  be  measured  by tne
ability of test substances to promote  morphological  alterations  in membrane
structures  (e.g.,  giant  cells)  or decrease  macrophage  phagocytic activit;-
(Warheit et al., 1984).  Although these i_n vitro assays have many applications,
they are not without  limitations.

-------
  glass fibers incubated with trachea! mucosa  cells  from 6- to 8-week-old male
  Syrian hamsters produced a significant (p <0.05) cytotoxic response (Woodworth
  et al.,  1983).   Fifty percent  of the fibers  had  diameters <0.2  urn,  and
  94 percent were no greater than  1.0  urn wide.  The majority of the  fibers were
  long, with 54, 78, 87, 94, and 100 percent having lengths <10,  20,  30,  50, er.c
  100 urn,  respectively.  Approximately 1  and 2 percent  of  the  cultured  cells
  either died or were  sloughed  off following exposure to 4 or 16 mg  fibers/ml,
  respectively.   A similar response was recorded  for  short  crocidolite asbestos
  fibers.   A nonfibrous analog  (glass  particles)  failed  to  produce any signifi-
  cant changes in the  cell  cultures.   The  authors  suggested that  fiber geometry
  is critical  in  evaluating  the  toxic potential of  fibers.
       Maroudas  et  al.  (1973)   incubated AAA glass fibers  (average  thickness
  0.6 um;  length, <10  to 320 um)  with  cultures of 3T3 or BHK21 fibroblasts for
  6 days.   A sigmoid relation was  observed between length and the probability cf
  inducing growth.    Both  cultures showed maximum  growth  on 200  urn  and longer
  fibers and no growth  on  fibers  with  lengths of 20 urn or  less.   These resets
  correlated with u> vivo  studies  showing increased incidence  of  mesotheliorr.as  in
  rats treated intrapleurally with AAA glass fibers  as the  percentage  of  longer
  fibers (20 to  320  urn) in  the  implants increased.
       The authors  postulated that fibers  with lengths  ^40 um stimulate fibrc-
  blastic  growth i_n  vitro  by  inducing  an anchorage-dependent response (normal
  anchorage-dependent cells  cannot grow in  suspension  unless the  cells  can  attach
  to solid substrates  above a  critical mass).   The induction of  this  response
  i_n vivo may then  lead to  the development  of mesothelioma.  Persistent hyper-
  plasia in fibroblasts or in reticuloendothelial cells may therefore constitute
  a preinitiation and  is probably  an  essential stage in the promotion of solio
* implant-induced carcinogenesis (Maroudas  et al.,  1973).
       Vincent et al.  (1983,  in  abstract form)  compared the  cytotoxicity of  glass
  fibers (codes 100, 108, and 110) toward human fibroblasts with that of several
  forms of asbestos  (UICC amosite, anthophyllite,  crocidolite, and chrysotile A
  and B),  and found  that long,  thin glass fibers were  as  toxic  as  asbestos  fibers
  of the same dimensions.   Cytotoxicity was evaluated  by  measuring cell viability
  and  the release  of  labeled cellular DNA  into  the  culture medium.    The  two
  active materials  were the similarly  sized UICC  chrysotile B and hand-milled
  code 100 glass fiber.  Results  of the viability assays showed  that  100 ys'ir.L
  UKC chrysotile B  or code  100  glass  fibers induced  death in 40  to 60  percent o*

-------
  the  original  fibroblast population.  The release of radioactivity observed in
  cultures  exposed to  either substance was  associated  with the reduction of
  numbers of  attached cells  but  not with [H]decay-induced damage to the nucleus.
  The  authors concluded that tritium  leakage  was  not due to DNA degradation in
  viable  cells  or to artifact accumulation from  radiolabeling;  rather,  fH]DN-.
  leakage was attributed to  the  direct cytotoxic  attack of  the fibers on plasma
  and  nuclear membranes.  Structural changes observed in exposed cells supported
  this conclusion.
       Tilkes and 8eck  (1980, 1983a,t>)  investigated  the cytotoxicity of well-
  defined mineral  fibers on  macroph»ges (1983a) and phagocytic epitheloid ascites
  tumor cells (1980, 1983b).  The size  distributions of the glass and asbestos
  fibers  are  presented  in Table  7-1.   For the macrophage assays, guinea pig  a-;
  rat  lung  macrophages  were  obtairiicJ py pulmonary  lavage and allowed to attacr.  tc
  culture  dishes  prior  to  fiber tntsiment.   Fiber doses were equivalent  on e
  gravimetric basis  in  most  expe-Mise'-.ts and  on  a numerical  basis in additional
  experiments.   The  release  of   ?arta.t€ dehydrogenase  and  6-glucuronitiase w = s
  measured  as  an indicator  of  Eyu/Lwcity.   Assays  measuring carboftyarste
  metabolism  and phagocytic.  abilfoj-**fe also conducted.  The asbestos end c'ess
  fibers caused  lactate  dehydrogenej*  and p-glucuronidase release from the  c.-ires
  pig  macrophages at  a relatively Tow  concentration (100  fibers  per cell).
  Similar responses  were elicited fry crocidolite, amosite,   and glass fiibers when
  the  doses were  numerically equivalent (100 fibers per cell), but the responses
  were dependent on fiber diametf- and  length.   For fibers of similar physical
  dimensions, the extent  of cytoiD*icity of  all asbestos  and  glass  samples
  detailed  in Table  7-1  were comparable when  doses were  graviwetrically equivalent
  (100 mg  dust per million  cells),  frelinrinary  experiments further showed  tnat
"  asbestos  and  glass fibers  reducs-d  the phagocytic  abiHty  of the unacrophages.
  The  authors concluded that if  geometr?r dimensions are similar, fibers with
  different physicocherrncat   compusntions  induced comparable responses.   Long,
  thin fibers of  glass, chrysolite,,  crocidolite,  and synthetic  fluoroamphiboles
  were all  toxic  in  the test system.   These results support  previous findings
  using established  macrophage cell  lines  (Beck et a!.,  1972).
       Similar  results were  obtained when the  fibers listed in Table "'-1 were
  incubated with cultures of phagi»cytic epitheloid  ascites  tumor cells  (Tilkes
  and  Beck,  I960,  1983b).    The  authors  also measured  DNA, RNA,  and prcteir.
  synthesis in  these  cultures and noted that  fibers  from diverse sources showed

-------
     Mammalian cells can  adjust to fiber-induced damage through various mecha-
nisms that are  related  to cell type and the conditions of the culture medium
(Mossman et a!.,  1983).   Cytotoxicity  elicited by certain fibers can be merr-
brane specific; therefore,  disruption  of the plasma membrane does not assure a
similar response on surface membranes or other  intracellular  membranes  (Davies,
1980; Yano et a!., 1986).
7.2.1.2  Test Material  Composition.   The physical  state of the  test material  is
a major consideration  in  the  evaluation of cytotoxicity.  The Stanton hypothe-
sis  (Stanton, 1973)  holds that the pleura! sarcoma-inducing behavior of fibers
is a  function of  fiber length and diameter.    On  a  per weight  basis,  "thick"
fibers generally  are less cytotoxic  and tumorigenic in the pleura  of  rets  and
harr.sters than "thin" fibers (Smith et  a!., 1980; Harington, 1981).  Althojg" e
close association  between the cytotoxicity  of  various fiber types and  their
ability to induce  neoplasia exists,  it is also recognized that fibers with  a
minimuir, length of  8  urn  and a diameter of 1.5 pm or less are  cytotoxic rec-s-d-
lesi of chemical composition (Archer, 1579).
     Reason dictates that short fibers with  a small  diameter  are more  cytctcxi:
than  longer  fibers because  they  are  more readily  engulfed by phagocytes.
However, Gormley  et  al.  (1980) showed an association between increasing fiDe-
length of 12 naturally  occurring mineral dusts  and cytotoxicity.   In  P3653,  snc;
A549 cells, the greatest  cytotoxicity effect was induced  by fibers  >8 urn  in
length, whereas in the  V79-4  cell line,  fibers that were  >10 urn in length were
the most cytotoxic.  In the same  study,  the  association between fiber diameter
and cytotoxicity  was assessed.  For  P388D, and A549 cells, a fiber diameter  of
0.2  to  0.3 urn was  significantly  correlated  with cytotoxicity.  However, no
correlation between  fiber diameter and cytotoxuity could be  established for
the V79-4 cells.
     The number of fibers that can be potentially  phagocytized  and  elicit  a
cytotoxic response should also be taken  into account  before  a  fibrous material
is considered biologically  inactive  (Hesterberg and Barrett, 1984).  The point
is  illustrated  by the  findings of Brown  et  al. (1979b),  who calculated the
number of fibers  in  a  given weight  of code  100 and code  110 glass fibers a-.:
found that  code 110 glass  fibers,   although  much  less cytotoxic  than the
code 100 glass  fibers  on a per weight basis,  were more  cytotcxic on  a  pe-
fiber basis.

-------
     Most studies correlate  fiber  geometry with cytotoxicity and deemphasize
fiber composition.  However,  the cytotoxicity of a fiber does appear to depend
somewhat on chemical  composition,  since  several types  of  absorption  phenomena
are known to  develop  between cellular membranes and fiber components.  Accord-
ing to  Mossman  et al.  (1983), the release of oxygen  free  radicals  is a poten-
tial result of  this  type of interaction  that can,  in turn,  lead  to pronounced
membrane alterations  (see Section 7.2.7).   In  other cases, membrane—fiber
interactions  can  lead to  cell  depolarization that may  produce  changes  in
membrane permeability (Mossman et al.,  1983).

7.2.2  Manrr.ade Mineral Fibers
7.2.2.1  Glass  Fibers.   Several  studies   have  confirmed  the cytotoxicity of
glass fibers.   Equivalent masses  and numbers  of  various  glass and asbestos
fibers  induced  comparable cell  responses,  but the  extent  of toxicity was
primarily dependent on  fiber size.   Long,  thin  (diameter ^0.3 pm and length  ID
to 40 pm for £3 and  74 percent of the fibers, respectively)  and long,  ve-y  tMr.
fibers  (diameter  <0.2 pm  and length 10  to  60  pm  for  9B.1 and 50 percent.
respectively) glass  fibers were as cytotoxic to epitheloid ascites  tumor  eel's,
mouse peritoneal  macrophages,  hamster  embryo cells,  and guinea pic, rat, enc1
dog  lung  macrophages  as similar-sized chrysotile, crocidolite,  amosite, ani
synthetic fluoramphiboles.   Long,  thin asbestos and  glass fibers  also  reduced
the phagocytic  ability  of  macrophages  and  inhibited DNA,  RNA,  and  protein
synthesis in  the  above-mentioned cells.   These glass fibers  (mean  diameters
between  0.12  and  1.9 pm)  also slowed the  growth  of  adult  and fetal human
fibroblasts and human  bronchial  epithelial  cells.   The cytotoxic behavior  of
the long, thin  glass  fibers  corresponds  to  results of in vivo  studies in which
rats and  hamsters treated intrapleurally and  rats treated  intraperitoneally
showed an increased incidence of mesotheliomas.
     Hesterberg and Barrett  (1984) reported that glass fibers (0.5,  1.0, 2.5,
5.0, and  10 pg/cm )  were  as potent in reducing survival of  Syrian  hamster
embryo cells  as were  the positive controls, UICC  chrysotile  and  crocidolite
asbestos fibers.  Consistent with the earlier introductory statements regardinc
fiber length, thin glass fibers  (diameter 0.13  ± 0.005 pm)  were more cytotcxic
than thick glsss  fibers  (diameter 0.8 ± 0.06 pm).   A dose  of  1 pg/cm  of the
thin glass fibers resulted in 5 percent cell survival; the same  dose of  tnick
glass fibers  resulted  in 86  percent  cell  survival.   In another study, cede  100

-------
TABLE 7-1.  SIZE DISTRIBUTION OF SOME ASBEST1FQRM FIBERS
         USED IN IN VITRO CYTOTOX1CITY STUDIES3
Fiber type
Johns Kanville
respirable glass
(code 100)

Johns Manville
respirable glass
(code 110)



Short, very thin
glass fibers
(4105-23-7)

Lc~:, ve-y thin
class fibers
(4105-24-2)



Short, thin
glass fibers
(4106-4-3)


Long, thin
glass fibers
(4106-19-2)



Short, thick
glass fibers
(4106-5-1)


Length
Length
(pm)
2-5
5-10
10-20
>20
2-5
5-10
10-20
20-30
30-40
>40-
<1
1.0-1.9
>1.9

1.0-2.9
3.0-4.9
5.0-9.9
10-19
20-29
30-59
1.0-1.9
2.0-2.9
3.0-4.9
5.0-9.9
10-22
<5
5.9-9
10-19
20-39
40-59
60-99
1.0-2.9
3.0-4.9
5.0-9.9
10-19
20-29
distribution
% of
fibers
60.6
34.8
2.5
2.1
18.6
28.1
30.6
12.8
3.6
6.3
83.6
11.5
4.8

9.2
15.8
25.0 '
30.0
9.2
10.8
14.1
35.2
35.2
8.5
7.0
0.7
7.0
37.5
36.0
14.8
4.0
22.4
33.5
31.5
10.5
2.1
Diameter di
Diameter
(pm)
<0.1
0.1-0.3
0.3-0.6
0.6-1.0
<0.1
0.1-0.3
0.3-0.6
0..6-1.0
1.0-1.6
>1.6
<0.1
0.1-0.14
0.15-0.19
0.2-0.29
<0.1
0.1-0.14
0.15-0.19
0.2-C.2S
0.3-0.59
(0.41)
0.15
0.15-0.29
0.3-0.59
0.6-0.99
1.0-1.99
<0.15
0.15-0.19
0.20-0.29
0.30-0.39
0.40-0.59
0.60-1.49
<0.4-0.59
0.6-0.99
1.00-1.49
1.50-1.99
3.00-3.95
Strifcutioi
% cf
fibers
17.6
53.7
24.3
4.4
0
2S.4
20.0
25.2
15.2
9.2
52.0
35.4
5.7
6.9
62. 7
25.6
£.£
i. c
0. 3

c Q
16.4
35.4
26.2
16.1
51.4
15.6
11.9
6.1
6.2
4.6
3.£
5.£
15.2
2£.6
17.6
                            (continued on the following page)
                         7-5

-------
TABLE 7-1.   (continued)
Fiber type
Long, thick
glass fibers
(4106-19-la)



Jeffry fibrils
(4173-46-1)



Je'fry fibrils
(4173-46-2)




Sno't Mjnroe fibrils
(4173-60-1).




Long Munroe
chrysotile
(4173-60-2)



Short crocidolite
(S. African)




Length
Length
3.0-4.9
5.0-9.9
10-19
20-29
30-59
60-99
<1.0
1.0-1.9
2.0-4.9
5.0-9.9
10-19
0.10-0.99
1.0-1.9
2.0-4.9
5.0-9.9
10-19
20-39
0.40-C.95
1.0-1.49
1.5-1.99
2.0-2.99
3.0-4.99
5.0-9.99
3.0
3.0-4.9
5.0-9.9
10-19
20-60
6D-SO
0.40-0.99
1.0-1.9
2.0-4.9
5.0-9.9
10-19
20-29
distribution
% of
fibers
9.9
14.6
28.9
20.1
22.1
4.4
29.5
34.7
27.6
6.6
1.6
11.8
20.0
24.1
20.5
17.7
5.9
22.5
29.7
21.0
15.3
10.0
1.7
0.0
0.7
4.3
26.4
65.7
2.9
17.6
44.4
19.8
6.2
1.6
0.4
Diameter di
Diameter
(urn)
<0.2-0.59
0.6-0.99
1.0-1.49
1.5-1.99
2.0-2.95
3.0-5.95
<0. 03-0. 05




0.10





0.03-C.05





<0.1
o!l-0.29
0.3-0.59
0.6-0.99
1.0-1.99
2.0-2.5
0.03-0.04
0.05-0.09
0.10-0.19
0.20-0.39
0.40-0.59
0.60-0.99
stribution
/t of
fibers
4.1
16.0
35.2
22.7
14. e
5.2
99.9




9r. S





c: c





4C.2
20.2
e.6
13.6
4.2
13.3
1.2
7.5
43.3
32.4
4.4
1.2
           (continued on the following page)

-------
                            TABLE 7-1.   (continued)
Fiber type
Length distribution
LengthS~oT"
 (urn)         fibers
Diameter distribution
Diameter% of'
  (urn)          fibers
Short crocidolite
(4106-23-9)




Long crocidolite
(4106-22-4)




Short f luoroamphiboTe
(4106-31-3)



Respirable
f luoroamphibole
(4106-30-1)



0.1-0.69
0.7-0.99
1.0-2.9
3.0-4.9
5.0-9.9
10-19
3.0-4.9
5.0-9.9
10-19
20-29
30-59
60-129
<1
1.0-1.9
2.0-4.9
5.0-19.0

0.5-1.49
1.5-2.9
3.0-4.9
5.0-9.9
10-19
20-39
47.8
22.3
24.6
3.5
1.5
0.3
4.2
13.6
28.3
23.0
24.1
6.8
89.9
5.8
3.6
0.75

8.2
21.5
27.0
27.5
12.5
3.5
<0. 05-0. 09
0.10-0.14
0.15-0.19
0.20-0.29
0.30-0.39

0.10-0.14
0.15-0.19
0.20-0.29
0.30-0.39
0.4-1.49

0.03-0.09
0.10-0.19
0.20-0.39
0.40-0.99
1.0-2.95
0.03-0.04
0.05-0.14
0.15-0.39
0.40-0.99
1.0-2.0

25.3
40.0
19.4
13.3
2.0

27.8
15.0
3C.S
13.1
7.9

46.0
40.9
c ~
2.9
0.6
44.7
25.3
20.9
5.6
3.5

 Dsta on the size distribution for Johns Manville glass codes  100 and  110
 are from Brown et al. (1979); all other data are from Tilkes  and Beck
 (19833,b).
the same trend, i.e., toxicity increased with increasing fiber length  and  dose.
In general, DNA,  RNA,  and protein synthesis of the cultures  were  inhibited  by
long, thin  fibers  of glass,  asbestos, and fluoroamphibole (a synthetic  amphi-
bole).  Cytotoxicity also  increased  when long, thin fibers were used in this
test  system.   The  authors  concluded  that the toxicities of fibers from differ-
ent sources but  with similar fiber dimensions were generally  comparable, anc
for  long  fibers  (>20 urn) of equivalent  weights,  the  thinner the fiber, the
greater the toxicity.   Additionally,  a glass fiber fraction with a  conficL-a-
tion  similar to  a  UICC fraction  of chrysotile  asbestos  exhibited the  sare nigh
toxicity (Tilkes  and Beck, 1983b).

-------
     The importance  of  fiber length and diameter has been  demonstrated  with
other glass  fibers  in different jin vitro cytotoxicity assays.  After milling
and collecting  respirable  fractions of codes  100  and 110 Johns Manville  fiber-
glass (see Table 7-1 for size distributions),  both fiber types were found to be
cytotoxic to cultured V79-4 cells, mouse peritoneal  macrophages,  and A549 cells
(Brown et a!.,  1979a,b).   The thinner fibers (code 100) were more cytotoxic
than  the  thicker fibers  (code  110) when fibers  were applied at  equivalent
gravimetric doses.  The authors suggested, however,  that the increased cytotox-
icity of the code 100 fibers may have been due to the  larger number of fibers/c
           10                                       Q
(78.2 x  10A ),  as  compared  to  code 110 (25.2 x  10 fibers/g);  numerically
equivalent  doses  were  not used  in these experiments.   Similar results  were
obtained by  Hesterberg  and Barrett (1984) with codes 100 and  110  fiberglass ir
Syrian hamster embryo cells and by Pickrell  et al. (1963) using glass fibers of
specified dimensions  in pulmonary alveolar macropnage cells  of  beagle  docs.
Longer,  thinner glass fibers were more cytotoxic  and induced a higher incidence
c: cell transformation  than  shorter,  thicker  fibers (the materials  were  cra\->
metrically equivalent;  see Section 5.1 for details).
     In contrast,  Nakatani •(19E3)  reported that  glass  fibers with  a  larger
diameter (code 110) were more cytotoxic than smaller diameter fibers (code 10C-)
when gravimetrically equivalent doses were applied to  cultures of mouse perito-
neal  macrophages.   The  size  distributions of  the  glass  fibers used  by Nakatani
(1953) were similar to  these used by Brown et al.  (1979b) (see Table 7-1).
     Fiber diameter appeared to play a role in inhibiting the growth of primary
adult and fetal  human fibroblasts following exposure  to 50 and 100 ug/mL code
100 glass  fibers (Sincock  et al., 1982).   Human cells  were  incubated  with
either code  100 or  code  110 glass  fibers  for 4 days.   Mean  fiber  lengths
measured between  2.7 and 26 urn;  mean  fiber diameters  were between'  0.12  and
1.9 um.   Significant  inhibition  of cell  growth was  apparent in both  cell  lines
within 2 days  of exposure  to code 100 fibers;  inhibition persisted  through  the
fourth day  of  the experiment.   Fetal  fibroblasts  appeared  to  be more resistant
to the cytotoxic  effects  of the fiber (code 100) than did mature cells.   Code
110 glass fibers  had no effect on  the growth rate  of fetal  or  adult fibre-
blasts.   Although cell  growth was severely blocked after exposure to the code
100 glass fibers, cell  viability was not affected by either fiber.  The authors
concluded that  the  differences  were probably related to  particle  diarr.ete".
since the code 100 fibers were much thinner than the code 110 fibers.

-------
     The  growth  rate of  human  bronchial  epithelial cells exposed to glass
fibers  (0.3,  1.0, 3.0,  10,  30, or  100 ug fibers/ml;  fiber dimensions  not
reported) was  inhibited  in a dose-related manner (Haugen et al., 1982).  The
number  of cell  population doublings/day (approximately 0.75  to  0.80) declined
slowly  but  was  comparable to the control   values for cells  exposed to <10 ing
glass  fibers/ml.   However,  as   fiber  concentration increased  from  30  to
100 ug/mL,  the  number  of dividing  cells dropped to  0.6 and 0.45 doublings/dsy,
respectively.  Although cell numbers were  reduced,  glass  fibers  were  considered
to be  only  mildly cytotoxic when compared to  asbestos fibers, which were 100
times more  cytotoxic.   Under similar conditions,  human bronchial  fibroblastic
cells exposed to 100 ug glass fibers/ml showed no  cytotoxic  effects.
     Jaurand and  Bignon  (1979)  indicated  that glass fibers  of  an unreported
source  and  physical  dimension were weakly hemolytic; however,  the degree  of
hemolytic activity was  similar  to  that of UICC crocidolite, but  less intense
than that of UICC chrysotile.
     Chang  et al.  (1963) incubated samples of  glass fibers and  chrysotile w-:t"
macromolecules (bovine  serum albumin,  DNA, and RNA) and  human  dermal  fife-
blasts.  Results  indicated that the glass fibers were neither  cytotoxic  nor
able to  bind  to any appreciable degree to tissue  macromolecules.   In contrast
to the  negative  results  with glass fibers, the asbestos  sample  was both adsc^-
bent and cytotoxic.   The  authors  drew a  correlation between tissue binding
capacity and cytotoxicity.  The data, however, are insufficient to support  this
conclusion  because no information on fiber dimensions was reported.
1.2.2.2  Mineral Wool/Rock Wool/Slag Wool.  Nadeau  et  al. (1963)  reported that
"type 1  and 2"  mineral wool  fibers  (diameter  3.3  urn and length 221.4  urr:  for
type 2)  were  not cytotoxic to  rat pulmonary  alveolar  macrophages or erythro-
cytes.   However, other investigators have  demonstrated that  rock and  slag wods
can induce  slight  to considerable  cytotoxicity (as measured by the  release  of
lactate dehydrogenase and £-glucuronidase) in V79-4 Chinese  hamster  lung cells,
A549 human  tumor  cells (Brown et al.,  1979b),  and  mouse  peritoneal macropr-ages
(Davies, 1980).
     Respirable rock and slag wool fibers (160 mg/ml)  induced  an  11  to  16 pe~
cent increase  in  the release of lactate  dehydrogenase and  a 26 to 36 percevt
increase in the  release of 6* glucuronidase from mouse peritoneal mac replaces
after an 18-hour  incubation period (Davies,  198D).  The authc^ conclude:  that
the fibers were weakly cytotoxic to the murine cells.

July 19£E                           7-13

-------
     Rock wool  and slag wool also wrre cytotoxic  to  cultured V79-4 cells and
induced giant  cell formation in A545 human tumor  cells  (Brown  et  a!., lS79b,
1980).  Although the exact size distributions of the fibers were not given,  the
authors indicated  that a "significant portion" of  the fibers were less  than
1.5 urn in diameter and had lengths greater than 8 urn.
7.2.2.3  Ceramic Fibers.   Several  inv-esti gators  (Gormley et a!., 1980; Wright
et a!., 1980;  Nadeau ct al., 1987) -have reported that ceramic (aluminum  sili-
cate) fibers  are  neither hemolytic toward erythrocytes  nor  cytotoxic  in rat
pulmonary alveolar macrophages  or marrophage-like P338D, cells.   The ceramic
fibers used  had median diameters raiding from 0.2 to 0.6 urn and approximate
lengths of  2  to 100 urn.   In all cas*=.,  the  data  showed minimal  cell  damage  aio
cell death.
     Potts et  al.  (1976) reported th?t 100 g of Dow Chemical Company "ceraT-c
foam" with  a mean diameter  of  5  urn  was not hemolytic in  Sprague-Dawley  rat
erythrocytes.   Less  than 1 percent irf  the  rat erythrocytes  were hemclyzec1  by
ceramic foam;  positive controls exhibited 94.7 percent hemolysis.   The ai-thcrs
noted that  the low  rate of hemolysvi for ceramic  foam  was  consistent wit".
j_n vivo data showing that ceramic fri^s are not fibrogenic.
     Two other  types of ceramic fibtts, Saffil  alumina and Saffil* zircor.ia,
were found  to  be  minimally cytotoxic and  nonfibrogenic when  incubated  with  rat
peritoneal  macrophages in vitro  (Styles  and Wilson,  1976).   However,  the
diameters of both  fibers (2 to 6 urn) were probably beyond the respirable range,
since Sebastien et al. (1979) found Irat 99 percent of the mineral fibers fojns
in lung tissue of necropsied animals' •had diameters less than 0.33 urn.   There-
fore, the relevance of the  in vitro aata reported by Styles and Wilson (1976)
is questionable.

7.2.3  Synthetic Fibers
7.2.3.1  Aramid Fibers.   Little  information  is  available on  the  i_n  vitro
cytotoxicity of  aranritJ fibers.   One study, however,  showed  that short arax-id
fibers incubated with  cultures of rai .pulmonary alveolar macrophages can induce
leakage of  cytosol and lysosomal ma"-ker enzymes  with a  concomitant decrease
in ATP cell  content (Dunnigan et al., 1984).  Aqueous solutions (25, 50, ICC.
                                                 ff
and  250 ug  fibers/ml) of commercial-grade Kevlar   fibers  were  incubate: for
18 hours with  macrophages obtained  '•••-om  adult male  Long-Evans black-hocce:

-------
rats.   Approximately  90 percent of the fibers were  less  than 5 um long and
0.25 pm wide;  the  average length and diameter were 2.72 and 0.138 pm,  respec-
tively.   Dose-related  increases in lactate dehydrogenase and p-glucuronidase
release and  a dose-dependent  drop in cellular  ATP  were observed  for the
aramid-treated cultures.   The  response of freshly harvested aramid-exposec
alveolar  cells was comparable to that of freshly harvested  chrysotile-exposed
cells.  When  the  respective cultures were grown 24 hours prior to fiber expo-
sure,  the cytotoxic response was highest with aramid.  The  authors concludes
that  aramid  fibers are biologically active and  suggested that  the cytotoxic
behavior  of these  fibers should be confirmed in animal studies.
7.2.3.2   Carbon Fibers.  Carbon fibers of various sizes (diameter 0.2 to 15 ur.)
were  only minimally cytotoxic to rat peritoneal and pulmonary macrophages an:
induced no signs  of  toxicity following ip injections  to rats.   Styles anc
Wilson  (1976)  exposed  peritoneal  and alveolar macrophages from  male and female
Wistar  rats  to approximately  0.5 mg  (3.75  x 10  particles) carbon  fibers
(diameter  0.2  to  15 pm)  for 2 hours.  Several  other  fibers,  polymers, ar.d
mineral dusts  were investigated to determine their relative  cytotoxic pcte.r.-
tial.  The overall cell  mortality was highest in cells incubated with  asbestos
or polyurethane  foam  dust;  at least 10 and 60 percent of  peritoneal anc1 pulr.c-
nary  macrophages,   respectively,  were  killed following exposure  to  these
materials.  Carbon fibers  were only minimally toxic  to cells:   less than 2  and
5 percent  of  the  peritoneal and pulmonary cell  populations,  respectively, were
killed after phagocytosis of the carbon dust particles.
     In a separate experiment,  six male  and six female Wistar rats were
injected  intraperitoneally  with 50 mg carbon fibers/kg  in  a 10-mg/mL saline
suspension.  No  lesions,  fibrosis, or granulomatous  reactions were observed  in
any  of  the dosed animals.  The authors found a good correlation between the
jm vivo and  jjj vitro  activities  of the materials tested and concluded  that
carbon  fibers  exert minimal uj vivo and HJ vitro toxicity (Styles and Wilson,
1976).
7.2.3.3   Polypiefin Fibers.   In a study conducted by  Styles and Wilson (19/6),
polyethylene "AUathene" and polypropylene exhibited low cytotoxic activity and
produced  no  i_n vivo histopathological or  clinical  signs  of toxicity  in rats.
Peritoneal and pulmonary alveolar macrophages obtained  from male ana fe-=ie
Wistar  rats  were  incubated with  0.5 mg of either  type of  polyolefin fiber
(diameter  3  to 75  um for polyethylene and 4  to  50 urn  for polypropylene).  Tre

July ISEc                           7-15              DRArT--D:  NCT £'*'" C-~ --":

-------
cell response to both polymers was similar:   less than 1 percent of the  perito-
neal macrophages and 4 percent of the pulmonary macrophages were killed  follow-
ing phagocytosis of the test materials.   Male and female Wistar rats (six/sex),
administered  a  single ip  dose  of 50 ing  polyethylene or  polypropylene/kg,
exhibited no  lesions, fibrosis, or granulomatous reactions; no adverse clinical
symptoms were noted  during the 3-month postdosing period.   In the same  study,
other fibers, polymers, and mineral dusts were evaluated, and the two polyolefin
fibers were judged to be relatively inert.

7.2.4  Minerals
7.2.4.1  Attapulgite/Sepiolite.  The  naturally  occurring fibrous  clay miners1,s
attapulgite and  sepiolite  generally  appear  to exhibit cytotoxic and hemolyfic
activity when incubated  with  peritoneal  and pulmonary alveolar macrophages or
erythrocytes.   However,  a cytotoxic  response  was not  observed in  trachea
cells.   I_n vitro studies  showing that long  and long, thin clay mineral  fibers
are mere active  than short,  short and wide,  or wide fibers, suggesting that
fiber dimension plays a role in the cytotoxicity of attapulgite and sepiolite.
     Chamberlain et al. (1982) investigated  the cytotoxicity of several  natural
asbestiforrr: fibers in three different cell lines:  mouse peritoneal macropr.ac-ss,
A549 cells, and  V79-4 cells.   Short-fibered  sepiolite (PU  7) and short-fibered
attapulgite (palygorskite)  (PU 7) were  nontoxic in  all three  test  systems.
Meerschaum (PU  0, palygorskite),  the short-fibered  PU  44  attapulgite, a no
another short attapulgite fiber (mean diameter 0.06 urn and length 0.77 urn) were
cytotoxic toward  both mouse peritoneal and  rabbit alveolar macrophages  but  net
toward the A549,  V79-4,  or rat pleura! mesothelial cells (Chamberlain et al.,
1982; Jaurand et al.,  1987).   Long-fibered sepiolite  (PU  00),  long-fiberec
attapulgite (PU  33),  and fibrous diatomaceous earth (PU 55) were cytotoxic in
all  three  cell   lines.   The  authors  concluded  that  ^n  vitro cytotoxicity
increases as  fiber  length increases.  Long  fiber lengths  generally exceeded
25 urn; other  fiber  dimensions were not  reported.  Chamberlain et al.  (1962)
made the following  predictions  based on  the correlation between fiber  length
and the cytotoxic responses observed:  (1) short-fibered sepiolite and attapul-
gite, which were nontoxic nj  vitro, would be  similarly nontoxic M} vivo;  (2)
test material  PU 0  and  PU 44,  which induced toxic  effects in macrophages,
would be fibrogenic  but not carcinogenic i_n  vivo;  and (3)  PU OC, PU 33, a-.3
PU 55 dust, which were cytotoxic to all  cell  lines, would  be fibrogenic arc
carcinogenic.   Jaurand et al. (19S7)  concluded that there  wes  net  a  strc-.c

-------
correlation between the JD  vitro  cytotoxicity  and  the  i£> vivo carcinogenicity
of attapulgite fibers; these  authors  noted,  however,  that  a lack of consis-
tently positive  effects  may  have  been related to the short length  of  the
fibers.
     Jaurand et  al.  (1979)  reported  that  attapulgite fibers were cytotoxic to
rabbit  alveolar  macrophages  and  caused  lysis  of  human  red blood  cells.
Macrophages were incubated  with 50,  100,  or 300 ug fibers/ml (dimensions net
reported)  for  2, 4,  6,  8,  20, and  23 hours.   Lactate  dehydrogenase  and
3-galactosidase activities  were measured  at  the end of incubation.   Hemolytic
activity was  determined  1  through 10  and  15,  30,  45,  and  60 minutes after
exposure;  at least  two fiber  concentrations  (50 and 100 pg/mL) were evaluates.
Significant cellular  enzyme  leakage was  observed  20  hours postexposure  to
300 ug/mL;  media concentrations of lactate dehydrogenase and p-galactesitiase
were increased by 45.5 and  31 percent,  respectively.    Approximately 11 perce-.t
of the human erythrocytes exposed  to  50 ug/mL were  lysed; 100 percent hemclysis
••ci observed  at  100 ug/mL.    The  authors   reported that  acid-treated fibe-s
exhibited greater hemolytic  activity,  but  supporting data were  not prese~tec.
     Bignon et  al.  (1980)  found  that attapulgite formulations (mean fiber
length 0.53 to 0.59 urn)  in  drugs  used to  treat gastrointestinal diseases  were
more hemolytic and  cytotoxic  than UICC chrysotile asbestos.  They concluded
that attapulgite may be  carcinogenic.  Supporting evidence of attapulgite-
induced  (diameter 0.2 urn  and  length  11.4  pm) hemolytic activity comparable  to
UICC chrysotile  B   fibers was furnished by Nadeau et  al.  (1963).  Parallel
cytotoxicity studies  further  indicated that attapulgite fibers produced a
"selective" leakage of  enzymes,  reduced  both cellular ATP and lactic  acid
production, and stimulated oxidant production in macrophages.
     In  a  similar study  (Woodworth et  al., 1983),  attapulgite fibers (1, 4,  or
16 mg/mL) were incubated for 96 hours with trachea! mucosa  cells harvested from
6- to  6-week-old male Syrian  hamsters.  Approximately  94 percent of the fibers
were no more than 1 pm long; 25, 60,  and 100 percent had  diameters  less  than  or
equal  to 0.6,  0.8,  or 1.0 |jm,  respectively.  Cytotoxic  effects  were observed  at
the two  highest  doses of  attapulgite, but the  results were not significantly
(p <0.05) different from the negative control.
7.2.4.2  Eriom'te.    Erionite  is highly toxic to  human  polymorphonuclear leuko-
cytes   and  caused  a  dose-related  increase  in the amount of  lysesonal
p-glucuronidase  released  from cells  (fiber  dimensions  not  reported)  Os*:
et al.,  1966).   Release  of the  cytoplasir.ic  enzyme  lactate denyc-ooe"=se.
• -  •"—                           7-17             D^T—D:  N:7  •*•:": -'•  £—

-------
however, was  not observed.   Normal -human serum treated with  100  ug  erionite/mL
generated  chemotactic  activity,  but no activation of  complement  (i.e.,  no C3
breakdown  products)  was detected at levels up  to  500  ug/mL.   Similar  results
were  obtained from  crocidolite-exposed  leukocytes.   No other  details were
given.
     The hemolytic  and cytotoxic activities of erionite (fiber  size unspeci-
fied)  were investigated by Nadeau et al. (1983).  Samples of the mineral were
incubated  with  either  rat pulmonary alveolar macrophages or rat erythrocytes.
Erionite exhibited  weak hemolytic activity.  Cytotoxic behavior was evident,
however, in the  release of lactate dehydrogenase and cr-galactosidase and in the
reduction  in  cellular  ATP and lactic acid  production.   No other details we-e
provided.

7.2.5  Talc
     I_n  vitro cytotoxicity  assays conducted with natural  or  acid-purified talc
(diarr.eter  from  0.2  to  20 um)  indicate  low-level  cytotoxic  activity in pe--'-
toneal and alveolar Wistar  rat macrophages (Styles  and Wilson,  1973), whic*
corresponded  to  the minimal  toxicity exhibited by talc JU vivo.   It is  note-
worthy that within  1 hour of exposure,  approximately 50 percent  of  the treatec
macrophages contained  talc  particles.   Other investigators have reported that
talc  particles   (50  ug/mL)  within a respirable range  were not cytotoxic to
V79-4  or A549 cells  (Chamberlain  and Brown, 1978).
     In  contrast,  intratracheal   instillations  of asbestos-  and  quartz-free
Vermont  talc  (0.15, 0.75,  or 3.75 mg/100  g bw; median  aerodynamic diameter
7.5 urr)  produced a  dose-related  cytotoxic  effect in hamsters,  as evidenced by
elevated  lactate dehydrogenase,  B-N-  acetylglucosaminidase, and  peroxidase
levels  in  bronchoalveolar lavage fluid  (BAL)  (Beck  et al.,  1987).   Cellular
damage was also  indicated  at 1  day postexposure by significantly  increased
BAL albumin  levels  in  higtrdose  animals when  compared to controls. Althougr.
an  acute response  was  observed, the  large diameter of the talc  particles
precludes extrapolation to inhalation data.
     The  biochemical  response  of rabbit  cartilage  tissue  to  talc-induced
synovitis  was  investigated  by Gerstuni  et al.  (1981).   In this study,  app"cx-;-
mately 0.25 cm   of  a talc-saline suspension was injected into  the right  hip  of
30  immature  (1-month-olti) New Zealand White rabbits  (sex and weight net  speci-
fied).   Left  hips  and  saline-injerted animals  served  as  controls.  Groups CT"
rabbits  were  sacrificed 1,  2, 3,  and  4 weeks  after injecfic-.   Sectic-=  o:

-------
cartilage  were  removed,  and levels of water,  hexuronic  acid,  hexosamine, and
collagen were  measured.   Cartilage thickness was measured directly on corre-
sponding hemisections  from left and right femoral heads.  Significant (0.05 <
p <0.005)  swelling  of  the medial joint space of the right hip was apparent  in
treated animals  at  all sacrifice intervals.   Water content was  significantly
(p <0.05)  increased  at weeks 1 and 2; hexuronic  acid  and hexosamine  content
were  decreased (p <0.05)  but  only  at  the end of the  first week.   Collagen
content was  not  affected.  The 15  percent  increase in  cartilage  thickness
(compared  to control cartilage thickness) could not be attributed solely to  the
3 percent  increase  in  hydration.   The  authors concluded that  the talc-induce:!
synovia! inflammation  caused a local  increase in temperature,  hyperemia, a^.c
venous  congestion that,  in  turn, facilitated  anabolic  cellular  activity ar.i
increased  production of cartilage.

7.2.6  Venriculite
     No information  on the i_n vityp cytotoxicity of  vermiculite was  feu"  *>
the available literature.

7.2.7  Mode of Action
7.2.7.1  Cellular Responses.   Cells respond  to mechanical or  chemical stir.li
caused by  the  presence of the fiber; the  response  is  primarily defensive in
nature.  Nevertheless, the stimulation of  protective mechanisms  can eventually
lead to cell disintegration, necrosis,  and disease.  The response of  cells  to
the physical presence  of fibers is somewhat analogous  to cellular responses
elicited in  the early  stages  of tuberculosis where  hypercellular activity
occurs and ultimately  facilitates  tubercule  formation  and disease  progression.
It would,  therefore, appear that the factors influencing cellular 'response  are
of great  importance if  the relationship between fiber-cell  interaction  and
pathogenesis is to be clearly understood.
     Four  of the major mechanisms by which  cells respond to  fibers  and  the
implications are discussed below.
7.2.7.1.1   Development of ferruginous  bodies.    Mineral  fibers,  including
asbestos,   are  often  found encapsulated  in a cellular  gel coat.   These cc-atec
fibers are referred to  as ferruginous,  asbestos, or  pseudossfaestos  bcc^ei.
Ferruginous bodies have  been observed  in guinea pigs and hamsters, but  n:t  ir.
rats,  following 90-day  inhalation exposures (6 hours/day, 5 csys/week) tc c'sss

-------
fibers  with  an average  diameter of 1.2 um  (Lee  et al.,  1973).   Time frorr,
exposure  to  recovery of  ferruginous bodies was  6 months  in  hamsters  and
12 months  in  guinea pigs.   Histology  revealed that most of the  gel-coated
fibers  developed  from fibrous particles; few contained dust particles.   The
significance of the findings was not discussed.
     Davis et al.  (1970) reported that following single intrapleural  injections
of glass  or  ceramic fibers (25 mg), ferruginous bodies were observed in giant
cells.  Glass  fibers with average  diameters of 0.05 to 0.99 um and ceramic
fibers  (median  diameter,  2 um)  were taken  up by the giant cells, and both  we<-e
coated  with  iron-containing  material by 2 weeks postexposure.   At 6 weeks, a
large  number  of  ferruginous  bodies, some of  them beaded,  were present  in
excised gran^iomss.   Encapsulated particles were  generally  reduced  in  size.
The authors estimated that about 1 percent of the injected fibers  were iruc'vec
in the process.
     KcL'se peritoneal  macrophages  stimulated with thioglycolate  broth  we^e
exarrinec!  following  ip  injection of  carbon  fiber particles less than  40  \;r  1o~:
and approximately 7 .to  8 urn wide (Hinson et al.,  1983).  The strain, sex, and
age of  mice  and fiber  concentration were not reported.  Aliquots of  stimulate::
macrophages were plated 24, 48,  120, and 144 hours posttreatment.   Phagocytoses
progressed over the specified  time period,  and by  120 and 144 hours,  most
fibers  were completely  covered  and  a mini-encapsulated cyst had formed.   Tne
authors noted that the carbon fibers generated a phagocytic  response  similar  to
that of abestos.
     As suggested by the above  studies, ferruginous  body development  occurs
frequently; however,  most investigators  have  failed to find an association
between fiber-induced  pathology and development  of ferruginous (or others)
bodies  (Davis et al., 1970).
7.2.7.1.2  Effects  of mineral fibers on  cell membranes.  The classical  method
used  by cells  to defend against foreign  substances is phagocytosis.   In
general,  phagocytosis tends  to  clear the surrounding tissues,  but the ability
of cells  to  engulf  solid particles  is  largely  dependent on  particle  dimension.
The presence  of  long,  thin fibers  can  result  in  "frustrated"  phagocytosis
whereby cells  can only  partially engulf the fibers.   Incomplete phagocytes1!:
with  subsequent leakage of  tissue-damaging  enzymes has been  proposer  as  e
possible mechanism of fiber-induced  disease.
     Frustrated phagocytosis  was exhibited in  a  study  conojctei  t}  De"':e=-
et el.  (19£5fc).   Cultures  of primary hepatocytes  (animal  species  ret  re::-tec)

-------
were exposed  to  attapulgite  and sepiolite fibers  (respective  diameters  ranged
from 0.01  to 0.1 urn;  average  lengths  were 0.8 and 2.04 urn,  respectively).
Phagocytosis  was  evident within 20 hours; both fibers  were  found in plasma
membrane invaginations and deeper in the cytoplasm.
     The effects  of  silica  and talc on cellular and  artificial membranes we^e
examined using three  test  systems:   wonolayer cultures  of  trachea! epithelial
cells,  suspensions   of  sheep  erythrocytes  (sRBC),  and preparations  of
phospholipid-cholesterol vesicles  (Woodworth et  al.,  1982).   Each  mineral
preparation was at least 99 percent pure,.and approximately 60 to  70  percent  o*
the talc or  silica fibers were 3.0  urn  or  smaller.  Particles were suspended  in
veronal-buffered 0.95  percent  NaCl  and added  to the test systems at concentra-
tions ranging  from 0.003 to 20.0 mg/mL.  Activity was monitored via radioa:-
tively labeled sodium  chromate (specific  activity of   Cr, 200 to 500  uCi/g).
At the end  of a  24-hour incubation  period, both minerals were cytotoxic to t^e
trachea!  epithelial  cells.   Phagocytosis  of  talc and  silica  particles  was
evident within  4 hours, and cells containing talc  or silica demonstrate:
retraction  of lamellipodal  extensions, surface blebbing,  and morphologies"
changes.   Release of    Cr  by trachea!  epithelial  cells was  higher  for cells
exposed to  0.1 mg/mL  silica (20 percent) and talc (35 percent)  than  fc>-
nonexposed  cells.  As talc  and silica concentrations  increased  from 1 to
20 mg/mL,  percent hemolysis  of sRBC increased steadily  from approximately  5  to
90 percent following  a 60-minute  incubation  period.   Membrane distortion  anc
sRBC ghosts were  observed.   Percent spontaneous leakage of [   Crjchromate  fror
liposomes  and  liposomes  exposed to  10  mg  silica particles/mL  was approximately
4 to 6 percent;  following a  1-hour  exposure to 10 mg talc/rnL,  chromate release
rose  to  about  20 percent,  suggesting an increased permeability of  the
liposomes.
     For additional  information on  the cytotoxic  and hemolytic activities cf
specific fibers, see  Section 7.2.
7.2.7.1.3  Release of superoxide.   The release of  oxygen free radicals  by cells
of the  respiratory  tract has  been implicated 1n the  pathology of  inhaled
mineral fibers.   Hansen  and  Mossman (1967) incubated erionite or glass f"ibe-s
(in unspecified  "nontoxic"  concentrations) with  alveolar macrophages isolate:
from young adult male Syrian hamsters or F344 rats.   All fibers were  dete —''ner
to be of  respirable  quality (<4 urn in diameter).   Mineral  dusts  and  rat  cells
were incubated  for 1 hour.   Hamster cells and minerals were inr'jtate:  :' =
serurr.-containing medium  for 2  hours;  the  medium  was removed er,d ce:1f <-?'"£
- -   • ' —                            7-21             D^"--:: s:~ c-"^  ::' ::~:

-------
exposed to zymosan  (25  rog/mL).   In both cell  systems,  erionite  and  fiberglass
caused significant  (p <0.05)  increases  in the production  of  oxygen  free radi-
cals when compared to untreated controls or nonfibrous  analogs (i.e., mordenite
and glass  beads,  respectively).   Superoxide  release  by  hamster  cells  was
directly related  to zymosan  concentration,  but zymosan  activation was not
required for superoxide generation.
7.2.7.1.4  Development of proteinosis.   Alveolar proteinosis  developed in  rats,
hamsters, and guinea  pigs  exposed  by inhalation  to 0.42 mg/L glass fibers at a
rate of 6 hours/day, 5 days/week (Lee et a!.,  1979;  see Section  7.2).  Proteir,-
osis developed  in all animals 90 days postexposure and persisted for 6 months.
A massive macrophage  reaction accompanied the alveolar proteinosis  but  disap-
peared within  12  months.   The  light and  ultra  structural alterations  were
similar to these seen in other experimental and human  alveolar proteinosis.
     Pulmonary alveolar proteinosis (lipoproteinosis)  was  noted  in 24  individu-
als exposed to  fine particulate matter  (Abraham  and McEuen, 1985).  In all but
twa cases,  more than 40 percent of  the  particles were  less than  1  urn wide.   The
concentration of inorganic  particulates  in the  areas of lipoproteinosis range:
              793
from 1.3 x 10   to 1.02  x 10 /cm .   The  lungs  of  three  infants contained most'y
talc;  approximately  97  percent  of  the  particles found in  the lungs  of  5  s = ~c-
blaster were silica.  Because high concentrations of  inorganic  particles  were
found within the pulmonary lesions,  the  authors  suggested that  the presence cf
talc or silica triggered the  lipoproteinosis.
     The pathogenesis of alveolar proteinosis is. unclear,  but Lee  et al.  (1979)
suggest  several  possibilities:  an overproduction of normal   or  abnormal
phospnolipids by granular  pneumocytes;  desquamation of granular pneumocytes;
primary failure of  lung clearance; or passage of blood proteins into  alveola"
spaces due to the damage of alveolar basement membrane.
7.2.7.2   Physical  Properties of Fibers.    As   discussed   in  detail  in
Sections 7.2.1 through 7.2.6,  the physical properties  of fibers  govern  the type
of cytopathic responses  elicited in mammalian cells.   An  awareness  of  how the
solubility properties,  geometry, flexibility, and size of fibers  relates  to
toxicity is, therefore,  an important aspect of the modes  of  action.   Asbestos
fibers are not  readily  degraded by interaction with biological  syster.s.   Tr.i;
ability to  maintain structural  integrity has  been implicated  in asbestcs-
induced cytotoxicity, genotoxicity,  and carcinogenicity  (see sections  7.1 er.:'I
7.2).   By contrast,  manmade mineral fibers are  relatively soluble  ip PV-^"'
locical solutions  but appear  to have low carcinogenic  pcte-tiel.

-------
     Klingholz  and  Steinkopf (1984)  noted  that the mechanical corrosion of
nonasbestos fibers, especially when accompanied by the formation of  a  gel  coat,
shows a  weakening  and increased plasticity of  the  fibers.   As fibers become
smaller, they  are  more compatible with and less harmful  to cells.   The glass-
like  nature of  manmade mineral  fibers  generally permits only  transverse
fragmentation  into shorter fibers unlike asbestos and other naturally  occurring
fibers, which  split lengthwise to produce large numbers of thinner fibers.
     Forster  (1984)  reported that  several  factors,  such  as time,  pH, anc!
temperature, influence  the  dissolution and  corrosion of mineral  fibers,  lor,
exchange (e.g.,  leaching) was  greatest  in  an  acidic solution  (pH  <7),  and
chemical reactions  (e.g., etching  and network  dissolution)  occurred under
alkaline conditions  (pH >9).   At pH between  7  and  9, the  two processes over-
lapped.   In all cases, fiber corrosion was time-dependent, and as exposure time
increased,   a thick  colloidal  surface formed around the core particle.  Corrc-
sion was evident  in  most fibers  held  at  37°C;  few changes were  observed et
20°C.    In  experiments  conducted  at 60°C,  the  rate of leaching  was  rapid.
Significant corrosion  and leaching occurred in all  fibers  (except asbestcs)  at
95°C.
     The reactions of slag, glass, and rock wools (diameter, 0.2 to 14 prr.) with
physiologic model  fluids  and water were examined by  Klingholz and  Steinkcrf
(1S84).   The model fluids contained various inorganic salts and nonelectrolytes
(e.g., glucose and  amino acids).   Leaching patterns and chemical and physical
changes were followed for 25 days for fibers in stationary fluids or for 2-  and
60 days  for fibers in continuously replenished solutions.  Structural changes
were  observed  via election microscopy.   Fibers with mass ratios of  1:40 or
1:100 were  kept in the stationary fluid; up to 10 mg of fibers (with the same
mass  ratios)   were  used  in  the  continuous leaching experiments. . Fibers
suspended in distilled water served as controls.
     All fibers showed distinct signs of attack.  After 25 days in the station-
ary fluid,  calcium Ions were leached from  slag wool, alkali ions were leached
from glass wool, and both calcium and alkali ions were released from rock woe1.
A shell-like gel  coat, generated from the  fiber core, formed around slag woc-1
fibers held  in the replenishable leaching  solutions.  At  24 days,  the alkali
alumino-silicate gel  was  about  2 ym thick;  after 60 days, the thickness nso
doubled.  As calcium and magnesium  ions were  leached  from  slag wool fibers into
the surrounding model  fluid,  approximately  1 urn of the original  fiber's  d'ia~e-
ter was  converted into  the  gel.   Similar  (bjt much  weaker) reactions *e--e
July 19 = 5                           7-23              DRA'T —DC N" CvC" C-=-  CIT:

-------
observed with  slag  wool  suspended in water.  Initially only alkali ions were
leached from glass  wool  into the replenished model solution.  At  the  end  of
24 days, some  calcium ions  had also been  released  from  the glass fibers.
However, leaching and gel  formation were slow.   After 24 and 60 days, fibers
were coated with  gels having a thickness of 0.1 to 0.25 urn  and 1.0 to  1.5  uir,,
respectively.   The  gel-forming  process was  much  slower for aqueous glass fiber
suspensions.   Leaching of  ions  was less intense for rock  wool fibers than  for
the other two mineral fibers.  Gel layers formed around the  rock  wool,  however.
The chemical  composition of  rock  wool  fibers  was not significantly altered
following suspension  in  water.   The authors concluded that  through corrosion,
fibers  lose structural  integrity and become more likely to  break into  shorter
lengths and to be cleared from  the  body;  in  addition, the  formation of gel
layers  may  render  fibers more susceptible  to  removal  (Klinghclz  anc! Steink:::',
1964).
     In a similar experiment, slag wool,  four types of glass wool,  two  types of
ceramic fibers  (Fiberfrax  ),  erionite, and  crocidolite and  chrysctile  asbestos
were suspenaed  in  a simulated extracellular fluid under  constant  flow condi-
tions  (Scholze  and  Conradt,  1987).   The durability of the various  fibers were
ranked.  Kanmade  mineral  fibers did not hold up well  in  the model  fluid. *it-
the fiber networks  dissolving at rates of  0.2  to  3.5 nm/day.  Natural fibers
(chrysotile, crocidolite,  and erionite)  were  very resistant to  attack by  the
"biological  fluid"  and had network dissolution  velocities  of 0.01 nm/day  or
less.    It was  estimated  that a  manmade  fiber (diameter 1 urn) would dissolve
completely in  approximately  0.5 to 6.5 years  and that natural fibers (diameter
5 um) would require a minimum of 100 years before dissolution was completed.
     Spurny et  al.  (1983) also  investigated the  rn vitro  chemical -and physical
durability of glass fibers, slag wool, and rock wool  held  for prolonged periods
in acid solutions (2N HCL and 2N H2$04),  a basic solution  (2N NaOH), water, and
human blood serum.  The results  indicated that all fibers  corroded  in the acios
and the base,  fibers with a  diameter less  than  1.5 urn were  degraded rapidly,
and glass fibers  were the roost  resistant to  corrosion.   Minimal leaching  of
ions from fibers suspended in water also was observed. After acidification,  Ce
and Zn ions were leached  from  the particles  but electron microscopy  showed
little  evidence of surface change.  In a similar experiment, a  significant  orc?
in the  Mg,  Al, and  Ca ion  content of acid-treated (1.4  K  HC1)  104-E  g'ess
ir.icrofibers  was reported (Bellmann et al.,  1967).   Fiber  etching,  a  32 percent
overall reduction in fiber weight, and fiber swelling accompanied tnese c"=ices

-------
in elemental composition.   However, no alterations were  observed  in other types
of  fibers  (glass, glass  microfibers,  and ceramic wool)  suspended  in acid
(Bellmann et a!., 1987) or in human serum for 1 year  (Spurny  et a!., 19E3).
     Several  investigators  have studied  fiber  morphology in whole animals
following intratracheal instillation  of  glass  fibers into rabbits and/or rats
(Spurny et  al.,  1983;  Bellmann et a!., 1987) and glass,  rock, or ceramic wool
fibers into rats (Bellmann et al., 1987).  Considerable  ni vivo leaching of Al,
Ca, Ba, Na,  Mg,  and Si from most manmade mineral fibers  occurred.  A  relative
increase'in  Mg,  Al,  and Ca  ion  concentrations  was  reported  for  104-E glass
fibers (Bellmann  et  al.,  1987).   The length of  glass, rock, and ceramic wool
fibers recovered from animal tissues remained fairly  constant throughout the  2-
to 5-year followup  periods.   A relative  increase in  the number of long (>5 pr)
glass microfibers was  observed by Bellmann et  al. (1987), who suggested that
short fibers  were cleared from the lungs more  efficiently than  long  fibers.
Diameters  of  glass  and ceramic  wool fibers remained almost  unchanged  over
2 years,  but  the  proportion of thicker fibers increased in animals  exposed tc
glass microfibers or  rock wool.   On the  basis  of _in  vitro and ^n vivo stusies,
the authors  concluded  that  none  of the fibers  was entirely resistant  to attack
by biological  fluids  or  inorganic solvents, although glass fibers  appease; tc
be more  resistant than, rock wool  and  slag wool.   It  was suggested tnst
durability and flexibility may be related to the replacement  of  leached  ions  by
hydroxyl   groups,  which could  decrease fiber  stiffness  and  increase fiber
deformability  (Spurny et al., 1983).
7.3  CARCINOGEN1CITY
7.3.1  Manmade Mineral Fibers
7.3.1.1  Fibrous Glass
7.3.1.1.1   Inhalation.   Several  studies  have been conducted to  determine the
carcinogenic potential of  glass  fibers  in five  laboratory  animal  species.  The
results  of  these studies  indicate  that  glass  fibers of varying  lengths ar.d
diameters were not carcinogenic in any of the animal species tested.
     Lee et  al.  (1981) exposed groups of  32 to 46 young adult male Sprague-
Dawley rats, hamsters  (strain and sex not  specified),  or  male albino guinea
pigs to aerosols of fiberglass insulation dust.   The diameter was about 1.2  urn.
the  length  <2  pm,  and only 15 percent of  the particles had a fibrous shape.
     15£=                           7-25

-------
Animals  were  exposed at  a concentration of  420 mg/m  for 6 hours/day for
3 months, and  groups  of animals were sacrificed  at  6-month  intervals for up to
24.months.  Only  a  few pulmonary tumors occurred in dosed animals;  the number
was too  small  to  derive any conclusions on carcinogenicity.  The glass fiber
used was not fibrogenic.
     Muhle  et  al. (1986,  1987)  exposed  108  female  Wistar rats nose-only to
glass  wool  JM 104  (Tempstran  475)  at  aerosol  concentrations of 3 mg/m  ,
5 hours/day, 4 days/week  for 1 year.  A control  group  of 55  rats was  exposed
nose-only to clean  air, and 50 remained  unexposed.   The diameter of the fibers
used was 90 percent  <0.80 urn  and  50 percent <0.42 pm, and the length was
90 percent  <12.4  pm  and  50 percent <4.8 pm.   In exposed  rats,  12  broncho-
alveolar hyperplasias,  1  squamous  metaplasia, and 1 squamous  cell tumor cf the
lung were observed.   No tumors were observed  in  the sham-exposed controls, t-t
4/55 and 3/50 from  the cage  and  unexposed   control  groups,  respectively,
exhibited bronchoalveolar  hyperplasia.   The  authors concluded that the  lo*
carcinogenic potency, which may be associated with  thin  and  durable MKK',F, is
difficult to demonstrate in an  inhalation experiment.
     Wagner et  al.  (1984)  reported similar  tumor incidences (2 percent)  in
groups of Fischer 344 rats exposed to 10 mg/m  of glass microfibers (JM IOC'},
glass wool  with resin,  or  uncoated glass wool.   The fiber dimensions ano'  tuner
incidences  are presented  in  Table 7-2.   Exposure  was for  6  hours/day,
5 days/week for  24 months;  animals were  followed for  life thereafter.   The
turners observed  in the  glass-exposed  animals were  all adenocarcinomas.  but
tumor  incidences  were not  significantly different  from those among controls.
In  rats  similarly exposed to  chrysotile asbestos,  12/48  developed  pulmonary
tumors (11 were adenocarcinomas).  Although the pattern of tumor development  in
this study  was  similar, the carcinogenic potential  of  glass  fibers was much
lower than that of asbestos.
     Smith  et  al.  (1984)  exposed 100-day-old male  Syrian hamsters  or  female
Osborne-Mendel rats to aerosols containing glass microfibers (diameter 0.45 urn)
at concentrations of  3  (3000 fibers/in ) or 0.3 mg/m  (300 fibers/m  ).   Exposure
was via  nose  only,  6 hours/day, 5  days/week for  24  months;  sham  control groups
were exposed to clean air.   No primary lung tumors or unusual pulmonary lesions
were found  in  exposed animals  of  either species.   Three  of 54  rats sitr.il arly
exposed  to  crocidolite  developed primary lung tumors:  one  developed  e mesc-
theliorr.a and two developed bronchoalveolar tumors.
i..v, TOCC                           7-9P.

-------
       TABLE 7-2.  NEOPLASMS AND HYPERPLASIA IN THE  LUNGS  OF  RATS AFTER
                      INHALATION EXPOSURE TO VARIOUS MMMF
Sample
Canadian chrysotile
Glass microfiber
Rock wool without resin
Glass wool with resin
Glass wool without resin
Control
Fibers/cm38
(d <3 \im,
I >5 pm)
3832
1432
227
240
320
Number of rats with:
Number of
animals^
examined
48
48
48
48
47
48
Broncho-
alveolar
hyperplasia
5
2
1
3
1
0
Neoplasm
12**
1
2
1
1
0
 ad = diameter, 1 = length.
  Eight animals/group were killed after 3 months of exposure and were not
 included in the evaluation of carcinogenicity.
"'Significantly different from control incidence at p <0.01.
Source:   Wagner et al.  (1964).
     Sir.ith et  al.  (1965,  1967)  exposed  (via  nose  only)  100-day-old male Svr-£-
harrste^s and  female  Osborne-Mendel  rats' to fibrous glass with mean diameters
of 0.45, 3.1,  5.4,  or 6.1 urn.  Details of this study are presented in Secticr
5.2.1.1 (see  Tables  5-4  and 5-5).  Animals were exposed to  aerosol  concentra-
tions  of  the  0.45-um  mean diameter (MD) fibers  at  0.3 or  3 mg/m  (300 or
3000 fibers/m3); the  3.1-um  MD  fibers at 10 mg/m3  (100 fibers/m3); 5.4-unr KD
fibers at 1.2  or 12 mg/m  (10 or 100 fibers/m ); or 6.1-um MD fibers  at ?  mg/r,
(25 fibers/m  ), 6 hours/day, 5 days/week for 24 months.   Surviving animals  were
observed for  life.   No primary lung tumors were observed in  any of the animals.
     No  lung  tumors were  induced in F344 rats or  cynomolgus  monkeys afte-
chronic inhalation  of  short or  long glass  fibers  (commercially available glass
wool  insulation)  with diameters  ranging from  0.5 to 6 urn (Mitchell  et al.,
1986).  In this  study,  50 male and  50  female  rats and 15 male monkeys were
exposed for  7 hours/day,  5 days/week.  Exposures were conducted for 18 months
in monkeys and 21  months  in rats.   In another  study,  exposure  of  10 baboons to
fibrous glass  dust  at a concentration 5.8 mg/m   for  7  hours/day,  5 days/week
for 34  months, resulted  in fibrosis but no  neoplasms in the lungs  (Goldstein
et al., 1983).
     However,  significantly  increased  incidences  (p<0.05)  of  mononu:"ie = '
cell  leukemia  (MCL)  were  observed in rats from all glass fibe"-expcsec CTJ:S
   r compared  to control incidences  (Mitchell et  al., 19Sc).   Fiber c^rre^' cr-s.

-------
aerosol  concentrations,  and incidences  of  MCL are presented  in  Table  7-3.
Since MCL  occurs  spontaneously in aged F344 rats,  the biological  significance
of this  finding  is  not presently known.   However,  this data suggests  that  the
presence of  glass fibers  in lymphatic  or  pulmonary  tissue may  later  the
cellular immune response.  Since no glass fibers were  found in  the spleen.

           TABLE 7-3.  INCIDENCE OF MONONUCLEAR CELL LEUKEMIA (MCL)
               IN F344 RATS EXPOSED TO FIBROUS GLASS INSULATION
                          VIA INHALATION FOR 2 YEARS
Fiber
description
Glass-coetec

Glass-coated

Glass-uncoated

Glass-uncoated

Control
Fiber Aerosol
dimensions (urn) concentration (mg/m3
d: 4 - 6
1: >20
d: 0.5 - 3.5
1: >10
d: <3.5
1: >10
d: <3.5
1: <10
--
15

15

5

5

0
ye
MCL
) incidence
37.0*

37 . 0*

35.4*

42.4**

21.2
 ad = diameter, 1 = length.
 *
  Significantly different from controls (p <0.05)..
 ""Significantly different from controls (p <0.01).
Source:   Mitchell et al. (1986).

7.3.1.1.2  Intratracheal.  The results from five studies with hamsters, rats,
and guinea pigs  provide equivocal  evidence on the  carcinogenic  potential  of
glass fibers  following  intratracheal  administration.   The results from one
study with  rats and  hamsters  indicate that microfibers  cause  an increased
incidence of respiratory tumors.   However, other studies with all three species
provided negative results.
     Preliminary results with  guinea  pigs  dosed twice  intratracheally witn
12.5 mg of long, thin glass fibers indicated fibrosis of the lung but no turrets
6 months after  instillation (Kuschner  and  Wright,  1976).   The  fibers  hei  a   (
diameter of <1  yrr,  and  13 percent were >10 urn in length.  Animals were k;1lei


-------
6, 12, and 24 months after the last injection.   Feron et al.  (1985)  instilled 1
oig glass  fibers  1n >0.2 ml of a 0.005 percent  saline solution  of  gelatin  into
the trachea  of  34 male and 35 female  Syrian hamsters once every 2 weeks for 52
weeks  and followed the animals  until  week 85.   The length of glass  fibers
(Johns Manville  code  104)  used was  58 percent <12 pm and 25 percent <5 pm; the
diameter  of  the  fibers used was  88  percent <2 pm with 35 percent <0.25 urn.  No
tumors were  induced.   Feron et al.  (1985) suggested that repeated acute reac-
tions  due  to continued treatment for 52 weeks did  not  allow fibers to settle
and have  constant chronic  contact with the target cells.   In contrast, unpub-
lished results of Pott et al. (1984b) indicated that the  same  type of glass
fiber  induced mesotheliomas  in the  same strain of  hamster by 113 weeks after
administration of 1 mg once weekly for 8 weeks.    However, very  few  tumors
appeared before week 85.  The results of Pott  et al. (1984b)  were  summarize:  by
Mohr et al.  (1984).   Groups of Syrian hamsters received once weekly instilla-
tions  of  1  mg fiber dust  for 8 weeks.   Fibrous glass of two different lengths,
UICC crocidolite  (positive  control),  and a negative control  group (TiCL)  we~e
studied.   Approximately 50  percent  of the hamsters survived for  2  years,  a-.a
the study was  terminated  at 130 weeks.   Fiber sizes  and tumor incidence  are
summarized in Table  7-4.   No tumors were found before 18 months.  Morphologi-
cally, the tumors were epithelial.   The microvilli on  the surface  were plec-
morphic under  scanning electron  microscopy,  whereas these projections were
uniform in normal mesothelial cells.
     A significant increase (p value not reported) in pulmonary tumors was also
observed in female Wistar rats following intratracheal  instillation  of the same
glass  fiber  (Johns  Manville code 104/475) (Pott et  al., 1987a).   Fifty percent
of the fibers had a diameter of <0.18 urn and a length of <3.2 urn.  A total dose
of 10  mg  (20 x  0.5 mg) was  injected  into each  of 34  rats.   Forty animals
receiving saline  (20  x 0.3 ml)  served  as  controls.  Five  of 34 glass-treated
rats  developed  pulmonary  tumors  (1 adenoma,  2  adenocarcinomas,  2  squamous
cystic  carcinomas) compared  to 0/40  for the control  group.   In  animals
similarly dosed  with  crocidolite asbestos (10 mg), 15/35  developed pulmona-y
tumors; however,  no mesotheliomas were found in any of the animals tested.
     Smith et al. (1986,  1987)  administered,  by  intratracheal  instillation,
2 mg of glass  microfibers  (mean  diameter 0.45  urn  and count mean  length 7.5 pm)
suspended  in  0.2 mL saline to male  Syrian hamsters and female  Osborne-Kenoel
rats.   The animals  were dosed once a  week for 5  weeks  (total  dose  10 mg) and
observe:? for life.  No primary lung tumors were observed in either  species.
July 19££                           7-29             DRAFT — DO N"  Q'JCTE  0* CITE

-------
    TABLE 7-4.  TUMORS IN SYRIAN HAMSTERS AFTER INTRATRACHEAL  INSTILLATION
                        OF GLASS FIBERS OR CROCIDOLITE
50% of fiber
less than
Material
Glass 1
Glass 2
Crocidolite
Ti02
Length
(pm)
7.0
4.2
2.1
Diameter
(urn)
0.3
0.3
0.2
-granular-
Number of
hamsters
136
138
142
135
Lung
carcinoma
5*
6*
9**
0
Tumors
Meso-
thelioma
37**
26**
8**
0

Thora*/
sarcorr.a
6
6
1
2
 *Significantly different from control  incidence at p <0.05.
'"Significantly different from control  incidence at p <0.01.

Source:   Mohr et al. (1984).

7.3.1.1.3  Intrapleural.  Unlike  inhalation or  intratracheal instillation into
the lungs, the  direct  administration of glass fibers, especially micrcfibers,
onto the pleura!  surface  in  rats and hamsters has been shown to result in  tk:e
development  of  mesotheliomas.   Stanton and Wrench (1972) found  that  various
fibrous  glass samples  (coated  and uncoated glass wool  insulation)  that  we^e
reduced  to a  distribution  of sizes approaching that  of conventionally mil lee
asbestos were carcinogenic when implanted intrapleurally into Osborne-Menoe1,
rats.   Partially pulverized  glass  fibers  with mean  lengths  of about 5 urn anc1
diameters ranging  from  0.06  to  3  urn  caused a  significant increase (p 
-------
that  fibers £1.5  pm  in diameter and  >8  pm in length produced the highest
Incidence  of  pleura!  sarcoma »fter  25 months.   Analysis  by  statistical regres-
sion  technique of  tumor  probability  in  relation  to fiber size, using the
logarithm  of  the number of  fibers  in  discrete size groups as the variable,
Indicated  good correlation coefficients (>0.85) for two size categories:   (1)
diameters  <0.25 urn  and  lengths >8 urn,  and (2) diameters between 0.25  and  1.5 urn
and  lengths >64 pm.  When the percent of total weight of fibers in  discrete
size  groups was the variable, the best correlation coefficient was with diame-
ters  between  0.25  and  1.5 pm and  lengths of  64  urn.  This reinforced the
authors' hypothesis that  carcinogenicity  depends on dimension and durability
rather  than on physicochemical  properties for  all  fibers  including asbestos.
Fibers  with a  length <8 urn are inactivated by phagocytosis.
     Wagner et al.  (1976)  also  tested the  carcinogenicity of glass  fibers
(Johns  Manville)  of different diameters by  intrapleural  injection into  rats.
For  the microfibers (code 100), 99  percent  had a diameter <0.5 pm (median,
0.12 (jm),  and 2 percent had  a  length  of  >20 urn (median, 1.7  urn).   For  the
coarser fibers  (code 110), 17 percent  had diameters <1 pm (median,  1.8 prr,), a"
10 percent  were  >5 pm in length  (median,  22 pm).   Groups  of male and fe-,a"e
Wistar  rats were  injected with 20 mg of fiber in 0.4 ml saline and observed fcr
life.   Four of 32 rats  receiving code 100 microfibers, but  none  of 32 controls
or 32  receiving code 110 glass fibers, developed mesotheliomas.   However, tne
increased   incidence was  not significantly  different  from controls.   Mean
survival  in all  groups was  Btout 700 days.  The  authors  reported  that  the
microfibers were  less  carcinogenic  than various samples of asbestos similarly
tested.
      In  another  long-term  study conducted  by Wagner et  al.  (1984), the
carcinogenicity  of several  MMMF  Including  glass  microfibers  (JM 100)  and
'uncoated glass wool were  evaluated  In Sprague-Dawley  rats.  Groups of 56 rats
received a single intrapleural  injection of 20 mg of each respirable  fiber ana
were  observed for  life.   Results indicated that  the  glass microfibers  were
carcinogenic;  4/48 rats developed  mesotheliomas.   In rats dosed with glass
wool,  1/48 developed a mesothelioma.   Six  of  48  rats similarly dosed with
chrysotile  asbestos developed mesotheliomas.
     The carcinogenic  potential  of  JM 104 glass microfibers in Sprague-Da»ley
rats  was  determined following intrapleural injection by Monchaux et  al.  (1531)
in studies  designed primarily to test  acid-leached chrysotile.  The mear.  length
Julv 19E:                            7-31              DRAFT—DO  NOT QuT'E 0

-------
I
of  the  glass microfibers  (JM 104) was 5.89 pm,  and  the mean diameter was
0.229 urn.  About 11 percent of the glass fibers conformed to a length >8  pm and
diameter <0.25 pm.   The  rats (3 months old) received  injections  of  a single
20-mg sample  (1.8  x 10   fibers) in 2  ml  saline  and  then were followed for
life; 32 controls received saline.   Thirteen percent (6/45)  of the rats receiv-
ing the  microfibers developed mesotheliomas compared to  0 percent of  controls.
Furthermore,  the  increased incidence was  significantly different from  the
controls (p <0.05).  Mean time-to-tumor was 513 days for glass-treated rats.
     Smith et al.  (1980) studied the effect of glass fiber size on induction o*
tumors in Syrian hamsters after intrapleural injection of 25 mg/animal.  Turc'-s
occurred in  9/60  animals when the fibers  had  a mean  diameter of 0.1  pir and
82 percent were 20  urn in length; the first tumor  appeared at day 284.  Fibers
with a mean  diameter of  0.33 urn and  46  percent >20  urn  in length  induced  turr.c1';
in 2/60  hamsters;  the first tumor was  found at day 408.  Fibers with a  mesn
diameter of  1.23  pm and 34 percent  >20 urn in  length  induced tumors  in  2/60
hamsters at 533 dsys.  No tumors were found in groups  treated with prepa'ef'c-
having mean  diameters  of O.D9, 0.41, and 0.49 pm  but  with only 0 to  2 perce'.
having lengths >10  prc.   Smith and  Hubert  (1974) found  no tumors  in 60 harr.ste-s
injected with 25  mg of an insulation grade glass  fiber with a  mean diarnete'- of
5 pm.
     Davis (1976) was unable to induce  tumors in groups of 25 BALB/c  mice after
intrapleural injection of 10 mg of glass insulation fibers with 0.05- or  3.5-prr
diameters having  lengths less than or greater  than  20  pm.   Surviving  mice  were
killed at 18 months.   The authors reported that only  2/150 mice treated  simi-
larly with  chrysotile  or crocidolite asbestos developed mesotheliomas.  Tney
theorized that  the pleura!  cavity of  mice may be  very resistant to  tumor
induction by  any  mineral fiber.  It is also possible  that  the length of the
study (18  months)  was not  sufficient  to  adequately assess  the  carcinogenic
potential of glass  fibers or asbestos in mice.
7.3.1.1.4   Intrapen'toneal.   Glass microfibers have  also  been  shown to be
carcinogenic after  intraperitoneal administration  to female Wistar and O.sborne-
Mendel rats.  The  predominant tumor  type  was mesothelioma.   The  results  o* c^e
study indicated  that the intraperitoneal   injection of  glass microfibe-s w=s
not carcinogenic in hamsters.
     A dose-related increase  in the  incidence of tumors was reported b\
and Friedrichs  (1972)  after intraperitoneel injection of fibroji c~e=s (S
105) into female  WistS'-  rets.  At dcses  of  2, 10,  ar.d IOC-  (- *  25)  &;,  -  ~

-------
4/40, and  23/40  rats  developed  tumors.  No tumors occurred in 80 control rats.
The  appearance of  the first tumor was at day  762,  350,  and  197  at the  low-,
mid-, and  high-dose  levels,  respectively.   Pott et al.  (1974,  1976) else
conducted  studies  with  fibrous  glass MN 104  (50 percent of the fibers having a
diameter <0.2 \im and  and 50 percent with  lengths <11 M1")  and fibrous  glass MN
112  (50  percent  of the fibers  having a diameter <1 ym and 50 percent  with
lengths <28 pm).   In  female Wistar  rats-injected  intraperitoneally with MN 104
at doses of 2, 10, and 50 (2 x 25) mg (in  2  ml  saline),  7/80,  23/80,  and 47/60,
respectively,  developed mesotheliomas; 1, 3, and  6  rats also developed  spindle
cell sarcomas, respectively.  Time to first  tumor was 576, 210, and 194  days  at
the  low-,  mid-,  and high-dose levels,  respectively. Following a single 20-mg
injection  of  fibrous  glass  MN 112,  11/40  rats  developed tumors (10 mesothelic-
mas), and  the  first tumor appeared at day 390.   The authors stated that NMRI
mice developed tumors  at  comparable rates after  injection of fibrous  glass,
whereas treated  hamsters  and guinea pigs  did not  develop  tumors; however, dcta
were not  presented.   The tumors  found following fibrous  glass  or asbestos
injection were histomorphologically indistinguishable.
     Results  reported  by Miihle et  al.  (1986,  1967) indicate a  significant
(p £0.05)  increase in the  tumor  rate (17  percent) of female Wistar. rats
injected with  glass wool JM 104  (Tempstran 475) when  compared  to controls
(6 percent).   Thirty  rats received  0.5  mg of  JM  104 glass wool  suspended  in
1 ml saline by  intraperitoneal  injection.  A control group of 32 rats received
saline only.  The  fiber length  was <8.2 pm  (90 percent),  and the diameter  was
<0.42 MFT, (90 percent).  Similar treatment  with  crocidolite yielded a  55 percent
tumor rate.   Smith et al.  (1986,  1987) administered 25 mg fibrous  glass (mean
diameter 0.45 urn)  suspended in 0.5 ml saline to male Syrian hamsters  and ferr.ale
Osborne-Mendel rats by  intraperitoneal  injection and observed the  animals  for
life.  The mean  lifespan  of the treated  hamsters was  similar  to that of the
saline and unmanipulated  cage controls.   However, treated rats  had  a reduced
mean lifespan compared  to the controls, due to  the occurrence of a  statisti-
cally significant  increase  (p <0.01) in abdominal tumors:  8/25  glass-treated
rats developed abdominal  mesotheliomas  compared to 0/25 saline  and 0/125 c£?e
controls.  No tumors were observed  in hamsters.
     In a  series of experiments conducted by  Pott  et  al.  (19£7a). 50 ma^sse
and  naturally occurring  mineral fibers,  metal  compounds, and other dusts were
tested  via the  intraperitoneal  route  for  carcinogenicity  in  Wists"  cr
Sprague-Dawley rats.   A total  of  five fibrous  glass, Uc  slag- nocl.  f:-"
July 1928                         '  7-33             DRfc"--DO NCT C

-------
attapulgite, and  two  erionite  samples  were tested;  in addition,  1  sample each
of granular glass,  rock wool  (one portion processed to produce  finer  fibers),
                  £
and aramid  (Kevlar  )  were used.   Groups of 28 to 115  rats (the majority were
females 3 to  9 weeks  of age) received  doses of a given test material suspended
in saline.  Fiber dimensions,  total  dose, tumor incidences,  and lifespan are
presented  in  Table 7-5.   (Rock  and slag  wool,  Kevlar ,  attapulgite,  and
erionite are  discussed  in Sections 7.3.1.2.,  7.3.2.1., 7.3.3.1, and 7.3.3.2,
respectively.)  All fibrous glass  samples were  found to be carcinogenic, and
administration of the longer  fibers  resulted  in a  slightly higher  tumor inci-
dences.  As expected,  granular glass did not cause  a  significantly increased
tumor incidence when compared to tumor incidences of rats  injected  with  saline.
In general,  Sprague-Dawley rats were more sensitive to the  carcinogenic  effects
of glass fibers than Wistar rats.
     The carcinogenic potential  of 11  naturally occurring  and manmade mineral
fibers including  glass  microfibers (104/475)  was compared  in  another  study  in
fer-='e wistar  rats  by Pott et  al.  (1967b).   The size distributions and  nur.ser
of fibers intraperitoneally injected are  presented  in  Table 7-6, and the life-
span  and  turner incidences observed  are presented  in Table 7-7.   Preliminary
results indicated that the microfibers  were carcinogenic;  34/53  rats dosed *-ith
5 mg  (990 x 10  fibers) glass  fibers  and  observed  for life developed tumors
(mesotheliomas or sarcomas).   The first tumor  occurred at  approximately  day  258.
7.3.1.2  Mineral  Wool.   Inhalation studies  with hamsters  and two  strains of
rats  indicate  little,  if any,  evidence of carcinogenicity for  slag wool and
rock  wool.  Increased incidences of  mesotheliomas  and sarcomas  were noted for
rock  wool following  intrapleural or  intraperitoneal administration.   However,
rock  wool  has a  low  carcinogenic  potential when compared  to  asbestos.   The
tumor  incidences  for  slag wool  following  intrapleural  or intraperitoneal
administration were low (0 to 6 percent), thus providing equivocal  evidence  for
carcinogenicity.
     No primary lung  tumors were observed in male  Syrian hamsters or female
Osborne-Mendel rats following  inhalation  exposure to slag wool  at  a concentra-
tion  of 10  mg/m  (200 fibers/m  ),  6 hours/day,  5 days/week  for 24  months (see
Section 5.2.1.2  and  Tables  5-4  and 5-5).   The mean diameter of the fibe-s
2.7 urn, and the animals were observed for life (Smith et al., 1985, 19S7).
     Wagner et al.  (1984) performed  carcinogenicity  studies  in  rats exposed b\
inhalation or  intrapleural  injection to glass wool, rock wool,  and slag wo:":
(intrapleural  administration only).  Glass microfiber (JK 100) ana Canada-:

-------
    TABLE 7-5.  SIZE DISTRIBUTION, TUMOR INCIDENCE, AND L1FESPAN FOLLOWING 1NTRAPER1TONEAL
        ADMINISTRATION OF MANMADE AND NATURALLY OCCURRING FIBERS TO FEMALE WISTAR  RATS
Lifespan
(weeks) after first treatment of
Dimension (urn)
Fiber
Glass
Glass-104/475

Glass-104/1974


Glass-lOO/Pen

Rock wool

Slag wool

Kevlar-
Attapulgite
(Georgia)
(Mormoiron)
(Lebrija)
(Caceres)
Erionite
(Oregon)

(Turkey) -



Diameter
50%<
Length
50%<
--granulai —
0.18
0.18
0.29
0.29
0.30
0.33
0.33
0.64
1.54
1.5
2.6
-c

0.04
0.07
0.07
0.07

0.21
0.21
0.38
0.38
0.38
0.38
3.2
3.2
4.8
4.8
3.5
2.4
2.4
4.1
23
14
26
..c

0.8
0.7
0.5
1.3

1.8
1.8
2.9
2.9
2.9
2.9
Total
dose (ing)
40(2x20)
0.5
2.0
5
5
10
2.0
10
10
75(3x25)
40(2x20)
40{2x2£)
10

60
60
£0
10

0.5
2.0
5.0
1.25
5.0
20
% Rats
with
tumors
4.4
16.7
25.8
81. 5b
44.4
50.0
38. 9b
45. 3b
13. 3b
71.4b
2.1.
6.1
12.9

3.6
3.5
3.5 .
40

48.4
90.3
70 8
71. 7*
81.1*
69. 8b
All rats
20*<
102
95
84
52
7
6
79
68
88
61
94
94
109

89
92
95
94

81
60
50
64
45
35
50*<
119
116
110
64
34
37
90
79
97
77
107
111
121

108
116
116
IDS

108
79
61
83
52
41
6Q%<
136
138
127
79
53
54
112
91
115
56
126
127
139

129
13S
134
132

129
93
76
97
68
51
Rats
with tump_r
First
121
EC
64
40
24
17
53
53
6:
39
67
£3
12:

75
47
9£
74

62
51
40
41
32
30
Me a
129
113
107
67
49
51
95
83
105
75
77
114
13E

IOC
92
114
116

9£
8C
6£
82
54
44
aAnimals having abdominal sarcomas, mesotheliomas, or sarcomas were included.
 This experiment was conducted in Sprague-Dawley rats.
cNot reported.
Source:   Pott et al. (1987a).

-------
            TABLE 7-6.  SIZE DISTRIBUTION AND NUMBER OF MANMADE AND
                    NATURALLY OCCURRING FIBERS ADMINISTERED

                    INTRAPERITONEALLY TO FEMALE WISTAR RATS
Fiber length
(urn)
Fiber
Glass -
Ceramic Fiberfrax
Ceramic MAN
Kevlar6
Polypropylene
Wollastonite
Chrysotile
10%<
1.4
2.6
2.1
1.5
2.9
2.4
0.25
50«<
3.2
8.3
6.9
3.9
7.4
5.2
0.61
90%<
8.4
34
47
11
23
13
1.9
Fiber diameter
(um)
10%<
0.09
0.44
0.56
0.28
0.58
0.62
0.03
50X<
0.18
0.91
1.1
0.47
1.1
1.1
0.06
90/t<
0.40
2.4
2.9
0.75
2.1
2.3
0.12
Number of
fibers
injected'
(x 10e)
990
173
105
1260
409
430
202
Source:   Pott et al. (19£7b).
     TABLE 7-7.   LIFESPAN AND TUMOR INCIDENCES IN FEMALE WISTAR RATS AFTER
   INTRAPERITONEAL ADMINISTRATION OF MANMADE AND NATURALLY OCCURRING FIBERS



Fiber
Glass (104/475)
Ceramic r
Fiberfrax
Ceramic MAN
Kevlar*
Polypropylene
Wollastonite
Chrysotile


Saline

Total
dose
(nig)
5

45
75
20
50
100
0.05
0.25
1.00
2 mL
Rats
with
tumors/
rats
tested
34/53**

33/47**
11/54**
3/53
2/51
0/54
11/36**
21/34**
30/36**
1/102
Lifespan
(weeks) after first treatment of:
rats

20%<
53

38
71
78
86
88
81
80
52
85
an
50!t<
67

51
91
106
108
107
105
90
63
111
rats
805U
185

63
106
126
130
126
128
115
72
127
with tumors
1005e<
126

95
128
130
130
130
130
130
130
130
first
37

30
.64
35
126
--
61
58
35
93
mean
65

65
S3
62
126
--
IOC
92
62
92
**Significantly different from control incidence at p <0.01.


Source:  Pott et al. (19S7b).

-------
chrysotile samples were  used  for comparison.   In the inhalation  study, groups
of 56  Fischer  344 rats were exposed to a  nominal  concentration  of 10 mg/m
respirable dust  for  7 hours/day, 5 days/week for 12 months, and then exposed
to clean  air  for lifetime observation.   Size distribution of the  fibers wes
determined  by scanning  electron microscopy.   With the exception  of the
chrysotile  and glass  microfibers,  sample  counts  of nonfibrous particles
exceeded fibrous particles.
     Table- 7-2 summarizes the incidence of pulmonary neoplasms  and  broncho-
alveolar hyperplasia  observed  in this study.  Tumors were induced in 2/48 rats
exposed to  uncoated  rock wool:   one benign  and  one  malignant adenoma.  A total
of 11  rats  exposed to chrysotile asbestos  developed adenocarcinomas, and one
developed a pulmonary adenoma.   Pulmonary  adenocarcinomas  occurred in one  ret
exposed to the glass microfibers and in one exposed  to the glass  wool wit!--
resin.   All neoplasms in the lung occurred between days  500 and 1000.  The
difference in  tumor  incidence  between rock  wool  and asbestos may have bee" due
to the differences in the numbers of respirable fibers within each  dust cicuc.
     In the  intrapleural  study, groups  of 48 Sprague-Dawley  rats  we^e
inoculated with  20 mg of respirable dust  in 0.5 ml  of saline.  Table  7-6 shows
the estimated  numbers of particles  and  fibers  injected  and  the tumor  respo~.se.
Tumor  incidences  for the rock wool with and without  resin  coating were 6 sr.c
4 percent, respectively.   The  low tumor incidences may have been  due to  the
high percentage  of nonfibrous  particles administered.   No tumors were induced
in rats injected with slag wool.
     Pott et  al.  (1984a) studied the carcinogenic  potential  of rock wool, slac
wool,  and  basalt wool  samples  in rats.   Groups of 45 female Wistar rats
received  a  single intraperitoneal injection of 5 mg of the basalt  or  slag
wool,  or  10 mg of rock  wool.   An additional group of 63 Sprague-Dawley  rats
received three 25-mg injections of rock wool.   The  rats were then observed for
life.  The diameters of the basalt wool (50 percent  of fibers <0.52 urn)  and
rock wool  samples (50 percent of fibers <0.64  or  <1.90 urn)  were greater tna~
those  of  the   JM  104 glass microfibers  (described previously)  that  caused
sarcomas or mesotheliomas.   The diameter  and length of 50 percent  of the  sis:
wool fibers were <0.18  and <2.7 pm,  respectively.   The  tumors noted were
reported as sarcomas or mesotheliomas  and  occurred  at  low  rates,  accc-T.tirc
for 5  (2/41) and 16  (10/63) percent in rats receiving slag wool  (5  mg) and
Ji.lv 198E'
7-37
DRA"--DC'N:

-------
     TABLE  7-8.   INCIDENCES OF  LUNG  MESOTHELIOMAS  IN RATS AFTER 1NTRAPLEURAL
                          INSTILLATION  OF  VARIOUS  MMMF
Fibers
or particles (x 108)
Sample
Canadian chrysotile
Glass microfiber
Rock wool with resin
Rock wool without resin
Glass wool
Slag wool
Control
Fibers
>5 urn
196
30
1.6
2.6
4.2
1.3
Particles
>1 urn
8.8
25.2
29.4
47.0
19.6
55
Number
of rats with
mesotheliomas
6
4
3
2
1
.  Forty-eight animals/group.
  Twenty-four control  animals.
 Source:   Wagner et al.  (1984).

 rock wool (25  mg  x  3),  respectively.   No  tumors  occurred  in  animals  receiving
 5 mg basalt  or 10 mg rock wool.  The  authors  indicated that the slag wod
 fibers  were  not durable  i_n viv o  during  the testing  period.
      In a series  of  studies  conducted  by Pott et  al.  (1987a)  and describe:
 earlier (see  Section  7.3.1.1.4),  female  Wistar  or Sprague-Dawley rats were
 administered intraperitoneal  injections of rock  or slag wool and  observed  for
 life.   Fiber  dimensions, total dose,  tumor incidences,  and lifespan are
 presented in  Table  7-5.  As  seen in  the previously  mentioned  study (Pott
 et al., 1984a), administration  of slag wool  did  not result  in a  significantly
 increased incidence of  tumors in  Wistar rats (2.1 to 6.1 percent incidence).
 No tumors were  found  in saline  controls.   However, tumor  incidences  of 13  a-,d
 71 percent were obtained  in Sprague-Dawley rats administered rock wool samples
 having  diameters of 0.64  and  1.94 pm,  respectively (see Table 7-5).   An  inci-
 dence of 5.6  percent  was  noted  in Sprague-Dawley  rats  injected  with saline.
 The tumor counts included abdominal  sarcomas, mesotheliomas,  and/or  carcinomas.
 7.3.1.3  Ceramic Fibers.  The results  of  animal studies on the carcinogeni'city
 of ceramic  fibers suggest that  the  carcinogenic  potential is similar to that
 obtained for  other  vitreous fibers.   Inhalation  of  aluminum  silicate ana
 refractory ceramic fibers (Fiberfrax )  have been  shown to  be  noncarcinogeri: ii
 rats and hamsters  in  chronic  toxicity studies,  and intratracheal
 of Fiberfrax  did not result in pulmonary tumors.   In contrast.  intreperito-e;"'
 injection of refractory ceramic fibers (Fiberfrax*) has bee-  sncv. t:  res.":  i"

-------
high  incidences  of mesotheliomas  in  rats and hamsters.  The  incidences  of
mesotheliomas  following intraperitoneal  or  intrapleural administration  of
aluminum silicate microfibers were 3 and 10 percent,  respectively.   Considerably
lower tumor incidences were obtained after Intrapleural  administration  of  large
diameter alumina and zirconium oxide fibers.
     Pigott et al.  (1981)  found  that  refractory A1203 fibers  (Saffil& fibers
I.C.I), as manufactured or thermally  aged, were not carcinogenic by inhalation
In  Wistar-derived  rats after 86 weeks  of exposure.   The median diameter of
fibers prior to processing (ball  milling) was about 3 urn, and the  median  length
varied between 10.5 and 62 urn; ball milling  did not change the median  diameter
(3 urn), but the  length of 6.6 percent  of the  fibers  was  reduced to <13.3 prr.
Animals were  exposed 6 hours/day, 5  days/week at  concentrations  of 2.16 c-
2.45 mg/m .   The cumulative exposure was about 6000 mg-hour/m .
     Davis et  al.   (1984)  studied the  carcinogenicity  of fibrous  ceramic
aluminum silicate glass after inhalation exposure  and intraperitoneal  injection
in rats.   In the inhalation experiment, 48 Wistar  AF HAN rats were exposes tc c
respirable dust  cloud  of the ceramic  glass for 12 months at  8.4 mg/m  (actua"
respirable concentration);  the rats  were observed for  up to  20  additional
months.  The dust  cloud consisted of  large numbers of short, thin fibers, sic"
only 95 fibers/mL  were  >5  urn in  length and  <0.3 urn  in  diameter (aspect ratio
>3:1).   Approximately  90 percent of  the  fibers were  <3 urn in length.   Some
animals were sacrificed early to determine lung dust burden (see Section  4.2.3).
Eight pulmonary  neoplasms  were found  in animals exposed to ceramic  fibers; one
adenoma, three carcinomas, and four malignant histocytomas.   No pulmonary  tumors
occurred  in  40  control  rats.   However,  the biological  significance of  the
.reported pattern of tumor  development was unclear since the  finding of malig-
nant histocytomas in rats exposed to ceramic fibers was not consistent  with  the
expected  pattern of  tumor  development resulting from  asbestos  exposure.
Peribronchiolar  fibrosis  did not  occur  in  the treated  rats, but  alveolar
proteinosis was  found.   In contrast,  the authors  found that fibrous  ceramic
glass was slightly  carcinogenic  in rats  after intraperitoneal injection.   The
rats received  a  single 25-mg dose of  ceramic dust.   Three of 32 rats developed
peritoneal tumors  (one  mesothelioma).   The first  tumor  did  not  appear until
day 850.   Similar doses of chrysotile asbestos reportedly produced mescthelio-ss
in  90  percent  of the rats, with  the  first tumor  appearing at 200 days.   Hc»-
ever, no data were presented.

.L.IV tote                           7.39              nRA-T--n: N:~ OJ:~E  Cr  '.'.'-.

-------
     In  one  of a series of  studies  conducted  by Smith et al.  (1966, 1967),
male Syrian  hamsters  and female Osborne-Mendel  rats were exposed to  refractory
ceramic  fibers  (Fiberfrax  )  via the  inhalation,  intratracheal , or intraperito-
neal routes  (see sections 5.2.1.1, 5.2.1.3, and Tables 5-4 and  5-5).   Nose-only
inhalation exposure was  for 6 hours/day, 5 days/week for 24 months.   The
diameter  of  the fibers  was  1.8 urn,  and the concentration was approximately
12 mg/m   (200 fibers/cm  ).   Controls  were  exposed  to air, and the animals were
observed  for  life.   Only 1/70 hamsters  developed a pulmonary mesothelioma, and
although  spontaneous  mesotheliomas  in Syrian  hamsters  are rare, the possiblity
that this  tumor was incidental cannot be ruled out.   No pulmonary tumors were
observed  in rats.   No primary tumors were  induced in  groups  of hamsters and
rats dosed intratracheally  once weekly  for 5  weeks with 2 nig of ceramic  fipe-s
suspended in physiological saline.  In contrast, intraperitoneal  administref'c-
of 25 mg  of ceramic fibers  in 0.5 ml  physiological saline resulted  in  signifi-
cant increases  (p  <0.. 05) in  the incidences of abdominal mesotheliomas  in both
rats and  hamsters.   Tumors  were observed in 7/36 hamsters and 19/23 rats  dcsec!
with ceramic fibers.  No tumors occurred in control animals  (0/25 rats  and G/23
hamsters).  In animals similarly treated with asbestos, 8/25 hamsters  anc! 20/2:
rats developed mesotheliomas.
     Comparison  of  the  carcinogenic  potential after intraperitoneal injection
of 11  naturally occurring  and manmade  mineral fibers  was  reported by Pott
et al.  (19£7b).  Groups  of  8-week-old female Wistar rats received five weekly
intraperitoneal  injections  of  1 mg  104/475 glass  microfibers,  9 mg ceramic
Fiberfrax15', 15 mg ceramic MAN, 4 mg Kevlar  fibers, 10 mg polypropylene fibers,
or 20 mg  wollastonite (India) suspended in 2 ml of sodium  chloride solution.
The  si2e  distribution and  number of fibers  administered  are presented  in
Table 7-6.  Surviving  animals were killed 130 weeks  after the last injection.
Preliminary results indicated that glass  microfibers  and ceramic Fiberfrax"
induced tumors  in  at least 64  and 70 percent of  rats injected,  respectively
(Table 7-7).    The   tumors  reported were either abdominal  mesotheliomas  or
sarcomas.   Ceramic Fiberfrax6, which is a relatively  thick fiber  (0.44 to
2.4 urn in  diameter),  induced a similar  tumor  incidence  as  compared to glass
microfibers although the number  of   fibers injected was substantially less
(Table 7-6).   The  second ceramic fiber  sample (MAN) evoked  moderate induction,
resulting  in  the development of  tumors  in 20 percent of the rats  injectec.
Rats injected with  0.05, 0.25, or 1.0 mg chrysotile asbestos (postive  controls)
July 1S = £                           7-40             DRAFT — DO N-~ C;.

-------
exhibited a  dose-related Increase  in  tumor incidence:   11/36,  21/34,  and
30/36 rats developed tumors  in  the  low,  mid-, and high-dose groups, respec-
tively.   Only 1/102 rats in the saline control  group developed a  tumor.
     Stanton et al. (1981)  studied  the carcinogenic potential  of two ceramic
glasses  after  intrapleural  injection  into  female  Osborne-Mendel rats  (see
section  7.3.4  for details).  These large-diameter fibers were  composed of
>80 percent microcrystalline Al^O,  (glass 21)  or 90  percent microcrystalline
2r02 (glass 22).   A  single  40-mg injection  of  each ceramic  fiber was  adminis-
tered, and animals were observed for life.   Tumor  incidences for both ceramic
glasses  were  low; 2/47  rats  receiving glass  21 and  1/45  receiving  glass
22 developed pleura!  neoplasms compared to 0/115 controls.
     Low  tumor  incidences were  obtained  after  intrapleural inoculation of
ceramic  (aluminum silicate) fibers  into  Wistar rats  (Wagner et al., 1972).
Ceramic  fibers with diameters of 0.5 to 1 urn (length was not reported)  induced
mesotheliomas in  3/31  rats.   The fibers were  administered  in a  suspension o*
physiological  saline  at a concentration of 50 mg/mL, and the rats were observe:
fcr life.  In  rats treated  similarly with ultrafine chrysotile (dimensions net
reported), 44/68 developed mesotheliomas.
7.3.2  Synthetic Fibers
7.3.2.1  Aramid and Pol^olefin Fibers.  Aramid fibers are carcinogenic  fclio*-
ing inhalation  exposure and have a low carcinogenic  potential   after intra-
peritoneal administration.  Stadler (1986)  reported the results of a chronic
inhalation study  in rats conducted  at Haskell  Laboratory.   Groups  of 10G male
and 100  female  CrV.CD  BR rats were  exposed  to 0,  0.08, 0.32,  or 0.63  mg/m
Kevlar   fibrils  (ultrafine) for  6  hours/day,  5 days/week  for 2 years.  In
addition, 100 males and 100 females were exposed  to 2.23 mg/m   under similar
conditions for 1  year  and observed for a recovery period of 1 year.   This  was
instituted after  study  initiation  because  of  high mortality  in the high-
exposure group; the majority of these early deaths occurred between  9  months
and 1 year into the study.  The fibers had a diameter  of 90  percent <1.5 um ar.s
length of 75 percent <20 pm.   Exposure concentrations of 0.08, 0.32,  0.63, and
2.23 mg/m  were approximately equivalent to  2.4, 25.4,  111.2, and 435.5  Kevlar5
fibrils/cm ,  respectively.
     Cystic keratinizing  squamous cell carcinomas  were  induced  in  1/36  surviv-
ing males and  6/56 surviving females  exposed  to 2.23 mg/m   for  1  year  ar.c in
4/69  surviving  females exposed  to  0.63 mg/m   for  2  years.   In  females, the
increased incidence  was dose  related, statistically sie*';fica-.t  (:  
-------
and showed  a  positive trend (p <0.01).   In addition, 1/56 females exposed to
2.23 mg/m   for  1 year  and 6/69 females exposed to  0.63 mg/m  for 2 years
developed squamous  cell  metaplasia,  which  was associated with alveolar bron-
chiolarization.    No tumors were observed  in control animals.  The  authors
suggested that  the  low incidence of metaplasia observed  at the high-exposure
level  as  compared to the  higher incidence at  the  0.63-mg/m   level may have
been due to the reversibility of this lesion after  a  1-year recovery  period.
     Pott et  al.  (1987a)  conducted  a series of experiments to determine  the
carcinogenic  potential  of 50 materials including aramid  (Kevlar  ) following
intraperitoneal  administration  to female Wistar  rats (see section 7.3.1.1.4).
A total of  10 mg  of Kevlar  was administered; the sample was prepared using
ultrasonic  treatment,  but  the  authors stated  that the test material was fle^y.
and its dimensions  apparently  were  not  measured.  The tumor incidence  obtained
                            s
in  rats  administered Kevlar  was 13 percent  (see  Table 7-5)  compared to an
incidence of  6  percent in rats  injected with  saline.  The tumor types  reported
included sarcomas, mesotheliomas, and/or carcinomas.
     Kevlar"  and  polypropylene  fibers were found to have  a low  carcinogenic
potential  in  female Wistar rats after  fiv« weekly intraperitoneal injections
of  4  and  10  mg,  respectively  (Pott et al.,  19£7b).   Fiber   dimensions £-c
details of the study are presented in section 7.3.1.1.4  and Tables 7-6 and 7-7.
Three of 53 and 2/51 rats developed tumors  (mesotheliomas or sarcomas) efter
                    £
injection of  Kevlar  and  polypropylene fibers,  respectively.  The increased
incidences were not statistically different from controls.
     Similar  results  were  obtained  in a second intraperitoneal study  in Wistar
rats by Davis (1987).   Groups  of 48, 32,  and 32 rats were administered single
intraperitoneal  injections of 0.25,  2.5, or 25 mg of  Kevlar  in saline, respec-
tively, and observed for life.   An additional group of 48 untreated rats served
as  controls.  The diameter of  96 percent of  the fibers was reportedly <1 urr,.
The fiber length was not reported,  but the author stated that  the test substance
was prepared  from Kevlar  pulp material and contained  both fibers and aggre-
gates.   Intraperitoneal  injection resulted in  the development of  mesotheliomas
in 2/32 high-dose rats (not significantly different from controls).  Peritoneal
sarcomas were found in 3/48, 2/32,  and 1/32 rats receiving 0.25,  2.5, end  25  me
of  Kevlar ,  respectively.   However,  2  of 48 controls also developed  se-cc^:;,
and since the incidences were neither statistically different  frorr controls  nor
dose-related, the author considered this tumor type to be spontaneous.
     iqe.=                           7-42             DKA-T--C: N:

-------
7.3.2.2  Carbon Fibers.   Only one dermal  carcinogenicity  study  on carbon fibers
was  found  in the  available  literature.   In  a study conducted at Bushy Run
Research Center, OePass  (1982)  evaluated the chronic dermal toxicity of four
carbon  fibers  in  male  C3H/HeJ  mice.   The four  carbon fibers tested were
"Thornel" type  P,  high-modulus,  pitch-based  continuous fibers  (CF);  "Thome"!"
pitch-based  fibers  (MAT);  "Thornel" polyacrylonitrile-based  continuous fibers
(PAN);  and  oxidized polyacrylonitrile-based  continuous fibers  (PAN-oxidized).
Approximately 1.9, 2.0, 1.1, or 1.3 mg of CF, MAT,  PAN, and PAN-oxidized carbon
fibers, respectively, in  25  uL  of a 10 percent (w/v)  benzene  suspension were
applied dermally three times a week for life  to groups  of 40 animals.
     Survival was  not  affected  in any of the  carbon-dosed groups compared tc
controls.   Administration  of  CF  fibers resulted a  tumor  incidence  of 7.5  per-
cent  (3/40  mice).   The  tumors  were all  located at  the  test site  and were
described as a papilloma, a  squamous  cell  carcinoma,  and a hemangiosarcorr.c
(subcutaneous).  Similarly,  MAT carbon  fibers induced tumors  in 3/40 mice.
These  tumors consisted  of  a  squamous  cell  carcinoma,  fibrosarcoma, and
re"2igiosarcoma; all  were located on  the skin in  areas  other  than the test
site.   One  mouse from the PAN-oxidized  carbon fiber group developed a skin
tumor.  This was described as a  leiomyosarcoma of the  skin  and subcutis of the
shoulder region.   No tumors  were induced in  the mice  dosed with PAN carbo-
fibers  or in control  animals receiving benzene alone.  Although  dermal  turners
were  found  in  this study,  the incidences were not  significantly different fro-
those of controls,  and the tumors observed were types frequently  seen in tierrcsl
studies.  Therefore, the  tumor  incidences were not considered  to be treatment
related.   In addition,  the  incidences  of hyperkeratinosis  and dermatitis
observed in  every   group  including  vehicle controls were  attributed  to the
benzene vehicle.

7.3.3  Minerals
7.3.3.1  Attapulgite.  No  studies on  the carcinogenic  potential of  attapulgite
following inhalation exposure were  found in  the available  literature.  Limited
evidence suggests  that  attapulgite  is carcinogenic following intraperitonecl,
but  not intrapleural,  injection.  Pott et al.  (1974) found that  in  addition tc
glass  fibers,  palygorskite (attapulgite) was tumorigenic when injected intra-
peritoneally into female Wistar rats.   Three 25-mg injections were  ad~inisterec
over a 3-week period to a group of 40 rats.   The lengths  of 37.5 percent c'  the
July 19ES                           7-43             DRA-T--D:

-------
 fibers  were <2 pm,  and 70.0 percent were <5 pm.  The tumor rate was 65 percent,
 and the first tumor appeared at  day 275.
      In a more recent, comprehensive  study by Pott et al. (1967a), the carciro-
 genicity of  attapulgite  obtained from four  different  geographical regions
 (Caceres, Spain;  Georgia,  United States; Lebrija, Spain; Mormoiron, France) was
 tested  via  the intraperitoneal  route  in  female Wistar rats  (see section
 7.3.1.1.4).   The  fiber dimensions, dose levels,  tumor incidences, and lifespans
 are presented in  Table 7-5.   Results  indicated that the attapulgite sample fror.
 Caceres was carcinogenic; a  tumor incidence  of  40  percent was obtained,  com-
 pared to a 6  percent  incidence  for saline controls.  Animals having abdominal
 sarcomas, mesotheliomas,  or  carcinomas were included.   The tumor incidence for
 each of the three  other  samples, which were all <1 pm in length, was approxi-
 mately  4 percent.
      In contrast,  Stanton  et al.  (1981) found that two preparations of attapul-
 gite were not  carcinogenic  when  implanted  intrapleurally  into Osborne-Mendel
 rats;  tumor  incidences  of  7 percent  were  obtained for each  sample.   Beth
 preparations  consisted of  small-diameter,  short fibers;  the  common log  of
 fibers  of the  dimensions ^0.25 x >8 um was 0.  Similar results were obteinec by
 Jaurand et al. (1987)  following  the  intrapleural  administration of 20 mg of
 attapulgite into  male  Sprague-Dawley  rats.  The  mean diameter and  length of the
 attapulgite fibers were 0.06 and 0.77 pm, respectively.  No tumors were fcjnd
 in  36 rats dosed with attapulgite,  and  survival was reportedly comparable to
 controls.
 7.3.3.2  Erionite and Other Zeolites.  Natural  fibrous  zeolites, especially
 erionite, are highly  carcinogenic in rats  via  the  inhalation, intratrachesl,
 intrapleural,  and intraperitoneal routes of administration.   Evidence suggests
•
 that erionite may be  more  carcinogenic following deposition in  the lungs  tha.-.
 asbestos.   Zeolites have also been shown to be carcinogenic after intraperito-
 neal injection in mice,  but no information was found in the available literature
 for other animal  species.   Nonfibrous, synthetic zeolites do not  appear to be
 carcinogenic following intrapleural or intraperitoneal administration.
      Palekar and  Coffin (1986)  reported in a recent abstract that fewer fibers
 of  erionite  produced  more  mesotheliomas  in  F344 rats  after  intratrachecl
 injection as  compared  to  chrysotile or crocidolite  asbestos fibers.   Adcitic'-ici
 details were  not  presented.

-------
     Wagner et al.  (1985)  tested  two samples of natural  fibrous  erionite, and
nonfibrous natural  and  synthetic  erionite,  for carcinogenicity  in  Fischer 344
rats using both  inhalation exposure and intrapleural injection.   Dusts of a
rock sample  from Karain,  Turkey,  which was characterized as an  incompletely
formed erionite, and a sample of Oregon erionite were prepared by disc  milling.
Groups of 20 male and 20 female rats (57 days old) were  exposed  to  mean respir-
able dust  concentrations  of  10 mg/m  Oregon erionite,  nonfibrous  synthetic
erionite, or  crocidolite  asbestos for  7 hours/day, 5  days/week for 1 year, and
were observed until death.  The size distribution of  the Oregon  erionite fibe-s
was as  follows:  a diameter of <0.2 urn  for 56  percent  and <0.4 urn for 92.6 per-
cent of  the  fibers, and a length of 3 to 5 urn for 40.2  percent  and 3  to 10 v~
for 66.7 percent of the fibers.   The particle size of the nonfibrous syr.thef'c
erionite was  >0.5  utr..   Twelve rats/group were sacrificed prior  to  1 yes-  tc
determine dust retention.   Mesotheliomas were found  in  27/28 of  the surviving
rats exposed  to  Oregon  erionite.   The average time-to-tumor was  580 days.   N:
mesotheliomas occurred  in  the asbestos-exposed rats;  1/28 developed a  squares
carcinoma (517 days).   Unexposeti  controls had no tumors (0/28),  and 1/26 rats
exposed to nonfibrous synthetic erionite developed a. mesothelioma (784  days).
     In the  intrapleural  study,  20 mg  of Karain (Turkey) rock fiber (d.-icmete-
of  59  percent <0.2 urn,  length of  4.9  percent >4 urn),  Oregon erionite, c-
nonfibrous natural  zeolite (Japan)  were administered to  rats  in a saline
suspension.    Chrysotile  was used  as a  positive  control.   Mesotheliorr.ss were
found in 40/40  rats injected with  Oregon erionite and  mean time-to-tumsr WES
390 days.  Karain  rock  fiber  caused mesotheliomas  in  38/40  animals with a mesr
time-tc-tumor of 435 days, and 2/40 rats receiving nonfibrous zeolite  developed
mesotheliomas at a mean  time of  715  days.   Nineteen of 40 rats  receiving
asbestos developed  mesotheliomas  (678  days),  and one  mesothelioma  was  fount? in
40 controls (720 days).
     Johnson et  al.  (1984) examined seven pleura! tumors induced  in  rats  b>
inhalation of erionite  using Mstochemical,  immunocytochemical, and electro-
microscopic techniques.   The  mesothelial nature of the  tumors was  confirmee by
a mucopolysaccharide and  keratin  staining pattern and was almost identical  to
human mesotheliomas.  All  of  the  tumors had  both  epithelial  and spindle ce-1
(sarcomatous) components;  the proportions  of each cell  type  varied  wit*-
tumor.   The  pluripotential nature  of  the  tumors  was  indicated by ver-ic
JuU  lSc£                           7-45

-------
degrees of  differentiation;  two tumors had dense cord granules  staining  posi-
tively for neuron-specific enolase, an indicator of endocrine cells  (epithelial
components).  One tumor had osteoblastlike cells, and two were lipid rich;  both
features were  indicators of  mesodermal  expression.   Spindle cells  in most
tumors were  indications  of  fibroblastoid differentiation, although the  cells
were not true fibroblasts.
     Maltoni et al.  (1982a)  tested the carcinogenicity of sedimentary erionite
in groups of 20 male and 20  female 8-week-old Sprague-Dawley  rats after  intra-
pleural or  intraperitoneal  injection.  The animals received  a  single dose of
25 mg  erionite  in 1 ml  water.   By week  53, mesotheliomas developed  in  the
pleura of 9  of  10 animals (6 males  and  3 females) that  died.  Survival was
13/20  for  males  and  17/20  for  females.    None  of the vehicle  controls  and
0/40 rats receiving  intraperitoneal doses of erionite developed  tumors.   IT,
contrast to  these results,  crocidolite caused mesotheliomas  following i-ntra-
peritoneal  but  not  intrapleural  administration.   In an  update  of this study,
Maltoni et  al.  (19825) reported  that  by week 67, 12  of 20  males and 9 of
20 females   that  died developed pleura! mesotheliomas  after  receiving irtra-
pleural injections of  erionite.   None  of the 40 control  animals (20 males  a-.d
2u females)  developed turners.   The final  report  of  this study has not  bee"
published.
     Pott et al.  (1984a) tested fibers obtained  from a  piece  of stone  fror
Ankara, Turkey,  in Sprague-Dawley rats.    Fifty  percent  of  the fibers had a
diameter <0.36 urn, and  90 percent were longer than  15.1 urn.   Intraperitoneel
doses  of 5  and  20 mg resulted  in  tumor  rates  of 65 and  72  percent,  respec-
tively.  The tumors  were sarcomas or mesotheliomas.   The first tumor appeared
at 288 and  222 days  for the low  and  high dose,  respectively.   Mineralogic
testing of  the  sample,  which was  incomplete  at the  time  of publication,
suggested that  the  sample was not a zeolite.  A sample of zeolite from Nevaoe
was not carcinogenic following similar testing.
     In a more recent, comprehensive study by Pott et al. (1967a), the carcino-
genicity of  erionite  samples from Karain  (Turkey) and Oregon were  tested  in
female Wistar  and Sprague-Dawley  rats via  the  intraperitoneal route (see
section 7.3.1.1.4).   The fiber  dimensions,  dose  levels,  tumor incidences.
and lifespans are presented  in Table 7-5.   No  information  on the  composition
of the fiber samples was reported, however.  Both  samples  were carcinogenic.
A dose-related  increase  in  tumor incidence was  noted in  Wista"  rets dose:  *:t ,

-------
erionite  from  Oregon;  incidences of 48 and 90 percent  were  obtained in the
low- and  high-dose  groups,  respectively.   The mean times-to-tumor were  98 and
80  weeks.   Sprague-Dawley rats proved to  be  slightly more  sensitive to the
carcinogenic effects of the Turkish erionite than Wistar rats  (see Table 7-5).
     Ozesmi et al.  (1985)  also tested  a zeolite  from the village of- Karain for
induction  of  tumors in Swiss  mice  using  the  intraperitoneal  route.   A rock
from Karain containing  both fibrous and nonfibrous  zeolite was ground to dust.
Characterization of  fiber  content and  size were  not reported.  One  milli'lite"
of  a suspension  containing 5, 10, 15,  20, 30, or 40 mg  of the dust  in physic-
logic saline was  injected  intraperitoneally into groups of 37 to 98 6-week-old
mice, which were  followed  until  death  (up to  32  months).  Tumor incidences ere
presented  in  Table 7-9.  The  predominant tumor  type was  mesctheliome.  The
incidence  of tumors  did not appear to  be  dose related.  However, this may have
been due  to a  disparity in survival at the higher dose  levels;  survival after
6 months was not  reported.   The authors stated that no  turners occurred before
9 months.   Three mesctheliomas and one lymphoma occurred in  55 saline cor*re's.
     Suzuki (19£2)  studied  the carcinogenic potential of two  natural  zeolites,
erionite and mordenite,  after intraperitoneal injection into male Swiss mice.
The  erionite  sample consisted  of fibers  that were 0.4 to  24.urn in  length
(mean length  1 urn)  and  0.01 to 2.5 urn in diameter (mean diameter  0.1 urn).
The  mordenite  sample,  which  consisted of granular and fibrous particles.
was  0.33  to  5.7 um  in  length  and  0.05 to 1.67 urn  in diameter.  A  total  of
45  mice  received a  single  intraperitoneal  injection  of 10  or 30 mg  zeolite
in  1 ml  of  saline  solution.   Six of  10 mice  receiving 10  mg erionite
developed  tumors  (4 mesotheliomas).   Fibrosis  occurred in mice  dosed witn
both zeolite samples.
     Suzuki  and Kohyama  (1984)  further   studied  the careinogenicity  of
natural  erionite  and mordenite, and of a synthetic zeolite,  after a  single
intraperitoneal  injection   in  male BALB/c mice.   In two natural  erionite
samples (from  the western  United States), 90 to 95 percent of the  fibers
had  a  length  <8 um, and 82 to 100 percent had  a diameter  <1 um.   Erionite
II  was  purer  than  erionite  I.   Mordenite was  a mixture  of  fibrous  anc
granular  forms;  94 percent of the  fibers  had lengths <3 um, and 89  percent
had diameters  <1 um.  Synthetic zeolite 4A was about 2.4 x 2.24 um.
I..TV. ictc                           7-i7             DRArT--c:

-------
             TABLE 7-9.  PERITONEAL TUMOR INCIDENCES IN MICE AFTER
                   INTRAPERITONEAL INJECTION OF KARAIN DUST
Amount of
dust (mg)
Control
(saline)
5
10
15
20
30
40
Total
No. Of
mice

55
69
97
98
45
45
37
391
No.
surviving
6 months

46
55
81
73
43
39
30
321
No.
Meso-
thelioma

3
6
12
10
6
3
4
41
of tumors
Lymphoma

1
0
11
7
5
4
4
31

Both

0
4
4
0
0
1
2
11
Percent
with
tumors

8.7
18.1
33
22.3
2r
20.5
33
26.1
Source:  Ozes-i et al. (19£5).

     Groups of 20,  50,  and 75 mice  received  0.5,  2.0,  and  10  mg erionite  II,
respectively.   Mortality was excessive at the high dose; consequently, a seccr.c
group of  50 mice was dosed with  10  mg  erionite  I.  Between 7  and  23 months
after exposure,  6/18,  24/44,  and 3/8 mice developed  peritoneal  tumors at  acse;
of  0.5,  2.0,   and  10 mg erionite  II, respectively;   in  the second  group.
21/42 mice developed tumors at a  dose  of 10 mg of erionite I.   Six cf the
54 tumors  were plasmocytomas, and 48 were  mesotheliomas.   The mesothelion-cs
resembled  asbestos-induced tumors in their  gross  appearance,  histology,  and
close association with fibrosis.  No  tumors were observed in groups of 118 rr.-ice
receiving  saline, 44 receiving 10 mg  mordenite, or 46 receiving 10 mg synthetic
zeolite 4A.
7.3.3.3   Talc.   Talc is  not carcinogenic  following  inhalation exposure  or
intraperitoneal  and  intrapleural  administration  to rats, hamsters,  and mice.
Evidence  from  one  study suggests that talc  may  be an effective cocarcinoge-.
when administered with benzo(a)pyrene (BaP).  Ingestion of talc  (100 ing/ret/is..)
for 101  days  resulted in  the development  of  leimyosarcome  of the storr.*:* i-i
1/32 Wistar rats observed  for life (Wagner et al., 1977).  No centre's oeve'opec
tumors.    Inhalation  exposure  to 10.8 mg/ir,  talc  for  7.5 hours/day.  5 dcys-»ee<
                                    7-ic

-------
for 12 months, resulted in a pulmonary adenoma in 1/24  Wistar  rats observed for
life (Wagner et al., 1977).  No pulmonary tumors  were noted  in 48 control rats.
Wehner et al. (1977) exposed groups of 50 male and 50 female Syrian  hamsters to
8 mg/m  talc  2.5  hours/day,  5 days/week for up to 300  days.   The animals were
then killed.  No carcinogenic effects were found.
     In a chronic study by Stenback et al.  (1986), groups  of 48  Syrian  hamsters
received  weekly  intratracheal injections of  either 3 mg talc  alone or in
combination  with  3 mg BaP in a  saline vehicle.   Talc  alone did not induce
tumors.   However,  in  combination with BaP,  talc  was found  to be an effective
cocarcinogen, inducing increased incidences  of alveolar  tumors  when compare:
to BaP  or talc alone.  Single  intrapleural  administration  of 40  mg talc tc
Osborne-Mendel rats (Stanton et al., 1981) or 20  mg talc  to  Wistar rats (Warper
et al., 1977) did  not result in tumor  induction  after lifetime observation.
Similarly,  no tumors were found after four intraperitoneal  injections  of  25  rr.g
in 40  female  Wistar rats  (Pott  et  al.,  1974)  or  after  a single  intraperitciea"
injection of  20 mg  talc  in 24 Swiss  mice  (Ozesmi  et al.,  19E5) observe:!  *f
life.   Hamilton  et al.  (1954)  injected talc  (10 mg)  intrabursally on tne
surface of ovaries of rats; the animals were observed  for  18 months.  Nc turcr-s
were found.
7.3.3.4   Vermiculite.  Results  from one study indicate that vermiculite is net
carcinogenic  following intrapleural  administration.  Hunter and Thomson (1873)
tested the carcinogenicity of vermiculite after intrapleural injection  of  25  me
in 0.2 ml saline  into a  group of  21-day-old  female Sprague-Dawley  rats.  The
vermiculite sample  consisted  of  amorphous particles and had  a  particle  sire
93 percent  <5 um  and 37  percent <2 urn.  The  animals  receiving vermiculite
developed granulomas  in the  lung and viscera, but  no  tumors  developed after
104 weeks.   A comparable  dose of  chrysotile  asbestos  produced  mesothelio^as
in 48 percent of the rats.

7.3.4  Other Fibers
     Potassium titanate was   carcinogenic after  intrapleural  administration,
                                         £
and limited  evidence  suggests that Fybex , a commercially available potass^u-
octatitanate, may  be  carcinogenic after inhalation  in  hamsters.    Calciun-
silicate  fibers and wollastonite (a natural calcium silicate) have  bee" s".cw.
to be  noncarcinogenic following inhalation exposure and  intraperitoneel  e-=
intrapleural administration in rats.  This may be due  to the lack  of d-re:-'"iity
  •iv i9:S                           1-^             Ds
-------
of these  fibers In vivo.  Several  other  manmade  and natural  mineral fibers
including  dawsonite  and aluminum oxide whiskers induced tumors  after  intra-
pleural administration,  but  studies  by  other routes  of  exposure were not
available.   Ferroactinolite,  another fibrous mineral, was  found to be more
carcinogenic after intratracheal than intrapleural  administration.
     Lee  et  al.  (1981)  studied the pulmonary responses of  potassium octatita-
nate  (Fybe*  )  after inhalation  exposure  in male  Sprague-Dawley  rats,  male
albino guinea  pigs,  and male hamsters (strain not specified).   Exposure  was
for  6  hours/day, 5  days/week  for 3 months, at concentrations  ranging  fror
2.9 x  106 to  41.8  x 106 fibers/L.   Fybex®  had  an average size  of  6.7 by
                                                                 B
0.2 urn, and  animals were  examined at 18  and 24  months.   Fybex  produced
mesotheliomas  in  1/12  hamsters exposed at 2.9 x 10  fibers/L, 1/13 exposed at
13.5 x 106 fibers/L, and 1/16 exposed at 41.8 x 106 fibers/L.   No mescthe 1 iorr.ss
were found in  control  hamsters (0/34), and  no tumors  developed  in  any  of the
Fybex^-exposed  rats  or  guinea  pigs.   Although the  development of  mesothel iorr;=s
ir. hsrr-sters  cannot  be  ignored, the short-term exposure, the  small  numbe- c*
animals tested,  and  the smell  number  of tumors produced  in  the  lungs precluoed
any conclusions on the carcinogenic-potential of Fybex .
     Stanton et al.  (1981) reported tumor incidences of 69 and  72  percent  in
Osborne-Mendel  rats  after single  intrapleural  injections  of 40 mg of  two
potassium octatitanate samples (Table 7-9).
     The chronic inhalation toxicity of manmade calcium silicate (wollastonite)
insulation materials (Ca^OH^SigO^-H^O)  in rats was reported by Bolton et al.
(1985).  Groups of 48 male Wistar rats were exposed for 7 hours/day, 5  days/week
for 12 months, to 10 mg/m (respirable fibers only)  of three  different  sar.ples
(A, B, and C) of calcium silicate.   Fiber diameter ranged  from approximately 1
to 10  urn.   Control animals were  exposed  to  filtered air.   All   animals were
observed  for an additional 12 months postexposure.   Pulmonary neoplasms occurred
in 2/139  rats  exposed to calcium silicate (both rats were  exposed to sample B)
and were  described  as  a focal  squamous carcinoma of  low-grade malignancy  and a
benign adenomatous  lesion.   No pulmonary  tumors were found in  control  rats.
One rat from the calcium silicate-exposed group developed a peritoneal  mesctne-
liorna, which the authors considered to be spontaneous.
     No tumors were  induced  after intraperitoneal  injection of a single 2r-rcc
dose of each of three calcium  silicate samples  in groups  of 36  male  Wists"
rats (Bclton  et al. ,  1986).  Wcllastonite was  also found to be

-------
in  female  Wistar rats  after  intraperitoneal  injection of 20 mg weekly  for
5 weeks  (total  dose  100 mg)  in saline (Pott et al., 1987b).  No tumors  were
found.   The  authors hypothesized that wollastonite may not  be  durable  in a
saline vehicle.
      Pott et al.  (1974)  found  that nemalite was tumorigenic  after  four  intra-
peritoneal injections of 25 mg in Wistar rats.   For the nemalite fibers  used,
91.6  percent  had diameters <2 urn and 96.4 percent had lengths <5 urn.   The
tumor  rate  was 62.5 percent,  and the first tumor appeared after  249 days.
Gypsuir, (4 x  25 mg)  produced  tumors in 5 percent  of the rats. Granular  dusts
(actinolite, biotite, hematite,  pectolite,  and  sanidine)  were inactive  after
similar exposure.
     The most comprehensive survey to date on  fiber induction of tumors  fo1"!ow-
ing  intrapleural  injection has been published  by Stanton et  al.  (19E1).
Seventy-two  experiments were  conducted  with different  fibrous materials.
Dimensional  ranges  of   fibers  were  tested  for  many groups:   13 samples  cf
crocidclite asbestos,  22 samples  of  classes,  8  samples  of aluminum oxiae.
(AI^CO whiskers, 7  samples  of talc, 7 samples  of-dawsonite,  and 4 samples  cf
wollastonite.    Also  studied  were two  samples  of  tremolite,  attapulgite.
potassium octatitanate,  and halloysite  (a  natural fibrous aluminum silicate),
anc! single samples  of  nickel  titanate, silicon  carbide,  and amosite  asbestos.
A standard dose  of  40 mg dispensed  in hardened  gelatin was applied to the left
pleura!  surface  of  groups of  30 to  50  female Osborne-Mendel  rats  (12  to
20 weeks old), and the rats were followed for  2  years.   Control  groups  included
untreated rats, rats  receiving thoracotomies but  no pleura!  implants, and rats
receiving pleura! implants of  nonfibrous material.  Size  distribution  of each
of the 72 preparations was presented, and fiber  counts  in  discrete-size  catego-
ries  were reported  as  the common log; electron-microscopic photomontages with
magnification of  about  5,000  to 100,000 were  used for  size measurements. The
actual incidence  of  tumors was counted,  and life-table analysis was used to
calculate percent tumor  probability.  Table 7-10  presents  selected data.
Fibers  of  potassium octatitanate,  silicon carbide, synthetic  and  natural
dawsonite (expt.  21),  fibrous  glass,  and Al.O,  whiskers were  a!! carcinogen:,
and the  predominant  fiber size was <0.25  x >8.0  urn.   Representative date fc-
some  asbestos  fibers are  also presented  in  Table 7-10.   Nickel  tita-ste.
halloysite,  attapulgite,  and  talc samples   (common log  fibers/ug = 0) we'e all
negative;  Al.O,  (log = 2.60)  was marginal  (4/25 tumors).   Glass fibers  wif

-------
          TABLE 7-10.   SELECTED DATA ON CARC1NOGENICITY  OF  DIFFERENT
         FIBROUS MATERIALS FOLLOWING JNTRAPLEURAL IMPLANTATION  IN  RATS
Compound
Nonasbestos
Potassium octatitanate
Potassium octatitanate
Silicon carbide
Dawsonite 5
Dawsonite 1
Dawsonite 7
Dawsonite 4
Dawsonite 3
Glass 1
Glass 2
Glass 3
Glass 4
Glass 5
Al-,03 1
AUO-: 2
AiJDs 3
Asbestos
Tremolite 1
Crocidolite 1
Crocidolite 3
Amosite
Control s
Glass 18
Untreated
Expt.
No.

1
2
3
4
7
20
21
22
13
15
16
17
19
18 .
27
28

5
e
10
11

67
* "
Actual
tumor
incidence

21/29
20/29
17/26
26/29
20/25
16/30
11/26
9/24
9/17
12/31
20/29
18/29
16/25
15/24
8/27
9/27

22/28
18/27
15/23
14/25

0/115
3/488
Percent tumgr
probability
(± SO)

95 * 4.7
100
100
100
95 ± 4.8
68 ± 9.8
66 ± 12.2
66 4 13.4
85 4 13.2
77 4 16.6
74 4 8.5
71 4 9.1
69 1 9.6
70 4 10.2
44 4 1.7
41 ± 10.5

100
94 i 6.0
93 4 6.9
93 4 7.1

0
••
Common log
fibers/ug
(<0.25 x >6 prr.)

4.94
4.70
5.15
4.94
4.66
4.71
4.01
5.73
5.16
4.29
3.59
4.02
3.00
3.63
2. 95
2.47

3.14
5.21
5.01
3.53

0
• •
Calculated by life-table analysis.

bDawsonite (NaAl(OH)2C03) samples 5, 1, 7, and 3 were synthetic,  and
 sample 4 was a natural fiber from Tanzania.

 Source:  Stanton et al. (1981).

-------
log  -  0 to 1.30 were  negative,  as  were samples of dawsonite, wollastonite,
attapulgite, and crocidolite with a log of 0.   Two .ceramic  glass  samples, glass
21 (80  percent  A1203)  and glass  22  (90 percent  Zr02), were also inactive.   The
correlation coefficient of a graph of logit (In p/l-p)  of tumor probability  (p)
versus  the  log  of  the number of particles/mg with dimensions  of £0.25  x >£  ur
was  0.80,  but  relatively  high correlation coefficients (0.68  and 6.63) were
also observed  in categories  having dimensions of 1.5 x >8.0 pm and  of  <0.25 x
>4 to 8 pm.
     Coffin et  al.   (1982,  1983)  compared ferroactinolite and  amosite  asbestos
(cummingtonite-grunerite)  for tumorigenic response  after  administration  to
Fischer 344 rats via the  intrapleural or  intratracheal  routes and  conclude:!
that in this  case,  intrapleural  administration failed  to predict the  carcino-
genic potential of  ferroactinolite in the lungs.  The fibrous  classification c*
ferroactinolite is  apparently  debatable (Zoltai,  1986),  but for the purpose cf
this document  it is considered to be an asbestiform mineral.   Single  dcses  cf
20 mg were  administered via  the  intrapleural route; 12 doses  of  either C.E  rr;
             te or 0.2 mg amosite were administered via the intratraches"!  rojte
for  a  total  of  6.0  or  3.08 urn, where*s
only O.OS  x  10   fibers/ing  ferroactinolite  were in  the  same  size  category.
However, the incidence  of pleura!  tumors  was  approximately equal (Table 7-11).
There was a  significant increase in mesotheliomas in both the ferroactinolite
and  amosite  groups  and  a slight  but significant (p = 0.05) increase  in pooled
primary  lung tumors  for ferroactinolite  in the  intratracheally dosed group.
Ferroactinolite was more carcinogenic on the basis of total fiber dose or  fiber
dose of  the  selective  size  category (^0.25 x  >8 urn) following  intratracheal
instillation.  When  the lungs of  rats  were examined at various times after
intratracheal injection, the ferroactinolite fibers present were split,  and the
number of thinner fibers increased (Cook et al.,  1982).   The number  of long,
thin fibers  present  after a year was  similar  for both amosite and ferroactinc-
lite groups (Table 7-12).  The ferroactinolite sample had essentially no fibers
meeting Stanton's criteria at the time of intratracheal instillation,  but  fiber
splitting produced 24.6 x 10  of such  fibers  after  1 year whereas  the ccr'-e-
sponding amosite dose had 18.6 x 10  fibers at 1 year.
     A tabular summary  of carcinogenicity data  for  each  material  is  found  in
Table 7-13.
Jjiy 1965                           7-53             DRArT--:: rC" QL/1TE Z-". CITE

-------
           TABLE 7-11.  INCIDENCE OF PLEURAL TUMORS INDUCED IN F344
                      RATS BY FERROACTINOLITE AND AMOSITE
Route/fiber
Intrapleural
Ferroactinolite
Amosite
Sham control
Untreated control
Intratracheal
Ferroactinolite
Amosite
Sharr control
Untreated control
No. of
animals
135
137
134
219
561
139
212
219
Percent of
Meso-
theliomas
23
33.6
1.5
0
0.89
1.4 '
0
0
animals
Lung
tumors
1.5
0.73
0.75
0.4
3.9
0.71
0.6
0.43
with tumors
Metastatic
tumors
0
0
0
0.46
2.1
0
0.9-:
Source: Coffin et al. (1963).
      TAS.E 7-12.  SIZES OF MINERAL FIBERS RETAINED IN THE LUNGS OF RATS
             AT VARIOUS INTERVALS AFTER INTRATRACHEAL INSTILLATION
Fiber/time
Ferroaciinolite
1 dayc
4 months
12 months'"
Amosite.
4 months
12 months0
Fibers/mo
(x 1C6)"
107
262
366
585
393
202
Mean length
2.86
2.00
1.77
2.60
3.02
3.62
Mean width
(MB!)
0.37
0.17
0.11
0.24
0.22
0.21
Asrecj
10.6
10.9
15.7
20.1
8A function of length and width.
 Single instillation of 0.5 nig ferroactinolite or 0.25 mg amosite.
GTwelve weekly instillations for a total of 6 mg ferroactinolite  or 3 mg
 amosite.

Source:  Cook et al. (1982).
     ~\ CiC.
                                    7-54
DR-r7--:o N:T £J:~E --•  -• • =

-------
1AIIII  7-1.1.  SUMMARY 01  URClNOr.l Nlcl IV  MIIIMIS  C1N HHMI .  SYMIMItM. I HUN',. AND  HINIRnlS
i.il 1 ihi-r Rinile
i'ii'. ijl.r.s inhalation
Inhalal ion
Inh.ilation
Itihalat ion
Inhalation
Inhalation
Inhalation
Inhalation
Idlralracheal
Intratracheal
lot.ralr.icheal

Inl ralrarhral
Sp«r. les/slrain
H.tt/f. Os,horne-
Mpn.lel;
h,imsler/M, Syrian
flat/M. Spraque-
llawley;
nuinpa pig/M;
hamster
Rat/F. Osborne-
Mendel;
hamsler/H. Syrian
Rat/F. Wistar
Rat. Fischer 344
Rat. Fischer 344
Monkey /N, cynowolgus
Baboon/N
Hamster, Syrian
Hamster. Syrian
Ral/f , Oshorne-
li'imvter/M, Syrian
Rat/f. Wislar
n,,,. 	
0. .1 at- 1 mii/m1,
f> h/.l. S d/wk,
inn wk
470 mij/m1
h h/il. S  ini o»t lustvp
mi'.iii il:0 4S. 3.1. No lung tumors - 70
V 4 , or f>. 1 |im
•IM P»4 IK/perplasI*. - 70
,1 'MH '0.11 |im I liimor
Mrt '11.47 (im
l:'Mit . 17.4 nm
'.lit • 1 n |im
.IM ion |/4fl developed tumor. - 10
il • .1 |im 7/4R developed hyper-
1: - S inn plasia
70 |i»
il •l.'j-C. |im No lung tumors - 73
1:. Ill ami -70 |im
d:ft.l-4 |im No lung tumors - 79
1 . S- 10(1 |im •
il: ISt '07'. |im terminated at Week - 6
l'>ialiX <7.fl |im developed respiratory
tumors, mostly
mpsutheliomas
W...HI 
-------
(AMI I  Ml    (mill iim<-il)

iin-i.it 1 ilinr Rnulp Ippr IpS/stralll
tl.ii.u-, yl.i-.s lntr.ipleiir.il R.tt/F. ffsliorne-
<"»'l » Mendel



Inlraplpiiral Rat/F . Wistar





Inlraple.ural Ri«* » SP*«q«e-0.i««ley


lntr,ipletiral Rat. Spraque-Oawley


Intrapletiral Mouse. BAlB/c



Inl.rapleural Hamster, Syrian









Intr.tplpiiral Hamster, Syrian

Inl raperitoneal Rat/F, Wistar



litlraprrilnneal Ral/F , Wlslsr




1 ilii-i SI /i'/ll<'si ri|.l mil
Ilits.iilP (il ili.imi'lri . 1 lrn<|lli)
10 *M|. l» il O.I. /'. |in:

lll |i*i
7S *H|, IK mi-an il: b |im ( insul.il inn
l|l Mill')
0 ', IM| .IM 101
il •I0t-0. mil .w.ii l.ililc
7% mi|. 1«



Comwenls Carclnoqpn Roferrncp
Proh^liillly of I
plpnral sarcoma:
M-RSX »
M-77X •
«-nx
7
0/3? lunq tumors


4/J? developed •
mpsothel lomas
4/4R developed » III
mesotheliomas

6/4S developed • 1
mpsolheliomas

No tumors - .11



9/60 developed « 4
tumors. 1st at d 784
7/60 developed « 4
1 timers, 1st at d 4 OH
2/60 developed • 4
tumors, 1st at
ri i>11
No tumor* In - 4
Kin animals

No tumors In 60 - 5
animals
in tumor rate . • 71
evposed
compared to 6Y
in controls
1/40. - 1?
4/40, and •
7.1/40 ilpve loped •
IIIMI) tumors.
rrspiM. lively
                                                         (unit iniii-il 
-------
IAHII /-It    (mill i mini)

HiniM.il 1 ilirr Rmilp
1 ilimii', ijl.r.s Inlrapprilnnpal
(mill )








Inlraperitoneal



Inlraperitoneal

Inlraperitoneal

Inlraperitoneal


i.l.iss wool Inhalation



Inlrapletirat
Miin-i.il wiml
(fill I Wllll I
Inhalation



Inlraplpwal
Intraperiloneal





Species/tlr.iin
Ral/f , Wistar









Hamster/M, Syrian

Rat/F . Osbornp-
Mpndpl
Rat, Spraqiie-
Oawley
Rat/F, Wlslar

Rat/F. Wistar


Rat, Fischer 344



Rat. Spr ague -Daw ley


Rat. Fischer 344



Ral, Spraqiie-Oawlpy
Ral/t, Spraipie-
Pawley



1 iln-r '.i/i'/Di-M i ipt inn
nn'..it|i- (il ili.im-ti-i , 1 li-ni|lh)
7 mil. l« HM 1114
.1 MIT .0 7(1 |im
I iirr. • 1 1 |tm
10 lm|. l»

7'.» mil. ?«

70 «M|. Ic HN 1 1?
d I |im
I:, 'II |im
?b «M|, Ix mr.in il 0 4'i (in
1: / '• |.«
?S «M|. Ix nn-.iii il:O.4S \tm
\:l ', |im
7-10 «H|. l» il:-0.:H |im
1 '" " 1™
(1 ',-10 «i|. Ix •
S d/wk. 17 IMI 1: •'• |im
nh'.prve 74 «HI

70 M|. Ix il:« .1 l«"
10 «M|. Ix i! Vlt OT ,11 im
7*> M|. .Ix il MIT • 1 sothel iomas
or sarcomas
2/95 developed
limq neoplasms
4/9'j developed
hyperplasia
1/4R developed
Mfsothet iooa


7/4B developed
liMors; I/4B
tlpyp loped
hyperplasla
S/% developed
•psolhellonas
0/4S developed
Immirs
|(,X liinor inriilpncp
(inpsolhpl ltK«*s or
•..tr i >HI>.I\ ). I1. 1 limnr
al .1 7/4

Carcinmjen Rplerente
i 33


'

•

« 13


70

« 20

« 39

« 39

« 7?


i 10



J 10


10



• 10
II
• II



                                                   (i mil i mini  im Mn> liillnwim) p*<|i>)

-------
I All! I  Ml    (l 1.1,1 ilHM.I)
.il 1 ilii-r Rnulp
• 1 1 Wllltl
1 wniil Inl r.ippri Inneal
i.nl (
Inl r.ippri loneal
"1 win.) lnli.il.it inn
Idl.raplpiiral
Inlrapleurat
1 nt raper i loneal
nil 1 iliprs
mi in um lnli.ll.il Ion
. t 1 ii .it P
Intraplpural
Intraperiloneat
, .11 In, v Inhalation
cr.imii
ih...-.
ii. i 1 r.ii""' )
Inl r.il r.trhpal
Inl r.i|n%r ilni.pal
Inl r..|ipr itnnpal
Species/strain
R.»l /F . Spraf|up-
ll.iwlpy
R.it /' . Spra<|up-
ll.iwlpy
Rat/I . Osborne-
Henile 1 ;
hamslpr/M, Syrian
»at/F. Wistar
Rat. Sprague-Dawley
Ral/F. Wislar
Ral. AF Han Wislar
Ral. Wislar
Ral. AF Han Wislar
llamsler/M. Syrian
Ral/F , Osborne-
Mpmle I
Hat/F. Osbornr-
Hrnile 1 ;
li.imslpr/M, Syrian
II.IBC. tpr/H. Syr i. in
H..I/I . nsln.rnp-
1. 	
10 «i|. Ix
7'. mi|. wprkly,
Ix
10 mil/Hi1, f, hAI,
', rl/wk. 104 wk
5 «|. Ix
70 Ml). Ix
70 m<|. wppkly.
?x
H 4 MCI/Hi1
(ar.lual rps-
litr^lilp conr. . )
/ hAI. 5 d/wk.
1? wo
70 m<|. Ix
7S mi|. Ix
C. h/.l. 5 d/wk.
1(14 wk
17 mil/in1
104 wk
7 m>t. wppkly, '•*
?', m,,
r, .,
1 ilic. '•i/c/llcsi cipl inn
(il ili.imi-lr, , 1 l«.i.i|lli) CoMRpnls Carclno«|pn
il n 1.1 |im ITX liMor incidence •
11 I |.m
•I I '14 |im /|X liMior incidence •
I / 1 |im
MP.IK d 7. / |im Mo lunq linnors
1 40 ,iin
il SOT . o in |im SX tunor rate. >
1 :M« • 7 / |im 1st Itmor at
d bZI
il • '. |im No tumors In 48 rals
d 1 '. -7 «• |im 7-6X liMMr Incidence 1
1 11 71. |im
.1:0 1 |im 1 adenoma. 3 carcinomas.
1 : 'HO; . .1 |im 4 hisliocylomas of
lungs in 40 rats
il:O.S 1 n pm 3/31 developed t
mesolheliomas
il 0 1 |im 1/37 developed t
1 '10% • 1 pm mesolhelloma.
7/37 developed
f il.rosarcomas
mc.nt .1: l.ll |m 1//0 developed t
1 . .IS |im mpsothelinma
mi-.in .1 l.ll pm No tumors
1: IS |i«
•M'.III il 1 . II |im No tumors
1 I'. |,m
mi-.m il l.ll pm //.If. developed •
1 C, ,,m mpsolliel iiMias
mi-.in il 1 II |im 11/7 1 il.wplopeil '
I |i, |,m liimni". (M mi-siillii'l
Rplpll'lli p
3'l
3'»
7(1
II
10
39
13
311
11
70
70
711
70
"'
                                                               iiim.r.)
                                                               i	i,.,n..l  mi  I IIP  InllnwiiKl  l>.»i|r)

-------
1AIIII  Ml    (MIII! iniM-il)

Mififr.fl 1 iluT
1 ilii-rl r.l«M


MAN


•..ill ilw 1 ilii-rs
(Al-.ll,)

Milt Al .O ,

•Mil /«(!

I'nl .<-.-. mm
in t .it 1 1 .iii.il i*
1 yl.l-.M






Al .11,
wlif.li-l •.

Ai .imiil 1 HUM s
tl>vl.llW













*""'•*
lntr.i|iprltnneal


Inlr.ipprilnneal


Inhalation


Inlrapleurat

liilrapleiirat



Initiation



Inlrapleural


Inlrapleural



Inhalation













Spp» les/s train
Ral/l . Wistar

,
Rat/r. Wistar


Rat. Wislar


Ral/f , Osborne-
Mpnde 1
Rat/r. Osborne-
Mende 1


Rat/H, Sprague-
Oawley;
hamsler/H;
quinea pig/H
Rat/r . Osborne-
Hende 1

Ral/r, Oshorne-
Hendel


Rat. Crl:CD BR












1 ilift '.i /iVIIr'.c ript iiitt
I»«S.I,,|. (.1 di.nw-lr,. 1 l.-»H|lh)
4't mi. wrrkly. 1 Mill • 14 \im
'.« ll.'MIl -74 (in

t'j nil). wrrHy. 1 ••Wii: • M |im
SK  d/wV. 1 III .S-1,7 |im
lid wk
40 «H|, IK l.ni|i> ili.imrlrr

40 *H|. l« l.in|p .ll.im.-trr



7 9-41 fl x III •••.in d II 7 |if"
tiltprs/l mi-.in II, ; |im
fi h/d. 'j d/wk.
3 mo
40 f*f|, Ix l'"| (it>P.rs/|ii|
• »l 7?i « •» |im:
4 /ll-4.')4
40 *N|, IX llH| 1 il>ITS/|l<|
• II . 7S x -R |im:
0 1 /«

« h/day. 'j d/wk 1: /'.Y -70 |M
1114 wk d: 'Hit • 1 ') |im
II. OR mi/m*
0. 17 HHi/di1
n 1,1 •!)/•'





ft li/.l.iy. S il/wfc
'.7 wit
7.71 M|/«I'>

Conwenl*
.11/47 developed
atidnninal nesothp-
liomas or sarcomas
11/54 developed
atidtwinal mesolnp-
lii>*MS or sarciMs
Ho, pulmonary tumors


7/47 developed
pleural sarcomas
1/45 developed
pleural sarcomas


Small numbers of
wsolhpl innias
in hamstprs

69-77X tumor
incidence

4-f.Jt tumor
incidence




No tumors
No tumors
4/69 females oVvet-
oped cystic keratin-
i/ini| squamoiis re 11
carcinomas and d/h'l
fpmales developed
siniamnus cell
7/V7 rats developed
cystic keratinl/imi
sifiMmnus cell tar-

tare Inoqen Rpfprpnrp
» 7?


« 77


- 14


f 17

1 17



i 'i



• \7


• 1?





74
74
« 74





« 74


                                                     t tin MM s and  I /Mi
                                                     tpn.ilPS developed
                                                     Miu.inmus rpll
                                                     m,.|.,|.|.r. U

                                                     (lonl imird on MM-  !«>ll»wiiw|  |MI|P)

-------
                                                            I Milt  /  II    (< mil in	I)
1 ilwr Roul.p SpPC IPS/sl rain
il.i-is •
<•> Inl i.iprrilonpal Ral./l , Wistar
lul i .ipiT ilonpal Rat. Wislar

liilr.ipprtloneal Ral/l , Wislar
1 i n
tupvlpiip Intratrachpal Ral/M. Long-Evans
Inlrappr itoneal Rat/I, Wislar
Normal Hnuse/H, C3M/IIP.I
rirrm.il Mnuse/M. C3ll/llp.t
Oi'im.il Moiise/M, C3H/IIP.I
.i.i.li/nl llr.m.il MicP/N. C3M/HP.I
lijiti- lnlr.iplpur.il H.il/M. SpraqiH--'
|,M k ili-) D.iwlpy
Oos.,,1,
70 mi|.
wppk ly . 'IK
II 7'. m<|. IK
7.S m>|, IK
7S m|, IK
10 mi|, IK
1.7V m|. IK
(0.7'jt in O.f» ml
of .1 II It «.olulilp
III IwPI-ll Ml/W.ltl'l )
10 mi|. fm, wpi-k ly
1.9 m| (7% |il
ol a lilt (w/v)
susppnsion),
3x/wk lor 1 ilp
?. 0 HHJ. .Ix/wk
lor 1 ilp
II m.|. .Wwk
lor lilp
1.3 m|. IK/W.
lor lilp
70 m|. IK
1 iltrr '• i /p/lli«M.r i|il iim
( as ,I|>I>VP 1 sqiiamous cell car-
cinoma, 1 flhro-
sarcoma. 1 heman-
qiosarcom* on skin
at nontesl site
in 40 mice
•..imp a', alutvp No tumors
'..imp as almvp 1/40 leimyosarroma
on skin at nonlesl
si IP
tin nil |im No tumors developed
1 II // |im
Ri-lfrcn
77
on

3')
7'i
77
?n
7f.
7fi
7f.
I/
Inl i .iplpiiral         H.I I/I, Oslmrnp-            111  mi|.  l»
                       Hi-iiilp I

11,11 ,I|IPI iliiiip.il     II,11/I. Wi'.l.u               7S  m<|.  IK
lin| I ilipr-./mi)  • II. 7'i «  -M |im:
II

I  I/  '.•!. .  7  |im.  /lit • '.  |im
                                                                                                                        ft tumor  im:ldpnti>
li'it ImiHir  r.«|p,
l-.l  .il  71'« il
                                                                                                                        (i mil initpil mi  tin-  lol liiwuii) |i.H|p)

-------
                                                                                      I Atll I  M I.    (niiillmii-il)
 il  I iln-1





 I I.-









 I.I Ml)
   I'll)
ill. II. I)
il iir.i I )
Until P


Int r.iper i lonpa 1
InlrAppriloneal
Int.rappriloneal
1 nlraper i toneal

Int taper itoneal
Inh.italion
lnh.il all on
Int ralracheal
Int.r.iplpiiral
Inl r.iplpurat
In! r.ipprilnnpal
lin|i'-.l inn
•.per |p pm
7.S h/d. 5 d/wk.
1110 il
3 m|, wppkly p.irlir. IP si/p: '.in^im
for lilp
(nil 1 ilifi-s/mq' -0. 7Sx -M pm:
40 m|, IK II
70 mi|. Ix mp.ni p.irl ii IP sirr ?'• |im
?!> mi. 4» •.!/!• nut. i h.ir.n IPI i/nl
Mill mi|/r.il/il.iy mi-. in p. irtii.tr '.I/P: ;"> pm
I'll |ll| ll.ly.
Comments Carcinogen Hpfprpnce

31/371 lymphomas
11/171 lymfitiomas and
mpsnlhelionias
7 IX tumor Incidence » V)
77X tumor Incidence • .I'l
BIX tumor incidence • J'l
/OX tumor incidence * 3*1
4BX tumor Incidence • 3'l
91/X tumor Incidence • 31
6SX tumor Incidence. * II
1st at 7R8 d
77X tumor Incidence. • II
Isl al 77? d
6/18. * I/
74/44. and
71/4? most tumors
were mesotheliomas,
respectively
No tumors - 34
No tumors - 3fc
No tumors - Vj
No tumors - 17
No timmrs - 14
7.SX tumor Incldpnce - IS
I/I? dpveloppd - '4
stiim.uli Inmir
                                                                                                                                                   (i mil i illicit  nil  Utf  Inllnwimi  p.ii|p )

-------
Mill)  7-11.    (timl imii-.l)
iii't.il 1 il>«-r Rniilr
Itilr.iperiloneal
"•I 1 ili« Intraperitoneal
I'.llv'l"' '«.k ill-)
W.Ofl i 1 l»
•.yiilhi-l it Int r.ipleural
n.iliit.il Inlrapleural
li um -,i 1 ii ,i|p
ii.tltu.il Inlrapleural
(Will I.V.I III! Ill-)
Intraperl loneal
•.ynllii-l ii. Inhalation
Inlraperl loneal
IMI ii.ii 1 iniilit<> Inlratracheal
Intrapleitral
i i if nt 1 1*
(niiM)nn) Inhalation
(K.i. mi) Intrapleiiral
Ilii .-1)1111) Inlrapleural

V,«e,/s,,,,,,
H.tl/f , Wislar
Ral/f. Wislar
Ral/f , Osbornp-
Mpnilel
Ral/f , Osborne-
Rat/F , Osbornp-
Mendel
Rat/r, Vlstar
Rat/M. Hlslar
Rat/M, Wislar
Rat. Fischer 344
Rat, Fischer 344

Rat, Fischer 344
Rat, M setter .144
Rat. fischer 144

„„, 	
II) «H|, 1"
40 m.|, |«
40 ni<|, 1*
40 imi. In
40 «H|
70 m<|, wrpkly.
bit
10 mM/m-1 rps-
pi rattle p.irti-
5 d/wk. 17 mo
75 *H|. In
o .; IHI, I?K
70 UNI. Ix

10 mq/m1. 7 I.AI.
5 d/wk. 17 mn.
ohservwl f:>n |im 77/Sfil developed
lunq tumors
%,imp ,is attnvf 7 IX developed
mesoOtellomas

 7H

                                                          rm.ili-s  itiwpln|ip.«|«-)

-------
 1.11  I ilii-r              Rntilr

i i. ii I ill*
                   1 nl r.i|.l rural
                                                                        I AH! I  /-M.   (mill I i«ii-il|
                                          type les /strain
Hal /I .  S|>rai)in<-
Oawlpy
                                                     I il.fi- M/I-/HI-M i
                                                  (il   ili.imi-lpi .  I
                                                                                          |>.it I ii Ii-  -, i/f:  'I'.T
                                                                                                    ts
                                     Ho timors
                                                               Cavclnm|*n
                                                                                                                                                                           ;m
M.ml in, i-|  at.  ( !•!//).
W...HI.-.  i-1 .il.  ( H/f.).
HIIIII I..111. pt  *\.  ( I'HI I)
•.mi Hi .-I .il   (  I'lllll).
'.mi Hi .mil MuliiMt  ( O/4).
I ITIIII i-1 .1).  (  1'in!!).
Mi.ht  c I. jtt.  ( I'Mil).
'.ml til i-t rfl.  (  I'NKI).
!••••  ••«..  .ll   
 Ib.   Pull el  al. (I'lM)
 !»..   Collin el al. (I'm?,
 I/.   Sii/nki and Hohy/WM (I'HIH).
 IR.   Ol«"i«i Pt al. (I'm1!)
 in.   Maqnpr Pi al. (I'HIS).
 70.   Smith  el al. ( I'ltlb.  I'm/).
71.
i? .
71
71
7'..
71..
71 .
711.
?'».
Miilitf p|. .il. (!'«».. I'm/)
r»ll P| al. (I'lll/li)
Hili tn-11 rt .il. ( l'lll(i).
M.Hiifi ( rum).
H..II-IIII ii
llfl'.l-.S (
Id. Hun i-
H.i 1 1 MI, i
l.it lif-.l p i
ill Miiii-n.i ( tW>).
•IH7).
al. (I'MM.).
•1. al. ( I'm7li).
> «-l al. ( I'MII).
til
     W...|.,,..  rl .il.  ( I'lM)
                                                                                                                                        .11.   Davis
                                                                                                                                        37..   Poll «nrt (rip.lrick-i (19/7).
                                                                                                                                        .11.   Poll el  a I  (11/fc).
                                                                                                                                        3«.   Waqner el al.  (!•»//).
                                                                                                                                        3S.   Stenback el al.  (IIRft)
                                                                                                                                        36.   Wrhnpr el al.  (I1//).
                                                                                                                                        37.   Jaiiranil  et si   ( 1'HU)
                                                                                                                                        38.   Hunter ami Ilionson (19/3).
                                                                                                                                        39.   Putt el  al. (llC/a).
                                                                                                                                        40.   Davis (19fl/).

-------
7.3.5  Summary
     The  results  of  several  chronic inhalation studies indicate that glass
fibers  of varying  lengths  and  diameters  do not induce lung tumors  in  five
animal  species  tested  (Table 7-13).   Intratracheal  instillation studies  in
which animals were  observed for 18 or more  months  provide  conflicting evidence
on  the  carcinogenic  potential  of glass microfibers (diameter <0.45, length
<12 M"0.   Pott  et al.  (1984b;  Mohr et al., 1984;  Pott et al., 1987a) found
increased  incidences  of lung carcinoma and mesothelioma in Syrian hamsters as
well as  increased incidences of lung tumors (carcinoma but not mesotheliomO
in  female  Wistar  rats following intratracheal  instillation of  microfibers.   In
contrast,  Smith  et al. (1986,  1987)  found  no  primary  lung tumors in ei'ther
male Syrian  hamsters  or female Osborne-Mendel  rats.   The fiber dimensions and
total dose used  with hamsters  were approximately the  same in both  studies
(Pott et  al., 1984b;  Smith et al., 1987).  The  fibers  used with Wistar rats
(Pott et  al., 1984b)  were significantly shorter and  thinner (diameter 0.18 um,
ler.gth 3.2 urn)  than those used  with the Osborne-Mendel  rats (diameter 0.45 urn
length 7.5 urn)  (Smith et  al.,  1967), but the total dose was the same.   Unlike
inhalation or intratracheal  instillation  studies,  intrapleural  or  intraperitc
neal administration of glass fibers has been associated with increased lung and
peritoneal tumors in  various strains of rats.   All  intraperitoneal  studies we^e
conducted  with microfibers  (diameter 0.18 to 1  um,  length 7.5 to  2.8 urn) at
total doses ranging from 0.5 mg to 25 mg, whereas intrapleural  studies utilized
glass fibers  of  various lengths and diameters  at  total doses  of 20  to 40  mg.
No tumors were noted  in mice dosed intrapleurally (Davis,  1976).
     Smith et al.  (1980)  reported tumors  in hamsters dosed intrapleurally  with
fibers having diameters  of 0.1 to 1.23 urn and lengths  >20 urn.   No tumors were
noted with shorter  (length <20 um) or with thicker fibers (diameter ^5 urn).
     Carcinogenicity  data  for  mineral  wool administered  by   inhalation or
intrapleural  injection  were similar to those obtained  for fibrous glass with
the  same  dimensions  (see Table 7-13).  Inhalation studies with  hamsters and
two strains  of  rats  indicate no evidence of Carcinogenicity for slag wool  or
rock wool  (diameter <3 um).  Intrapleural administration of rock wool (diameter
<3 urn) or  slag  wool (diameter <0.18 um) to  rats resulted  in a  low  incidence cf
pulmonary  tumors, whereas  thicker fibers  (diameter <5  um)  did  not.   Intrase-i-
toneal studies  conducted  by Pott et al.  (1984a, 1987a)  in rats  indicate that
slag wool  (diameter  1.5  to 2.6 um) is not carcinogenic,  whereas  ro:^  wool
(aiar->eter 1.94 um)  increased the incidences of pulmonary tumors.
JL.-I,- 19::                           7't'-             DRAFT —D:  N"~ 0*0':  OF. CI'E

-------
     The results  of  animal  carcinogenicity  studies with ceramic fibers suggest
similar carcinogenic  potential  to that obtained for fibrous glass  and mineral
wool.  Inhalation  of  refractory ceramic  fibers  and  inhalation or intratracheal
administration of  aluminum  silicate fibers  in rats and hamsters indicate  nc
evidence of  carcinogenicity.    Intraperitoneal  injection  of various  cerarr.ic
glass  fibers  results  in  high incidences of mesothelial tumors in  rats  and
hamsters.  Considerably  lower incidences of tumors were observed after intra-
pleural administration.
     Carcinogenicity  studies  with synthetic fibers are limited (Table 7-13).
In an  inhalation  study with  Kevlar® fibers in  rats,  increased  incidences  of
squamous cell  carcinomas  were noted at exposure levels of 0.63  and 2.23 mg'rr.
but  not  0.32  mg/m .   Increased incidences of mesothelial tumors and  sa~co~.cs
were noted  in rats  following  intraperitoneal  injection  of Kevlar"  fibers
(diameter 0.75 to 1.0 urn).   The  increase, however, was  relatively  low (6  to
13 percent).    Similarly,  2  of 51  rats  developed tumors following intre-
peritoneal  injection  of  polypropylene  fibers.   No tumors  were observed ir  re's
following  intratracheal   administration  of  polypropylene (unsized) fibers.
Results from  a dermal carcinogenicity study with  four  carbon  fibers  in mice
were negative.  Carcinogenicity studies  by  other routes  of exposure  were  not
found.   The carcinogenic  potential  of attapulgite was  investigated following
intraperitoneal and  intrapleural  administration (see  Table  7-13).   Two samples
(of  unspecified composition)  were found to be  carcinogenic in  rats,  whereas
three  other   samples  from  other  regions were  not.   Attapulgite  was net
carcinogenic  following intrapleural administation to rats.
     In contrast  to  the  available data  for manmade  and synthetic   fibers and
minerals discussed in this  chapter,  erionite  has  been shown to be  highly
carcinogenic  in  various  strains  of rats and  mice following inhalation or
intratracheal, intrapleural,  and  intraperitoneal  administration at relatively
low  doses  (Table  7-13).   All  the fibers tested had diameters ranging from 0.2
to 1.0 um and lengths ranging from 1.8 to 15.0 um.
     Talc powder  is  not carcinogenic following inhalation  exposure or intra-
tracheal and  intrapleural administration to rats, hamsters,  and mice.   Evidence
from one study  suggests  that talc may be an effective cocarcinogen wher. admin-
istered with  benzo(a)pyrene.  Carcinogenicity data for vermiculite  were lirr.-itei
to one  intrapleural   study  (Table 7-13).  No tumors  were observed  following
administration of 25  mg;  animals  were observed for 2 years.

Julv 1SEE                           7-65              DnA-T--D: tCT  CJ:" C-r CI":

-------
7.4  REFERENCES
             N

Abraham, J.  L.;  McEuen, D. D.  (1986)  Inorganic  participates  associated with
     pulmonary alveolar proteinosis: SEM and x-ray microanalysis results. Appl.
     Pathol. 4: 138-146.

Archer,  V.  E.  (1979)  Carcinogenicity of  fibers  and films: a  theory.  Meci.
     Hypotheses 5: 1257-1262.

Beck,  E.  G.; Holt,  P.  F.;  Manojlovic,  N.  (1972) Comparison of effects  on
     macrophage cultures of glass fibre, glass powder, and chrysotile asbestos.
     Br. J. Ind.  Med. 29:  280-286.

Bellmann,  B. ; Muhle,  H. ;   Pott, F. ;  Konig,  H.;  Kloppel,  H.;  Spurny,  K.  (19S7)
     Persistence of  man-made mineral fibres (MMMF) and asbestos in rat lungs.
     Ann. Occup.  Hyg. 31:  693-709.

Bignon,  J. ;  Sebastien, P.;  Gaudichet,  A.; Jaurand, M. C.  (1960)  Biological
     effects of  attapulgite.  In: Wagner,  J.  C.; Davis, W., eds.  Biological
     effects of  mineral  fibres  = Effects biologiques des  fibres  minerales:
     proceedings  of  a  symposium, v.  1;  September 1979; Lyon,  France.  Lyon.
     France: World Health Organization,  International Agency for  Research on
     Cancer; pp.  163-161.  (IARC scientific publication no.  30;  INSERK sycpcsis
     series: v.  92).

Eclton,  R.  E. ;  Ao'dison,  J. ;  Davis,  J.  M.  G. ; Donaldson,  K. ;  Jones,  A. D. ;
     Killer, B.  G. ;  Wright,  A.  (1965) Effects of  inhalation  of dusts fro,?
     calcium silicate  insulation  materials in laboratory  rats.  Environ.  Res.
     3S: 26-43.

Brown,  R.  C. ;  Chamberlain, M.; Davies,  R. ; Gaffen,  J.; Skidmore, J.  W.  (1979a)
     In  vitro biological  effects  of glass  fibers. J.  Environ.  Pathol.  Toxicol.
     2: 1369-1383.

Brown,  R.  C. ;  Chamberlain, M.; Skidmore,  J.  W.   (1979b) In vitro  effects  of
     man-made mineral fibres. Ann. Occup. Hyg. 22: 175-179.

Chamberlain, M.;  Brown, R.  C.  (1978)  The  cytotoxic effects of asbestos  and
     other  mineral  dust  in  tissue  culture cell  lines.  Br.  J.  Exp.  Pathol.
     59: 183-189.

Chamberlain, M.;  Tarmy,  E. M. (1977) Asbestos  and  glass fibres in bacterial
     mutation tests.  Mutat. Res. 43: 159-164.

Chamberlain, M. ;  Davies,  R.; Brown, R.  C.;  Griffiths,  D.  M.  (1982) In vitrc
     tests  for the  pathogenicity of mineral dusts.  In:  Walton,  W., ed.  Inhaled
     particles V:  proceedings  of  an international  symposium;  Auoust  I960;
     Cardiff, Wales, United Kingdom. Ann. Occup.  Hyg. 26: 563-592."

Chang,  K.  J. W. ;  Joseph,  L.  B.; Stephens,  R.  E.; Hart, R.  w.  (19S3) Adsorption
     of macromolecules on  mineral fibers. J. Am.  Coll. Toxicol.  2: 167-152.

Coffin,  D.  L. ;  Palekar,  L.  D. ; Cook,  P.  K.  (1982)  Tumorigenesis  t\  £
     ferroactinolite mineral. Toxicol. Lett. 13:  143-15C.

-------
Coffin, D.  L.;  Palekar,  L. D.; Cook, P. M. (19B3) Correlation of in vitro and
     uj vivo  methods  by means of mass dose and fiber distribution for amcsite
     and fibrous ferroactinolite. EHP Environ. Health Perspect.  51:  49-53.

Cook,  P.  M.; Palekar,  L.  D.;  Coffin,  D.  L.  (1982)  Interpretation  of the
     carcinogenicity  of  amosite asbestos and ferroactinolite on the  basis  of
     retained  fiber  dose  and  characteristics  in vivo.  Toxicol.   Lett.
     13: 151-158.

Davies, R.  (1980)  The effect of mineral fibres on macrophages.  In:  Wagner, J.
     C., ed.  Biological  effects of mineral fibers:  proceedings  of a  symposiu";
     v. 1;  September  1979. Lyon, France:  World Health  Organization,  Regions'
     Office for Europe;  IARC  Sci. Publ. 30: 419-425.

Davis,  J.  M.  G.  (1976)  Pathological aspects  of  the injection of glass fibe<-
     into  the pleura!  and peritoneal cavities of  rats  and mice.  In: Occupa-
     tional exposure  to  fibrous glass: proceedings of a symposium;  June 197-;
     College  Park,  MD.  Cincinnati,  OH: U.  S. Department of Health,  EcJucaticr.,
     and  Welfare,   National  Institute  for Occupational  Safety and  Healtn;
     pp. 141-149;  HEW publication no.   (N10SH) 76-151.  Available  from: NTIS,
     Springfield, VA:  PB-258869.

Dans,  J.  M.  G.  (1987)  Carcinogenicity of Kevlar aramid pulp following ir.tre-
     peritoneal   injection  into  rat  {unpublished study].  Edinburgh,  Sect's":
     Institute of  Occupational  Medicine; Report  no.  TM/87/12 UDC  676.6:ol;.27;
     Available from: NTIS, Springfield, VA; PBS8-109004.

Davis,  J.  M.  G. ;  Gross, P.;  de Treville,  R. T. D. (197C) "Ferruginous bore-"
     in guinea pigs. Arch. Pathol. 89:  364-373.

Davis,  J.  M.  G. ; Addison,  J.;  Bolton, R.  E.; Donaldson,  K.;  Jones, A.  D. ;
     Wright,  A.  (1984)  The  pathogenic effects of  fibrous ceramic  alurr.inur
     silicate glass administered to  rats by inhalation or peritoneal injection.
     In: Guthe,  T., ed.  Biological effects of man-made mineral fibres: proceed-
     ings  of  a WHO/1ARC conference  in  association  with JEMRB and TIMA,  v.  2;
     April  1982; Copenhagen,  Denmark.  Copenhagen,  Denmark:  World  Health  Organ-
     ization, Regional Office for Europe;  pp. 303-322.

Denizeau,  F.;  Marion, M.;  Chevalier,  G.;  Cote,  M.  G.  (1985a) Absence cf
     genotoxic effects  of  nonasbestos  mineral fibers.  Cell  Biol.  Toxicol.
     1: 23-32.

Denizeau,  F.; Marion,  M.;  Chevalier, G.; Cote, M.  G.  (1985b) Uptake of mineral
     fibers by hepatocytes in vitro and investigation of concomitant genctc*ic
     effects. Environ. Mutagen.  7(suppl. 3): 26.

DePass, L.  R.  (1982)  Evaluation of  the  dermal  carcinogenicity of fou'-'carbon
     fiber  materials  in male C3H/HeJ  mice.  Final  report by the Bushy  Rur.
     Research Center for the  Union Carbide Corporation;  Project no.  45-12.

Dunnigan, J. ; Nadeau,  D.;  Paradis, D.  (1964) Cytotoxic effects of ara-ic  fi&ers
     on  rat pulmonary  macrophaoes: comparison  with  chrysctile  asbes*.:s.
     Toxicol.  Lett. 20:  277-282."

-------
  Feron,  V.  J.; Scherrenberg,  P.  M. ;  Immel,  H.  R.;  Spit, B. J.  (1985)  Pulmonary
       response  of  hamsters to  fibrous  glass:  chronic  effects of  repeated
       intratracheal  instillation with or without benzo[a]pyrene. Carcinogenesis
       6:  1495-1499.

  Forster, H.  (1984) The  behaviour  of mineral fibers  in  physiological  scluticr,.
       In: Biological  effects  of man-made mineral  fibers:  proceedings  of  a
       WHO/IARC  conference; v.  2; April  1982. Copenhagen, Denmark: World Health
       Organization,  Regional Office for  Europe; pp. 18-26.

  Gershuni,  D.  H.;  Amiel,  D.; Gonsalves,  M.;  Akeson, W. H. (1981) The biochemical
       response  of  rabbit  articular  cartilage matrix to an induced  talcum synovi-
       tis.  Acta Orthop. Scand.  52:  599-603.

  Goldstein,  B.  ; Rendall,  R.  E.  G.;  Webster,  I.  (1983) A  comparison  of the
       effects of exposure  of  baboons to crocidolite and fibrous-glass  dusts.
       Environ.  Res.  32: 344-359.

  Gorir-ley, I.  P.; Wright,  A.;  Ceilings,  P.;  Davis,  J. M. G.  (1980) The cytcto-
       icity  of  U1CC and  modified asbestos  fibres in vitro. In: Wagner,  J. C. ,
       ed.  Biological  effects  of mineral  fibers:  proceedings  of  a WHO/I ARC
       conference;  v.  1; September 1979.  Lyon, France: World Health Organize*.icn.
       Regional  Office  for Europe; pp.  427-434.

  Hcrilton,  T.  C. ;  Fox,  H. ; Buckley,  C.  H. ; Henderson,  W.  J. ;  Griffiths,  H.
       (1954)  Effects of talc on  the rat  ovary.  Br.  J. Exp.  Pathol.  65:  iCl-i:=.
 Hsnsen,  K. ;  Mossman, B. T. (1987) Generation of superoxide (02-) frorr. eU-ec'e"
      macrophages  exposed to asbestiform  and  nonfibrous  particles.  Cancer Res.
      47:  1661-1686.

 Harington,  J.  S.  (1981) Fiber  carcinogenesis:  epidemiologic  observations and
      the  Stanton  hypothesis. JNCI  J.  Natl.  Cancer  Inst.  67: 977-966.

 Haugen,  A.; Schafer,  P.  W. ;  Lechner,  J. F.; Stoner, G.  D.;  Trump,  B.   F. ;
      Harris,  C.  C.  (1982) Cellular  ingestion,  toxic effects,  and lesions
      observed  in human  bronchial  epithelial tissue  and cells cultured  witr,
      asbestos  and glass  fibers.  Int.  J.  Cancer  30:  265-272.

* Hesterberg,  T.  W.;  Barrett, J.  C.  (1984) Dependence of asbestos- and miners!
      dust-induced transformation of mammalian cells  in culture on fiber  dimen-
      sion.  Cancer Res.  44:  2170-2180.

 Hinson,  T.  R.;  Schol,  H. M.; Pipkin,  J.  L.  (1983)  A preliminary ultrastructursl
      study  of  macrophage  response to  carbon fiber  particles. EHP  Environ.
      Health  Perspect.  51:  373.

 Hunter,  B. ; Thomson,  C.  (1973)  Evaluation of  the tumorigenic potential c*
      vermiculite  by intrapleural  injection  in  rats.   Br.  J.  Ind.  Me:.
      30:  167-173.

 Jaurand,  M.  C.  ; Bignon, J.  (1979) Interaction  of  fibres with human en:  6~irr,=
      cells.  Thorax  34:  694.

-------
Jaurand, M.-C.;  Bignon,  J.;  Magne, 1.; Renier, A.; Lafuma, J.  (1979) Interac-
     tion des  fibres  avec les globules rouges et  les  macrophages  alveolaires
     in vitro. Rev. Fr. Mai.  Resp. 7: 717-722.

Jaurand, M.-C.;  Fleury,  J.;  Monchaux, G.;  Nebut, M.; Bignon, J.  (1987)  Pleura!
     carcinogenic  potency  of mineral  fibers (asbestos, attapulgite) and their
     cytotoxicity  on cultured cells. JNCI J. Natl.  Cancer Inst.  79: 797-804.

Johnson, N.  F.;  Edwards, R.  E.; Munday, D. E.; Rowe, N.; Wagner, J. C.  (1954)
     Pluripotential nature of  mesotheliomata  induced by inhalation of erionite
     in rats. Br.  J. Exp. Pathol. 65: 377-388.

Klingholz, R.; Steinkopf,  6. (1984) The reactions of  MMMF in a  physiological
     model fluid and  in  water.  In:  Biological effects of man-made mineral
     fibers:   proceedings  of a  WHO/IARC  conference;  v.   2;  April  1982.
     Copenhagen,   Denmark:  World  Health Organization,  Regional  Office  for
     Europe;  pp.  60-86.

Kuschner, M.  ;  Wright,  G.  (1976) The  effects  of  intratracheal  instillation cf
     glass fiber  of  varying  sizes in guinea pigs.  In:  Occupational exposure  to
     fibrous  glass: proceedings  of a symposium; June  1974;  College Park,  fC.
     Cincinnati,   OH:  U.  S.  Department of  Health,  Education,  and Welfare.
     National  Institute  for  Occupational  Safety and Health; pp.  151-165;  HE*
     publication  no.  (N10SH) 76-151. Available  from:  NT1S,  Springfielc.  \A:
     PE-256669.

Lee, K.  P. (1973) Lung response to particulates with emphasis on  asbestos a~c
     other fibrous dusts. CRC Crit. Rev. Toxicol. 14: 33-66.

Lee, K.  P.;  Kelly, D.  P.; Kennedy,  G.  L.,  Jr.  (1973)  Pulmonary  response  to
     inhaled Kevlar aramid  synthetic fibers in rats. Toxicol. Appl. Pharrr,a:ol.
     71: 242-253.

Lee, K.  P.;  Barras,  C. E.;  Griffith,  F.  D.;  Waritz, R.  S.  (1979) Pulmona-y
     response to glass fiber by inhalation  exposure. Lab.  Invest.  40: 123-133.

Lee, K.  P.;  Barras,  C. E.;  Griffith, F.  p.; Waritz,  R.  S.; Lapin,  C.  A.  (1961)
     Comparative  pulmonary responses  to inhaled inorganic fibers with asbestos
     and fiberglass. Environ. Res. 24: 167-191.

Lord, G. H.  (1978) The biological effects of talc in the  experimental animal: a
     literature review. Food Cosmet. Toxicol.  16: 51-57.

Maltoni,  C.;  Minardi,  F.;  Morisi,   L.  (1982a) Pleura!  mesotheliomas  in
     Sprague-Dawley  rats by erionite:  first  experimental evidence. Environ.
     Res. 29: 238-244.

Maltoni, C.;  Minardi,  F.;  Morisi, L.  (1982b)  The relevance of the experimental
     approach  in the  assessment  of  the  oncogenic  risks  from  fibrous  and
     non-fibrous  particles.  The ongoing project of  the Bologna Institute of
     Oncology. Med. Lav. 4:  394-407.

Maroudas, N.   G. ;  O'Neill,  C. H.;  Stanton,  M.  F.  (1973) Fibroblest anchcrace  in
     carcinocenesis bv fibres. Lancet 1: 807-8D9.

-------
Mitchell,  R.  I.; Donafrio, D.  J. ;  Moorman,  W.  J. (1986) Chronic  inhalation
     toxicity of fibrous glass  in rats and monkeys. J.  Am.  Coll.  Toxicol.
     5: 545-575.

Mohr, U. ; Pott,  F. ; Vonnahme, F.-J. (1984) Morphological aspects of mesothelio-
     mas after  intratracheal  instillations of fibrous dusts  in  Syrian  gcltien
     hamsters. Exp. Pathol. 26: 179-183.

Monchaux, G. ; Bignon, J. ; Jaurand, M. C. ; Lafuma, J. ; Sebastien, P.; Masse, R. ;
     Hirsch,  A.;  Goni ,  J.  (1981) Mesotheliomas in rats  following  inoculation
     with acid-leached  chrysotile  asbestos and other mineral fibres.  Carcino-
     genesis  2:   229-236.

Moreno, 0.  M. ;  Moreno,  M.  T. (1980) Twenty-one month observation after single
     intratracheal  insufflation [unpublished  study].  Spinnertown, PA:  M.E.
     Research Laboratories, Inc.; Project no. 78-2806.

Mossman, B. ;  Light,  W. ; Wei, E. (1983)  Asbestos:  mechanisms of toxicity a-c
     carcinogenicity in  the  respiratory tract.  Annu. Rev. Pharmaccl.  Tcxicc".
     23: 595-615.

Mossman, B. T. ;  Cameron, G.  S. ; Yotti,  L.  P.  (1985) Cocarcinogenic and .tumor
     promoting properties  of  asbestos  and other minerals in tracheobronchial
     epithelium. Carcinogenesis 8:  217-238.
MuhTe, H. ;  Pott, F. ; Bellmann, B. ; Takenaka, 'S. ; Ziem, U. (1986) Inhalaticr.
     injection experiments  in rats for testing man-made mineral fibers in the
     working  environment.  Presented at:  international  symposium on ir.ar-r.ace
     mineral  fibres  in  the working environment; October; Copenhagen, Den-=rK.
     Copenhagen,  Denmark:  World  Health  Organization,  Regional Office  for
     Europe.

Muhle, H. ;  Pott, F.; Bellmann, B. ; Takenaka, S. ; Ziem, U. (1987) Inhalation and
     injection experiments  in rats to test  the  carcinogenicity of MMMF.  Ann.
     Occup. Hyg. 31: 755-764.

Nadeau, D.  ; Paradis, D.  ; Gaudreau, A.; Pele, J. P.; Calvert, R. (1983) Biologi-
     cal evaluation of  various natural and man-made mineral fibers: cytoto*ici-
     ty, hemolytic  activity and chemi luminescence study. EHP  Environ. . Health
     Perspect. 51: 374.

Nadeau,  D. ;  Fouquette-Couture,  L. ;  Paradis, D. ;  Khorami ,  J.  H. ;  Lane, D. ;
     Dunnigan,  J.  (1987)  Cytotoxicity  of respirable  dusts from  industrial
     minerals:  comparison  of two  naturally  occurring  and  two  man-made sili-
     cates. Drug Chem.  Toxicol. 10: 49-86.

Nakatani,  Y.  (1983) Biological  effects  of mineral fibers  on  lymphocytes  j_n
     vitro: comparative studies  of  asbestos and  glass  fibers.  Sangyo  loa**
     25: 375-386.

Czesmi,  M. ;  Patiroglu,   T.  E. ; Hillerdal, G. ;  Ozesmi ,  C.   (19£5)  Peritoneal
     mesotheliorr.a and malignant  lymphoma  in  mice caused  by  fibrous zeolite. Br.
     J. Ind. Med. 42: 746-749.
                                    7-70             DRAFT--D:  N:~  Q::" c: ::TE

-------
Palekar, L.  D.; Coffin, D.  L.  (1986) Significance of mineral  fibers  in  mesothe-
     lioma induction and V79 cytotoxicity.  lexicologist 6:  6.

Palekar, L.  D.;  Eyre,  J.  F. ;  Most, B. M.;  Coffin,  D.  L. (1987) Metaphase and
     anaphase analysis  of  V79  cells exposed to erionite,  UJCC chrysotile and
     UICC crocidolite.  Carcinogenesis (London) 8:  553-560.

Pickrell, J. A.;  Hill, J.  0.; Carpenter,  R.  L.;  Hahn, F.  F.; Rebar,  A.  H.
     (1983) In vitro and |n vivo response  after  exposure  to man-made  mineral
     and asbestos insulation fibers. Am.  Ind.  Hyg. Assoc.  J.  44:  557-561.

Pigott, G. H.; Gaskell, B. A.; Ishmael,  J.  (1981) Effects  of long term inhala-
     tion of alumina fibres in rats. Br.  J. Exp.  Pathol.  62:  323-331.

Poole,  A.;  Brown, R.   C.;  Turver,  C.  J.; Skidmore, J. W.; Griffiths,  D.  K.
     (1962)  In vitro genotoxic  activities  of fibrous  erionite. Br. J.  Cance-
     47: 697:7or

Pott,  F.;  Friedrichs,  K.   H.  (1972) Tumoren  der Ratte nach i.p.-lnjekiion
     faserfoermiger Staeube [Tumors in rats after  i.p.  injection  of  fibrous
     glass].  Naturwissenschaften 59: 318.

Pott.  F.;  Huth,  F.; Friedrichs,  K.  H.  (1974) Tumorigenic effect of fibrous
     dusts in experimental  animals. EHP Environ.  Health Perspect.  9: 313-215.

Pott,  F.; Huth,  F.;  Friedrichs,  K. H.  (1976) Results of animal carcinogenesis
     studies after  application of fibrous  glass  and their  implications regard-
     ing human exposure. In:  Occupational  exposure to fibrous glass:  proceed-
     ings of a symposium;  June 1974; College Park,  MD.  Cincinnati, OH:  U. S.
     Department  of  Health, Education, and Welfare,  National  Institute for
     Occupational Safety and  Health;  pp. 183-191; HEW publication no.  (NIOSH)
     76-151.  Available from:  NT1S, Springfield, VA; PB-258869.

Pott,  F. ;  Schlipkoeter, H. W.;  Ziem,  U.;  Spurny,  K.; Huth, F.  (19B4a) Ne«
     results from implantation studies  with  mineral fibers.  In: Guthe,  T. ,
     ed. Biological  effects of  man-made mineral  fibres:  proceedings  of  a
     WHO/IARC conference in association  with  JEMRB and TIMA,  v.  2;  April  19E2;
     Copenhagen,   Denmark.  Copenhagen,  Denmark:   World  Health  Organization,
     Regional Office for Europe; pp. 286-302.

Pott,  F.; Ziem,  U.; Mohr,  U.  (1984b)  Lung  carcinomas and mesotheliomas follow-
     ing intratracheal  instillation of glass  fibers and asbestos.  In:  Procees-
     ings of  the Vlth international  pneumoconiosis conference, v. 2;  19£2;
     Bochum, Federal  Republic of  Germany.  Bremerhaven, Federal Republic  of
     Germany: Verlag fuer neue Wissenschaft; pp.  746-756.

Pott,  F.; Ziem,  U.; Reiffer,  F.-J.; Huth,  F.; Mohr, U. (1987a) Carcinogenicity
     studies on  fibres, metal  compounds, and some  other  dusts in  rats. Ext.
     Pathol.  32:  129-152.

Pott,  F. ; Roller,  M.;  Ziem,  U. ; Reiffer, F. J.;  Bellmann, B.; Rosenbruch, K.;
     Huth, F.  (19£7b)  Carcinogenicity studies on natural  and  man-made fise-s
     with the  intraperitoneal test  in  rats. Proceedings  of a Sympcsiur: c-
     mineral  fibers  in the  non-occupational  environment;  Lyor,  France:
     September e-10.

-------
Potts, W.  J.;  Lederer,  T.  S.; Gehring,  P.  J.  (1978) Hemolysis of washed rat
     erythrocytes  in vitro  by  dust  particles. Am.  Ind.  Hyg.  Assoc.  J.
     39:  497-502.

Richards,  R. J. ;  Hunt,  J.  (1983) Reaction  of  mineral  dusts  with primary lung
     fibroblast cultures. EHP Environ. Health Perspect. 51:  61-65.

Scholze,  H.; Conradt, R. (1987) An in vitro study of the chemical durability of
     siliceous fibres. Ann. Occup. Hyg. 31: 683-692.

Sebastien,  P.;  Bonnaud,  G.;  Bigfnon,  J. (1979) Assessment of  health  risks  of
     exposure to  mineral  fibres:  contribution  of measurements  in environmental
     and biological samples.  Thorax 34: 694.

Sincock,   A.  M. ;  Delhanty, J.  D. A.;  Casey, G.  (1982) A comparison  of  the
     cytogenetic  response  to asbestos and glass fibre  in Chinese hamster  end
     human  cell  lines.  Demonstration  of growth inhibition  in primary  hums'
     fibroblasts. Mutat. Res. 101: 257-268.

Smith, W.  E. ;  Hubert,  D.  D. (1974)  The  intrapleural  route  as  a  means  for
     estimating carcinogenicity.  In:  Karbe, E. ; Park,  J.  E. , eds.  Experimental
     Lung Cancer: carcinogenesis and bioassays. New York: Springer; pp.  93-1C1.

Srr.nh, W.  E. ;  Hubert,  D.  D. ; Sobel,  H.  J.  (1980) Dimensions  of  fibres  in
     relation to  biological  activity.  In:  Wagner,  J.  C. ;   Davis, W. ,  eds.
     Biological   effects  of  mineral  fibres =  Effets  biologiques des  fibres
     minerales:  proceedings of a symposium, v.  1;  September 1979;  Lyon,  France.
     Lyon, France: World Health Organization, International  Agency for Kese = -c.~-
     on  Cancer;  pp.   357-360.  (1ARC  scientific publication  no.  30;  INSERM
     symposia series: v. 92).

Smith, D.  M. ; Ortiz,  L.  W. ;  Archuleta, R. (1984) Long-term exposure of Syrian
     hamsters and Osborne-Mendel rats  to aerosolized 0.45  urn mean diameter
     fibrous glass.  In:  Guthe,  T.,  ed. Biological  effects of man-made mineral
     fibres: proceedings of a WHO/IARC conference in association with JEMR=  anc
     TIMA,  v. 2;  April  1982; Copenhagen, Denmark.   Copenhagen,  Denmark:  World
     Health Organization, Regional Office for Europe; pp. 253-272.

Smith, D.  M. ;  Ortiz, L. W.;  Archuleta, R. F.; Johnson, N.  F.  (1986) Long-
     term  health  effects  in  hamsters and rats exposed chronically to man-made
     vitreous fibers. Presented at:  international  symposium on man-made mineral
     fibres  in  the   working environment;  October; Copenhagen,  Denmark.
     Copenhagen, Denmark: World Health Organization, Regional Office for Europe.

Smith, D.  M. ; Ortiz,  L.  W.;  Archuleta, R. F.;  Johnson, N.  F. (1987) Long-term
     health effects  in hamsters  and  rats  exposed  chronically  to man-made
     vitreous fibers. Ann.  Occup. Hyg. 31: 731-754.

Spumy, K. R. ;  Opiela, H.; Schoerman, J.; Weiss, G. (1983) Methods and investi-
     gations relative to  the stability of mineral  fibers in  vitro and in vivc.
     Staub Reinhalt.  Luft 44: 169-180.

Stadler,  J. C.  (1986) Two-year inhalation  study in rats with  Kevlar fibrils
     [unpublished study].  Michigan:  E.I.  du Pont de Nemours  and Corro=";>.  In:.;
     Report no.  290-86"  MR no.  4442-001.

-------
Stanton, M.  F.  (1973)  Some  etiological  considerations of fibre carcinogenesis.
     In: Bogovski, P.;  Gilson, J.  C.; Timbrel!, V.;  Wagner,  J.  C.,  eds.  Biolog-
     ical  effects  of  asbestos:  proceedings of a working conference;  October
     1972; Lyon, France. Lyon, France:  World Health  Organization,  International
     Agency  for  Research  on Cancer;  pp. 289-294.  (IARC  scientific  publication
     no. 8).

Stanton, M.  F.;  Wrench, C.  (1972) Mechanisms  of mesothelioma  induction with
     asbestos and fibrous glass.  J. Natl. Cancer Inst.  U.  S.  48:  797-816.

Stanton, M.  F.;  Layard, M.; Tegeris, A.; Miller,  E.;  May,  M.;  Kent, E.  (1977)
     Carcinogenicity of fibrous glass:  pleura! response in the rat  in relation
     to fiber dimension. JNCI J.  Natl.  Cancer Inst.  58:  587-603.

Stanton, M.  F.; Layard,  M. ;  Tegeris,  A.; Miller,   E.; May,  M. ;  Morgan, E. ;
     Smith,  A.  (1981)   Relation of particle dimension to carcinogenicity  in
     amphibole  asbestoses  and other fibrous  minerals.  JNCI J. Natl. Cance"
     Inst.  57: 965-975.

Stenback,  F.;  Wasenius, V.-M.;  Rowland,  J. (1986)  Alveolar  and  interstitial
     changes in silicate-associated lung tumors in Syrian hamsters.  Cancer Res.
     Mono.  2: 199-213.

Styles, J.  A.; Wilson,  J.  (1973) Comparison  between j_n vitrc toxicity  c*
     polymer  and  mineral  dusts and  their  fibrogeriitity.  Ann.  Occu?.  hvc.
     16: 241-250.

Styles, J.  A.;  Wilson, J. (1976)   Comparison between  ir\  vitro  toxicity  e*  U:
     novel  fibrous  mineral  dusts  and  their tissue reactions  i_n  viy_c.  Ar.r.
     Occup. Hyg. 19: 63-66.

Suzuki, Y.  (1982)  Carcinogenic  and fibrogenic  effects  of zeolites:  prelimins^
     observations. Environ.  Res.  27:  433-445.

Suzuki, Y. ;  Kohyama, N.  (1984)  Malignant mesothelioma induced by asbestos anc
     zeolite in the mouse peritoneal cavity. Environ.  Res.  35:  277-292.

Tilkes, F.;  Beck,  E.  G. (1980) Comparison of  length-dependent cytotoxicity of
     inhalable  asbestos and man-made mineral  fibres.  In: Wagner,  J.  C.; Daos,
     W. , eds.  Biological  effects  of mineral fibres s  Effets biologiques des
     fibres  minerales:  proceedings of  a symposium,  v.  1;  September 1979;  Lyon,
     France.  Lyon,  France:  World  Health Organization,  International Agency fcr
     Research  on  Cancer;  pp. 475-483.  (IARC  scientific publication  no.  30;
     INSERM symposia series: v.  92).

Tilkes, F.;  Beck,  E. G.  (1983a) Macrophage functions  after exposure to  mineral
     fibers. EHP Environ.  Health Perspect. 51: 67-72.

Tilkes, F.;  Beck,  E.  G. (1983b)  Influence  of  well-defined mineral fibers on
     proliferating cells.  EHP Environ.   Health  Perspect.  51:  275-279.

Vincent, R.  A.,  Jr.;  Hintz, D.  S.;  Harley,  R. A.   (1962) Association c'  DN-
     degradation in cultured human fibroblasts with the cytotoxicity  c' fir.rcj$
     substances. EH? Environ. Health Perspect. 51:  275.


J .-I-.- 195:                           7-72             DF.A"--DC NT  :~~: C:- --"'

-------
Wagner,  J.  C.; Berry,  G.;  Timbrel 1, V.  (1973)  Mesotheliomas  in rats after
     inoculation with asbestos and other materials. Br. J. Cancer 28: 173-185.

Wagner,  J.  C. ;  Berry,  G.; Skidmore, J.  W.  (1976) Studies of the carcinogenic
     effects  of fiber  glass  of different  diameters  following intrapleursl
     inoculation  in  experimental  animals.  In:  Occupational  exposure to fibrous
     glass: proceedings  of  a symposium; June  1974; College  Park, MD.  Cincin-
     nati,  OH:  U.  S. Department of  Health,  Education,  and Welfare, National
     Institute  for Occupational Safety and Health; pp. 193-197; HEW publication
     no. (NIOSH) 76-151. Available from: NTIS, Springfield, VA; PB-258869.

Wagner,  J.  C. ;  Berry,  G.; Cooke, T. J.; Hill, R. J.; Pooley, F. D. ; Skidmc'-e,
     J.  W.  (1977)  Animal experiments with talc.  In: Walton,  W.  H., ed.  Inhaled
     Particles  IV: v. 2. Oxford, England: Pergamon Press; pp. 647-654.

Wagner, J. C.;  Berry, G. B.; Hill, R. J.; Munday, D. E.;  Skidmore, J. W.  (198-)
     Animal experiments  with MMM(V)F--effects of  inhalation  and intraplejre'i
     inoculation  in  rats.  In:  Guthe, T., ed.  Biological  effects of ma^-mece
     mineral  fibres: proceedings  of a WHO/IARC  conference in association wif
     JEMRB  and  TIMA, v.  2; April  1982;  Copenhagen,  Denmark.  Copenhagen,  Den-
     mark: World Health  Organization, Regional Office for Europe; pp. 2D9-223.

Ws:?e-,  J.  C. ;  Skidmcre, J  W. ; Hill,  R. J.;  Griffiths,  D. M.  (19E5) Erioiite
     exposure and mesotheliorr.as in rats. Br. J.  Cancer 51: 727-730.

Warheit,  D.   B. ;  Hill,   L. H. ;  Brody,  A.  R.  (1984)  In   vitro  effects  of
     crocidolite asbestos  and  wcllastonite  on pulmonary  macrophages and  se-u^
     complement. Scanning Electron Microscopy  2:  919-926.

Wehner,  A.  P.;  Zwicker, G.  M.; Cannon,  W.  C.; Watson, C. R. ;  Carlton, W. u.
     (1S77) Inhalation  of  talc baby powder  by hamsters.  Food Cosmet. Toxicc".
     15: 121-129.

Woodworth,  C.  D.;  Mossman,  B.  T.; Craighead,  J.  E. (1982) Comparative effects
     of  fibrous and  nonfibrous minerals on  cells and liposomes.  Environ. Res.
     27: 190-205.

Woodworth, C.  D. ; Mossman, B. T.; Craighead, J.  E. (1983) Induction  of squamous
     metaplasia in organ cultures of hamster trachea by naturally occurring a->d
     synthetic  fibers.  Cancer Res. 43: 4906-4912.

Wright,  A.;  Gormley, I.  P.; Ceilings,  P.  L.;  Davis, J.  M.  G. (1980) The
     cytotoxicities  of  asbestos and other  fibrous dusts. In:  Brown,  R.  C. ; .
     Chamberlain, M.; Davies,  R.;  Gormley,  I.  P., eds.  The in vitro effects of
     mineral  dusts.  New  York: Academic  Press;  pp. 25-31.

Yano,  E.;  Takeuchi,  A.; Yukiyama, Y.;  Evans,  P. H.  (1986) Carcinogenicity cf
     mineral  dusts:  II.  in  vitro  cytotoxicity  of erionite  to  PMN.  Toxicol.
     Lett. 31:  207.

Zcltai,  T.  (1986) Review  of summary health assessment  document for miners"
     fibers other  than  asbestos [letter to  Dr.  Nicolas  Hajjar]. Minneapolis,
     MN: University  of Minnesota, Department of  Geology and Geophysics; Ksy 24

-------
                             8.  EFFECTS ON HUMANS
This chapter  presents  a critical  review and analysis of the carcinogenic and
noncarcinogenic effects  in  humans  following exposure to manmade mineral fibers
(mineral  wool,  glass wool,  continuous  glass  filament,  and ceramic fibers),
synthetic fibers (aramid, carbon,  and polyolefin fibers),  and minerals  (attapul-
gite, sepiolite, erionite, talc, and vermiculite).   An  analysis  of  the  data  for
each fiber from cross-sectional studies, clinical evaluations,  and  case reports
is presented first, followed by reviews of the epidemiological  studies.
8.1  MANMADE MINERAL FIBERS
8.1,1  Glass Fibers and Mineral Wool
8.1.1.1  Cross-Sectional Studies, Clinical Evaluations, and Case Reports
8.1.1.1.1   Eye and  skin irritation.   Skin  and  eye  irritation  have been  associ-
ated with  occupational exposure  to fibrous glass and  mineral  wool  fibers.
Mechanical  irritation  of the  skin and eye  is  common,  although immunologic
effects in  a small  number of exposed workers have also been observed.
     In an  early  summary  (Milby and Wolf, 1969), skin and eye irritation were
reported  in 653 of 691 cases  of occupational  disease attributed to  fibrous
glass exposure  in  California during a 30-month  period  between  I960 and 1962.
.The nature  of these effects was not described.
     The prevalence of work-related dermatitis in approximately 4000 workers
employed at fibrous glass manufacturing plants in the United States  was  reported
by  Fossick  et  al.  (1970).   Skin irritation, usually  found  on the arms, face,
and neck, occurred mostly in new workers.   The workers'  most  common complaints
included  burning,   itching,  or  prickling  sensations  of the skin.  Fibers  of
large diameters were  more likely to  cause  these irritations.   Other  workers
were found  to  have intermittent itching and dermatitis.   The primary lesions
that developed  in  workers with dermatitis included papules and  papulovesicles.
Secondary  lesions  were common  and resulted from bacterial infection of the

July 1988                           8-1           DRAFT—DO NOT QUOTE OR  CITE

-------
primary  lesions.   In  addition,  thickening  of the skin was observed in a number
of exposed workers.   The  cause  of the dermatitis was  attributed  to mechanical
Irritation and not to an allergic response.
     Heise and Mattheus (1971) reported complaints  of  Uching  skin and a number
of cutaneous eruptions in workers using insulation  containing  rock wool.   Patch
tests performed  on these individuals and  a group  of  control  subjects  using
various types of rock wools revealed a higher  frequency  of  $kin reaction in  the
exposed workers.   These effects  were attributed to  mechanical  irritation of  the
skin; no allergic reactions were noted.
     Arbosti et  al.  (1980) performed dermatologic  examinations  and allergy
tests on 135 workers  from an  Italian  glass fiber factory.  Irritative dermato-
sis was  found  in  workers  from  various  areas  of the plant.   Skin disorders
appearing in  19 percent  of the production workers had not  only a primary
irritative etiopathogenesis but  also  an  immune  pathogenesis.   It was concluded
that the  risk  of developing dermatologic  disorders with immune  pathogenesis
existed  mainly in those workers involved in  spinning  and spooling of the  glass
fibers,  while  the risk of irritative dermatosis  existed at  all production
steps.
     An immunologically based topical  eczema was also  noted in 3  of 427 workers
from a plant that produced spun glass fibers  (Maggioni  et al.,  1980).  Skin
disease,  mostly  primary irritation  dermatitis,  was observed  in 14 percent  of
the workers.  Total dust  concentration  in the  plant averaged  1 mg/m  , and the
percentage by weight  of glass fibers in the total  dust ranged from 1 to 7
percent.
     In a recent review of glass fiber dermatitis, Bjbrnberg (1985) reported
that 65 percent  of 700 workers  in the Swedish  building  industry  had.cutaneous
lesions from glass  or mineral wool.   Similar results were found among 62  glass
wool workers who were evaluated at the end of  the working day for skin prob-
lems.  Of these  workers,  27 had both  Uching and skin lesions, 10 had itching,
and  1  had lesions.   Skin  lesions  were  usually small, erythematous papules.
Other  lesions,   occurring  less   frequently, Included  urticaria,  petechiae,
erosions, and  nummular eczema.    Dermatitis was generally localized to the
forearms, hands,  face, neck, and  flexural folds.   Covered areas were also
affected, since  glass fibers  can penetrate clothing.  The  severity  and number
of the reactions varied directly with the size and  coarseness  of  the  fiber.
     Twenty-five percent  (79/135)  of  tested  subjects  exhibited positive  skin
reactions when  patch  tested  with  mineral wool (rock wool)   (Bjbrnberg and
i..i,, IQOO                          ft-?          n.RAPT--nn NOT  nnniF  OR  CITE

-------
Lowhagen, 1977).   Coating  the  fibers with phenol-formaldehyde did not change
the skin  reactions.   It was postulated that the skin reactions were mechani-
cally  induced  since the  fibrous,  but not the  powder,  form of the mineral
produced  the positive  skin response.   Clinically, the reactions are character-
ized by  either  an allergic macroscopic appearance or an erythematous  dermal
infiltration.  The  findings  for  tiineral  wool  correspond to those in a subse-
quent  study  (Bjbrnberg,  1985), where approximately 25 percent of  a  nonexposed
population showed a positive  response to patch tests  with  glass  fibers.
Clinically, the  lesions  resembled  allergic contact dermatitis.  Workers in the
glass  wool industry had the same  level of patch test reactivity, and chronic
exposure  to  glass  fibers  did  not change  the  original  level of  reactivity
(Bjornberg, 1985).
     The  ocular  effects  of MMMF  were investigated by  Stokholm et  al. (1982)  in
a  cross-sectional  study of  15 workers exposed to mineral  wool.  A matched
reference group  consisted  of 15  nonexposed persons.   Ocular examinations  were
performed on the exposed group just before beginning work after the  weekend and
again  4  days  later.  A significantly (p <0.001) higher frequency  of eye symp-
toms (running,  red, and smarting eyes) related to work conditions  was found
among  exposed workers.   The number of microepithelial  defects  on the  medial
bulbar conjunctiva  increased significantly (p * 0.009) after 4 days  of exposure
in  comparison  to those  in  the reference  group.   A pathologic  increase (p =
0.006)  in the neutrophil  count  of  the  conjunctiva!  fluid occurred in six
exposed  workers  after 4 days.   This  effect was  observed 1n only one worker
before  the  exposure.  An  increased  amount of mucus  was  found in 10 of  15
conjunctiva! fluid  samples taken prior to exposure and in 8 of 12  samples  taken
after  exposure.   In addition,  there was a significant  (p  <0.01)  .correlation
between  the  number of microepithelial defects  on the  medial bulbar  conjunctiva
and the  number  of  fibers accumulated in the opposite eye of the exposed work-
ers.   There  was also a  significant  (p <0.01)  correlation  between epithelial
defects  and  the  total  airborne dust.  The authors suggested that  the symptoms
and cellular changes observed  1n the exposed workers  were  a result  of  mechani-
cal irritation of the eye similar to that observed following skin exposure.
     Alsbirk et al. (1983)  examined 39  workers for ocular irritation.   The
workers  were  exposed for  an unspecified  duration  to  mineral  wool  fibers
released  from ceiling  boards at their workplace.  A  nonexposed control group
consisted of 23 individuals of the same  average age and sex, and  with  the same
smoking  habits.   Total  airborne  dust concentrations  at the workplaces ranged
July 1988                          6-3          DRAFT—DO NOT QUOTE OR CITE

-------
from 0.14 to  0.27 mg/m  for the exposed group, and from 0.10 to  0.45  mg/m3  for
the controls.   Airborne mineral fiber concentrations were not reported.   In the
exposed group, a  significant increase was observed  in  the incidence of burning
eyes, watery  eyes,  swollen eyelids,  disturbances  of sight, and smarting of the
nose.  In addition,  conjunctiva!  hyperemia was significantly increased and, in
five exposed  workers, «ineral  fibers were  found  1n the conjunctiva! mucous
threads.
     A study  by  Rindel  et al.  (1987) reported that  symptoms of  eye  and  skin
irritation among  persons at 24 Danish kindergartens were more  frequent  in
institutions  with ceilings covered  by  MMMF products  produced  with either
water-soluble or resin binder.   The concentration  levels in  the  indoor  environ-
ment were more  than 10,000 times less than the Threshold Limit  Value (TLV) in
Denmark,  and  there  was no significant difference  between the mean concentra-
tions of respirable  MMMF from control and study  institutions.   In  addition,
measured levels of  fibers/cm  (both respirable and  nonrespirable) were often
below the  reported  analyzed  detection  limit of  40 to 80 fibers/cm .  The
authors concluded  no association  between MMMF and  the  reported  symptoms  and
diseases.
8.1.1.1.2  Respiratory effects.   Several  studies  have  been  conducted to deter-
mine the  possible respiratory  effects  of fibrous  glass and mineral  wool.
Respiratory manifestations including bronchitis,  pharyngitis, rhinitis, asthma,
laryngitis, and  sinusitis  have been associated with exposure to MMMF.  These
effects are apparently reversible when  exposure to  fibrous  glass is  discontin-
ued.  No clear  evidence of pulmonary abnormalities  among exposed workers  was
found in several  studies.   However, increased prevalence of small  lung opaci-
ties and higher  incidences of chronic and dysplastic  pharyngolaryngitis have
been noted in a few workers exposed to fibrous glass for more than  5 years.
     During a 30-month period  from  1960  to  1962, 691 cases of  occupational
disease attributed to fibrous glass exposure were  reported in California (Milby
and Wolf, 1969).  Of these cases, 38 involved respiratory tract irritation. An
additional 28 case  reports describing adverse respiratory reactions  to fibrous
glass were  reported for a 13-month  period from 1967 to 1968.   Among the com-
bined 66 cases,  11 were females and 55 were males.  The individuals ranged in
age  from 17  to  68 years.  Occupational designations revealed that most of the
complaints were  made  by  individuals who were working  directly  with  fibrous
glass or fibrous  glass plastics, especially in dusty operations that involved

                                   *>-*          np&FT—nn NOT QUOTE OR CITE

-------
 cutting,  sanding, or machining  of the material.   However,  a number of  case
'reports  involved Individuals who worked in the general area where fibrous glass
 dust was being generated.  Chest x-ray evaluations were  performed  in 13  of  the
 28 cases that  occurred during 1967-68,  and none showed changes  that were
 attributable to fibrous glass exposure.  Three case reports, resulting from the
 same exposure, Involved maintenance employees who  had removed  fibrous  glass
 Insulation  from  steam  pipes.   Adverse effects developed  by these workers
 included sore throat,  nasal congestion, laryngeal pain,  and cough.  A summary
 of the respiratory manifestations in the 66 case  reports attributed to fibrous
 glass  exposure included bronchitis in 64 percent,  pharyngitis  in  38 percent,
 rhinitis  in  30  percent,  asthma  in 9 percent, laryngitis in  6  percent,  and
 sinusitis  in 5  percent of the  cases.   There were no reports  of  permanent
 disability  in any  of  the exposed  individuals.  However, the extent  of  the
 followup  period was not reported.  Recently,  Finnegan et al. (1985)  reported
 that work-related rhinitis, developed by 20  percent of the workers in a contin-
 uous filament glass fiber plant, was  due to an  irritant reaction  to  airborne
 glass  fibers.
     No  unusual  pavtems  of radiologic densities  were noted in  chest  x-rays of
 1389 employees  «ftga%fd in the **m;facture of  glass wool for 10 or more  years
 (Wright, 1968).   Tt>e workers were divided into four exposure categories ranging
 from minimal  to heavy exposure.   Total dust concentrations  varied  from 0.93 to
 13.3 mg/m .   Particle  concentration  ranged  from  0.09 to 0.32  million
 particles/ft   (3.3 to  11.9/cm ), and fibers made  up less than 1 percent  of  the
 total  particles.   The  median diameter of the  fibers was 6 urn, with 85 percent
 being  in the range of  2  to 10 urn.  The average values from samples throughout
 the  plant  indicated that  16  percent of the fibers were less than 40 urn in
 .length and  6  percent were  less than 20 urn.
     Utidjian and de Treville (1970) examined a number  of respiratory parame-
 ters,  including  chest x-ray* and pulmonary function tests, in 2332 workers from
 a  fibrous  glass manufacturing plant.  The workers  were  categorized  into age
 groups (<30,  30  to 49,  and >50 years) and further subdivided Into  three catego-
 ries based on  exposure levels (least, medium, and highest exposure).   The
 environmental  concentrations of fibrous glass  and duration of exposure were not
 reported.   No relationship was  observed between the degree  of exposure and
 respiratory history or radiologic findings.  In  addition, deviations  froir.
 predicted pulmonary function values did not  correlate with  exposure.

 July 1988                           8-5          DRAFT—DO NOT QUOTE OR CITE

-------
     In a followup  of the above study, a more detailed evaluation  was made of
30 of  these  workers who were over 50  years of age, 15  from the least exposed
group  and 15 from the most exposed group (Hook tt al.,  1970).  After obtaining
medical histories  from respiratory  disease questionnaires, the investigators
conducted physical  examinations,  fluoroscopies of the  chest,  electrocardio-
grams, hematologic  tests,  pulmonary  function  tests, and tests  for  lung  diffu-
sion capacity.   No  adverse effects  were observed that  could be attributed to
exposure to  fibrous glass dusts.   Moreover,  the Incidence of  ventilation or
diffusion abnormalities  found  in  the highest  exposed  group was similar  to the
incidence in the least exposed group.
     No evidence  of a respiratory hazard attributable to glass fiber exposure
was noted in a group of 70 workers who had a  mean exposure of  almost 20 (12 to
24) years in the manufacture of insulation-type fiberglass (Hill  et a!., 1973).
Chest  x-rays,  pulmonary function  tests,  and a questionnaire regarding respira-
tory symptoms  performed on the exposed workers were  compared  with  those of a
control group  living in the same geographic  area and matched  for  age, sex,
height, and  weight.   Environmental investigations found that total  dust in the
workers'  breathing  zone ranged from  0.4  to 12.7 mg/m  .   Respirable  size fibers
ranged from  1.0  to  4.8 fibers/cm .   Approximately  75  percent  of  fibers
collected at the site  of  emission were  less  than 4 urn  in diameter, and  35
percent were  less  than 2 urn in diameter.   Over 50 percent of these  fibers were
less than 50 pm long.
     Nasr et al.  (1971) conducted a  chest x-ray survey of 2028 male employees
of a factory manufacturing glass fibers.   The  employees (196  office workers and
1832 production  workers) ranged in  age from 20 to 59  years,  and approximately
two-thirds of  the  workers were employed  for 10 years  or longer.  The degree of
exposure of each worker  to glass fibers was classified as low,  average,  or  high
based  on previously performed industrial hygiene surveys.  No information on
airborne dust,  particle, or fiber concentrations was  provided.  Approximately
16 percent of the workers were found  to have radiographic abnormalities, the
most common being "increased lung markings," emphysema, and abnormal shadows  of
aortas and  hearts.    Although  there  was  a correlation between duration of
employment and age  with the occurrence  of abnormal  chest radiographs, the
prevalence of all abnormalities was similar among office workers and production
workers.
     A clinical  evaluation of 427 workers with an average of 12 years'  experi-
ence  in  a spun-glass fiber plant was performed by Maggioni  et al. (1980).
Julv 1988                          8-6          DRAFT--DO NOT QUOTE OR CITE

-------
Results were  compared  to a group of 40 nonexposed workers of similar age and
smoking habits.  Airborne  dust  concentrations  in the plant averaged 1 mg/m .
The percent by weight  of glass  fibers  In the total dust was about 1 percent in
the  production units  and 3  to  7  percent  in the spinning  department.
Respirable-size glass particles were detected following morphologic  analysis of
glass fibers.   No cases of pneumoconiosis  end/or pulmonary fibrosis  were  found.
The incidence  of chronic bronchitis in subjects exposed to higher concentra-
tions of  glass fibers  was similar to that  of the controls.   However, the
incidence did  increase  with  years of service  and smoking.  It  is worthwhile
to note that  the comparison  group of unexposed workers was extremely small.  A
higher incidence of  chronic  and dysplastic pharyngolaryngitis was observed in
workers who had worked more than 5 years  in departments where  exposure  to glass
fibers was high.
     Hill  et al. (1984) conducted a stratified cross-sectional morbidity  survey
at a  plant that manufactures  fibrous glass  wool  insulation  in the United
Kingdom.   The  study  involved 340 men and women, including all  current living
and former employees aged 55 to  74 years  with over  10 years  of employment,
workers of the same  age with 1 to 10 years of employment, and two  groups aged
45 to 54  with either I to 10 years or over 10 years  of employment.   There wes
no information on  exposure concentrations.   All  subjects  were given a  respira-
tory  questionnaire  and  a  physical  examination.   In  addition,  lung function
tests and  chest x-rays  were  performed.  The  prevalence  of small  lung opacities
on chest x-rays of category 1/0 or more (using International  Labor  Organization
(ILO) classifications)  was 10.6 percent  and was significantly related  to age,
sex,  previous  employment in  dusty occupations, and to never having  lived in a
smoke-free zone.  There was a greater prevalence of small  opacities  >1/0  in men
with  more  than 15  years of prior dust exposure in comparison  to those  who had
worked  less  than 15 years (38.7 percent  versus 9.6 percent).   A  residual
prevalence of  small opacities was observed among 9.3 percent of  the  men with no
previous  dust exposure.   However,  no  relationship was found between  small
opacities  in  men with  less  than 15 years or  no previous  dust  exposure  and
duration  or  intensity  of  exposure  to MMMF.   There was also  no significant
relationship  with  duration of exposure for any lung function index.  Respira-
tory  symptoms  were  not significantly associated with  employment in the  MMMF
industry.
     Weil! et al.  (1983)  conducted  a  cross-sectional  investigation  of  the
respiratory health of  1028 male  workers at seven plants  (five  glass fiber and
i..i., IODO                          fi-7          DRAFT--nn NOT niJCTE OR CITE

-------
two mineral wool)  in the  United States.  The average age of the workers was 46
years, and  the median  length of employment was  18 years.   Employees were
classified, based  on smoking habits,  as smokers,  ex-smokers, and nonsmokers.
Health assessment was performed through respiratory questionnaires,  ILO  classi-
fication of  chest x-rays,  and pulmonary function tests.  Concentration  of
airborne fibers  in the plants  ranged  from  0.001  to >1.0 fiber/cm , and the
sizes of the fibers were classified as ordinary (>3 urn  in diameter),  fine  (1  to
3 pm),  very fine  (<1 pm),  and mineral wool fibers (>3 urn).  The prevalence of
respiratory symptoms,  Including dyspnea, chronic  bronchitis,  and shortness of
breath, was  not related  to KMMF exposure  but was significantly (p <0.05)
related to cigarette smoking.   Although lung function tests did not indicate
any adverse  effects  of exposure, measurements  were  related  to age, height,
race, and cigarette  smoking.   Of 941 chest x-rays examined,  the prevalence of
small opacities in categories 0/1,  1/0,  and 1/1 was, respectively, 7, 3,  and  1
percent with no opacities >1/1 being  observed.  The prevalence  of small opaci-
ties  increased  as both age and smoking increased.  Among the  smokers,  the
probability of small  opacities was  positively  associated with  exposure dura-
tion.   However, among ex-smokers and nonsmokers, there  was no association  with
exposure measures.   In  addition,  exposure  to a mixture  of ordinary  and fine
fibers significantly increased the probability of  low-level  small  opacities
among current  smokers.   Among  ex-smokers  and   nonsmokers, this effect was
observed only  in  a plant  having higher airborne concentrations of fibers  less
than 1 urn in diameter.
     Carpenter and  Spolyar  (1945)  evaluated chest x-rays of  84 workers  in the
rock and slag  wool industry with 7 to 26 years of exposure.   The workers  were
exposed during the major  portion of their  industrial employment  to  airborne
dust  concentrations  ranging  from 12  to 36  million  particles/ft  (444  to
1333/cm3).   No evidence of silicosis or fibrosis was found in the exposed group
when compared  to  a control  group of nonexposed  workers from the  same plant and
living in the same community.
     Skuric  and  Stahuljak-Beritic  (1984)   reported  results  of  ventilatory
function tests performed  on 106 male  and 56 female workers exposed to airborne
rock wool fibers  at three production plants.   Airborne  dust  mass had a mean
range of 1.8 to 28 mg/m  with an average mass concentration varying from 2.8 to
13.1 mg/m3.   The  range  of mean respirable   fiber  concentrations  was  0.003 to
0.463 fiber/cm3.    The results of the pulmonary function tests were signifi-
cantly (p <0.05)  lower than  the expected  reference values  in  both  men and
1..1,, TOOO                          *»-fi          noftCT._nr\ UHT AMHTC no PTTF

-------
women.  No relationship, however, was found between the magnitude  of  changes in
ventilatory function and  the  cumulative exposure of workers to total dust or
fibers.   Since  the workers 1n  this survey  had  previous  exposure to other
occupational dusts before starting employment at  the rock  wool  plant,  a  followup
survey was performed 5 years  later.  Ventilatory function tests were.performed
on  102 workers (63 men and  39  women)  from the  original  survey population
exposed only  to MMMF.   The authors concluded that the observed  changes  in
ventilatory function occurring during the 5-year  period could not  be  attributed
to MMMF exposure.   Moreover,  indices of lung function deficit and indices of
exposure  (duration, fiber years,  and total particle years) were not  signifi-
cantly related.
     Respiratory abnormalities in 21 male workers involved in the  production  of
rock wool  were investigated by Ma1mberg et al.  (1984) using lung function tests
and chest x-rays.   Mean duration of employment was  17.6 years.  There was no
history of previous exposure  to asbestos or other  potentially harmful dusts,
and no history  of  lung disease.  The results were compared to a control  group
of 43  male  individuals, matched  for age and smoking  habits,  from the  same
county as the workers.   Fiber levels in the plant averaged 0.21 fibers/cm ,  and
about  92  percent  of these fibers were  respirable.   There were no significant
differences between the workers and the controls  in  any  of  the lung  function
tests.  In addition, all men had normal chest x-rays.
     Newball  and  Brahim  (1976) reported  that severe respiratory symptoms
developed in  a  family  of  four during a  3-year exposure to fibrous glass  fibers
liberated from  the  fibrous glass lining of the central air-conditioning  system
in their home.  Analysis of dust samples in the home revealed glass  fibers with
diameters ranging  from 1 to  15 urn.  Respiratory  symptoms  included  sinusitis
with obstructed nasal  passages, nosebleed, shortness of breath, and  cough.  In
one family member,  an  11-year-old boy, a  chest  x-ray revealed an edematous,
erythematous, tracheobronchial mucosa with marked tracheobronchitis.
In addition,  bronchial  washings contained fibrous  glass with diameters  ranging
from  0.5  to 0.7 urn.  Three months  after the exposure, the family showed no
objective signs  that could be  attributed to fibrous  glass  exposure.   Chest
x-rays were  normal,  and pulmonary function tests indicated only mild obstruc-
tions  in two individuals.
8.1.1.2   Epidemiological  Studies.   Prompted by  evidence  from animal studies
indicating that MMMF  induce  both carcinogenic and  fibrotic  responses in the
respiratory  system,  numerous  epidemiological studies  have  been  conducted to
July 1988                          8-9          DRAFT--DO NOT QUOTE OR CITE

-------
 identify potential  relationships  between  exposure  to MMMF and malignant and
 nonmalignant respiratory disease  (NMRD).  Three major studies, each consisting
 of a series of historical prospective mortality studies with cohorts of occupa-
 tional ly exposed workers from  the United  States,  Canada, and Western Europe,
 have been conducted.  Subsequent studies In  each series reported findings after
 allowing additional years of followup of  the cohort under study.  Enterline
 and March (1982) and Enterline  and colleagues (1982,  1983, 1986, 1987) examined
 the mortality experience of U.S.  workers, Shannon  and colleagues  (1982, 1984,
 1986, 1987) examined  the mortality experience of Canadian workers, and Saracci
 and colleagues  (1984) and subsequently  Simonato and  colleagues (1986a, 1986b,
 1986c, 1987) examined the mortality experience of workers  from several European
 countries.   Recently, Doll  (1987) examined  the mortality experience  of the
 combined U.S.,  Canadian, and European cohorts.
 8.1.1.2.1  Epidemiological  studies in U.S. workers.   Enterline and March (1982)
 and Enterline and  colleagues (1983,  1986, 1987) conducted a series of studies
 on the mortality experience of production  and maintenance workers in 17 U.S.
 plants that produced MMMF at some time between 1940 and 1963.  Special emphasis
 was placed  upon workers  exposed to small-diameter MMMF, i.e.,  less than 1.5  urn,
 because they are most similar  to  asbestos fibers.   The maximum followup period
 for the cohort  was  37 years.   It was  found that the greatest risk  of respiratory
 cancer occurred among workers  exposed to mineral wool.   Fibrous  glass workers
 had much lower  rates  of respiratory  cancer.   No excess carcinogenic or noncar-
 cinogenic risk  was  found among  workers exposed to small-diameter  fibers.
      Enterline  and Marsh (1982)  conducted a historical  prospective mortality
 study of 7049 male  production  and maintenance employees who had worked for 1 or
•more years  in  any  one of eight fibrous  glass (FG)  or mineral  wool (MW) plants
 between January 1,  1945, and  December 31,  1963.   For employees  in one plant
 producing small-diameter fibers,  the cohort  included  those  with  6 or more
 months of exposure.   After the cohort from all eight plants was assembled from
 company employment  records, vital status  was determined through  U.S. Social
 Security Administration  (SSA)  records and  other unspecified files. After using
 a modified  life-table technique to obtain person-years  at risk  of dying  to
 compute the standardized mortality  ratio  (SMR),   the  expected  numbers of
 cause-specific  deaths were calculated from  U.S. age-,  race-,  and sex-specific
 rates for  the  years  1945  to   1973,  presumably for appropriate  5-year time
 intervals.   The investigators  did not state  whether nonwhite males were exclud-
 ed from analysis.   By the  end  of the followup  period, December  31, 1973, 14
 • •   	•-                          o-in         noACT—nn wnr  nuniF nR CITE

-------
percent of  the  cohort had died.   Vital status was determined  for all but 31
(0.4 percent) of  the 7049 cohort members.   Of the  1006  deaths, death certifi-
cates were  obtained  for all  but 39 (3.9 percent),  Indicating a high degree of
followup and  a  high  trace rate.   The 7th Revision of the International  List of.
Diseases and Causes of Death (ICO) was  used to code causes  of death.
     Workers  were exposed to MMMF 1n  either  FG  or MW plants.   To test the
hypothesis  that exposure  to  respirable WMF results  in malignant respiratory
disease, including mesothelioma,  and NHRD,  excluding  influenza and pneumonia,
the investigators developed  an  exposure index for  each  worker.   As of the end
of 1972, concentrations of respirable  fibers  to  which FG workers were exposed
were  very   low;  the   estimated  mean exposure  level was 0.023 fibers/cm .
Exposure data for MW  workers  were unavailable  because of the  confounding
presence of asbestos  in mineral  wool  plants.
     Overall, investigators  found  no significant excess in mortality, a mani-
festation of  the  healthy  worker effect in this group  of factory  workers (1006
observed, 1081.7 expected, SMR = 93).   When investigators analyzed the data  for
workers  from  all  eight plants  by time since first exposure (surrogate for
latency  period),  they found  no  significantly  increased   SMRs  for any
cause-specific deaths  among  those with less than 20 years  since  first exposure
or among those  with  20 or more years since  first exposure.   However, there was
insufficient power to  detect digestive cancer, respiratory system cancer,  and
all other cancers among those with 20  or more years  since  first  exposure,  and
there was insufficient power to detect NMRD in those  with  either less than  20
or 20 or more years since first exposure.
     When the data were analyzed by plant,  type  of fiber,  and  time since first
exposure for malignant  and nonmalignant  respiratory disease,  significant
excesses were limited to  malignant respiratory  disease among  MW workers in
Plant 7  with less than 20  years  since first exposure (2 observed, 0.794
expected, SMR =  251.6, p <0.05) and those  with  20 or more years since  first
exposure (2 observed,  0.22 expected,  SMR « 905.7, p  <0.05), even  though  the
power to detect  a 50 percent increased incidence of respiratory  cancer was low
(12 and 4 percent, respectively).  However, the small  numbers of  deaths  and  low
power (except for those workers from all FG plants combined with less  than 20
years since first exposure who  died from respiratory  cancer) precluded drawing
any conclusions  from  these  data.   Of  the  7049  cohort members,  4120 (58.4
percent) had 20 or more years of observation.   It was not possible,  however,  to

-------
determine the percent of the cohort with 20 or more years of exposure  and  30  or
more years of observation following Initial exposure to MMMF.
     To further  determine  whether exposure to respirable FG fibers (<3 urn in
diameter)  resulted  in malignant  and nonmalignant respiratory  disease, the
investigators analyzed the data by cumulative exposure to respirable FG fibers.
To avoid  overlapping  exposure and follovup periods, the  investigators calcu-
lated each worker's cumulative exposure as of December 31, 1962, thereby giving
11 years  of  observation without overlapping exposure years.  No relationship
was found  because  the power was  low.  Generally,  such  a  technique alleviates
underestimation  of  cause-specific risks  in the study group.  However, the low
power to  detect  malignant  and nonmalignant respiratory disease might  account
for the failure to find a relationship between exposure to respirable  FG fibers
and either malignant or nonmalignant respiratory disease.
     Enterline and  Marsh  (1982)  concluded that (1) the  presence of asbestos  in
one MW  plant where  significantly increased deaths  occurred  from  malignant
respiratory disease confounds the interpretation of data on workers employed  in
MW plants; (2) no  direct  relationship  exists between exposure to respirable  FG
fibers and  the  incidence  of either aalignant  or nonmalignant  respiratory
disease;  (3) there  is  no  clear evidence that exposure to MMMF produced malig-
nant or nonmalignant respiratory disease; and (4) studies of workers  exposed  to
such low  levels  may be inappropriate for testing hypotheses regarding MMMF and
disease.
     The conclusion of the investigators that there is no causal relationship
between MMMF,  particularly respirable  FG fibers,  and disease is weakened by
several limitations.   First, there  was  insufficient power to  detect a  50
percent increased  incidence of either malignant or nonmalignant respiratory
disease.   Second,  although no cases of mesothelioma were reported, because no
cause-specific expected values  were reported,  it was not  possible  to  determine
whether the group of exposed subjects was large enough to detect an increase  in
mesothelioma incidence  of  50 percent above that  of  U.S.  white  males.  Third,
the small  number of deaths and low power preclude drawing the conclusion  that
exposure to MMMF is not causally related to adverse health effects.   In addi-
tion, the exposure levels might have been too low to detect a weak carcinogenic
effect if, in  fact, it existed.   Furthermore,  under these conditions,  6 months
of MMMF exposure might have been  too  short a period to  consider.  Usually,
workers with less  than 1 year of exposure are  excluded because  the briefer the

	                          ft-19          HRAFT--nn NOT QUOTE OR CITE

-------
 period of exposure to  a  carcinogen,  the  longer the latent period and the more
 likely the association will go  unrecognized,  Finally, besides asbestos expo-
 sure, smoking is another  variable  that night  confound interpretation of the
 data because of Us influence  on cancer «t several sites.
      When data  from  additional  MMMF  production  facilities became available,
 Enter!ine et al. (1983) conducted  a  historical prospective mortality study of
 16,730 male production and maintenance employees who  had worked for 1 or more
 years in any one of 17  FG (14,884 workers) or MW plants (1,846 workers) between
 January 1, 1945, and December 31, 1963.   For employees in two plants producing
 small-diameter fibers  (<1.5 urn), the cohort included those  with  6 or more
 months of exposure.  The  date of initiation of the study varied:   (1) 1945 was
 selected for 16 plants because of problems in tracing terminees prior to 1945;
 and (2) 1940 was selected for  1 plant to permit comparison with a previously
 conducted study at that  plant.  After the  Thermal  Insulation Manufacturers
 Association (TIMA)  and its member  companies abstracted  employment records to
 assemble the cohort, vital status  was identified through a search of company,
 SSA, state drivers' license, and Veterans Administration (VA)  files, as well  as
 personal  contacts.    A modified  life-table technique was used to  obtain
 person-years at risk  of dying.   The U.S. white male average annual age-specific
 death  rates  for 5-year  time  periods from  1940  to 1974  were applied to
 person-years to obtain expected deaths  for computing SMRs.   The  1975 death
 rates were applied to  person-years accrued  from  1975 to 1977.  Because non-
 whites constituted  less than  2  percent of  the cohort, investigators believed
 that the use of U.S.  white male  death rates  did not bias results.  To determine
 whether geographical  variations  in expected mortality rates  existed, investi-
 gators compared age  and  calendar-time adjusted  state and national  mortality
'rates and found similar results  (results not reported).   The years of followjp
 were  from  January  1,  1941, for  1  plant  and from January 1,  1946,  for the
 remaining 16 plants to  December 31, 1977, allowing a  maximum  followup of 37  and
 32 years,  respectively.   By the end  of  the followup period, 12,633 of the
 16,730 (75.5 percent) cohort members  were alive; 3,761 (22.5  percent) had died;
 and 336  (2  percent)  were  lost to followup.   Death certificates were obtained
 for all  but 108  (3 percent).   Deaths were coded by a nosologist according to
 the 7th Revision of the ICD.
      Exposures to MMMF  less  than 3  urn in  diameter  for each worker  were  estimat-
 ed by  combining  environmental survey results and  individual job histories.
 Exposure occurred  from either January 1,  1940 (1 plant), or January  1, 1945
 July 1988                          8-13         DRAFT—DO  NOT QUOTE  OR  CITE

-------
(16  plants),  to December  31,  1963.   Each worker's  occupational  history was
characterized by an  estimate  of cumulative exposure and average intensity of
exposure to MMMF.   For all FG  plant workers, mean cumulative exposure was 4.47
fibers <3 urn/cm -months,  and  average  Intensity of exposure was 0.039 fibers
<3 Mm/cm -months;  for  all MW  plant  workers,  mean  cumulative  exposure  was
53.26 fibers  <3 urn/cm -months,  and  average  Intensity  of exposure  was
0.353 fibers <3 urn/cm -months.   A  survey  of  certain MMMF plants conducted by
the  University  of  Pittsburgh  served  as a basis for reconstructing historical
exposures of  workers to respirable fibers.  Personal samples were collected
from within the breathing zones of employees  and analyzed by phase contrast and
transmission electron microscopy.  The results  of this  survey were published in
Esmen et al.  (1979b).   Enterline et al. (1983)  stated that the results of  this
survey indicated that past exposure levels did  not differ from  present exposure
levels and  thus were equivalent.   However, because  Esmen et al. (1979) made no
attempt to estimate past exposure levels,  it  is unclear how Enterline and Marsh
(1982) reached  the  conclusion  that  past  and  present  exposure levels  were
equivalent.
     Neither observed  nor  expected numbers of  deaths were  reported for several
causes.   Consequently, the overall mortality  experience of the  cohort, the risk
of bronchial, trachea!,  and  lung cancers  combined,  and the risk of laryngeal
cancer cannot  be  determined.    No  deaths  from  mesothelioma  were  reported.
Overall,  investigators  found  significantly elevated mortality  from  NMRD  (150
observed, 113.5 expected,  SMR  - 132.1, p  <0.01).  When the data were analyzed
by plant and  time  since first exposure, there  were significant excesses  from
respiratory cancer among  those ever  employed in MW plants (45 observed,  28.1
expected, SMR = 160.1,  p <0.01), especially  among those with 20 or more  years
since first employment (30 observed,  19.0  expected,  SMR = 157.2, p <0.05), even
though the  power to  detect a 50 percent increase In respiratory cancer was 64
percent  or  less.   The  greatest excess of respiratory cancer in MW workers
occurred among  those ever employed  1n Plant 7 (6  observed, 1.32 expected,
SMR = 451.3,  p  <0.01,  Power * 14  percent), especially among those with 20 or
more years  since  first employment (4  observed,  0.46  expected,  SMR  = 868.5,
p <0.01,  Power = 3 percent).   There was also  an excess emong  those ever  employed
in Plant  12 (16 observed, 2.04 expected, SMR  =  293.6,  p <0.05,  Power  =  8
percent).  The  investigators  stated that  the excesses  in  Plant 7 were probably
not attributable to  employment in the MW plant because the workers were  hired
late in life and had short exposure.   However,  the power to detect a  50  percent
                                   "-if         nparr—nn NOT nuniF OR CITE

-------
increased  incidence  of  respiratory cancer was only 4 percent among  those with
less than  20 years since first employment, 3  percent  among  those with  20 or
more years since  first  employment,  and  14 percent  overall.   Nevertheless,
significant  Increases  were observed.   Because of an absence  of  quantitative
data on  exposure to HW and because of  possible asbestos exposure, these data
provide  limited  evidence of a causal association between respiratory cancer  and
MW exposure.
     Similarly,  when the  cohort was analyzed for NMRO by plant and  time since
first exposure,  there  were significant  excesses of death from  NMRD among those
workers  ever  employed  in FG plants (129 observed, 99.53 expected, SMR = 129.6,
p <0.01),  particularly  among FG  workers with 20 or more years  since  first
employment  (85  observed,  62.73  expected, SMR = 135.5, p <0.01).  The signifi-
cant excesses  of death from NMRD among all MMMF workers (150 observed,  113.55
expected,  SMR  =  132.1,  p <0.01) were concentrated among those with  20  or more
years since first employment  (100 observed,   72.93  expected,  SMR « 137.1,
p 
-------
     To  determine whether  workers exposed  to small-diameter FG  were at
increased risk of respiratory cancer or NMRD, the investigators examined plants
where workers were most likely to have been exposed to fine-diameter fibers.
In Plant  9,  a significant excess of NMRD occurred among workers  never  exposed
to fine  fibers (71 observed,  52.07 expected,  SMR « 136.4,  p <0.05).   Small
numbers of deaths and low power precluded meaningful analysis  of  fiber  exposure
and respiratory cancer.
     Next, the  Investigators  reanalyzed  the  data from Bayliss et al.  (1976),
who studied  the  mortality experience  of workers  In Plant 9.  Neither Bayliss
et al. (1976)  nor interline  et  al.  (1983)  found a significant relationship
between exposure  to  fine-diameter FG  particles and either NMRD or  respiratory
cancer.    However, the  lack of statistical significance may be related to the
insufficient power to detect either respiratory cancer or NMRD in Plant 9  among
those ever exposed to fine fibers.
     When the data on  respiratory cancer were analyzed by cumulative exposure
to various  lengths of  MW, the investigators  did not find a  clear relationship
between exposure  to  respirable fibers  and respiratory  cancer  excesses among  MW
workers.   A  significant excess of respiratory cancer deaths  was found  among
workers with a cumulative  exposure  to 21.88 to  99.99 fibers/cm -months  (17
observed, 9.54 expected, SMR = 178.2,  p <0.05), even though  the power to  detect
a 50  percent increase  in respiratory cancer was low.   Small numbers of deaths
and low power  preclude meaningful analysis  and do  not support the  conclusion
that  no relationship exists between respiratory cancer and  cumulative exposure
to KW fibers.   When  the data on NMRD were analyzed by cumulative exposures  to
various MW fiber  lengths, no significant excess deaths were  found in any  of  the
four  fiber  categories.   However, there was  insufficient power to detect  a 50
percent increased incidence of NMRD in each  category.
     Finally,  no significant  excess  of  respiratory cancer  deaths  was found
among MW  workers  with  20 or more  years  of  exposure and  3D or more years of
observation  following  initial exposure.  However,  only two of the seven MW
plants had deaths in this category, and there were too few workers  and  insuffi-
cient power  to  detect  a 50 percent increase  in respiratory cancer in  either
plant or  even in  all MW plants.
     The  investigators  concluded that no consistent evidence of  a respiratory
hazard exists among MMMF workers, but there is some evidence, although  weak,  of
excess NMRD  among both FG and MW workers.  The authors correctly concluded that

July 1988                          8-16         DRAFT--DO NOT QUOTE OR  CITE

-------
more  followup,  smoking habit  histories,  and  information  on the presence of
asbestos  and other possibly  confounding  substances,  e.g., silica, asphalt,
phenol, formaldehyde, ammonia, carbon monoxide, solvent vapors,  and heavy metal
fumes, are needed.
     However, as  1n the previous  study, this  study  1s  limited by small numbers
of  deaths  and low power to detect  respiratory  cancer  and NHRO  when the  data
were  subdivided  into  many categories.   Although data  were analyzed by latency
and  length of exposure, 1t was not possible  to determine the percent of the
cohort with  20  or more years  of  exposure  and 30 or more  years  of  observation
following  initial  exposure  to HMMF.   In fact,  only  two MW plants had  deaths 30
years after  initial  exposure  among workers with 20 or more years  of exposure.
The  average  duration  of exposure  was 11.67 years for  all  FG plants and  11.08
years for  all  MW plants.   In addition, the selection  of 1945 rather than 1940
for  initiation of the study because of problems in tracing terminees  in  16  of
17 plants may have underestimated the true risk of respiratory  cancer and NMRD.
Based on  these  data,  one cannot  conclude  that  exposure to respirable MMMF  is
not associated with respiratory cancer and NMRD.
     Prompted  by  the  need for  more conclusive epidemiological  evidence,
Enter!ine et al.   (1986) extended the followup period of their 1983 study  for an
additional 5 years and  conducted a nested case-control study to  control  for  the
effects of cigarette smoking.   Because the start of the followup period for  one
plant was delayed  from  1941 to 1946, the cohort was reduced from 16,730 workers
in the 1963  study  to 16,661 in this study.  Therefore, this cohort consisted of
16,661 white male production  and  maintenance  employees who had  worked for 1 or
more years in any  1 of  17 FG or MW plants between January 1, 1945, and December
31,  1963.   For  employees  in two  plants  producing small-diameter  fibers
(<1.5 urn), the  cohort  included those with  6  or more  months of exposure.   A
modified  life-table technique was used to obtain person-years at risk of dying.
The  U.S.  white  male mortality experience  adjusted  for age and time was applied
to  person-years  to obtain the expected number of deaths.   Because  the previous
study (Enterline  et al.,  1983) showed no significant differences between SMRs
based on  state  versus  national mortality  rates, U.S.  national  rates  had been
selected as  the reference population.  However,  subsequent investigations found
that  local death rates from the  areas  in which the plants were located were
more  appropriate  for calculating  expected  mortality  from respiratory cancer
(Enterline et al., 1987).  The investigators estimated that for FG plants,  mean
exposure  levels  were  0.03 fiber/cm .   For MW plants,  mean exposure levels were
July 1988                          8-17          DRAFT—DO  NOT QUOTE OR CITE

-------
1.5 fiber/cm  before  1945,  Q.3 between 1945 and 1960,  and  0.03  after 1960.
Besides exposure  to MMMF, workers  had potentially  confounding exposures  to
silica, asphalt,  asbestos,  phenol, formaldehyde, ammonia,  carbon monoxide,
solvent vapors,  and heavy metal fumes.
     The years of  observation  for workers from all 17 plants were from January
1, 1946, to December 31, 1982,  allowing a  naximum followup of 37 years, as well
as 19  years  of  observation  without overlapping exposure years.  By the end of
the followup  period,  December 31, 1982, 11,328 of  the  16,661  workers (68.0
percent) were alive;  4,986  (29.9 percent) had died; and 347 (2.1 percent) were
lost to  followup.   Death certificates  were obtained for all  but 145 (2.9
percent).   The investigators were  successful  in  determining vital status and
obtaining death certificates for some  workers who had been lost to followup in
the 1983 study.   All  deaths  were coded to  revisions in effect at the time of
death  so that U.S.  death rates for corresponding time  periods were  used to
generate expected  numbers of  deaths.   For ease of presentation,  investigators
used the codes listed in the 8th Revision.
     Significantly increased causes of  death for 1978 through  1982 and for the
entire period of  1946 through  1982 are listed in Tables 8-1 and 8-2, respec-
tively.  Significant deficits during 1946  through 1982 are  listed  in  Table  8-2.
     When the cohort  was  analyzed by time sine* initial employment,  investiga-
tors found among those with  less than 20 years  since first employment a  signif-
icant  excess  from all  other  heart disease  (56  observed,  30.48  expected,
SMR =  183.7,  p <0.01).   However,  among those with  20  or  more years since
initial employment, there were significant  excesses from all   causes' of death
(3263  observed, 2974  expected, SMR  = 109.7, p  <0.01); all malignant  neoplasms
(735 observed,  658.01 expected, SMR  = 111.7, p <0.05); respiratory system
cancer (301 observed,  242.15 expected,  SMR  = 124.3, p <0.01);  lung cancer (26£
observed, 229.66 expected,  SMR = 125.4, p <0.01); all other heart disease (139
observed, 120.24 expected,  SMR = 151.6, p <0.01); NMRD  including influenza and
pneumonia (236 observed, 197.15 expected,  SMR ~ 119.7, p <0.01);  NMRD excluding
influenza and pneumonia  (172 observed,  127.50 expected, SMR «  134.9, p <0.01);
and nephritis and nephrosis  (24 observed,  14.77 expected, SMR = 162.4, p <0.05);
and  a  significant  deficit  from accidents  (94  observed, 116.04  expected,
SMR ~ 81.0, p <0.05).   These findings indicate  that  20  or more  years  of  latency
are needed to develop NMRD.  Except for laryngeal cancer, there was  sufficient
power  to  detect  a  50 percent  increase  in cause-specific  deaths among the
exposed workers.
July 1988                          8-18         DRAFT—DO NOT  QUOTE OR CITE

-------
         TABLE 8-1.  SIGNIFICANTLY ELEVATED SMRs BY CAUSE FOR ALL MMMF
                         WORKERS BETWEEN 1978 and 1982
Cause of death
All causes
All cancers
Respiratory system cancer
Lung cancer
NMRD
Observed deaths
1,251
320
135
130
69
SMR
114.7*
118.7*
129.7"
130.8*
128.5*
•Significantly different from national mortality rates:  p <0.05.

Source:  Enterline et al. (1986).
         TABLE 8-2.  SIGNIFICANTLY ELEVATED SMRs BY CAUSE FOR ALL MMMF
                         WORKERS BETWEEN 1946 and 1982
Cause of death
Respiratory system cancer
Lung cancer
Neoplasms of other lymphopoietic
tissue (residual)
Other heart disease (residual)
NMRD
Nephritis and nephrosis
Observed deaths
391
370
36
195
230
47
. SMR
115.6**
116.0**
144.2*
159.6"
132. 3«*
144.2*
 "Significantly different from national mortality rates:  p <0.05.
"•Significantly different from national mortality rates:  p <0.01.

Source:  Enterline et al. (1986).
July 1988                           8-19          DRAFT—DO NOT  QUOTE OR  CITE

-------
        TABLE 8-3.  SIGNIFICANTLY DECREASED SMRs  BY CAUSE  FOR  ALL MMMF
                         WORKERS BETWEEN 1946 and 1982
Cause of death
Ischemic heart disease
Influenza and pneumonia
Cirrhosis of the liver
External causes of death
Accidents
All other causes (residual)
Observed deaths
1,686
64
99
405
279
437
SMR
94.9*
76. 5*
77. 6»
85.1"*
85.9"
91.0*
 •Significantly different from national mortality rates:   p  <0.05.
"Significantly different from national mortality rates:   p  <0.01.
Source:  Enterline et al.  (1986).

     In  an  update of their  1986 unpublished study,  Enterline et al. (19E7)
found that after 20 or more years of latency, the observed excess  from respira-
tory cancer among  MMMF  workers was significant compared to  national  mortality
rates  (SMR  =  124.3,  p  <0.01),  but not  compared to local  mortality rates
(SMR = 112.3).   The  observed excess from lung cancer among  MMMF workers after
20 or  more  years of latency, on  the other  hand, was significant  compared to
national  (SMR =  125.4,  p  <0.01) and local  rates (SMR = 112.7,  p <0.05).   This
shows  the strength  of the lung cancer hazard, which appears to be responsible
for the  respiratory  cancer  excess among MMMF workers (288 lung cancers  out  of
301 respiratory cancer deaths).
     To  explore  a  potential  relationship between respiratory cancer, latency,
and type of fiber produced,  the investigators next subdivided the respiratory
cancers  among  the cohort  by type  of  process ((1) fibrous  glass filament,
(2) fibrous glass wool,  (3)  fibrous glass filament and  fibrous glass wool, and
{4) mineral wool) and time  since initial employment.  Among those with 20 or
more years since first employment, significant excesses from respiratory cancer
occurred among  workers exposed  in plants that  produced  both  fibrous  glass
filament and fibrous  glass  wool (155 observed,  119.73  expected,  SMR = 129.5,
p <0.01) and mineral  wool  (45 observed,  30.80 expected,  SMR =  146.1, p <0.05).
Neither  excess  remained significantly elevated when compared  to  local  rates
(Enterline et al., 1987).   Only one plant had sufficient power to detect a  50
percent  increased  risk  of death from respiratory cancer.   Despite  low  power,
statistically significant  excesses  from  respiratory  cancer  were found  in  three
plants.

July 1988                          8-20         DRAFT--DD NOT QUOTE OR CITE

-------
     When latency was  further  subdivided  into four time periods since first
employment, there was  a  clear  trend of increasing SMRs with  increasing time
since first employment  in  plants that produced both fibrous glass filament and
fibrous glass wool.  Among workers  with 30 or more years of  latency, signifi-
cant excesses were found in fibrous  glass  plants  producing  continuous filaments
and  glass  wool  (90 observed,  65.29  expected,  SNR =  137.8. p <0.01) and  in
mineral wool  plants (25 observed,  16.48  expected, SMR *  151.7, p <0.05).
Neither excess  remained  significantly  elevated when  compared to local rates
(Enterline et al.,  1987).   Enterline et  al.  (1987)  found  that the greatest
excess of  respiratory  cancer  occurred  among MW workers compared to either
national  (SMR = 148.4,  p <0.01)  or  local  rates  (SMR = 133.6, p <0.05).  This
shows that respiratory  cancer  among MMMF  workers  is truly elevated in the MW
subcohort.   Although the investigators  claimed that  no trends existed  between
latency and fibrous glass  filament  and  mineral wool plants, there was insuffi-
cient power to  detect  a 50 percent  increased incidence in  respiratory  cancer,
if it did exist  in these types  of plants.
     When the cohort  was analyzed by process  and by duration of employment
group (surrogate  for  length of  exposure),  no positive relationship between
respiratory cancer  and  duration  of  exposure  by process was observed.  However,
except for workers  in  the  less than 10 years'  exposure category,  there was
insufficient power  to  detect a 50 percent increase  in  respiratory  cancer in
workers with either 10  to  19 years, 20 to  29 years, or 30 or more years of
exposure  to MMMr.
     When respiratory  cancer deaths were  analyzed by  latency and  length of
exposure, among  workers with 20 to 29 years of exposure and 30 or more  years of
latency,  there was  a  significant excess among FG  workers  (26 observed, 16.24
expected, SMR = 160.1,  p <0.05), but not  among MW workers.  However, the power
to detect a 50 percent increase of respiratory cancer among workers  was only 14
percent.   Among FG  and  among MW workers exposed for  less  than 10 years,  there
were increasing  SMRs with increasing latency.
     To determine whether  exposure  to  small-diameter fibers produced  in four
different plants was related to an excess in mortality from respiratory cancer,
the  investigators examined respiratory  cancer by plant among  those  ever  and
never exposed.  The only significant increase in  respiratory  cancer occurred
among workers in one plant who were  never exposed to  small-diameter fibers (169
observed, 140.58  expected,  SMR = 120.2, p <0.05).  The small number of deaths
(22) among those  ever  exposed  contributed to the low power and lack of excess
July 1988                          8-21         DRAFT—DO NOT QUOTE OR CITE

-------
 deaths  in this group.  When  those  22  respiratory cancer deaths were further
 categorized  in workers  ever exposed to fine-diameter fibers by both latency and
 length  of exposure,  the  investigators stated  that  the data supported the
 hypothesis  that exposure to  small-diameter fibers is associated with respira-
 tory cancer.   However, there were  too few deaths and  insufficient power  to .
 permit  meaningful  analysis of the data.
      To explore a potential relationship between NMRD (excluding influenza and
 pneumonia),  latency,  and type of fiber produced, the investigators subdivided
 the  data for NMRD by type  of process  and time since first employment.  Among
 those with  20 or more  years  since  first  employment, there was a significant
 excess  from  NMRD in plants that produced fibrous glass filament and fibrous
 glass wool   (93  observed,  65.60 expected, SMR ~  141.8,  p  <0.01),  which was
 attributable to a  significant excess  among those with  20  or  more years  of
 latency in  one plant producing fibrous glass  filament  and fibrous glass wool
 (87  observed,  63.26 expected, SMR  = 137.5, p <0.01).  This was the only plant
 with sufficient power to detect a 50 percent increase in NMRD.   Despite only 15
 percent power,  in  one continuous glass filament plant a  significant excess from
 NMRD was found among those with less  than 20 years since  first exposure  (8
 observed, 2.37 expected,  SMR = 338.1, p  <0.01).  The low power to detect NMRD
 in individual  plants may have  precluded obtaining excess deaths.
      When  further subdividing  latency  into four time  periods  since  initial
 employment,  significant excesses were found in FG plants producing both contin-
 uous filaments and glass wool among those with 10 to 19 (24 observed, 13.97
 expected, SMR  = 171.8, p <0.05) and 20  to 29 years since  first exposure (46
 observed,  28.14 expected,  SMR =163.5,  p <0.01).   Although there does not
 appear  to be an overall pattern of response  for NMRD with increasing latency
"period, with the exception of workers  in plants producing both types  of  FG
 fibers  with  30 or  more  years  since  first  exposure, there was insufficient power
 to detect a  50 percent  increase 1n  NMRD.
      When  the  cohort was  analyzed  by  process and by duration  of  employment
 group,  the  investigators concluded that  a positive  relationship between NMRD
 and  duration of exposure appeared  1n  none of the five processes.   Except  for
 workers in  the less than  10  years' exposure category exposed to both types of
 FG  (69  observed, 44.73 expected,  SMR ~ 154.3, p <0.01), there  was insufficient
 power to detect a 50 percent increase in NMRD  among workers with  either 10 to
 19,  20  to 29,  or 30 or  more years of exposure  to  MMMF.

 July 1988                           8-22         DRAFT—DO  NOT QUOTE OR CITE

-------
     When NMRD  deaths  were analyzed by both latency and  length  of  exposure,
among workers with  20  to 29 years  of exposure and 30 or more years of latency,
there was a  significant excess among FG workers (19 observed,  10.91  expected,
SMR = 174.1,  p  <0.05), but not among MW workers.  However, the power to detect
a 50 percent  Increase  in NMRD  among MW workers  with 20 to 29 years of exposure
and 30 or more years of latency was only 17 percent.
     The investigators  also examined NMRD  In four plants  to  determine whether
exposure to small-diameter fibers  produced in four was related  to  an excess
mortality among those exposed and never exposed.  The only significant increase
in  NMRD occurred among workers in one plant who  were  never exposed  to
small-diameter  fibers  (the same plant  where there was a  significant excess  of
respiratory cancer:   107 observed,  77.19 expected, SMR »  138.6, p <0.01).  This
excess also resulted  in a significant  excess of NMRD  for  all four plants among
those never  exposed (115  observed,  66.15  expected, SMR  = 133.5,  p <0.01).
The small number of deaths  (11) among those ever exposed  contributed to  the  low
power to detect NMRD  in that group.  The small  number of deaths and low power
precluded meaningful analysis of these  data by further categorizing the  11 NMRD
deaths among workers ever exposed  to fine-diameter fibers by both latency and
length of exposure.
     When investigators  examined cumulative  exposure to  respirable fibers and
respiratory cancer  or  NMRD, the small   number of  deaths  precluded meaningful
analysis of  these  data.   In  addition, except for FG  fibers less than
2.14 fibers/cm -months,  however, there was insufficient  power to detect a  50
percent increase in respiratory cancer or NMRD.   Similarly,  the low power in FG
and/or MW plants  precluded meaningful  analysis  of the data.   Therefore, it  is
difficult to conclude with  certainty that no direct relationship exists  between
respiratory cancer or NMRD  and exposure to MMMF.
     The investigators further explored the relationship  between fiber  exposure
and  respiratory disease by examining  average  intensity  of  exposure.   They
stated that a  nonsignificant  dose-response relationship  existed between fiber
exposure and  respiratory cancer for plants producing  continuous glass  fila-
ments.  Once  again,  larger numbers of  deaths would have  made the analysis more
meaningful.
     To  control  for smoking,  investigators  conducted a  case-control  study
nested within the historical prospective mortality study.  All MMMF workers  who
died of  respiratory cancer or NMRD between January 1, 1950,  and December 1,
1982, were cases.   Controls for both causes  of  death were a  5 percent randomly
July 1988                          8-23         DRAFT—DO NOT  QUOTE OR CITE

-------
selected sample of all workers stratified by plant and year of  birth.  Excluded
from controls were  members of the cohort (1) who  died  of respiratory cancer or
NMRD;  (2) who  died of  any cause prior  to  January  1,  1950;  (3) who had an
unknown date of  death;  (4) who died before age 45; (5) who were less than 45
years  of age at  termination of study (December 31, 1982); (6) who terminated
prior  to January 1, 1956; (7) whose termination minus year of birth was less
than 45; (B) whose date of termination was unknown;  and (9) whose date of birth
was  unknown.   Through telephone  interviews, smoking histories were  obtained
from the worker  himself,  1f still alive,  or a knowledgeable informant, ideally
a member of  the  worker's immediate family, for 287 out of 386 (74.4 percent)
respiratory cancer cases, for 150 out of 205 (73.2 percent) NMRD cases,  and for
426 out of 596 (71.5 percent) controls.   Of those  426  controls  for  whom  smoking
histories were available, the percentage who had ever  smoked was similar to the
1976 U.S.  white male population.
     The  investigators  performed  a backward stepwise logistic regression
analysis in which  ages  at termination of study or  at  death and  date of birth
were used as control  or matching variables and smoking  as the possible con-
founding variable.   For  respiratory  cancer cases among FG and  MW workers,
smoking rather  than time-weighted  cumulative  exposure to  respirable glass
fibers was an  important factor in respiratory cancer.  For NMRD  among FG or MW
workers, smoking appeared to be responsible for  NMRD.   However, the goodness  of
fit value of <0.0001 indicates that the logistic model may be inappropriate for
this data set.
     The authors concluded that  conflicting  evidence exists  regarding  the
relationship between  exposure to  MMMF and respiratory disease.  In support of
this  relationship,  increasing  SMRs  were associated with  time since first
exposure, with duration group, and among those ever  exposed to  smaTl fibers.
Against  this relationship,  neither duration of  exposure, a  time-weighted
measure of fiber exposure, nor measure of  intensity  was  related to SMRs for
respiratory cancer.  This lack of a dose-response persisted when smoking habits
were taken into  account.  In support of a relationship between exposure  to  MMMF
and  NMRD, SMRs  for NMRD  are  slightly higher than for respiratory cancer and
there  is some evidence of a dose-response relationship among smokers exposed to
FG  and MW  combined when  smoking  habits  are taken into account.  Against this
relationship, the  low exposure levels to MMMF relative to asbestos raises  tfH
question of  whether a dose-response relationship  is attributable to  MMMF or tc
low  levels of asbestos.

-------
      This study was strengthened by  the  addition of new members of the cohort
 lost to followup in the  previous  studies.   However,  many of the limitations
 that existed in the previous  studies,  such  as  low power in many subcategories
 and length of  employment used as  a surrogate for duration of  exposure,  are
 pertinent here as well.  In this particular study,  because historical  exposure
 data were unavailable,  a  conclusion that  there  
-------
Because of  small  numbers,  women were excluded from analysis.  After  using  a
modified  life-table  technique to obtain  person-years  at  risk of dying, for
computing the  SMRs,  expected deaths were calculated from U.S. age, race, and
sex-specific rates for the years 1945 to 1972 for appropriate 5-year intervals.
By the end of the followup period, December 31, 1972,  111  (26.6 percent) of  the
416 retirees had  died and death certificates were located for 103  (92.7 per-
cent).  Deaths were  coded  by a nosologist according to the 7th Revision of the
ICO.  The mean followup  was  approximately 30 years.  Overall mortality was low
and there was  no  excess  from respiratory cancer mortality,  a manifestation  of
the healthy worker effect.   When the cause of disability was  examined among  127
workers who had retired  early on a disability during  the  same period from the
same six plants, the cause was clearly stated for 115  (90.5  percent) of the  127
disabled workers.   Causes of disability were compared  with those  expected based
on the percent of disability retirements by age and by cause among U.S. males
in all occupations  as reported by SSA.   This comparison  showed no evidence  of
any unusual health  hazards among FG workers.  The  study  was limited by  the
small  numbers  of  deaths, the low power  to detect  lung  cancer and  NMRD,  the  low
trace  rate, and the  restriction  of  mortality observations to those  aged 65  and
above.
     Bayliss et al.  (1976) conducted a  historical prospective mortality study
of 1448 white  males  first employed between January 1, 1940,  and  December 31,
1949,  and who subsequently achieved 5 or more years of employment in FG produc-
tion,  packing, or maintenance of an FG  manufacturing  facility.  The cohort was
assembled from company employment files.   A modified  life-table  technique was
used to obtain person-years  at risk of dying  by  5-year calendar-time periods
and 5-year  age groups  based on  U.S.  white male  age,  calendar-time,  and
cause-specific mortality.  Vital  status was determined through records main-
tained by Federal,  State,  and local government agencies,  including those from
SSA, city and  county health departments, and area funeral homes.   Deaths were
coded by  a  qualified nosologist  according to the  Revision of the  ICD  in effect
at time of  death.  Total airborne dust in the plant averaged 0.3 mg/m  with a
mean concentration of 0.08 respirable fibers/cm .  The collected glass fibers
had a median  diameter of 1.6 urn and a  length of  28 urn.  Smoking  histories were
not provided.
     By the end of the followup period, June 1, 1972,  1072 (74 percent) employ-
ees in the  cohort were alive; 376 (26  percent) had  died.  No workers  were lost

July 1988                          8-26         DRAFT—DO NOT QUOTE OR CITE

-------
to  follow-up,  and death  certificates  were obtained  for  all  376 decedents.
Therefore, there was 100 percent followup and 100 percent trace  rate.
     Whereas overall mortality was  1«ss  than expected, Investigators found a
significant excess  from  NMRD  (19  observed, 10.04 expected, SMR « 189, p <0.05,
Power *  34  percent) and significant deficits from tuberculosis (0  observed,
4.69 expected,  SMR  « 0,  p <0.05) «nd cirrhosis of the liver  (2  observed,  8.93
expected, SMR = 22, p <0.05).   When mortality from  NMRD was examined by time
since first employment,  a  significant excess was observed among those with 10
or more  years  since first  employment (19 observed,  9.51 expected, SMR = 200,
p <0.01).   Despite  the  low power,  the observation  of any excess probably
reflects the investigators' excellent followup and trace  rates.
     To  evaluate  the pathogenicity of  smaller diameter  glass  fibers,  the
investigators performed  a  nested  case-control  study  of malignant and nonmalig-
nant respiratory disease.  They found no association between  cases  of malignant
and nonmalignant  respiratory  disease  and  work  in operations  involving exposure
to SfTiol 1-diameter fibers.
     Morgan et al.  (1984) conducted e historical prospective  mortality  study of
4399 males  who  had worked at least 10  years in FG  production  and  who  were
employed  at  some  time between January 1, 1968,  and December 31,  1977.   The
cohort was  assembled from  company lists of active employees  and from lists of
former employees with 10 or more years of service.   Since the company could net
identify  all former employees  and retirees who had  died  prior  to  January 1,
1968, person-years  for  the cohort accrued from this  date.   These  sources of
data were augmented by  employment  applications, job  histories,  and death
certificates.    Employees'  work  histories  were reviewed to exclude those with
less than 10 years  of employment in  FG production,  those  who left employment
prior to  1968,  and  those not  exposed  to FG.  Job descriptions of interest  were
classified  into  nine exposure categories.  A modified life-table technique was
used to  calculate  SMRs  based on U.S. white  male mortality.  Vital  status  was
determined through  SSA  records.   Deaths were coded according to the 8th Revi-
sion of  the 1CD.   Actual  exposure levels  and smoking  histories  were  not
provided.
     By the end of  the followup period, December 31, 1977, 294 (6.7 percent) of
the 4399 cohort members had  died;  84 (1.9 percent)  were  lost  to  followup.
Death certificates   were  obtained for  all  but  six  (2.1  percent)  of  the
decedents.

.lulv 1988                          B-27         DRAFT—DO  NOT QUOTE OR CITE

-------
     For the entire  cohort,  Investigators  observed significant deficits from
all causes of  death  (SMR « 85, 95 percent  CI  = 76,  96, p <0.05,  Power  = 100
percent) and all  external  causes  of death,  particularly all accidents (includ-
ing motor vehicle accidents) and suicide.   No significantly  increased  SMRs were
found for any  category of mortality.   Even when the mortality experience of
those with  at  least  30 years of observation  and  20 years  of exposure was
examined, no significantly elevated causes  of death were identified.
     This study was limited by concurrent exposure  and  observation periods,  the
small numbers of deaths, and the low power  to detect  cause-specific deaths from
lung cancer, buccal  cavity and pharyngeal  cancer,  laryngeal cancer, and NMRD,
excluding pneumonia.    Although  subdividing  the cohort  into those with at least
30 years  of latency  and 20 years of exposure  is  most likely to result in
detecting the  effect  of  a deleterious  exposure followed  by a long  latency
period,  the small  number of deaths and low power preclude meaningful  analysis
of these data.
     Robinson  et  al.  (1982) conducted an  industrial  hygiene  survey  and  a
historical  prospective  mortality  study of 596  male production workers employed
for at  least 1 year  between January 1, 1940,  and  December  31, 1948,  in a MW
production facility.    The  cohort  was assembled from company lists  of  individu-
als who had ever worked at the plant.   Detailed work  histories, job titles,  and
dates of specific jobs held by each employee were coded.  A  modified  life-table
technique was used to calculate SMRs based  on U.S.  white male  mortality.  Vital
status was determined  through records  maintained by various government agen-
cies, directories,  and other  sources.   Death certificates  were  coded  by a
diagnostician  according  to the  Revision  of the ICD in  effect  at  time  of death
and subsequently converted into corresponding 7th Revision codes.
     By the end  of the followup period,  December 31, 1974,  188 (31.5  percent)
of the  596  cohort members had died;  less than  2 percent were  lost  to  followup.
Four war casualties  were excluded from the cohort.  Death  certificates were
obtained for all but one (0.5 percent)  of the decedents.
     Exposure  was assessed by  job  description  and location,  from personal
samples taken  at  the time of study, and from historical exposure data.  These
data suggested that  the average historical  airborne fiber  concentrations did
not exceed 2.5 fibers/cm3 before  1935  and 1.0 fiber/cm3 after 1935.   All jobs
in the  plant were assigned to one of  eight potential  exposure categories to
assess the extent and  severity of MW' exposure on employee  mortality.

July 1988                          8*28         DRAFT—DO  NOT QUOTE OR CITE

-------
     Overall  mortality was  less  than expected  but  was not  statistically
significant.  Although  no  cause-specific  excesses  or deficits were observed,
except  for  deaths  from overall  mortality  and  heart disease, there was insuffi-
cient power to  detect any excess mortality for the remaining  causes of  death.
When the  cohort was analyzed by latency for digestive cancer, lung cancer, and
NMRD, small numbers of deaths  and low power precluded meaningful analysis.
When the  cohort was analyzed by duration  of employment, however, a significant
excess  of digestive cancer was found among those in the  highest  duration group,
i.e., more  than 20 years  (5 observed, 1.5 expected,  SMR *  333,  p <0.05).  The
investigators concluded that the  results  were  not consistent with those of
other studies.   While  this  study was strengthened by the 27 years without
overlapping  exposure  and  observation periods, the small number  of deaths and
the  low power  to  detect  cause-specific  deaths  hampered  more  meaningful
analysis.
8.1.1.2.2  Epidemiological studies in Canadian workers.   Shannon and colleagues
(19S2,   1984,  1986, 1987)  conducted  a series  of studies  on  the mortality
experience  of current  and former  employees of an Ontario FG wool plant  who  had
worked  for  more than 3 months  between 1955  and 1977.   The maximum followup
period  for the cohort was 30 years.  Jt was found that an excess of lung cancer
was concentrated  in production  workers with the highest exposure, the longest
duration of  exposure,  and the longest time since  first  exposure.   No excess
noncarcinogenic risk was detected.
     Shannon et al.  (1982,  1984)  conducted a historical prospective mortality
study of  2576 current and former male employees  who  had worked  for more than  3
months  in a Canadian plant manufacturing FG wool between January 1,  1955, and
December  31, 1977.  After  the  cohort was assembled from  company employment
records,  vital  status  was  determined through income tax  records  for those
employed between 1960 and 1969, union records, group insurance records,  senior-
ity  lists,  driver's  license  bureaus, Canadian Vital Statistics and  Disease
Registry,  U.S.  Department  of Immigration and  Naturalization records,  and
personal contacts.  A modified life-table technique was applied to person-years
to calculate expected age,  sex, and  calendar-specific deaths  using  the  Ontario
male population.  By the end of the followup period, December 31, 1977,  2414 of
the 2576  cohort members were alive (93.7 percent); 88 (3.4 percent) had died;
74 (2.9 percent)  were lost to followup.   Death certificates were obtained for
all but one  (1 percent), who was classified and used in the analysis as a death

i..i.< looc                          ft-29         DRAFT—DO  NOT QUOTE OR  CITE

-------
from ischemic  heart  disease (IHD).   The Investigators  did not  state  who  coded
the death certificates.
     Historical exposure  data  were  unavailable  for the period -of the  study.
Samples taken  in  1978 and 1979 suggest that optical microscopy counts rarely
exceeded 0.2 fibers/cm.  Mean levels  in most areas were  reported to have been
                        3
less  than  0.1 fiber/cm .   Based on  employment  histories,  workers were
classified  into one  of three exposure  categories:  plant  exposure only (2239);
office exposure only (234); and mixed plant and  office  exposure (103).
     No statistically  significant excesses  in mortality were Identified.  As
might  be  expected from a group of  factory  workers,  there was a significant
deficit from all  causes of death  (88 observed,  113.3 expected, SMR = 77.6, p =
0.02,  Power =  99  percent).  The authors stated  that the deficit persisted when
the subcohort  of  those with 10 years  since first employment was analyzed (56
observed, 75.54 expected,  SMR  = 75.1,  p <0.05).  When  the cohort was analyzed
by job classification among plant-only workers,  investigators found  significant
deficits from  circulatory diseases  (30 observed, 40.74 expected,  SMR =  73.6,
p <0.05,  Power - 89 percent)  and  from  accidents and  violent deaths (13
observed, 22.57 expected,  SMR = 57.5, p - 0.04,  Power = 64 percent).   According
to the  investigators,  deaths from NMRD and  gastrointestinal  (GI) tract cancers
were limited to plant-only workers and were equal to expected values.   However,
there  was only 22 percent power to detect a 50  percent increased  incidence  of
NMRD and 15 percent  power to  detect a  50  percent increased incidence of GI
tract cancers  among plant-only workers.  A significant  deficit of  IHD was found
among  plant-only  workers  (20  observed, 29.92 expected, SMR =  66.8,  p <0.05,
Power  =  79  percent), but  a significant excess  among mixed exposure  workers  (7
observed, 2.94 expected, SMR = 238,  p <0.05, Power = 16 percent).  The investi-
gators concluded that three suspicious cancer deaths,  upon pathological  review,
were not attributable to mesothelioma.
     When lung cancer  mortality  among plant-only workers  was analyzed by
duration of exposure (<5  years or >5 years) and time since  first exposure (<10
years  or >10 years), too  few  deaths and low power precluded meaningful  analy-
sis.   Only  502 men  (19 percent) had  accrued 20 or more years since first
employment.
     While  the overall mortality of these workers appears to be low, this study
is  limited  by  the following:    (1)  the  small number of deaths in the cohort
(88);  (2) the  lack  of direct  exposure and  the  unavailability  of quantitative
exposure measurements  to  FG wool  fibers (exposure  to  FG was only inferred);
July 1988                          8-30          DRAFT--DO NOT QUOTE OR CITE

-------
(3) among workers classified as in-plant, the power to detect  lung  and Gl tract
cancer were  only 19 and 29  percent,  respectively;  (4) the  exclusion of 363
workers  employed  between 1948  and  1955 for  whom  records were incomplete,
potentially  underestimating  risk;  and  (5) observation and  exposure years
overlapped, thereby underestimating risk.
     Shannon et al.  (1986, 1987)  extended the followup of their previous study
for 7 years  by conducting  a  historical  prospective mortality study of 2557 men
who had  worked at least 3 months in an Ontario  FG  wool plant  between 1955  and
1977.   The  cohort was  reduced from 2576 in the previous study  to 2557 by
excluding women and  some  men who were counted twice.   No  deaths were observed
among the 19 excluded  individuals.   Since historical  data on exposure  levels
were unavailable,  exposure was  inferred by categorizing workers  as plant
exposure only,  office  exposure  only,  or mixed plant and office exposure.
According to the investigators,  hygiene measurements taken since 1978 have
shown fiber  levels rarely greater  than 0.2 fiber/cm   (<3.5 diameter)  using
optical   microscopy  counts.   Mean fiber  levels  in most  areas  were below
0.1 fiber/cm ,  but anecdotal  evidence  suggests  that dust concentrations were
previously higher (Shannon et al., 1987).
     The Ontario  male  cause-specific  mortality  rates adjusted for  age and
calendar period were applied to person-years to  calculate  expected deaths.   The
investigators  stated  that both county  and  provincial  age-adjusted mortality
rates for  lung cancer were  similar to the Canadian  national rates  for the
period 1966 to 1976.   Therefore,  the use of  Ontario  cause-specific  mortality
rates would  not  underestimate  the  risk of  lung cancer in  the plant-  only
workers.
     Vital status was  determined  through  the Canadian Vital Statistics and
Disease Registry and by linkage to the Ontario driver's license records.  Years
of observation  accrued  from  January 1, 1955, to December 31, 1984, allowing a
maximum  followup  of  30  years, only 7 of which did  not overlap exposure  years.
By the end  of  the followup period, December  31, 1984,  2312  of the  2557 (90.4
percent) cohort members  were alive; 157 (6.1 percent)  had  died;  88  (3.4 per-
cent) were lost to followup.   Death certificates were obtained for ell  but one
(1.1 percent),  which was categorized as death from IHD.
     Although  a significant deficit  in overall mortality  was observed,  a
manifestation  of  the healthy worker effect in this group of factory workers
(157 observed,  187.54  expected, SMR =  84,  p  <0.05),  there  was a significant

July 1988                          8-31         DRAFT—DO NOT QUOTE  OR  CITE

-------
excess of  lung  cancer (21 observed, 11.90 expected, SMR = 176, p <0.05).   Nc
cases  of mesothelioma  were  reported.   However,  because  no cause-specific
expected values were  reported  for mesothelioma, it was  not  possible to  deter-
mine whether  there  were enough exposed workers to detect  an increase  in meso-
thelioma incidence of 50 percent above that of Ontario males.
     When  the data  were analyzed by job classification, plant-only workers,
presumably those with the highest exposure, appeared to  account for the  signif-
icant excess  of  lung  cancer, representing  all  but 2 of the overall lung cancer
cases  (19  observed, 9.53  expected,  SMR =  199,  95 percent  Cl = 128, 311,
p =  0.002,  Power  =  36 percent).   Except for all cancers, accidental or violent
death, and  circulatory  disease,  especially IHD, among plant-only workers,  the
power was  too low to  detect site-specific causes of death of  interest.   When
the  lung cancer deaths  were  analyzed by duration of exposure  (<5 years or >$
years) and time since first employment (<10 years or >10 years), too few deaths
and  low  power precluded  meaningful  analysis  in the 1986  unpublished study.
However, in the 1987 published version, there  was an additional death  from  lung
cancer, raising the SMR above  the 0.05 level  of significance.   Thus,  contrary
to the  claims by  Shannon et al.  (1986, 1987),  a  significant excess  of lung
cancer was  found  in workers  with the longest  duration of  exposure  and longest
time  since  first  exposure (13 observed, 7.12  expected, SMR = 182, p <0.05,
Power s 28  percent).   When the lung cancer deaths were analyzed by length of
employment  in 5-year  increments  or by time since first employment in 5-year
increments, the numbers  of death again were too small to  analyze.  It was  not
possible to determine the percent of the cohort with  20 or more years of
exposure and  30 or  more years of observation  following  initial exposure to FG
wool.
     The investigators  concluded the  following:  (1) although  a  significant
excess of  lung  cancer was observed, SMRs  by length of exposure and time since
first employment  were  not consistent with exposure to FG  wool; and (2)  deaths
from NMRD  were  less than expected.  These conclusions are weakened by several
limitations that  also  existed  in  the previous  study, I.e.,  low  power to detect
site-specific cancers  and NMRD (30  percent for  lung cancer, 33 percent for
digestive system cancer, and 11 percent for NMRD), small cohort size,  and small
percentage of the cohort that died (6 percent).
8.1.1.2.3  Epidemiolooical studies in European workers.   Saracci and  colleagues
(1984) and  subsequently  Simonato  and colleagues (1986a,b,c,1987) conducted a :
series of  studies on  the morbidity and mortality  experience  of 21,967 male and
Julv 1988                          8"32         DRAFT--DO  NOT QUOTE OR CITE

-------
female production workers from 13 European plants  that  produced MW, FG wool, or
continuous FG  filament  at  some time between 1933 and 1961.  Maximum followup
varied among the  plants.   MW workers had  excesses of lung cancer, particularly
if first employed during the early technological phase.  This  same group had an
elevated Incidence  of  buccal  cavity/pharyngeal cancer.   FG wool workers had an
excess of  lung cancer  limited to those first employed  during  the intermediate
technological  phase, but only when compared to national nortality rates.   FG
wool workers also had  an excess of laryngeal cancer, but this was limited to
male production workers  in an Italian FG wool plant.   An  excess of NMRD was
limited to short-term MMMF workers.
     The following  section  begins  with critical reviews of each European study
and culminates in critical  reviews of the collated results  presented by Saracci
and colleagues (1984) and Simonato and colleagues  (1986a, 1986b,  1986c, 1987).
     Andersen  and  Langmark  (1986)  conducted  a historical prospective mortality
and morbidity  study of  2361 men who had worked at any  time between January 1,
1935, and  December  31,  1977, in one of four Norwegian  plants  that manufacture
either mineral  rock wool  (MW; Plants 1,  2,  and 3) or  fibrous glass wool (FG;
Plant 4) and who were alive on January 1, 1953.  The  investigators excluded 20£
women, 25  men  for whom essential  information (date  of birth  and/or  date  of
first employment) was  missing,  as  well as those  ever employed in polystyrene
production.  The  cohort  was assembled  from employment  records from each of the
four plants containing  names, dates of birth, all periods of employment,  and
months per worksite within each plant.   Vital status end cancer morbidity  were
determined through  a countrywide  automated  vital statistics  system  and the
Norwegian  Cancer  Registry.   A modified  life-table procedure was used to calcu-
late person-years  at risk,  expected numbers  of deaths,  and expected numbers of
cancer cases.   The  Norwegian national  age-specific  mortality and incidence
rates for 5-year time periods from 1953 to 1983 were  applied to person-years  to
obtain the expected number of deaths or  cancer cases.  The years of followup
were from January 1, 1953, to December 31, 1983,  allowing  a maximum followup  of
31 years.  Since  exposure  occurred between 1935  and 1977, there were 6 years
without overlapping exposure and observation periods.   Neither the degree of
completeness of ascertainment of cause of death or morbidity nor the  percent  of
workers  lost  to  followup  was reported.   Because  all death certificates are
coded by the  Norwegian Central  Bureau  of Statistics  and because  reporting  from
hospitals  and  histopathological  labs is  compulsory,  presumably the potential

July 1988                          8-33         DRAFT—DO NOT QUOTE  OR CITE

-------
 bias  from lack  of  disease ascertainment among workers lost to  followup  is
 minimal.  Actual exposure  levels were not reported.
     One  of three MW plants  employed 1815  workers (77 percent), and the FG
 plant  employed 546  workers (23 percent).  Only 49 percent  of  the cohort  was
 employed  for 1 or more years.  Overall,  • significant  excess of  death occurred
 from  accidents,  poisoning, and violence (37 observed, 24.58 expected,  SMR =
 151,  p <0.05), which was  also found among those  with less than 1 year  of
 employment  (23 observed,  10.04  expected,  SMR = 299, p <0.01).  Although no
 excess NMRD  deaths occurred,  the power to detect NMRD among those employed less
 than 1 year, 1 or more years, or ever employed was low.
     When  cancer incidence  data were analyzed,  a significantly  increased
 incidence of buccal cavity and pharyngeal cancer was found among employees from
 all factories  combined  with  1 or more years  of employment (7  observed, 3.00
 expected, Standardized  Incidence  Ratio  (SIR)  = 233, p <0.05,  Power = 17  per-
 cent).   While  the power to detect  all  cases  of cancer was sufficient  among
 those  with  1 or more years of employment (99  percent)  and for  the total cohort
 (100 percent), there was insufficient power to detect *11  cases of cancer among
 those employed less than 1 year (78 percent) and lung cancer among those in any
 category.
     When cancer  incidence data among those with  1 or  more  years of employment
 were analyzed  by years  since first employment  and period of first employment
 (either  before  1960  or I960  to 1977), a significantly increased incidence of
 intestinal  cancer was  found when both periods  of  MW production  were combined
 (14  observed,  7.14  expected, SIR = 196, p  <0.05).  When  the 20  lung cancer
 cases  were  analyzed  by length of employment  (less than 1 year or 1 or more
^years) and by plant, in Plant 2 (rock wool) a significantly increased incidence
 of  lung  cancer was  found among those ever employed (8  observed,  2.68 expected,
 SIR =  299,  p <0.05,  Power = 11 percent), which was more pronounced among  those
 with  1 or more years  of  employment (8  observed,  2.07 expected,  SIR *  385,
 p <0.01, Power = 10 percent).  Comparing lung cancer incidence  among MW produc-
 tion and nonproduction  employees,  a significantly elevated incidence of  lung
 cancer occurred  among  production and nonproduction employees combined who had
 20  or  more  years since first exposure (11 observed, 5.52  expected, SIR = 199,
 p <0.05,  Power  = 21 percent), which was concentrated among those employed for
 1 or  more years (9 observed, 4.36  expected,  SIR = 206,  p <0.05,  Power = 21
 percent).   According  to the  authors,  the use  of  county rather than national

 Julv 1988                          8-34         DRAFT—DO NOT QUOTE OR CITE

-------
rates  increased  the expected  lung  cancer  incidence,  thereby decreasing  the
SIRs.   Nevertheless,  the trends  1n SIRs persisted,  but  the data  were  not
reported.
     The  Investigators  concluded  the following:   (1)  no consistency occurred
across  the  four plants  in  cancer  morbidity  or  nortality;  the only excess
mortality was  from  violent  deaths among  those with less than 1 year of employ-
ment;  (2) because the  significantly elevated incidence of buccal cavity and
pharyngeal cancer was limited to one plant  and because neither  human nor  animal
evidence suggests involvement  of  this site and MMMF,   it appears that factors
other  than  MMMF are responsible; (3) the  increased incidence of intestinal
cancer  is  attributable to  random variation;  (4) because  the increased lung
cancer  incidence was  limited to  one plant, agents  other  than  MMMF might be
responsible;  (5) no association  exists  between  lung  cancer incidence  and
duration of employment  or type of work;  and (6)  the risk of  lung cancer proba-
bly is related to more hazardous environmental conditions  present before  1960.
     However,  these conclusions are tempered  by  the  low power to detect NMRD
mortality and  lung  cancer mortality and morbidity for  the entire cohort, the
small  cohort,  the small  number of deaths,  and the short exposure period  (over
half of  the cohort  was employed for less than 1  year).  Although the investi-
gators  stated  that  no  positive association existed between  the  incidence  of
lung cancer and  duration of employment, they did find a significant excess  of
lung cancer morbidity  among those with 20  or more years since first exposure,
even though the  power  to detect  lung cancer for  the entire  cohort  was only  21
percent.  It was not possible to determine  the percent of the cohort with 20 or
more years of  exposure and  30 or more years  of  observation  following initial
exposure to  MMMF.   In  view  of these deficiencies, demonstration of  a dose-
response relationship between MMMF and lung cancer or  NMRD would  be  difficult.
     The authors'  suggestion  that  agents  other  than  MMMF  might have been
responsible for  the increased lung cancer  incidence  in one  plant is  not sub-
stantiated in this study.  The investigators made no mention of the presence of
other  chemicals  associated  with  lung  cancer,  and the increased lung cancer
observed  in  this plant  could have  resulted  from higher exposures to MMMF
compared to the  other  three plants.  Concerning  the investigators'  claim that
there  is  an absence of human evidence to suggest involvement of  buccal  cavity
and pharyngeal  cancer  and MMMF, aside from this  study, significant  excesses  of
buccal  cavity  and  pharyngeal  cancer were  reported by Moulin et al.  (1986),
1,,1,, 1OOO

-------
Olsen  et  al.  (1986), and  Simonato  et al.  (1986b,c, 1987).  Therefore,  the
authors'  conclusion  is  weakened by the finding of  statistically  significant
excesses from this site in three other epidemiological  studies.
     Westerholm and  Bolander  (1986) conducted a historical prospective mortal-
ity and morbidity  study of 3600 male  and female (20 percent) employees who had
worked for  at  least  1 year In one of  two MW plants or one FG wool plant.  From
personnel records  available  from the  start of production for each of the three
plants, 1933, 1938,  or  1943,  working  groups consisting of employer representa-
tives, unions,  and former employees  extracted  the  names,  Swedish  identification
numbers, and work  histories  of the  cohort.   Past or present exposure  levels
were not  discussed.   Expected numbers of death were based on mortality statis-
tics for  all Sweden  by age,  sex, and cause of  death  for each  calendar year.
Expected numbers of cancer incidence were based  on rates  for all  Sweden by  age
sex, cancer diagnosis, and calendar  year.
     Vital  status  was determined through record linkage of the  computerized
data bases from which dates and causes of death  were coded according  to the 1CD
in effect at the time of death were  extracted.   Cancer morbidity  was  determined
through the  Swedish   Cancer  Registry, whose  compulsory  notification system
requires Swedish identification number,  name,  diagnosis, site  and histopath-
ology according to the  7th Revision  of the ICD.  The  years of exposure and
observation were concurrent.
     By the end of the followup period, December 31, 1982,  3014 of the  3600 (84
percent) of  the  cohort  members were alive;  524  (15  percent) had  died;  62 (1.7
percent) were lost to followup.  Of  the 230  (6.4 percent) who had emigrated,
147 (64 percent) were successfully  traced through the vital statistics regis-
tries of other Nordic countries.  It is unclear,  however, whether the remaining
83 emigrants accrued person-years  until  their dates of termination or  whether-
they were totally  excluded from the analysis.   Adding the 83 emigrants to  the
62 other persons lost to followup,  it appears  that 145 (4 percent) were lost to
followup.
     Overall,  a significant  excess  of accidents,  poisoning,  and violence
occurred among  MW  workers  (35 observed, 25.7 expected,  SMR =  136, p <0.05).
Since the cohort was too small to detect a 50 percent increase in lung cancer
(Power =  61 percent) or NMRD (Power = 76 percent),  subdivision of the 18 lung
cancer deaths and  20  NMRD deaths failed to yield useful information.   Neverthe-
less, when lung cancer mortality data were analyzed by length of employment and

Julv IQRR                          8-36         DRAFT—DO NOT QUOTE OR CITE

-------
 years since first exposure, an excess  of lung  cancers occurred among those with
 1 to A  years  of employment and 30 or wore years  since  first exposure in MW
 production (4 observed, 0.5  expected,  SMR * 888,  p <0.005, Power = 4 percent)
 or in FG  production  (4 observed,  1.2  «xpected,  SMR = 345, p  <0.05, Power = 11
 percent).   It was  not  possible to determine the percent of the cohort with 20
 or nore years of exposure  and 30 or more years of  observation following Initial
 exposure to MMMF.
      The  investigators concluded  that CD both FG and MW workers who developed
 lung cancer were exposed  for short periods  of time, i.e., during the first 4
 years of  employment; (2)  the  observed excess  cancer deaths nay be attributable
 to agents other  than MMMF; and (3) insufficient evidence exists to conclude
 that exposure to MMMF  contributed to the  lung cancers observed in this cohort.
 This study is limited  by  the low power to detect  lung cancer and NMRD and the
 small number of  lung cancer  and NMRD deaths,  precluding meaningful analysis.
 Although a high  degree of followup was accomplished among Swedish nationals,
 failure  to determine the  vital status of 36  percent of  the  emigrants could
 result in  an underestimation of. the risk of  lung cancer and NMRD.   Although  the
 investigators speculated  that  agents other than MMMF might have been responsi-
 ble for the  increased  lung cancer deaths, the investigators made no mention of
 the presence  of other  chemicals  associated  with lung cancer;  furthermore,
 historical exposure data  were not available.   Therefore, the increased  lung
 cancer mortality might  have resulted from MMMF exposure.
      Olsen and Jensen (1984) conducted a historical  prospective  morbidity study
 of 53ES male  and female employees (8 percent) who had worked in a Danish MW
 plant at  any  time  between the start of  production in 1937  and  December 31,
 1977, allowing  a maximum  exposure to MW  of up  to 40 years.   After using a
»
 jnodified  life-table  technique to  obtain person-years at risk,  the expected
 numbers of cancer  cases were  derived  from Danish  national age, sex, and calen-
 dar time-specific  cancer  incidence rates for  5-year time intervals.  Through
 computerized record linkage with  national and municipal  files of  the Danish
 Cancer Registry, persons 1n the cohort were  followed  from date of first employ-
 ment to date  of death, emigration, or the end of  the foTlowup period (December
 31, 1977).  Because the Danish Cancer Registry  began reporting  incidence data
 in 1943,  the years of  observation were  from  1943 to  1977.   Although  maximum
 followup was 35  years, exposure and observation  periods were concurrent.  Of
 the 5369  cohort  members,  174 (3.2 percent)  were lost to followup.   Since many

 .I'.iw IQRR                           8-37         DRAFT--DO NOT QUOTE OR CITE

-------
of these workers  were not Danish nationals, their vital  status  could  not  be
determined and, therefore, they were excluded from analysis.
     The concentration of  respirable  fibers in the plant at the  time  of the
study  was  between 0.1 and 0.001 fibers/cm3.  Although  previous  measurements
were unavailable,  1t  was  estimated  that the concentration of respirable fibers
at the production lines 1n the past was between 9.8 and 26.5  fibers/cm3.
     Overall, no  statistically  significant  excess  cancer cases  (144  observed,
136  expected,  SIR -  108)  were reported.   However, when cancer cases  among
production employees  were  analyzed  by site and time since first  employment,
cancer of the digestive organs and peritoneum was  significantly  elevated during
the  first  10 years after  initial   employment  (21  observed,  13.8 expected,
SIR = 152,  p <0.05).   Among those with 20  or more  years since first employment,
there was significantly  increased  incidence of lung cancer  (9  observed, 4.3
expected, SIR =  209,  p = 0.02, Power = 20 percent).  When lung  cancer  and  all
cancer cases  among those with 20 or  more  years since  first employment were
analyzed by category of employment (male,  female,  production, and nonproductior
workers), a  significantly  increased incidence of  lung cancer was found among
male production workers (8 observed, 4.1 expected,  SIR  = 195, p  = 0.05, Power  =
17 percent)  and  among all  employees (10 observed, 4.8  expected, SIR =  208, p =
0.02, Power = 19 percent).   When cancer cases among those with 20 or  more year;
since first employment were analyzed by duration of employment,  no relationship
was  found between  cancer  incidence  and duration of employment.   However, this
finding is  not surprising because (1) there was only 38 percent  power to detect
a 50 percent  increased incidence of lung  cancer among  workers with 20  or more
years of latency;  (2) only 25 percent of  the cohort was employed before 1960,
i.e., had 18  or  more  years of  latency; and (3) only 25  percent of the cohort
had  been employed  for more than 2  years.    These observations indicate a high
turnover rate among employees and a short  duration of  exposure to MW,  suggest-
ing  that there might  have been insufficient exposure or followup to  detect an
excess risk  of  lung cancer.   One case of mesothelioma  was observed in  a person
13 years  after first  employment.   Because expected values  for mesothelioma
cases were not  reported,  it was not possible to ascertain the power  to detect
the  risk of mesothelioma in this cohort.
     To control for confounding variables, the investigators found that the us«
of local lung  cancer  incidence rates  increased  SIRs  by 8.3 percent.  They  also
determined that  SMR  for  respiratory system  cancer  among unskilled  workers

-------
compared to Danish males  was  less than the SIR for respiratory system cancer
for male workers  in  this  analysis.   Finally,  in the absence of smoking histo-
ries, the  investigators surveyed  smoking  habits at the time of the study and
found 10 percent  more  smokers but 7 percent less heavy smokers compared to the
general population of eastern Denmark.
     The investigators concluded  that  (1) a  twofold increase  In lung cancer
Incidence  among MW workers  with 20  years since first employment supports the
hypothesis that  an  association exists between lung cancer  and  mineral  wool
production when the  data  are  analyzed'by latency period; (2) It is unlikely
that this  association  could be  ascribed  to either disease ascertainment bias
because of the adequate degree  of followup or to  smoking; and (3)  the  lack  of
a dose-response relationship requires additional followup to increase the evi-
dence of causality.
     To further study  the relationship between occupational exposure to MW and
lung cancer, 01 sen et  al. (1986)  extended the followup of their 1984 study for
an additional 5 years  (December 31, 1982).  The cohort of 5317 male and female
(8 percent) employees  was assembled  from manual files  at the  plant containing
information on dates of initiation  and termination of employment  and type of
position.   The 52 individuals  included  in the 1984 study who  were  on  file but
had never  worked were deleted  from the  current cohort.   In addition, three
alternative methods  for calculating  expected  lung cancer cases among male MW
workers were used:   (1) age-specific local  rates;  (2)  weighted  rates  based  on
place of  residence  (urban/rural) at time of  first employment;  and (3) age-
specific smoking-induced  lung cancer risks  for men aged 30 to  59  to  account
for tobacco consumption.   Through computerized record  linkage  with  national  and
municipal files of the Danish Cancer Registry, analysis of cancer morbidity  was
for the period 1943 to 1982, allowing a maximum of 40  years  of observation,  but
only 5 years without overlapping exposure and observation periods.   Of the 5317
cohort  members,  173  (3.3  percent) individuals who were  non-Danish nationals
were lost to followup.
     As  in the previous  study, no statistically  significant excess  cancer
Incidence was observed among  the  total cohort (240 observed, 211.0 expected,
SIR = 114,  95  percent  CI  ~  100, 129).  When the cohort was  analyzed by sex and
job  classification  (production or  office  staff), however,  a  significant
increase in  total cancer incidence  was  found among production workers  (222
observed,  191.9,  SIR = 116, p <0.05, Power = 100  percent), which appears as a
result  of  the  increase among  male production workers  (213 observed,  182.0
July 1988                          8-39         DRAFT—DO NOT QUOTE OR CITE

-------
expected, SIR =  117,  95 percent CI = 102,  134,  p <0.05,  Power = 100 percent).
When investigators analyzed  site-specific  cancer incidence among male workers
by time  since  first employment,  they found a significantly increased incidence
of buccal cavity and  pharyngeal  cancers (13 observed, 6.3 expected, SIR = 210,
95 percent  Cl =  110,  340, p <0.05, Power * 24 percent) and of digestive organ
cancers  (64 observed, 43.9 expected, SIR * 150,  95  percent  CI  * 110, 190,
p <0.05, Power s 90 percent),  but no trends of  Increasing SIRs occurred with
increasing  time  since first  employment.  Although no significant elevations of
lung cancer incidence were found among males with 20 or more years since first
employment,  the  power to  detect  lung cancer In this group was only 53 percent.
In addition, the power to detect mesothelioma  cases could not be calculated
since  cause-specific  expected incidence was  not reported  for  this site.
Finally, even when adjustments to expected lung cancer incidence rates were
made to  account  for geographical differences  in local and national rates, in
urban and rural  places  of residence, and  in regional tobacco consumption, no
significant  elevations in lung cancer SIRs  were  identified.
     The investigators  concluded the following:   (1)  the overall  risk of  lung
cancer  appears restricted to the largest  subgroup, male  production workers;
(2) there was an increasing  trend with time since first employment for cancers
of the buccal cavity and pharynx  and respiratory organs,  but  a decreasing  trend
for cancer  of the digestive  organs;  (3) there is a moderately elevated risk of
cancer of the respiratory tract, especially lung cancer  and  buccal  cavity and
pharyngeal   cancer in  workers with 20 or more years  since first employment;
(4) adjustments for local lung cancer incidence,  place of residence, or smoking
habits did  not result in significantly  elevated lung cancer incidence among
male production  workers;  and (5) although  it  is unlikely that the  lung cancer
excess is attributable  to disease ascertainment bias, life style, or occupa-
tional exposures  unrelated to  the plant,  the  number of cases was  too small  to
exclude the effects of chance.
     Although this  study indicates  an overall  risk of lung cancer in male
production workers, contrary to the conclusion of Olsen et al.  (1986), when the
data are analyzed by latency  period neither significantly  increasing  nor
decreasing  trends in lung  cancer incidence are identified.   Furthermore,
although the excess lung  cancer  incidence among those with  20  or more years
since first employment  was  significantly  elevated in the previous  study, the
lung cancer incidence was reduced in this study.   It was not possible  to

July 1988                          8-40         DRAFT—DO NOT QUOTE OR CITE

-------
determine the  percent  of the cohort with 20 or more years  of  exposure  because
the data  were  not analyzed by  duration group.  Despite the limitations of this
study, such  as the short exposure period, the small number of cases, and the
small percentage  of  the  cohort with 30 or wore years of  latency,  an  excess  of
lung cancer was found in KMMF workers.
     Teppo and Kojonen  (1986)  conducted  •  historical prospective mortality and
morbidity study of 941 male (85 percent) and  female (35 percent) workers who
were employed  at least 3 months in a Finnish FG plant between  1941 and  1951  and
were alive on  January  1, 1953,  as well as  all  those employed at least 3 months
between January 1,  1953, and December 31, 1977.  A list of workers assembled
from company employment records was compared and supplemented  with a  membership
list from a  local  trade  union.  Job  histories  were then extracted from employ-
ment records.   If  exposure periods  were discontinuous, then the  duration  of
exposure  was  defined as  the sum of  an  individual's exposure periods.  In
addition  to  the presence of FG wool, asbestos  had been used in both  insulation
and personal protective  equipment  as well as  in production from  1946  through
1948.   Actual exposure measurements were unavailable.
     The  Finnish  national  age,  sex,  and  calendar-time  specific mortality rates
from  the Central  Statistical  Office and the  national  age,  sex,  and
calendar-time  specific  cancer  incidence  rates from the Cancer  Registry  were
applied  to  person-years to  obtain the expected  numbers  of deaths or  cancer
cases.   The  years  of followup  were  from  January  1,  1953, to December 31, 1981,
allowing  a maximum followup of 29 years,  only 4 of which  did not overlap the
exposure  period.   Of the 941 cohort members,  13  (1.4  percent)  were lost to
followup, mainly because they emigrated to Sweden.
     Although  no  significant excesses from specific causes  of death were
observed, only 133 of  the 941 (14 percent) cohort members  had died.   Further-
more, among  male  workers, the  power  to detect all malignant neoplasms  and NMRD
was only  47  percent and 26 percent,  respectively.   There were no  trends in  the
SHRs with either  length of exposure or  time  since  first exposure.   When the
data were analyzed by  length of employment,  among males and females  with 20  or
more years  of exposure,  the power to detect  all  cancer deaths was only 2
percent and  to detect  all cancer incidence was only 6 percent.   When the data
were analyzed  by years since first exposure, among males and females with 30 or
more years since first exposure, the power to detect all  cancer deaths was  only
62  percent  and to  detect all  cancer  incidence  was 28 percent.  No  excess
cause-specific cancer morbidity was identified.  No mesotheliomas were reported.
July 1988                          8-41         DRAFT—DO NOT QUOTE OR CITE

-------
Since  local  lung  cancer incidence approximated national rates and the urban/
rural  distribution  of  the  cohort  corresponded  with  the distribution of the
Finnish population, neither affected the SIRs.
     The investigators  concluded  that  the  results argue against an etiological
role between  lung cancer and MMMF exposure for  several  reasons:  (1) mortality
analysis did not reveal any statistically significant risks;  (2)  no excess  risk
of lung cancer  was  identified (three of the four observed cases occurred among
those  who worked  less than 1 year); and (3) no dose-response relationship or
Increasing  relative  risk with  increasing  years of  followup  was  found.  The
investigators correctly noted that the lack of  positive evidence was attribut-
able to the small cohort and small number of deaths.   In fact,  only 1.8 percent
of the person-years  at risk had accrued among  those with 30 or more years of
followup;  mean  followup was 12 years.   When a  cohort  is  small,  they pointed
out,  mortality and cancer morbidity will be too small to draw conclusions about
the presence or absence of an excess risk.
     Claude  and  Frentzel-Beyme  (1984) conducted a  historical prospective
mortality study of  2096 male production and maintenance employees who  had  ever
worked in a  West  German MW plant  between  January 1,  1942,  and December 31,
1977.  After  the  cohort was assembled  from company payroll records,  a  modified
life-table technique  was  used to obtain person-years at risk of  dying, begin-
ning  with  date  of  first employment or January 1, 1942, whichever occurred
later.   Individuals with  unknown  vital status  contributed to person-years at
risk  until  the  date  of last  information.   Those for  whom  death  certificates
were  unavailable  were assumed to  have  died from unknown causes.  For an  exter-
nal  comparison  group,  the  West German national age, sex, and calendar-time
specific mortality rates for 5-year age groups  and 5-year calendar-time periods
were  applied  to person-years to obtain the expected numbers  of cause-specific
deaths.  For  an internal  comparison group, the number of deaths  stratified by
10-year age  groups  among 1,77 male cork and styropor panel production workers
of the same plant were compared with deaths among  the study group.
     Although a steam-blown rock wool  manufacturing  process was in  operation  at
the plant since 1942, measurements of  respirable fibers in  production areas
were  unavailable.   In  addition,  because this  operation was  discontinued  in
1977,  no current  exposure data were available.  Consequently, the cohort  was
classified by main  job category as a  surrogate for  degree of  exposure to  MW.
Workers were  possibly  exposed  to asbestos  during the  manufacture of  other

July 1988                          8-42         DRAFT--DO NOT QUOTE OR CITE

-------
products within the  plant,  as well as the presence of other potentially con-
founding substances, including silica, formaldehyde,  oil binders, and coal tar.
Exposure occurred between 1942  and 1977, allowing a maximum exposure period of
36 years.   Since  the observation period was  from January  1, 1942, to December
31, 1979,  allowing  a maximum followup of 38 years, there were  only 2  years
without overlapping exposure and observation  periods.
     Vital  status was  determined through personal  contacts and  through  inqui-
ries made  at  registrar's offices regarding a worker's last residence or birth-
place.  Death  certificates  were obtained from local government health offices,
but ascertainment of cause of death was impeded  by recordkeeping laws requiring
the maintenance of  death  certificates for only 5 to 10 years.  By the end of
the followup period,  December 31, 1979, 1655 (79 percent) of the 2096  study
group members  were  alive;  242 (11.5 percent) had died;  199  (9.5 percent)  were
lost to followup.   Among  the 1778 internal  controls,  1372 (77.2 percent)  were
alive; 219  (12.3  percent)  had died; 187 (10.5 percent)  were  lost to followup.
According to the investigators, missing information on 7 percent of  the  deaths,
62 percent  of which  occurred in  migrant workers,  may have underestimated
cause-specific cancer mortality.
     Although  no  significantly  increased overall or cause-specific mortality
was found  in  the  study group when compared to West German males, 43 percent  of
the cohort  were  exposed for less than 1 year;  only 3 percent were  exposed for
30 or more  years.   There  was insufficient power  to detect either lung cancer
(46 percent) or NMRD (52 percent).  When the  exposed production  and  maintenance
workers were  analyzed  by  time since first employment, a significant excess  of
malignant  neoplasms  was found  among  those with 20 to 29 years  since  first
employment  (27  observed,  17.5 expected, SMR =  154, 95  percent  CI = 104,  221,
p <0.05).   When  the same  analysis was  restricted to production workers,  a
significant excess  of  stomach cancer was observed among  those  with 20  to 29
years since first employment (5 observed, 1.8 expected, SMR * 273,  95  percent
CI * 102, 616, p <0.05).  When restricted to maintenance workers, n  significant
excess from all causes of death was found among those with 20 to 29  years  since
first employment  (37 observed,  25.3 expected, SMR « 146, 95 percent CI  =  106,
200, p <0.05).  When internal controls were used as the reference  population,
investigators found  a  significantly elevated relative risk (RR) of  neoplasms of
unspecified nature  (RR =  7.03, 95  percent  CI  = 1.17,  42.29,  p <0.05) and
pancreatic  cancer (5  cases,  0  controls, p  <0.05).   No mesotheliomas  were

July 1988                          8-43         DRAFT—DO NOT QUOTE OR CITE

-------
observed In the  study  groups;  however, one mesothelioma occurred in the inter-
nal control group.   Using  one  as an expected value,  the power to detect meso-
theliomas was  only 7 percent.
     The  investigators concluded that (1) no  significantly increased risk
exists for any  cause of death among persons engaged in rock wool production,
packaging, and  maintenance activities compared  to  the general population;
(2) using job  classification  as a  surrogate for Intensity  of  exposure,  a
significant excess  of stomach cancer existed  among  production  workers;  and
(3) interpretation of positive findings  is complicated  by the presence  of other
substances,  such as  asbestos,  silica, formaldehyde, oil binders, and coal  tar,
the short exposure period,  and the  potential  disease  ascertainment  bias.  Thus,
the adverse effects  identified in  this  study may not  be causally  related  to
exposure to MMMF.
     This study has several limitations.   While the power to detect deaths.from
all causes  was  sufficient (100  percent),  the  cohort was  too small to have
sufficient power to  detect either  lung cancer  or NMRD.  More important, incom-
plete enumeration of  the  causes  of death in  the study  group, especially among
migrants, may have  underestimated  the  cancer mortality risk in this  cohort,
especially for  rare  tumors.  Coupled witti the  laro/e percentage of workers  lost
to followup (10  percent),  the  potential  disease ascertainment bias may have
greatly  underestimated the risk  of lung cancer and  NMRD in this study.    In
addition, because the  majority of  workers were exposed for  less than  1 year
(median duration = 1.6 years),  there .nay  have beeri too  short an  exposure period
for this  small  a cohort to demonstrate a dose-response relationship  between
MMMF and either lung cancer or NMRD.
     To  increase the  size  of  the cohort  with 30 or more years of observation,
Claude and Frentzel-Beyme  (1986) conducted a historical prospective.mortality
study of  2092 male  production  and  maintenance  employees who had ever worked  in
a West German MW plant between January 10,  1941, and  December  31, 1977.   An
investigation into  past exposure by the Edinburgh Institute of  Occupational
Medicine between 1983  and  1984 revealed  that production began in 1941.  There-
fore, the exposure period and person-years at risk were calculated  from date  of
first employment or January  10,  1941, whichever occurred  later, rather than
January  1, 1942,  as in the previous study.  Except for extending the  followup
by 3 years  from December 31,  1979, to December 31,  1982,  both the  study group
and the reference group of 1775 cork and styropor panel production workers  used

Julv 1988                          8-44         DRAFT—DO  NOT QUOTE OR CITE

-------
as internal  controls  were  identified through the same procedures used in the
1984 study.  However,  the  study group was reduced from 2096  to  2092  and the
reference group from 1778 to 1775 because of earlier  errors  in coding  sex.
     Since local mortality  rates  did not  differ significantly from West German
national rates, national rates  were used to calculate expected deaths for an
external comparison  group.   After  using  a modified life-table analysis to
obtain  person-years  at risk, age,  sex,  and calendar-time specific national
mortality  rates for  the years 1952  to 1982  were  applied  to person-years to
determine expected deaths.   The 1952 rates were applied to person-years accumu-
lated prior  to the  availability  of the  rates.  For an internal  comparison
group,  relative risks  were  estimated for  causes of five or more deaths.  Where
cause-specific mortality was large,  deaths  were stratified  in  10-year  age
groups;  where  cause-specific mortality was  small,  in  25-year age groups.
Because Claude and Frentzel-Beyme (1986)  stated that occupational exposure of
the reference  group  to low levels of MMMF cannot be  excluded, the suitability
of this  group as an internal control is questionable.
     Although exposure  data  were  unavailable,  mineral  oil was  used as a binder
from 1941  until  1954 when  resin  binders  were  added, indicating a potential
change  in  exposure conditions.   As  in the previous  study,  coal  tar,  silica,
urethane formaldehyde  foam,  and polystyrene were used during  production pro-
cesses.   Exposure occurred  between  1941  and 1977, allowing  a maximum  exposure
period of 37 years.   Since  the observation period was from January 10, 1941,  to
December 31, 1982, allowing a maximum followup of 42 years, there were only  5
years without overlapping exposure and observation periods.
     By the  end of the followup period, December  31, 1982,  1564  (74.8 percent)
of the  study group members were alive, 315 (15.1 percent) had died;  213  (10.1
percent) were  lost to followup.  Among the 1775 internal  controls, 1300  (73.2
percent) were  alive;  290 (16.3  percent)  had died; 185  (10.4 percent)  were  lost
to followup.  The proportion of persons lost to followup and through emigration
was large; 62 percent were  migrant workers.
     Compared to West  German males, significantly increased mortality occurred
from all malignant neoplasms (80 observed, 63.1 expected,  SMR = 127,  95 percent
CI = 101,  157,  p  <0.05, Power « 97 percent), benign and neoplasms of unspeci-
fied nature  (10 observed,  3.0 expected,   SMR « 336,  95 percent CI * 171,600,
p <0.05),  mental disorders  (7 observed,  3.1 expected,  SMR  = 228, 95 percent
CI = 100,  452,  p  <0.05) and symptoms and ill-defined conditions (23 observed,

July 1988                          8-45         DRAFT—DO NOT QUOTE OR CITE

-------
 8.6 expected, SMR = 268, 95 percent CI = 174, 397,  p <0.05).   Although  investi-
 gators found  no  overall  excesses from either lung  cancer or  NMRD,  there was
 insufficient  power  to  detect both lung cancer (53  percent) and NMRD (53 per-
 cent).  To  reduce  the  high number of deaths  from  111-defined  conditions for
 which  death certificates were unavailable, the investigators  contacted  last.
 attending physicians and family to obtain Information on causes of death.  When
 additional  cases  identified in  this  Manner were  included in  the number of
 observed deaths, deaths from accidents, poisoning,  and violence became  signifi-
 cantly elevated  (56 observed,  39.8 expected, SMR =  143, 95 percent CI = 110,
 184, p <0.05).
      Since  those employed  for  less than 1 year contributed one-third  of the
 total person-years  at  risk,  the  mortality of the  entire cohort was  influenced
 by this group.  The mortality experience of those  employed less than 1  year was
 significantly increased overall (SMR = 140, 95 percent CI =  113,  173),  but this
 excess was  not  attributable  to any particular cause of  death.  When short-term
 workers were  excluded,  investigators  found no significantly elevated causes of
 death, but  19 of the 21  lung cancer deaths occurred *(nong those with 1 or more
 years of employment.
      When the cohort was  analyzed by time since  first exposure,  investigators
 found significant excesses of  death from  all malignant  neoplasms among those
 with 20 to  29 years since first  exposure  (39 observed,  24.3  expected,  SMR =
 161, 95 percent  CI  = 116, 217,  p <0.05)  and  from  benign and unspecified neo-
 plasms among  those  with  9 or less years since first exposure  (4 observed, 0.7
 expected, SMR = 576, 95 percent CI = 184, 1388, p  <0.05).  Although  it  appeared
 that the  excess  mortality from malignant  neoplasms was  unrelated to  latency
.period or  length of exposure,  the power  to  detect site-specific cancers was
 low, perhaps  because only 1.4  percent person-years had  accrued in the category
 of 20 or more years of exposure and 30 or more years of observation  since first
 exposure.
      To determine whether the introduction of resin binders  in 1953  altered the
 mortality experience of  the  cohort, analyses were performed by date of entry,
 i.e., before and after 1953, and years since first exposure.   Among  those first
 employed  before  1953,  investigators found significant excesses from all malig-
 nant  neoplasms  (24  observed, 14.5 expected,  SMR - 165, p <0.05) as well as
 stomach cancer  (6  observed,  2.3  expected, SMR = 261,  p <0.05)  among those with
 20 to  29  years  since first exposure;  lung cancer  among those with 10 to 19
 years since first exposure (6 observed, 2.2 expected, SMR = 272, p  <0.05, Power
                                    B-46         DRAFT—DO NOT QUOTE  OR  CITE

-------
= 12  percent);  benign  and unspecified neoplasms among those with less than 9
years since  first  exposure  (2  observed, 0.2 expected, SMR » 888, p <0.05); and
unspecified  neoplasms  overall  (6  observed,  1.5 expected,  SMR &  406, p <0.05).
Among those  first  employed  after  1953, a  significant excess was observed  from
all causes of death (163 observed, 137.7 expected,  SMR = 118, p  <0.05).
     When  the  cohort was analyzed by  job classification  as  a  surrogate for
degree  of  exposure,  production workers demonstrated a significant excess  of
deaths  from  benign and unspecified neoplasms  (6 observed,  1.9 expected, SMR =
313, p <0.05), especially among those with 9 or less years  since first exposure
(3 observed, 0.5 expected,  SMR =  666,  p <0.05).  Maintenance workers exhibited
significant  excesses from all  malignant neoplasms  among those  with 20  to 29
years since  first  exposure  (16 observed,  8.1 expected, SMR = 197, p <0.05),
lung  cancer  among  those with 20 to 29 years since  first exposure (6 observed,
2.4 expected, SMR  «  249,  p  <0.05,  Power = 16  percent), and benign and unspeci-
fied neoplasms (4 observed,  1.1 expected,  SMR  = 379, p <0.05), especially  among
those with 20 to  29 years  since  first exposure (2  observed, 0.3  expected,
SMR = 618,  p <0.05).   Therefore,  both production and maintenance workers  had
significantly elevated deaths from benign  and  unspecified  neoplasms, especially
among those  with 20  to 29 years  since  first  exposure.  Lung cancer mortality
appeared to be higher  in maintenance workers,  especially  among those with  20  to
29 years since first exposure.
     When  the  data were  analyzed using internal controls as the  reference
population,  there  were significantly increased relative  risks  from pancreatic
cancer  (RR = 6.76, 95 percent C!  = 1.01, 42.50, p  <0.05) and  neoplasms  of
unspecified nature (RR * 2.89,  95 percent  CI = 1.14, 7.32, p  <0.05).   All  seven
pancreatic cancers ard all! twat one of the  unspecified neoplasms  occurred in the
exposed group.  The  investigators stated  that no cases of mesothelioma were
found.  It Is  unclear  whether  the mesothelioma found in the control group  from
the 1984  study  was not listed here because 1t occurred among one of the cases
excluded due to errors 1n coding sex.
     The investigators concluded  that (1) no  significantly elevated mortality
from  lung  cancer  was  observed  in  the  cohort; (2)  most of the  excesses from
other cause-specific  deaths appear  unrelated to MW production  because they
occurred among those employed  for short periods or as frequently among internal
controls;  (3) there was 95  percent power  to detect a twofold  lung  cancer  risk,
but an  increased risk  of  lung  cancer was not  significant and related neither to
time  since  first  exposure nor to duration of employment; (4) as stated in the
Julv IPPft

-------
1984  study,  -Interpretation of  positive  findings nay be complicated by  the
possibility of  multiple  exposures  and lack of information  on previous  occupa-
tional exposures, especially among short-term workers;  and  (5) the  unavailabil-
ity  of  death certificates  for 7 percent  of the decedents  resulted in  an
underestimate of cause-specific risks.
     This  study suffers from  the  same limitations as its predecessor.   In
addition,  although  the  investigators  correctly  concluded  that there was 95
percent power to detect a twofold risk of lung cancer,  the  cohort was too small
to have  sufficient  power to detect a  1.5-fold risk  of either lung cancer or
NMRD.  For example,  when the data were analyzed by latency,  duration of  expo-
sure, and period of  employment,  only 1.4 percent of  the  person-years were
accrued by individuals with 20 or more years of exposure and 30 or more  years
of latency.  Nevertheless, with only 12 percent power to detect a 1.5-fold risk
of lung  cancer, compared to West German males,  there  was  a  threefold  excess
among production  and  maintenance  workers employed before 1953  with  10 to 19
years of  latency.   Furthermore, without extensive exposure  and  latency  for  a
large percentage of the cohort, it is especially important  to adequately  follow
exposed individuals,  whether  terminated  or retired,  to determine whether they
develop cancer.   Thus, the large  percentage  of the cohort  lost  to  followup may
also explain why no significantly elevated lung cancer mortality was detected.
     Bertazzi et  al.  (1984)  conducted a  historical prospective  mortality study
of 1158 male production workers who were employed at  least  1  year in an Italian
FG plant  between  1944 and  1973.  The  cohort  was  assembled  from  company person-
nel  records containing names,  sex,  place and date of  birth,  residence and/or
last address, chronological  work  history,  and dates  of employment.   A  total  of
2434 workers -(2100  males and 344  females)  were employed  between 1944 and 1973.
Analysis was restricted to 1158 male production workers with  at least.1 year  of
employment.  After  using a computer program to obtain person-years at  risk  of
dying, the expected numbers of  deaths were calculated by cause, sex, and  5-year
calendar-time and age groups based  on three  different  sets of mortality  rates:
(1)  Italian national  rates for the years 1942 to 1977;  (2)  Lombardy regional
rates for the  years  1969  to 1978;  and (3) rates from a nearby  city with a
population of about 100,000 for the years 1954 to 1978.
     Exposure was classified  according to 20 different  jobs,  each implying  a
different  level of  exposure to FG as  estimated  in 1980 by  the Edinburgh  Insti-
tute of  Occupational  Medicine.   Personal samples from a random sample of 166
workers  reported  concentrations ranging from 0.007 to 0.023  fiber/cm  for the
July 1988                          8-48         DRAFT-DO NOT QUOTE OR CITE

-------
respirable  fiber  fraction  and 0.002 to 0.10 fiber/cm  for the  nonrespirable
fraction.   From scanning electron  microscopy,  median fiber  length  ranged  from
3.7  to  7.5 pin and median  diameter  ranged  from 0.3 to 0.7 pm.   There was a
maximum  exposure  period of 30 years from 1944 to  1973.  For each worker, two
exposure  indices  were  calculated:   a cumulative Index obtained by summing up
fibers/cm-months  and  a weighted  cumulative  Index.   Since the observation
period was  1944 to 1980,  «1lowing a maximum followup of 37 years, there were
7 years without overlapping exposure and observation periods.
     Vital  status was  ascertained  through  the Population Statistics Office
nearest  a  worker's  residence  or place of birth.   By  the  end of the  followup
period,  December  31, 1980, 1027 (88.7  percent)  of the 1158 male production
workers  were  alive;  106 (9.2  percent) had died; 25 (2.1 percent)  were lost  to
followup.   It was not  possible to determine the percent of  the death certifi-
cates obtained for  decedents.   Deaths were  coded according to  the  8th Revision
of the ICD.
     There  appeared  to be  a  gradient  of  SMRs based on choice  of  reference
population:  SMRs were  highest compared to national rates, intermediate  coir-
pared to local rates, and lowest compared to regional rates.   The  significantly
increased mortality  from all  cancers (34 observed, 22.90 expected, SMR = 14£,
95 percent  CI =  105,  205,  p  <0.05) and laryngeal cancer (4 observed,  0.9£
expected, SMR = 408, 95 percent CI  =  130,  985,  p  <0.05)  compared  to nations!
rates were  no longer  significant  when local  or regional rates were used.
Deaths  from cardiovascular disease  were significantly  decreased  compared to
national  rates (20  observed,  31.39 expected,  SMR « 64,  95 percent  CI =  40,  97,
p <0.05),  local  rates   (38.13  expected, SMR =  52,  95  percent  CI  =  33, 80,
p <0.05), and regional  rates (36.90 expected, SMR = 54, 95 percent-CI = 34,  82,
p <0.05).   Using  national, regional, or local  rates,  there was insufficient
power to detect  a 50 percent  increased mortality  from lung cancer,  laryngeal
cancer,  or  NMRD.   No deaths from mesothelioma were reported.
     Since  cancer mortality 1n the area where the plant was located was known
to be higher than that of the national population, the Investigators used local
rates  as the comparison group  for analyses of cancer mortality data.  When
cancer mortality  data  were analyzed by latency period (<5,  5 to 15, 15 to 25,
and  >25  years),  duration  of exposure (<10,  10 to  20, and >20 years), age at
first exposure (<25, 25 to  34, 35 to 44 and >45 years), and cumulative exposure
(unexposed, <0.03, 0.03 to  0.10, 0.10 to 0.30, and >0.30 fibers/cm -months), no

Julv 1988                          ft-^s          n&AFT--nn urn nnnrp  OP nTF

-------
significant trends were  found except for laryngeal  cancer,  which was  signifi-
cantly elevated among  those  with  25  or more years of latency (3 observed, 0.32
expected, SMR = 938,  95 percent CI = 238, 2551,  p  <0.005); with  20 or more
years of exposure  (2  observed, 0.24 expected,  SMR * 833,  95 percent CI = 140,
2753, p  <0.05); whose first  exposure  was before age  25  (2 observed, 0.13
expected, SMR = 1,538,  95 percent CI * 258,  5083, p <0.01); and whose cumula-
tive exposure was  greater than or equal  to 0.30 fibers/cm -months (3 observed,
0.20 expected, SMR e  1500, 95 percent CI * 381, 4082, p <0.005).   The investi-
gators also  found that  all  four  workers with laryngeal  cancer  were first
employed before  1960  when FG wool rather than  continuous  FG filament was
produced.  In addition,  the  results  showed  a consistent,  significant  increase
among those most heavily exposed.
     The investigators  concluded   that the  excess risk of  laryngeal  cancer
following exposure to  FG seemed biologically plausible and was estimated  in
comparison with a  population having similar  socioeconomic  conditions as well  as
an elevated background  risk.   According  to the investigators, the small number
of laryngeal cancer cases upon which the excess  was based, the absence of an
excess in previous epidemiologies! studies,  and incomplete control  of  confound-
ing variables limit the  findings  of  this study.   Generally, SMRs based on five
or fewer observations  are subject to chance  fluctuation.  This also applies  to
cases where, because  no  deaths were  observed,  SMRs  are zero.  In both of these
cases, the 95 percent  confidence   intervals  reflect a wide  range  of possible
risk.
     Bertazzi  et al.  (1986)  extended the followup  period of their previous
study for an additional  3 years (December 31, 1983).  However,  it  cannot  be
determined why the size  of the cohort was reduced  from  1158 male  production
workers  in the previous  study to  1098  in this  study.  Because there are marked
variations In cancer  mortality rates across  different regions in Italy, inves-
tigators used  local  rather  than  national rates  to  study cancer  mortality.
Between  1944  and   1960,  FG wool was  produced.   After  1960, production was
limited  to continuous  FG filament.  Resin binders were used since 1944.  There
was no  reported  use  of  asbestos.  The  concentrations of respirable  fibers
reported  ranged  from  0.005  to 0.048  fiber/cm  with a median  length,  as
determined by  scanning  electron  microscopy,  of  3.7 to 7.5 urn and  a  median
diameter of 0.3 to 0.7 urn.  No asbestos  fibers were detected in any of the
samples.   There was a  maximum exposure  period of 30 years  from 1944  to 1973.

July 1988                          8-50         DRAFT—DO NOT QUOTE OR CITE

-------
Since the observation period was from 1944 to 1983,  allowing  a maximum followup
of 40 years,  there were 10 years without overlapping exposure and  observation
periods.
     Vital status  of  former employees was ascertained through the Population
Statistics Office  nearest  • worker's nost recent residence or place of birth.
Of the  1098  workers,  956 (87.1 percent)  were  alive;  130 (11.8  percent)  had
died; and only 12  (1.1 percent)  were lost to followup.   Death certificates
were obtained  for  all  130 of the decedents.   Therefore,  there was  a very high
degree  of  followup and •  100 percent trace rate.  Deaths were coded according
to the 8th Revision of the ICO.
     SHRs were not consistent between national  and local  rates.  They tended to
be higher when national rates were used.   The significantly increased mortality
from all  causes compared  to national rates (130 observed,  106.4  expected,
SMR = 122, 95 percent CI « 102,  145) was  no longer significant when local  rates
were used  (126.1 expected,  SMR  = 103, 95 percent CI « 86, 122).   Using either
national or  local  rates,  there  was  insufficient power to detect a 50 percent
increased mortality from  lung cancer, laryngeal  cancer,  and  NMRD.  Deaths from
cardiovascular disease (CVD) were significantly decreased using  either  national
(21 observed,  35.0 expected, SMR = 60, 95 percent CI = 37, 92)  or local  rates
(42.1 expected, SMR =  50,  95 percent  CI  = 31,  76),  as one would expect for a
group of factory workers.   No deaths from mesothelioma were  reported.
     Using local  rates as  the  comparison group, cancer mortality data were
analyzed by  latency,  duration of exposure, period of exposure,  end cumulative
exposure.  No  significant  trends  were observed except for  laryngeal cancer
among those  with 25 or more years of latency (3 observed, 0.5 expected,  SMR =
591, 95  percent  CI * 150, 1608, p <0.05, Power = 4 percent). When the cohort
was  analyzed by period of  first employment,  all laryngeal  cancer occurred
before  1960  when  glass wool was  produced.  When the data were  analyzed  by
cumulative exposure, three  of the four laryngeal cancers were observed  in the
highest exposure group.
     The investigators concluded that (1) mortality from all  causes was  signif-
icantly elevated based on national  but not local mortality rates;  (2) analysis
by latency,  duration  of exposure, period of exposure, and cumulative exposure
failed  to  confirm  the hypothesis of  an association between lung cancer deaths
and exposure  to  MMMF;  (3) increased  laryngeal  cancer mortality was  statisti-
cally significant  after 25 or more  years  of latency and 15 or more years of

July 1988                          8-51         DRAFT—DO NOT QUOTE OR CITE

-------
exposure  among  those hired  before  aged 25,  having  the highest cumulative
exposure, and first employed before 1960 when airborne  concentrations  of  fibers
were higher  than  during the production of continuous filament;  and  (4) exoge-
nous risk  factors  associated with laryngeal cancer,  such as  smoking,  alcohol,
and asbestos, do  not appear to be present.   The investigators cautioned that
although biological  plausibility  exists  for the excess laryngeal  cancer  risk,
the study  findings  are  limited by the small  number  of laryngeal cancers, the
lack of  histopathological characterization of  these cases,  and the lack of
information on an individual's other possible risk  factors.
     This  study  was strengthened by  the  excellent followup  and 100 percent
trace rate.  Despite the low power and small number of  deaths,  the detection of
rare tumors  such as laryngeal cancer  attests to the importance of thorough
disease ascertainment among decedents.  It is not possible to conclude that  the
excess laryngeal cancer is  causally related to the  low  levels  of FG  exposure
because four cases are too few to exclude the effects of chance.   Because other
studies have  lacked  this degree  of followup and disease ascertainment,  laryn-
geal cancers associated with FG exposure may have gone  undetected.
     Gardner et al.  (1986)  conducted a historical  prospective mortality  study
of two FG  factories  in  the United Kingdom.  One cohort consisted of 4734 male
and female  (25  percent) workers  employed for at least  1 year between April  1,
1946, and  December  31,  1978, in an English FG wool factory.   After  the cohort
was assembled  from  payroll  records,  data were supplemented by  pension  and
medical records.   Vital  status was  determined from the National Health Service
Central Register and Department of Health and Social  Security.
     The other  cohort consisted  of  1910  male  and female  (37.4 percent) workers
employed  at  any time between January  1,  1956, and  December 31,  1978, in  a
Northern  Ireland  continuous glass  filament factory.  After the  cohort  was
assembled  from  job application forms, personnel files,  and  medical  records,
vital  status was determined  from the Central  Services  Agency in  Belfast.
Copies of death certificates were obtained through  the  General  Register Office.
Persons not  identified from these sources were traced through the plant medical
officer and/or their own physicians.  For those who emigrated to England, vital
status was  determined through the National Health  Service Central Register  and
Department of Health and Social Security.
     After  using  a  computerized  life-table  method developed to  obtain
person-years at risk of dying,  the expected  numbers of deaths were calculated

-------
by age,  sex,  cause,  and calendar-year specific national  rates.   Local  adjust-
ment factors were applied to txpected numbers for the  English  cohort  using  SMRs
for the  period  1968  to 1978 from the  locality containing the factory.  Since
local  rates  were unavailable  for  the Irish cohort, the national rates  for
Northern Ireland were used to calculate expected deaths.
     From  employment  records,  dates  of  employment and descriptions of  all
recorded jobs  in the  factory  were  abstracted.   Workers  were  categorized as
preproduction, production, secondary processes, less dusty  secondary  processes,
maintenance, general staff,  and  office staff.   Short-term  industrial hygiene
surveys  of dust  levels were performed by the  Edinburgh  Institute of Occupa-
tional Medicine  during 1977 for  the English factory and during  1979 for the
Irish  factory.   Samples were  collected  on  membrane filters  and analyzed by
"interference" microscopy.   Fiber  levels were  reported  to have been low,
especially for the continuous  filament process in Northern Ireland,  with mean
concentrations of less than 0.1  fibers/cm   for each job  category.  Mean  respi-
rable concentrations were  highest  in the secondary processes category of the

                                    3
English factory at  0.073  fibers/cm   and  in the secondary processes category of
the Irish factory at 0.002 fibers/cm'
     By the end  of  the followup  period,  December 31, 1984, of the 4734 English
cohort members,  3932  (83.1  percent) were alive; 622 (13.1 percent) had died;
180 (3.8  percent) were lost to followup.  Of  the  1910  Irish cohort members,
1,729  (90.5 percent)  were alive;  156 (8.2 percent) had  died; 25 (1.3  percent)
were lost to  followup.   For the  English  cohort, exposure occurred between 1S46
and 1978, for a  maximum exposure of 33  years,  and followup  occurred  between
1946 and  1984, for  a maximum observation period of  39  years.   For  the Irish
cohort, exposure occurred between  1956  and 1976 for a maximum exposure  of
23 years, and followup occurred  between  1956 and 1984,  for a maximum  observa-
tion period of 29 years.  For both cohorts, there were  5 years without overlap-
ping exposure and observation periods.   Deaths  were  coded according to the  8th
Revision of the  ICD.
     For both sexes of the English cohort, total deaths  were similar to national
mortality rates  but lower than local mortality rates.  Among male  workers, a
significant deficit was  observed from accidents, poisoning,  and violence when
compared to national rates (21 observed,  33 expected, SMR = 64,  95 percent CI -
40, 98, p <0.05).  Among female workers,  a significant excess of stomach cancer
was found compared  to national rates (6  observed, 2 expected,  SMR  = 277, 95
percent CI =  102,  604, p <0.05).   Because  there were  few lung cancer deaths
                                   8-53         DRAFT—DO NOT QUOTE  OR CITE

-------
among  female  workers, analysis  of  lung cancer mortality was restricted  to
males, who  were further  restricted  to exclude office workers.   Among male
production workers, a significant excess  of lung cancer was found only when
compared to national  rates  (66 observed,  50.8 expected, SMR = 130, p <0.05).
When this group was analyzed by latency and duration of  employment, none of the
excesses was  significant  compared to local rates.   No trends were associated
with latency  or duration  of exposure.   In addition, when data for both sexes
were analyzed by degree of exposure,  no statistically significant excess deaths
from  lung  cancer occurred.   Although  the cohort was large  enough to have
sufficient power to detect lung cancer or  NMRD, the  power was too low  to detect
either  lung  cancer  or NMRD  when the  data were  divided  into multiple
subcategories.
     Among males in  the  Irish cohort,  significant excesses were found from all
other  causes  of death (112 observed, 89 expected,  SMR = 125, 95 percent  CI =
103, 151,  p  <0.05) and accidents, poisoning,  and violence  (27  observed,  15
expected, SMR = 174,  95 percent CI =  115,  254,  p <0.05).  Among women,  investi-
gators found  a  significant  deficit  from all  causes  (21  observed, 32 expected,
SMR 65,  95 percent CI = 40,  99, p <0.05).   There was  insufficient power to
detect either  lung cancer or  NMRD in either sex.   When  lung cancer mortality
among  males was  analyzed  by duration of exposure,  length of employment,  and
occupational   category,  no trends were  identified.   Of  the  nine  lung cancer
deaths,  four  occurred among males who worked  in the  factory for less than 1
year (4  observed,  1.1 expected,  SMR  = 378,  p  <0.05).  No mesothelioma deaths
were reported in either factory.
     The investigators concluded, for the  English cohort (1) no excess mortal-
ity existed,  except  for  lung  cancer  when  compared  to national rates;  (2) no
relationship existed  between  length  of  employment,  duration of exposure,  level
of exposure,  or occupational  classification; (3) the absence of any important
relationship between  lung  cancer and latency,  duration,  or degree  of  exposure
suggests that  the  excess  using national rates was  unrelated to employment at
the  factory;  and (4) the excess  stomach  cancer among women  was  significant
compared to national  but  not  local  rates  because of high local stomach  cancer
rates.    They  concluded that in the  Irish  cohort, no excess mortality existed
for  either  sex, except for accidents,  poisoning,  and violence  among males,
which  can be  ascribed to the high rate of violent deaths in  Northern  Ireland.
They concluded  that  for  both  cohorts (1)  low FG fiber levels were  measured in
both factories  as  compared  to U.K. asbestos  control  limits; (2) no excess lung
July 1988                          8-54         DRAFT—DO NOT QUOTE OR CITE

-------
cancer mortality  was  found for workers with  long followup; and (3) the statis-
tical power  to detect laryngeal  cancer was only 38 percent  (level of  risk not
reported).
     The  conclusion that  no relationship exists between lung cancer  and job
classification,  latency,  duration,  and  degree  of  exposure  1s weakened  by
several limitations.  First, although there was  sufficient power to detect a SO
percent increased risk  of lung  cancer among English males,  there was  insuffi-
cient power  to detect  lung cancer  among English females or Irish males or
females.  The  deficits  that occurred when the data were analyzed by duration,
latency,  and degree  of exposure  could t»ave  resulted from the low power to
detect a  50  percent excess of lung cancer deaths,  if  such  an excess  really
existed.  The  same holds  true for NMRD.  Second,  although  the investigators
claimed no excess of  lung cancer with long  followup,  it was not possible  to
determine the percent of the cohort with 20 or more years of exposure  and 30  or
more years of observation.  Third, fiber levels may have been too  low  to detect
a weak carcinogenic response, if it existed.
     Saracci  et al.  (1984) -rondvrrtrt t tisttm'ca* prospective mortality and
morbidity  study   of  25,146  persons  employed between  1900  and  1955  and
December 31,  1977.  The cohort  essentially consisted of workers employed in  1
of 13 plants producing  MMMF in  the seven European  countries described above.
Selection criteria  for  including  a  plant's work force  in the study were (1) no
loss or destruction of  personnel  records,  so that  the  total  cohort was identi-
fiable;  (2) MMMF  production for at least 20 years;  (3)  facilities  available for
mortality and  morbidity ascertainment;  and (4) exclusion of plants that pro-
cessed asbestos.   As  a result of mailed questionnaires  followed by  site visits,
13  factories (7  rock wool, 4 glass  wool,  and  2 continuous filament) were
selected  for the  study.   In each country, a national research team supervised
the  identification  and  transcription  of workers' names  and  job histories  at
each plant as  well as the  followup.  A  worker  was  considered to have entered
the cohort on the date of first employment 1n one of the factories  ranging  from
1900 to 1955.  At four plants,  however, because of missing  essential  informa-
tion on short-term employees, those employed for less than 1 year  were excluded
from the cohort.
     After using  a modified life-table procedure to generate person-years at
risk, each  country's  national  age, sex, cause, and calendar-time  specific
mortality rates and age,  sex, site, and  calendar-time  specific national cancer
incidence rates where available were  applied to person-years at risk  to obtain
July 1988                          8-55         DRAFT—DO NOT QUOTE OR CITE

-------
the  expected  numbers of deaths or  cancer  cases.   Person-years  and expected
values for 5-year  age  and calendar-year groups were computed from the first
year in which national  rates were available for each country, which ranged from
1942 to 1953 for mortality and from 1943 to 1958 for cancer  incidence.  For  all
but  one plant,  the Edinburgh Institute of Occupational Medicine conducted an
environmental survey to measure concentrations  of fibers present at the time of
the  study collected  from  personal samples  during an 8-hour work shift.  Within
each factory, jobs were  classified  into six occupational  categories:  (1) not
specified; (2)  office;  (3) preproduction;  (4)  production; (5) secondary  pro-
cesses; and  (6) maintenance.   Concentrations of respirable  fibers  at  the time
of the study  analyzed  by  phase-contrast microscopy were below 1 fiber/cm  and
ranged from  0.01   to 0.1  fiber/cm .   Exposure  and observation periods were
concurrent.
     Vital status  was  ascertained  through national facilities or  local con-
tacts.   The  followup period  ended  on  December  31,  1977,  for  10 plants;
December 31,  1978, for  1  plant; and December  31,  1979, for 2 other plants.
Of the 25,146 total  cohort numbers. 23,487 (93.4  percent)  were alive; 1,659
(6.6 percent) had  died;  and  745  (3.0  percent) were lost to followup.  The
followup of  individual plants ranged from 3 to 10 percent.   Causes  of death
and  cancer diagnosis were  coded according to the ICD in use at  time  of death,
but for ease  of presentation, all  codes were converted to the 7th Revision  of
the ICD.
     Among males,  a  significant excess of deaths occurred  from  accidents,
poisoning, and  violence  (246  observed,  205.8 expected, SMR = 120, 95 percent
CI = 105, 135, p  <0.05),  but  significant  deficits  from  all causes  (1505
observed,  1666.4 expected, SMR * 90, 95 percent CI = 86, 95, p <0.05), prostate
cancer (14 observed,  25.9 expected,  SMR = 54, 95 percent CI  * 30,  91,- p <0.05),
circulatory system disease  (538 observed,  715.1 expected, SMR » 75, 95 percent
CI = 69,  82), and  bronchitis  and asthma (36  observed,  50.7 expected, SMR = 71,
95 percent  CI s 50,  98).   Although laryngeal  cancer  was not significantly
elevated  among  workers  from the combined  13 MMMF  plants,  four  of the seven
deaths occurred in one factory (4 observed, 1.0 expected,  SMR =  408,  95 percent
CI = Hi, 1024, p  <0.05,  Power = 7 percent).  The investigators did  not state
the  location  of the  plant or the type  of fiber produced there.  One death from
mesothelioma was reported.  Among females, significant deficits  were  found frotr,
all  causes (154  observed,  219.6 expected, SMR =  70,  95 percent CI = 59, 82,
p <0.05),  all malignant  neoplasms  (49 observed, 70.1  expected, SMR  ~ 70, 95
Julv 1988                          8-56         DRAFT—DO  NOT QUOTE OR CITE

-------
percent CI =  52,  92,  p  <0.05),  intestinal cancer (0 observed, 5.1 expected,  95
percent C.I = 0, 72), and circulatory system  disease  (44 observed, 76.9 expected,
SMR s  57 95 percent  CI *  42,  77,  p <0.05).   However,  because of the small
number of cancer deaths in women, further analysis was  restricted to men.
     When the cohort was analyzed by time since  first employment and by process
(rock wool, glass  wool,  or continuous filament), neither  significant excesses
nor deficits  were  associated  with  increasing latency.  Among glass wool work-
ers, significant deficits  were  associated with all causes (487 observed, 679.2
expected, SMR »  72,  95 percent CI  * 66, 78,  p <0.05),  which  remained signifi-
cantly elevated among those with 19 or  less, 20 to 29,  or 30 or more years
since first exposure.   There  also  were  significant deficits  from all neoplasms
(119 observed, 155.5  expected,  SMR = 77, 95  percent CI = 63, 97,  p <0.05),
especially among those with 20 to 29 years since first  employment (28 observed,
45.0 expected, SMR =  62,  95 percent  CI  s 41, 90).  Among continuous filament
workers,  however,  there  were  significant  excesses from  all  causes  (212
observed, 171.6 expected,  SMR = 124,  95 percent CI = 108, 141, p <0.05), espe-
cially among  those with 19 or less years since  first employment (173 observed,
137.1 expected, SMR = 126, 95 percent  CI =  108,  146,  p                            e_n         tNBurT. . r\*> »i*» »• i^^-r «n rTTC

-------
 the low power to  detect  lung cancer precluded meaningful analysis, presumably
 because of an absence of  airborne fiber measurements in one factory and missing
 job classifications for others.   To further explore a potential dose-response
 relationship, the  investigators  performed  a  nested case-control  study using
 male lung cancer  deaths observed  in the study as cases.   Controls  surviving to
 each case were matched by  factory and date of birth (up to 40 controls  per
 case).   The  population  from which  controls  were drawn was not  specified.
 Duration of exposure of each case was compared with that of controls  up to the
 time of death of  the cases.  The investigators reported that  they found no
 statistically significant trend with increasing  risk and increasing duration of
 exposure in any  of the three production processes.
      When  cancer  incidence  was analyzed by time since first employment in rock
 wool or glass wool  factories,  investigators  found a significantly decreased
 incidence  from all  neoplasms  in  FG  wool workers  (62 observed,  104.3 expected,
 SIR = 59, 95 percent CI  =  46,  76, p <0.05), especially among those with  19  or
 less years since  first employment (32 observed,  54.4 expected, SIR =  59,  95
 percent CI = 40,  83,  p <0.05) and 20  to 29 years since first exposure  (19
 observed,  34.4 expected,  SIR = 55, 95 percent CI = 33, 86, p <0.05).
      Finally, when lung cancer mortality data were pooled from  all 13 factories
 and analyzed by latency, a significant excess  of lung cancer was found among
 those with 30 or  more years since first exposure (17 observed, 8.9 expected,
 SMR = 192,  95 percent CI  =  117, 307, p <0.05, Power * 36 percent).
      The investigators concluded  that  (1)  the mortality and  cancer  incidence
 analyses showed  no  consistent  trends  except for  lung cancer; (2) the signifi-
 cant excess of lung cancer observed 30 years after first employment when data
 from the 13 plants  were  pooled suggests that an  increased risk was associated
•with exposure to  MMMr  30 or more years ago; (3)  participation of  five more
 factories  would  have increased the  study population  by 30 percent;  (4) while
 this study used  national  rates to assess cancer mortality, and  while the  use of
 local rates tends to  lower overall  SMRs,  local  rates do not appear to affect
 SMRs in the  longest periods since first employment; and (5) although informa-
 tion on potential confounding variables,  such  as smoking,  was unavailable,
 asbestos does not  appear  to be a  confounding variable because asbestos  exposure
 would have excluded participation of a plant  in  this  study.
      Pooling the  data  increased  the  sample  size  so  that  sufficient power
 existed among the males to detect lung  cancer or  NMRD.  However, when the data
 were subdivided  into multiple  categories,  the power was too low to detect lung
 July 1988                           8-58         DRAFT—DO NOT QUOTE OR  CITE

-------
cancer or NMRD.  This study was further limited by the  overlapping exposure and
observation periods,  Inadequate  followup  in  some  plants (as high as 10 per-
cent), the small number of deaths fn the cohort, «nd the failure  to analyze the
data by length of exposure.  All of these deficiencies  could result in underes-
timating the  risk  of lung cancer and NMRD 1n the  study  group.  In addition, it
was not possible to determine the  percent of the  cohort with 20  or more years
of exposure and  30 or more years of observation since  first exposure to MMMF.
Nevertheless,  despite only 2.9  percent  of the person-years for both sexes
accumulated in the  30 or more years since first exposure category, and despite
low power,  a twofold excess  of lung  cancer  mortality was  found among male
cohort members after 30 years of latency.
     Simonato et al.  (1986a)  extended  the followup of the  cohort studied  by
Saracci et al. (1984) for 5 additional years by  conducting a historical pro-
spective mortality study of 21,967  male and female workers  ever employed  in the
13 European factories manufacturing various types  of MMMF.   As in the previous
study, those  employed for less than 1 year at four plants  were excluded.   The
cohort was reduced  from 25,146 to  21,967  by calculating person-years from  date
of first exposure  to MMMF rather than  from the date of  first employment in one
of the 13 plants and by excluding office workers from the cohort.  As  described
earlier, factories in countries where no followup  could be  performed,  that  were
recently opened, or that had asbestos exposures were excluded.  Expected  deaths
were  calculated  using age, sex, cause, and calendar-year  specific mortality
rates  for each  of  the seven countries.   In addition, local  lung  cancer mortal-
ity rates were  estimated by adjusting  the expected numbers  by a  factor equiva-
lent to the local lung cancer SMR.
     Because  past  exposure data were  unavailable,  based on the  use  of dust
suppressants  (mineral  oil,  resin binders, and  pitch) and potential exposure to
MMMF  in  different  periods of  industrial activity,  production processes were
divided into  three  qualitatively distinct technological  phases.  In the first,
the early  technological phase, airborne concentrations  of  MMMF  were  highest
because of  the absence  of dust suppressants  and/or batch  processing with
labor-intensive methods.   In  the second,  the  -intermediate  technological phase,
production methods with and without dust suppressants were used.   In  the  third,
the late technological  phase, mineral  oil and resin binders were used as  dust
suppressants with modern mechanized production methods.
     Simonato  et al.  (1986a)  reported  significant excesses from all causes
(2719 observed, 2457.0 expected, SMR = 111, 95 percent CI =  107,  115,  p <0.05),

-------
all malignant neoplasms (661 observed, 597.7 expected,  SMR = 111,  95 percent CI
= 102, 119, p <0.05), and lung cancer (189 observed,  151.2 expected, SMR = 125,
95 percent  CI  = 108, 144,  p <0.05).   One  death from mesothelioma was reported.
Expected values  were estimated to have ranged from one to  three  deaths from
mesothelioma.
     When the  cohort was  analyzed by latency, investigators found significant
excesses from  all  causes  among those with less than 10 to 19,  and 20 to 29
years since' first exposure, from all malignant neoplasms among those with 10 to
19 and 20  to 29 years since first  exposure,  and from lung  cancer among those
with 30 or  more years since first exposure (29 observed,  16.8 expected, SMR =
173, 95 percent  CI  * 116, 248, p <0.05, Power * 54 percent).  The excess lung
cancer mortality was further  analyzed by time since first exposure and by
process using both national and local rates.   While the excess among those with
30 or more  years since first exposure  persisted even  in  comparison to local
rates (19.1  expected,  SMR  = 152,  p  <0.05), the excess lung  cancer mortality in
the whole cohort was no longer significant when compared to local rates.  No
significant excess lung cancers were found among rock wool workers, but signif-
icant excesses  of  lung cancer  were found among all glass wool workers, espe-
cially those with  30 or more years  since  first exposure  compared to national
rates (17 observed,  9.8 expected, SMR = 173,  p <0.05, Power = 40 percent), and
among continuous  filament  workers with less than 10 years  since first  exposure
compared to national rates (9 observed, 4.0 expected, SMR  = 224, p <0.05, Power
= 15 percent).
     When lung  cancer  mortality among those with 30  or more years since  first
exposure were  analyzed  by year of followup, lung cancer mortality was  signifi-
cantly elevated  among the  total  cohort  for  either  period, up  to 1977 (14
observed, 5.7  expected, SMR *  246,  p <0.05)  or up to 1982  (29 observed, 16.8
expected, SMR  =  173, p <0.05).   Lung cancer  mortality was  significantly ele-
vated up to 1977 (7 observed,  2.8 expected, SMR *  251, p  <0.05),  but not up to
1982 in rock wool  workers,  and up to the  end  of 1982 in glass wool workers (17
observed, 9.8  expected, SMR -  173, p <0.05).   Although no  lung  cancers were
observed among continuous  fiber  workers  with 30 or more years since  first
exposure, the power was only 8 percent or less.
     When  lung cancer mortality  was further analyzed by  time since  first
exposure and by technological  phase  at first exposure,  significant  excesses
were found among rock wool workers of the early technological phase compared  to

'•••••• •>no°                          e-ftn         nRAFT--DO NOT QUOTE OR CITE

-------
 either  national  rates (10 observed, 4.7 expected, SMR = 214, p <0.05,  Power =
 18 percent) or  local rates  (3.9 expected,  SMR  = 257,  p <0.05, Power = 24
 percent),  especially among those with 30 or more years  since first exposure  (6
 observed,  2.0  expected,  SMR « 295, p <0.05, Power « 8 percent), and among glass
 wool  workers in  the  intermediate technological phase compared to national rates
 (68 observed,  48.8 expected, SMR = 139,  p <0.05,  Power = 93 percent), espe-
 cially  among those with 10 to 19 years since first exposure  (29 observed, 17.7
 expected,  SMR  * 164, p <0.05, Power * 57 percent).  The authors reported that
 among those with 20  years  since  first exposure, based on national rates,  a
 significant excess of  lung  cancer  existed in the early  technological  phase
 (SMR  -  254; 95  percent  CI =  122, 467, p <0.05), and 1n the intermediate phase
 (SMR  =  176, 95  percent, CI  = 88,  315, p <0.05), but not  in the late  phase
 (SMR  = 92, 95 percent,  CI = 49, 157).  They  also reported  a statistically
 significant (p <0.01) decreasing trend of SMRs  from early to late phase (data
 not provided)  and  an  80  percent power to detect  an 80 percent increased risk of
 lung  cancer among  the late phase subcohort.
      Within rock wool processes, a batch process used in the early technologi-
 cal  phase  at  four plants  in  three countries was considered to have had poten-
 tially  high historical  fiber levels.   Among  223  workers employed at these
 plants  during  the period of batch operation, investigators reported a  signifi-
 cant  excess of  lung  cancer  using national rates for those with 20 to 29 years
 (4 observed, 1.0 expected, SMR = 402, p <0.05) and 30 or more years since first
 exposure (6 observed, 1.8  expected, SMR =  330, p <0.05).
      When  lung cancer mortality among those with  20 or more years of  latency
 were  analyzed  by duration of employment per process, a significant lung cancer
 excess  was restricted to glass wool workers  employed from 1 to 4 years. (19
 observed,  9.7  expected, SMR * 197, 95 percent CI = 119, 308, p <0.05)  and  frorr.
"10 to 19 years (15 observed,  7.9 expected, SMR * 190, 95 percent CI = 106, 314,
 p <0.05).   The authors stated that duration of  employment did not appear to
 increase the risk  of lung  cancer.
      The  investigators concluded  that (1) the  elevated  mortality  from lung
 cancer  among rock wool workers was associated with employment during the early
 technological  phase  when dust suppressants were  not  1n  use;  (2)  after consider-
 ing potential  confounding factors, such  as  smoking and asbestos, the concen-
 tration of excess lung  cancer among  those with  the  highest  levels of  exposure
 to respirable  fibers points  to  MV as  a  risk factor;  (3)  no  increased  risk was
 found  in  the  late  technological  phase  after  full  introduction of  dust
 July  1988                          8-61          DRAFT—DO NOT QUOTE  OR  CITE

-------
suppressants;  (4)  the  absence of an excess of the  same  magnitude  present in
glass wool  workers is  consistent with the presumed low exposure to respirable
fibers; and (5) there  is no increased risk of mesothelioma.
     This  study was enhanced by pooling data  from the 13 plants, thereby
increasing  the number of deaths, analyzing  of  the data from first exposure
rather than first  employment, and adding 5 years to the followup.   On  the other
hand, although there was sufficient power to detect  lung cancer overall, the
low power to detect  lung cancer in glass wool workers when they were subdivided
into multiple  categories may  have impeded the detection of an increased risk of
lung cancer  in the early and  late phases  of  glass wool production.  Similarly,
the conclusion that no evidence exists for  an  increased risk of death  from
mesothelioma is weakened by the low power (17 percent  based on an estimated
maximum expected  value  of  three deaths) to  detect a 50 percent  increased
incidence of mesothelioma. "
     Simonato  et al. (1986b)  further analyzed the mortality  and cancer morbid-
ity experience of  the  European cohort of  21,967  male  and female (14.6  percent)
production workers.  The 21,967 production workers  consisted of 10,115 workers
in the rock wool/slag wool subcohort, 8,286 workers in the glass wool  subcohort,
and 3,566 workers  in the continuous glass filament  subcohort (Simonato et al.,
1987).   Expected  deaths and  incident  cancer cases  were derived by applying
national mortality rates adjusted  for age, sex, and  calendar-year to  person-
years at risk.   Correction factors for regional  lung cancer mortality  were also
used,  except for  two factories where no adjustments were made because  of small
national populations,  and  for two others where immediate local area mortality
rates were used.   The exposure assessment was based on the results  of  a histor-
ical environmental  investigation reported by Cherrie  et  al.  (1986)  and Cherrie
and Dodgson  (1986).  Simonato et al. (1987) estimated the highest  cumulative
dose in the early technological phase of  rock wool/slag  wool  processes to have
been in the range of 27 to 70 fibers/cm  .   After  a conversion  table was pre-
pared at  the  International Agency  for Research  on Cancer (IARC) to  allow
pooling of results over different ICO Revisions, deaths were reported  according
to the  8th  Revision of the ICD.  National cancer morbidity data were  reported
according to the 7th Revision of the ICD.
     The end of the followup  period varied by plant from 1981 to 1983.   Of the
combined 24,609 office  and production workers,  by the  end  of the followup
period,  83.3 percent were alive, 11.5 percent had  died, and 5.2 percent were
lost to  followup.    Since  vital status was not  reported for the  production
.!•••»» IQBO                          8-62         DRAFT—DO NOT QUOTE OR CITE

-------
workers, It was  not possible to determine the  percent  alive, dead, or lost to
followup.  However,  death certificates were obtained  for all  but  46  (1.7
percent) of the  2719 decedents who had been production workers.  Less  than 1
percent  of the  cohort had accrued person-years at risk 30 or more years since
first exposure with 20 or more years'  exposure to MMMF.  The average duration
of employment was  3.8 years  for rock wool workers and 5.6 years for glass wool
workers.
     Because the investigators  Included both males and females  in this analy-
sis, all data were presented for both  sexes.   Although this analysis focussed
on production workers, the investigators  also examined  the overall mortality  of
office workers compared  to  national  mortality  rates and found  a significant
excess  of  bladder  cancer (4  observed, 1.0 expected,  SMR = 396, 95 percent
CI =  108,  1014, p <0.05, Power = 7 percent)  and  significant deficits from
accidents,  poisoning,  and violence  (13 observed,  23.9 expected, SMR =54, 95
percent CI  = 29, 93, p <0.05, Power = 71  percent), and  all causes (117 observed,
158.6 expected,  SMR =74, 95 percent CI = 61, 88,  p <0.05, Power « 100 percent).
Among the production workers,  compared to national rates investigators found
significant excesses  from all  causes  (2719 observed,  2457.0 expected, SMR =
111, 95 percent CI  = 107, 115, p <0.05),  all malignant  neoplasms (661 observed,
597.7 expected,  SMR = 111, 95 percent CI  = 102, 119, p  <0.05),  lung  cancer (189
observed, 151.2 expected, SMR = 125, 95 percent CI « 108, 144,  p <0.05, Power =
100 percent),  and  accidents,  poisoning,  and  violence  (406 observed, 264.8
expected, SMR =  153, 95 percent CI «  139,  169,  p <0.05),  including  suicides
(110 observed, 84.6  expected,  SMR = 130, 95 percent CI  = 107,  157, p <0.05).
While there was sufficient power to detect lung cancer  or NMRD  among production
workers, there  was only 24 percent power  to detect laryngeal  cancer  and 42
percent power to detect buccal cavity and pharyngeal cancers.
     To  examine  the  mortality experience  of short-term workers, the data  were
analyzed by duration of employment of less than  or more than  1 year.  Among
short-term workers,  investigators  found  significant excesses  from all  causes
(741 observed, 504.08 expected, SMR = 147, p <0.05),  malignant neoplasms  (143
observed, 120.17 expected,  SMR * 119,  p <0.05, Power  = 80 percent), NMRD (36
observed, 26  expected, SMR  - 150, p <0.05, Power  - 71 percent),  digestive
diseases (40 observed, 23.39 expected,  SMR * 171, p <0.05,  Power =  66 percent),
violent  deaths  (200 observed,  82.64 expected,  SMR  = 242, p <0.05), and other
known causes  (80 observed,  60.15 expected, SMR = 133,  p <0.05).  Compared to
local rates,  these excesses  were reduced, but  remained significantly elevated
July 1988                          8-63         DRAFT—DO NOT QUOTE OR  CITE

-------
at the p  = 0.05 level in a  subsequent update by Simonato et al. (1987):   all
causes (SMR =  142);  malignant neoplasms  (SMR =  115);  NMRD  (SMR  = 146); diges-
tive diseases (SMR = 165); and violent deaths (SMR  = 227).  Among those employed
for more  than  1 year. Investigators found significant excesses  from malignant
neoplasms  (518  observed, 475.23  expected,  SMR  *  109, p <0.05,  Power = 100
percent),  lung  cancer (154  observed,  120.31 expected, SMR * 128, p <0.05,
Power e 100  percent), and violent deaths  (206 observed, 176.07 expected,  SMR =
117, p <0.05).  While the highest risk of death among short-term workers was
from violent  causes  of death, the highest  risk among those  employed  for at
least 1  year was  from lung cancer.   According to the authors, short-term
workers included  from 9  of  the  13  factories constituted  one-third  of the
cohort.
     Next, the investigators analyzed lung cancer mortality by time since  first
exposure and found  significant  excesses  from all causes  among those with  less
than 10,  10  to 19, and  20  to  29 years since first exposure, all malignant
neoplasms among those with  10 to 19 and  20  to 29 years since first exposure,
and lung  cancer  among those with 30 or more years  since first  exposure   (29
observed,  16.8  expected,  SMR =  173,  95 percent CI = 116, 248, p <0.05, Power =
54 percent),  which remained significant even when compared  to local rates  (19.1
expected,   SMR = 152,  p <0.05, Power = 57 percent).   Although there  were no
significant excesses  from NMRD,  there was only 71 percent power to detect NMRD
among those  with 30  or  more years  since first exposure.   Except  for lung
cancer,  the authors  stated  that no relationship existed  between all malignant
neoplasms and time since first exposure.
     When  lung  cancer mortality compared  to both national  and local rates was
further analyzed  by  time  since first exposure  and by process, significant
excesses were found among glass wool workers (93 observed,  73.3  expected,  SMR =
127, p <0.05, Power = 99 percent), especially among those with 30 or more  years
since first exposure  (17 observed, 9.8 expected, SMR = 173, p <0.05, Power = 40
percent) when compared to national,  but not local,  rates and among continuous
filament  workers  with less  than 10 years since first exposure compared  to
national  rates  (9 observed,  4.0 expected,  SMR =  224, p <0.05, Power =   15
percent).   When  lung  cancer mortality was analyzed by factory and by  process,
significant excesses  were found  in  Factory 4 (rock  wool) compared to both
national  (7 observed, 1.9 expected,  SMR  *  360, p  <0.05) and local rates  (2.3
expected, SMR = 300, p <0.05); in Factory 9 (rock wool) compared to  local  rates
(8 observed, 2.9  expected,  SMR  = 275,  p <0.05); and 1h Factory  10 (glass  wool)
v.v. -.000                          6-64         DRAFT—DO NOT QUOTE  OR CITE

-------
compared  to  national  rates (71 observed, 50.6 expected, SMR = 140,. p <0.05).
When  lung cancer mortality was analyzed by occupational category among those
with  20 or more  years  of latency,  investigators found a significant excess
among  production/preproduction  rock  wool  workers compared to local rates (11
observed,  5.3  expected, SMR = 209, p <0.05, Power *  28 percent).   For  glass
wool  workers  in all job categories combined, •  significant  excess  was  found
only  when compared to national rates (46 observed, 33.1 expected, SMR = 139,
p <0.05,  Power = 81 percent).  When the  excess  lung  cancer among rock wool
workers with  20 or more years of latency were analyzed by duration of employ-
ment,  a significant excess was found among those employed in early and interme-
diate  technological phases  compared to both  national (21  observed,  10.2
expected,  SMR = 206, p <0.05, Power = 37 percent) and local rates (9.4 expected,
SMR s 223, p <0.05, Power =  34 percent), showing a  dose-response  relation-
ship,  particularly among those employed from 1  to  4  years when compared to
local  rates (7 observed, 2.7 expected, SMR = 257, p <0.05, Power = 11 percent).
Similarly, when the excess lung cancer among glass wool workers with 20 or more
years  of  latency was analyzed by duration of employment, significant excesses
were  found among those employed from 1 to 4 years (19 observed, 9.7 expected,
SMR =  197, p <0.05) and 10 to 19 years (15 observed,  7.9 expected, SMR = 19C,
p <0.05)  only  when compared to national rates.   All  of these  deaths  occurred
among  those first  employed in the early and intermediate phases.   Those employed
from  1 to 4 years during  the  early  and intermediate  phases  had  a  significant
excess  only  when compared to national  rates  (19 observed,  10.0 expected, SMR =
191,  p <0.05).   While  sufficient  power existed to detect lung  cancer  among
those  with 20 or more years  since first  exposure,  the insufficient power to
.detect  lung cancer  in each of the duration of employment categories may account
for the absence  of trends.
     When  lung cancer  mortality  was analyzed  by latency and technological
phase,  the excess was concentrated among rock  wool workers  first  employed  in
the early phase with 30 or more years of latency when compared to local rates
(6 observed, 2.0  expected, SMR * 295, p <0.05, Power « 8 percent).   The authors
reported,  for  those with 20 or more  years of latency, a significant decreasing
trend  for lung cancer from the early to late technological phases  of rock wool
production (chi-square   s  6.5 based on national rates and 9.8 based on local
rates).   Because none  of the SMRs was  significant, however, the most that can
be concluded  is that the  data  suggest  a  trend.   For  glass wool workers, the
significantly  elevated  lung cancer was restricted  to those  first employed  in
i..i,. ieoo                          p-ce,          nctcr—nn  wnr nunrp f)R CITE

-------
the  intermediate  phase with 10 to 19 years  of latency compared to national
rates.  Only  one  case of mesothelioma was  found.  Because referent values were
unavailable, the authors estimated expected values of mesothelioma  ranging from
one to three deaths.
     In addition  to lung cancer,  the investigators  analyzed bladder cancer by
latency and by  process.   The Investigators reported  a  significantly Increasing
trend  (p <0.05) with  Increasing latency among  rock wool workers.  Because none
of the  SMRs was significant, however,  the most that can be concluded is that
the data are suggestive of a trend.   Concerning other elevated causes  of  death,
mortality from  laryngeal  cancer was  not significantly  elevated among the total
exposed cohort  (9 observed,  6.3 expected,  SMR  = 143, 95 percent CI « 66, 272).
However,  not  only  did all  nine  deaths  occur 1n male production workers
(Simonato et a!.,  1986b),  but Simonato et  al.  (1986c)  reported that six  of the
nine  deaths  occurred  in  the Italian FG  wool  cohort previously studied by
Bertazzi et al. (1984,  1986).   This  excess from laryngeal cancer was  statisti-
cally  significant (6  observed, 1.5 expected,  SMR =  400,  p  <0.05,  Power = 8
percent).   In  the  other .plants,  there  were  3 observed deaths versus  4.8
expected with only 19 percent power.  Simonato et al.  (1986c) concluded that
these data confirm the previous findings  by  Bertazzi et al.  (1984, 1986) of an
elevated laryngeal cancer risk in this  Italian FG  wool  cohort.
     When cancer   incidence  was examined, a  significant  excess of buccal
cavity/pharyngeal   cancer was found  (28  observed,  17.0 expected,  SIR ~ 164,  95
percent CI =  109, 238, p <0.05, Power  = 57 percent).  Analysis of these  cases
by latency  and by  process  revealed significant  excesses  in the rock wool
workers (24 observed,  12.2 expected, SMR = 197,  95  percent,  CI  = 126,  2S3,
p <0.05, Power = 47 percent), especially among those with  10 to 19 (11 observed,
4.6 expected, SMR = 238,  95  percent  CI =  119,  426, p <0.05)  and 20 to 29 years
of latency (7  observed,  2.6 expected,  SMR =  272,  95 percent CI = 109,  55S,
p <0.05).   In the subsequent study,  however, Simonato et al.  (1987)  did not
explain why  there  were  two  less  observed cases of buccal  cavity/pharyngeal
cancer.  Nevertheless,  the  excess  remained  significant (26  observed,  17.0
expected, SMR = 153,  95 percent CI = 100,  224, p  <0.05) especially among those
with 20 to 29 years of latency (9 observed, 3.9 expected,  SMR = 231, p <0.05).
     Although there were  no  trends when lung cancer incidence was analyzed  by
latency and by  process,  there was low power  in each of the  four  categories
since  first  exposure.  Similarly, when  further  analyses were  performed by

July 1988                          8-66         DRAFT—DO NOT QUOTE OR CITE

-------
technological phase  at  first exposure,  there were no visible  trends  with  such
low power.  Moreover, analysis of lung cancer for the few cases  where histology
was reported by country did not demonstrate any patterns.
     To  analyze  potentially confounding variables,  Simonato  et al.  (1986c)
presented additional  data  whereby  lung  cancer aortality was analyzed in rela-
tion  to  the use  of slag  as  a raw atatefial,  bitumen or pitch as a dust
suppressant, minor use of asbestos in some plants, and formaldehyde.
     First,  because  of  potential  exposure to arsenic,  lung cancer mortality
among rock  wool  workers was analyzed  by latency and  by  use of slag,  especially
copper slag, as a raw material.  Although the investigators  stated that the use
of slag  appeared  to be  associated with increasing lung cancer mortality after
20 years  of latency, the data did not  support  this  claim.  In  the subsequent
1987  study,  this analysis  was confined to those  with  20 or  more years  of
latency.   Simonato  et al.  (1987)  found  an increased risk of lung cancer  among
those with 20 or more years since first exposure to rock wool  during  the  use of
slag  (23  observed,   12.2  expected,  SMR =  189,  95 percent  CI « 120,  283,
p <0.005),  especially ««*ong t^o*e -witt,  30 cr inort ye-ars  since  first exposure
(9 observed, 3.5  expectad, SMR = 254, 95 percent CI « 116,  481, p <0.0 1) and
among those  with  20 or  acre years since first exposure during the use of  slag
without copper (18  observed,  10.2 expected, SMR « 176, 95  percent CI =  104,
278, p <0.05).   Furthermore,  when  Simonato «t al. (1987) examined the use  of
slag  as  a  raw material during the  early  technological phase, there  was  a
significant excess of lung  cancer among workers in factories with limited  or no
use of slag (6  observed, 1.8  expected,  SMR = 341, p <0.0 1),  but not in  those
factories which  used only slag as a  raw material  (4 observed, 1.4 expected,
SMR = 295).
     Second, because of potential exposure to polycyclic aromatic hydrocarbons,
lung  cancer mortality was analyzed by  latency  and  use  of bitumen as  a dust
suppressant.  Although  the  investigators  claimed that there  was no  clear
difference with or without  the presence of bitumen after 20 years of latency in
either rock wool or -glass wool workers, among glass wool workers with potential
exposure to bitumen, there was a significant excess of  lung cancer (56 observed,
38.5  expected, SMR  = 145, 95 percent CI =  110,  189,  p <0.05),  especially  among
those with  10 to 19 years  of latency  (23 observed, 12.8 expected, SMR = 183, 95
percent,  CI = 114,  270,  p <0.05).   In  the subsequent  1987 study  where this
analysis was confined to those with 20  or more years of latency, no  significant
excesses were found  (Simonato  et al., 1987).
 + i+i*r**                          f\  f^         nn J» tTT_ _r\^ *ir«T rsitnTr  Att ^ T TT

-------
     Third, when  lung cancer and potential  exposure to asbestos were examined,
there was a significant excess among rock wool workers with 30 or  more years  of
latency without  the potential presence of asbestos (10 observed, 4.6 expected,
SMR = 218, 95 percent CI = 105, 401, p <0.05) and increasing SMRs  with increas-
ing latency in the  same group.  In the subsequent study,  the excess  was  slightly
higher  (SMR «  242, p <0.01) when confined  to  those with 20 or more years  of
latency (Simonato  et a!.,  1987).   A similar excess was apparent in  glass wool
workers (17  observed, 9.4  expected,  SMR * 181, 95 percent CI = 105,  290,
p <0.05).   One would expect to observe these excesses because the presence of
asbestos was a criterion for excluding a plant from participating  in the study.
However, in the  subsequent study, this excess  was  no  longer significant (SMR =
142) when  analysis was restricted to those with 20 or more years since first
exposure (Simonato  et al., 1987).
     Finally,  the  investigators  examined the potential effect of  formaldehyde
on lung cancer mortality and found no  excess  mortality  among  those with high
(up to  5  pptn)  or  intermediate levels of formaldehyde.  In fact, a significant
excess was seen  among those with low  levels  of  exposure to formaldehyde (61
observed,   62.13  expected,  SMR  = 130,  95 percent  CI  « 104, 162, p <0.05),
especially those  with 10 to 19 years  of latency  (38  observed,  22.6 expected,
SMR =  168,  95  percent CI =  119,  231)  and those  never exposed  with  30 or more
years of  latency (11 observed, 5.2 expected,  SMR  = 208,  95 percent CI  = 104,
373, p  <0.05).   However,  in  the subsequent study where this  analysis was
confined to those with 20 or more years of latency,  no significant excesses
were found (Simonato et al., 1987).
     The investigators concluded  the  following:   (1)  no excess mortality from
NMRO was  found  in any of three production processes;  (2)  an elevated mortality
.from lung  cancer was associated with  employment  during the  early  technological
phase of rock wool  production (exposures such as smoking or previous employment
were unlikely  to  account  for  this  excess); (3)  the  concentration of a lung
cancer  excess  in the group with  the potential for highest levels of MMMF dust
Indicates  that the association may be of a  causal  nature;  (4)  since respirable
fibers were a  significant component of workroom air,  1t 1s possible that MMMF
exposure during  the early production phase of rock wool,  either alone  or in
combination with other  factors,  may  have  contributed to  the observed  lung
cancer  excess  (no  excess  was observed  after  the  full  introduction of  dust
suppressants); (5) there was no  excess  of  the same  magnitude  in glass wool
workers; and (6)  there is no evidence of an increased risk  of mesothelioma.
     i ooo

-------
     The  limitations  listed for the previous study (Simonato et al., 1986a)
such as  low power in wany  subcategon'es, are pertinent  here  as well.   With
improved ascertainment of cause of death from death certificates and  additional
followup,  additional  lung  cancer  deaths  among  glass  wool workers might be
detected.
     Two additional epidemiologies! studies  were conducted in  European workers.
These studies were not part of the category of  European cohorts reviewed by
Saracci et al. (1984) and Simonato tt al.  (1986a,b,c,  1987).
     Moultn et al.  (1986)  conducted a historical prospective  morbidity study
with 1374  active  or retired male employees  who  had worked in  a  French FG wool
plant for at least 1 year between January 1, 1975,  and December  31, 1984.  From
company  records,  the  investigators  assembled the  cohort  and reconstituted
workers'   occupational  histories in  the  factory from  date of  hire.   After
computing person-years  at  risk of dying for this period, the  expected  numbers
of cancer  cases  for calculating SIRs were  derived  from  French  regional  age,
sex, and calendar-specific cancer incidence  rates for  5-year age groups  for  the
period 1975 through 1984.  Cancer cases in this  cohort were identified  from  the
company's social insurance records.  Diagnoses were ascertained  by  histologies!
reports,  hospital  records,  or from attending physicians'  reports.  Of the 1374
cohort members,  101  (7.35 percent) were lost to followup.  These individuals
were considered  to be  alive and contributed to person-years through  1984.
Cancer cases were coded according to the 8th Revision  of the ICO.
     Production  at  this FG thermal insulation factory started  in 1940.   The
manufacturing  processes generate  discontinuous FG  fibers coated  with a
thermosetting binder  consisting of water, phenol,  formaldehyde, ammonia, urea,
and oil.   The  mean diameter of the fibers produced at this factory  is  6.4 uir,;
30 percent  have  diameters less than 3 pm,  and  10  percent have  diameters less
than 1 urn.   Two exposure  monitoring surveys performed  1n 1981 identified
occupational groups according to place of work.   Results of personal  monitoring
samples  from  the exposure monitoring surveys showed low fiber concentrations:
the  average respirable  fiber  concentrations  were less  than  0.2 fibers/cm  .
Tobacco  consumption  was estimated from Information on 966 active employees  in
1983.  No data on alcohol consumption could be obtained.
     In  1984,  the  mean age  of  the cohort was 44.4 years (SO =  11.3), the mean
length of employment  was  16.6 years  (SO  s 8.8),  and 25  percent  of  the
person-years  had  accrued among those with 20 or more  years since first  employ-
ment.  Results of the  inquiry on tobacco consumption indicated  no difference in
July 1988                          8-69         DRAFT—DO  NOT QUOTE OR CITE

-------
tobacco consumption between  the  cohort and the French male population of the
same age in 1979 and no differences inside the factory according  to  occupation.
     Of 41  cancer cases  identified among the 1374 cohort members, 19 were
located  in  the  upper respiratory  or  alimentary  tract.   Histopathological
reports available  for six cases  found squamous cell  carcinomas.  Of these 19
cancer cases,  9  were  buccal  cavity cancers, 5 were pharyngeal cancers, and 5
were laryngeal cancers.   Compared  to French regional  cancer incidence rates,
there was  a significant  excess  of  upper  respiratory and alimentary tract
cancers (19  observed, 8.71 expected,  SIR *  218,  95 percent CI = 131, 341,
p <0.005, Power  =  33  percent), particularly from  buccal  cavity and  pharyngeal
cancers (14 observed,  4.41 expected, SIR * 317, p <0.001, Power = 22 percent).
When the  upper  respiratory  and  alimentary tract cancers  were examined by
specific cancer  site,  there  was  a  significantly elevated incidence of buccal
cavity cancer  (9  observed, 2.99 expected,  SIR =  301, p <0.005, Power =  17
percent).   However, the  increased  incidences  of laryngeal  cancer (5 observed,
2.17 expected,  SIR - 230, Power = 11 percent) and  pharyngeal  cancer  (5  observed,
3.57 expected, SIR =  140, Power =  17 percent)  were  not significantly elevated,
perhaps due to low power to detect tumors at these sites.   In addition, although
there was no significantly elevated lung  cancer incidence, the power to detect
lung cancer in this cohort was only 32 percent.
     When the  cohort  was  divided into  administrative  staff, maintenance staff,
and production workers,  the  excesses of  upper  respiratory and  alimentary tract
cancers were confined to  production workers, particularly among those  exposed
for more than 10 years (Table 8-4).
     The investigators concluded that (1) a significant excess  of cancer  of the
upper respiratory  and alimentary  tract  was observed, which appears to be
associated with duration of exposure to FG wool; (2) the  excess may  be  underes-
timated because  of the 101 workers  lost to  followup; and (3)  the increase in
cancer incidence cannot  be  explained  by an  excess of alcohol  or  tobacco
consumption.
     The evidence  of  a  causal relationship between  upper  respiratory and
alimentary tract cancers  and exposure  to FG wool was strengthened  by data on
smoking histories  that indicated no significant differences  1n tobacco  consump-
tion between the cohort  and the French male  population  or  between  different
segments within the cohort itself.   Despite low power to  detect a 1.5-fold risk
of buccal cavity cancer,  there was  a  significant excess  of cancer from  this
site alone.   All other studies have reported observed and expected deaths  from
July 1988                          8-70         DRAFT—DO NOT QUOTE  OR  CITE

-------
               TABLE 8-4.   CANCER INCIDENCE ANALYSIS  BY  SITE AND
                             DURATION OF EXPOSURE
JJpj>er respiratory and alimentary tract cancers
Group
Administrative staff
Maintenance staff
Production workers
Unexposed
Exposed 1-9 years
Exposed 10-19 years
Exposed >70 years
Total
Observed
.
1

1
3
7
7
19
Expected
••
1.41

0.60
1.49
2.30
2.10
8.71
SIR
0
71

167
202
304**
333***
218***
95% CI
0-461
2-398

4-929
41-584
122-627
134-687
131-341
Power (X.)
•
6

6
6
14
10
33
 ""Significantly different from regional  cancer incidence rates:   p <0.01.
***Significantly different from regional  cancer incidence rates:   p <0.005.
Source:   Moulin et el. (1986).

the combined category of buccal cavity and pharyngeal  cancers.   In addition,  as
the investigators pointed  out,  the large number of workers  lost  to followup
(7.4 percent) might account' for an underestimation of  the true  risk of laryngeal
and pharyngeal  cancers.   However,  this also  holds  true  for lung cancer anc may
account for the absence of a significant excess of cancer from  this site.
     Engholm  et al.  (1987) conducted a  historical prospective  mortality  and
morbidity study and  a nested case-control study designed to control  for con-
founding variables, such as asbestos and silica exposures, smoking, and popula-
tion  density.   The cohort  consisted of  135,026  male Swedish  construction
workers who  had first participated in regular health examinations from  1S71
through 1974.   It was estimated that approximately 80  percent  of all workers
undergo such examinations.  There was no discussion of past or  present exposure
levels to MMMF.   Expected numbers of death were based on mortality statistics
for Sweden by age, sex, cause of death, and calendar year.  Expected numbers  of
cancer incidence  were based on rates for Sweden by age,  sex, cancer diagnosis,
and calendar year.
     Vital status was determined through record linkage  of  computerized data
bases.  The  cohort  was followed for mortality to  December  31,  1983, and for
cancer morbidity to December 31, 1982, by linkage to the Swedish  national  vital
statistics  file and  the  Swedish Cancer  Registry.   The  average followup for
mortality was 10.4 years  and  for cancer morbidity was  9.4  years.  The  8th
     TQRR
8-71
DRAFT-DO NOT QUOTE OR CITE

-------
Revision of  the  ICD was used to code mortality, and the 7th Revision  was  used
for cancer morbidity.
     By the  end  of the mortality followup period,  December 31,. 1983, 127,009
(94.1 percent) of  the 135,026 cohort members were  alive; 7,356 (5.4 percent)
had died.   Eleven  subjects lost to followup were excluded  from analysis.  It
1s  unclear  whether the 661  (0.5 percent) emigrants accrued person-years or
whether they were excluded from analysis.
     Overall mortality 1n this  cohort  was significantly less  than in the
general Swedish  population (7,356  observed,  10,896 expected, SMR *  68,  95
percent CI = 66,  69, p <0.05).  The healthy worker effect was further demon-
strated by the significantly lower deaths  from all  cancers (2153  observed,  2553
expected, SMR = 84, 95 percent CI * 81,  88, p <0.05),  lung cancer (444 observed,
513 expected,  SMR *  86,  95  percent CI « 79,  95,  p <0.05),  and  NMRD (193
observed, 418  expected, SMR  = 46, 95 percent  CI = 40, 53, p <0.05).  For
overall mortality  and for  each of the above causes of  death, there was 100
percent power to  detect a  50  percent increase  in mortality.   The  only signifi-
cant excess was from industrial accidents  (data not reported).
     When cancer  incidence was examined,  there was  a significant deficit from
all cancers  (3910  observed,  4067 expected, SIR = 94, 95  percent  CI = 91,  97,
p <0.05, Power =  100 percent), but a significant excess  from cancers of the
pleura, all  of which were classified by  the Cancer Registry  as mesotheliomas
(23 observed, 11  expected, SIR = 213, 95  percent CI * 135,  320, p <0.05, Power
= 39 percent).
     In the  nested case-control study, cases  were  all  subjects  who after a
first health examination had been diagnosed with primary cancer of the respira-
tory system.  For each of the 518 cases, five controls  were  selected from among
those who had not developed respiratory cancer with Individual matching for age
and year  of first  examination.   Controls were selected from a file  of  all
subjects who had developed respiratory cancer,  sorted by year of  first examina-
tion, and by date of birth.    For  each  case,  controls  were first selected at
random from among those with the same year of examination and year of birth and
alive up to  the  time of diagnosis  of respiratory cancer  in the corresponding
case.   Where five controls could not be  found in this group, controls were
drawn  from  those born  1  year  earlier  than the case.   A subject could  be
selected as  a  control for only one case.   Cases and controls  were then classi-
fied into categories based on self-reported exposure to MMMF and asbestos and
on  estimates  of  average  intensity  of  exposure to  MMMF and  asbestos  in  160
July 1988                          6-72         DRAFT—DO NOT QUOTE OR CITE

-------
different types of  construction  tasks.   To control  for confounding variables,
detailed smoking habit Information was obtained and  each  subject was classified
according to population density.
     A total of 18  cases of Individuals diagnosed with a primary tumor of the
respiratory tract  included 23 cases  of mesothelioma  and 424 cases of lung
cancer; 16  subjects  diagnosed  with lung cancer following primary diagnosis of
another  cancer were  excluded.   These  Investigators  found  that  neither
self-reported exposure to  HMMF nor exposure to asbestos  was associated with
lung cancer.   In  fact, subjects  in the  cohort  appeared to have been unaware of
their exposure to asbestos.  The strong association between  asbestos exposure
and MMMF exposure made  it difficult to separate the effects of the two expo-
sures upon respiratory cancer.   Subjects categorized as heavily exposed to MMMF
did not  have  a significantly  higher incidence  of  respiratory cancer than
unexposed subjects.   The  investigators  concluded that  the  results did  not
support a relationship between MMMF and increased risk of lung  cancer.  Because
the risk  appeared to be related  to  asbestos  exposure rather  than to MMMF
exposure, they concluded that  the results appeared  compatible with the known
effect of asbestos exposure on lung cancer incidence.
     Although  there  appeared  to  be a  high  degree  of followup,  failure to
determine the vital  status of the 661 emigrants could  result  in an  underestima-
tion of the risk  of lung  cancer  and  NMRD.   Furthermore, additional deaths or
tumors could have occurred among 20 percent of  the construction workers who
chose not to participate in the  regular  health  examinations,  thereby further
underestimating the  risk of lung cancer and NMRD.   These potential biases were
minimized, however,  because even though only 5 percent of the  cohort had died,
there was sufficient  power to  have detected excesses  from lung cancer and NMRD
in this cohort.
8.1.1.2.4  Epidemiological study combining U.S.. Canadian,  and  European workers.
In an overview of the hazards of MMMF following exposure  in  U.S.,  Canadian, and
European workers,  Doll (1987) collated the results from the  three major cohorts
of MMMF workers previously studied by Enter!ine et  al.  (1987), Shannon et  al.
(1987), and Saracci  et al. (1984) and Simonato et al. (1987).  Taken together,
these  studies  provide data  on 7,682 deaths In 41,185 workers in  the  U.S.,
Canada, and Europe,  many of whom were first employed  in  the  1940s.  To include
workers with presumed exposure to MMMF in production plants,  i.e., production
and maintenance employees, Doll  (1987)  excluded office workers,  who had previ-
ously been  included in  the Canadian  cohort.   Because the European work force
July 1988                          8-73         DRAFT—DO NOT  QUOTE OR CITE

-------
included women  and because  it  was  not possible to determine men and women
employed  in European plants  for more  than  1 year,  Doll  (1987) excluded
short-term workers, i.e.,  those exposed for less than 1 year.   This  served to
make the European  data  more comparable with the U.S.  data.   Expected deaths
were based  on  national  or  local rates  when available  for the U.S. and European
cohorts and  provincial  rates for the Canadian cohort.  Doll (1987) pointed out
that the SMRs  for  specific causes of  death were  low  because  expected deaths
prior to 1950  were based on the national mortality rates for  the early  1950s.
This resulted  in a potential  underestimate of causes  of death that became more
common, such as  lung  cancer, and a potential underestimate of  diseases that
became less common (Doll, 1987).
     All workers included  in the study were observed  for at least 5 years.  Of
the 41,185 workers, 7,862 (19 percent)  had died.   According to Doll  (1987), all
three studies  succeeded  in determining the mortality  experience of  96 percent
of the  defined populations,  i.e.,  4  percent were lost to  followup.  Death
certificates were obtained for all but  191 (2.4  percent) of the 7862  decedents.
     U.S.,  Canadian,  and European workers experienced mortality rates similar
to those expected for all causes, cancers other  than lung cancer, and nonmalig-
nant diseases other than NMRD (Table 8-5).  Lung cancer mortality, however, was
significantly elevated for  the  U.S.  cohort compared  to  national  rates  (SKR =
116), for the Canadian cohort compared  to provincial  rates (SMR = 174),  for the
European cohort  compared to  national  rates  (SMR * 128), and  for the three
cohorts combined (SMR = 121).
     To control  for potentially  confounding  variables,  such  as smoking  and
other etiological  agents that may vary with geography, Doll (1987)  used local
mortality rates  available  for  the  U.S.  and  European cohorts to calculate
expected deaths (Table 8-6).  When  compared  to  local   rates, the excess lung
cancer mortality was  reduced below the level  of significance  for the U.S. and
the European cohorts  but remained significantly  elevated for the combined U.S.
and European cohorts (SMR * 109).
     Next, Doll (1987)  analyzed the lung cancer mortality  by  process,  i.e.,
rock wool  or slag  wool  (MW),  fibrous glass  wool  (FG wool),  or continuous
fibrous glass filament (continuous FG filament).   For  this analysis,  Doll (1967)
categorized U.S. workers in plants where both FG wool and continuous FG fila-
ment were made concurrently with workers making  FG wool only.   It appears that
July 1988                          8-74         DRAFT—DO NOT QUOTE OR CITE

-------
                            smart? cw»sr-swrinr wmiAiHv 'on Nmniir.MnN wtwKiir, tnnN tin  H.V. tANAniAN. uranrtAN. AWI cnmimf> conoirts
United Stales C.uia.l.i

ise of death Ohs.
causes 4906
>q canter 3/0
wr raiMrrrs
'ticliiiliixi lour
•snlhel ionias) 677
«er respiratory
isease 314
ler disease 30/5
lernal causes 405

-------
the greatest risk of lung cancer appears in MW workers  (SMR =  128,  p <0.01), an
intermediate, nonsignificant  risk in FG  wool workers  (SMR = 110),  and an
absence of risk in continuous FG filament workers (SMR  « 93; see  Table  8-7).
     To determine whether  the  lung cancer excess could be ascribed to  occupa-
tional hazards,  lung cancer mortality  in  the  three  cohorts was  analyzed by
process and  time  since  first exposure (Table 8-8).   Compared to local rates,
there was a  significant excess  of lung  cancer  among MW workers for all periods
since first  exposure  (SMR  = 128,  p <0.01), particularly among those with 30 or
more years since first exposure (SMR = 148, p <0.05).
     To compare relative excesses among those holding  different  jobs within  a
plant, Doll  (198?)  presented data from Enterline et  al.  (1987) on  respiratory
cancer among workers  exposed to small-diameter FG fibers compared  to  those  in
the same  plants,  as well as those employed in other  FG plants, not exposed  to
small-diameter FG fibers.  No significant excesses were found  among those
        TABLE 8-6.   COMPARISON BETWEEN U.S.  AND EUROPEAN SMRs FOR LUNG
                 CANCER DERIVED FROM NATIONAL AND LOCAL RATES
Cohort
location
U.S.
Europe
Combined
Observed
deaths
370
189a
559
National
Expected
deaths
319
151
470
Rates
SMR
116**
125**
119***
Local Rates
Expected
deaths
339
172
513

SMR
109
110
109*
  alncludes those with less than 1 year of employment.
  *Significantly different from comparison group:   p <0.05.
 **Significantly different from comparison group:   p <0.01.
***Significantly different from comparison group:   p <0.001.
Source:  Doll (1987).

exposed  to  small-diameter fibers,  and no increasing  trends  of respiratory
cancer were  associated  with  increasing time  since first  exposure  when compared
to  local  rates  (Table 8-9).   However,  the small number of deaths  (22 observed)
and  insufficient power  (57 percent) to detect respiratory cancer among  those
exposed to small-diameter fibers precluded meaningful analysis of these data.

July 1988                          8-76         DRAFT—DO NOT QUOTE OR CITE

-------
                                     tMUl fl-r.  HINT. CANTIR WIRIAI IIY ItY PHOtl S%
Rock or Slag Wool
ml
it inn Otis.

iria
e
l» 1 ant s
<|t. SM«
74 YO
HI. 9
90. 1
144 . %
or,
01
l»9
l/4»
KM
III)


Pnwf r
too
i;
inn
iwi


(III', (up.
r.4 r.'i 6
..
IS 15. 5
f> 114 9


il.HW»nts
SMR
9?
--
9/
91



(X)
98
..
55
99


•ce: Doll (1987)
1AHIE fl-fl.
irs
n e
•si
'Mire
0
19
?'
r
per i ods



Obs.
19
43
4?
37
141
HINT. CANCER MORI All IV

Rock or

E«p.
in. 6
33.1
33.6
75.0
110. Z

Slag Wool

SMR
10?
130
175
148*
i?a"
AMONG THE


Power
(X)
60
81
83
74
100
iiinrf coMntNio COINIRIS nv



Otis.
76
88
140
175
1/9
Proress
1 ilirmis (ilA

f«P
37 I
77.9
179. f.
105 0
344.5

PROCISS ANO

ss Won!

SMH
ni
113
Km
119
no
Power
(*)
/4
99
too
inn
ion
1IHI SINCE rtRSI

Glass E

obs. r»p.
IS 10.9
1', 74.6
Jh 37.4
11 If. 5
79 H5.0
EXPOSURE

llanenls

SNR
138
61
111
79
93



Power
(X)
„
71
76
59
100
;pir*t«ry cancer for U.S. plants.
iqntricantly different fro* local »orlallty rate*:  p <0.05.
i«|nificatitly different fro* local Mortality rates:  p 
-------
                  tAmr ni    BISNRAKWV r.wir.m MINIM m nv IXI-OMIRI  in  '.MAIJ-IHAMMIN WAV, wmii  nnrns ANII tint Mucr HUM  IMPIOVMNI
                     Ivpr exposed in
•••»          niip nf four plants  MitufKluHmf     Movri- r»piiM-s.       I «p      '.MH       (t)      01.s       Op.      SMR      (X)      Obs.       l«p      SHU       (t)
10             I        I M       f.1       II                            -
II             /        S.«1      17H       ?l                            -
">             n        / ».?      iii'i       3f.                            -
•               h        3.01      |9fl       IB                            -
              ??       II. 1       |?4       5;         mi      I7S ?      I OS       100
::•»:   I>«M  (I9W7); pKpPclprt  values for respiratory  canrrr t.,i-,c(l mi  lo«..il imirt.il il.y r.il«-s.
-
-
-
119
4?
96
9/
94 ?4S
39.fi
nn.r
H/.4
2?6.9
lor.
ion
in
ion
im
ion
ino
too

-------
     To compare relative excesses of lung cancer among those  exposed  to  differ-
ent amounts  of  MMMF,  Doll  (1987)  presented data  from  Simonato et al.  (1987) in
which  significantly elevated  SMRs were found in workers  employed  in  the early
and  intermediate  technological  phases of the MW industry (21 observed, 9.4
expected, SMR * 223,  p 
-------
                            tAtiir n-in   siilctrn tAMrin MOIMAIIIY urn u.v Ktm MMIOPMN r.oiicwis
Mle
a
iiral cavity/
'iiaryttx
i»|>httf]iis
.nmach
iliin
•i linn
nynx
os 1 ate
l.idili-r
>nrs
•itl«>*i a
liter lymphatic
.mil lii'm.ilo-
Itoielic
,i
Ofos.
370

35
75
47
81
36
15
68
79
2
35


68

f.p
319

?8. T
71.9
43.5
91.0
30.8
14.4
63.6
77.6
5.7
38.9


57.1

SMR
II6«*

17?
114
108
89
117
104
107
105
35
90


119

P
100
t
75
65
91
99
80
41
97
7?
74
88


96

Ol.s
189

11
14
f,4
37
3.1
1
31
74
r,
I/


33

1.,..
IM 7

10.6
II 0
Ml II
10 7
77.0
(>. 3
31.4
17.9
1.1
71 1


79.5

'.Mil
175"

171
KIM
'II
91
177
M.I
91
1 11
IM
HO


II7

P
lim

41
49
'III
ll'l
/I
74
117
59
71
66


76

Dliv
559

40
.19
III
tin
!>')
74
99
5.1
8
57


101

I.,,.
4/07

39 3
14.0
117. 1
111. 1
58 0
70. 7
97 1
45. .1
9.6
60 5


86 3

'.Mil
118-**

177
117
99
90
119
116
102
117
8.1
86


117

P
100

Rfl
87
100
100
96
67
99
97
37
97


99

            i ohorl  includes those employed less than I year; U.S. anil I uri»|MMn exper.trd vslup% lor limq cancer based
  on national  rales.
**Si(|nif icanlly different frow CMpartson group p <0.0).
**Sii|nif icantly different fro» conparison group p 
-------
                                                               IAIIII  n n   MnniAtm i ROM NMRti nv cmmnl


Cause of Oe.ilh
Acute upper respiratory
inter V. inn
Influrn/a ami pneumonia
other respiratory
disease (includinn,
piMMMMx: on i os i s )
(Ironed it is, emphysema.
and asthma
All NHRD
United States
|>
Obs. fxp. SHR (*.)

--
B4 109. 1 77* 100


730 174.7 13?" 10(1


314 787.8 111 100
1 .m.iil.l | nrni
P
Otis. l»|i SHR (1) 01. s. tup.

7 75
1 5.f. in»* 7.1 7J 75. J


74 M.I

1 7.4 171 If. Mi 7.1.3
4 fl.7 4T .IS Ifi5 lfi!»
1*

SHR

80
97


170*
.<
90
100

P
(t)

18
99


46

99
100
If S. ami flirope combined
~ ' ' •"••' P
Ohs. t«p. SHR (X)

-
157 184. 7 85 ion


370 767.3 177'" 100


479 447.7 107 |OO
    *%ii|iti f iranlly di((rr«nt fro« national norlality rates fnr llii> llnltril States ami I nrope ami nrnvlncial •orlallly rates  for Canada:  p  <0.05.
   **Sii|nificantly different fro«i national Mortality rales for tli« United St.ilrs and liirnp* and provincial •ortallly rates  for Canada:  p  <0.01.
  •"•Sii|ni f itanl ly different Iron national Mortality rates for the United Sidles ami furnpe and provincial mortality rates  for Canada:  p  
-------
of  respirable  fibers  were probably tenfold higher than current levels.  Con-
cerning  asbestos  exposure as a confounding variable, Doll (1987) stated that
because  neither  excesses  of lung cancers nor mesotheliomas were  found  among
those employed when asbestos exposure was most likely to have  occurred,  asbestos
is  not  likely to have  accounted  for the excess lung cancer  in MW  workers.
While other  carcinogens may have been present, 1t fs unclear to  what extent
they could  have  accounted for the difference between MW  and  FW workers.   In
addition, it was not  possible to determine the confounding effects  of bitumen
and mineral  oil,  arsenic  in slag, and polycyclic aromatic hydrocarbons  from
furnaces upon the excess  lung cancer observed in MW workers.
     Conflicting evidence exists regarding the relationship between  exposure to
MMMF and  respiratory  disease.   In favor of a causal  relationship  between MMMF
and lung cancer, Doll (1967) pointed out the positive animal data, the gradient
of  fiber  counts  in  MW,  FG wool,  and  continuous  FG  filament,  and the higher
mortality among  FG  workers exposed to small-diameter fibers compared to those
who were  not.   Against a causal  relationship are the failure to observe  a
relationship between risk and duration of employment, lack of  clear  evidence  of
a quantitative relationship  between  mortality and estimated  cumulative dose,
and low  level  of fiber counts in  the  workplace.  Doll  (1987) concluded that
(1) the risk of lung cancer among those employed in the  early  periods of MW and
FG wool  sectors  of  MMMF  production is 25 percent above  that expected after 30
years since  first employment;  (2) the risk  of lung cancer is  greater  in the MW
than in  the  FG industry;  (3) no  risk  of  lung cancer was demonstrated in the
continuous FG filament  sector; (4) a variety of carcinogens might  have contrib-
uted to  the  lung cancer  hazard;  (5) uncertainty of  fiber counts  in the early
production  periods  and the  extent of the contribution  of other  carcinogens
precludes a  precise  quantitative  estimate of the likely effect of exposure to
current  fiber  levels;  (6) no specific hazard other  than  lung cancer  has been
established; and (7) MMMF are less carcinogenic than asbestos  fibers.
     However, nonsignificant  increases  from lung cancer based on  small  numbers
of  deaths  and  low power  preclude  the  purported  greater risk  of  lung cancer
among those  exposed  to  small-diameter FG fibers.  Similarly,  the  small  numbers
of  deaths  and low  power  preclude Doll's claim that no risk  exists  in the
continuous FG  filament  sector.   In addition, the low power to detect NMRD in
the Canadian cohort  contributed to the inability  to  detect a  hazard from NMRD.
Despite the  significant excesses from other respiratory disease in the combined

.i,.iv, IP**                          8-82         DRAFT—DO NOT  QUOTE  OR CITE

-------
U.S./European cohort, it is unclear why Doll  (1987)  concluded  that  no NMRD risk
exists.  Aside  from  the limitations discussed above, the healthy worker effect
in this study was minimized by a long  observation period.
8.1.1.2.5  Integration of human health effects  data  by  fiber and  by carcinogenic
and noncarcinogenic effects.  Table 8-12 summarizes the najor  epidemiological
studies available on MMMF, placing special  emphasis  upon malignant  and  nonmalig-
nant respiratory  disease  (NMRD).   All  of these studies are historical pros-
pective  in  design and  have been critically  reviewed and analyzed  above  in
depth.
     Concerning  noncarcinogenic  effects following  exposure of  workers  to
microfibers  (fine-diameter  FG,  i.e.,   <1 un>  in diameter), Enterline et  al.
(1983)  found an  excess  of  NMRD  among workers never exposed to  microfibers
(SMR =  130.1, p  <0.05),  especially  among  those  in Plant 9 (SMR  = 136.4,
p <0.05).  The  apparent absence  of a nonmalignant  respiratory disease risk
among those  ever  exposed to microfibers was  from small numbers of  NMRD deaths
and insufficient  power  to detect a 50 percent increased  risk  of NMRD.   In  a
followup study,  Enterline  et al.  (1986)  again  found that en excess  of NMRD was
concentrated  among those  never  exposed to  microfibers  (SMR =  133.5, p  <0.01).
Bayliss et al.  (1976)  evaluated  the  pathogenicity of microfibers  in a nested
case-control   study of NMRD.   There was no association between  cases of  NMRD  and
work in  operations  involving exposure to microfibers.   Thus,   it appears that
there was no association between NMRD  and  exposure  to  microfibers.   In the few
available studies that  explored  noncarcinogenic  effects  following  exposure  to
microfibers,  the  small number of deaths among those  ever  exposed contributed to
the low power to  detect NMRD in that group.
     Concerning noncarcinogenic  effects  following exposure to  respirable glass
fibers  (FG wool  <3  urn 1>n diameter), Enterline et al.  (1983) found  excess  NMRD
among those employed in FG wool and continuous FG filament plants combined (SMR
= 129.6, p <0.01),  which increased after 20 or  more years  of latency  (SMR =
135.5, p <0.01).  The highest risk of NMRD (SMR * 136.1,  p <0.05),  was  found in
those with  the lowest  cumulative exposure to  respirable FG  fibers,  i.e.,
2.14 fibers/cm -month.  This  seems  to indicate that there was  no  association
between NMRD  and  exposure to FG wool  and/or  continuous  FG filament.   In the
followup study,  Enterline et al.  (1986) found a  slightly  higher excess  from
NMRD in  plants  that produced FG wool  and  continuous  FG  filament after 20 or
more years of latency (SMR = 141.8, p <0.01).  The  low power  to detect NMRD in
individual  plants may have precluded detecting NMRD in these studies.   Bayliss
July 1988                          8-83         DRAFT—DO NOT  QUOTE OR CITE

-------
I iftatnrs
I I her
I i nr and
'» (I'm?}
line el al.
                MW and FG
                                           n-i?   SUMMARY nr rARtiNotiiMit AND NONrAPr.iNnr.rNlr Uriels mi*
                                          PlNIM'f CIIVI MOHIIIIIIIY  AMI) HdMIAIIIY MllHM S IN WIIHKI RS I xl'OSI II  III
Cohort si/r and
minimum iluYatioti
  of exposure
 Mi-.ui I ilii-r
   Invi-ls
 (I il.rrAm')
                       7049 male production
                       and maintenance
                       workers  employed
                       for at  (past. I  ye.ir
                       (6 monlhs for llmsn
                       in Iwn plants
                       exposed  to tine-
                       diameler 1C)
                       16,/SO Male produc-
                       tion and n«inlpnancp
                       workers (l.fUf, MW;
                       I4.RR4 TO) rm|ilnyrd
                       (or at least 1 ypar
                       (6 months for those
                       in two plants
                       to fine-dianeter
II. Jin lor MW;
(Ml? I lor Hi
Mean ciimilat ive exposure
{I iher/em" •mntlis):
SJ 71. for MW, 4.47  for If,
   tare Inoifenic endpolnl*
Ixiess of respiratory syste*
t:ancer (site-specific cancers
not reported) limited to
Plant /.  Increased risk was
cfreater amonq those with ?0
or more years since first
exposure |ntficant excess for
combined MW and FG workers.
Noncarc Inoitenlc
   enilpnints
None delected.
                                                                                                                                       [xcess  nf  NMRO in
                                                                                                                                       combined MW and 1C.
                                                                                                                                       workers (SMR = 1.1?. 1")
                                                                                                                                       appeared unchanged
                                                                                                                                       after 20 or metre years
                                                                                                                                       since first employment
                                                                                                                                       (SMR =  1.37. !••).
                 1C wool and
                 1C. filament
                                                                               Ixcess of respiratory system
                                                                               cancers (site-specific cancers
                                                                               not re-reported) found In MW
                                                                               workers (SMB = 160.!»•).   luces*
                                                                               was qreatest In Plant 7 (SMR =
                                                                               4M.J") and Increased 70 or
                                                                               more years since first employ-
                                                                               ment (SMR-fl6B.5**).   the highest
                                                                               risk occurred In those with 21.B8
                                                                               to 99.99 fiber/cn'-months of
                                                                               cumulative exposure to resplrable
                                                                               MW fihers (SMR = 178.2«).

                                                                               None delected.
                                                                                                                                       Excess of NMRO  In MW
                                                                                                                                       plant 3 (SMR =  ?2I«)
                                                                                                                                       appeared unchanged
                                                                                                                                       after 20 or more years
                                                                                                                                       since first employment
                                                                                                                                       (SMR - 219.9*).
                                                                       Excess of NMRO In all  FG
                                                                       planls (SMR = 129.6")
                                                                       Increased after 20 or
                                                                       •ore years since lirst
                                                                       employment (SMR =
                                                                       I.1S.S**).  Ihe hlohesl
                                                                       risk of NMRD (SMR -
                                                                       13n.I*) was found in
                                                                       those with lowrst
                                                                       exposure to rpspiralile
                                                                       1C, fillers. <2. M
                                                                       f ibers/c.m^-monlhs.
                                                                                                         (continued on  the  following paqe)

-------
                                                                       1MIII n  17    (r.nnliniHMl)
Invesl iq.tlnrs
Interline pt al.
(U.S.)
(ti.nt. )
Cohort sire ami
• iniwM ilnral inn
f lh?r of exposure
IG fine dlaaeler 6 (mnths «iini«nini
duration of pw)ili>y-
•ent tor those in
two plants
HP.III fil.ri
I.-V.-I-.
|l ilwr/im'

Iliter Iine p|
 ( 11111,)
 (H. V)
at.
        MW and fG
on
 i
CO
en
                      fG wool and
                      IG filament
                      rr. rine-
                               16.661 nalr produc-
                               tion antt •ainfrii.tricp
                               workers raploypit  lor
                               at Ipast I y«-.>r (f.
                               «mnth« tor thnsp  in
                               two plants
                               to
                               fG).
Not fi*piiff.pil
                               6 Month*
                               duration of p«i|>lfiy-
                               •pnt for those  in
                               two plants
                                                                                          t.irtlnof|enlc endpolnt*

                                                                                       None
       of respiratory tystc*
cancer (SMR = IIS. 6**)  Increased
altrr 70 or Mire years  after
since first exposure (SHU =
                               fxcess of lung cancer (SMI «
                               IK.*) Increased after 20 or
                               •ore years since first exposure
                               (SUB = m.«").

                               txcess of respiratory tystew
                               cancer found in IW workers
                               after fO or anre years since
                               first exposure (SUB « M6. I*).
                               fur.ess was greater anong those
                               MM workers with 30 *ore years
                               since first exposure (SMI =
                               I5I./M.

                               F»cess of respiratory systea
                               cancer after 20 or More years
                               since first exposure (SHR -
                               129.%*).   Slonlficant
                               increasino trend of SHBs with
                               increaslnq latency; excess
                               aoomi those with 30 or awre
                               years since first exposure
                               (SMI - 137.0**) increased
                               after 20 to 29 years of
                               e«|>loyiwnt (SHR = 160.1*).

                               Nonp rfptpcled.
                                                                                                                                             MoncarcInoqenlc
                                                                                                                                                 erw (points
(xcess of NNRO  l
to fG workers nevpr
exposed to I Inp-iliameler
fG (SMP = 130. I*).
especially thoip  in
Plant 9 (SMS =  1.16.4*).
Iherefore. no associa-
tion between NMRD and
f ine-diaaeler filters.

(xcess of MHffO  (SMW  =
132.3**) remains
unchanged after 20 or
•ore years since  first
exposure 
-------
                    fiber
 i*rl inp pi at.
 inn
 '.S.)
MW and
                        Cohort sl/e •niil
                        • I n I KIM duration
                          nl  exposure
                           (Mill  It-1?,  (rnnllnupil)

                           Mr .in I ilipr
                             ir ve Is
                           (I ihri/in')
                                  CarclnnctPnlc emfpotntt
                                         Honearclnoi|pnlc
                                            endpoinls
16,f>M m*\r prnilur-
llon ami Maintenance
workers pmploypd (or
at Iras I I yp.»r
(f> i*nnths  (or  those
in two plants
exposed to (inp-
         If.)
     p«pnsurp  li-wpls:
It;  - ll.tll; MW  -  !.!»
l.rlorr  !•»/!,. H 1
IM1f, IIMI. (I. III .liter
I'M.O
                                                                     0. 3S?
                    FR wool ami
                    in filament
                                                 0. OVJ
tier I inp Anil
•tiller', on
I I'I/S»
[H S. )
-Insulation"
|rr. wool]
with  a  mean lpni|lh
of  SPrvie* of ypvirs
?.1.<  ypars
                          Not roportpil.
Allpr ?n< ypars of  latency,
ol»sprvp«l excpss of  respiratory
syslpi* canrpr slqnlfleant
«:niii|>arpd to national  rates
C.MH = l?«.3"). but  not to
IIM.A! rales (112.3).  t«cess
ol liini) cancer after  20* years
of latency siqnlf leant coaipared
to national (SMR =  I2S.4**) or
loral rates (112.7").  this
shows lite strength  of the
lunq cancer hazard, which
appears to be responsible for
Hie respiratory cancer excess.

I xcess of respiratory tystew oncer
found In MW workers after 20*
ypars of latency significant
compared to national (SMA =
Mr. I**), but not to  local rates
(SUB = 130.9).  fhe greatest
PKCPSS of respiratory cancer
occurred aamng MW workers
compared to either national
(SHK = I4R.4**) or  local rates
(SMR = IJ3.6*). this shows the
respiratory cancer haiard awong
MMMI  workers  Is concentrated In
the MW suhcohort. •

lucess of respiratory syste*
cancer found after 20 years
significant compared  to
national (SMR = 129.S**) but not
local rates (3MM =  110.6).
liicess of respiratory cancer aanng
those with 30* years of latency
siipilf leant compared  to national
(SMR = I.1/.8**) but not local
ralPS (SMR =  110.8).

Nonp detected.
txcess of NMRD (SMR =
132. .!••).  Ho further
analyses of NMRR.
                                                                                                                                            None detected.
                                                                                                             (continued on the following pane)

-------
                                                                   IW1K  n 17.  (rontlttupil)
eslli)at«r«
llss ft al.
9/f.)
S. )
Hber
"Insulation"
(ff, wool]
Cohort si/e ami
Minimum dural inn
of exposure
I44B while males
employed for S or
more years
Mfitn 1 ilwr
Inypls
(lil male production
workers employed (or
at .least 1 year
2S76 current and
former male workers
employed for more
than 3 months

2557 current and
former male workers
employed for more
than 3 months
                       ?3Si males without
                       regard for minimum
                       duration of employ-
                       ment:  I.BIS (//»)
                       In one of three MW
                       plants; S46 (7«)
                       in one fO plant
                         Not
                                               l year In any four
plants (SIK = 233*).
                                        None detected.
None deleeled.
                                                                                                                      None deleeled.
None detected.
                                                                                                                                        None delected.
                                                                                                         (continued on the following page)

-------
                                                                   lAlllf  II I?   (rnnl iniii-d)
  gators
f ihrr
 •ii and
 >rk (nnr.)
                MW
 holm and
 i|pr (1'JHn)
 i-.i)
 and Jensen
 )
 .irk)
MW or fC wool
 rl  al
.)
PI.II k )
 and
it™ ( I'tW,)
I .in.l)
                 1C wool
 Cohort si/c 4i«l
 Mi ni mini rlnr.il inn
   of exposuYf
3600 male and  female
(2HX) p>'odiirlion
employees who  had
worked »t  leasl
1 year in one  of two
MW plants or  in a 10
wool p!int

5169 «ale and  fewtle
(flX) employee! with-
out regard  (or
• iniwMi duration of
e«ploynent
                       5317 *ale ami fenale
                       (ftT) pmfiloypp*. with-
                       out rei|.»ril lor
                       minimm ittir.il inn of
                       enpfoynienl.

                       94| ma IP and fpn.ilp
                       worlriT < V.il 1 wm-llrr'.
                       p*i|ilnypil .ls
 ( 1 ilirr/i «')
Not ic|iorlp eslim.iti
                          for |i.ist <*»poMirrs
                          Hot rpported
   r.ir« lom|piilc  endpnlnls
I »c«".s of  intestinal cancer
iui idrnce  in MW workers
pn>|ili>ypd >l yi>ar (MR = 196*).
I xi rss nf  luiii| cancer inei-
iIt-in I? in one MW plant (rnck
w,,ol) (MR = 799*).  which was
wiiri1 prnnniinced when shorl-
Iprn ('I yr) workers were
excluded (SIR  = 3n6**).
Ixcrss of  lung cancer Incl-
dpnr.c in the three MW ptants
appeared after 70» years since
first exposure (SIR  = 199*),
which was concentrated among
those employed for 1 or more
years (SIR = 706*).

Ixcess of  lung cancer among those
with I to 4 years of employment
and 30 or more years since first
exposure In MW production (SMR =
HHII""") or in  fG wool production
(SMR = 345").
                               Ixcess  of  digestive organ/
                               peritoneum cancer Incidence
                               amniHf male and female pro-
                               duction workers during the
                               first  10 years after Initial
                               employment (SIR = 15?*).

                               Excess  of  lung cancer Incidence
                               among male and female production
                               workers after 70 or more years
                               since  first employment (SIR =
                               ?0'l*) was  concentrated among
                               male production workers (SIR =
                               19V).

                               Among male production workers.
                               excess  incidence from huccal
                               i:avily/pharyni|pal cancer (SIR =
                               710*)  and  digestive organ cancer
                               
-------
                                                                  lAlllt  n  17   (mnllnued)
itigators
fiber
 Cohort si/e and
 Minimum duration
   of exposure
le  and
il/el-Beyme
14)
.1  del-many)
le and
iliel-fleymo
If,)
•st Germany)
                 HW
ittl el a).
»«»
• ly)
FG wool before
I960; continuous
10 filament  after
i960
 Mi et
 i.)
 IV)
        al.
fG woof before
 I9MI; continuous
 1C  filament
 alter  |<»r.O
2096 male prndmlion
and maintenance
employees without
regard for minimum
duration of employ-
menl

2092 male product loo
and mainlenance
employees without
regard for minimum
duration of emplny-
menl
11S8 oale production
workers e«ptoyed at
least 1 year
i098 male firmhirtion
workers eiqiloyrd  at
at least  I year
                                                                 Mean 1 ilier
                                                                    levels
                                                Not  reported
                                                Not  reported
                         Resplralile  fraction:
                         II.not  to 0.0?3  (median
                         diamvler 0.)  to 0. 1 )«•)
                                     fraction:
                          II  ("I*. I  It.018 (median
                          diam-I.er 0. .1 to O. 7 |IM)
                                                           r.arclnoc|«nlc endpolnts
                                        NoncarcInoqenic
                                           endpolnts
                                                       I »te-.s of stomach cancer  among
                                                       production workers after  ?0  to
                                                       79 years since  first employment
                                                       (SMH = 2/34).
      production and maintenance
workers hired before the  intro-
duction of resin binders  In 1969.
excesses of stomach cancer after
70 to 79 years since first e«po-
sure (SHU = ?6I«) and lung
cancer after 10 to 19 years
since first exposure (SMA = 272*).
Maintenance workers had • tlgnl-
f leant excess of lung cancer
after 70 to 79 years since first
exposure (SMR = 249*).

Using Internal controls as the
reference group, there was an
increased relative risk fro*
pancreatic cancer (MR = 6.76*).

txcess of larynqeal cancer compared
to national rates (SMR = 400*) but
not local or regional rates.
Compared to local rates, exposure
of taryngeal cancer among those
with 75 or more years of  latency
(SMR = 938***). with 20 or more
years of exposure (SMR * 833*).
whose first exposure was before
aue ?S (SMR = 1*3B**). and whose
cumulative exposure was >0. 30
f ibers/cm'-wmths (SMR =  1SOO*").
All four cases were first
employed before  I960 when FG wool
ralher than FG filament was
produced,  these results  show  •
consistent, significant Increase
among those most heavily exposed.

Using local rales, excess of
larynqeal cancer among those with
?S or more years of lalency (SMR =
                                                                                               All laryngeal cancer deaths
                                                                                               occurred lielore  I'»M> when If. wool
                                                                                               was |irn
-------
                                                                    lAPIf «- \f.  {continued)
    (colors
f Ihpr
  irr Pt al.
  Mi)
  I ami-IG wnnl;
  Ihprn Irplaitd-
  limmtis
  .imrnt)
in wool or
continuous fG
I ilament
 ace I  el al.
 '.)»* )
 irrnpe )
MM and FG
       el
                   MM ami
I in opp)
 Cohort 'sl/P ami
 minimum duration
   of exposure
4734 Males anil
females (?SX)
employed lor at
least I ypar in
an I no, list* 1C wnnl
factory; 1910 males
and (pica les (.17 «»
employed at any
limp in • Norlliprn
Ireland cont unions
1C filament factory
75.146 »ale  and
female and employees
of one of 13 HMHF
plants (7 MM plants.
4 fG wool. ? contin-
uous ff,  fllampnt);
for four plants.
those caploypd !p fraction:
                                                ' I »;
                                                        fl (11 to 0. I
   CarctmMienlc endpolnts
Noiw.«rclno<|pnlc
   pndpnints
Iirpss of stomach cancer aeong
tltp fpiiiale work force of the
Lnqlish factory compared to
national (SMR = 777*) hut not
local rates.  Excess of lung,
cancer aiwinq male production
workers of the English factory
('.MB = 110*) only when compared
to national rates.  In this
sa*p qroini, there were some
excesses of lung cancer when
the data were analyied by
latency and duration, but
these were not significant
compared to local rates.

No site-specific cancer
excesses were detected In the
Northern Ireland factory.

When the data fro* all 1) plants
were pooled, compared to
national rates, excess of lung
cancer annmf men with 30 or (rare
years since first employment
(SHR = 1931).
                               Excess  of  lung cancer compared
                               to national  but not local  rates
                               (SNR -  175*).   Excess of  lung
                               cancer  Increased after 30 or
                               more years since first exposure
                               to M** compared to cither
                               national (SMR = 183*) or  local
                               rates (SHU = I5?«).

                               Excess  lung cancer among  MM workers
                               compared to national (SMR = 714*)
                               and local  rates (SHU = 757*) during
                               the early  technological phase.
                               especially among those with 30 or
                               more years since first exposure  to
                               MM diirimi the early phase (SMR =
None detected In either
InglIsh or  Irish
cohorts.
                                                                                                                      None detected.
                                        Not applicable because
                                        analysis restricted to
                                        lung cancer mortality.
                                                                                                           (continued on the following

-------
                                                                      I Mil I  Ml?   (con! I mini)
                     Htipr
 Cohort *!/«• and
 minimum dur.it {mi
   ol exposure
 M«».in I ilii-r
   li»v«» Is
 (I ilirr/rm')
         et
( I 'IW..1 )
(I |.p)
( • tint  )
                     If. wool
                     rr. wool
Simonaln fl a I.
 (i'mr.h.c;
 (I urnpi*)
                     Continuous
                     IT. filament
                     MM and
2I.967 male and
female (14. fit) pro-
duction workers;
worker employed ' I
year  In four plants
were excluded.
in, i is MM, B.?n6 to
wool, and 3.bM> con-
tinuous fC
workers
                                                                                <)I>SP in
                    of
MM
                                                                       iMK ?/ In /(I
                                                                                                                   pndpolnt*
                                        Noncarcoints
l«r.pss  lunq cancvr aw>ng fC wool
wnrkrrs rmnparpd to national rates
only (SMR = 1?^*). especially
.iminci Ihnsp with 30 or »or« years
sinrp first exposure to 1C wool
(SMH =  113").

fxcess  lim«j cancer Mom) FC wool
workers compared to national
ral.ns only during; the Intermediate
tprhnoloqical phase (SMR = 139*),
particularly mong those after
in to 19 years since first
exposure  and local rates
(SMR =  IS?*),  fhe highest risk
of ileath aiming short-tera
workers was from violent causes.
whereas the highest risk amng
those employed for at least
I year was lung cancer.
Increased Incidence of hoccal
ravity/pharyngeal cancer (SIR =
IM*) among production workers
exposed to MMMf, remained
siipiK itanl compared to local
r;il.-% (SIR - IV>«).
(•cess of MMRD among
short-term (employed

-------
                                                                  (AMI I fl-l?    (ronlliuipil)
sliqatnrs
I I her
Cohort slfe .mil
minimum duration
  of
Mi« an I iliri
  N've Is
(ftl>.-t/<«.'
nato el at.
"Mt.c; I'm/)
rope)
nl )
                  rr. wool
Carcinogenic enrfpolnts
                                                                              fxcess of  lung cancer In MW
                                                                              factory 4 compared to national
                                                                              (SMR = jr.O*) and local rates
                                                                              (KM)*) and in factory 9 compared
                                                                              to local rates (SMR = 140*).
                                                                              Among those with 20 or more
                                                                              years of latency, excess of lung
                                                                              cancer in preproduction/
                                                                              production rock wool workers
                                                                              compared to local rates (SMR =
                                                                              711')*).  Among workers with ?0 or
                                                                              more years of latency, excess of
                                                                              lung cancer during the early and
                                                                              intermediate phases compared to
                                                                              national (SMB = 706*) and local
                                                                              rates (SMR = ??3*). showing a
                                                                              dose-response relationship of
                                                                              decreasing SMRs with Increasing
                                                                              duration of employment,   this
                                                                              lung cancer excess was concen-
                                                                              trated in rock wool workers first
                                                                              employed In the early phase
                                                                              (SMD = ?S/*), particularly among
                                                                              those with 30 or more year* of
                                                                              latency (SMR = 295*) compared to
                                                                              local rates.

                                                                              txcess Incidence of buccal cav.lly/
                                                                              pharynqeal cancer was limited to
                                                                              rock wool workers (SIR * 19?*).
                                                                              especially among those with 10 to
                                                                              19 years of latency (SIR = 238")
                                                                              and 70 to 29 years of latency
                                                                              (SMR = 2/2").

                                                                              Ixcess of  lung cancer compared to
                                                                              national rates only (SMR * 177*).
                                                                              especially among those with 30
                                                                              or more years since first
                                                                              exposure (SMR = 1/3*).
Noncarcinoqenlc
   emlpoinls
                                                                                                         (continued on the following page)

-------
estimator*
 iber
 Cohort sl/r ami
 minimum idir.il ion
   of e«|i«Mirr
                                                                    Mfll I  fl-17.   (r:nnt innwtf

                                                                    Mo.ln t ilii-r
                                                                      li-wrls
                                                                    (lih.-r/.V)
wtti.lto ft  a I.
•iiiMi.r.;  inni)
nrn|>f*)
mil. )
                   fO wool
I In et  al.
inr.)
ranee)
                   Continuous
                   rfi filament
FC wool
1)74 active or
retired •Ales
••ployed for at
least I year
                                                                   Respiralilr (railIon:
                                                                   07 (mr-aii diameter -
                                                                   6.4 inn;  JirX < I |im;
                                                                   Kit ' I |im)
   C*rclnoi|enlc endnolnts
None arc InoqenIc
   emlpoints
         to national rates,  excess
of liini| c.incer  United  to  those
•••tiluypil in the  intermediate
l<-(hmilo<|iral phase (SMR =  139*).
p.irt it ul^rly annnq those with
III lo \>) years  since first
ritposiire (SMU =  I6«s).  No
r«i rssrs of lung cancer Mono,
It; workers were  found using
local rates,  txcess of
larymieal cancer (SHft = 400*)
limited to Male  production
workers in an Italian FC wool
cohort.   Intension of follow-up
continued previous excess of
larynqeal cancer.

I xcess of lung cancer umony  those
with  less than  10 years of  latency
was significant  only when compared
to national rates (SMR  = 224*).

Increased Incidence of  upper
respiratory and  alimentary
tract cancers (SMR = 2I8"«)
particularly buccal cavity/
pharyngeal cancer (SIR  =
3 1 /••••) compared to regional
rates.  Increased Incidence  of
nuccat cavity cancer (SIR -
Not applicable because
study was limited to
cancer incidence.
holm el al.
TO/)
                                          135.0?6 aale SwedUh
                                          construction workers
                                          without reqarrf to
                                          •titralinn of enptny-
                                          •ent
                                                Hot rppnrtPil
                                                        Excesses of upper  respiratory
                                                        and alimentary  tract cancers
                                                        were confined to production
                                                        workers exposed fro* 10 to
                                                        19 years (SIR = 304**)  or  for
                                                        20 or more year* (SIR = 3)3***).

                                                        Increased  Incidence of
                                                        mesolhrliomas (SIR = 213*).
                                                                                                                                           None detected.
                                                                                                           (continued on the  following pane)

-------
Investigators
I iber
Cohnrl sl/f> .«»IIS.  Co^iared to local
rales, excess lunq cancer
rcnainerf significantly elevated
lor combined U.S. and European
cohorts 
-------
et al.  (1976)  found  an excess  of  NMRD  in U.S. male FG wool workers (SMR = 189,
p <0.05, Power = 34 percent), which increased among those  with  10  or more years
of  latency (SMR = 200, p <0.01).   Several  studies,  however, were unable to
detect  excesses  from NMRO  even if such excesses did, in  fact, exist.  These
include Enterline  and Henderson (197S),  Morgan et •!.  (1984),  Shannon 
-------
to im'crofibers,  the  small  numbers  of  deaths and  low power precluded meaningful
analysis of microfiber exposure and respiratory cancer.
     To further  explore  whether exposure to small-diameter  fibers  produced  in
four different plants was related to an excess of respiratory cancer,  Enter!ine
et al.  (1986,  1987)  examined mortality from respiratory cancer by  plant  among
those exposed  and  never  exposed.   Contrary to expectations,  the only  signifi-
cant excess  of respiratory cancer appeared in workers  1n one  plant who  were
never exposed  to microfibers as compared to  national mortality  rates  (SMR  =
120.2, p <0.05).  The apparent absence of a respiratory  cancer risk among those
ever exposed to microfibers was from small numbers of respiratory cancer  deaths
and insufficient power to  detect a 50 percent  increased  risk of respiratory
cancer  in  this group.   When investigators further categorized the  respiratory
cancer  deaths  among those  ever exposed to microfibers  by both  latency  and
length  of  exposure,  they found that this analysis  supported  a  causal  relation-
ship between exposure  to respiratory  cancer and microfiber  exposure.  On the
contrary,  because  there  were too few deaths  and  insufficient  power to permit
meaningful  analysis  of the data in this  manner,  the  author's  hypothesis of a
causal  relationship  is not supported  by  the data.  In addition,  Bayliss  et  al.
(1976)  evaluated the pathogenicity of  microfibers in a nested  case-control
study of malignant respiratory disease.   No association  was  found between cases
of malignant respiratory disease  and  work in operations involving  exposure  to
microfibers.    Thus,  it  appears  that  no  association  existed between  excess
respiratory cancer and exposure to microfibers.
     Concerning  carcinogenic  effects  following  exposure of workers to  FG wool,
Enter!ine  et al. (1986)  found an excess  of respiratory  system cancer  in U.S.
male  production  and maintenance workers  after  20 or more years since first
exposure to  both FG wool and continuous  FG filament  (SMR =  129.5,  p <0.01).   A
significant, increasing  trend of  SMRs was associated with increasing latency.
An excess  of  respiratory cancer among those  with  30  or more years of  latency
(SMR =  137.8,  p <0.01)  increased after  20 to 29 years  of employment  (SMR  =
160.1,  p <0.01).   In the subsequent update, Enterline et al.  (1987) found that
the excess of  respiratory system cancer after 20 years  of latency  was reduced
and no  longer  significant compared to local  rates (SMR = 110.6).  Similarly,
the excess of  respiratory cancer seen among  those with 30  or more years of
latency was  reduced  and  no  longer  significant  compared to  local rates (SMR =
110.8).  This  suggests that  no relationship existed between exposure  to FG wool
and continuous FG filament and respiratory cancer.

-------
       In  Canada,  Shannon et al.  (1986, 1987) found an excess of lung cancer  in
  male  production workers with  the highest  exposure  to  FG wool (SMR = 199,
  p  <0.005).   In addition, an excess of lung cancer was  found  among  those with
  the  longest  duration of  exposure and the longest tine since first exposure  (SMR
  =  182,  p <0.05).  This  suggests a causal  relationship  between exposure to FG
  wool  and lung  cancer that was correlated with latency and duration of exposure.
       In  Europe,  Bertazzi tt al. (1984) found a fourfold  excess  of  laryngeal
  cancer  among male FG wool production workers compared  to national  rates, but
  not  regional or  local  rates.  However, using local rates, excesses of laryngeal
  cancer  were  found among those with 25 or  more  years of  latency  (SMR = 938,
  p  <0.005), with  20 or more years of exposure (SMR =  833,  p <0.05),  whose first
  exposure was before age  25 (SMR =1538, p <0.01), and whose cumulative exposure
  was  >0.30 fibers/cm -months  (SMR = 1500,  p <0.005).   All  four cases were first
  employed before  1960  when  FG wool  rather than  FG  filament  was produced.
  Despite  the  small number of  observed deaths upon which  these analyses were
  based,  the results show a consistent, significant increase among the  FG wool
  workers  who were most  heavily exposed.  In a subsequent update, Bertazzi  et al.
  (1986)  found a sixfold  excess of laryngeal cancer among those with 25 or more
  years of latency.   All  laryngeal cancer deaths occurred  before  1960 when FG
  wool was produced.
      Gardner  et  al.  (1986) found  an excess of stomach cancer among female
  employees of an  English  FG wool plant compared to national  (SMR = 277, p  <0.05)
  but  not  local  rates.   In the same factory, an excess of  lung cancer was  found
  among male production  workers (SMR = 130,  p <0.05)  when compared to national
  rates.   Additional excesses  of lung cancer were  found when the data from male
  production workers were analyzed by  latency and  duration of  employment,  but
•  these were not significant compared to local rates.
       Simonato  et al.   (1986a) found  an excess of lung  cancer among FG wool
  workers  compared to national rates (SMR * 127, p <0.05), especially among those
  with  30  or more  years  since first exposure  to FG wool (SMR = 173, p <0.05).   An
  excess  of lung cancer was found for the intermediate phase of FG wool produc-
  tion  (SMR = 139, p <0.05),  particularly after 10 to  19  years  of  latency  (SMR =
  164,  p  <0.05), but these excesses were no longer significant  compared to  local
  rates.   Among  those with 20 or more years  of latency,  excesses of  lung cancer
  were  found for those employed  from 1 to 4  years  (SMR = 197, p <0.05) and from
  10 to 19 years  (SMR * 190, p <0.05); local  rates not reported.   In subsequent
  updates,  Simonato  et  al. (1986b,c, 1987) found that  the above excesses of  lung
  July 1988                         8-97         DRAFT—DO NOT  QUOTE OR CITE

-------
cancer were  significant  only when compared to national rates.  However, con-
cerning other  elevated causes of death, Simonato et al. (1986b) found that all
nine laryngeal  cancer  deaths occurred  in male production workers, and that six
of the nine  deaths occurred fn the Italian FG wool  cohort  studied by  Bertazzi
«t al.  (1984,  1986)  and  Simonato et al.  (1986c).   This excess from laryngeal
cancer was  statistically significant  (6 observed,  1.5  expected,  SMR  « 400,
p <0.05, Power  =  8 percent).  In other plants,  the  apparent  lack of an  excess
from laryngeal  cancer  mortality  could  be  attributed to the insufficient power
to detect a 50 percent increased risk of death from  laryngeal  cancer.  Based  on
the  additional  deaths identified  after  increased  followup,  Simonato et al.
(1986c) concluded  that these data confirmed the previous findings by  Bertazzi
et al.  (1984, 1986) of an elevated laryngeal cancer  risk in the Italian  FG  wool
cohort.
     Moulin et  al.  (1986)  reported an  increased incidence  of  upper respiratory
and alimentary tract cancer among active or retired  French  FG wool workers  (SIR
= 216, p <0.005),  particularly buccal  cavity/pharyngeal cancers (SIR = 317,
p <0.001) compared  to  regional  rates.   There was a  threefold excess of  buccal
cavity cancer alone.  Excesses of upper respiratory  and alimentary tract cancer
were confined  to  production workers exposed for 10 to  19  years (SIR = 304,
p <0.01)  or for 20 or more years (SIR = 333, p <0.005).
     Thus,  it  appears that  among FG  wool  workers, there  were significant
excesses  from  respiratory  cancer  (Enterline  et al.,  1986),   including  lung
cancer (Shannon et al.,  1986,  1987;  Gardner et al.,  1986; Simonato et al.,
1986a); from laryngeal cancer (Bertazzi  et al., 1984, 1986;   Simonato et al.,
1986a,b);  from stomach  cancer  (Gardner  etal.,  1986);  and  from buccal
cavity/pharyngeal  cancer  (Moulin et al.,  1986),  especially from buccal  cavity
cancer (Moulin et al. , 1986).
     Concerning carcinogenic  effects following  exposure of workers to continu-
ous  FG filament,  Simonato et al. (1986a)  found  an  excess of  lung cancer among
male and female Europe production workers with less  than  10 years of latency
compared to  national rates (SMR =  224, p  <0.05), but not  compared to  local
rates  (Simonato et al.,  1986b.,C, 1987).  However, there was  insufficient power
to detect  lung cancer in this cohort.   Enterline  et al.  (1986, 1987)  found no
excess respiratory  cancers  among those producing  continuous  FG filament,  even
when the cohort was subdivided by  latency  and  length of  exposure.   Although
there was sufficient power to detect respiratory cancer in  the total  continuous
FG filament cohort (Power = 91 percent), there was  insufficient power to detect
.1,-iw TQfip                          8-98         DRAFT—DO NOT QUOTE  OR CITE

-------
a  50  percent Increase in respiratory cancer when the cohort was  subdivided by
latency  and  length of exposure.   Bertazzi et al. (1984, 1986) found that the
excess laryngeal cancer cases were first employed before 1960 when  only  FG wool
was  produced rather  than  after  1960 when  only continuous  FG filament was
produced.  There was  insufficient power to detect site-specific  cancers among
those employed after  1960.  Gardner et al. (1986) found no  site-specific cancer
excesses  in  a Northern  Ireland  continuous FG  filament factory.   There was
insufficient power to detect lung cancer in this cohort.   Thus,  at this time,
it appears that  no relationship  exists  between  cancer and exposure to continu-
ous FG filament.
     Concerning carcinogenic effects following exposure to  MW (mineral wool and
rock and/or  slag wool),  interline and Marsh  (1982) found an  excess of respira-
tory system  cancer in one MW plant of eight U.S. MW or FG  plants.   A ninefold
risk was  found  among  jnale production or  maintenance workers  with  20 or more
years of  latency compared to a 2.5-fold  risk among  those  with less than 20
years of  latency,  even though the power  to  detect  a 50 percent increase in
respiratory  cancer mortality was low (12  and 4  percent, respectively).   In the
second in a  series of four  studies  of  U.S.  MMMF production and maintenance
workers,   Enterline et al.  (1983)  reported  an  overall  excess of respiratory
system cancer among all  MW workers (SMR = 160.1,  p <0.01).   As in  the previous
study, the excess  was greatest in Plant 7 (SMR = 451.3, p  <0.01) and resulted
in a  ninefold  increase after 20  or more years of latency even  though there was
insufficient power to detect respiratory  cancer  in MW  workers.  In the third
study, despite  low power,  Enterlin* et al.  (1986) found a  46 percent  increase
in respiratory  cancer among  workers after 20 or  more  years of latency. The
excess was slightly higher after 30 or more years  of  latency (SMR « 151.7,
p <0.05).  In  the   fourth  study,  Enterline et al. (1987) found  that the 46
percent  increase in respiratory  cancer  among MW workers after  20 or more years
of latency was  no  longer significant when compared  to  local mortality  rates.
However,   as  in  the previous  studies, there was Insufficient power to detect  a
50 percent increased  risk of death from respiratory  cancer  among those  with 20
or more  years of latency.  Compared to  FG wool  and continuous  FG filament, the
greatest  respiratory  cancer  hazard was  found among MW  workers (SMR « 148.4,
p <0.01), which  remained significantly elevated  compared  to local mortality
rates (SMR * 133.6, p <0.05).  The respiratory cancer excess among MW workers
is above  background rates, thereby taking into  account etiological agents such
as smoking  and  regional  mortality differences,  which  could  influence  the
July 1988                          8-99         DRAFT--DD NOT QUOTE OR CITE

-------
magnitude of  the  risk.   Robinson  et  al.  (1982) found a 3.33-fold increase from
digestive tract cancer  among Bale MW production workers after 20 or more years
of employment.  There was insufficient power to detect excess mortality from
any site-specific cancer mortality.
     In  Europe,  Andersen and  Langmark (1986) found a  twofold increase in
Intestinal cancer  Incidence  among male -MW workers employed at least  1  year.
There  was  insufficient power  to  detect this excess  as well as excess  lung
cancer.  Nevertheless,  there was  a threefold increase in lung cancer incidence
in one  rock  wool  plant, which became more pronounced when  short-term  workers,
i.e., those  employed  for less than 1 year, were excluded (SIR = 386, p <0.01).
A twofold increase  in lung cancer incidence three MW plants after 20 or more
years of  latency  was  concentrated among those with 1 or more years  of employ-
ment.  Westerholm and Bolander (1986) found an 8.8-fold excess  from  lung cancer
among male and female MW production workers with 1 to 4 years of employment  and
30 or more years of latency, even though there was insufficient power  to detect
lung cancer.   Olsen and Jensen (1984) found a 52 percent increased  incidence  of
digestive organ/peritoneum cancer among male and female MW production workers
during the first  10 years after  initial  employment.  They  also found  a  twofold
increase  in  lung  cancer  incidence «»ong male and female  production workers
after 20  or  more  years  of latency (Power  =  20 percent).   This increase was
concentrated  among  male production workers.   In a  followup  study, Olsen et al.
(1986) found an increased incidence from buccal cavity/pharyngeal cancer (SIR =
210, p <0.05,  Power = 24 percent) and from digestive organ cancer  (SIR  = 150,
p <0.05,  Power =  90  percent) among  male  production workers.   Claude  and
Frentzel-Beyme (1984) found  an excess of stomach cancer among  male  production
workers  after 20  to 29  years of  latency (SMR = 273, p <0.05), despite low
power.  In a followup study among male production and maintenance workers hired
before  the introduction of resin binders in 1953, Claude  and  Frentzel-Beyme
(1986)  found  excesses  from stomach cancer after 20 to 29 years of latency (SMR
= 261,  p <0.05),  thereby confirming  the previous  finding,  and  from  lung cancer
after  10 to  19 years of  latency  (SMR = 272, p <0.05,  Power « 12  percent).
Maintenance workers had a 2.5-fold excess of lung cancer  after 20  to  29 years
of  latency.   Using internal  controls as  a  reference  group,  there was a
sevenfold increase  in  relative risk  from pancreatic  cancer.  Pooling  data from
the European studies, Simonato et al. (1986a) found excess lung cancer among MW
workers  first  employed  during the early technological  phase of MMMF production

July 1988                          8-100             DRAFT--DO NOT  QUOTE OR  CITE

-------
compared to  both  national  (SMR * 214, p <0.05, Power =  14 percent) or local
rates  (SMR = 257, p <0.05, Power = 24 percent).   This excess was even greater
after  20 or more years since first exposure to MW  during  the early phase (SMR =
295,  p <0.05,  Power «  8 percent).    In  followup  studies,  Simonato et-al.
(1986b,c, 198?) found lung cancer excesses in Factory 4 (rock wool) compared to
both national (SMR = 360, p <0.05) and local rates (SMR e 300, p <0.05).  Among
those with 20 or more years of latency, a twofold  excess  of cancer was found in
preproduction/production rock wool workers  compared to local rates (Power = 28
percent).  Among  rock wool  workers with  20  or more years of latency, a twofold
excess  of  lung  cancer was  found for  the early and intermediate technological
phases  compared to  both national  and  local  rates.  This  lung cancer excess was
concentrated among  rock  wool  workers  first employed  in the early phase (SMR =
257,  p <0.05),  particularly after 30  or more  years  of  latency,  compared to
local rates (SMR = 295,  p <0.05).   Despite low power, Simonato et al.  (1986fc,c,
1987)  also  found  a  twofold increased incidence  of buccal  cavity/pharyngeal
cancer  in rock wool workers (SIR - 197, p <0.05),  which increased with increas-
ing  latency.   The low power to  detect site-specific  cancers in many subcatego-
                                                       *
ries that was  found in  several  of the individual  European studies pertained to
the Simonato studies even when the data were pooled from  the 13 plants.
     Combining U.S. and European MW workers, Doll  (1987)  found that  compared-to
local  rates, the  lung  cancer excess among MW workers was greatest  among  those
with  30 or more years  of latency (SMR = 148,  p  <0.05,  Power  * 74 percent).
This excess was concentrated among those employed  in  the  early and  intermediate
technological phases but only significant  when compared  to national rates (SMR
= 223,  p <0.001).
     Thus, it  appears that among  MW workers,  significant excesses existed from
respiratory  system  cancer  (Enterline  and Marsh,  1982;  Enterline et  al.,  1983,
1986,  1987), including lung cancer (Andersen and  Langmark,  1986; Westerholm  and
Bolander,  1986;  Olsen and  Jensen,  1984;  Claude  and Frentzel-Beyme,  1986;
Simonato et  al.,  1986a,b,c, 1987; Doll, 1987);  from digestive  tract cancer
(Robinson et al.,  1982), including  stomach cancer (Claude  and Frentzel-Beyme,
1984,  1986), digestive  organ/peritoneal cancer   (Olsen  and  Jensen, 1984),
digestive organ cancer  (Olsen et al., 1986), Intestinal  cancer (Andersen and
Langmark, 1986),  and pancreatic cancer (Claude and Frentzel-Beyme,  1986);  and
from  buccal  cavity  and  pharyngeal cancer (Olsen et al.,  1986; Simonato et al.,
1966b,c, 1987).  Additional excesses  of buccal cavity/pharyngeal  cancer occurred

-------
in Norwegian MW  and FG wool workers (Andersen and Langmark,  19B6),  but  it was
not possible to attribute the excess to one or the other group.

8.1.2  Ceramic Fibers
8.1.2.1  Cross-Sectional Studies. Clinical  Evaluations,  and Case Reports
8.1.2.1.1  Respiratory effects.   Limited information was  found  on the  health
effects of  ceramic fibers.   Gantner  (1986) determined the  exposure  levels
resulting from conversion  of ceramic  fiber to  cristobalite and  the dust gener-
ated  by  removal  of  ceramic fiber insulation  from heat  treatment  furnaces.
Ceramic fiber  insulation consists of  alumina-silicate fibers that are vitreous
in their  initial,  manmade  state.  When  subjected to high temperatures over
extended periods of  time,  however, the fibers  devitrify and  are converted into
cristobalite, a  form  of free silica.   Temperature appeared to be the  control-
ling  factor  in transformation of ceramic  fiber, with highest percentages  of
Cristobalite present  closest to  the  hot  face and none  found beyond the  4-inch
layer.  During  removal of  the insulation, personal respirable dust  levels
exceeded the threshold limit value (TLV)  in. 75 percent of the  samples.  Per-
sonal total  dust samples  exceeded the TLV  in  80  percent  of  the samples.  TLV
was calculated using the average percentage of cristobalite found in total  dust
and respirable dust  samples.   Workers appeared to be at  significant  risk for
cristobalite exposure.  According to Ziskind et al.  (1976), who  studied  workers
occupationally exposed to   silica,  silicosis  may be caused  by  exposure to
elevated levels  of cristobalite  over  extended periods  of time.   Silicoses  are
associated with  rapid deterioration  of  pulmonary function  characterized by
restriction  of  lung  volume  with a reduction  of total  lung  capacity,  vital
capacity,  and pulmonary diffusing capacity.
8.2  SYNTHETIC FIBERS
8.2.1  Aramid Fibers
     No  information  was  available concerning the human health effects of this
fiber.
i..i.. 1000                          8-102 .            DRAFT--DO NOT OUQTE OR CITE

-------
8.2.2  Carbon Fibers
8.2.2.1  Cross-Sectional Studies, Clinical  Evaluations,  and  Case  Reports
8.2.2.1.1   Respiratory  effects.   Jones et  al.  (1982) conducted radiograph!c,
spirometric, and  respiratory  questionnaire  exams  of 88 workers from a carbon
fiber plant.   Little  variation had occurred over the years in the  levels  of
respirable  dust concentrations,  which  varied from a mean of 0.03 to 0.16 tng/m
in different production areas.   Diameters ranged from 8 to 10 urn.  No evidence
existed  that  carbon fibers caused  any 111  effects  in the lungs of exposed
workers.
8.2.2.1.2   Effects from surgical introduction.   Rosenthall (1987) studied  the
effect of  a knee  prosthesis with carbon fiber-reinforced polyethylene surfaces
on production  of  an inflammatory reaction.   A foreign body reaction in three
patients  with  total  knee  arthroplasties was  demonstrated  by technium-99rc
methylenediphosphonate  (MDP) deposition in  the  intra-articular space.   However,
there was  no visible  uptake of   Ga citrate.   The inflammation was shown to  be
due to a synovia! joint cell  foreign body reaction  to particulate carbon fiber
debris in  one  patient with correlating data from a  histological  examination  of
the synovium.

8.2.3  Polyolefin Fibers.  No  information was  available concerning the human
health effects of this  fiber.
8.3  MINERALS
8.3.1  Attapulgite and Sepiolite
8.3.1.1  Cross-Sectional Studies, Clinical Evaluations, and Case Reports
8.3.1.1.1  Respiratory effects.   Limited  information  exists  on the effects of
attapulgite  and  sepielite  in humans.  In a  cross-sectional  morbidity study,
Gamble et al. (1985) examined the prevalence of respiratory symptoms, pneumoco-
niosis, and  reduced  pulmonary function 1n U.S.  attapulgite workers.   The  study
population consisted  of 705  men from two  attapulgite mining  and milling
companies (355 men from Company A and 310 men from Company B)  who  were  similar
in age, tenure,  and smoking habits.   Over 50 percent  of the  workers  in  Company
6  had  been employed in  dusty occupations other than  attapulgite  mining  and
milling, most  for less  than 5 years.  Mean  estimates  of cumulative exposure  to
total  and  respirable dust  were calculated  from  environmental survey date.

July 1988                          8-103             DRAFT-DO NOT QUOTE OR CITE

-------
Total dust  concentrations  varied between 0.09 and 22.96 mg/m ,  and respirable
dust varied  between  0.05 and 2.71 mg/m3.  Medical examinations, performed by
the two  companies,  consisted of a respiratory  questionnaire, chest x-rays, and
spirometry.   Lung  function data  included  the  maximal  forced vital capacity
(FVC), forced  expiratory volume in 1 second (FEV^, and FEVj/FVC x 100 (FEVj
percent).  Logistic regression was used for analysis of the association between
exposure  and respiratory symptoms  and  radiographic findings.   The observed
prevalence  of  symptoms  and  the odds ratio  (OR)  by exposure category were
calculated,  controlling  for  race,  age,  height, and smoking in each model.   No
consistent relationship  was  found between  exposure and respiratory symptoms.
FVC and  FEV- were generally lower in the medium*  and high-exposure groups,  but
the differences were not statistically significant.
     A total of  701  x-rays  obtained from both  companies were read  for  evidence
of pneumoconiosis.   The  prevalence of pneumoconiosis was elevated  in the older
age categories and  the prevalence was consistently increased in the  highest
exposure groups, but these  trends  were  not  significant.  There was  no  associa-
tion between reduced  FVC  and FEV- and  radiographic findings.  The authors
concluded that their data  did not provide evidence for a relationship between
attapulgite exposure and respiratory morbidity.
     Bignon  et al.  (1980)  described case reports  of two  individuals exposed  to
attapulgite fibers, one by inhalation and the other by ingestion.   A 41-year-old
French worker  with  lung  fibrosis had been exposed  to  attapulgite  dust for  3
years while  mining  and processing the mineral for  use  in  the  pharmaceutical
industry.  Attapulgite fibers were  found  in  lung  washing fluid  following
bronchoalveolar  lavage at a  concentration  of 42,000 fibers/cm .   The mean
length and  diameter of  the fibers were 1.5 pm and 0.11 urn, respectively.
Attapulgite  fibers, at a concentration of 300,000  fibers/cm , were  also detected
in the urine of  a 60-year-old woman who had been  treated for gastrointestinal
disease  for  6  months with 9 g/day of a  drug containing  attapulgite (Attapul-
gite® consisting of  3 g  activated Mormoiron attapulgite In  a 6-g  pack).   No
adverse effects attributable to the fibers were reported.
     Baris et  al.  (1980) performed mass  photofluorograms of  the inhabitants  of
four Turkish villages where mining and processing  of sepiolite  had occurred  for
more than 100 years.  A clinical and radiologic survey was  also conducted on 63
sepiolite workers whose exposure duration ranged from 1 to 30 years with a mean
of 11.9 years.   Mineral fiber-related pleura! diseases were not detected in  the
villagers  exposed  to  sepiolite dust.   Calcified  pleura!  plaques, pleura!
July 1988                          8-104             DRAFT—DO  NOT QUOTE OR  CITE

-------
thickening, and mesothelioma  observed  in the villagers  were  attributed to the
home  use  of a  stucco finish that contained tremolite  asbestos.  Pulmonary
fibrosis was found  in 10 of 63 workers.  However, more than half of the 10
workers resided  1n dusty  rural  regions where  asbestos and zeolites were
present, 3  had  previous  occupational  histories that  nay have been responsible
for or  contributed  to the  pulmonary disorders,  and all  were  smokers exhibiting
symptoms of chronic bronchitis.  Therefore,  no causal  relationship between
pulmonary fibrosis and sepiolite dust inhalation could be established.
8.3.1.2  Epidemiologies!  Studies.   Waxweiler et al.  (1985) conducted a histori-
cal prospective mortality study of employees of a large  U.S.  attapulgite  mining
and milling facility.  The cohort consisted of 2302 white and  nonwhite  male
miners  and  millers employed  at least  1 month  between  January  1, 1940,  and
December 31, 1975.   Expected  deaths were based on U.S.  age,  race, and calendar
year-specific mortality rates.  Vital  status was determined through  information
obtained from  the Social  Security Administration, Internal  Revenue Service,
state bureaus  of motor  vehicles,  state vital  statistics offices,  and other
sources.  Of  the 2302 workers,  1847  (80.2  percent) were alive, 315 (13.7
percent) had died, and 140 (6.1 percent) were lost to followup.   Death certifi-
cates were obtained for all but 19 (6.0 percent) of the  decedents.  Deaths  were
coded by a  qualified nosologist and converted to the 7th Revision of the ICD.
Jobs were classified  from  analysis by dust exposure levels into high-exposure
(1.15  to  1.60 mg/m   respirable  dust),  low-exposure  (0.45 mg/m  respirable
dust),  and  intermittent-exposure  (occasional   work  in  high-exposure jobs)
categories.   Although  there was a maximum followup  of  36 years,  exposure and
observation periods  were concurrent,  thereby underestimating the risk of lung
cancer.
     Personal exposure data were  unavailable for the cohort, but area monitor-
ing data were collected  in 1976.  The monitoring survey indicated that airborne
fibers  in the  plant were generally short 1n length, I.e., 0.1  to 2.5 um, and
narrow  in  diameter,  I.e.,  0.02 to 0.1 pm.   Fiber aspect ratios (length to
diameter) were  generally greater  than  10:1.   Respirable dust concentrations in
all  of  the monitored work areas  were  less than the  Mine Safety and Health
Administration (MSHA) nuisance dust standard of 5 mg/m  .
     There  were  no  significant Increases in cause-specific  mortality for the
whole cohort.   When the  cohort was divided  into white  and nonwhite workers, a
significant excess  of lung cancer was  seen  in white males  (16 observed, 8.3
expected, SMR = 192, p <0.05, Power =  37 percent), but  a  significant deficit of
July 1988                          8-105             DRAFT—DO  NOT  QUOTE OR CITE

-------
lung cancer  was  seen in nonwhite males (5 observed,  9.4  expected,  SMR =  53,
Power = 34 percent).  This excess in white males persisted after 10  years  since
first  employment (16 observed,  7.2 expected,  SMR = 222, p <0.05), but was
limited to  those employed from 1 month to 5  years (10 observed,  4,3 expected,
SMR  =  233,  p <0.05)  rather than those with 5  or »ore years of employment
(long-term  workers).   A threefold  excess of  lung cancer was observed among
white  males  employed In  the high-  and intermittent-exposure job categories
after  20  or more  years of  latency (6 observed, 2.0 expected, SMR = 300,
p <0.05), which was greater among long-term employees (3 observed, 0.7 expected,
SMR =429, p <0.05, Power * 9 percent).
     The investigators concluded that (1)  whereas white employees experienced  a
twofold excess of  lung  cancer, nonwhites  had a lung cancer risk approximately
half of that expected based on U.S. mortality  rates;  (2)  lung cancer was not
altered by  dust  exposure  level,  latency  period, or duration of employment,
except for  those employed at least 5 years  in high-exposure jobs; and  (3) the
deficits from NMRD are consistent with the hypothesis that short fibers  are  not
fibrogenic.  The  investigators  stated  that the study was  limited by  (1)  the
absence of  smoking histories;  (2) the incomplete enumeration of  the  cohort;
(3) the large number  of individuals lost  to  followup; (4) the missing  race  on
50 persons; (5) the lack of longitudinal  person-specific dosimetry;  (6)  the  fe«
expected deaths  among those with 20 or more years  of  latency; and  (7) use of
national rather than  local mortality rates as a comparison group.
     In addition to  the above  limitations,  the  power  to  detect  site-specific
cancers, such as lung cancer,  as well as NMRD,  was  insufficient; there was a
low  trace  rate;  and exposure and observation periods were concurrent.   All  of
these limitations served to underestimate  the true risk of lung cancer and NMnD
among this group of workers.  Despite the  low power, since exposure  to asbestos
or other  lung  carcinogens  was  not mentioned, it  appears  that attapulgite was
associated with  excess  lung cancer  mortality.   Without local mortality rates
for  lung cancer,  however, this cannot be stated with certainty.

8.3.2  Erionite
     Results from  several  cross-sectional morbidity and mortality studies have
shown that  erionite  induces pleura! and peritoneal mesotheliomas among adult
inhabitants  from three  Turkish  villages.  The  annual  crude death  rate from
malignant mesothelioma  (both pleura!  and peritoneal)  was 8 deaths per 1000

July 1988                          8-106             DRAFT—DO NOT QUOTE OR  CITE

-------
inhabitants  of  the three affected villages.   In addition,  radiological data
Indicate a  higher  prevalence  of pleura! changes in villages affected  with a
high prevalence of mesothelioma as compared to a control  village.  The  presence
of asbestos  did not correlate with the distribution of mesothelioma in these
villages.   Turks who  emigrated to Sweden exhibited  immunological  effects
similar to those found in asbestos-exposed Individuals.
8.3.2.1  Cross-Sectional Studies. Clinical Evaluations,  end Case  Reports
8.3.2.1.1    Respiratory  effects.    Baris   et al.   (1987)  conducted   a
cross-sectional  Morbidity  and mortality  study  and  reported 16 new coses of
pleura! disease  between  197S  and 1976  from Karain,  a remote village located in
the Anatolian  region  of  Turkey.   The study group consisted  of  575 residents of
Karain, whose  name  translated from Turkish means "abdominal pain."  Because no
formal records  were  kept by a vital statistics office  for the period up to
1970,  the  investigators  relied upon information obtained  from relatives  and
from hospital admissions  from Karain.   Survey radiographs taken in 1971  and
available information  on  deaths  between 1970 and 1974  were reviewed.   For the
period from  1975 to 1976,  the investigators  studied physical and socioeconomic
conditions and  performed  geological, biological,  and chemical  tests to detect
environmental  toxicants  in Karain and neighboring villages.   These  tests
included sampling  airborne dust;  performing elemental  analysis  of soil  and
rock;  sampling  food,  drink, and water  supplies; and conducting a health survey
of the inhabitants of  Karain and neighboring  villages.   The  health  survey
included chest  x-rays  and  physical  examinations and clinical tests  of  hospital
admissions.    Occupational,  domestic,  and familial  medical histories were
obtained for  new cases of chest disease that developed between 1975 and 1976.
The  investigators  stated that  causes  of death  during  1975 and 1976  were
recorded and verified to the maximum extent possible.
     Although most  of  the  information  sources up to 1975 were  anecdotal,  more
recent  information  from medical  records  showed   that  serofibrinous  or
serohemorrhagic pleural  effusion was present among respiratory disease admis-
sions.  Tests for tuberculosis were negative, and antituberculosis  treatment up
to 1975 proved  ineffective.   In 1974,  for Karain's 604 inhabitants, 11 of the
18 deaths that  occurred were  attributed to pleural mesotheliomas.   Comparison
of total  deaths in  Karain and eight  neighboring villages  in  1974 revealed
significant  excesses  from all causes  of  death in  Karain  (18  observed, 6.60
expected, SMR » 273, p <0.001) and in Akcaoren  (11  observed, 5.94 expected.  SMR
e 185, p <0.05).   Although the source  of  the expected  number of  deaths was  not
July 1988                          8-107             DRAFT--DO NOT QUOTE OR CITE

-------
identified, the  investigators  stated  that these two villages had  a  signifi-
cantly higher  death  rate (chi-square  * 30.45, 8 degrees of freedom) than the
other villages.  Of  greater importance,  however,  was the concentration of 11
pleura! mesotheliomas in Karain and the absence of mesotheliomas  in each of  the
eight other villages.   When the investigators  compared results of  chest x-rays
taken during health  surveys of 1971 and 1975,  they found no significant change
in the general pattern of chest diseases 1n Karain.
     The  investigators  also reported that some of the men  from Karain were
smokers; women never smoked.  The village has no asbestos  deposits  and  asbestos
brought in from  other areas has not been  processed.  Arsenic, uranium, nickel,
and chromium were present in trace amounts.  Extracts of  soil,  dust,  and stucco
were negative  for  benzo(a)pyrene.   No significant radioactivity was found  in
the village  or surrounding soil or rocks.   Tests  for aflatoxins  in various
foods were negative.   Some water supplies contained small  quantities of uniden-
tifiable fibers.   Actual exposure levels were unavailable.
     The mass  radiographic  survey  showed  no significant change  in the preva-
lence of  chest disease between 1971 and  1975.  During 1975  and 1976,  16  new
cases of  pleura! disease were found.   The final diagnosis of mesothelioma was
based on  radiological evidence  in  11 of the  16  cases  and on histologies!
evidence for 7 of  the cases.   The  five other cases were diagnosed  with chronic
fibrosing  pleurisy.    The  cases mainly   had  agricultural   occupations.
Geneological  information from  medical  histories indicated that blood relation-
ship was  not  a prerequisite for mesotheliomas within a  family; the authors
suggested a common environmental etiology.
     The investigators  concluded that  apart  from a few asbestos  fibers identi-
fied in the drinking water and in two of five samples  of pleura!  tissue, no
obvious environmental  or occupational  source of asbestos exposure existed  in
Karain.   They suggested that a patient might have been exposed to asbestos frorr,
an undetected source or that tapwater might be the source of  asbestos.   However,
they did  not  explain why,  if  asbestos were  the causative agent,  the  excess
malignant respiratory  disease  was  limited to  Karain despite the proximity  of
other villages.
     Ban's et  al.  (1979)  reported  the  results  of cross-sectional morbidity  and
mortality  studies  of asbestos-related diseases, primarily calcified pleura!
plaques (CPP), chronic  fibrosing pleuritis  (CFP), and malignant pleura!  meso-
thelioma  (MPM) in  Turkish villages  with  or without asbestos  deposits.  Because
certain parts of Turkey are rich in asbestos minerals, this study was to expand
July 1988                          8-108             DRAFT--DO NOT QUOTE OR CITE

-------
previously conducted surveys in areas surrounding asbestos  mines  and mi 11s  into
rural regions  of  Turkey.   The investigators summarized their findings  for  six
Anatolian regions  of  Turkey by stating that CPP, CFP, and MPM are endemic to
some villages with or without asbestos deposits.   Although  asbestos in  building
materials had  been postulated as a cause, environmental samples  in villages
where CPP 1s endemic revealed the absence of asbestos.   In  fact,  the prevalence
of  CPP  did not  appear to  be  correlated with asbestos-containing building
•aterials.  Because erionite was  detected  1n soil, rock,  and lung tissue
samples  in areas  where asbestos is absent and where the diseases have  shown  a
geographical distribution,  the  investigators  stated that CPP, CFP, and MPM may
be  related to  zeolite minerals rather than to asbestos.  When additional  soil
samples  near Tuzkby revealed respirable zeolite  fibers, Baris et al.  (1979)
concluded that tumor  induction by asbestos contamination of drinking  water,
previously postulated  ty  Baris  et al.  (1978), was highly  improbable.  Actual
exposure levels were not reported.
     Baris et  al.  (1981)  conducted a cross-sectional  morbidity and mortality
study of Karain (population 554) and its neighboring village of Karlik  (popula-
tion 479) to determine whether erionite  might be responsible for  the very  high
rates of mesotheliomas and at which concentrations erionite might be  associated
with this  excess.  The  investigators  compared  the  inhabitants  of the two
villages for several  demographic characteristics, assessed  the prevalence of
radiological chest abnormalities, and  estimated concentrations  of  mineral
fibers  in air  and water samples and their  size  and, composition.  Demographic
data were abstracted  from health center registers, which  contained birth  and
death rates.   Death rates were obtained by a direct age-adjustment  procedure,
with the  combined populations of the two villages as a standard.  No medically
certified cause  of death  was  recorded.   All  persons aged 20  and over were
invited  to participate in the interview, chest x-ray,  and  physical examination
portions of the  study:  89.8 and  90.8 percent of the  men and women,  respec-
tively, in Karain, and 83.4 and 77.1 percent of the men and women, respectively,
in  Karlik  participated.   Interviews  conducted by a  local   resident lasted
approximately  4.5 minutes  and Included questions on personal,  family,  residen-
tial, and occupational history.   Less than 2  percent  of the interviews were
discarded as  unreliable by  an  observer.   Full-size posterior-anterior chest
x-rays  were  taken; films  were read blind  by Baris.   Respirable and  total
environmental  dust was measurable for 8 to 10 hours per day  over a minimum of 3
days.  Samples were analyzed by optical and electron microscopy.
July 1988                          6-109             DRAFT-DO NOT QUOTE OR CITE

-------
     The sex  and  age compositions of the two  villages  in  1978 were  similar.
Whereas no  significant  differences  were found for educational  status,  occupa-
tion,  lifelong residence,  proportion  of consanguinous marriages,  original
residence of spouse, and cigarette smoking habits, life expectancy was  signifi-
cantly reduced in Karain (p <0.01) by 14 years and adult mortality was  70  to  80
percent higher than 1n  Karlik.   However, Investigators found no statistically
significant differences  In  pleura!  radiological abnormalities.  The prevalence
of mesothelioma diagnosis by x-ray in Karain adults was 3.4 percent (11/325)  in
1979.  Concentrations of respirable fibers from the street  atmosphere, from
work and recreational areas,  and from indoor  locations were uniformly low in
Karlik (at  or below 0.1 fibers/cm ) and higher in  some  samples from Karain,
with  erionite representing  80  percent  of  the fibrous material.   Neither
asbestiform nor erionite fibers were detected  in  the three water  samples.  The
investigators concluded that these findings were compatible with the hypothesis
of a causal association between endemic mesothelioma and inhalation of erionite
fibers, but the  fiber  concentrations in all samples were so low that the  role
of erionite as an etiological agent is questionable.
     Ban's  et al.  (1987) updated and completed previously  published results
(Baris et al., 1981) of an  association  between exposure  to  naturally occurring
fibrous erionite  and malignant  mesothelioma.   In 1979, the  adult populations
(aged 20 or more) of Karain, with  known high  numbers  of mesothelioma, and of
Karlik,  with  no   known cases  of  mesothelioma,   were selected  for  a
cross-sectional morbidity  and  mortality  study.   In 1980,  a third village,
Sarihidir,  was  included because  of  the  presence  of erionite fibers in soil
samples but with  no  mesotheliomas reported up to 1980.   Adults were interviewed
and  given  chest  x-rays  and physical examinations.  X-rays  were read without
knowledge of  village at the International  Agency  for Research  on  Cancer (IARC)
in  Lyon.   Although  there  was no followup of  study subjects,  health center-
records, hospital  clinical  records, and other  sources  were  used to establish a
list  of  deaths that occurred between June  1979 and July 1983  in  Karain  and
Karlik and  between July 1980 and July  1983 in Sarihidir.   For the mortality
portion  of  the study,  a fourth  village in the same province, Turkey, was
included for  deaths that occurred between  1980 and 1983.   Only deaths among
those  aged  20 and older were included in the analysis.   Comparison groups were
the  1978 populations of Karain and  Karlik,  the  1979 population of Sarihidir,
and  the 1980  population  of Tuzkb'y.

July 1988                          8-110             DRAFT—DO  NOT QUOTE OR CITE

-------
     Airborne  levels  of  respirable  (diameter <3 um) and total dust were mea-
sured  in  all  four villages.   More than  150  samples  were obtained by the method
recommended by the Asbestos International Association (AIA) and analyzed by
transmission  electron microscopy.   The U.S.  Environmental  Protection  Agency
(U.S.  EPA)  analyzed  two  water samples from  Karain, one from  Karlik, and two
from Sarihidir.   Lung tissues from  sheep 5  or »ore years  old from all area
villages except Karain, which had no sheep,  were analyzed  for  mineral content.
     The  study group consisted of 32? people from Karain, 228 from Karlik, and
337 from Sarihidir.  Participation 1n the study was  89.8 and 90.8 percent males
and females,  respectively,  1n Karain; 83.6  and 77.1 percent males and  females,
respectively,  in  Karlik;  and 93.3  and 83.2  percent  males  and  females,  respec-
tively, in Sarihidir.  The age structure of  the three  villages was  similar.   No
major  differences  were found between the villages,  except that consanguineous
marriages and  illiteracy  were higher in Sarihidir.   The major occupation was
farming, which  ranged from  89.6 to  100  percent  in the three villages.  None of
those  interviewed  had experienced either  known or suspected exposure to asbes-
tos.    In  all  three villages, more  than 50  percent  of  the males were  either
current or  ex-smokers; only  2  of 482 (0.4   percent) of the females  smoked.
Among current smokers, 70 percent smoked 20  or more  cigarettes per  day.
     The entire population  was  x-rayed  in Karlik except for  1.8 and 3.0 pe'—
cent,  respectively,  who  were x-rayed in Karain and  Sarihidir.  Round opacities
of category 1/0 or more  were more  common in males  from all three villages;
males from Sarihidir exhibited a significantly higher  prevalence (p <0.01)  than
those  from  Karain and Karlik.   Calcifications of  the  diaphragm were more
frequent  among both males and  females  from Karain  and Sarihidir than among
those  from  Karlik;  this  difference  was  statistically significant (p <0.01) for
males from Sarihidir as compared to males from Karlik.
     During the study period, 141  deaths were recorded in  the four villages:
29 from malignant pleura! mesotheliomas,  4  from malignant peritoneal  mesothe-
liomas, 17  from  lung cancers,  1 from laryngeal  cancer, 21 from unspecified
cancers, 59 from  other causes, and 10  from  unknown causes.   Irrespective  of
sex, the  largest  number  of deaths  from pleura! mesotheliomas  (21  out of 29)
occurred in Karain.  There were three pleura! mesotheliomas from Sarihidir,  but
no peritoneal mesotheliomas, and five pleura! and four peritoneal mesotheliomas
limited to women  from Tuzkby.  Lung cancer prevalence was highest in Sarihidir.
No cancer deaths  occurred in Karlik.  When  age-  and  sex-specific  death  rates

July 1988                          8-111             DRAFT—DO NOT QUOTE OR CITE

-------
 were  calculated, Karain  had  the  highest overall  death rate and the  highest
 death  rate  from  malignant mesothelioma.
      Zeolite  fibers constitute the  major portion of dust  1n  the  clouds of
 Karain,  Tuzkby,  and Sarihidir.  Tremolite was observed 1n Karain and  Karlik at
 less than 0.001  fiber/cm3.  In Sarihidir, chrysotile and tremolite fibers <5 »n
 in  Itngth  contributed up to  10 percent  of  the fibers  in this  size range.   In
 Tuzkby,  no  asbestos fibers were detected; the shorter fibers aainly consisted
 of  zeolite, calcite, and Muscovite.   Thus, airborne fiber concentrations were
 low but  showed a higher proportion of erionite In villages affected by mesothe-
 lioma  than  in the control village.   The same pattern was found in  fiber levels
 from the lung tissues of sheep from villages affected and unaffected by meso-
 thelioma.   Analysis  of water  samples performed by the U.S. EPA showed that none
 of  the samples  contained asbestos, zeolite, or other mineral fibers above the
 detection limit  of  50,000 fibers per liter.
      The investigators  concluded  that the results of this study confirm previ-
 ous  findings  of  an extremely high rate  of  malignant  mesotheliomas in some
 villages of central  Turkey.   They further concluded that (1) the cancer mortal-
 ity excess  was  not  limited to pleura! mesotheliomas (peritoneal mesotheliomas
 and lung cancer  were also found); (2) the similarity of the demographic charac-
 teristics  of the villages  enabled  morbidity  and mortality  comparisons;
 (3) radiological  data indicate a  higher prevalence of  pleural changes  in
 villages affected by high prevalence of mesotheliomas compared to the control
 village;  (4)  the annual  crude death  rate from malignant mesothelioma  was  6
 deaths  per 1000 inhabitants  of  the  three  affected villages; (5) because
 measurements  showed higher airborne  erionite levels in affected versus control
 villages, the prevalence of mesothelioma and lung cancer was related to expo-
• sure to  erionite  fibers,  and  (6) some documented exposure to asbestiform fibers
 does not appear  to  correlate  with the distribution of the disease.
      Artvinli  and  Ban's  (1979) conducted  a cross-sectional morbidity  and
 mortality study  and  described three hospital admissions from Tuzkby.  Tuzkby is
 50  km  from  Karain,  the  village where Baris  et al.  (1978) previously found
 excesses from malignant pleural mesothelioma (MPM).  The reference population
 was drawn  from the  neighboring village  of Kizilkb'y.  Both Tuzkby  and Kizilkby
 are located in the  city of Nevsehir.  For the cross-sectional study, persons at
 least  25 years  old  were  selected and divided by age and sex into  four 15-year
 age groups, i.e., 25 to  39,  40  to  54,  55 to 69,  and 70*.   Through stratified
 sampling, 312 persons (145 males, 167 females) from Tuzkby and 95 persons  (45
 v.i.. 1000                          8-112             DRAFT--DO  NOT QUOTE OR CITE

-------
males, 50  females)  from Kizilkby were included in the study.   A  questionnaire
on smoking  habits  and personal and family medical history completed for each
person was  followed by physical examinations, x-rays, and lung function stud-
ies.  In  addition,  chemical  analyses  of water samples and mineralogical analy-
ses of building materials and soil samples were performed.
     Of the three  hospital  admissions,  one 42-year-old male  fanner had MPM,  a
31-year-old male salt-miner  had a malignant epithelial tumor, and a 36-year-
old male  had chronic fibrous pleuritis.   No  ferruginous bodies were found in
the lung  tissues;  among the study group, x-ray examinations yielded 53 cases
(17 percent) of pleura!  calcification,  33 cases (10.5 percent)  of  pleura!
thickening,  47  cases (15 percent) of  obscured costodiaphragmatic  angle, and  3E
cases (12.1 percent)  of  interstitial  pulmonary fibrosis.   In addition, two
cases of  MPM were  found in this group.   Among the control  villagers, no  cases
of malignant diseases  or pathological conditions upon x-ray examination were
found.
     No differences between the  water  samples from  the  study and  control
villages  were found.  No fibrous  minerals were found in  the water  samples.
Soil samples  from Tuzkb'y, on the other hand, contained multiple fibers, usually
30 to 40  pm long  and  1 to 2 pm in diameter,  which were classified as  zeolite
minerals, i.e., mordenite, erionite, or chabayite.  These  same fibrous minerals
had been  identified in soil  samples from Karain and were  also present in lung
samples from the three hospital admissions.   However,  no  fibrous  zeolites were
present in  the  control village.  No types of asbestos were  present  in either
village.   Exposure  concentrations were unavailable.
     The investigators concluded that zeolites were responsible for the MPM and
other respiratory  disorders  in Tuzkb'y.  They  estimated an  incidence  and preva-
lence of  MPM in Tuzkoy of 6.5  cases  per  year or 22 cases per 10,000 people.
Compared  to the annual incidence of MPM in Pennsylvania (1 case  per million),
in Finland  (1 per  million), and in the United Kingdom (2.3 per million),  they
further estimated  that the  annual Incidence  of MPM  in Tuzkb'y  1s  at  least 942
times higher than expected.
     Artvinli and Baris (1982)  updated their previous cross-sectional morbidity
and mortality study  (Artvinli  and Baris, 1979)  of  312 residents of Tuzkoy
compared to  95  residents of Kizilkb'y.  In addition, they analyzed the number of
deaths among those aged 15 or  older  in  both  villages from  January  1,  1976,
until December  3,  1980.   The methodology  remained the same as  that used  in  the

July 1988                          8-113             DRAFT—DO NOT QUOTE OR CITE

-------
previous study; only  new results win be reported here.  Exposure concentra-
tions were unavailable.
     Physical examination  revealed  that the Tuzkby group had diminished breath
sounds  (p <0.01).   Ronchi and  rales  occurred  more  frequently  in the Tuzkby
group than  in  controls (p <0.05).   All abnormal  x-ray findings,  except  inter-
stitial  pulmonary fibrosis,  were  significantly  less  common  in housewives
compared  to other  occupational groups,  e.g., farmers,  salt-miners, civil
servants, and other unspecified occupations.   Among  1780 Tuzkb'y inhabitants  who
were at  least  15  years old, 67 died between 1976 and 1980;  of  these, 41 (61.2
percent) of  the deaths  resulted from malignant  diseases  (15  from  pleura!
mesothelioma,  12  from peritoneal mesothelioma,  8 from lung cancer,  2  from
gastric  cancer,  2 from  colon cancer, 1 from  laryngeal  cancer,  and 1  from
leukemia).   Of  the  remaining 26 decedents,  19 died  from nonmalignant causes,
and 7 died from unknown causes.   In the control village during  the  same  period,
only 2  of 325  individuals aged 15 years and over died; both deaths  were from
nonmalignant causes.  No x-ray abnormalities or deaths resulting from malignan-
cies were found in the control group.
     The investigators  concluded that (1)  pathological x-ray findings,  i.e.,
calcified pleura!  plaques, pleura!  thickening, obscured costophrenic angles,
and diffuse  interstitial  pulmonary  fibrosis,  were observed  more frequently  in
the Tuzkb'y  group  than in the control  village;  (2) the majority of  deaths frorr
Tuzkby  were attributable  to malignant  diseases (pleural  and  peritoneal
mesotheTioma and lung cancer);  (3) the incidence  of  cancer  in Tuzkby is  greater
than that of asbestos insulation workers;  and (4) because  zeolite  rather than
asbestos was found  in soil samples and in  building  materials from  Tuzkby, but
not  in  the  control  village,  and because zeolite was detected in lung  and
pleural  biopsies  of  patients from  Tuzkby,  zeolite  was considered the  agent
responsible for these pulmonary and pleural  malignancies.
     Artvinli  and  Ban's  (1982)  reported  the  results  of a cross-sectional
morbidity study of 327 persons over the age of 20 from Karain,  312  over  the  age
of 25  from  Tuzkby,  and 302 over the age of 25 from  Sarihidir.   Study subjects
were identified through  x-ray screening  studies.  For each  subject, question-
naires,  physical  examinations,  and  respiratory function tests were  obtained.
Those with  abnormal  findings were admitted to a hospital for further clinica!
evaluation.   Environmental  and  tissue samples  were  also obtained.   Deaths from
malignant diseases were reported for those  over age 15 from Karain between 1975

-------
and 1980,  for  those over age 15 from  Tuzkby  between  1978 and 1980, and for
persons from Sarihidir  of unspecified ninimum age at death between  1978 and
1979.   All  death  certificates  were evaluated in these villages  by an unspeci-
fied  person.   For  each  of the three  villages  under  study,  an unspecified
control village with a similar population was chosen as  the comparison group.
      In Karain, 64  (19.6 percent)  of  the  327  study subjects had abnormal x-ray
findings;  specifically,  51 of  the  327 (15.6 percent) had pleura! thickening.
Malignant pleural mesothelioma accounted for 38 of the 47 (80.8  percent) deaths
from  1975  to  1980 among those 15 and older.   In Tuzkb'y, 171  (54.8 percent)  of
the 312  study subjects  had abnormal   x-ray  findings;  specifically, 53 (17
percent) had pleural  calcifications,  33  (10.5 percent)  had pleural thickening;
47 (15.0 percent)  had obscured costophrenic angle; and  38 (12.1  percent) had
diffuse  interstitial  pulmonary  fibrosis.   Malignant  pleural mesothelioma
accounted for 15 (22.3 percent) and malignant peritoneal mesothelioma accounted
for 12 (17.9 percent) of the 67 deaths  from  1978 to  1980 among those 15  and
older.  In addition, eight lung cancer deaths occurred.   In Sarihidir, 65  (21.5
percent) of the 302 study subjects had abnormal x-r«y findings;  specifically,
25 (8.1 percent)  had  pleural thickening.   Investigators found three malignant
pleural mesotheliomas, two malignant peritoneal  mesotheliomas, two gastrointes-
tinal  malignancies, and  one lung cancer for the period between 1978 and 1979.
However,  neither  abnormal  x-ray findings nor deaths  from  malignant diseases
(period unspecified)  were  found in control villages.    While  neither asbestos
nor erionite was  detected in any control village, environmental  samples  and
tissue samples  from  patients  of the  three  villages  under study  contained
erionite.   No environmental concentrations were reported.
     The investigators  concluded that (1) the  existence of  asbestos-related
radiological findings and  diseases in three villages  in the absence of asbes-
tos suggests a  cause  other than asbestos;  and (2) since all  samples  have shown
the presence of erionite fibers, erionite rather than asbestos was  responsible
for the same pathological changes seen in those exposed to asbestos.
     Recently,  a  number of  Investigators  have  found  several  different
asbestiform minerals  in  addition to zeolite  in  environmental and lung tissue
samples from Karain and Tuzkb'y.  These findings  have  raised the question of
whether asbestos  alone  or a synergistic effect  involving both  asbestos and
zeolite may be causing  the  high prevalence of respiratory diseases  in this
region.

July 1988                          8-115             DRAFT—DO NOT QUOTE OR CITE

-------
     Rohl et al.  (1982)  conducted a  detailed analysis of environmental samples
(including house  dust,  soil  samples,  and building materials) obtained from the
Turkish villages  (Karain,  Tuzkb'y, Sarihidir)  where  mesothelioma was known to
occur.  In addition,  lung  tissue specimens of patients diagnosed with mesothe-
lioma who  lived in these villages were also analyzed.  Detailed examination of
environmental  samples  snowed that they contained several different forms  of
fibrous minerals, Including  asbestos minerals and fibrous erionite.  In  one
sample  collected  near Karain,  up to 30 percent,  by weight, consisted  of
tremolite and chrysotile.  Analysis of pleura!  and parenchyma! tissues revealed
that the  lungs  contained both asbestos and eripnite fibers.  Approximately 90
percent of the  fibrous particles in the lung tissue had a  composition consis-
tent with that  of erionite.   A  small amount  (approximately 1 to 5 percent by
volume) of fibrous mineral was  identified  as tremolite with a length typically
less than 10 urn.   Chrysotile,  both in fiber bundles and unit fiber form,  was
also found in  small  amounts.  In addition, tremolite fibers were observed in
dust samples  from Tuzkby.    In  lung tissue specimens obtained  from  pleura!
disease patients  from  this  village,  zeolite fibers  predominated but tremolite
fibers were also  found.
     Sebastien  et al.  {1981)  analyzed lung tissue samples from two malignant
pleura! mesothelioma  patients from  Tuzkby  and  found ferruginous bodies  and
uncoated  fibers.   The  numerical  concentration of the ferruginous bodies  per
gram of dried  lung tissue  was in the range of 10 .   Characterization  of  these
bodies  revealed that  they contained a central  core of zeolite.  Analysis
                   p
indicated about 10  uncoated fibers  per gram of dried lung tissue, and  93
percent were identified  as  zeolite fibers  with a mean  length of about 3.7 urn.
Only 3  percent  of the fibers were longer  than  8 urn or thinner  than 0.25 urn.
Two  other types  of  fibers,  found less  frequently,  were also characterized.
These were identified  as titanium oxide and an  amphibolelike fiber  having a
chemical composition similar to that of tremolite and actinolite.
     Boman et al.  (1982) analyzed lung tissue samples from two mesothelioma
patients  who  had  emigrated  from Karain to  Sweden.   Transmission electron
microscopic and microcheirrical analysis revealed that 64 and 75 percent,  respec-
tively, of  the fibrous  minerals  present in  the tissue samples  were zeolite
fibers.   In addition,  several asbestiform  minerals  including chrysotile  (9 and
5 percent) and  tremolite (27 and 9 percent) were also  identified.  The authors
suggested a synergistic effect involving both zeolite and asbestos.

.lulv 1988                          8-116             DRAFT—DO  NOT QUOTE  OR CITE

-------
     In the  United  States,  fibrous  erionite noted in samples of road dust from
•n area  in Nevada prompted  Rom et al.  (1981) to evaluate chest x-rays obtained
from a  local  community hospital.   All chest x-rays  (275)  of  patients over the
•ge of  25 in 1968 and 1975  and over the  age of 50 during the first 8 months of
1980 were examined.   Seventeen (6.2 percent) of the  radiographs  showed signs of
pleura! thickening  that  Included  blunting of one  or both  costophrenic angles,
and  five  (1.8 percent) had pleural  plaques.   No  pleural  tumors were noted.
Although  the  prevalence  of  the pleural changes was age related,  the authors
found no  unusual occurrence of pleural chest disease.
     Casey  et al.  (1985) described a  case report of  a 52-year-old man with
pleural thickening  and pulmonary  fibrosis.   The patient,  who lived and  worked
in an  area of  Nevada rich in zeolite deposits,  had no known occupational
exposure  to  asbestos.  A  lung biopsy revealed both fibrous  and  nonfibrous
particles.  The particles were studied using energy-dispersive x-ray analysis;
the analytic  pattern  of the  particles was  identical  to that produced  by  a
sample  of fibrous erionite.   No  asbestos fibers  could be identified.   The
authors suggested that exposure to erionite may be associated with the  patient's
pulmonary disease.
8.3.2.1.2   Immunological effects.   Ozesmi et al.  (1986)  conducted a cross-
sectional  morbidity  study of  74  individuals who  emigrated from Karain, the
Turkish village with the high incidence of MPM, to Sweden.  Because the  minere-
logical characteristics and the adverse effects of fibrous zeolite (irionite)
are similar  to  those of  asbestos, and because  immunological  aberrations have
been found  in those with asbestosis,  the investigators searched  for possible
disturbances  in the  immune  systems  of the former  residents of Karain.   From  a
list of participants in  an earlier study identified by Boman et  al.  (1982),
former  residents  of  Karain currently living in  Stockholm were invited to
participate.  The 74 study  subjects underwent clinical examinations and chest
radiography.  Blood  samples and personal and family medical histories were  also
taken.   Healthy blood donors were used as controls.
     Of the 74 study group members (36 males and 38 females), seven had pleural
plaques upon  x-ray  examination.   Compared to controls, the mean percentages of
Leu4* cells  and  Leu 3a* cells were  significantly decreased  (p <0.01, respec-
tively), whereas  the mean percentage of  Leu 2a   cells was normal.   This lead
to a significant  reduction  (p <0.05) of  the Leu  3a:Leu 2a ratio  in the study
group.   In  addition, the mean percentage of  B  cells (Leu  12* cells) was signi-
ficantly  decreased  (p <0.05)   in  the  study  group.   No correlation  was  found
* '•• -«"                          8-117             DRAFT—DO NOT QUOTE OR CITE

-------
between percentages of different lymphoid cells and age,  sex,  family  history of
mesothelioma, or duration of residence in Karain.
     The investigators concluded  that these findings indicate that exposure to
fibrous zeolite causes a numerical imbalance  between two  phenotypically  differ-
ent T cell subsets similar to that found in Individuals exposed  to  asbestos.
6.3.2.2  Epidemiological Studies.  No analytical epidemiological studies were
available for erionite.

8.3.3  Talc
8.3.3.1     Cross-Sectional Studies.  Clinical  Evaluations,  and Case Reports.
Increased respiratory  symptoms, higher  prevalence  of  pleural thickening or
calcification and  pneumoconiosis,  and  decreased pulmonary  function have been
reported in workers  exposed to talc  containing various  amounts of tremolite,
anthophyllite, or  silica  fibers.   The extent of these effects  increased with
age of workers,  intensity of smoking, and  duration  of exposure.  These  effects
were less pronounced with exposure to talc free of  asbestiform fibers.
8.3.3.1.1  Respiratory effects.  The  results  of a  study  of 93 New  York  miners
and millers  exposed to  talc containing tremolite  and anthophyllite fibers
indicated increased  respiratory symptoms,  x-ray abnormalities,  and  decreased
pulmonary function when compared  to  a population  of potash  miners  (Gamble
et a!., 1979).   Among  the talc workers, higher respiratory symptom rates  were
found for those  who  had worked for less than 15 years  and  those with 15 years
or more of  employment.   However,  most of the  workers with  symptoms were eithe-
smokers or  ex-smokers.   Mean pulmonary  function values (FEV., FVC, and  maximum
respiratory  flow at 50  and 75 percent) of talc workers were significantly
decreased in  comparison  to  those of the  potash workers.   These  decreases
remained after  controlling for age,  height,   smoking, and  number of. years  of
employment.   An  association was found between decreased  pulmonary  function and
fiber and particulate  exposure and the  number of years worked.  The  prevalence
of pleural  calcification  and pneumoconiosis  1n talc workers  with  15 or more
years of employment  was  higher than  1n  potash miners.  The greatest  difference
between the  two groups was the Increased  prevalence of  pleural thickening in
talc workers.   Among  the  talc workers, pleural thickening  occurred in 1.6
percent of  workers with less than 15 years of employment and in 31 percent of
those  with  more than  15  years employment.  Corresponding values  for potash
miners were 0.5 percent and 4.0 percent, respectively.

July 1988                          8-118             DRAFT—DO NOT QUOTE OR CITE

-------
     A  cross-sectional  study of 299  workers  from Montana,  Texas,  and North
Carolina,  who were exposed  to talc  containing  low levels of  silica and
tremolite  and  antigorite  fibers, was  conducted by Gamble et al. (1982).  The
mean age  of the study population was  40, and the nean  duration  of  exposure  to
talc dust  was  less than 10 years.   The average concentration  of airborne  dust
ranged  from 0.28  to 2.64 »g/m .   Respiratory symptoms did  not  show any
consistent association with years worked or cumulative  exposure.  There  were no
significant  Increases 1n  pneumoconiosis  aimong the study group or significant
reductions  in  lung function.   An  increased prevalence in bilateral pleura!
thickening was  found  in some exposed workers.  The prevalence was  related  to
age and  smoking habits  and was  greater (23  percent) in  those employees working
10 years  or  more when compared to  those  working  less than 10 years (2.5  per-
cent).   Moreover,  workers  with  bilateral  pleura!  thickening had higher  average
and cumulative  exposures  than those with no pleura! thickening.  All pulmonary
function values  of these  affected  workers were reduced when compared  to unaf-
fected workers.
     Evidence of fibrosis in workers exposed to talc containing  minimal  amounts
of crystalline  silica and asbestiform minerals  was  reported in a  pulmonary
pathology  study by Vallyathan  and  Craighead  (1981).   Pulmonary tissue from
seven deceased  Vermont  talc  workers  exhibited various degrees  of  pulmonary
fibrosis.  The  severity  of the fibrosis increased with duration of employment
and concentration  of  heavy mineral  deposits in fibrotic  lesions of the lung.
X-ray crystallographic studies of lung digestates revealed that  the predominant
mineral  in the lungs was talc.
     A  study of 202  talc  (purity  not specified) workers in  India  showed a
significant  increase  (p <0.001)  in  respiratory symptoms among the workers when
compared to  101 matched control subjects of the same socioeconomic status but
not engaged  in any dusty  trade  (Bachani  and Agarwal,  198S; Bachani, 1984).
Exposure  levels were not  reported.  Signs and  symptoms  Included pain and
burning  of the nose  and  throat,  increased phlegm production, difficulty in
breathing, and  chest  pain.   Clinical  examinations revealed a higher number of
workers  with nasal and  pharyngeal   iiucosal  congestion  and scattered rhonchi.
Over 13  percent of the workers  »nd 9 percent of the  controls  had clinical
evidence  of  chronic  airflow  limitation (CAL).   The prevalence of CAL was
related to  age  and smoking history.  A positive correlation was found between
the prevalence  of  CAL and duration  of exposure to talc dust.  In addition,  the

July 1988                           8-119             DRAFT—DO NOT QUOTE OR CITE

-------
presence of  eosinophils  in sputum samples of 13 of 38 workers  and none  in  the
control  subjects  suggested that  talc  exposure may  induce  an  allergic-type
pathology.
     Changes in ventilatory functions in response to industrial  grade  talc  dust
exposure (no levels  given) were investigated in 202 workers (Oamodar et al.,
1983).   Smoking habits were graded by duration In years  times  frequency per
day.   FVC,  FEV^,  and FEV. percent were measured.  All indices decreased with
increasing age, intensity of smoking, and duration of exposure.  A significant
decrease in  pulmonary  function was found for workers who smoked  compared  to
control  subjects  who smoked.   No significant difference was  found between
workers  and  control  subjects  who did  not smoke.  The authors postulated  a
complementary effect of smoking on talc-induced changes in lung function.
     A cross-sectional study  of the  respiratory function  of 116 Vermont miners
and millers  of talc ore, free from asbestos and silica,  was  performed  by Wegman
et al. (1982).  Exposure  levels were found to be below 3.0 mg/m   respirable
dust with a  geometric mean exposure of 1.8 mg/m  respirable  dust.   In  addition,
a 1-year followup  evaluation of pulmonary function was performed  on  103 work-
ers.   A  significant  reduction in pulmonary function was  found in workers  who
smoked.  After adjusting  for smoking,  the effect of exposure  to  talc was  not
statistically significant,  although  there was evidence of an  exposure-related
effect in  workers with  an annual dust  exposure of 1.5 mg/m .  Effects on
pulmonary  function  in nonsmokers  was  not associated  with  exposure.   Talc
exposure was also  associated with small,  rounded, and  irregular opacities  seen
on chest x-rays.   An annual loss in  pulmonary function occurred in the workers
at  a  rate  greater than expected.  However, the  loss did  not correlate with
exposure to  talc dust.
     A microscopic analysis of the lung and  lymph  node dust  samples  from  12
deceased Austrian  workers  occupationally  exposed  to talc  dust  revealed   a
slightly  increased  concentration of real  fibers and elongated particles  in
comparison to normal  individuals (Friedrichs, 1987).  The ratio  of length to
diameter was inconsistent throughout the diameter range  examined and was  not
comparable to that  found in asbestos workers with mesothelioma.   It was con-
cluded that  the carcinogenic potential  of the fibers was not  very high since
neither of the two factors postulated to be important in assessing the carcino-
genic  hazard of a fiber, i.e., increased numbers of very fine fibers  <0.01 urn
in diameter  and high numbers of fibers  in the Stanton range,  appeared to  be
related to the disease state.

-------
     Talc  is  considered  to be  hazardous when inhaled in large quantities.   The
hazard  is  acute in cases  of accidental  Inhalation and chronic in cases of
industrial exposure or prolonged  use of commercial products.   However, since
talc is  not  always a uniform substance, the association between talc exposure
and adverse  health effects nay be related  in some  instances to the presence of
contaminating ninerals such as asbestos and silica.   Most reports describing
acute and  chronic  effects  of talc exposure Include little characterization of
the talc involved.
     Infants  and  small  children  have died  from  cardiorespiratory arrest,
pulmonary  edema,  and pneumonia within  hours after  inhaling  talcum powder
(Cotton  and  Davidson,  1985; Motomatsu et el.,  1979;  Gould and Barnardo, 1972;
Anonymous, 1969).  The powder  dries the mucous membranes  of  the bronchioles,
thus disrupting  pulmonary  clearance and clogging  the small airways.  Victims
usually display dyspnea,  tachypnea, tachycardia, cyanosis, and fever.   Broncho-
pulmonary lavage is not effective, since talc is  insoluble in  water.   Gould and
Barnardo (1972)  described  the  case of a 7-year-old girl who survived a massive
aspiration of  powdered talc and  later  developed  bronchiectasis  of the left
lung.
     Lund and Feldt-Rasmussen (1969) described  the case of a child who survived
a massive aspiration of talc but only after exhibiting cyanosis,  tachypnea  with
intercostal retractions, and tachycardia.   There was fine crepitation over the
lungs,  and x-rays  revealed signs  of bronchitis and peribronchitis with atelec-
tases.   Clinical treatment provided a gradual regression back  to  normal.
     In  addition to  the  effects of a massive acute  inhalation, cases of talc
pneumoconiosis or  chronic  sarcoidosis characterized  by foreign body granulomas
and interstitial  fibrosis  have been  reported as a result of  chronic use of
cosmetic talcum powder (Tukiainen et el.,  1984;  Wells et al.,  1979).   Signs and
symptoms  included  dyspnea,  productive  cough,  and weakness.   Abnormal chest
radiographs and decreased pulmonary function were  also observed.
8.3.3.1.2   Effects from  dermal application.   An  analogy   between  asbestos-
produced mesotheliomas and talc-produced  ovarian  cancers  was  originally based
upon the  close  relationship between talc and asbestos.   Additional supporting
evidence for  this  analogy  is derived from  the similarity  1n origin and histol-
ogy between mesotheliomas and ovarian cancers (Longo and Young, 1979).
     In  a  case-control study  of  215 white  females  with  epithelial  ovarian
cancers  and  215  controls from the  general  population,  92/215 (42.8 percent)
cancer cases  regularly  used talc  either as a dusting powder on the perineum  or
July 1988                          8-121             DRAFT—DO NOT QUOTE OR CITE

-------
on sanitary  napkins  compared with 61/215 (28.4 percent) of controls  (Cramer
«t al., 1982).  The  relative risk adjusted  for parity and menopausal state was
1.92 (p <0.003).  A  total  of 32/215 (14.9  percent) of cancer cases regularly
used talc  both  on the perineum and on  sanitary  napkins  compared with 13/215
(6.0 percent) of  the controls.   Women with both exposures therefore  had  the
greatest adjusted relative risk of 3.28 (p <0.001).
     An extraction-replication  technique (Griffiths «t al.,  1973; Henderson,
1972;  Henderson etal., 1973;  Henderson and Griffiths, 1975)  was used  to
examine tissue  from  patients with ovarian tumors.   No  asbestos  particles  were
found  in any of the tissues studied (Henderson  «t  al.,  1971).   Particles of
talc were identified in approximately 75 percent (10/13)  of the  ovarian tumors.
Talc particles  were  found  localized deep within  the tumor tissue;  some were as
small  as 1000 angstroms in size, but they  were  generally within a range  from
1000 angstroms to 2 pm.
     In a  letter  to  the editor, Hartge  et  al.  (1983) refuted the  association
between talc use and  ovarian cancer.    The data collected were  part  of  a
case-control  interview study of  epithelial ovarian cancer conducted on  197
women with pathologically confirmed primary epithelial  ovarian cancer from 19J
to 1977.   The control  population consisted of 197  women treated at the same
hospital for conditions other  than gynecologic, psychiatric,  or malignant
diseases or  pregnancy.   Questions about talc were  added to the questionnaire
after  the  study began so that 135 cases and 171  controls were questioned  about
talc exposure.  No  increase was found  in the  relative risk to  women  who  used
talc as a  body  powder or on diaphragms.  A reliable estimate  of relative  risk
in women  who used talc  specifically  on sanitary napkins,  underwear, or  the
genital area was  not possible, however, because of the small  numbers of women
involved.
     Granulomas of the skin postulated to be caused by entrance of talc through
the  skin at sites of draining or incised furuncles  were  reported by  Tye et  al.
(1966).
8.3.3.1.3   Effects  from intrapleural  Instillation.  Adverse  systemic effects,
including  dyspnea  followed  by acute respiratory failure and bilateral pleura!
effusion with interstitial  infiltrates, was reported 1n a patient a few hours
after  talc pleurodesis  (Bouchama  et  al., 1984).  The patient  improved with
corticosteroid  therapy.
     A  total of  210  cases of  patients who had undergone pleurodesis  with
iodized talc or kaolin 14 to 40 years previously were traced, and the number's
July 1988                          8-122             DRAFT—DO NOT QUOTE OR CITE

-------
of deaths  from  all  causes  and  cancer  of the  lung and pleura were compared to a
control death rate  derived from  age-  and year-specific tables from England and
Wales  (Research Committee  of the British  Thoracic  Association and  the Medical
Research Council  Pneumoconiosis  Unit, 1979).  No  increase  was  found in the
incidence  of  lung  cancer in these patients,  and  no cases  of mesothelioma were
reported.
8.3.3.1.4   Effects from surgical  "introduction.   Inadvertent Introduction  of
talc  into  the body has been shown  to produce granulomas.   A granuloma that
completely obstructed  the  ureter of a 53-year-old  woman was attributed  to  the
introduction of talc  as a glove powder 20 years previously during a hyster-
ectomy operation (Joannides, 1978).   However, the possibility  that  the causitive
agent may also have been either starch granules or  Lycopodium  spores was noted.
Sahi et al.  (1967)  reported that granuloma formation on the bowel  and perito-
neum was  believed to  be the result of talcum powder use during a previous
surgical intervention.   Fienberg  (1937) reported two cases  of granuloma forma-
tion caused by talc introduced during surgical procedures.
8.3.3.1.5   Effects from intravenous administration.   Intravenous   abuse  of
stimulant  drugs intended for  oral administration can lead to the embolization
of insoluble  tablet fillers such as  talc to the pulmonary  vasculature.  The
vasculature acts  as a filter in most cases,  thereby restricting talc-induced
pathologic  lesions  to  the  lungs.   Blockage  of  pulmonary perfusion by the
passage  of intravenously  injected  talc  particles  can result in  pulmonary
hypertension with the  formation of arterio-venous. shunts.   The venous blood can
then  bypass  the  lungs and enter the  arterial system, lodging  particles  in
various organs  such as the eye.   The  resulting peripheral retinal  neovascular-
ization has  been  noted after injection of talc  emboli  (Lederer and Sabates,
1982).  Particles  smaller  than 7 own can travel to  the retina  via the pulmonary
capillary  bed.  Bilateral  intraretinal talc emboli  were reported in 23 patients
by  Tse and Ober (1980).   A case of talc  retinopathy that  included vitreous
hemorrhage and  retinal detachment was reported by Bluth  and  Hanscom (1981).
The effects of talc particle injection resulted 1n blindness In this case.
     Granulomatous  lesions found in  lungs,  liver, spleen, bone marrow, and
systemic lymph nodes  as  a  result of systemic talc embolization were  reported by
Mariani-Costantini  et al.  (1982).   Larger talc crystals  (10 to  17  urn) remained
localized  in  the  lungs; small  crystals (2  to 10 urn) spread systemically.   No
crystals were  found in numerous  sections  of the central  nervous  system, and

July 1988                           8-123             DRAFT-DO NOT QUOTE OR CITE

-------
 only a  few  were  present in the myocardium, retina, optic nerve, and kidneys.
 The talc particles were present in urine,  demonstrating  that they may have been
 able to pass through  the  glomerular capillary bed.   Self-induced pulmonary
 granulomatosis was described  by Waller  et al.  (1980).   Cases of intravenous
 talc granulomatosis  (Tao et al.,  1984),  liver  talc granulomas (Molos et al.,
 1987), and systemic talc granulomatosis  (Lewis  et al., 1985) have been reported
 in association with intravenous drug abuse.
 8.3.3.2  Epidemiological Studies.    Several  historical  prospective mortality
 studies have been  conducted to identify potential  relationships between  expo-
 sure to talc and  nonmalignant respiratory disease (NMRD).   An increase  in
 respiratory cancer and  NMRD was observed  in several studies.   However, workers
 were simultaneously  exposed to talc  containing asbetos, silicates, carbonates,
 or radon.   Results from one  study indicate that  exposure  to pure talc was net
 associated with increased NMRD or  respiratory cancer.
      Kleinfeld et al.  (1967)  conducted  a  proportionate  mortality  study  of 220
 talc miners  and  millers employed  between 1940 and 1965 who  had achieved at
 least 15 years of  exposure to talc  dust.   In addition to talc, exposure also
 included tremolite and  anthophyllite (asbestiform  and nonasbestiform  fibers),
 carbonate dusts,  and a small  amount of  free  silica.   Data for this  cohort
 concerning employment and health histories as well  as deaths were  obtained from
 a number of  sources,  including plant records,  death certificates  (when  avail-
 able),   private physician records,  state  health  department files,  hospital
 records, and  autopsy records  (procured  for 35/91 deaths).   The  degree of
 completeness for  disease ascertainment of the study population and  the criteria
 for classifications  of  morbidity  and mortality were not specified.  The refer-
, ence population  used  for calculating expected  mortality was  U.S.  white males
 for the year 1957, which was  the  median year of  death for the 91  total  deaths
 occurring out of 220 workers.  An analysis  of the 91 deaths by cause yielded a
 significant increase  in the proportional  mortality of the  total  observed deaths
 from cancer of the lung and pleura  (11 observed, 3.2 expected, Proportionate
 Mortality Ratio (PMR) = 344,  p <0.01, Power « 21  percent).  When the  cohort was
 analyzed by  age  group (<40,  40 to  59,  60 to  79, 80+ years),  a significant
 increase in cancer of the  lung and  pleura  occurred in  the  subgroup  60  to 79
 years old (17 observed,  3.9 expected, PMR « 436, p <0.01, Power = 24 percent).
 The authors  expressed concern that this finding  was  at  variance with the
 results from asbestos  workers  who had a similar duration  of  exposure  during a
 similar period.   Increased respiratory  cancer  incidence was  found  in  both the
                                    8-124            DRAFT—DO NOT  QUOTE  OR CITE

-------
40 to  59  and 60 to 79 age  groups  of  asbestos  workers.  A possibility exists
that the  discrepancy  -in  these findings could be  due  to  a greater carcinogenic
potential  and/or  greater exposure  levels  in the case of asbestos workers.
However,  since  the  power to detect a carcinogenic risk in the  40  to 59 age
group  of  talc workers was  inadequate  (21  percent),  no  definite conclusions
could  be  drawn.  An  analysis  of  Individuals with carcinoma  of the  lung or
pleura, by duration of txposure,  found that all these  Individuals were Ini-
tially  exposed  prior  to the institution of  wet  drilling (a  period of heavy
exposure  before 1945) with  an average duration  of exposure of 14.6 years.
Environmental dust exposure concentrations  before 1945 ranged from  120 to 818 x
106 particles/ft3 for wines and 69 to 1227 x 106 particles/ft3  for mills,  and
between 1946  and  1965 ranged from 5 to 9 x 10   particles/ft   for mines  and 25
to 73  x 10  particles/ft3  for  mills.   No  direct relationship was observed,
however,  between  the  duration  of  exposure  before wet drilling  and the occur-
rence of pulmonary carcinoma.
     Of the  91  total  deaths, 28 were  attributed  to pneumoconiosis  or complica-
tions  or  both.   Nineteen  of the 28  were  due  to cor pulmonale, which  was
reported  as  a major cause  of death  among talc  workers.  Because of the absence
of adequate  data, the role of smoking in the causation  of  pulmonary  carcinoma
could not be ascertained in this study.
     Findings from  an update  of  this previous  study through 1969  further
supported the increased  incidence  of  cancer  of the lungs and pleura  (Kleinfeld
et al., 1974).  The criteria for  inclusion in  the cohort remained  the same for
the  followup study.  The  total number of  workers studied was  260,  and the
analysis  used was a comparison of observed  proportional mortality rates from
specific  causes with  expected rates.   Expected mortality for the  total group
represented  the proportion of deaths  among white men in the U.S.  due to spe-
cific  causes  in 1955  (median year of  death  among the total  108 deaths).   For
each of  the  5-year  periods, median years 1942, 1947,  1952, 1957, and 1962  were
selected.  Data for the  cohort were obtained from the same sources as those in
the previous study.   The total number of deaths out of 260  workers was 108, and
•n analysis  of these deaths by cause yielded results  consistent with the
previous  study; there was a  significant  increase in proportional mortality from
respiratory  cancer  in the 60  to 79 age group  (16.6 observed,  3.6 expected,
PMR =461, p <0.05).
     Insufficient power  to detect  an  increase in other age groups was still a
confounding problem in this  analysis.   An analyses of carcinoma of the lung and
July 1988                          8-125             DRAFT—DO NOT QUOTE OR CITE

-------
pleura as  related  to six 5-year periods (1940-44, 1945-49, 1950-59, 1960-64,
and  1965-69) revealed  significant increases in  the  observed proportional
mortality  for respiratory  cancer for  three periods:  1945-49  (observed * 11.7,
expected = 3.1, PMR  * 377,  p <0.05,  Power  = 18.9 percent);  1950-54 (12.5
observed,  4.3 expected,  PMR « 291, p <0.01, Power * 20 percent); and 1955-59
(observed  =  20.0,  expected « 5.7, PMR * 351, p <0.05,  Power  * 24 percent).
Investigators found  no  differences in mortality  for  respiratory  cancer  occur-
ring after 1960 and no correlation between respiratory cancer  deaths  and age or
years of exposure.   Ten of the 13 deaths from respiratory tract cancer occurred
in the period between 15 and 24 years  of exposure to talc dust.   However,  a
prolonged exposure of >25 years did not produce any additional risk.
     Although they remained major  causes  of death, deaths attributable  to
pneumoconisis and  its  complications decreased by more than 50 percent  in the
last 5-year period studied (decrease from nine to four deaths).
     The only historical prospective  mortality  study conducted with  very pure
talc was reported  by Rubino et al. (1976).   The  cohort consisted of  1514 talc
miners and 478 talc millers, who began work in the years between  1921 and 1950,
and  who  had been  employed for  at least  1  year  in talc  operations  of  the
Germanasca and Chisone Valley.  Because millers were  exposed to very  pure talc,
whereas miners'  exposures  included a  certain amount of  inhalable  silica parti-
cles, miners and millers were analyzed separately.  Data for  each worker were
obtained  from  payrolls  and  records  of a  mining company of Pinerolo.
Age-matched control  subjects  were  chosen  from the population  of  the town of
Alba,  which represented similar ethnic,  social,  and economic conditions.
Follow-up  for the  study cohort was 88.9 percent  complete for miners and 91.6
percent  complete  for millers.   The cause  of death was obtained  from death
certificates for exposed and control persons and  was  coded according  to  the  8th
Revision of the ICO.  For both internal and external  controls, an age- adjusted
modified life-table  technique  was  used to compute expected deaths.  The  age
intervals  used were  <30, 30 to 50, 51 to 70, and >70, and the latency periods
were 1 to 10, 11 to 20, 21 to 30, 31 to 40,  41 to 50, and >50.
     Cumulative exposure (millions particles per  ft /years, mppcf) was computed
by multiplying  dust  level  for a period by number of  years exposed and summing
the  resulting values  for each period.   Three exposure  levels  were then  classi-
fied  for  miners (exposure  1 s 566 to  1,699; exposure  2 = 1,700 to 5,665;
exposure 3 = 5,666 to 12,750) and  millers  (exposure  1 - 25 to 141;  exposure
2 = 142 to 424;  exposure 3 = 425 to 906).   Analysis of mortality  data indicated
     1988                          8-126             DRAFT—DO NOT QUOTE OR CITE

-------
a  significantly  lower mortality  for  talc miners and millers than expected
(miners « 704 observed, 791.2 expected,  SMR * 89,  p <0.01, Power = 100 percent;
millers = 227 observed, 258.4 expected,  SMR = 88,  p <0.01, Power = 100 percent).
When the overall  Mortality  was  subdivided by cause of deaths, deaths from all
NMRD 1n miners were significantly higher  than in controls (140 observed, 101.8
expected, SMR = 136, p <0.01, Power = 99.8 percent).  Among malignant diseases,
deaths from  lung  cancer were significantly  lower than 1n controls (9 observed,
19.7 expected, SMR  * 46,  p  <0.01, Power * 63.7 percent).  An analysis of cause
of death by  exposure levels in  miners showed a trend of Increasing deaths with
increasing exposure  from  NMRO (Exposure  Level  1  * 26 observed, 36.9 expected,
SMR =70,  p  <0.05,  Power  = 86  percent; Exposure  Level  2 = 38 observed,  36.3
expected,  SMR  s  105, not significant, Power * 85.8  percent;  Exposure Level
3 = 76  observed,  67.1 expected, SMR * 113,  p <0.05, Power « 97.9 percent).  No
clear trends, by  exposure level,  were observed for deaths from any other cause
in miners or millers.  When mortality by cause of death  was analyzed by  latency,
the only  significant increase  found  was  for NMRD after 20 to 40  years of
latency (76  observed,  54.4  expected,  SMR  * 140,  p <0.01, Power « 95 percent);
and after >40 years of latency (36 observed, 23.9 expected, SMR = 151, p  <0.01,
Power = 151).  No significant increase in observed deaths by latency was found
for millers.
     The choice of  the control  population was discussed by the authors  as  a
confounding  factor  that could contribute  to the  observed differences.   It was
not possible to verify the  appropriateness of this reference group.   In  addi-
tion,  smoking data were unavailable; therefore,  it was assumed that the  control
and exposed  groups  were  similar in this  respect.  An internal comparison of
groups of  workers  by increasing dose and  latency period produced a pattern  of
excess mortality  for respiratory causes  of death in  miners.  This trend, plus
the different incidence of silicosis 1n miners (62 observed, 30.9 expected,  SMR
=  201,  p <0.01,  Power =  80 percent),  led the authors  to conclude that  the
increased risk of death from NMRO was due  to silica rather than talc.  The lack
of increased risk  from any cause of death 1n talc-exposed millers led  the
authors to conclude that  no evidence  from the study supported the role of pure
talc in carcinogenesis.   It should  be noted, however, that the power to detect
lung cancer  in the study  population of millers was inadequate.
     Brown et al.  (1979)  conducted a historical  prospective mortality  study
involving 398 white males initially employed some time  between  January 1,  1947,
and December 31,  1959, at  a talc  mine   and mill in upper  New York State
July 1988                          8-127             DRAFT—DO NOT  QUOTE OR  CITE

-------
(Gouverneur  region).   Vital  status  for cohort members as of June 30, 1975, was
determined through government records -Including the SSA,  State Vital  Statistics
offices, and State Motor Vehicle Registration.   Death certificates  were obtained
and causes of death were interpreted by a qualified nosologist and  converted to
the  7th Revision of  the  ICD.   A wodified life-table technique was  used  to
obtain person-years at risk of dying by 5-year time periods,  5-year age groups,
duration  of  employment, and number of years since Initial employment.  The
reference population  used  was  U.S.  white Bales.  Vital status was determined
for 96 percent of the exposed cohort.
     No significant difference  in  overall  mortality was  observed.   However, a
significantly higher  mortality  from all malignant neoplasms was detected  (19
observed, 10.6 expected,  SMR  = 180, p <0.01, Power = 42  percent).   This  was
ascribed, in part,  to the significant  increase in deaths  due  to bronchogenic
cancer  (9 observed, 3.3 expected,  SMR  = 270, p  <0.05, Power =  13 percent).   In
addition, investigators  found an  increase  in  deaths from NMRD, excluding
influenza, pneumonia,  bronchitis,   and  acute upper respiratory infection  (5
observed, 1.3 expected,  SMR  = 380, p <0.05, Power = 13  percent).   An analysis
of the  association  between bronchogenic cancer  and  latency revealed  a  signifi-
cant  increase  in the  interval 20  to 28 years  since onset of  employment  (6
observed, 1.3 expected, SMR = 460,  p <0.01,  Power = 13 percent).
     The  study  of mortality  patterns  in workers  exposed  to talc from the
Gouvernour mining area  demonstrated an excess 1n mortality from both NMRD and
bronchogenic cancer.   However,  this study included several confounding vari-
ables.  Smoking  histories were  unavailable  for this  cohort.   Data on other
occupational exposures  were  also missing.  In addition,  an industrial  hygiene
analysis  showed  that  the amphibole  fiber exposure characteristics  between a
neighboring  talc  operation  known to have asbestiform talc and the  study mine
and mill  were  essentially the same.  Fiber  exposures, when measured in 1975,
were  found  to exceed  Occupational  Safety and  Health Administration (OSHA)
asbestos  standards.   The  authors postulated that the exposure to asbestiform
tremolite and anthophyllite may have been responsible for the results observed.
A confounding problem with the study cohort was the relatively short duration
of employment for many of the workers:  50  percent were employed  less than  1
year, and less  than 25 percent were employed >10  years.  Of  the  10  deaths  from
bronchogenic cancer,  4  were  found   1n workers employed less than 1 year.   Only
two of these four were known to have been previously exposed to talc containing
asbestiform minerals.
                                    •> •-*•             r\fM>rT. _nr\ uAT AllATC  HP

-------
        Selevan et al.  (1979)  conducted  a historical  prospective  mortality  study
   of 392 white male talc workers wt>o had been radiographed as  part of  the Vermont
   Health Department  annual  surveys.  The workers were employed by five companies
   in the Vermont  talc industry for at least 1 year between January 1, 1940, and
   December 31, 1969.   Follow-up was  begun  from the Initial  radiographic examina-
   tion, when 1 year of employment was achieved, or January 1,  1940,  whichever was
   later, and was continued through December 31, 1975.   Vital status ascertainment
   was performed through  government records  including Social Security  Administra-
   tion, State  vital   statistics  offices, and State motor vehicle registration
   files.  Vital status was  unknown for  1.0 percent  (4  of 392 workers) of the
   cohort.   Death  certificates  were obtained,  and causes of  death were classified
   by a  qualified  nosologist and converted to  the 7th  Revision  of the ICD.  A
   modified life-table technique was used to obtain person-years  of observation  by
   5-year time  periods,  5-year  age groups,  5-year  exposure groups, and 5-year
   latency groups.   The study  cohort  was divided Into two groups,  milling  (225
   workers) and mining (163  workers).   A minimum of  1 year  of employment in the
   work  area  was  required for  inclusion; 47 workers (12  percent  of the  total
   population) qualified  for inclusion  in both groups.   U.S. rates were used for
   calculation of  expected  values.   In  addition,. Vermont rates  were  used  for
   comparison in the  cases of NMRD and  respiratory cancer deaths,  since Vermont
   rates are higher for these causes.
        An analysis of exposure revealed that both airborne  dust  and bulk samples
   of talc from the mines and  mills were similar  in composition:   no asbestos was
   detected; free  silica  was below 0.25  percent for  nearly  all samples; and the
   minerals magnesite,  chlorite,  and  dolomite were present in  significant quanti-
   ties.  Although no cumulative exposures  could  be calculated,  it was not  uncom-
   mon  for  past  levels  to  exceed the   OSHA  and Mine  Safety and Health
   Administration  (MSHA)  standard of  20  nppcf.  Miners,  but  not millers, may also
   have been exposed to serpentine  rock and radon.
        The mortality  analysis  of the total cohort showed that the only signifi-
   cant excess occurred from NMRD (excluding influenza and pneumonia)  (11 observed,
   1.79  expected,  SMR =  615,  p <0.01,  Power « 14 percent), particularly among
|  millers (7 observed, 0.89 expected,  SMR  =787,  p  <0.01, Power  = 5 percent).   A
   significant increase in respiratory  cancer mortality was observed  only  among
   miners (5 observed, 1.15  expected, SMR =  435,  p <0.05, Power =  10  percent).
   These two significant  excesses were  maintained when  expected  numbers of  deaths

   July 1988                          8-129             DRAFT—DO NOT  QUOTE OR CITE

-------
from the  Vermont  population were used (NMRD, 1.33 expected,  p  <0.01;  respira-
tory cancer,  1.22 expected,  p <0.05).   To confirm that the excess of  NMRD
deaths in millers was from talc exposure, chest radiographs  for 8 of the 11 men
who  died  from this disease  were  examined  for pneumoconiosis.   Of the  eight
radiographs, six (75 percent) had indications of pneumoconiosis.   When 2 of the
11  workers  with prior  dust exposure were  included,  five of the six (83.3
percent)  radiographs were  read as having pneumoconiosis.   Qualitative  radio-
graph readings  by  the  Vermont Health Department provided  information  on three
deaths with  missing radiographs;  all  three had evidence  of pneumoconiosis,
which provided  a  total  of  9 of 11 deaths with readings consistent with  pneumo-
coniosis.   This supported the association between talc exposure and NMRD.
     The  authors  stated that the  possibility of  a selection bias was  small
because workers who were missed by the radiographic surveys  were for  the most
part short-term workers.   An examination of NMRD  mortality by  latency resulted
in a pattern  of significant excess after 15 years (2 observed, 0.31 expected,
SMR = 645,  p  <0.05,  Power = 8  percent), and after 30 years  (5 observed,  0.41
Expected,  SMR =  1220,  p <0.01, Power «  3  percent),  but not before 15 years.
This suggested  that no  selection bias  toward less healthy workers existed.  No
data were available on  the  smoking  histories of workers.   In addition,  the
confounding problem of  small cohort size resulted in a lack of adequate power
to detect increased risk in many of the disease categories.
     Stille and Tabershaw  (1982)  conducted a historical prospective mortality
study of  655  white male workers employed between  January 1,  1948, and December
31, 1977, at  one  talc  mine  and mill  (TMX)  in upper New York State.   Exposure
data were not  included.   Due to small  numbers and lack of talc exposure,  women
were excluded from analysis.  The cohort was followed until  December  12,  197S.
Data on the  workers were retrieved from employment records.   Vital  status was
unknown for 36  members  (5.5 percent) of the cohort.  Death  certificates  were
obtained  and  coded according  to the 8th  revision of the  JCD by a trained
nosologist.   Expected values were based on  the U.S.  white male  population by
means of a modified life-table method.
     The  mortality analysis of the cohort  revealed no significant increases in
deaths from all  causes, from cancer of  the respiratory tract  and lung, from
NMRD, or  from  other causes of death.   The  cohort was divided into workers with
a prior work history before TMX employment and workers with no known  prior work
experience.   Investigators  found  no significant difference in mortality rates
for  all  disease categories  in workers  without  prior employment.   However,
July 1988                          8-130             DRAFT--DO NOT QUOTE OR CITE

-------
workers with  prior work histories  did  show significant  increases  in deaths
attributable to  a  number of  causes  including respiratory cancer (9 observed, 4
expected, SMR *  228,  p <0.05,  Power « 15 percent) and NMRD  (9 observed, 2.9
expected, SMR = 307, p <0.01, Power « 15 percent).  The authors postulated that
the observed mortality  cancer was from exposures that occurred in the workers'
previous jobs.
     An analysis of the dose-response relationship of the lung cancer cases
observed  in  the entire cohort  and  cumulative  dose as Measured by years of
exposure  resulted  in an Inverse  dose-response relationship.  Investigators
found a clustering  of 9 out  of  12 cases within the first 5 years of employment
at TMX.   In addition, a calculation of the average  latency  time for  lung  cancer
among TMX  workers   was  19.9  years.   The authors cited the  lack  of  a dose-
response relationship and  the  short latency period as evidence that the  prior
occupational  exposures  and not  exposures at TMX may  have  been  the  cause of
subsequent cancers.
     Confounding variables  in  the  study  included  the small  cohort  size and
inclusion of  short-term (<1 year)  employees.  Although  the  U.S.  white male
population was used as  a reference  group, the  authors pointed out that the  New
York State  cancer   death rate  is substantially  higher  (199.24 compared to
174.04).   Lack of  data  concerning smoking histories and  incomplete work histo-
ries were also limitations  of this study.
     Thomas and  Stewart (1987) conducted a  historical prospective  mortality
study of 2055 white males employed for at least 1 year between  January 1,  1935,
and January 1,  1966,  at three plants of  a  ceramic plumbing  fixture company.
More than 60 percent of the cohort was employed for 10 or more  years.   Nonwhite
males and females  were  excluded from the cohort due  to  small numbers.  Vital
status was  determined  from  company,  Social Security Administration, credit
bureau,  and department  of  motor vehicle records.   As  of  January  1,  1981,  1394
(67.8 percent) of  the 2055 cohort members were alive; 578 (28.1 percent) had
died; and  83 (4.0  percent)  were  lost to followup.  Death certificates were
obtained for all but 1.7 percent of  the  decedents and  coded by a  qualified
nosologist using the  8th Revision of  the  ICO.  Expected  numbers of  deaths were
calculated using rates  for U.S. white males.  Because exposure data  for talc
dust were  unavailable,  exposure classifications were based  on  the  potential
exposure to silica  dust (none,  low,  high).  All jobs involving talc  exposure
had high  silica  exposure  and were further classified into no talc,  nonfibrous

July 1988                          8-131             DRAFT—DO NOT QUOTE  OR  CITE

-------
talc, and  fibrous  talc exposures;  no  talc was the lowest exposure, and fibrous
talc was the  highest exposure.   No information on smoking patterns was known
for this cohort.
     The mortality  analysis  showed a  significant  increase  in  deaths  from lung
cancer  (52  observed,  36.3 expected, SMR * 143,  p  <0.05,  Power « 85.5 percent)
and NMRD (64  observed, 37.0  expected, SMR s 173, p <0.05. Power * 86 percent).
The  increase  1n NMRD  deaths was mainly from  respiratory disease  other than
pneumonia or  emphysema,  which  Included  23 deaths  nosologically classified as
pneumoconiosis  from  silica and  silicates (41  observed,  14.1 expected, SMR =
290, p <0.05, Power = 46 percent).   An analysis  of lung cancer and NMRD by year
of  hire showed a  significant  increase   in  lung cancer  among  workers hired
between 1940  and  1949 (22 observed, 11.4 expected, SMR = 193, p <0.05, Power =
45 percent) and a significant increase  in NMRD  among workers  hired before 1940
(55 observed, 24.5 expected,  SMR = 224,  p <0.05, Power =  68 percent).  Analysis
of mortality  by exposure category  for lung cancer and NMRD revealed a signifi-
cant excess only  in workers  in  the high silica  exposure category  (44 observed,
24.3 expected,  SMR =  181, p <0.05, Power » 68  percent), particularly among
those exposed to nonfibrous talc  (21  observed,  8.3  expected,  SMR  = 254,
p <0.05, Power  =  37 percent).   In  addition, a significant  increase in NMRD was
found among those  exposed to high silica with  no talc exposure (36 observed,
13.7 expected,  SMR  = 264, p <0.05, Power = 49 percent) and to nonfibrous talc
(16 observed,  7.3 expected, SMR  = 220, p <0.05, Power  =  31  percent).   An
analysis by duration of exposure and years since first silica  exposure  revealed
an increase in  NMRD mortality  by duration of exposure (15 to 29  years:   21
observed, 12.1 expected, SMR = 173, p <0.05,  Power =  45 percent; and  30* years:
36 observed,  12.8  expected,  SMR = 281,  p <0.05, Power =  49 percent)  and years
since first exposure  (50 observed, 22.0 expected, SMR *  227,  p <0.05,  Power  =
66 percent).  Lung cancer mortality did  not increase  with duration of exposure,
but there  was  an  increase  in years  since  first  exposure (30* years:  32
observed, 20.6  expected, SMR = 156,  p  <0.05,  Power =  65.8 percent).   The
analysis by duration  of exposure  and years since  first exposure to nonfibrous
talc revealed an increased mortality from lung cancer with  duration of  exposure
(5 to 14 years:  11 observed, 4.0 expected,  SMR =  276, p  <0.05, Power = 15 per-
cent; and 15* years:   8 observed,  2.2 expected, SMR = 364, p  <0.05,  Power = 12
percent) and  with  years since  first exposure  (5 to  14 years:   8 observed, 2.9
expected, SMR = 281,  p <0.05, Power = 15 percent; and 15*  years:    13 observed,

July 1988                          8-132             DRAFT-DO NOT QUOTE OR CITE

-------
4.7 expected, SMR = 275, p <0.05, Power = 16 percent).  Mortality from NMRD was
not affected by duration or time since first exposure  to  nonfibrous talc.
     The excess NMRO was elevated among workers  with exposure  to high levels of
silica regardless  of  nonfibrous  talc  exposure.  This risk decreased over time,
suggesting a reduction to silica exposure.   Risk of  NMRD  was also unaffected by
duration or  time  since  first exposure to nonfibrous talc.  The authors con-
cluded an  association between  silica  dust rather than talc exposure with this
disease.    The  excess in lung cancer  found  in this  study, however, occurred
primarily  in workers  exposed to  both  nonfibrous talc and  silica.  Risk of lung
cancer occurred only ajnong those hired after 1940 and  increased with  increasing
duration of  exposure  to nonfibrous talc but not with  duration of silica expo-
sure.  This  suggested that exposure  to  nonfibrous  talc  was  related  to lung
cancer risk.  However,  the confounding problem  of simultaneous silica exposure
prevented any conclusions as to their respective roles in the  disease.

8.3.4  Venrnculite
8.3.4.1     Cross-Sectional Studies. Clinical Evaluations, and Case Reports.
Results of two studies indicated an association  between past  fiber  exposure  and
parenchyma!  and  pleura! radiographic  abnormalities, but  no correlation with
pulmonary function tests.
8.3.4.1.1  Respiratory effects.   In a review on nonasbestos fibrous materials,
Lockey (1981) concluded that vermiculiHe itself did not  produce adverse health
effects but  noted that  vermiculite ore may  contain  asbestos fibers.  In a more
recent study (Lockey et al., 1984), workers exposed to vermiculite  contaminated
with  fibrous  tremolite  were examined.  A total  of  530 employees  in  a plant
processing  vermiculite  ore to  its expanded  form were studied.   Industrial
hygiene sampling  of  airborne fibers was initiated in 1972,  15 years  after  the
plant had  begun  using vermiculite.  Particles with a length >5 urn, a diameter
<3 urn, and  aspect  ratio  of 3:1  or  greater were  counted as fibers.   A
high-exposure group was Identified and compared with  the chemical processing
facility where exposure was low.  The highest cumulative fiber exposure for an
employee was  39  fibers/cm -year; only 9.6 percent had an exposure  greater than
10 fibers/cm3-year,  and 10.7 percent  had been employed 20 years or more since
initial exposure.  A cluster of  12  cases of pleura! effusions was identified in
the  high-exposure group of 194  workers.  Radiographic changes correlated  with
exposure when age-matched  groups and groups with comparable smoking habits  were

Ju1y i988                           8-133             nRAFT-.nn NOT nnniF np ri-rr

-------
compared.   The  prevalence of pleuritic changes was significantly related  to
fiber exposure, but no correlation was found between fiber exposure  and  various
pulmonary function tests.
     Amandus at al.  (1987b)  conducted a cross-sectional morbidity study of 191
men employed for  at least 5 years between  1975 and 1982  at a vermiculite ore
mine and mill  near Libby, MT.   Radiographic examinations  had been administered
by a local  hospital to all active workers  in  1959  and annually since 1964.
Questionnaires on smoking habits and respiratory symptoms  had  been administered
by the  company  to most active workers employed after 1975.  Chest radiographs
were available  for 184 workers, and questionnaires  about  smoking and respira-
tory symptoms  were available  for 121 workers.  Radiographic findings  were
independently  interpreted by  three   readers  blinded to  other  data.  The
radiographic readings  indicated that the  prevalence of small  opacities  was 10
percent, any pleura!  change  was 15 percent, pleura!  calcification  was 4 per-
cent, and pleura!  thickening on the wall was  13 percent. Fiber exposure,  as
measured by fiber-years, was  significantly related to small  opacities, any
pleura! change, and pleura!  thickening on the wall  (p <0.05).   The  prevalence
of small opacities  was related to age and  fiber-years  but not  significantly
related to  smoking.   The confounding effect of smoking  could  not be accurately
assessed, however, due to the small number of nonsmokers (25 overall).
8.3.4.2  Epiderciologica! Studies.   Amandus  and Wheeler  (1987)  conducted  a
historical   prospective mortality study of  575 men  hired prior to  1970  and
employed at  least 1 year at the vermiculite ore mine and  mill near  Libby,  KT.
Vital status was  ascertained for 569 of the 575 cohort  members  (99  percent) as
of December  31,  1981.   Death certificates were obtained for all but two of  the
decedents (1.2  percent)  and  coded according to the 8th Revision of the 1CD.
Expected deaths were calculated from the U.S. white  male death rates.  Individ-
ual  cumulative  fiber  exposure  estimates  (fiber-years) for the  cohort  were
computed for 25 "location-operations" (LO) for the years after 1968  by using an
arithmetic  average  of fiber  concentrations (fibers/cm )  and for the years
before  1968 by using  an arithmetic  mean  of dust  concentrations  (mppcf).
Vermiculite ore was found to be contaminated with fibrous tremolite-actinolite
(Amandus et al., 1987a).
     The results  of the mortality analysis  indicated  a  significantly increased
risk of mortality for lung cancer (20  observed, 9.0 expected,  SMR  = 223.2,
p <0.01, Power  =  38 percent) and NMRD (20  observed,  8.2 expected, SMR = 243.0,
p <0.05, Power = 35 percent).  An analysis by exposure, measured as  fiber-years
July 1988                          8-134             DRAFT--DO NOT  QUOTE OR CITE

-------
 (f*y)i  showed a  significant  increase  in mortality in the highest exposure
 category  (>399 f-y) for both lung cancer  (10  observed,  1.7  expected,  SMR =
 575.5,  p  <0.01,  Power  = 12 percent) and NMRD (7 observed,  1.8 expected;  SMR =
 400.7,  p  <0.1, Power = 14 percent).  Mortality from NMRD  was  also  significantly
 elevated  in  the  lowest exposure category (<150  f-y) (8 observed, 3.6 expected,
 SMR  « 220.0, p <0.05,  Power *  18  percent).  Mortality from  lung  cancer and
 NMRD  was  further evaluated by  f-y and years of latency.  No  clear pattern
 emerged with respect to  latency for lung cancer, although  the significant
 increases occurred  in  the highest f-y exposure category (>399) at <10 years
 latency (2 observed, 0.2  expected,  SMR = 1370.2. p <0.05, Power * 4 percent)
 and £20 years  latency  (7  observed,  1.0 expected, SMR = 671.3,  p <0.01, Power =
 7 percent).   No  exposure-response  association  was  found  for  NMRD;  significant
 effects  occurred after  10 to  19 years  of  latency for 100 to 399  f-y
 (3 observed, 0.6  expected, SMR  = 459.9, p <0.05, Power «  6  percent) and >399
 f-y  (4  observed,  0.5 expected,  SMR * 774.5, p  <0.01, Power  =  4 percent)-  At
£20 years,  only   the  lowest exposure  category  (<50  f-y)  was significant
 (7 observed, 2.1  expected,  SMR  = 327.8, p <0.05, Power  =  10 percent).   There
 was  insufficient  power  to detect significant increases  in risk at other expo-
 sure  and  latency  periods.   In addition  to small cohort  size and  insufficient
 power to.detect  lung cancer and NMRD among the  whole cohort, another confound-
 ing  variable  in  this  study included the effect of smoking.  The proportion of
 current and  former smokers among 161  vermiculite  workers was found  to  be
 15.5 percent higher than  that  among U.S. white males in  1975.  Another source
 of bias may have  been the use of the national mortality rates rather  than state
 or county rates.
8.4  REFERENCES

Alsbirk,  A.   K.  E.; Johansson,  M.;  Petersen, R.  (1983) Ojensymptomer  og
     eksponering for mineralfibre fra lyddaempende loftsplader [Ocular symptoms
     and  exposure  to  mineral fibres In boards for  sound-insulation of ceil-
     ings]. Ugeskr. Laeg. 145: 43-47.
Amandus,  H.  E.;  Wheeler, R.  (1987) The morbidity and mortality of vermiculite
     miners  and  millers exposed to tremolite-actinolite: Part  II.  Mortality.
     Am.  J. Ind. Med. 11: 15-26.
July 1988                          8-135             DRAFT-DO NOT QUOTE OR CITE

-------
 Amandus, H. E.; Wheeler, R.; Jankovic, J.; Tucker,  J.  (1987a) The morbidity and
      mortality  of  vermiculite  miners and  millers exposed  to  tremolite-
      actinolite: part 1. exposure estimates.  Am.  J. Ind.  Med.  11: 1-14.

 Amandus, H.  E.; Althouse, R.; Morgan, V.  K.  C.; Sargent, E. N.;  Jones,  R.
      (1987b)  The  morbidity and  mortality of vermiculite miners and millers
      exposed  to tremolite-actinolite:  Part  III.  Radiographic findings.  Am.  J.
      Jnd.  Med. 11:  27-37.

 Andersen,   A.;  Langmark, F.  (1986)  Incidence of cancer  1n the  mineral-wool
      producing industry  in Norway.  In: Saracci,  R., ed. Contributions to the
      IARC   study on mortality and cancer incidence  among  man-made mineral fiber
      production workers. Scand.  J. Work Environ.  Health 12 (suppl.  1):  72-77.

 Anonymous.  (1969)  Accidental inhalation of talcum powder. Br.  Med.  J.  4:  5-6.

 Arbosti, G.;  Lo Martire,  N.;  Bonari, R.   (1980) Patologia dermatologies  e
      allergica in lavoratori di  una fabbrica del meridione per  la produzione
      di filati di fibre minerali artificial!' (fibre di vetro) [Dermatologies!
      and allergic pathology  in  workers of a glass  fibre factory]. Med.  Lav.
      1: 99-105.

 Artvinli,  M.; Baris,  Y.  I.  (1979) Malignant mesotheliomas in  a small  village in
      the Anatolian  region of Turkey:  an  epidemiologic study. JNCI J. Nail.
      Cancer Inst.  63:  17-22.
                                                        •
 Artvinli,  M. ; Baris,  Y.  I.  (1982) Environmental  fiber-induced pleuro-pulmona-y
      diseases in an Anatolian  village: an epidemiologic study.  Arch.  Environ.
      Health 37:  177-181.

 Bachani, D. (1984) Chronic airflow limitation in  talc  industry:  role  of  age,
      smoking  habits  and dust exposure.  Indian   J.  Chest Dis. Allied  Sci.
      26:  220-224.

 Bachani, D. ;  Agarwal, D.  (1985) Impact of talc  dust on the respiratory syster
      of working population. J.  Indian Med. Assoc. 63:  41-43.

.Baris,  Y.   I.; Sahin,  A. A.; Ozesmi,  M.;  Kerse,  I.; Ozen, E.;  Kolacan,  B.;
      Altinors, M.;  Goktepeli, A. (1978) An outbreak of pleura! mesothelioma and
      chronic  fibrosing  pleurisy  in  the village  of Karain/Uerguep in Anatolia.
      Thorax 33:  181-192.

 Baris,  Y.   I.; Artvrtnli, M.; Sahin, A. A.  (1979)  Environmental mesothelioma in
      Turkey. Ann.  N.  Y.  Acad. Sci. 330: 423-432.

 Baris,  Y.  I.; Sahin,  A.  A.; Erkan, M. L.  (1980)  Clinical and  radiological  study
      in sepiolite  workers.  Arch. Environ.  Health 35: 343-346.

 Baris,  Y.   I.; Saracci,  R.; Simonato, L.;  Skidmore, J.  W.; Artvinli, M. (1981)
      Malignant  mesothelioma and  radiological  chest  abnormalities  in two
      villages in Central Turkey:  an epidemiological and environmental  investi-
      gation. Lancet 1: 984-987.
                                                      ncar-r-.nn wivr nilDTE OR CITE

-------
Baris, I.;  Simonato,  L.;  Artvlnll, M.; Pooley, F.; Saracci, R.;  Skidmore,  J. ;
     Wagner, C. (1987) Epidemiological and environmental tvidence of the health
     effects  of exposure  to  erionite  fibres: a  four-year study  in  the
     Cappadocian region of Turkey. Int. J. Cancer 39:  10-17.

Bayliss, D.  L.;  Dement,  J. M.; Wagoner, J. K.; Blejer, H.  P.  (1976) Mortality
     patterns among fibrous  glass production workers. Ann. N. Y.  Acad.  Sci.
     271: 324-335.

Bertazzi, P.  A.;  Zocchetti.  C.; Pesatori, A.; Radice,  L.;  Riboldi,  L.  (1984)
     Mortalita per turnorl  In una coorte di produttori  di fibre vetrose  [Cancer
     mortality among workers In the glass fibre production Industry]. Med.  Lav.
     75:  339-358.

Bertazzi, P. A.; Zocchetti, C.; Riboldi, L.;  Pesatori, A.; Radice, L.; Latocca,
     R.  (1986)  Cancer  mortality of an  Italian cohort  of  workers in man-made
     glass-fiber production.  In:  Saracci, R., ed. Contributions  to the IARC
     study  on mortality and  cancer incidence among  man-made mineral  fiber
     production workers.  Scand. J. Work Environ.  Health 12 (suppl. 1):  65-71.

Bignon,  J.;  Sebastien, P.;  Gaudichet,  A.; Jaurand,  M. C.  (1980)  Biological
     effects  of  attapulgite.  In:  Wagner,  J.  C.;  Davis, W.,  eds.  Biological
     effects  of  mineral  fibres  * Effects biologiques des  fibres  minerales:
     proceedings of  a symposium,  v.  1; September 1979; lyon, France.  Lyon,
     France: World Health  Organization, International Agency for  Research on
     Cancer; pp. 163-181.  (IARC scientific publication no.  30;  1NSERM symposia
     series: v. 92).

Bjornberg,  A. (1985) Glass fiber dermatitis.  Am.  J. Ind. Med.  8:  395-400.

Bjornberg,   A.;  Lowhagen,  G.-B.  (1977) Patch testing with  mineral wool
     (rockwool). Acta Dermatov. 57: 257-260.

Bluth, L. L. ;  Hanscom, T.  A.  (1981) Retinal  detachment and vitreous hemorrhage
     due to talc emboli.  JAMA J. Am. Med. Assoc.  246:  980-981.

Boman, G. ;  Schubert,  V.;  Svane, B.; Westerholm,  P.; Bolinder, E.;  Rohl, A.  N. ;
     Fischbein, A, (1982) Malignant mesothelioma in Turkish immigrants residing
     in Sweden. Scand. J. Work Environ. Health 8:  108-112.

Bouchama, A.;  Chastre, J.; Gaudichet, A.; Soler,  P.;  Gibert, C.  (1984) Acute
     pneumonitis with  bilateral  pleura! effusion  after talc pleurodesis. Chest
     86:  795-797.

Brown, D. P.;  Dement,  J.  M.;  Wagoner,  J.  K.  (1979) Mortality patterns among
     miners and millers  occupationally exposed to asbestiform talc. In:  Lemen,
     R.;  Dement, J. M.,  eds. Dusts and disease:  proceedings of the  conference
     on  occupational  exposures to fibrous and particulate dust  and  their
     extension into the  environment; 1977; Washington, DC. Park Forest South,
     II:  Pathotox Publishers,  Inc.; pp. 317-324.

Carpenter,   J.  L.;  Spolyar, L. W.  (1945)  Negative  chest findings in a mineral
     wool industry. J. Indiana State Med.  Assoc.  38: 389-390.
July 1988                          6-137             DRAFT-DO  NOT QUOTE OR CITE

-------
Casey, K. R. ; Shigeoka, J. W.; Rom, W. N. ; Moatamed, F.  (1985) Zeolite exposure
     and associated pneumoconiosis. Chest 87: 837-S40.

Cherrie, J.;  Dodgson, J. (1986)  Past exposure  to airborne fibers and  other
     potential risk  factors  1n the European Ban-made mineral  fiber production
     industry. In: Saracci, R., ed. Contributions to the 1ARC  study on mortali-
     ty and cancer  incidence among man-made mineral fiber production workers.
     Scand. J. Work Environ.  Health 12 (tuppl.  1): 26-33.

Cherrie, J.;  Dodgson,  J.; Groat,  S.;  MacLeren,  W.  (1986)  Environmental  surveys
     in the European man-made  mineral fiber production industry.  In: Saracci,
     R. , ed.  Contributions to  the IARC study on mortality and cancer Incidence
     among man-made  mineral  fiber production workers.  Scand.  J.  Work Environ.
     Health 12 (suppl. 1): 18-25.

Claude, J.; Frentzel-Beyme,  R. (1984) A mortality study of workers employed  in
     a German rock wool factory. Scand. J. Work Environ. Health 10:  151-157.

Claude, J.; Frentzel-Beyme, R.  (1986) Mortality of workers in  a German rockwool
     factory--a  second look with  extended follow-up.   In: Saracci,  R. , ed.
     Contributions to  the IARC study on mortality  and  cancer incidence among
     man-made mineral  fiber  production workers.  Scand.  J.  Work Environ.  Health
     12 (suppl.  1):  53-60.

Cotton, W.  H. ;  Davidson,  P.  J. (1985)  Aspiration of .haty powder  [letter]. N.
     Engl.  J.  Med. 313: 1662.

Cramer, D.  W.; Welch, W.  R.;  Scully,  R. E.; Wojciechowski, C.  A.  (1982) Ovarian
     cancer and talc: a case-control  study. Cancer 50:  372-376.

Damodar; Agarwal, D. ;  Sharma,  P.  N.  (1983)  Changes  in  ventilatory  capacities
     due to exposure to talc dust. Indian Chest Dis. Allied Sci.  25:  21-24.

Doll,  R.  (1987)  Symposium on  MMMF,  Copenhagen,  October  1986:  overview and
     conclusions. Ann. Occup.  Hyg. 31: 805-819.

Engholm, G.;  Englund,  A.; Fletcher,  T.; Hallin,  N.  (1987) Respiratory  cancer
     incidence  in Swedish construction  workers exposed  to man-made mineral
     fibers. Ann. Occup.  Hyg.  31:  663-675.

Enterline,  P.  E.; Henderson,  V.  (1975) The health of  retired fibrous  glass
     workers.  Arch.   Environ.  Health 30: 113-116.

Enterline,  P.  E.; Marsh, G.  M. (1982) Mortality of workers  in the  man-made
     mineral fibre industry.  In: Biological effects of man-made mineral fibres:
     proceedings  of  a WHO/IARC conference;  April  1982; Copenhagen,  Denmark.
     Geneva,  Switzerland: World  Health  Organization,  Regional  Office  for
     Europe; pp.  965-972.

Enterline,  P. E.; Marsh,  G.  M. (1984) The health of workers in the MMMF indus-
     try.  In: Guthe, T.,  ed.  Biological  effects  of man-made  mineral fibres:
     proceedings  of  a WHO/IARC conference  in association  with JEMRB and T1MA,
     v. 1;  April  1982; Copenhagen, Denmark. Copenhagen,  Denmark: World Health
     Organization, Regional Office for  Europe; pp. 311-339.


.i.iiw iQftR                          8-138             DRAFT--DO NOT QUOTE OR CITE

-------
Enterline,  P.  E.;  Marsh, G. M.; Esmen, N. A. (1983) Respiratory disease among
     workers  exposed to Ban-made  mineral  fibers. Am.  Rev.  Respir.  Dis.
     128: 1-7.

Enterline,  P.  E.;  Marsh, C. M.;  Henderson,  V.  (1986) Mortality update of  a
     cohort of U. S. man-made mineral fiber workers. Presented at:  internation-
     al  symposium  on man-made  mineral  fibres  in  the working  environment;
     October;  Copenhagen,  Denmark.  Copenhagen,  Denmark:  World Health Organiza-
     tion, Regional Office for Europe.

Enterline,  P.  E.;  Marsh, 6. M.; Henderson,  V.;  Callahan,  C.  (1987)  Mortality
     update of •  cohort of U.S. man-made mineral  fiber  workers.  Ann. Occup.
     Hyg. 31: 625-656.

Esmen, N. A.;  Corn, M.; Hammad, Y.  Y.;  Whittier,  D.;  Kotsko, N.;  Haller,  M. ;
     Kahn,  R.  A.  (1979) Exposure of employees  to man-made mineral  fibers:
     ceramic fiber production. Environ. Res. 19: 265-278.

Fienberg, R. (1937) Talcum powder granuloma. Arch.  Pathol.  24: 36-42.

Finnegan, M. J. ;  Pickering, C.  A.  C.; Burge, P.  S. ; Goffe, T. R.  P.; Austwick,
     P. K. C.; Davies, P. S. (1985)  Occupational asthma  in a  fibre glass works.
     J. Soc. Occup. Med. 35: 121-127.

Friedrichs,  K. H.  (1987) Electron microscopic analyses  of dust from the lungs
     and  the  lymph nodes  of talc-mine  employees.  Am.  Ind.  Hyg.  Assoc.  J.
     48:  626-633.

Gamble, J.  F.;  Fellner,  W. ; Oimeo,  M.  J.  (1979) An epidemiologic study of a
     group of talc workers. Am. Rev.  Respir. Dis.  119: 741-753.

Gamble,  J. ;  Greife,  A.; Hancock,  J.  (1982) An epidemiological-industrial
     hygiene study  of  talc -woHters.  In: Walton, W. ,  ed. Inhaled particles V:
     proceedings of  an  international symposium; August  1980;  Cardiff, Wales,
     United Kingdom. Ann. Occup. Hyg. 26: 841-859.

Gamble,  J.;  Sieber, W.  K.; Wheeler,  R.  W.; Reger, R.; Hall,  B.  (1985) A
     cross-sectional study  of  U.  S.  attapulgite workers.  Presented  at:  sixth
     international symposium on inhaled particles; September; Cambridge, United
     Kingdom.

Gantner,  B.  A.  (1986)  Respiratory hazard from removal  of ceramic fiber insula-
     tion from high temperature industrial furnaces.  Am.  Ind.  Hyg.  Assoc.  J.
     47:  530-534.

Gardner,  M.;  Winter, P.  D.;  Pannett, B.; Simpson, M. J.  C.; Hamilton, C.;
     Acheson,  E. D.  (1986) Mortality study of workers in  the man-made mineral
     fiber  production  industry in  the  United Kingdom.  In: Saracci, R.,  ed.
     Contributions to the IARC study on mortality and cancer incidence among
     man-made mineral  fiber production workers. Scand. J.  Work Environ.  Health
     12 (suppl. 1): 85-93.

Gould, S. R. ;  Bernardo,  D. E.  (1972) Respiratory  distress after talc inhala-
     tion. Br. J. Dis. Chest 66: 230-233.


July 1988                          8-139             DRAFT—DO  NOT QUOTE OR CITE

-------
Griffiths,  K. ;  Chandler, J.  A.;  Henderson,  W.  J. ; Joslin, C.  A.  F.  (1973)
     Ovarian  cancer:  some  new  analytical  approaches. Postgraduate  Med.  J.
     49: 69-72.

Hartge,  P.;  Hoover, R.; Lesher,  L.  P.;  McGowan,  L.  (1983) Talc and  ovarian
     cancer. JAMA J. Am. Med. Assoc. 250: 1844.

Heise,  H.;  Mattheus,  A.  (1971) Hautreaktionen auf Isolierstoffe. I.  Epikutane
     Testung von Mineralfasern [Skin reactions to Insulating materials.  I. Skin
     testing of mineral fibers]. Dermatol.  Monatsschr. 157: 52-61.

Henderson, W.  J.  (1972) The use of analytical electron microscopy  in  the  study
     of  foreign  and naturally occurring  minerals  in  biological  tissues. Micron
     3:  131-133.

Henderson, W.  J.;  Griffiths, K.  (1975)  Identification of particles in ovarian
     tissue.  In:  Brush,  H.  G. ; Taylor,  R.  W., eds. Gynaecological malignancy:
     clinical  and  experimental  studies.  Baltimore, MO: Williams and  Wilkins;
     pp. 225-240.

Henderson, W.  J. ;  Joslin,  C. A.  F. ; Turnbull,  A.  C.; Griffiths, K.  (1971) Talc
     and carcinoma  of the ovary and cervix. J. Obstet. Gynaecol. 78:  266-272.

Henderson,  W.  J. ;  Chandler,  J.  A.;  Blundell,  G.; Griffiths, C.;  Davies,  J.
     (1973) The  application of analytical  electron microscopy  to the study of
     diseased biological tissue. J. Microscopy 99: 183-192.

Hill, J.  W. ;  Whitehead,  W.   S. ; Cameron, J.  D.;  Hedgecock, G.  A.  (1973) Glass
     fibres: absence  of  pulmonary hazard in production workers.  Br.  J.  Inc!.
     Med. 30:  174-179.

Hill, J. W.; Rossiter, C. E. ; Foden, D. W.  (1984) A pilot respiratory morbidity
     study of  workers  in a  MMMF  plant in  the United Kingdom.  In:  Guthe,  T. ,
     ed.  Biological effects  of  man-made mineral  fibres:  proceedings of  a
     WHO/IARC  conference  in association  with JEMRB and TIMA,  v. 1; April  1SS2;
     Copenhagen,  Denmark.   Copenhagen,  Denmark:   World Health  Organization,
     Regional Office for Europe; pp. 413-426.

Hook,  H.  L. ;   Morrice,  G. ,   Jr.;  deTreville,  R.  T. P. (1970)  Fibrous  glass
     manufacturing  and  health:  results of  a  comprehensive physiological study,
     part  II.  In:  Transactions  of the 35th  annual  meeting of  the Industrial
     Health  Foundation.  Pittsburgh,  PA:  Industrial  Health   Foundation;
     pp. 103-111.

Joannides, J.  I.  (1978) Talc granuloma  of the  ureter:  a case  report. Br. J.
     Surg. 65: 883-885.

Jones,  H.  D.;  Jones, T. R.;  Lyle, W.  H. (1982)  Carbon fiber:  results  of a
     survey of  process  workers  and their environments in a factory  producing
     continuous filament. Ann. Occup. Hyg.  26: 861-867.

Kleinfeld, M. ;  Messite,  J. ; Kooyman, 0.; Zaki, M.  H.  (1967) Mortality among
     talc  miners and  millers  in New York  State.  Arch.  Environ. Health
     14: 663-667.


July 1988                          8-140             DRAFT—DO  NOT QUOTE OR CITE

-------
Kleinfeld, M.; MessUe, J.; Zaki, M. H. (1974) Mortality experiences among taU
     workers: a follow-up study. JOM J. Occup. Med. 16: 345-349.

Lederer,  C.  M.,  Jr.;  Sabates, F. N. (1982) Ocular findings in the intravenous
     drug abuser. Ann. Ophthalmol. 14: 436-438.

Lewis,  J.  H.; Sundeen,  J.  T.; Simon, G.  L.;  Schulof, R.  S.; Wand, G.  S.;
     Geifand, R. L.; Miller, H.; Garrett, C. T.; Jannotta, F. S.; Grenstein, J.
     M.  (1985) Disseminated  talc granulomatosis. Arch.  Pathol.  Lab.  Med.
     109: 147-150.

Lockey, J. £. (1981) Nonasbestos fibrous minerals. Clin. Chest Med.  2:  203-218.

Lockey, J.  E.;  Brooks, S. M.; Jarabek, A.  M.; Khoury, P. R.; McKay,  R.  T. ;
     Carson, A.; Morrison,  J.  A.; Wiot, 0.  F.;  Spitz, H.  B.  (1984) Pulmonary
     changes after exposure to vermiculite contaminated with fibrous tremolite.
     Am. Rev. Respir.  Dis. 129: 952-958.

Longo,  D.  L. ;  Young,  R.  C.  (1979)  Cosmetic talc and  ovarian  cancer.  Lancet
     2: 349-351.

Lund,  J.  S.;  Feldt-Rasmussen, M. (1969) Accidental  aspiration of talc. Acta
     Poediat. Scand.  58: 295-296.

Maggioni, A.; Meregalli, G.  ; Sala, C. ; Riva, M. (1980) Patologia  respiratoMa  e
     cutanea negli addetti alia produzione di  fibre di vetro  (filato)  [Respira-
     tory and skin diseases in glass fibre workers].  Med. Lav. 3: 216-227.

Malmberg, P.;  Hedenstrom,  H. ; Kolmodin-Hedman, B.; Krantz,  S. (1984) Pulmonary
     function  in  workers of a mineral  rock fibre plant. In:  Guthe,  T. ,  ed.
     Biological effects  of  man-made mineral fibres:  proceedings  of a WHO/1ARC
     conference in association with JEMRB  and T1MA,  v. 1; April  1982;  Copenha-
     gen, Denmark.  Copenhagen,  Denmark: World Health Organization, Regional
     Office for Europe; pp.  427-435.

Mariani-Costantini,  R. ;  Jannotta,  F.  S. ;  Johnson,  F.  B.  (1982)  Systemic
     visceral  talc  granulomatosis associated  with miliary  tuberculosis in a
     drug addict. Am.  J. Clin. Pathol. 78:  785-789.

Milby,  T.  H.; Wclf,  C.  R.  (1969) Respiratory tract  irritation  from fibrous
     glass inhalation. JOM J. Occup. Med. 11:  409-410.

Molos,  M.  A.;  Litton, N.;  Schubert,  T.  T.  (1987)  Talc  liver.  J. Clin.
     Gastroenterol. 9: 198-203.

Morgan,  R.  W.; Kaplan, S.  D.;  Bratsberg,  J. A. (1984)  Mortality in fibrous
     glass production workers.  In:  Guthe,  T.,  ed. Biological effects of man-
     made mineral  fibres:  proceedings of a WHO/IARC conference  in  association
     with JEMRB  and  TIMA, v.  1;  April  1982; Copenhagen, Denmark. Copenhagen,
     Denmark:  World  Health  Organization,   Regional   Office   for Europe;
     pp. 340-346.
July 1988                          8-141              DRAFT-DO  NOT  QUOTE OR CITE

-------
Motomatsu, K.; Adachi, H.; Uno, T. (1979) Two Infant deaths after inhaling baby
     powder.  Chest 75: 448-450.

Moulin, J.  J.;  Mur.  J. M.; Wild, P.; Ferreaux, J. P.; Pham, Q. T. (1986) Oral
     cavity  and  laryngeal cancers  among man-made irineral  fiber production
     workers. Scand.  J. Work Environ. Health 12: 27-31.

Nasr, A.  N.  M.;  Ditchek, T.; Scholtens, P. A. (1971) The prevalence of radio-
     graphic  abnormalities  1n  the chests of fiber glass  workers.  JOM J.  Occup.
     Med.  13: 371-376.

Newball,  H.  H.;  Brahim,  S. A. (1976) Respiratory response to domestic fibrous
     glass exposure.  Environ. Res. 12: 201-207.

Olsen,  J.  H. ; Jensen, 0. M.  (1984)  Cancer incidence among employees  in one
     mineral  wool  production  plant in Denmark. Scand. J. Work Environ. Health
     10: 17-24.

Olsen,  J.  H. ; Jensen,  0. M. ; Kampstrup, 0. (1986) Influence of smoking habits
     and place of  residence on the  risk of lung cancer among workers in one
     rock-wool producing  plant  in Denmark. In: Saracci, R., ed. Contributions
     to the IARC study on mortality and  cancer incidence among man-made mineral
     fiber production  workers.  Scand.  J. Work Environ.  Health 12 (suppl.  1):
     46-52.

Ozesmi, M. ;  Karlsson-Parra,  A.; Hillerdal, G.;  Forsum,  U.  (1986) Phenotypic
     characterization  of  peripheral  blood lymphoid cells in people exposed to
     fibrous zeolite. Br. J. Jnd. Med. 43: 830-833.

Possick, P.  A.;  Gellin,  G.  A.;  Key,  M.  H.  (1970) Fibrous glass dermatitis.  Am.
     Ind.  Hyg. Assoc. J.  31: 12-15.

Research Committee of the British Thoracic Association and the Medical Research
     Council  Pneumoconiosis Unit.  (1979) A survey of the long-term effects of
     talc and kaolin pleurodesis. Br. J. Dis. Chest 73: 285-288.

Rindel, A.; Bach, E.; Breum, N. 0.; Hugod, C.; Schneider, T. (1987) Correlating
     health effect with  indoor air quality 1n kindergartens.  Int.  Arch.  Occup.
     Environ. Health 59:  363-373.

Robinson, C.  F.;  Dement, J.  M.;  Ness,  G. 0.;  Waxweiler,  R.  J.  (1982) Mortality
     patterns of rock and slag mineral wool production workers: an epidemiolog-
     ical and environmental study. Br. J. Ind. Med. 39: 45-53.

Rohl, A.  N.;  Langer, A.  M.; Moncure, G.; Selikoff, I.  J.;  Fischbein, A.  (1962)
     Endemic  pleura!  disease  associated with exposure to mixed fibrous dust in
     Turkey. Science (Washington, DC) 216: 518-520.

Rom, W.;  Greaves, W.; Parry, W.;  Moatamed, F. (1981) Prevalence  of pleura!
     disease  in  a  Nevada community hospital  near  a  fibrous erionite deposit.
     Am. Rev. Respir. Dis. 123: 143.

Rosenthall, L. (1987) Radiophosphate visualization of the foreign body reaction
     to wear debris  from total knee prosthesis. J. Nucl. Med.  28: S15-917.


Julv 1988                          8-142             DRAFT--DO NOT QUOTE OR CITE

-------
Rubino,  G.  F.; Scansetti, G.;  Piolatto,  G.;  Romano,  C.  A. (1976) Mortality
     study of talc miners and millers. JOM J.  Occup.  Med.  18:  186-193.

Sahi, R. P.; Kapoor, V. K.; Seth, S. K.; Yadav, R.  V.  S.  (1967) Talcum granulo-
     ma. J. Ind. Med. Assoc. 48: 550-551.

Saracci,  R.  (1986) Ten  years of epidemiologic  investigations on man-made
     mineral fibers  and  health.  Scand. J. Work Environ. Health 12(suppl. 1):
     5-11.

Saracci, R.; Simonato,  L.  (1982) Man-made vitreous fibers  and workers'  health.
     Scand. J.  Work Environ. Health 8: 234-242.

Saracci,  R.;  Simonato,  L.;  Acheson,  E.  0.; Andersen, A.; Bertazzi,  P. A.;
     Claude, J.;  Charnay,  N. ; Esteve, J.; Frentzel-Beyme, R.  R.;  Gardner,  M.
     J. ; Jensen,  0.  M. ;  Maasing, R.; Olsen, J. H.; Teppo,  L.; Westerholm, P.;
     Zocchetti, C.  (1984)  Mortality and  incidence of cancer of workers in the
     man made  vitreous  fibres producing  industry:  an international investiga-
     tion at 13 European plants. Br. J. Ind. Med.  41:  425-436.

Jebastien, P.;  Gaudichet, A.; Bignon, J.; Baris, Y.  I. (1981) Zeolite bodies in
     human lungs from Turkey. Lab. Invest. 44: 420-425.

Selevan, S. G. ;  Dement,  J.  M.;  Wagoner, J.  K.; Froines, J. R. (1979)  Mortality
     patterns  among  miners and  millers of  non-asbestiform talc:  preliminary
     report. J. Environ.  Pathol. Toxicol.  2: 273-284.

Shannon, H. ; Hayes,  M. ;  Julian, J.  ; Muir,  D.  (1982)  Mortality experience of
     glass fibre  workers.  In:  Biological  effects of  man-made mineral  fibres:
     proceedings of  a  WHO/IARC  conference, v. 1; April; Copenhagen,  Denmark.
     Geneva, Switzerland:  World Health  Organization,  Regional  Office  for
     Europe; pp. 347-349.

Shannon, H. S. ;  Hayes, M.;  Julian,  J.  A.; Muir,  D.  (1984)  Mortality  experience
     of glass fibre workers. Br. J.  Ind. Med.  41:  35-38.

Shannon, H. S. ;  Jamieson,  E. ; Julian, J.  A.;  Muir, D. C.  F.; Walsh,  C. (1986)
     Mortality experience of glass fibre workers--extended follow-up. Presented
     at:  international  symposium on  man-made mineral fibres  in the  working
     environment;  October;  Copenhagen, Denmark.  Copenhagen,   Denmark:  World
     Health Organization, Regional Office for Europe.

Shannon, H. S.;  Jamieson,  E.; Julian, J.  A.;  Muir, D. C.  F.; Walsh,  C. (1987)
     Mortality experience of glass  fibre workers --  extended follow-up.  Ann.
     Occup. Hyg. 31: 657-662.

Simonato,  L.;  Fletcher,  A.  C.;  Cherrie,  J.;  Andersen,  A.; Bertazzi,  P.  A.;
     Charney,  N.;  Claude,  J.;  Dodgson,  J.; Esteve,  J.; Frentzel-Beyme,  R.;
     Gardner, M. J.; Jensen, 0.; Olsen, J.; Saracci,  R.; Teppo, L.; Westerholm,
     P.; Winklemann,  R.; Winter, P.  0.;  Zocchetti,   C.  (1986a) The  man-made
     mineral fibers  (MMMF)  European historical cohort study:  extension of the
     follow-up.  Presented  at:  international  symposium on  man-made  mineral
     fibres  in  the working environment;  October;   Copenhagen,   Denmark.
     Copenhagen, Denmark: World  Health Organization,  Regional  Office for Europe.


July 1988                          6-143              DRAFT-DO NOT QUOTE OR CITE

-------
Simonato,  L. ;  Fletcher, A.  C.;  Cherrie,  J.;  Andersen, A.; Bertazzi, P.  A.;
     Charney,  N. ;  Claude,  J.;-Dodgson, J.; Esteve, J.;  Frentzel-Beyme,  R.;
     Gardner, M. J. ; Jensen, 0.; Olsen, J.; Saracci, R.; Teppo, L.; Westerholm,
     P.; Winklemann,  R. ;  Winter, P. D.;  Zocchetti,  C.  (1986b) Updating lung
     cancer  mortality among a  cohort  of  man-made mineral  fibre production
     workers in seven European countries. Cancer Lett. 30: 189-200.

Simonato,  L.;  Fletcher, A.  C.;  Cherrie,  J.;  Andersen, A.; Bertazri, P.  A.;
     Charnay,  N.;  Claude,  J.;  Dodgson, J.; Esteve, J.;  Frentzel-Beyme,  R. ;
     Gardner,  M.  J.; Jensen,  0.  M.; Olsen, J. H.;  Saracci,  R.;  Teppo, L. ;
     Winkelmann, R. ;  Westerholm,  P.; Winter,  P. D.; Zocchetti, C. (1986c) The
     man-made mineral  fiber European historical cohort study.  Scand. J.  Work
     Environ. Health 12(suppl. 1):  34-47.

Simonato,  L. ;  Fletcher, A.  C. ;  Cherrie,  J.  W.; Andersen, A.;  Bertazzi,  P.;
     Charnay,  N. ;  Claude,  J.;  Dodgson, J.; Esteve, J.;  Frentzel-Beyme,  R. ;
     Gardner,  M.  J.;  Jensen,  0. M.;  Olsen,  J.;  Saracci,  R.; Teppo,  L.;
     Winkelmann, R.;  Westerholm,  P.; Winter,  P.  D.; Zocchetti, C.;  Saracci,  R.
     (1987) The  International  Agency for Research on Cancer historical  cohort
     study of MMMF production workers in  seven European countries: extension of
     the follow-up. Ann. Occup. Hyg. 31: 603-623.

Skuric, 2.;  Stahuljak-Beritic,  D.  (1984) Occupational  exposure and ventilatory
     function  changes in  rock  wool workers.  In:  Guthe,  T.,  ed.  Biological
     effects of  man-made  mineral fibres: proceedings of a WHO/IARC conference
     in association with JEMRB and  TIMA,  v. 1; April 1982; Copenhagen,  Denmark.
     Copenhagen,  Denmark:   World  Health  Organization,  Regional  Office for
     Europe; pp. 436-437.

Stille, W. T.;  Tabershaw,  I. R.  (1982) The mortality experience of upstate New
     York talc workers. JOM  J. Occup. Med. 24: 480-484.

Stokholm, J. ;  Norn,  M. ; Schneider,  T.  (1982)  Ophthamologic effects of man-made
     mineral fibers. Scand.  J. Work Environ. Health 8:  185-190.

Tao, L.-C.;  Morgan,  R.  C. ; Donat,  E. E. (1984)  Cytologic  diagnosis of intrave-
     nous  talc  granulomatosis by fine  needle  aspiration  biopsy.  Acta  Cytol.
     28: 737-739.

Teppo,  L. ;  Kojonen,  E.  (1986) Mortality  and cancer risk among workers  exposed
     to man-made mineral  fibers in  Finland.  In:  Saracci,  R.,  ed.  Contributions
     to  the  IARC  study on  moratality  and cancer  incidence  among man-made
     mineral  fiber production workers.  Scand.  J.  Work Environ.  Health  12
     (suppl. 1): 61-64.

Thomas, T. L.; Stewart, P. A. (1987) Mortality from lung cancer and respiratory
     disease  among pottery workers  exposed  to  silica and  talc.  Am.  J.
     Epidemic!. 125: 35-43.

Tse, D. T.; Ober, R. R. (1980) Talc retinopathy. Am. J. Ophthalmol. 90:  624-640.
July 1988                          8-144             DRAFT--DO  NOT QUOTE OR CITE

-------
Tukiainen, P., Nickels, J.; Taskinen, E.; Nyberg, M.  (1984) Pulmonary granulom-
     atous reaction:  talc pneumoconiosis  or chronic sarcoidosis? Br. J.  Ind.
     Med. 41: 84-87.

Tye, M.  J.;  Hashimoto, K.;  Fox,  F.  (1966)  Talc  granulomas  of  the  skin. JAMA  J.
     Am. Med. Assoc. 198: 120-122.

Utidjlan, H. M. 0.; deTreville, R. T. P. (1970) Fibrous glass manufacturing and
     health: report  of an tpidemiological  study, part I.  In: Transactions of
     the 35th  annual  meeting of the Industrial Health Foundation. Pittsburgh,
     PA: Industrial Health Foundation; pp.  98-102.

Vallyathan;  N.  V.; Craighead,  J.  E. (1981) Pulmonary pathology  in workers
     exposed to nonasbestifonn talc. Hum.  Pathol. 12: 28-35.

Waller,  B.  F. ; Brownlee,  W.  J. ; Roberts,  W.  C. (1980)  Structure-function
     correlations  in  cardiovascular and pulmonary diseases (CPC).  Self-induced
     pulmonary granulomatosis:  a  consequence of intravenous injection of drugs
     intended for oral use. Chest 78: 90-94.

Waxweiler, R.  J.;  Zumwalde,  R.  D.; Ness, G. 0.; Brown, D. P. (1985) Mortality
     among persons  mining and milling attapulgite clay.  Cincinnati, OH:  U.  S.
     Department of  Health and Human Services,  National  Institute for Occupa-
     tional Safety and Health, Centers for Disease Control.

Wegman, D. H. ;  Peters, J. M. ; Boundy,  M. G.; Smith,  T. J.  (1982)  Evaluation  of
     respiratory effects  in miners and millers  exposed to talc free of asbestos
     and silica.  Br. J. Ind. Med. 39: 233-238.

Weill, H. ; Hughes,  J.  M.  ; Hammad,  Y.  Y.;  Glindmeyer,  H.  W. ,  III; Sharon, G. ;
     Jones,  R.  N.   (1983)  Respiratory  health in workers exposed  to man-made
     vitreous fibers. Am. Rev. Respir.  Dis. 128: 104-112.

Wells, I.  P.;  Dubbins, P. A.; Whimster, W. F.  (1979) Pulmonary disease  caused
     by the inhalation of cosmetic talcum powder. Br. J.  Radiol. 52: 586-586.

Westerholm,  P.;  Bolander, A.  M. (1986) Mortality and  cancer incidence in the
     man-made mineral  fiber industry in Sweden.  In:  Saracci,  R.,  ed.  Contribu-
     tions to  the  1ARC study on mortality and cancer incidence among man-made
     mineral  fiber production workers.  Scand.  J. Work  Environ.  Health 12
     (suppl. 1): 78-84.

Wright, G. W.  (1968) Airborne fibrous  glass particles: chest roentgenograms  of
     persons with prolonged exposure. Arch. Environ. Health 16: 175-181.

Ziskind, M.; Weill,  H.; Anderson, A. E.; Samini, B.; Neilson, A.;  Waggenspak,
     C. (1976) SHicosis  in shipyard sandblasters. Environ. Res. 11: 237-242.
July 1988                          8-145              DRAFT-DO  NOT QUOTE OR CITE

-------
                  9.  QUANTIFICATION OF TOXICOLOGICAL EFFECTS
     The  quantification  of toxicological effects  of  a chemical  consists of
separate  assessments  of  noncarcinogenic  and carcinogenic effects.  Noncarcino-
genic effects are believed to occur via a threshold mechanism for which  concen-
trations  that produce  adverse health  effects can  be identified,  while
carcinogenic mechanisms are assumed to act without a threshold.   The  procedures
used to  quantify  toxicological  effects are based  on guidelines proposed by the
U.S. Environmental  Protection  Agency  (1986,  1988) for inhalation exposure or
oral intake.  No  guidelines  are available for intraperitoneal or intrapleural
administration.   These procedures are summarized below.
9.1  PROCEDURES FOR QUANTIFICATION OF TOXICOLOGICAL EFFECTS
9.1.1  Noncarcinogenic Effects
     In the  quantification  of noncarcinogenic effects, a Reference Dose (RfD)
is calculated.  The RfD is an estimate (with an uncertainty spanning perhaps an
order  of  magnitude) of a daily  exposure  of the human population  (including
sensitive  subgroups)  that  is  likely to  be  without an appreciable risk of
deleterious health effects during a lifetime.
     RfDs  are  typically  calculated using  a  single  exposure  level  and  uncer-
tainty  factors that account  for specific deficiencies in the toxicity date
base.   Both  the  exposure level and the uncertainty factors  are  selected  and
evaluated  in  the  context of all  available chemical-specific  literature.  After
all  toxicological,  epidemiologic,  and supporting data have  been  reviewed and
evaluated, a  key  study Is selected that reflects optimal data on the critical
effect.  Dose-response data points for all  reported effects are examined  as a
component  of this review.
     The threshold  concept  is the basis for the derivation of the RfD.  Essen-
tially, an experimental  exposure level is selected from the available studies
that represents the  highest level tested  in which  an adverse effect was  not

July 1988                          9-1          DRAFT—DO NOT QUOTE OR CITE

-------
demonstrated.  This  chosen  No-Observed-Adverse-Effect  Level  (NOAEL) represents
the first quantitative basis for the scientific evaluation of the risk posed  to

humans by systemic toxicants.  The RfD is operationally derived from this NOAEL
or from the Lowest-Observed-Adverse-Effect Level (LOAEL) by consistent applica-
tion of  general  order-of-magnitude uncertainty factors (UFs) that reflect the

second quantitative  basis  of this scientific evaluation of risk.  Uncertainty
factors  are associated  with various specifically  recognized uncertainties
(i.e., toxicity  data base  deficiencies) when extrapolating  from the  type of
study  from  which the  RfD  is estimated.  Suggested uncertainty factors for
extrapolation follow:
Human to Sensitive Human
Animal to Human
Subchronic to Chronic
LOAEL to NOAEL
Incomplete to Complete
Use a  10-fold  factor when extrapolating from valid
experimental results  of  studies  using  prolonged
exposure to average  healthy  humans.   This  factor is
intended to account  for  the  variation  in sensitivity
among the members of the human population.

Use an  additional  10-fold factor  when extrapolating
from valid  results of long-term studies on  experi-
mental  animals  when results of human exposure stucies
are not  available  or are inadequate.   This  factor  isi
intended to account  for  the  uncertainty in extrapo-
lating from animal data to the case of humans.

Use up  to  an additional  10-fold factor when extrapo-
lating from less than chronic results on experimental
animals  when there are  no useful  long-term human
data.    This  factor is intended to  account  for  the
uncertainty in  extrapolating from less than chronic
NOAELs to chronic NOAELs.

Use up  to  an additional  10-fold factor when deriving
an RfD  from a  LOAEL  instead  of  a  NOAEL.  This factor
is intended to  account for the uncertainty in extrap-
olating from LOAELs to NOAELs.

Use up  to  a 10-fold factor when  extrapolating  from
valid results  in  experimental  animals when the  data
are "incomplete."  This factor is  intended to account
for the inability of  any single  animal  study  to
adequately address all  possible adverse outcomes in
humans.
     An additional  modifying factor (MF) reflects professional judgment of the
entire data available on the specific agent and is usually >0  and <10.
July 1988
         9-2
DRAFT—DO NOT QUOTE OR CITE

-------
The inhalation RfD is derived from the NOAEL as:

                              RfD = NOAEL/UF x MF
where:
     UF « an uncertainty factor suited to the characteristics  of the  data.
     MF - a modifying factor, based on professional  judgment of the entire
          data base (e.g., sample size).


9.1.2  Carcinogenic Effects
9.1.2.1  Studies  in Humans.   Evidence of carcinogenicity from human studies
comes from case reports, descriptive or cross-sectional  studies, and  analytical
epidemiology or case-control  and cohort  studies.  The weight  of evidence for
carcinogenicity from studies in humans is classified as  follows:

     1.    Sufficient evidence  of carcinogenicity, which  indicates that  a
          causal  relationship  exists between the agent and human  cancer.

     2.    Limited  evidence  of  carcinogenicity,  which  indicates  that  a
          causal interpretation is credible, but that alternative explana-
          tions,  such  as  chance,  bias,  or  confounding,  could  not  be
          adequately excluded.

     3.    Inadequate evidence, which  indicates  that there were few avail-
          able  studies  with  little  pertinent  data, which while  showing
          evidence  of  association,  did  not  exclude  chance,  bijs,  or
          confounding,  and  therefore,  a  causal   interpretation  is  not
          credible.

     4.    No data, which  indicates that data are not available.

     5.    No  evidence,   which indicates  that  no  association was  found
          between  exposure  and   an   increased  risk  of  cancer in  well-
          designed  and  well-conducted  independent  analytical  epidemi-
          ologic studies.


.Inlv 1PRR                          Q-3          DRAFT—HO. NOT  nilDTF OR CITE

-------
9.1.2.2  Studies in Animals.  Evidence of carcinogenicity  in  animals  is classi-
fied into five groups:

     1.   Sufficient  evidence  of  carcinogenicity,   which  indicates  an
          increased  incidence of malignant  tumors  or combined malignant
          and benign  tumors (1)  in  multiple  species  or strains; or  (2) in
          multiple experiments (e.g., with different  routes of  administra-
          tion  or using  different   dose  levels);  or  (3) to  an unusual
          degree  in   a  single experiment with  regard  to   high  incidence,
          unusual site or type of tumor, or early age at onset.

          Additional   evidence may  be  provided  by  data on  dose-response
          effects, as well  as  information  from short-term  tests  or on
          chemical structure.

     2.   Limited evidence  of carcinogenicity,  which means that the  data
          suggest a  carcinogenic effect  but are limited  because (1)  the
          studies involve  a single  species,  strain,  or experiment  and do
          not meet  criteria for sufficient  evidence;  (2)  the  experiments
          are restricted by inadequate  dosage  levels,  inadequate duration
          of  exposure to the  agent, inadequate period of followup,  poo-
          survival,  too few animals, or inadequate reporting;  or (3) there
          is an increase in the incidence of benign tumors only.

     3.   Inadequate  evidence,  which  indicates that  because  of   major
          qualitative  or  quantitative  limitations,  the studies cannot be
          interpreted  as  showing   either  the  presence  or  absence of  a
          carcinogenic effect.

     4.   No data, which indicates that data are not  available.

     5.   No  evidence, which indicates  that  no  increased  incidence of
          neoplasms  has been found in  at  least two well-designed  and
          well-conducted animal studies in different species.
July 1988                          9-4          DRAFT—DO NOT QUOTE OR CITE

-------
     The EPA categorizes the carcinogenic potential  of  a  chemical, based on the
overall weight of evidence, according to the following  scheme:

     •  Group A:  Human Carcinogen.    Sufficient   evidence  exists  from
                  epidemiology  studies   to  support  a  causal  association
                  between  exposure  to  the chemical  and  human cancer.
     •  Group B:  Probable Human Carcinogen.    Sufficient   evidence   of
                  carcinogenicity in animals with  limited (Group Bl) or
                  inadequate (Group 62) evidence in humans.
     •  Group C:  Possible Human Carcinogen.     Limited     evidence    of
                  carcinogenicity in animals in the absence of human data.
     •  Group D:  Not Classified as to Human Carcinogenicity.    Inadequate
                  human  arid  animal  evidence  of  carcinogenicity or  for
                  which no data are available.
     •  Group E:  Evidence of Noncarcinogenicity for Humans.   No  evidence
                  of carcinogenicity in at least two adequate animal  tests
                  in different  species  or  in  both  adequate epidemiologic
                  and animal studies.

     If lexicological evidence  leads to the classification of the contaminant
as a  known,  probable,  or possible hunan tBrcinogsn, mathematical models  are
used to calculate  the  estimated excess  cancer  risk  associated  with exposure  to
the chemical.   The data used in  these  estimates usually come from lifetime
exposure studies  in  animals.  To predict the risk  for  humans  from animal date,
animal  doses  must be converted to  equivalent  human doses.  This conversion
includes correction for noncontinuous exposure, less-than-lifetime studies,  and
for differences  in  size.   The factor that compensates  for the size  difference
is the  cube  root of the ratio  of the  animal and human  body weights.  It is
assumed that the average adult human body weight is 70  kg.
     Chemical  levels  are correlated with a carcinogenic  risk estimate by
employing a  cancer  potency (unit risk) value together  with the assumption  for
lifetime exposure  via  inhalation or ingestion.   The cancer risk is usually
derived from the linearized multistage model with • 95  percent upper confidence
limit providing  a  low-dose estimate; that  is, the  true  risk to  humans, while
not identifiable,  is not likely  to  exceed  the upper limit estimate and, in
fact,  may be  lower.   Excess cancer  risk estimates  may  also be  calculated using
other models  such  as the one-hit, Weibull,  logit,  and  probit.  There  is  little
basis  in the  current understanding of  the  biological  mechanisms involved  in

i.,i,, TQDP                          Q-n          noj&rT—nn wni  nnnic nP fTTF

-------
cancer to  suggest  that any one of these  models  is  able to predict  risk  more
accurately than  any  other.   Because  each model is based on differing assump-
tions, the estimates that are derived  for  each model  can  differ by several
orders of  magnitude.   However,  a  model  that incorporates low dose linearity in
tumor  formation, such  as the multistage,  is preferable  as  it is more scientif-
ically plausible.
     The  scientific  data base used to  calculate  and support  the setting of
cancer risk  rate levels has an inherent uncertainty due to the systematic  and
random errors  in scientific measurement.   In most cases,  only  studies using
experimental animals have been performed.   Thus,  there is only  uncertainty  when
the data  are extrapolated to humans.  When  developing cancer risk rate levels,
several other  areas of  uncertainty exist,  such  as  the incomplete  knowledge
concerning the  health  effects of  chemicals  in the environment; the impact  of
the experimental animal's age, sex, and species;  the nature of  the target organ
system(s)  examined;  and the actual rate  of exposure of the internal target
organs in  experimental  animals or humans.   'Dose-response  data  usually  are
available  only  for high levels of exposure, not for lower  levels of exposure,
levels that  are closer to the standard and closer to levels to  which humans are
exposed.    When  there  is  exposure to more  than one  contaminant,  additional
uncertainty  results  from a lack of information about possible  synergistic  or
antagonistic effects.
9.2  MANMADE MINERAL FIBERS
9.2.1  Human Health Effects
     The  carcinogenicity  of crocidolite  asbestos  fibers in both humans and
animals  has  been well-established.   The  available epidemiological  evidence
clearly  demonstrates  an association  between exposure to  asbestos  and the
subsequent development  of  carcinomas  of the lung and diffuse mesotheliomas of
the pleura and peritoneum.   According to  the International  Agency for  Research
on Cancer (IARC), following a long latency period,  an increased risk of pleura!
and peritoneal  mesotheliomas with  increasing duration of exposure  has been
reported  in  persons occupationally or environmentally  exposed  to  asbestos
fibers  (National Research  Council,  1984).   The epidemiological study  of
respiratory  diseases  and asbestos  exposure has detected  low to moderately-
increased risk  ratios  (two- to fivefold) for  lung  cancer,  laryngeal cancer,

July 1988                          9-6          DRAFT—DO NOT QUOTE  OR CITE

-------
gastrointestinal cancer, and colorectal  cancer,  as  well  as high risk ratios
(ninefold) for  lung  cancer  among those with asbestosis.  In addition, there is
a  we 11-documented  synergistic  relationship  between asbestos  exposure  and
smoking for  the  development of  lung cancer.  However, the risk of mesothelioma
appears to be independent of smoking.
     Because of  the  physical similarities between WIMP and asbestos fibers and
because of the  widespread and Increasing use of HMMF, numerous epidemiological
studies have been  conducted to  determine whether a  causal  relationship exists
between exposure to  MMMF and  malignant and nonmalignant respiratory disease.
These studies,  all  of which are historical prospective in design,  have  been
critically reviewed in Chapter 8 and summarized in Table  8-12.   No  epidetniologic
studies were found for ceramic fibers.
     Limitations in  the  epidemiologies!  method in elucidating  occupational and
environmental hazards  are  more often  a result  of  an  absence of available
information  rather  than a  flaw in  methodology.   In the case  of MHMF, many
well-conducted studies exist when  evaluated  in the  context of adequate design
and  followup.   However,  factors such  as low estimated levels  of  exposure or
short duration  of  exposure, coupled  with insufficient  power to detect  a
50 percent increased  incidence  of  cancer,  NMRD,  or  mesothelioma when the data
are  divided  into multiple  subcategories, make the observed risk sufficiently
low to be masked by background mortality rates.   This especially holds true for
lung cancer, which  has a high background rate.   Therefore, since an epidemio-
logical study,  like any other  screening test, is an  instrument of limited
sensitivity, however well   a study  is  conducted it can only  detect effects
within the limits of sensitivity of the experimental design.
     While the  overall  quality  of  the  epidemiological  studies  on MMMF is  good,
because no actual  exposure  data on historical levels of MMMF  exist with  which
to establish a  dose-response  relationship,  these data are unsuitable for the
quantification of  risk.  A  summary of the weight of evidence  for causal  rela-
tionships between exposure to various MMMF and NMRD  and carcinogenic effects  is
presented in Table 9-1.
9.2.1.1  Fibrous Glass
9.2.1.1.1   Qualitative evaluation  of noncarcinogenic effects.   An  evaluation
of noncarcinogenic  effects  following  exposure of workers to glass  microfibers
indicates no association between  exposure to glass  microfibers and subsequent
development of NMRD.  Similarly, it appears that no association exists between

July 1988                          9-7          DRAFT—DO NOT QUOTE OR CITE

-------
    TABLE 9-1.  SUMMARY OF THE WEIGHT OF EVIDENCE FOR CAUSAL  RELATIONSHIPS
     BETWEEN EXPOSURE TO MMMF AND CARCINOGENIC AND NONCARCINOGENIC  EFFECTS
      Fiber
      Effect
   Weight
of Evidence
Glass microfibers
Fibrous glass wool
Continuous fibrous glass
  filament

Mineral wool
Ceramic fibers
NMRDa
Respiratory cancer

NMRD
Lung cancer
Laryngeal cancer
Buccal cavity cancer

NMRD
Lung cancer

NMRD
Respiratory cancer
Lung cancer
Digestive system cancer
Stomach cancer
Buccal cavity and
  pharyngeal cancer

         b
 Inadequate
 Inadequate

 Limited
 Limited
 Limited
 Limited

 Inadequate
 Inadequate

 Inadequate
 Limited
 Limited
 Limited
 Limited
 Limited
 Nonmalignant respiratory disease.

5No data are available.
exposure to  fibrous  glass  wool  and subsequent development  of  NMRD.   In both
cases,  the  apparent absence  of  an association,  however, is limited  by the
sensitivity  of  the  studies to detect  a  50  percent  increase  in  NMRD.   Enterline
et al.  (1983,  1986) found  excess  NMRD  in the combined  fibrous  glass wool/
continuous  fibrous  glass  filament  category,  but did not find an  excess  of
NMRD  associated with fibrous  glass wool or continuous fibrous glass  filament
when  the data  were  analyzed by fiber.  Only  one study (Bayliss  et al., 1976)
detected an  excess  of  NMRD in fibrous glass wool workers (SMR  =  189,  p <0.05,
Power = 34 percent), which increased among  those  with 10 or  more years  of
latency.  This  twofold  excess was  limited  to  one study and  is  insufficient for
establishing a  causal  relationship.   Therefore,  the weight of evidence for a

causal  relationship  between  fibrous  glass wool  and  NMRD is  limited.  The
evidence is  inadequate  to  establish a causal relationship  between continuous

fibrous glass filament and NMRD.

-------
9.2.1.1.2  Characterization of carcinogenic effects.   An  evaluation of carcino-
genic effects  following  exposure  of workers to glass  microfibers  indicates no
association between exposure to glass microfibers  and  subsequent development of
cancer.   Only  four studies focused on workers  in MMMF plants  who were  most
likely to  have been exposed to nicrofibers.  Neither  the series of studies by
interline  et  al.  (1983,  1986, 1987) nor the  study  by Bayliss  et  al.  (1976)
found significant  excesses of  respiratory or other cancers associated  with
exposure to  microfibers.  The  relationship between the  calculated SMRs for
respiratory cancer and  exposure  to nricrofibers reported in the  studies by
Enter!ine  et al.  (1986,  1987) and Bayliss  et  al. (1976) led the  authors to
conclude that  these data supported the hypothesis of an association  between
microfiber exposure  and  respiratory cancer.  Since  the excesses reported were
not statistically  significant,  however,  and nay have  been elevated because  of
random variation or as a result of chance, such excesses could not be used  to
establish  a causal  relationship.   The lack of  statistical  significance, how-
ever, might be attributed to the insufficient sensitivity to detect  a 50 per-
cent increase  in  respiratory cancer, if it did,  in fact, exist among workers
exposed  to glass   microfibers.  Thus,  the  weight of  evidence  for a  causal
relationship between microfibers and respiratory cancer was inadequate.
     Concerning carcinogenic  effects  following exposure  of workers to fibrous
glass wool,  investigators found excesses  from respiratory  cancer,  including
lung  cancer  and  laryngeal cancer;  from stomach cancer;  and  from buccal
cavity/pharyngeal   cancer,  especially buccal cavity  cancer.  Enterline  et al.
(1986) found an excess of respiratory system cancer  in U.S.  male production  and
maintenance workers  after 20 or more years of latency following  exposure  to
both fibrous glass wool  and continuous fibrous glass filament  (SMR = 129.5,
p <0.01),  which in the subsequent study (Enterline  et al.,  1987)  was  decreased
and  no  longer significant  compared to local  mortality  rates (SMR =  110.6).
Similarly, an  excess of  respiratory cancer among those with 30 or  more years of
latency  following  exposure to both fibrous glass wool and  continuous fibrous
glass filament (SMR = 137.8, p <0.01), and which increased after 20 to 29 years
of employment  (SMR = 160.1, p <0.01) (Enterline et al.,   1986), was no  longer
significant compared  to  local mortality rates (Enterline et al.,  1987).  Since
local mortality rates  are more appropriate than  national  rates as comparison
groups  for respiratory  cancer,  these results  suggest that no relationship
existed  between respiratory cancer and exposure of workers to  fibrous  glass

July 1988                          9-9          DRAFT-DO NOT QUOTE OR CITE

-------
wool In  combination  with continuous fibrous glass filament.   While there was
sufficient  power  to detect  respiratory  cancer among those exposed to both
fibrous  glass  wool  and continuous  fibrous  glass  filament,  there was
insufficient power when  the  cohort was subdivided by latency or  duration  of
exposure.   Therefore,  the  weight of evidence for a causal relationship between
combined exposure to fibrous glass wool  and continuous  fibrous  glass filament
and respiratory cancer is inadequate.
     In  Canada, Shannon  et al.  (1986, 1987) found a  twofold  excess  (p <0.005,
Power =  36  percent)  of lung  cancer in male  production workers with the highest
exposure to fibrous glass wool when compared to provincial  mortality rates.   In
addition, an excess  of lung cancer was observed among those with the longest
duration of exposure  (>S  years) and the longest time since first  exposure
(>10 years) (SMR = 182,  p  <0.05,  Power = 28 percent).   This  suggests a causal
relationship between exposure to fibrous glass wool  and lung cancer that  also
was associated  with  increasing  latency and duration  of  exposure.   In England,
Gardner  et  al.  (1986)  found a  30 percent  increase  in lung cancer mortality
(p <0.05, Power = 95 percent) among male fibrous glass wool production workers
when compared to  national  rates.   Although  additional excesses were found when
the data from male  production workers were  analyzed  by  latency and duration of
employment, these excesses were not  significant compared to local  mortality
rates.    However,  the low power  to detect lung cancer when the  data  among  male
production  workers were  analyzed by duration of employment and by latency may
have contributed  to  the apparent  absence of a  lung  cancer risk using local
mortality rates.  Pooling  data  from 13 European MMMF plants, Simonato et al.
(1986a,b,c) found an excess  of  lung  cancer among fibrous  glass  wool workers
compared to national rates (SMR = 127, p <0.05), especially among those with  30
or more  years  since  first  exposure to fibrous glass  wool  (SMR « 173, p <0.05).
There was  an  excess  of  lung cancer  during  the  intermediate  phase of fibrous
glass wool  production  (SMR » 164, p <0.05).  Among those with 20 or more years
of latency, lung  cancer did not  appear to  increase  with  increasing length of
exposure.   In  subsequent studies,  however,  Simonato et al.  (1986b,c,  1987)
reported that the above excesses from lung cancer were no longer significant
compared to local mortality  rates.   Again, the apparent absence of a  lung
cancer risk using local mortality rates may be attributable to  the insufficient
power to detect a 50 percent increase in lung  cancer when the  data were sub-
divided  into latency and duration groups.   Therefore, the data are  equivocal:

Julv 1988                          9-10         DRAFT—DO NOT QUOTE  OR  CITE

-------
a twofold  significant  excess  of lung cancer among Canadian  fibrous glass wool
workers, which  was associated with  latency  and  duration,  supports  a causal
relationship between lung  cancer and exposure to  fibrous wool.  The failure to
sustain significant excesses  of lung cancer among English and other European
fibrous glass wool  workers when compared to local rates refutes such a rela-
tionship.   Thus,  the weight  of evidence  for a  causal relationship  between
fibrous glass wool and lung cancer is limited.
     Concerning other carcinogenic effects following exposure to fibrous glass
wool,  in Italy,  Bertazzi  et al. (1984) found a  fourfold  excess  of  laryngeal
cancer among male  fibrous  glass wool production workers compared to national
but  not  local  or regional rates.   Despite  low power  and  only four observed
laryngeal   cancer  deaths,   compared to  local  mortality rates, a  consistent,
significant increase was observed among the fibrous glass  wool workers  who  were
most heavily exposed.   The excesses  ranged from eightfold among  those  with 20
or more years of  exposure to fifteenfold for those whose  first  exposure was
before age 25.   All  four  cases were first employed before  1960  when fibrous
glass wool rather  than continuous  fibrous glass  filament was  produced, thereby
indicating a temporally correct association.   This excess was confirmed in a
followup study  by Bertazzi  et  al.  (1986) that  detected  nine deaths  from
laryngeal  cancer, all of which occurred in male  production workers.   Six  of the
nine deaths  occurred in  the  Italian  fibrous glass  wool  cohort studies by
Bertazzi et al.   (1984,  1986)  and Simonato et al.  (1986a,b,c)  and resulted  in a
fourfold increase  (6 observed, 1.5  expected, SMR =  400,  p <0.05,  Power  =
8 percent).  In  other  plants, the  apparent lack  of an  excess from laryngeal
cancer mortality  could be  attributed  to the  insufficient  power to detect
50 percent increased  risk of death  from laryngeal cancer.   The additional
deaths  identified  by Simonato  et  al.  (1986c) after increased followup con-
firmed the previous  findings  by Bertazzi et al.  (1984, 1986) of an elevated
laryngeal cancer  risk  in  the Italian fibrous glass wool  cohort.   The strength
of  the association despite low power,  I.e.,  large SMRs that increased with
increasing latency and length  of  exposure, the  significant  findings  after
comparison with  high  local rates  of laryngeal cancer that  would control  for
exogenous  risk  factors  such as  smoking and  alcohol consumption,  the  additional
cases  among male  production workers  identified  by Simonato  and colleagues  that
were consistent with  previous excesses identified by Bertazzi and colleagues,
and  the biological  plausibility for the excess laryngeal  cancer risk,  suggest

-------
that a  causal  relationship existed between exposure to fibrous  glass wool and
laryngeal  cancer.   However, the  excess  is apparently  limited  primarily to
Italian  workers  and is based on  only  six  deaths.   Therefore,  the weight of
evidence for  a causal  relationship between exposure to fibrous  glass wool and
laryngeal cancer is limited at the present time.
     In  England, Gardner  et al.  (1986) found an excess  of stomach cancer among
female employees of a fibrous glass wool  plant  compared to national  (SMR  = 277,
p <0.05, Power = 8  percent),  but not local, mortality  rates.   No excesses of
stomach  cancer were found in other studies  following exposure to fibrous glass
wool.   A threefold  excess of stomach cancer limited to one  study is insuffi-
cient for  establishing  a  causal  relationship.   Therefore, the  weight of evi-
dence for  a causal relationship  between exposure  to  fibrous glass wool  and
stomach  cancer is inadequate.
     In  France, Moulin  et al.  (1986) found a twofold  increased incidence of
upper respiratory and  alimentary tract cancers among active  or  retired fibrous
glass wool  workers (SIR =  218,  p <0.005,   Power =  33  percent)  compared  to
regional rates.  Despite low power, investigators found a  threefold excess from
buccal  cavity  cancers  (SIR = 317, p <0.001, Power = 22 percent).   Significant
excesses were  greater among  production  workers  in the two  highest latency
groups,   10 to 19  years  (SIR =  304,  p 20 years (SIR  =  333,
p <0.005).   Although a threefold excess of buccal cavity cancer  provides  strong
evidence for  a causal  relationship, it is  limited  to  only one  study.  There-
fore, the  weight of evidence for a causal  relationship between exposure to
fibrous   glass wool  and buccal cavity cancer is  limited.
     Concerning carcinogenic  effects  following exposure of workers  to continu-
ous  fibrous  filament,  Simonato  et al.  (1986c)  found an excess  of  lung cancer
among male and female production workers  from the combined European cohort with
less than  10 years  of latency compared to national rates  (SMR = 224,  p <0.05,
Power =  15 percent) but  not compared to  local rates (Simonato et al., 1986b,c,
1987).   The  appearance  of an excess after  so  short a latency  period argues
against  a  causal  relationship.   Other studies of  continuous fibrous  glass
filament workers found no excess site-specific  cancers (Enterline et al., 1986,
1987; Bertazzi et  al.,  1984, 1986; Gardner et al., 1986);  none of these had
sufficient power to detect lung cancer or.laryngeal cancer  in  this group of
workers.  A twofold excess of lung cancer after an inadequate latency period  is
insufficient for establishing  a  causal relationship.   Therefore,  the weight of

- -  -««<•>                          O.T?         noarT-.nn wni nnniF DR  CITE

-------
evidence for a causal relationship between exposure  to  continuous fibrous glass
filament and respiratory cancer is Inadequate.
9.2.1.2  Mineral Wool
9.2.1.2.1  Qualitative evaluation of noncarcinogem'c effects.  An evaluation of
noncarcinogenic effects following exposure of workers to  mineral wool  indicates
that only one study (Enter!ine et a!., 1983} detected an  excess of NMRD  in male
workers employed  in  a single plant (SMR = 221, p <0.05, Power = 30 percent).
This excess  appeared  unchanged after  20 or more years  of latency (SMR = 219.9,
p <0.05,  Power  = 25 percent)  and did not persist  in  the  1986  study after
additional years  of  followup  (Enterline  et al., 1986).   A twofold excess
limited to one  study,  which did not  increase  after 20 years  of  latency, is
insufficient for  establishing  a causal relationship.   Therefore, the weight of
evidence is inadequate for a causal relationship between  mineral wool  and NMRD.
     Therefore, no association appears to exist between noncarcinogenic  effects
and exposure to glass microfibers, fibrous glass wool,  continuous  fibrous glass
filament, or mineral  wool.   There are numerous reasons why an association was
not detected.   First, there may  be  no such relationship.   Second, exposure
levels may have  been  too low to induce a response.   Third, a  relationship may
have been  undetected because  of  insufficient power to  detect a  50 percent
increase in  NMRD  if  it did exist.  Based  on these  limitations, the most that
can be  concluded is  that the available studies failed  to demonstrate  NtfRD
within the limits of  statistical  sensitivity imposed by  their respective study
designs.   The absence  of a dose-response relationship or  increased mortality
from NMRD does not establish the safety of the individual fiber types  studied.
9.2.1.2.2.  Characterization of carcinogenic effects.  An  evaluation  of carci-
nogenic effects following exposure of workers to mineral  wool  indicates  signif-
icant excesses  from  respiratory cancer including lung cancer; from digestive
tract cancer including  digestive  organ and peritoneal  cancer, digestive organ
cancer,  stomach  cancer,  intestinal  cancer,  and pancreatic cancer; and from
buccal cavity and pharyngeal cancer.   In the United States, Enterline and Marsh
(1982)  found a  2.5-fold excess  of  respiratory cancer  (p <0.05,  Power =
12 percent) among male  production and maintenance workers  in one  mineral wool
plant out of eight MMMF plants with less than 20 years  of latency.   After 20  or
more years  of latency,  this  excess  increased to ninefold  (p <0.05,  Power =
4 percent).  Enterline  et al.  (1983) found that the respiratory cancer  excess
was no longer  limited to Plant 7, but included all  mineral wool workers (SMR  =

July 1988                          9-13         DRAFT—DO NOT QUOTE OR CITE

-------
160.1, p <0.01,  Power  = 64 percent).   As in the previous study, the greatest
excess was found in Plant 7 (SMR = 451.3, p <0.01,  Power « 14  percent) and also
resulted  in  a ninefold  excess  after  20 or more years  of  latency  (p <0.01,
Power = 3  percent).   A  46  percent  excess of  respiratory cancer among  all
mineral wool  workers  after 20 or more years of latency found by Enter!ine et
al.  (1986) was  no  longer significant  when compared to local mortality rates  in
a  later analysis (Enter!ine et al.,  1987).   However,  as  in the previous studies
by Enterline  and colleagues,  there  was insufficient  power to  detect a 50  per-
cent  increased  risk of  death from respiratory cancer  among mineral wool
workers with  20  or more years of latency.   Evidence  of a respiratory cancer
hazard following mineral wool exposure was strengthened when  Enterline et al.
(1987) found  a  significant excess  of respiratory  cancer among mineral  wool
workers (SMR  =  148.4,  p <0.01,  Power  = 89 percent) even when  compared to  local
mortality  rates  (SMR  ~ 133.6, p <0.05, Power « 91 percent).  This  shows  that
with adequate power,  respiratory cancer is significantly elevated compared  to
background rates,  thereby taking into account etiologies! agents  such   as
smoking and local  mortality differences that could influence  the magnitude  of
the  risk as well as social, economic, ethnic,  and cultural factors related  to
disease.   The strength of the association despite low  power,  i.e., SMRs  that
increased  with  increasing latency, and the significant  findings  among  all
mineral wool  workers  compared  to both national and  local  mortality  rates,
suggest that a causal  relationship existed between mineral wool and respiratory
cancer.  However,  one  cannot  completely rule out other  etiological  agents that
might  be  capable  of  inducing similar  effects.  Therefore,  the weight  cf
evidence for  a  causal  relationship  between  mineral wool and respiratory cancer
is limited at the  present time.
     In Europe,  despite low power,  Andersen and Langmark (1986)  found a three-
fold increase in lung cancer  incidence compared to  national  rates  among  male
mineral wool  workers  in a rock wool  plant,  which became more pronounced  when
short-term workers,  i.e.,  those employed for less than 1 year,  were excluded
(SIR = 386,   p <0.01).   They  also  found  a  twofold  increase in  lung cancer
incidence  (p  <0.05,  Power = 21 percent) among those  employed for 1 or more
years  in one  of  three mineral wool plants  after 20  or  more years  of latency.
Westerholm and  Bolander  (1986)  found a ninefold excess  (p <0.005) of  lung
cancer compared  to national  rates in  male  and  female mineral  woo!  production
workers with  1  to  4 years of  employment and after 30 or more  years  of latency,

July 1988                          9-14         DRAFT-DO NOT QUOTE OR  CITE

-------
even though there  was  insufficient power to detect lung cancer (Power = 4 per-
cent).   Olsen  and Jensen  (1984)  found a  twofold  increase in lung  cancer
incidence among male and  female production workers after 20 or more years of
latency compared to  national  rates (p <0.05, Power = 20 percent).  This excess
was concentrated among  male  production workers  (SIR = 195, p <0.05).  Claude
and Frentzel-Beytne (1986)  found a threefold excess among male production and
maintenance workers who were  hired before  the introduction of resin binders in
1953 and who had 10 to 19 years of latency compared to national rates (p  <0.05,
Power = 12  percent).   Pooling data from the European studies,  Simonato et al.
(1986a) found  an excess  of  lung cancer  among  mineral wool workers  first
employed during the  early technological  phase  of MMMF production compared to
both national (SMR =214, p <0.05, Power * 14 percent) or local mortality rates
(SMR * 257, p  <0.05,  Power = 24 percent).   This excess was even  greater  after
20  or  more years  since first  exposure  to mineral  wool  during  the early
technological phase (SMR = 295, p <0.05, Power = 8 percent).   Subsequent  studies
by  Simonato et al. (1986b,c,  1987) confirmed these excesses even when  compared
to  local rates.   Combining U.S.  and  European mineral  wool  workers, Doll  (1987)
found that  compared  to  local  mortality rates,  the  lung .cancer excess  was
greatest among mineral  wool workers with 30 or more years of latency (SMR = 148,
p <0.05, Power = 74 percent).   Despite low power, the following factors suggest
that a  causal  relationship existed between exposure  to  mineral wool and lung
cancer:   the strength  of  the association, ranging from  two- to ninefold;  the
significant excesses in nine  studies,  even when compared  to  local  mortality
rates;  the  appearance  of  the excesses during the early technological phase of
MMMF production;  and the concentration of excesses  among  production workers
employed for at  least  1 year.   Therefore, similar to  respiratory cancer,  the
weight of  evidence for a causal relationship between exposure to mineral wool
and lung cancer is limited at the present time.
     Investigators found  numerous  excesses from digestive  tract cancer follow-
ing exposure to  mineral wool.   In the United States, Robinson et al.  (1982)
found a 3.3-fold  increase in digestive system  cancer among male mineral  wool
production  workers with 20  or more years of  employment  (p <0.05,  Power  =
8 percent).  In  Europe, Olsen and Jensen  (1984) found  a 52 percent  incidence
of  digestive organ and peritoneal cancer  among  male and female  mineral  wool
production workers during the first 10 years after initial employment (p <0.05,
Power =  53 percent).   In a  followup  study, Olsen  et al. (1986) found a

.iniu TQftfi                          Q-l*;         neari-.nn  WOT nunrc np TTTF

-------
50 percent increased  incidence  of  digestive organ cancer among male production
workers (p <0.05,  Power -  90 percent).  Claude and Frentzel-Beyme (1984) found
an excess of  stomach  cancer among  male production workers after 20 to 29 years
of latency (SMR  = 273, p <0.05, Power = 14 percent).   In  a followup study,
Claude and Frentzel-Beyme  (1986)  confirmed their various findings of  a  cancer
excess and further found  that  it was concentrated among  male production and
maintenance workers who had been hired before  the introduction  of resin  binders
in 1953 and who  had 20 to  29 years  of latency (SMR = 261, p <0.05, Power =
14 percent).   Using internal controls as  a reference  group,  a sevenfold  relative
risk  of  pancreatic cancer was  found.  Andersen  and  Langmark  (1986) found a
twofold increase  in  intestinal  cancer  among male mineral wool workers employed
at least  1 year  (p <0.05,  Power =  28 percent).  Based on excesses of digestive
system cancer among male  and female  mineral wool production workers from three
studies,  two  of  which showed  excesses ranging from 33  to  50 percent  in male
production workers alone,  a causal  relationship appears  to  exist  between
digestive system  cancer and exposure to mineral wool.  The excesses of  stomach
cancer found  in  two studies of  male  production workers with sufficient  latency
and employed  before the  introduction of  resin binders also support a causal
relationship between  stomach cancer  and  exposure to mineral wool,  but  these
excesses were limited to one factory.  Excesses from  pancreatic cancer and  frorr,
intestinal cancer  were each limited  to one study and are probably not causally
related to  mineral wool  exposure.   Therefore, the weight  of  evidence  for a
causal relationship is limited  for digestive  system  cancer and stomach  cancer
and inadequate for pancreatic  cancer and  for intestinal  cancer.
     In Europe,  Olsen et  al.  (1986) found an  increased incidence from  buccal
cavity and pharyngeal cancer among male production workers  (SIR = 210, p <0.05,
Power = 24  percent).   The low  power to  detect buccal  cavity and pharyngeal
cancer in many  subcategories,  which were  found in several  of  the  individual
European  studies,  remained even when the data from the  13  plants were pooled.
For example, despite  low power, Simonato  et al.  (1986b,c, 1987) found  a  twofold
increased incidence of buccal  cavity and  pharyngeal cancer  in rock wool  workers
(p <0.05),  which  increased with  increasing latency.  However, the twofold
excesses,  even  though detected  In multiple studies, are  insufficient  for
establishing a causal  relationship.   Therefore,  the  weight of  evidence for a
causal relationship between exposure to  mineral  wool  and buccal cavity  cancer
is limited.

July 1988                          9-16         DRAFT—DO NOT  QUOTE  OR CITE

-------
9.2.1.3  Ceramic Fibers.  No  information  was  found on the health effects of
ceramic fibers.
9.2.1.4  Quantification of Carcinogenic Effects of Manmade Mineral Fibers.  The
minimum requirements for an epidemiclogical study to qualify  for a quantitative
risk assessment  include  the measurement of risk  (SMR, PMR, SIR, RR) and corre-
sponding exposure levels, explicit or implicit, with length of exposure  or some
measure of  duration  of employment.   The study also  should show statistically
significant risk for the exposed compared to  reference population, and more
weight  should be  placed on  the  study  with,  larger  expected values  and
person-years if other factors are equivalent.
     The data  describing the  exposure  levels of  fibrous glass wool, continuous
glass wool filaments, and mineral wool  are based on much speculation concerning
past industry conditions (see also Chapter 3).   Accurate values for  the  concen-
tration of  airborne  fibers  to which workers have  been  exposed are  available
only since  the  late  1970s.   However, it is widely believed that the levels  in
earlier years  were  higher than  the  current levels.   Cherrie  and Dodgson (1985)
identified the two most  important factors  contributing  to release of MMMF into
the  air  as  nominal  fiber size  and  the addition of oil to MMMF wool.   They
estimated that both  factors  could alter the airborne concentrations at least
tenfold.   Cherrie et al.  (1986) reported past exposures to airborne fibers  by
conducting a historical environmental investigation in European MMMF factories.
The purpose of  the  investigation was to assess past exposures to MMMF  as  well
as  other  environmental  risk  factors.   They concluded  that  the production
history of  each  plant could be classified  into  two or  three distinct techno-
logical phases.  In  the  early technological phase,  batches were processed with
no  oil added  to  the  fibers.   The later technology included more modern produc-
tion techniques, and oil was  added  to  the  fibers to prevent  them from becoming
airborne.  An  intermediate  phase, in which a mixture of production  techniques
was utilized,  was  identified  at some plants.  It is assumed that during  the
early  phase,  airborne fiber  concentrations  in the mineral wool plants were
greater than  current concentrations.   Cherrie  et al.  (1986)  estimated that the
airborne concentrations  existing during  the  early technological phase were
greater than 10 times the concentrations measured during the  late technological
phase.   For glass wool plants, a thicker-diameter fiber was manufactured during
the early technological  phase.   It is assumed that this  change in  fiber size
would put a greater  number  of  smaller fibers  into the  air and perhaps  cancel

-------
any  effect  of lowering the  concentration  with  the addition of oil.  Hence,
advances  in  technology may not have changed the airborne concentrations at the
glass wool plants.   Cherrie et al.  (1986)  conclude  that it was reasonable to
separate  each plant history into an early, intermediate, and late  technological
phase, but they give no accurate estimate of past exposure  concentrations.
     Cherrie et  al.  (1987) describe a simulation of a mineral wool  production
plant operating  in  the 1940s to estimate the airborne fiber levels  before  oil
was added to the process to minimize airborne fiber levels.  They  conclude  that
the addition  of  the oil reduced the airborne fiber concentration  by three- to
ninefold  in different  scenarios.
     Dodgson et  al.  (1987) derive a model  to estimate  past exposure  at mineral
wool and  glass wool plants.  The model predicts mean airborne concentrations at
mineral wool plants of between 1 to 2 fibers/cm  and up to  10 fibers/cm  during
the early technological phase.  Estimates for glass wool plants remain at about
0.1 fiber/cm  for the  early and late technological phases.
     About the  most that  can be  gathered from these  data is a qualitative
assessment that  levels during the early technological  phase  were higher than
they are  today,  perhaps by one or two  orders of magnitude.  Oodgson  et al.
(1987) caution  that potential errors exist in  any estimation that  has  been
performed to measure  past exposure to MMMF.  Methods of gathering data (i.e.,
self-administered questionnaire)  and variability in various  factors  from plant.
to plant, the  extent  of which cannot be quantified, can exert great influence
on  any  estimates.   Doll  (1987) concluded  that  the uncertainty in the fiber
counts during  the  early technological phase and the extent of the contribution
of other  carcinogens  make it impossible to provide precise quantitative  esti-
mates of  the effects of exposure to MMMF.
     The  epidemiologic data likewise provide an inconsistent amount of evidence
to perform a  quantitative risk assessment.  There is no doubt from a qualita-
tive standpoint that exposure to MMMF 1s limited to various respiratory cancers
as can be seen in the conclusions reached In the previous  sections.  However,
the ability to quantify the human cancer risk for a given exposure level  is not
that straightforward.   The available epidemiologic data are overall an incon-
sistent set of data upon which to base quantification of human risk.
     Enterline et  al.  (1987) describe an investigation of workers employed at
17 U.S. manufacturing  plants.  This study updates earlier reports of the same
cohort (Enterline and  Marsh,  1982; Enterline et  al.,  1983,  1987).   They con-
cluded that  analyses  with various surrogates of fiber exposure  did  not support
July 1988                          9-18         DRAFT-DO  NOT QUOTE OR CITE

-------
a  dose-response relationship between  exposure and  respiratory  cancer  for
mineral wool workers,  fibrous glass wool  workers,  or continuous  fibrous glass
filament workers.  While  SMRs did increase with time since  first exposure  for
fibrous glass wool workers (though none were significantly increased), this did
not hold  for respiratory  cancer and duration of exposure for any of  the three
fibers.  Cumulative exposures by  fiber did not exhibit  any  increasing dose-
response  relationship,  and measurements  of fiber  concentrations during the
early technological phase were not available.
     Shannon et  al.  (1986) extended the followup of previously  reported data
(Shannon et al., 1982, 1984) from workers  in Ontario, Canada.  Analysis  of  lung
cancer deaths by duration of exposure and  time since first exposure yielded too
few deaths and low statistical power for meaningful evaluation.
     Bertazzi et al.  (1986)  found a fourfold excess of  laryngeal  cancer in a
factory in  Italy.  As  only four cases of  laryngeal cancer were described,  too
few cases were  available  to  adequately establish a  dose-response relationship.
However, three  of  the  four cases were observed in the highest exposure  group,
and a  significant increase was observed  in those  with 25 or more years  of
latency.
     Simonato et al.  (1987) extended a study of 13 MMMF plants in Europe.   This
cohort had  been  studied previously by  Saracci  et al.  (1984); the most convinc-
ing evidence of  a  causal  association would establish  a  relationship between
increasing lung cancer risk and increasing dose of MMMF.   No such dose-response
relationship was observed using length of employment as a surrogate  for dose.
The authors  do  qualify the use of length of exposure instead of dose by indi-
cating that  improvement in worker exposure may make this factor unreliable as
an estimate of cumulative dose.   No dose-response relationship was observed for
mineral wool or fibrous glass workers by duration of employment.
     Dose-response relationships  were observed  for mineral wool  workers  by
years since first exposure and also by technological phase, but these relation-
ships did not  hold for fibrous glass  wool  workers.   However, the effects  of
other environmental  conditions, such  as  the use of arsenic-containing slag
during the  early technological  phase,  poor ventilation, and exposure to PAHs,
could not be excluded.
     From these  data,  it  appears  that  the evidence for respiratory cancers  due
to exposure  to  MMMFs  is inconsistent.   It is not unreasonable to conclude  that
conditions during  the  early years of the MMMF industry have contributed to an
observed excess  of cancer  risk.   The  effects  of  exposure  control  methods
Julv 1988                          9-19         DRAFT—DO NOT QUOTE OR CITE

-------
employed  in  the  last 20 years may not have  been manifested as of yet.  It also
seems reasonable  to  conclude that exposure to mineral  wools  is  a  greater  risk
than exposure to fibrous glass wool, which carries a greater  risk  than continu-
ous glass filaments, with the latter being almost risk-free.   While qualitative
evidence  for cancer  has been demonstrated, consistent  dose-response relation-
ships necessary  for  quantification of risk have not been demonstrated in any
cohort investigated.
     Because of these inconsistent dose-response relationships,  the uncertainty
in exposure  concentrations  in the early part of the industry and  the unknown
effects of exposure  to other environmental  carcinogens make the quantification
of human risk inappropriate at this time.

9.2.2  Animal Toxicity
9.2.2.1  Characterization of  Noncarcinogem'c Effects.  Several  chronic inhala-
tion toxicity studies conducted with rats, guinea pigs, and  hamsters exposed to
MMMF (glass  fibers,  mineral  wool, and ceramic  fibers)  of varying diameters  and
lengths generally indicated some pulmonary damage,  but  no carcinogenic response.
These studies  were  reviewed  in Chapter 5.   The  severity.of the  pulmonary
response is  influenced  by  fiber dimensions. Thin  (<3 um), long  (>20  urn) fibers
elicit a  more severe response than do  either  thick (>5 um),  long fibers or
short (<5 um)  fibers of any diameter.  However, unlike asbestos  fibers  of
similar dimensions  that have  been shown to  be  toxic  and fibrogenic after
introduction  into  the  lung,  manmade mineral fibers generally do not elicit a
pulmonary response more severe than macrophage aggregation and alveolar pro-
teinosis following inhalation exposure or intratracheal instillation.  This  may
be due in  part  to the  proportion of nonrespirable-to-respirable fibers within
the aerosol  and  the  number  of respirable  fibers actually deposited in the lung
and alveoli.  Fibrous  glass  and mineral wools have been shown to  be cytotoxic
by several investigators,  although very limited evidence suggests  that ceramic
fibers are not cytotoxic.  Again, thin, long fibers proved to be more cytotoxic
than either  short fibers of  any diameter  or thick and  long fibers.   Therefore,
since manmade  mineral   fibers  have been  shown to  be cytotoxic in ni vitro
assays, questions  remain as  to why these fibers are not as toxic  as asbestos
following inhalation.   One possible explanation may be a lack of durability of
vitreous fibers  as  compared  to that of asbestos.   Following introduction  into
the lungs, glass fibers have been shown to erode,  becoming coarse  in appearance

I..1).. idee                          9-20         DRAFT—DO NOT QUOTE OR CITE

-------
after 12  to  24 months.   Leaching of components has also been shown to occur,
resulting in weakened  fibers  that become fragmented,  engulfed by macrophages,
and cleared  from  the lung  or  dissolved,  whereas asbestos (and erionite) fibers
are relatively insoluble and  are not readily cleared from the lungs.  Further-
more, the dissolution of  asbestos  fibers generally results in longitudinal
splitting, producing thinner fibers.  In addition,  inhaled  vitreous  fibers  have
not been  shown to migrate  to the pleura!  surfaces  of the lung or the tissues
other than the tracheobronchial  lymph  nodes.  Asbestos, on  the other hand, has
been shown to  readily  penetrate epithelial surfaces to the lung interstitium
and migrate  to pleura!  surfaces, thus  interacting  with  different target cells
than manmade mineral fibers and eliciting different pulmonary  responses  includ-
ing tumorigenesis.
     Results of the  inhalation  studies detailed in  Chapter  5 indicate that the
mechanism of toxicity  for  manmade mineral  fibers may not be comparable  to that
of asbestos, particularly  on  a dimensional basis alone.  Other fiber charac-
teristics, e.g.,  durability and migration within tissue, must also be consid-
ered.   However, to  completely assess  the toxicity  of manmade mineral  fibers
following inhalation exposure,  more testing of fibers with  similar  dimensions,
using a range of exposure levels, must be conducted. .
9.2.2.2  Quantification of Noncarcinogenic Effects
9.2.2.2.1  Fibrous glass.   Several  chronic inhalation  toxicity studies were
considered for  the  derivation of the  Reference Dose  (RfD)  for  fibrous  glass
(Table 9-2).    No  more  than two  exposure levels of  a specific fiber  were tested
and a dose-response  relationship was  not demonstrated in  any  one  study.  How-
ever, when these  studies  are  considered collectively,  it  is  apparent  that a
range of  exposure levels  using  fibrous glass  samples of similar dimensions
within the respirable range (<3.5 urn in diameter) were used.  Conse.quently,  the
results of these studies can be considered together for calculation of the RfD.
     The  pulmonary  lesions  resulting from fibrous  glass exposure in each study
have been scored  by the  reviewers using the grading system  of McConnell  et al.
(1984) described  previously (see Section 5.2.1.1.2., Table  5-2).  According  to
this grading system, a lesion score of 1.0 to 2.9 signifies either no pulmonary
tissue response or  scattered  macrophage accumulation in the alveoli.   Lesion
scores of 3.0  to  3.9 indicate some structural  change in the lung  architecture
such  as   development  of  cuboidal epithelium,  reticulin fiber  networks at
bronchoalveolar bifurcations,  and scattered focal peribronchial  fibrosis with

July 1988                          9-21         DRAFT--DO NOT QUOTE OR CITE

-------
TABLE 9-2.   SUMMARY OF LONG-TERM INHALATION STUDIES CONSIDERED FOR DERIVATION OF THE RfD FOR FIBROUS GLASS
Exposure Mean
level diameter
Species (mg/m3) (pm)
Rat/hamster 0.3 0.45
3.0
Rat/hamster 100 0.5

Rat 5 <3.5
15
Rat/hamster 10 3.1
Rat/hamster 1.2 5.4
12.0

Rat 15 4.6
Rat/hamster 9 6.1
Length
(pm) Pulmonary effects
<20 Dose- related increases in
macrophage aggregation.
Slight increases in bron-
choalveolar metaplasia in
hamsters at 3.0 mg/m-1.
5-20 Increased alveolar protein-
osis; sparse collagenous
fibrnsis in lung, pleura,
and lymph nodes of rats and
hamsters; and laminated,
calcific microliths in
hamsters.
>10 Severe, plaguelike pleural
lesions, mild granulomatous
inflammation, and macrophage
aggregation.
77% Increase in macrophage
>10 aggregation that was more
severe in rats than hamsters.
5-50 Increase in macrophage
aggregation that was more
severe in rats than hamsters.
Slight increase in broncho-
alveolar metaplasia in
hamsters exposed to 1.2 but
not to 12.0 mg/m-1.
>20 No pulmonary response.
50% Increase in macrophage
>80 aggregation.
Lesion
score References
2.1 Smith et al.
2.3 (1984. 1986,
1987)a
3.7-4.0 Gross et al.
(1970)°

3.2 Mitchell et al.
3.5 (1986)c
2.1-2.3 Smith et al.
(1984, 1986,
1987)
1.8-2.4 Smith et al.
1.6-2.1 (1984, 1986,
1987)

1 Mitchell et al.
(1986)
1.9-2.3 Smith et al.
(1984, 19R6,
1987)
exposed for 6 hours/day, 5 days/week for 24 m
                                                          and observed for life.

-------
no collagen  formation,  in  addition to moderate to severe dust-laden macrophage
aggregation.   Scores  of 4.0 or higher denote the  development  of minimal  to
severe collagenous  fibrosis,  blockage of  airways,  and destruction of  lung
architecture.
     Considering only glass microfibers  (diameters of 0.45 to 0.5 urn), a NOAEL
of 3.0 mg/m  was identified from the studies of Smith et. al.  (1984, 1986, 1987)
in which  lesion  scores  of  2.1 to 2.3 we're obtained in both  rats  and hamsters.
A LOAEL was identified from the study by  Gross et al.  (1970), in  which rats  and
hamsters were exposed to fibrous  glass  (0.5 urn  in diameter and 5 to 20 pm in
length) at  a total  respirable  concentration of  100 mg/m .   In  this  study,
exposed rats and  hamsters  developed  sparse focal collagenous fibrosis in the
lung,  pleura,  and lymph nodes;  in addition, hamsters  developed calcified,
laminated microliths.

     The RfD for a 70-kg adult is calculated as  follows:

                        RfD - <*•*> • 0.03
where:
     3.0 mg/m3 - NOAEL, based on  data  from the studies by Smith et al.  (19S4,
                 1985, 1987).
          100 =  uncertainty factor, chosen  in accordance with EPA guidelines
                 for use with a NOAEL from an animal study.

     For fibers  with  diameters  >0.5 but <3.5 urn, two  studies  were  considered
for derivation  of  a second RfD.  A  LOAEL  of 5 mg/m  was identified from the
study of Mitchell  et  al.  (1986) in  which a lesion score  of  3.2 was. obtained  in
rats.   In  the second  study (Smith et al.,  1986), lesion scores of  2.1 to 2.3
were obtained in rats and hamsters,  and a NOAEL of 10 mg/m  was identified.   In
this  study,  however,  the dust  cloud had a respirable mass fraction of  only
30 percent and contained a high percentage of nonfibrous particles.   Therefore,
the study  of Mitchell  et al.  (1986) was chosen for  derivation  of the RfD value
for fibrous  glass  having diameters  >0.5 but <3 urn.   The RfD for a 70-kg adult
is calculated as follows:
                  RfD =   (i0m   s 0-005 n>g/m3 (5.0 pg/m')

July 1988                          9-23         DRAFT—DO NOT QUOTE OR CITE

-------
where:
     5.0 mg/m3 « LOAEL, based on  the  development  of severe  plaque! ike  lesions
                 and  moderate  dust-laden macrophage  aggregation  in alveoli
                 and  mild granulomatous  inflammation In  rats exposed  to
                 fibrous  glass with a diameter of <3.5 and length  of  >10 urn
                 (Mitchell et al., 1986).
          1000 = uncertainty factor,  chosen  in accordance with EPA guidelines
                 for use with a LOAEL from an animal study.

     An RfD  value  was not calculated for glass fibers with diameters  >3 um,
since these  fibers are not considered to be  respirable.  The  only pulmonary
response observed  was  slight macrophage  aggregation,  and lesion scores of 1 to
2.3 were obtained  in rats and hamsters (see Table 9-2).
9.2.2.2.2  Mineral wool
     9.2.2.2.2.1   Rock wool.  Two long-term  inhalation studies were considered
for derivation  of  the RfD for rock wool  (Table 9-3).   As  with glass  fibers,
only one exposure  level of rock wool was tested, and since  the  fibers  tested in
each study  had  approximately the  same dimensions,  they were considered collec-
                                                                  3
tively.   The  data  indicate a dose response with a  NOAEL of  5 mg/m  (respirable
concentration),  identified from  the  study  of Le Bouffant  et al.   (1986).   A
LOAEL of 10  mg/m  was identified  from the study of Wagner et al.  (1984), based
on a slight cellular change in lung architecture.   The RfD  for  a 70-kg adult is
calculated as follows:
                   Rf0 =   .gm   = 0 05 mg/m3 (50
where:
     5.0 mg/m3 = NOAEL,  based  on data  from the study by Le Bouffant  et  al.
                 (1986).
           100 = uncertainty factor,  chosen in accordance with EPA guidelines
                 for use with a NOAEL from an animal study.

     9.2.2.2.2.2  Slag wool.  Only one long-term inhalation study was found for
use in  the  derivation of the RfD for slag wool.  In the study by Smith et al.
(1986,  1987),  a LOAEL  of 10 mg/m   was  identified, based on  a  significant
decrease in the lifespan and a slight increase (not statistically significant)
July 1988                          9-24         DRAFT—DO NOT QUOTE OR CITE

-------
               TABLE 9-3.  SUMMARY OF LONG-TERM INHALATION STUDIES CONSIDERED FOR DERIVATION OF
                                             THE RfO FOR ROCK WOOL


Species
Rat


Rat


Respirable
fiber con-
centration
(mg/m3)
10


5


Mean
diameter
(Mm)
87% <2


49% <2



Length
dim)
>5


88% >5




Pulmonary effects
Mild to moderate dust- laden
macrophage aggregation and
some nyperplasia.
Mild macrophage aggregation
and chronic inflammation of
nasal fossae.

Lesion
score References
2.7-3.3 Wagner et al.
(1984)a

2.3 Le Bouffant
et al. (1986b)

aAnima1s were exposed  for 7 hours/day, 5 days/week for 12 months and observed for life.

 Exposure was for 5 hours/day, 5 days/week for 12 or 24 months.  Animals exposed for 12 months were
 observed up to 16 months postexposure, and animals exposed for 24 months were observed up to 4 months
 postexposure.

-------
in pulmonary  fibrosis  in rats  exposed  to  slag wool.  The RfD for a 70-kg adult
is calculated as follows:
                   RfD =
where:
     10 mg/m3 =  LOAEL, based on a significant decrease in lifespan and slight,
                 nonsignificant increase in pulmonary  fibrosis  in rats (Smith
                 et al., 1986, 1987).
         1000 =  uncertainty factor, chosen in  accordance with EPA guidelines
                 for use with a LOAEL from an  animal study.

9.2.2.2.3  Ceramic fibers.  Three  studies  were  considered for calculation of
the RfD  value  for  ceramic fibers  (Pigott  et  al.,  1981;  Davis et  al.,  1984;
Smith  et al.,  1987).   The data and  lesion scores  derived using  the  grading
system of  McConnell  et  al.  (1984)  are  summarized in  Table 9-4.  Only  one
exposure level was  tested in each of the  chronic  studies considered, and no
other  suitable  studies were  found.   Therefore, the results  from all three
studies were used to assess to LOAEL and NOAEL for ceramic fibers.
     Since no effects  were observed  at  either fiber concentration in  the  stud.>
by  Pigott  et al.  (1981),  the higher fiber concentration  of  2.45 mg/rr,  was
chosen as  the  NOAEL  for ceramic fibers.   The results  of the study by  Davis
et al.   (1984)  indicated  a treatment-related effect  at 8.4 mg/m .   This
concentration could  be considered the  LOAEL  since no  other effects  such as
mortality  were noted  in the exposed animals.    The  RfD for a 70- kg adult is
calculated as follows:
                  RfD =  -(lo     = °-024 m9/n>3 (24.5 ug/m3)

where:
     2.45 mg/m  = NOAEL, based  on the absence of  increases  in  mortality and
                  pulmonary effects  in  rats  exposed to ceramic fibers (Pigott
                  et al., 1981).
            100 = uncertainty factor, chosen  in accordance with EPA  guidelines
                  for use with a NOAEL from an animal study.
July 1988                          9-26         DRAFT--DO NOT QUOTE OR CITE

-------
                    TABLE 9-4.  SUMMARY OF LONG-TERM INHALATION STUDIES CONSIDERED FOR DERIVATION OF
                                               THE RfO FOR CERAMIC FIBERS
Species
Rat
Rat
Rat/hamster
Respirable
fiber con- Mean
centration diameter
(mg/m3) (pm)
2.18-2.45 3
8.4 0.3
12 1.8
Length Lesion
(pm) Pulmonary effects score
10.5-62 Aggregation of dust- laden 2.3
macrophages. No fibres is
present.
90% <3 large areas of alveolar 3.2
proteinosis and dust- laden
macrophage aggregation.
Ferruginous bodies were
present.
35 Moderate dust- laden macrophage 2.1-3.0
aggregation and a slightly
increased incidence of fibrosis
were noted in rats. One
mesothelioma was found in
70 hamsters.
References
Pigott et al.
(1981)a
Davis et al.
(1984)°
Smith et al.
(1987)c
^Animals were exposed to refractory Saffilft fibers for 6 hours/day, 5 days/week for 86 weeks.
 Animals were exposed to aluminum silicate fibers for 7 hours/day, 5 days/week for 12 months and observed for
 ?0 additional months.
cExposure to Fiberfraxfc was for 6 hours/day, 5 days/week for 24 months, and the animals were observed for life.

-------
9.2.2.3  Characterization of Carcinogenic Effects
     Several chronic  inhalation  toxicity studies  with rats,  guinea pigs, and
hamsters exposed to MMMF of varying diameters and lengths indicate no carcino-
genic response.  These  studies  were  reviewed in Chapter 7.   The inability of
glass fibers to elicit a carcinogenic  response following inhalation exposure  to
animals  is  unclear.   However,  it  nay be  Influenced  by the proportion  of
nonrespirable to respirable fibers and the actual  number of  respirable fibers
retained in  the  lungs and alveoli.   In  addition, the biodisposition of these
fibers  from the  lung, as previously described for  noncarcinogenic responses
(see Section 9.2.2.1), may also influence the effects  noted.
     In  contrast,  either intratracheal  or  intraperitoneal administration  of
MMMF into several  animal species was  carcinogenic.  The  severity of the  carci-
nogenic  responses,  like that of the  noncarcinogenic  responses  characterized
earlier, was found  to be influenced  by  fiber dimensions.  Thin (<3 uir,), long
(>20 urn) fibers were  the most potent  carcinogens,  and  thick (>5 urn), long or
short (<5 urn)  fibers  were less  carcinogenic.  Stanton (1973) hypothesized that
durable  fibers, of  which asbestos  is  one example., cause  cancer  irrespective  of
their physicochemical  nature simply  because they are fibers,  provided  they
subscribe to well-defined ranges of  diameters and lengths.   Stanton1s  studies
(Stanton et al., 1977,  1981) indicated that in addition  to asbestos, several
other natural  and  synthetic  fibers  caused  malignant tumors  in rats after
intrapleural implantation,  and  found  a good correlation  between tumorigenicity
and a fiber length of >8 urn and a diameter <0.25 urn.   However,  the investiga-
tors did not find  a sharp demarcation line in fiber size, and  carcinogem'city
probably correlated with fibers of a  size that was  able  to escape  phagocytosis
and retain the ability to penetrate into and migrate in  tissues.
     The "Stanton  hypothesis"  has  been tested, reanalyzed, and  criticized, but
it has  not  been  discarded or disproven,  and many  but  not all difficulties have
been resolved.
     Bertrand  and  Pezerat  (1980) reanalyzed statistically some of Stanton's
data and concluded  from an analysis  of  the  length/diameter variables (aspect
ratio)  that carcinogenesis  is  the  result  of  a  continuous  function of  the
aspect ratios and that the effects of the two variables  cannot  be  separated.
     Wagner  et  al.  (1969)  were  the first to indicate a correlation between
fiber size  and carcinogenic potential in their studies  on amphibole  asbestos.
For a given weight of asbestos  injected  intrapleurally  in  rats and  assuming

July 1988                          9-28         DRAFT-DO NOT QUOTE OR  CITE

-------
that all  fibers  had approximately the  same length of 10 to 20 pm, the carcino-
genic  potential,  correlated with the  thinness of fibers,  was  as follows:
crocidolite  (riebeckite)>amosite>anthophyllite.  Milling  of fibers decreased
their  length  and  also decreased their carcinogenicity.  With  three samples of
tremolite, Wagner  et al.  (1982) found that the Incidence  of mesotheliomas was
14/47  in  Sprague-Dawley  rats  after  intrapleural Injection of 20 mg of a prepa-
ration  in which 56 percent of  the  fibers  were >8 pm and  had a diameter of
<1.5 pm (15.4 x  10  fibers/ug).  Two other samples of tremolite with 0.7 and
0 percent of  fibers  in this size category were inactive.   Pott  and colleagues
(1976) and Pott (1978) found that the maximum biologic activity  of glass  fibers
was among the  fibers with  diameters of about  0.25 mm and  lengths >10 pm.  More
recently,  Pott  et al. (1987a)  found that  the length and durability of  the
fibers are perhaps  of greater significance than diameter in  determining the
carcinogenic potential.  Several other studies reviewed in Chapter 7 indicate
that fiber size  is critical for explaining carcinogenicity of  natural and/or
manmade fibers, although all  investigators did not  give precise measurements,
and all results did not agree with the discrete fiber size.
9.2.2.3.1  Fibrous glass.   Several  long-term  studies  have been  conducted with
fibrous glass of varying diameters and lengths using several routes of  adminis-
tration in various  animal  species  (see Table  7-13).   No  carcinogenicity was
observed  in  at  least three animal species  following  inhalation  exposure (Smith
et al., 1984,  1986; Gross  et al.,   1970;  Goldstein  et al., 1983; Mitchell
et al., 1986), and the results for intratracheal  instillation were equivocal.
Negative results were obtained following intratracheal instillation in  rats and
hamsters  by  Smith et  al.  (1986, 1987).   Conversely,  results  of Pott et al.
(19845; 1987a) and Mohr et al. (1984) indicated that intratracheal instillation
of two  different  glass fibers resulted in mesothelial tumors  in both rats  and
hamsters.
     In contrast, fibrous glass has been shown to be carcinogenic in  laboratory
animals following  intrapleural  «nd intraperitoneal  administration.   After
intrapleural  injection,  glass  fibers were  found to be  carcinogenic in rats and
hamsters  but  not  nice.   Mesotheliomas and sarcomas  were the predominate  tumor
types  observed.   In general,  glass microfibers having  diameters  <0.5 um and
lengths >5 urn were  the  most  carcinogenic (Smith and Hubert,  1974; Wagner
et al., 1976;  Stanton et  al.,  1977;  Smith  et al.,  1980; Monchaux et al.,
1981).

-------
     Intraperitoneal injection of glass fibers has also been  shown  to  result in
carcinogenicity in  rats,  but limited evidence suggests that glass fibers are
not carcinogenic  in hamsters.   As seen following intrapleural  administration,
the primary  tumor types found were mesotheliomas  or sarcomas.  The majority of
the fibers tested were nricrofibers  having diameters <0.05 and lengths <10 um
(Pott and  Friedrichs,  1972;  Pott et al., 1976, 1987a,b; Muhle et  al., 19S6;
Smith et al., 1986).
     Thus, the weight  of evidence suggests that glass fibers having  diameters
<3 |jm and  lengths >5 urn are carcinogenic following direct administration  to
pleural and  peritoneal  surfaces.   In addition, tumor incidences  (mostly  meso-
theliomas  or sarcomas)  increase as  the fiber  diameters decrease,  with glass
microfibers  being the  most  carcinogenic.   The results of  the Pott  et  al.
(1984a,  1987a)  studies  suggest  that  intratracheal  instillation  of  glass
microfibers  results in  induction of tumors (mostly mesotheliomas) in  the  lungs.
However, no  strong  evidence  of carcinogenicity was observed  following inhala-
tion exposure.  This  may be  due to  small  total respirable fiber  concentrations
within the dust cloud resulting in small concentrations in the lungs  and/or the
biodisposition of the  fibers  following inhalation.   Due to the absence of
methodology  to estimate cancer risk from data obtained following intrapleurel
or intraperitoneal  administration and  the lack of evidence of carcinogenicity
following inhalation exposure, a quantitative analysis of  the carcinogenic  risk
for glass fibers cannot be made at the present time.
9.2.2.3.2  Mineral  wool.   Relatively  little  information on  the  carcinogenic
potential  of mineral  wool  was  found  in the  available  literature  (see
Table 7-13).   Rock  wool was  not carcinogenic  to rats  following inhalation
exposure (Wagner  et al., 1984), and no  information on  the carcinogenicity  of
rock wool  following intratracheal  instillation was found in  the  available
literature.   Only  one  carcinogenicity  study testing the effects of  rock wool
following  intrapleural  administration  was available.   In this  study,  rock wool
was found  to be  carcinogenic,  resulting in  the development of  mesotheliomas in
5/96 rats  (Wagner et al., 1984).  Rock wool was also found to be carcinogenic
in three  intraperitoneal  studies (Pott et al.,  1984a, 1987a);  the  tumors
observed were either  mesotheliomas  or sarcomas.  No conclusions  regarding  the
possible correlation between the physical characteristics and carcinogenicity
of rock wool  can be drawn  from the data obtained.
July 1988                          9-30         DRAFT—DO NOT QUOTE OR CITE

-------
     Slag wool was  not  carcinogenic in rats or hamsters following inhalation
exposure (Smith et  al.,  1986, 1987).   No information on the carcinogenicity of
slag wool following  intratracheal  or intraperitoneal administration was found
in the  available  literature.   Data on carcinogenicity  following intrapleural
administration were equivocal.   In a long-term study in rats  reported by Wagner
et al.  (1984),  no tumors were  induced  in 48 animals.   In a second intrapleural
study in rats,  however,  a  5 percent tumor  incidence of mesotheliomas or sarco-
mas was obtained  (Pott  et  al., 1984a).  Although  the tumor  incidence was  low,
the development of  mesotheliomas  and sarcomas following intrapleural adminis-
tration is  consistent with the  expected pattern of tumor development following
similar exposure  to  asbestos,  and therefore,  slag wool was  considered to be
carcinogenic in this study.  Since there is limited evidence  that rock  and slag
wools are not  carcinogenic following inhalation exposure,  and no methodology
exists for  estimation of the  carcinogenic  risk using  intrapleural or  intra-
peritoneal   data,  quantification of carcinogenic risk cannot be  performed at
this time.
9.2.2.3.3   Ceramic fibers.   Six studies  addressing the carcinogenic  potential
of ceramic  fibers  were  evaluated; the data are summarized in Table 7-13.  No
evidence of a significant increase in pulmonary  tumors was noted in  three
studies utilizing  inhalation as the route  of ceramic fiber  exposure.   Chronic
inhalation  exposure  to   refractory  ceramic fibers (Fiberfrax )  produced  no
tumors in rats  and one  mesothelioma in 70  hamsters (Smith  et al.,  1986).   The
mesothelioma was  noted  because  of its known  association with fiber exposure,
but the possibility  that the mesothelioma  was a spontaneous  event could not  be
discounted.    The  chronic  inhalation  exposure of  rats  to refractory ceramic
alumina (Saffil )  fibers in  the manufactured or thermally aged  form  did  not
result  in  any benign or malignant  pulmonary  tumors in exposed  rats (Pigott
et al., 1981).  In addition,  no significant differences were observed in  benign
or malignant  tumors  at  sites  other than the  respiratory tract.  In a chronic
inhalation  study  by  Davis  et al. (1984), eight pulmonary neoplasms consisting
of one  adenoma,  three  carcinomas, and four malignant histiocytomas were  noted
in 48  rats  exposed  to  aluminum silicate  ceramic  fibers.   An additional  six
animals exhibited  areas  suspected of being preneoplastic.  However, the impli-
cation  of  the reported  pattern of tumor  development  is unknown,  since  the
finding of  malignant histiocytomas in fiber-exposed animals is not consistent

-------
with the expected  pattern of tumor development resulting from asbestos expo-
sure.  Intratracheal administration  of  refractory ceramic fibers  (Fiberfrax^)
in  rats  and hamsters did not result  in any pulmonary tumors (Smith et al.,
1986).
     Administration of ceramic  fibers via the intraperitoneal  route yielded a
high incidence  of  mesotheliomas in rats and hamsters  (Smith et al., 1986; Pott
et al., 1987b).  However,  lower tumor Incidences  were noted in the study by
Davis et al.  (1984):   1 mesothelioma and 2 fibrosarcomas developed in  32  rets
injected with  aluminum silicate  fibers.   Surprisingly,  there  was no  strong
evidence of  carcinogenicity  in  rats for aluminum silicate and large-diameter
alumina and  zirconium oxide  fibers  administered  via the intrapleural  route
(Wagner et al., 1973; Stanton et al., 1981).
     The weight of evidence  suggests that  ceramic  fibers are  carcinogenic
following intraperitoneal administration, although no carcinogenic response was
noted in two strains  of  rats after intrapleural injection.  As stated  earlier,
no  methodology  exists for calculation  of carcinogenic risk  using data from
intraperitoneal studies.   In addition,  strong evidence indicates  that  ceramic
fibers  are  not  carcinogenic  following inhalation  exposure in two  animal  spe-
cies.  Therefore,  a carcinogenic  risk  estimate for  ceramic fibers was  net
calculated.

9.2.3  Characterization of Carcinogenic  Potential  Based on Human and Animal
       Data
     The weight of evidence from  both  human health effects  data  and  animal
toxicity data, based on the information described  in the previous  sections (see
also summary in Section 9.5), suggest the following carcinogenic categories for
the  MMMF:   glass microfibers, B2;  fibrous glass,  Bl;  mineral wool,  Bl; ceramic
fibers, C.
9.3  SYNTHETIC FIBERS
9.3.1  Human Health Effects
     No epidemiological studies were found for carbon and polyolefin fibers.

-------
9.3.2  Animal Toxicity
9.3.2.1  Characterization of Noncarcinogenic Effects.   The available informa-
tion is too  limited  to enable an evaluation of  the  effect of fiber dimensions
on pulmonary  response similar to  that determined  for  MMMF.   However,  the
combined evidence from  subchronic and chronic toxicity, as well as cytotoxici-
ty, studies  Indicate  that  respirable  aramid fibers  are nuch more toxic than
carbon or polyolefin fibers.   Aramid fibers  have been  shown to  be  cytotoxic and
to cause  fibrosis,  obliterative bronchiolitis  with collagen  formation, and
emphysema as  well  as lung carcinomas in  rats following repeated  or chronic
inhalation exposure.   Inhalation of  carbon fibers resulted  in macrophage
accumulation  but no  pulmonary damage in exposed rats,  whereas  intratracheal
administration of polypropylene  to  rats resulted in pulmonary granulomas only.
No pulmonary  tumors  were noted following  administration of either fiber.  Both
carbon  and  polyolefin  fibers exhibited  minimal  cytotoxic  and  hemolytic
activity.
9.3.2.2  Quantification of Noncarcinogenic Effects
9.3.2.2.1  Ararcid fibers.   Only one long-term inhalation study  was found in the
available literature (Stadler, 1986).   In this study,  a NOAEL  of 0.08 mg/tn3 was
identified in rats,  based  on cellular changes  in lung  architecture  following
exposure at 0.32 and 0.63 mg/m  for 2 years  or at 2.23 mg/m  for 1 year.   These
changes included  type II cell hyperplasia,  nodular thickening of ridges  at
alveolar duct bifurcations, and formation of ciliated columnar  epithelial  cells
in alveolar ducts.   Significant increases in absolute and relative lung  weights
were seen at  0.63  and 2.23 mg/m ,  and  emphysema and  significant  increases in
mortality were noted at 2.23 mg/m .   The RfD for a 70-kg adult  is  calculated  as
follows:

                 RfD s (O
where:
     0.08 mg/m3 = NOAEL,  based  on the  absence  of cellular  change in  lung
                  architecture (Stadler, 1986).
            100 = uncertainty factor, chosen  in accordance with  EPA guidelines
                  for use with a NOAEL from an animal study.
     1988                          9-33         DRAFT-DO NOT "QUOTE OR CITE

-------
9.3.2.2.2   Carbon  fibers.   No  suitable long-term  inhalation studies were founa
in the available literature.  In a subchronic inhalation study reported  by Owen
et al. (1986),  no  effects were obtained in  rats following exposure to 20 mg/m
for  6 hours/day,  5  days/week  for 16  weeks.  The rats  were  observed for a
recovery period of 32 weeks.   However, the  authors did  not report the respira-
ble  concentration,  and the majority of carbon fibers within  the aerosol cloud
were  beyond the respirable range.   Furthermore, a small  number  of animals were
exposed  for the entire 16-week period (20 males), and a range of doses  was not
tested.  Therefore,  this study was  not considered suitable for  derivation of  a
conservative estimate  of the  RfD.   Since no other studies  were found,  the  RfD
for carbon  fibers could  not be calculated.
9.3.2.2.3   Polyolefin  fibers.   No  inhalation studies were  found in the  avail-
able  literature,  and  therefore, an RfD for polyolefin  fibers  could not be
calculated.
9.3.2.3  Characterization and Quantification of Carcinogenic  Effects
9.3.2.3.1   Aramid  fibers.   In  a chronic inhalation study conducted by  Haskell
Laboratory  (Stadler,  1986),  cystic  keratinizing  squamous cell carcinomas were
exhibited  in female but not  in male rats  (see Table 7-12).   Groups of 100
animals (50 males, 50  females) were exposed to Kevlar  fibers at concentrations
of 0,  0.08, 0.32,  or 0.63 mg/m  for 2 years or 2.23  mg/m  for 1 year (followed
by a 1-year recovery period).   These exposure concentrations  were approximately
equivalent  to  2.4,  25.4, 11.2,  and  435.5 fibrils/cm  , respectively.  Since the
individual  animal  data were available, a survival analysis was performed  tc
determine  if the  survival  of  the  animals  could  have an influence  on  tumor
development.   The  survival  analysis showed  that animals exposed to  0.08, 0.32,
and 2.23 mg/m   exhibited significantly shorter lifespans (p <0.05)  compared  to
controls.
     Analysis  of  the  tumor  incidence data, accounting for  differences in
survival  between  groups, can be performed  following procedures described  by
International  Agency for Research  on Cancer (1980)  for tumors found  in an
incidental  context.  By  this method, the data are  separated into time intervals
exhibiting  a monotonic increase in tumor incidence across dose groups.  This
created  a  number of  time intervals resulting in  a time-weighted  statistic.
                                            £
Application  of  this  procedure  to the Kevlar  data resulted in the  formation  of
only one interval,  indicating  that there is no advantage in  using this method
to adjust  for  survival differences.   Hence, we will  not adjust  for  survival  in
the tumor incidence analysis.
July 1988                          9-34         DRAFT—DO NOT QUOTE OR  CITE

-------
     The incidence of  lung  carcinoma in female rats exhibited a  significant
•increasing trend with  dose  by the Cochran-Armitage  Trend  test  (p  <0.01).  The
incidence of tumors in the high-dose group was significantly  increased compared
to controls (6/69 versus 0/78) by Fisher's Exact test  (p <0.05).
     The multistage model was used to estimate  cancer risk.  No adjustments to
dose were  made to account  for  differences  in extrapolating  from animals to
humans.  For  purposes of this  extrapolation, the  high-exposure  value  was
reduced by half  to account  for exposure of only 1 year with  a  1-year recovery
period.  The  assumption  made  here  is that the animals were exposed to the same
number of fibers at 2.23 mg/m  for 1 year as they would have  been  had they  been
exposed to 1.12  mg/m   for 2 years.   The  cancer  unit risk, using the data shown
below,  is 9.4 x  ID*2  (mg/m3)"1 or 7.5 x 10~4 (fibrils/cm3)"1 .   This value  is
the 95 percent upper bound on risk.  The true value  could  be  as low as zero  but.
is unlikely to be higher.
                                            Exposure level  (mg/m3)
                                        0.08      0.32      0.63      2.23
Squamous cell carcinoma        0/78**   0/76      0/76      4/79      6/69**
**Sigm'ficant dose-related trend by the Cochran-Armitage Trend test denoted
  at controls (p <0.01), and significant increase compared to controls by
  Fisher's Exact test denoted at the high dose (p <0.01).

9.3.2.3.2  Carbon  fibers.   Only one dermal carcinogenicity  study  using  four
types of  carbon  fibers was found in the  available  literature  (DePass, 19£2).
In  this  study,  CF,  MAT,  PAN,  and PAN-oxidized carbon  fibers were  not
carcinogenic  in  mice  following  dermal  exposure,  although a few tumors  were
present following  the dermal  application of CF, MAT,  and PAN-oxidized carbon
fibers.   The incidences were not significantly different from controls, and the
tumor types observed  (see Table 7-13) have been shown to occur spontaneously in
this strain  of  mice.   Therefore, the tumor induction was not considered to be
due to  the  administration of carbon fibers.   Since  no other suitable studies
were found,  an  estimate  of  the carcinogenic risk of  carbon fibers  was  not
calculated.
9.3.2.3.3  Polyolefin fibers.  No inhalation or intrapleural studies were found
in the  available literature  (see Table 7-13).  Polypropylene was not carcino-
genic following  a  single  intratracheal  injection of 1.25  mg in Long-Evans  rats
July 1988                          9-35         DRAFT—DO NOT QUOTE OR CITE

-------
(Moreno and Moreno,  1986).   Conversely,  intraperitoneal  injection resulted in
the induction of  mesotheliomas  or sarcomas in 2/51 female Wistar rats (Pott
et a!., 1987b).   However, the increased incidence was  not significantly differ-
ent from that of controls.
     Very limited evidence  suggests  that polyolefin,  specifically polypropyl-
ene,  is  carcinogenic following  intraperitoneal  administration.   However,  the
lack of  inhalation  data  precludes the quantification  of  the  carcinogenic risk
associated with exposure to polyolefin fibers.

9.3.3  Characterization of Carcinogenic Potential Based on Human  and  Animal
       Data
     The weight  of evidence from both human  health effects  data and animal
toxicity data (see  also  summary in Section 9.5)  suggest  the  following carcino-
genic categories  for the synthetic  fibers:  aramid fibers, B2; carbon  fibers,
D; polyolefin fibers, D.
9.4  MINERALS
9.4.1  Human Health Effects
9.4.1.1  Attapulgite
9.4.1.1.1   Qualitative  evaluation  of  noncarcinogem'c and carcinogenic effects.
Only one  study  (Waxweiler et al.,  1985) detected an excess of lung cancer in
white male  workers  (SMR = 192, p <0.05, Power =  37  percent).  This  risk  in-
creased  to  threefold  (p <0.05) among  white  males  employed  in  high- and
intermittent-exposure job categories  after  20 or more  years of  latency.
Although  no excesses  from NMRD were detected, there  was  insufficient  power to
detect either NMRD  or site-specific cancers,  such as  lung cancer, among  the
entire cohort.  The small number of deaths and low power  and  absence  of  smoking
histories precluded meaningful analysis of the data by  cumulative dust exposure
and  latency or  duration of employment, especially among nonwhite  employees.
Therefore, the weight of evidence for a causal relationship between exposure to
attapulgite and  lung  cancer  or NMRD is inadequate.  The International Agency
for  Research  on  Cancer  (1987) has  determined  that there  is inadequate evidence
of human carcinogenicity for attapulgite.
                                                                       rTTC

-------
9.4.1.2  Erionite
9.4.1.2.1   Qualitative  evaluation  of  noncarcinogenic and carcinogenic effects.
Cross-sectional mortality studies have demonstrated a high  prevalence of deaths
from malignant mesotheliomas, to a greater extent of the pleura  arid  to a lesser
extent of  the  peritoneum,  in the adult populations  from three Turkish villages
where  individuals  were  exposed throughout their  lifetimes.  Erionite fibers
were identified in soil  and air samples from these three villages  as well  as in
the lung tissues  of patients with  malignant mesotheliomas.  No malignancies of
any type were found in a control village during the same period.   Smoking  could
not account  for  the excesses seen; less than  1 percent of the women from the
three contaminated villages smoked, yet a high prevalence of  malignant mesothe-
liomas was observed among women.
     No apparent  sex  differences were observed.   Tests  for other  environmental
carcinogens  showed  that  arsenic, uranium,  nickel,  and chromium were present in
trace  amounts.   Tests for aflatoxins and  radioactivity were negative  (Baris
et al., 1978).  None  of the  villagers interviewed had experienced either  known
or suspected exposure to asbestos.  Since the major occupation  in  the  three
villages was farming, an environmental  rather than an occupational  agent  would
be more likely  to be responsible for the malignancies.   Although causal  rela-
tionships  cannot  be  discerned through either cross-sectional studies or  cese
reports, the consistency of the findings from multiple studies  and the  presence
of erionite  rather  than asbestos suggests an  etiological  role  for  erionite.
The International  Agency for  Research  on  Cancer (1987) has determined that
there is sufficient evidence for the carcinogenicty of erionite  to humans.
     Because of carcinogenic effects, each of the studies examined nonmalignant
pleura! disease.  Tests  for  tuberculosis were negative.  Fibrosis of the  lung
and pleura were two  effects commonly noted.   The  research  by  Baris et  al.
(1987) culminated in  a  finding that  there was  a  higher prevalence  of pleural
radiological changes  in  villages affected by mesotheliomas as  compared  to the
control village.  Because  cross-sectional  studies provide a measure of  preva-
lence  rather than  a measure of  incidence  (i.e.,  a measurement  of risk),  they
cannot be used for quantitative risk assessment.
9.4.1.3  Talc
9.4.1.3.1   Qualitative  evaluation  of  noncarcinogenic and carcinogenic effects.
Health effects  seen following occupational exposure to  talc  included increased
mortality  from  respiratory  cancer  and nonmalignant respiratory  disease (NMRD).

     T98R                          9-37         DRAFT-DO NOT QUOTE OR  CITE

-------
In many of the studies, however, workers were simultaneously  exposed to contam-
inants such  as  asbestiform amphiboles,  silicates, carbonates, and radon.   This
lack  of  specificity made it difficult  to  ascribe any  of the health effects
solely to  talc exposure.   Therefore,  the weight of evidence  for a causal
relationship between  exposure  to talc  alone and lung cancer  or  NMRD is inade-
quate.   International  Agency  for Research  on Cancer (1987) has also determined
that there is inadequate evidence for pure talc.
9.4.1.4  Vermiculite
9.4.1.4.1   Qualitative evaluation of noncarcinogenic and carcinogenic effects.
Chest radiographs  from a cross-sectional morbidity  study showed some  evidence
of  pleura!   disease  in  workers  employed at least  5  years  in  a Montana
vermiculite  ore  mine  and  mill.   The effect of smoking  and  the  presence  of
tremolite-actinolite  fibers confounded  these  findings  with  respect  to  the
prevalence of pleura!  disease in this group of workers.
     Results from  a  historical  prospective mortality study of 575 male employ-
ees of the  same Montana vermiculite ore mine  and mill  (Amandus and Wheeler,
1987) found  significant excesses  from  lung cancer (SMR =  223.2, p  <0.01,
Power = 38 percent) and NMRD (SMR = 243.0, p <0.05,  Power = 35  percent)  despite
low power.   The highest excesses were found among those in the  highest exposure
category (>399  fiber-years) for both lung cancer (SMR  = 575.5,  p <0.01)  and
NMRD  (SMR =  400.7, p <0.01).   Because  vermiculite ore was found  to be contarr.i-
nated with fibrous tremolite-actinolite, it is not possible  to  attribute these
and other excesses to vermiculite exposure.  Therefore,  the  weight of evidence
for a causal  relationship  between exposure  to  vermiculite and  lung cancer  or
NMRD is inadequate.

9.4.2  Animal Toxicity
9.4.2.1  Quantification of  Noncarcinogenic Effects
9.4.2.1.1  Attapulgite.  No inhalation studies on attapulgite were found  in the
available literature,  and therefore, an RfD was not  calculated.
9.4.2.1.2  Erionite.   Although no suitable  inhalation  toxicity  studies  were
found for use  in quantification of noncarcinogenic  effects,  erionite  has  been
shown to be  carcinogenic in animals and humans  following inhalation  exposure
(see Section 9.4.2.2.2 and  Chapter 7).   Consequently,  an inhalation RfD  value
was not calculated.
9.4.2.1.3  Talc.   Only one inhalation  study was found  in the available litera-
ture  (Wagner et  a!.,  1977).  In  this  study,  groups of  rats  were exposed to a
July 1988                          9-38         DRAFT—DO NOT QUOTE OR CITE

-------
mean respirable concentration of 10.8 mg/m3 talc for 7.5 hours/day, 5 days/week
for 3, 6,  or 12 months.  The animals were then observed for  up  to  1 year.  At
this concentration, minimal to slight fibrosis was  observed in the  talc-exposed
rats; the severity increased with increasing length of exposure.  Thus, a  LOAEL
of  10.8  mg/m  was identified, and since  this  was  the only  inhalation study
available for consideration, this value was used to calculate the RfO  for  talc.
The RfD for a 70-kg adult is calculated as follows:
                 RfD =          m   = 0.011 mg/m3 (10.8 mg/m3)
where:
     10.8 mg/m3 = LOAEL, based on presence of pulmonary fibrosis in rats,
           1000 = uncertainty factor, chosen  in  accordance with EPA guidelines
                  for use with a LOAEL from an animal study.

9.4.2.1.4  Venm'culite.   No inhalation  studies  were found in  the available
literature.  Therefore, an RfD for vermiculite could not be calculated.
9.4.2.2  Characterization and Quantification of Carcinogenic Effects
9.4.2.2.1  Attapulgite..   No studies  on  the  carcinogenicity  of attapulgite
following  inhalation  exposure were  found in  the  available  literature  (see
Table 7-13).   Negative results were obtained following intrapleural administra-
tion  in  rats  by Stanton et  al.  (1981).   However,  Pott et al.  (1974, 19S7a)
reported  that  attapulgite  was carcinogenic when administered  via  the intra-
peritoneal route,  inducing  tumors  in 40  to 65 percent of the rats injected.
The majority of the tumors were reportedly mesotheliomas.  No methodology for
the quantification  of the carcinogenic  risk  using data  from intraperitoneal
studies is available and, therefore, a risk estimate was not calculated.
9.4.2.2.2  Erionite.   Unlike the other  naturally occurring  and manmade mineral
fibers mentioned  above (see also Table 7-13), erionite was found to be  a very
potent carcinogen following inhalation  exposure,  Inducing mesotheliomas  in
27/28 rats (Wagner  et al.,  1985).   Erionite has  also been shown to be carcino-
genic  after  intraperitoneal  administration in  rats and mice  (Suzuki  and
Kohyama, 1984; Ozesmi et al., 1985; Pott et al.; 1984a, 1987a) and intrapleural
injection in two  strains of  rats (Maltoni et al.,  1982; Wagner et al., 1965).
     Using the multistage model for carcinogenic risk  estimation, the unit risk
is 5.56  (mg/m3)"1,  based on the tumor incidences  obtained from  the inhalation

July 1988                          9-39         DRAFT--nn KIHT 'nnn-rc nc rm

-------
study of  Wagner  et al.  (1985) given below.   This value  represents the 95 per-
cent  upper bound  on  risk.   The  true value could be as  low  as  zero but is
unlikely  to  be  higher.   No adjustments were made to account for differences in
extrapolation between animals and humans.
                                          Exposure level  (mg/tn3)
                                                                       10
         Mesotheliomas          0/28**                               27/28**

""Significant dose-related trend by the Cochran-Armitage Trend test denoted
  at controls (p <0.01), and significant increase compared to controls by
  Fisher's Exact test denoted at the high-dose group (p <0.01).

9.4.2.2.3  Talc.   There is  no  evidence  that talc is carcinogenic  following
ingestion, inhalation exposure, or intratracheal, intrapleural,  or intraperito-
neal administration in rats, hamsters, or mice.   Negative results were obtained
following inhalation  exposure  in rats and  hamsters  by Wagner et  al.  (1977) and
Wehner  et  al.  (1977).   Similar findings were  observed after intratracheal
instillation in  hamsters,  although talc  proved  to be an effective  cocarcinogen
when  administered with  benzo(a)pyrene (Stenback eta!.,  1986).   No  tumor
induction was observed  following intrapleural administration in  two strains of
rats (Wagner et  al.,  1977; Stanton et al.,  1981),  and  negative results were
also obtained after  intraperitoneal  injection of talc  in  rats and mice  (Pott
et al., 1974; Ozesmi et al., 1985).
     The weight  of evidence indicates that  talc is  not  carcinogenic in labora-
tory  animals following  inges\ion, inhalation  exposure, or  intfatracheal,
intrapleural, or intraperitoneal administration.  Therefore, an estimate of the
carcinogenic risk for talc was not calculated.
9.4.2.2.4  Vermiculite.  The  only study on the carcinogenicity of vermiculite.
found in the available literature indicated that vermiculite is not carcinogen-
ic following intrapleural  administration in rats (Hunter  and Thomson,  1973).
However, since no  other carcinogenicity studies were found,  an  assessment of
the carcinogenic risk of vermiculite cannot be made at this time.
July 1988                          9-40         DRAFT—DO NOT QUOTE OR CITE

-------
9.4.3  Characterization of Carcinogenic Potential  Based on  Human  and Animal
       Data
     The weight  of evidence from both  human  health  effects  data and animal
toxicity data  (see  also the following summary) suggest the following  carcino-
genic  categories  for the minerals:   attapulgite,  C;  erionite,  B2; talc, D;
vermiculite, D.
9.5  SUMMARY OF CARCINOGENIC AND NONCARCINOGENIC EVIDENCE
     The  NOAELs  and LOAELs  identified from inhalation toxicity studies are
listed  in  Table  9-5.   These values serve as the  basis  for calculation of the
RfD values.   Suitable  inhalation toxicity studies were  not  found  for  carbon
fibers, polyolefin  fibers,  attapulgite,  and vermiculite.  Although an  inhala-
tion study was found for erionite, only the carcinogenic effects were described.
The weight of  evidence for  animal  and human carcinogenicity  are also listed  in
Table 9-5.  Most  fibers for which data were available  showed some  carcinogenic
effect  when  administered via the  intrapleural  and/.or  intraperitoneal  route,
except  for talc.   However,  only aramid fibers  and erionite are  carcinogenic  by
the inhalation route.
     The  weight  of  evidence  for carcinogenicity  in humans  was limited  for
fibrous glass  wool  (with diameters <3 pm), mineral wool, and erionite.   Esti-
mates of  cancer  risk were not  determined  because dose-related  increases were
not observed,  and exposure  data were  not  available.   Inadequate evidence was
found for glass microfibers, attapulgite, talc, and vermiculite.  Both talc  and
vermiculite were  found to  contain  asbestos fibers.   Epidemiological  studies
were not found for ceramic, aramid, carbon, or polyolefin fibers.

-------
              TABLE  9-5.   SUMMARY OF CARCINOGENIC AND NONCARCIMOGEN1C EFFECTS FROM ANIMAL  AND HUMAN DATA
Fiber
Glass microfibers
Fibrous glass
Rock wool
Slag wool
Ceramic fibers
Aramid fibers
Carbon fibers
Polyolefin fibers
Attapulgite
Erionite
Talc
Vermiculite
Noncarcinoqenic e
(mg/m-1 )
NOAEL LOAEL
3.0 100
~a 5
5 10
10
2.45 8.4
0.08 0.32
—
..
._
—
10.8
..
ffects

RfD Inh.
0.03
0.005
0.05
0.01
0.024
0.0008 *
—
--
—
4
0.011
—
Carcinogenic evidence in
Animal Unit risk Human
Intpl. Intrpr. (mg/m3) ' Inh.
+ * -- Inadequate
+ * Limited
+ + — . Limited
± - — Limited
i *
t 9.4xlO~2
_
-4
± — Inadequate
+ + 5156 Limited
Inadequate
Inadequate
Carcinogenic
classification
82
Bl
Blb
Blb
C
B2
0
D
C
Bl
0
0
 No suitable studies were found.
'Human evidence is for mineral wool, i.e., rock and slag wool combined.
 Abbreviations:  Inh. = inhalation; Intpl. = intrapleural administration; Intrpr.  = intraperitoneal  injection.

-------
9.6  REFERENCES


Amandus, H.  E.;  Wheeler,  R.  (1987) The morbidity and mortality of vermiculite
     miners  and  millers exposed to tremolite-actinolite:  Part II. Mortality.
     Am. J.  Ind. Med. 11:  15-26.

Andersen,  A.;  Langmark,  F.  (1986) Incidence of cancer in  the mineral-wool
     producing  industry in Norway.  In:  Saracci, R., ed. Contributions to the
     IARC  study  on  mortality and cancer incidence  among man-made  mineral fiber
     production workers.  Scand. J. Work Environ. Health 12 (suppl. 1):  72-77.

Ban's,  Y.  I.;  Sahin, A.  A.;  Ozesmi,  M.; Kerse, I.; Ozen,  E.;  Kolacan,  B.;
     Altinors, M.; Goktepeli, A. (1978) An outbreak of pleural mesothelioma  and
     chronic fibrosing pleurisy in the village of Karain/Uerguep in Anatolia.
     Thorax  33: 181-192.

Baris,  I.;  Simonato,  L.;  Artvinli, M.; Pooley, F.; Saracci, R.; Skidmore,  J.;
     Wagner, C. (1987) Epidemiological and environmental evidence of the health
     effects  of exposure  to  erionite  fibres:  a  four-year study  in  the
     Cappadocian region of Turkey. Int. J. Cancer 39: 10-17.

Bayliss, D.  L.;  Dement,  J.  M.; Wagoner, J. K.; Blejer, H.  P. (1976) Mortality
     patterns among  fibrous  glass production workers. Ann.  N.  Y.  Acad.  Sci.
     271: 324-335.

Bertazzi,  P. A.; Zocchetti,  C.; Pesatori, A.;  Radice,  L.;  Riboldi,  L.  (1964)
     Mortalita per  tumori  in una coorte di produttori  di  fibre vetrose  [Cancer
     mortality among workers in the glass  fibre production industry]. Med.  Lev.
     75: 339-358.

Bertazzi, P. A.; Zocchetti, C.; Riboldi, L.; Pesatori, A.; Radice, L.; Latocca,
     R.  (1986)  Cancer  mortality of an  Italian  cohort  of  workers in man-made
     glass-fiber production.  In:  Saracci, R.,  ed. Contributions  to  the IARC
     study  on  mortality  and  cancer  incidence  among man-made mineral  fiber
     production workers.  Scand. J. Work Environ. Health 12 (suppl. 1):  65-71.

Bertrand, R.; Pezerat, H.  (1980) Fibrous glass: carcinogenicity and dimensional
     characteristics.  In: Wagner,  J.  C.; Davis, W.,  eds.  Biological  effects  of
     mineral fibres = Effets biologiques des fibres minerales: proceedings  of a
     symposium,  v.  2;  September 1979; Lyon,  France.  Lyon,  France:  World Health
     Organization,  International  Agency for Research on Cancer;  pp.  901-911.
     (IARC scientific publication no. 30.; INSERM  symposia series: v. 92).

Cherrie, J.; Dodgson, J.  (1986)  Past exposure to airborne  fibers and  other
     potential  risk factors  in the European nan-made mineral fiber production
     industry. In: Saracci, R., ed. Contributions  to the IARC study on mortali-
     ty  and  cancer  incidence among man-made mineral fiber production workers.
     Scand.  J. Work Environ.  Health 12  (suppl.  1): 26-33.
July 1988                          9-43         DRAFT—DO NOT QUOTE OR CITE

-------
Cherrie, J. ;  Dodgson,  J.;  Groat,  S.;  MacLaren,  W.  (1986)  Environmental surveys
     in the  European  man-made mineral fiber production industry.  In:  Saracci,
     R., ed.  Contributions  to the 1ARC study on mortality and  cancer  incidence
     among man-made mineral  fiber production workers.  Scand.  J.  Work Environ.
     Health 12 (suppl. 1): 18-25.

Cherrie, J.;  Krantz,  S.;  Schneider, T.;  Ohberg, I.;  Kamstrup,  0.;  Linander, W.
     (1987) An experimental  simulation of an early rock wool/slag  wool produc-
     tion process. Ann. Occup. Hyg. 31: 583-593.

Claude, J.;  Frentzel-Beyme,  R.  (1984) A mortality  study of workers employed in
     a German rock wool factory. Scand. J. Work Environ. Health 10: 151-157.

Claude, J.; Frentzel-Beyme, R. (1986) Mortality of workers in a German rockwool
     factory—a  second look  with  extended follow-up.   In: Saracci,  R.,  ed.
     Contributions to  the  IARC study on mortality and  cancer  incidence  among
     man-made mineral  fiber  production workers.  Scand.  J.  Work Environ.  Health
     12 (suppl.  1): 53-60.

Davis,  J.  M.  G.; Addison, J.;  Bolton,  R.  E.;  Donaldson,  K.;  Jones,  A.  D.;
     Wright,  A.  (1984) The  pathogenic effects  of fibrous ceramic aluminum
     silicate glass administered to rats by inhalation or peritoneal injection.
     In: Guthe,  T., ed. Biological effects of man-made mineral  fibres: proceed-
     ings  of  a WHO/IARC  conference in association with JEMRB  and  TIMA,  v. 2;
     April  1982;  Copenhagen,  Denmark.  Copenhagen,  Denmark: World Health  Organ-
     ization, Regional Office for Europe; pp. 303-322.

DePass, L.   R. (1982)  Evaluation of the  dermal  carcinogenicity of  four carbon
     fiber materials  in male C3H/HeJ mice.  Final  report by  the  Bushy  Run
     Research Center for the  Union Carbide Corporation; Project no. 45-13.

Dodgson, J.;  Cherrie, J.;  Groat, S.   (1987)  Estimates of past exposure to
     respirable  man-made mineral  fibres in the  European insulation wool  indus-
     try.  Ann. Occup.  Hyg. 31: 567-582.

Doll,  R.  (1987)   Symposium  on MMMF,  Copenhagen, October  1986: overview  and
     conclusions. Ann. Occup. Hyg. 31: 805-819.

Enterline,  P. E.; Marsh, G.  M. (1982)  Mortality of  workers in the  man-made
     mineral fibre industry.  In: Biological effects of man-made mineral  fibres:
     proceedings  of  a WHO/IARC conference;  April  1982; Copenhagen, Denmark.
     Geneva,  Switzerland:  World  Health  Organization,  Regional  Office  for
     Europe; pp.  965-972.

Enterline, P. E.;  Marsh,  G.   M.; Esmen, N. A. (1983) Respiratory disease among
     workers  exposed   to  nan-made mineral  fibers.  Am.  Rev.   Respir. Dis.
     128:  1-7.

Enterline,  P. E.; Marsh, G.  M.;  Henderson,  V.  (1986)  Mortality update  of a
     cohort of U.  S.  man-made mineral fiber  workers.  Presented  at:  interna-
     tional symposium  on  man-made mineral fibres  in the  working environment;
     October; Copenhagen,  Denmark.  Copenhagen,  Denmark: World  Health Organiza-
     tion,  Regional Office for Europe.
July 1988                          9-44         DRAFT—DO NOT QUOTE OR CITE

-------
Enterline, P.  E.;  Marsh,  G.  M.; Henderson, V.;  Callahan, C.  (1987) Mortality
     update of  a  cohort of U.S. nan-made mineral  fiber  workers.  Ann.  Occup.
     Hyg. 31: 625-656.

Gardner,  M.;  Winter, P.  D.;  Pannett,  B.;  Simpson, M. J. C.;  Hamilton,  C.;
     Acheson, E.  0.  (1986) Mortality study of workers In the man-made  mineral
     fiber production industry In the  United Kingdom.  In:  Saracci, R., ed.
     Contributions to the IARC study on mortality and cancer incidence among
     man-made mineral fiber  production  workers.  Scand. J. Work Environ. Health
     12 (suppl. 1): 65-93.

Goldstein, B.;  Rendall,  R.  E.  G.; Webster,  I.  (1983)  A comparison of the
     effects  of exposure  of baboons to crocidolite  and  fibrous-glass  dusts.
     Environ. Res. 32: 344-359.

Gross, P.; Kaschak,  M.; Tolker, E. B.; Babyak,  M.  A.;  de Treville, R. T.  P.
     (1970)  The pulmonary reaction  to  high concentrations of  fibrous glass
     dust: a preliminary report. Arch. Environ. Health 20: 696-704.

Hunter,  B.;  Thomson, C.  (1973) Evaluation of the  tumorigenic potential  of
     vermiculite  by  intrapleural  injection  in  rats.  Br.  J.  Ind, Med.
     30:  167-173.

International Agency  for  Research on Cancer.   (1980)  Long-term and  short-term
     screening  assays for carcinogens: a  critical  appraisal.  Lyon, France:
     International Agency  for Research  on  Cancer; IARC Monographs,  Supplement
     2.

International  Agency for  Research on  Cancer.  (1987) IARC Monographs  on the
     evaluation of  the  carcinogenic  risk  of  chemicals to humans:  silica and
     other silicates, v. 42.  Lyon, France:  World Health Organization.

Le Bouffant, L.; Daniel, H.; Henin, J. P.;  Martin, J. C.; Normand, C.;  Tichoux,
     G.; Trolard,  F.  (1986)  Experimental study on long-term effects of MMMF on
     the  lung of  rats.  Presented at:  international symposium on  man-made
     mineral  fibres  in  the working environment; October; Copenhagen,  Denmark.
     Copenhagen,  Denmark:  World  Health Organization,  Regional Office  for
     Europe.

Maltoni,  C.;  Minardi,  F.;  Morisi,   L.  (1982)  Pleura!  mesotheliemas  in
     Sprague-Dawley  rats  by  erionite:  first  experimental evidence. Environ.
     Res. 29: 238-244.

McConnell, E. E.; Wagner,  J.  C.; Skidmore,  J.  W.; Moore, J. A.  (1984) A compar-
     ative study  of the fibrogenic and carcinogenic effects  of UICC Canadian
     chrysotile B asbestos and glass micrpfibre  (JM100).   In: Biological Effects
     of  Man-made  Mineral  Fibres:  proceedings  of a  WHO/IARC conference;  v.  2;
     April  1982.  Copenhagen,  Denmark:  World  Health Organization,  Regional
     Office for Europe; pp. 234-250.

Mitchell,  R.  I.; Donafrio,  D.  J.;  Moorman, W.  J.  (1986) Chronic inhalation
     toxicity  of  fibrous  glass  in rats and  monkeys. J.  Am.  Coll.  Toxicol.
     5: 545-575.
July 1988                          9-45         DRAFT—DO  NOT QUOTE OR CITE

-------
Mohr, U.; Pott, F.; Vonnahme, F.-J. (1984) Morphological aspects of mesothelio-
     mas after  intratracheal  instillations of fibrous dusts  in  Syrian  golden
     hamsters. Exp. Pathol. 26: 179-183.

Monchaux, G.; Bignon, J.; Jaurand, M. C.; Lafuma, J.; Sebastien, P.; Masse,  R.;
     Hirsch,  A.;  Goni,  J.  (1981) Mesotheliomas  in  rats  following  inoculation
     with acid-leached  chrysotile asbestos and other mineral fibres. Carcino-
     genesis  2: 229-236.

Moreno,  0.  M.;  Moreno,  M.  T. (1980) Twenty-one month observation after single
     intratracheal  insufflation  [unpublished  study].  Spinnertown,  PA:  M.B.
     Research Laboratories, Inc.; Project no. 78-2806.

Moulin,  J.  J.;  Mur, J.  M.; Wild, P.; Ferreaux, J. P.; Pham, Q. T.  (1985) Oral
     cavity  and laryngeal  cancers  among  man-made  mineral  fiber production
     workers. Scand. J.  Work Environ. Health 12: 27-31.

Muhle, H.; Pott, F.; Bellmann, B.; Takenaka, S.; Ziem, U. (1986) Inhalation and
     injection  experiments  in  rats for testing man-made mineral fibers in the
     working  environment.  Presented at:  international  symposium on man-made
     mineral  fibres in  the working environment; October; Copenhagen, Denmark.
     Copenhagen,  Denmark:  World  Health  Organization,  Regional Office  for
     Europe.

National Research Council.  (1984) Asbestiform fibers:  nonoccupational  health
     risks.  Washington, DC: National Academy Press.

Olsen,  J. H.  ; Jensen, 0.  M.  (1984)  Cancer incidence among  employees  in  one
     mineral  wool  production  plant in Denmark. Scand. J. Work Environ. Health
     10: 17-24.

Olsen, J. H.; Jensen, 0. K.; Kampstrup, 0. (1986)  Influence of smoking habits
     and place  of residence on the risk  of lung cancer among workers in one
     rock-wool  producing plant  in Denmark. In: Saracci, R., ed. Contributions
     to the IARC study on mortality and cancer incidence among man-made mineral
     fiber production workers.  Scand.  J.  Work Environ.  Health 12  (suppt.  1):
     48-52.

Owen, P. E.;  Glaister,  J.  R.; Ballantyne,  B.;  Clary, J.  J. (1986) Subchronic
     inhalation toxicology of carbon fibers. JOM J. Occup. Med. 28:  373-376.

Ozesmi,  M.;   Patiroglu,   T.  E.;  Hillerdal,  G.; Ozesmi,  C.  (1985) Peritoneal
     mesothelioma  and malignant lymphoma in mice caused by  fibrous zeolite.  Br.
     J. Ind.  Med.  42: 746-749.

Pigott, G.  H.;  Gaskell, B.  A.;  Ishmael, J.  (1981) Effects of long  term inhala-
     tion of  alumina fibres in rats. Br. J. Exp. Pathol. 62: 323-331.

Pott, F. (1978) Some aspects on  the  dosimetry of the carcinogenic potency  of
     asbestos and  other fibrous dusts. Staub Reinhalt. Luft  38: 486-490.

Pott,  F.;  Friedrichs,  K.  H.  (1972)  Tumoren der  Ratte  nach i.p.-Injektion
     faserfoermiger  Staeube  [Tumors in rats after  i.p.  injection  of fibrous
     glass].  Naturwissenschaften  59: 318.


July 1988                          9-46         DRAFT-DO NOT QUOTE  OR CITE

-------
Pott,  F.;  Huth, F.; Friedrichs, K.  H.  (1974)  Tumorigenic effect of fibrous
     dusts in experimental animals. EHP Environ.  Health Perspect.  9:  313-315.

Pott,  F.;  Huth,  F.;  Friedrichs, K. H.  (1976) Results of animal  carcinogenesis
     studies after application  of  fibrous  glass  and  their  implications regard-
     ing human  exposure.  In:  Occupational  exposure to fibrous glass:  proceed-
     ings  of a  symposium;  June 1974; College Park,  MO.  Cincinnati, OH: U. S.
     Department  of  Health, Education,  and Welfare,  National  Institute  for
     Occupational Safety and  Health; pp.  183-191; HEW publication no.  (NIOSH)
     76-151. Available from: NTIS," Springfield, VA; PB-258869.

Pott,  F.;  Schlipkoeter, H. W.;  Ziem,  U.;  Spurny, K.; Huth,  F.  (1984a)  New
     results from implantation  studies  with  mineral  fibers.  In: Guthe, T., ed.
     Biological  effects of man-made  mineral  fibres:  proceedings of a WHO/IARC
     conference  in association  with  JEMRB  and  TIMA,  v.  2;  April 1982;  Copenha-
     gen,  Denmark.  Copenhagen,  Denmark: World Health  Organization,  Regional
     Office for  Europe; pp. 286-302.

Pott,  F.;  Ziem,  U.;  Mohr,  U.  (1984b) Lung  carcinomas and mesotheliomas follow-
     ing intratracheal  instillation  of  glass fibers  and asbestos.  In:  Proceed-
     ings  of  the Vlth  international pneumoconiosis  conference, v. 2; 19B3;
     Bochum, Federal  Republic of  Germany.  Bremerhaven, Federal Republic  of
     Germany:  Verlag fuer neue Wissenschaft; pp.  746-756.

Pott,  F.;  Ziem,  U.;  Reiffer,  F.-J.;  Huth,  F.;  Mohr,  U.  (1987a)  Carcinogenicit;
     studies on  fibres, metal  compounds,  and.some other dusts  in rats.  Exp.
     Pathol. 32: 129-152.

Pott,  F.;  Roller, M.;  Ziem, U.; Reiffer,  F.  J.;  Bellmann, B.; Rosenbruch, K. ;
     Huth, F.   (1987b)  Carcinogenicity  studies  on natural  and man-made fibers
     with  the  intraperitoneal test  in  rats. Proceedings  of a  Symposium on
     mineral  fibers  in the  non-occupational  environment;  Lyon,  France;
     September 8-10.

Robinson,  C. F.; Dement,  J. M.;  Ness,  G. 0.; Waxweiler,  R.  J.  (1982)  Mortality
     patterns of rock and slag mineral wool production workers:  an epidemiolog
     ical and environmental study.  Br.  J.  Ind.  Med. 39: 45-53.

Saracci,  R.;  Simonato, L.; Acheson, E. D.; Andersen,  A.;  Bertazzi,  P.  A.;
     Claude, J.;  Charnay,  N.;  Esteve,  J.;  Frentzel-Beyme,  R.  R.;  Gardner, M.
     J.; Jensen,  0.  M.; Maasing, R.; Olsen, J. H.; Teppo, L.; Westerholm, P.;
     Zocchetti,  C.  (1984)  Mortality  and incidence of cancer of workers in the
     man made  vitreous  fibres producing industry: an international investiga-
     tion at 13  European plants. Br.  J. Ind.  Med. 41: 425-436.

Shannon, H.; Hayes,  M.; Julian, J.; Muir, D.  (1982) Mortality experience of
     glass fibre workers.  In:  Biological  effects of man-made mineral  fibres:
     proceedings of  a  WHO/IARC conference, v.  1;  April; Copenhagen,  Denmark.
     Geneva,  Switzerland:  World  Health Organization,  Regional Office  for
     Europe; pp. 347-349.

Shannon, H. S.;  Hayes,  M.; Julian, J.  A.;  Muir,  D.  (1984) Mortality experieno
     of glass fibre workers. Br. J.  Ind. Med. 41: 35-38.
July 1988                          9-47         DRAFT—DO NOT QUOTE OR CITE

-------
Shannon, H.  S. ;  Jamieson,  E.; Julian, J. A.; Muir, D.  C.  F.;  Walsh, C.  (1985)
     Mortality experience of glass fibre workers—extended follow-up.  Presented
     at: international  symposium on  man-made  mineral  fibres  in the working
     environment;  October;  Copenhagen,  Denmark.  Copenhagen,  Denmark:  World
     Health Organization, Regional Office for Europe.

Shannon, H.  S.;  Jamieson,  E.; Julian, J. A.; Muir, D.  C.  F.;  Walsh, C.  (1957)
     Mortality experience of  glass  fibre workers •• extended  follow-up.  Ann.
     Occup. Hyg.  31: 657-662.

Simonato,  L.;  Fletcher, A.  C.;  Cherrie,  J.; Andersen, A.; Bertazzi, P.  A.;
     Charney,  N.;  Claude,  J.;  Dodgson,  J.; Esteve, J.;  Frentzel-Beyme,  R.;
     Gardner, M.  J.; Jensen, 0.; Olsen, J.; Saracci, R.; Teppo, L.; Westerholrr,,
     P.; Winklemann,  R.; Winter, P.  0.; Zocchetti, C. (1986a) The man-made
     mineral fibers  (MMMF)  European historical cohort study:  extension of the
     follow-up.  Presented  at:  international  symposium on nan-made mineral
     fibres  in the working  environment;  October;  Copenhagen,  Denmark.  Copenha-
     gen, Denmark: World Health Organization, Regional  Office for Europe.

Simonato,  L.;  Fletcher, A.  C.;  Cherrie,  J.; Andersen, A.; Bertazzi, P.  A.;
     Charney,  N.;  Claude,  J.;  Dodgson,  J.; Esteve, J.;  Frentzel-Beyme,  R.;
     Gardner, M.  J.; Jensen, 0.; Olsen, J.; Saracci, R.; Teppo, L.; Westerholm,
     P.; Winklemann,  R.; Winter, P. D.;  Zocchetti,  C.  (1986b) Updating  lung
     cancer  mortality among a  cohort of man-made  mineral  fibre production
     workers in seven European countries. Cancer Lett.  30: 189-200.

Simonato,  L.;  Fletcher, A.  C.;  Cherrie,  J.; Andersen, A.; Bertazzi, P.  A.;
     Charnay,  N.;  Claude,  J.;  Dodgson,  J.; Esteve, J.;  Frentzel-Beyme,  R. ;
     Gardner,  M.  J. ; Jensen,  0.  M. ;  Olsen, J. H. ;  Saracci,  R. ;  Teppo,   L. ;
     Winkelmann,  R.;  Westerholm,  P.;  Winter, P. D.; Zocchetti, C.  (1986c) The
     man-made mineral fiber  European  historical cohort study.  Scand.  J.  Work
     Environ. Health 12(suppl. 1): 34-47.

Simonato,  L.;  Fletcher, A.  C.;  Cherrie,  J.  W.; Andersen, A.;  Bertazzi,  P.;
     Charnay,  N.;  Claude,  J.;  Dodgson,  J.; Esteve, J.;  Frentzel-Beyme,  R.;
     Gardner,  M.  J.; Jensen,  0. M.;  Olsen, J.;  Saracci,  R.; Teppo,  L. ;
     Winkelmann,  R.;  Westerholm,  P.;  Winter, P.  D.; Zocchetti, C.;  Saracci,  R.
     (1987)  The  International  Agency  for Research on Cancer historical cohort
     study of MMMF production workers in  seven European countries:  extension of
     the follow-up. Ann. Occup. Hyg. 31:  603-623.

Smith,  W.  E.;  Hubert,  D.  D.  (1974)  The intrapleural  route as a  means  for
     estimating carcinpgenicity.  In:  Karbe, E.; Park,  J.  E.,  eds.  Experimental
     Lung Cancer: carcinogenesis and bioassays. New York: Springer; pp. 93-101.

Smith,  W.  E.;  Hubert,  D.  D.;  Sobel,  H.  J.  (1980) Dimensions  of  fibres  in
     relation  to  biological  activity.  In:  Wagner,  J.  C.; Davis,  W.,  eds.
     Biological  effects of mineral  fibres  =  Effets biologiques des fibres
     minerales: proceedings of a symposium, v. 1; September 1979; Lyon, France.
     Lyon, France: World Health Organization, International Agency for Research
     on  Cancer;  pp.  357-360.  (IARC scientific  publication  no. 30;  INSERM
     symposia series: v. 92).
July 1988                          9-48         DRAFT—DO NOT QUOTE OR CITE

-------
                     Smith, D. M.;  Ortiz,  L.  W.; Archuleta, R.  (1984) Long-term exposure of
                          hamsters  and  Osborne-Mendel  rats  to  aerosolized 0.45 urn mean  dij
                          fibrous glass. In:  Guthe,  T.,  ed. Biological effects of man-made i
                          fibres: proceedings of a WHO/IARC conference in association with J
                          TIMA,  v.  2; April  1982; Copenhagen, Denmark. Copenhagen, Denmark:
                          Health Organization, Regional  Office for Europe; pp.  253-272.

                     Smith, D. M.;  Ortiz,  L.  W.; Archuleta, R.  F.; Johnson, N. F. (1986) Lo
                          health  effects  In  hamsters and rats exposed  chronically to  man-
                          vitreous  fibers.  Presented at:  International  symposium on  wan-
                          mineral fibres 1n the working  environment; October;  Copenhagen, D
                          Copenhagen, Denmark: World Health Organization, Regional Office fo

                     Smith, D. M.;  Ortiz,  L.  W.; Archuleta, R.  F.; Johnson, N. F. (1987) Lo
                          health  effects  in  hamsters and rats exposed  chronically to  man-
                          vitreous fibers.  Ann. Occup.  Hyg. 31:  731-754.

                     Stadler, J.  C.  (1986)  Two-year inhalation study  in  rats  with Kevlar f
                          [unpublished study],  Michigan:  E.I.  du  Pont de Nemours and  Compan'
                          Report no. 290-86; MR no. 4442-001.

                     Stanton, M.  F.  (1973)  Some etiological considerations of  fibre carcino
                          In: Bogovski,  P.; Gilson, J.  C.; Timbrel!, V.; Wagner, J. C., eds.
                          ical effects  of  asbestos:  proceedings  of  a working  conference; C
                          1972; Lyon, France.  Lyon, France: World Health Organization, Inter
                          Agency  for Research on Cancer; pp. 289-294. (IARC scientific pub!
                          no. 8).

                     Stanton, M.  F.;  Layard,  M.; Tegeris, A.; Miller, E.; May, M.; Kent, I.
                          Carcinogenicity of  fibrous glass:  pleura!  response in the rat in
                          to fiber dimension.  JNCI J. Natl. Cancer Inst.  58: 587-603.

                     Stanton, M.  F.; Layard,  M.;  Tegeris, A.;  Miller, E.;  May,  M.;  Morgar
                          Smith,  A.  (1981) Relation of  particle  dimension  to  carcinogenic
                          amphibole  asbestoses  and other fibrous minerals. JNCI  J.  Natl.
                          Inst.  67:  965-975.

                     Stenback, F.;  Wasenius,  V.-M.; Rowland, J.  (1986)  Alveolar and inte
                          changes in silicate-associated  lung tumors in Syrian hamsters.  C
I                          Mono.  2: 199-213.
I
I                     Suzuki, Y.;  Kohyama,  N.  (1984) Malignant mesothelioma  induced by asb
                          zeolite in the mouse peritoneal cavity.  Environ. Res. 35: 277-29

                     U. S. Environmental Protection Agency. (1986) The risk  assessment gui
                          1986.  Washington, DC:  Office  of Health and Environmental Assess
                          report no. EPA 600/8-87/045.

                     U. S.  Environmental  Protection Agency. (1988)  Interim  methods for  de
                          of inhalation  reference doses  [final  draft]. Research Triangle
                          Environmental   Criteria  and  Assessment   Offices;   EPA  report
                          EPA-600/8-88-066.
                                                                               i,ir\T

-------
Wagner, J.  C.;  Berry,  G.; Timbrell, V. (1969) Mesotheliomas in rats following
     inoculation with asbestos. Br. J. Cancer 23: 567-581.

Wagner, J.  C.; Berry,  G.; Timbrel!, V. (1973)  Mesotheliomas  in  rats after
     inoculation with asbestos and other materials. Br. J. Cancer 28: 173-185.

Wagner, J.  C.;  Berry,  G.; Skidmore, J. W.  (1976) Studies of the  carcinogenic
     effects of fiber fllass  of different  diameters  following intrapleural
     inoculation in  experimental  animals.  In:  Occupational  exposure to fibrous
     glass: proceedings  of a symposium; June 1974; College  Park, MD.  Cincin-
     nati,  OH:  U.  S. Department of Health,  Education,  and Welfare, National
     Institute  for Occupational Safety and Health; pp. 193-197; HEW publication
     no. (NIOSH) 76-151. Available from: NTIS, Springfield, VA; PB-258869.

Wagner, J.  C.;  Berry,  G.; Cooke, T. J.; Hill, R. J.; Pooley, F. D.; Skidmore,
     J. W.  (1977)  Animal experiments with talc.  In: Walton,  W.  H., ed. Inhaled
     Particles  IV: v. 2. Oxford, England: Pergamon Press; pp. 647-654.

Wagner, J.  C.;  Chamberlain, M.; Brown, R. C.; Berry, G.;  Pooley,  F. D.; Davies,
     R.; Griffiths,  D. M.  (1982) Biological effects of tremolite.  Br.  J. Cancer
     45: 352-360.

Wagner, J.  C.;  Berry, G. B.; Hill, R. J.; Munday, D. E.;  Skidmore, J.  W. (1984)
     Animal  experiments  with MMM(V)F--effects of  inhalation  and  intrapleural
     inoculation in  rats.  In:  Cuthe, T., ed. £ictogital  effects  of man-made
     mineral fibres: proceedings  of a: WHO/IARC  conference in association with
     JEMRB  and TINA,  v.  2;  April  1982; Copenhagen,  Denmark.  Copenhagen,
     Denmark:  World  Health  Organization,  Regional  Office  for  Europe;
     pp. 209-233.

Wagner, J.  C. ;  Skidmore, J. W.; Hill, R. J.; Griffiths,  D. M.  (1985)  Erionite
     exposure and mesotheliomas in rats. Br. J.  Cancer 51: 727-730.

Waxweiler,  R.  J.;  Zumwalde, R. D.; Ness, G. 0.; Brown, D. P. (1985) Mortality
     among  persons mining  and milling attapulgite clay.  Cincinnati, OH: U.  S.
     Department of Health  and Human Services, National  Institute for Occupa-
     tional Safety and Health, Centers for Disease Control.

Wehner, A.  P.;  Zwicker,  G. M.; Cannon, W.  C.; Watson, C. R.;  Carlton, W. W.
     (1977)  Inhalation  of  talc baby powder by hamsters.  Food Cosmet.  Toxicol.
     15: 121-129.

Westerholm,  P.;  Bolander,  A.  M. (1986) Mortality and cancer incidence in the
     man-made mineral  fiber industry in Sweden.  In:  Saracci, R.,  ed. Contribu-
     tions  to  the  IARC study on mortality and cancer incidence among  man-made
     mineral fiber  production workers.  Scand.   J.  Work Environ.   Health  12
     (suppl. 1): 7B-B4.
July 1988                          9-50         DRAFT--DO NOT QUOTE OR CITE

-------