6EFK
            United States
            Environmental Protection
            Agency
            Office of Health and
            Environmental Assessment
            Washington DC 20460
EPA-600/8-82-007
November 1982
Revised
            Research and Development
Health Assessment
Document for
Acrylonitrile
REVISED
DRAFT

-------
               UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

                    Environmental Criteria and Assessment Office (MD-52)
::  November 15, 1982

:T  Revised Draft of EPA OHEA Health Assessment
   Document for Acrylonitrile           „       n    U     	~?
                                      i £ ^^^j^^^^  ffi    /f    ^^^^^^^^^^
   Lester D. Grant, Ph.D., Director  X^W^• -^ta'"^
   Environmental Criteria and Assessment Office (MD-52)

   Dr. Roger McClellan, Chairman
   Health Committee,  EPA Science Advisory Board (SAB)

        Attached is a revised draft of  the EPA (OHEA) Health Assessment
   Document for Acrylom'trile, earlier  issued as an External Review Draft
   (EPA-600/8-82-007; March, 1982) for  public comment and  SAB review.

        The attached  (November, 1982) draft of the acrylonitrile document
   incorporates revisions which take into account both:   (1) public comments
   received on the March, 1982, External Review Draft; and  (2) recommendations
   provided by the SAB Health Committee under your chairmanship in connection
   the Committee's peer-review of the March External  Review Draft at the
   August 2-3, 1982,  Health Committee public meeting  in Washington.

        The main revisions which have been incorporated into the attached
   November, 1982, draft include the following:

   1.  The listing of Authors and Reviewers (on pp. xiii-xvii) has been
   revised to indicate clearly the individuals/groups responsible for
   authorship of specific chapters or sections of the document.

   2.   The Summary and Conclusions (Chapter 1) materials  have been revised
   to reflect more accurately (and consistently) key  information and con-
   clusions contained in later, detailed chapters of  the document.  Particu-
   lar emphasis is placed in the Summary and Conclusions Chapter on the
   discussion of key  findings and conclusions regarding studies of muta-
   genic and carcinogenic effects associated with acrylonitrile.  This
   reflects the focus of major attention at this time on the evaluation of
   such effects as likely being of pivotal importance in regard to resolution
   of whether or not  serious health effects are associated with chronic
   low-level exposure to acrylonitrile  in the ambient environment.  There
   have also been added to the Summary  and Conclusions discussion of
   carcinogenicity aspects certain important caveats  regarding the evaluation
   of acrylonitrile as a possible human carcinogen and the estimation by
   the EPA Cancer Assessment Group (CAG) of unit risk values representing
   the plausible upper bound of potential carcinogenic risk associated with
   acrylonitrile exposure.  Specific factors, e.g., the retrospective
   estimation of exposure levels for occupationally-exposed workers studied
   by O'Berg (1980),  contributing to uncertainties associated with the unit
   risk estimates are now more clearly  noted.
   -fj IK-v. 3 761

-------
3.   The air quality chapters addressing "Sources In the Environment"
(Chapter 5) and "Environmental Levels and Exposure" (Chapter 7) have
been updated to include the most recent information available with
respect to sources, emission estimates, and exposure estimates based on
a comparison between monitoring data and dispersion modeling data.  The
references principally used in updating these chapters were those by
R. V. Crume (1982) and B. E. Suta (1982).

4.  In addition to specific air quality chapters being revised and/or
updated there were considerable changes in the Chapter 13 section (13.5)
addressing "Carcinogenicity."  With regard to the epidemiologic aspects
of carcinogenicity (section 13.5.2), discussions of several newly available
studies were added, which included both published and unpublished studies.
The new studies added (unpublished studies are marked with an asterisk)
include:

     1.  Gaffey and Strauss (1981)*
     2.  Kisselbach, et al. (1980)*
     3.  Thiess, et al. (1980)
     4.  Herman (1981)*
     5.  Stallard (1982)*
     6.  Weinch and Carter (1981)

     In regard to the quantitative aspects of carcinogencity (Section
13.5.3) new information was added addressing the procedures and methodology
(Section 15.3.0) involved in the determination of unit risks and the
actual calculations of unit risk for acrylonitrile based on different
animal data sets and one human study (Section 13.5.3.2).  The procedures
and methodology section attempted to more clearly define the philosophy
of doing risk assessment, the basis for using the'linear non-threshold
multistage model, the mathematical formulations describing the latter
model and the selection of data sets from several studies to use in the
model.  There was a brief discussion of alternate methodological approaches
that have been considered by the Carcinogen Assessment Group and a
relative comparison of the carcinogenic potency of acrylonitrile to
other carcinogens based on unit risk calculations.

     Lastly, the CAG appendix was revised by the Carcinogen Assessment
Group and inserted into the main text of the document as Section 13.5
(pp 13-87 through 13-172), instead of appearing as an appendix to the
overall document.
5.  Time limitations precluded  listing at the end of each chapter the
references for that given chapter in the attached revised document
draft.  Rather all references presently appear at the end of the document
in the collated bibliography.  However, this will be changed in the
final document so that references for individual chapters will  appear at
the end of each chapter as appropriate.

-------
     We feel that the above major revisions and other changes appearing
in the attached draft document adequately address the major substantive
issues raised by public comments and/or the SAB Health Committee in the
course of its August, 1982, review of the March, 1982 External  Review
Draft of the EPA OHEA Health Assessment Document for Acrylonitrile.  We
look forward to SAB review of this revised draft at the upcoming December
8-9, 1982, SAB public meeting in Washington and wish to thank you for
the Committee's excellent and helpful recommendations concerning revision
of the earlier draft of the document.
Attachment


 cc:   Health  Committee members

-------
                                           EPA-600/8-82-007
                                              Revised Draft
                                              November 1982
                    DRAFT
            Do not cite or quote
       HEALTH ASSESSMENT DOCUMENT FOR
                ACRYLONITRILE
                   NOTICE

   This document is a preliminary draft.   It
   has not been formally released by EPA and
   should not at this stage be construed to
   represent Agency policy.  It is being
   circulated for comment on its technical
   accuracy and policy implications.
    U.S. ENVIRONMENTAL PROTECTION AGENCY
     Office of Research and Development
Office of Health and Environmental Assessment
Environmental Criteria and Assessment Office
Research Triangle Park, North Carolina  27711

   Project Coordinator:  Dr. Robert Bruce
              November 12, 1982

-------
                                  DISCLAIMER









     The report  is  an external  draft for  review purposes only and  does  not




constitute  Agency Policy.  Mention  of trade names or commercial products does




not constitute endorsement or recommendation for use.
                                       11

-------
                                    PREFACE








     The  Office  of  Health and  Environmental Assessment  has  prepared  this




health assessment  to serve as  a "source  document"  for EPA use.   The  health




assessment  document  was  originally developed for  use by  the  Office  of  Air




Quality Planning  and Standards to  support  decision-making  regarding possible




regulation of acrylonitrile as  a hazardous air pollutant.   However, the scope




of this document has since been expanded to address multimedia aspects.



     In the development  of the assessment document,  the scientific literature




has been  inventoried,  key studies have been evaluated and summary/conclusions




have been prepared so that the chemical's toxicity and related characteristics




are qualitatively  identified.   Observed  effect  levels and other  measures of




dose-response  relationships are  discussed,  where  appropriate,  so  that  the




nature of the adverse health responses are placed in perspective with observed




environmental levels.
                                        111

-------
                                TABLE OF CONTENTS

                                                                         Page

LIST OF TABLES 	     vii
LIST OF FIGURES 	     xii

1.    SUMMARY AND CONCLUSIONS 	     1-1

2.    INTRODUCTION 	     2-1

3.    PHYSICAL AND CHEMICAL PROPERTIES 	     3-1
     3.1. SYNONYMS AND TRADE NAMES 	     3-1
     3.2. STRUCTURAL AND MOLECULAR FORMULAS AND MOLECULAR WEIGHT 	     3-1
                                                                           3-1
                                                                           3-1
                                                                           3-1
                                                                           3-1
                                                                           3-1
                                                                           3-2
                                                                           3-2
                                                                           3-2
                                                                           3-2
                                                                           3-2
                                                                           3-3
                                                                           3-3
                                                                           3-3
                                                                           3-3
                                                                           3-3
                                                                           3-3
                                                                           3-4
                                                                           3-4
                                                                           3-5
                                                                           3-5
                                                                           3-6
                                                                           3-6

     SAMPLING AND ANALYTICAL METHODS 	     4-1
     4.1. AIR 	     4-1
          4.1.1.   Sampling 	     4-1
          4.1.2.   Analysis 	     4-5
          4.1.3.   Conclusions 	     4-9
     4.2. WATER 	    4-10
          4.2.1.   Sampling	    4-10
          4.2.2.   Analysis 	    4-11
          4.2.3.   Conclusions 	    4-14
     4.3. WASTEWATER 	    4-14
          4.3.1.   Sampling 	    4-14
          4.3.2.   Analysis	    4-14
          4.3.3.   Conclusions 	    4-16
3.3.
3.4.
3.5.
3.6.
3.7.
BOND ANGLES AND BOND DISTANCES
PHYSICAL PROPERTIES 	
3.4.1. Description 	
3.4.2. Boiling Point 	
3.4.3. Melting Point 	
3.4.4. Density 	
3.4.5. Refractive Index 	
3.4.6. Spectroscopic Data 	
3.4.7. Solubility 	
3.4.8. Volatility in Water ....
3.4.9. Volatility 	
3.4.10. Stability 	
3.4.11. Octanol-Water Partition
3.4. 12. Conversion Factor 	
CHEMICAL PROPERTIES 	
3.5.1. Reactivity 	
3.5.2. Polymerization 	
3.5.3. Reaction at the Nitrile
3.5.4. Reactions at the Double
3.5.5. Cyanoethylation Reactior
CHARACTERISTICS OF THE CHEMICAL
CONCLUSION 	












Coefficient 	




Groups 	
Bond 	
is 	
PRODUCT 	


-------
                            TABLE OF CONTENTS (cont.)
     4.4. SOIL AND SEDIMENT 	      4-16
          4.4.1.   Sampling 	      4-16
          4.4.2.   Analysis 	      4-16
          4.4.3.   Conclusions 	      4-18
     4.5. RESIDUE IN POLYMERS AND THE EXTENT OF MONOMER
            MIGRATION OF FOOD SIMULATING SOLVENTS 	      4-18
          4.5.1.   Analysis 	      4-19
          4.5.2.   Conclusions 	      4-23
     4.6. OTHER MEDIA 	      4-23
     4.7. GENERAL METHODS FOR THE ANALYSIS OF ACRYLONITRILE 	      4-23

5.    SOURCES IN THE ENVIRONMENT 	       5-1
     5.1. PRODUCTION PROCESSES 	       5-1
     .5.2. ACRYLONITRILE PRODUCERS 	       5-1
     5.3. ACRYLONITRILE USES 	       5-2
     5.4. CONSUMPTION OF ACRYLONITRILE BY PRODUCT 	       5-4
     5.5. SOURCES OF EMISSIONS 	       5-8
          5.5.1.   Monomer and Polymer Production 	       5-8
          5.5.2.   Emissions During Transportation 	      5-12
          5.5.3.   Emissions from End-Product Usage 	      5-12
          5.5.4.   Conclusions 	      5-15

6.    ENVIRONMENTAL FATE, TRANSPORT,  AND DISTRIBUTION	       6-1
     6.1. ATMOSPHERIC FATE, PERSISTENCE, AND TRANSPORT 	       6-1
          6.1.1.   Atmospheric Chemical Reactions 	       6-1
          6.1.2.   Photochemical Reactions 	       6-2
          6.1.3.   Atmospheric Persistence and Transport 	       6-2
     6.2. FATE, PERSISTENCE, TRANSPORT, AND BIOACCUMULATION IN
            AQUEOUS MEDIA 	       6-3
          6.2.1.   Chemical Reactivity in Water 	       6-3
          6.2.2.   Photochemical Reaction in Water 	       6-4
          6.2.3.   Degradation of Acrylonotrile by Microorganisms 	       6-4
          6.2.4.   Bioaccumulation in Water 	       6-7
          6.2.5.   Transport in Water 	       6-8
     6.3. FATE, PERSISTENCE, AND TRANSPORT IN SOIL 	      6-10

7.    ENVIRONMENTAL LEVELS AND EXPOSURE 	       7-1
     7.1. ENVIRONMENTAL LEVELS 	       7-1
          7.1.1.   Atmospheric Levels of Acrylonitrile Around Its
                    Major Production and Usage Facilities 	       7-1
          7.1.2.   Acrylonitrile Levels in Surface Waters 	       7-5
     7.2. ACRYLONITRILE LEVELS IN SOILS AND SEDIMENTS 	       7-7
     7.3. ENVIRONMENTAL EXPOSURE 	       7-7
          7.3.1.   Exposure From Air Polluted by Industrial Sources ...       7-9
          7.3.2.   Exposure From Drinking Water 	       7-9
          7.3.3.   Exposure From Foods  	       7-9
          7.3.4.   Exposure From Spillage During Transportation 	      7-14
          7.3.5.   Exposure From Thermal Degradation 	      7-15

-------
                            TABLE OF CONTENTS (cont.)

                                                                           Page

     7.4  CONCLUSIONS ................................................      7-17

 8 .   BIOLOGICAL EFFECTS ON MICROORGANISMS ............................       8-1

 9 .   BIOLOGICAL EFFECTS ON PLANTS ....................................       9-1

10.   BIOLOGICAL EFFECTS ON DOMESTIC ANIMALS ..........................      10-1

11.   BIOLOGICAL EFFECTS ON WILDLIFE ..................................      11-1
     11.1.   Insects .................................................      11-1

12.   BIOLOGICAL EFFECTS ON AQUATIC ORGANISMS .........................      12-1
     12.1. ACUTE TOXICITY ............................................      12-1
           12.1.1.   Freshwater Fish ..................................      12-1
           12.1.2.   Marine Fish ......................................      12-4
           12.1.3.   Freshwater Invertebrates .........................      12-5
           12.1.4.   Marine Invertebrates .............................      12-5
     12.2. SUBCHRONIC TOXICITY .......................................      12-6
           12.2.1.   Freshwater Fish ..................................      12-6
           12.2.2.   Freshwater Invertebrates .........................      12-7
     12.3. SUMMARY AND CONCLUSIONS ...................................      12-8

13.   BIOLOGICAL EFFECTS IN MAN AND EXPERIMENTAL ANIMALS ..............      13-1
     13.1. PHARMACOKINETICS ..........................................      13-1
           13.1.1.   Absorption and Distribution ......................      13-1
           13.1.2.   Metabolism .......................................      13-7
                    13.1.2.3.   Summary and conclusions ...............     13-16
     13.2. ACUTE, SUBCHRONIC,  AND CHRONIC TOXICITY ...................     13-17
           13.2.1.   Acute Toxicity ...................................     13-17
           13.2.2.   Subchronic Toxicity in Non-Human Mammals .........     13-43
           13.2.3.   Chronic Toxicity in Non-Human Mammals ............     13-45
           13.2.4.   Summary and Conclusions ..........................     13-50
     13.3. TERATOGENI CITY AND REPRODUCTIVE TOXICITY . ..................     13-51
           13.3.1.   Summary and Conclusions ..........................     13-57
     13.4. MUTAGENICITY ..............................................     13-60
           13.4.1.   Gene Mutation Studies ............................     13-60
           13.4.2.   Chromosomal Aberration Studies ...................     13-68
           13.4.3.   Other Tests Indicative of Genetic Damage .........     13-71
           13.4.4.   Summary and Conclusions ..........................     13-75
     13.5. CARCINOGENICITY ...........................................     13-87
           13.5.1.   Animal Studies ...................................     13-87
           13.5.2.   Epidemiologic Studies ............................    13-116
           13.5.3.   Quantitative Estimation ..........................    13-136
                    13.5.3.1.   Procedures and Methodology ............    13-136
                    13.5.3.2.   Acrylonitrile Risk Estimates ..........    13-151
           13.5.4.   Summary ..........................................    13-164
           13.5.5.   Conclusions ......................................    13-172
REFERENCES
                                       VI

-------
                                LIST OF TABLES

Table
 No.                                                                       Page

 4-1      Various Sorbents and Trapping Media for Collection
            of Acrylonitrile in Air 	       4-3

 4-2      Recovery of Acrylonitrile from Various Solvents 	       4-7

 4-3      Direct Analysis of Acrylonitrile 	       4-8

 4-4      Analyses of Acrylonitrile in Water 	      4-15

 4-5      Analysis of Acrylonitrile in Wastewaters 	      4-17

 4-6      Analysis of Acrylonitrile Residue in Polymers and
            Food-Simulating Solvents 	      4-20

 5-1      Producers of Acrylonitrile in the United States 	       5-2

 5-2      Distribution of Acrylonitrile in 1980 and Projected
            Growth Through 1984 	       5-2

 5-3      Primary Uses for Compounds Synthesized from
            Acrylonitrile-Containing Compounds 	       5-3

 5-4      Acrylonitrile Consumption and Project Growth of
            Products Using Acrylonitrile 	       5-4

 5-5      Producers of SAN and ABS Resins 	       5-6

 5-6      Producers of Acrylic and Modacrylic Fibers 	       5-6

 5-7      Producers of Nitrile Rubbers and Elastomers 	       5-7

 5-8      Producers of Acrylamide 	       5-7

 5-9      Estimated Atmospheric Emissions of Acrylonitrile from
            Monomer Production Facilities 	       5-9

 5-10     Estimated Acrylonitrile Emission Rates from ABS-SAN
            Resin Production 	      5-10

 5-11     Estimated Acrylonitrile Emission Rates from Acrylic
            Fiber Production 	      5-10

 5-12     Estimated Acrylonitrile Emission Rates from Nitrile
            Elastomer Production 	      5-11
                                      VII

-------
                            LIST OF TABLES (cont.)
Table
 No.
 5-13     Estimated Acrylonitrile Emission Rate from Adiponitrile
            Production 	      5-11

 5-14     Hazards of Acrylonitrile Transportation 	      5-11

 5-15     Monomer Residue in End-Products of Acrylonitrile 	      5-14

 7-1      Atmospheric Monitoring Data for Acrylonitrile 	       7-2

 7-2      Comparison of Monitoring and Dispersion Modeling Data 	       7-4

 7-3      Acrylonitrile Monitoring Data for Surface Waters 	       7-6

 7-4      Acrylonitrile Monitoring Data for Sediments 	       7-8

 7-5      Acrylonitrile Monitoring Data for Soils 	       7-8

 7-6      Estimates of Population Exposures to Atmospheric
            Acrylonitrile from Specific Emission Source
            Categories 	      7-10

 7-7      Acrylonitrile Migration Under Different Storage
            Conditions 	      7-12

 7-8      Amounts of Various Acrylonitrile Copolymers Used in
            Food-Contact Applications 	      7-13

 11-1     Lethal Dose Values for Insects Exposed to Acrylonitrile
            Fumigation 	   11-2

 12-1     Median Lethal Concentration (LC50) Values for Fish
            Exposed to Acrylonitrile	      12-2

 13-1     Recovery of Radioactivity from Rats Given Single Oral
            Doses of 0.1 or 10 mg/kg 14C-Acrylonitrile 	      13-2

 13-2     Recovery of Radioactivity from Rats Exposed by
            Inhalation to 5 or 100 ppm l4C-Acrylonitrile for
            6 hours 	      13-3

 13-3     Distribution of Radioactivity in Selected Tissues of
            Rats Given 14C-Acrylonitrile 	      13-5
                                     Vlll

-------
                            LIST OF TABLES (cont.)

Table
 No.                                                                       Page

13-4      Urinary Metabolites Following the Oral Administration
            of C-l (Cyano) Labeled Acrylonitrile 	      13-14

13-5      Metabolites of 14C-Acrylonitrile Separated from
            Various Fluids of Rats by High Pressure Liquid 	      13-18

13-6      Summary of Results of Exposures of Rats to
            Acrylonitrile 	      13-22

13-7      Minimal Lethal Concentration of Acrylonitrile During
            Four-Hour Exposure 	      13-24

13-8      Comparison of the Effects of Acrylonitrile and of
            Hydrocyanic Acid on Various Species of Animals 	      13-26

13-9      Cyanide and Thiocyanate in Blood of Animals Exposed
            to Acrylonitrile 	      13-28

13-10     Effect of Methemoglobinemia on Mortality Ratios in
            Albino Rats Poisoned with Acrylonitrile, Potassium
            Cyanide, and Acetone Cyanohydrin 	      13-33

13-11     Therapeutic Effect of SH and S-S Compounds on Acute
            Acrylonitrile Poisoning 	      13-38

13-12     Concentration of Protein (PBSH) and Nonprotein (NPSH) SH
            Groups in Normal and Acrylonitrile-Intoxicated
            Animals 	      13-39

13-13     Cumulative Mortality of Male and Female Rats Maintained
            for Two Years on Drinking Water Containing
            Acrylonitrile 	 	      13-49

13-14     Significant Changes Considered to be Secondary to
            Ingestion of Acrylonitrile 	      13-52

13-15     Incidence of Fetal Malformation Among Litters of Rats
            Given Acrylonitrile 	      13-54

13-16     Pup Weight on Days 4 and 21 of Lactation  	      13-58

13-17     Tissues Examined for Histopathologic Changes in the F3b
            Litter 	      13-59

-------
                            LIST OF TABLES (cent.)

Table
 No.                                                                        Page

13-18          Mutagenicity Tests of Acrylonitrile  	      13-78

13-19          Cumulative Mortality Data of Male Rats Maintained
                 for 2 Years on Drinking Water Containing
                 Acrylonitrile 	      13-89

13-20          Cumulative Mortality Data of Female  Rats  Maintained
                 for 2 Years on Drinking Water Containing
                 Acrylonitrile 	      13-90

13-21          Histopathologic Diagnoses and Tumor  Incidences
                 in Male Rats Maintained for 2 Years  on  Drinking
                 Water Containing Acrylonitrile 	      13-91

13-22          Histopathologic Diagnoses and Tumor  Incidences  in
                 Female Rats Maintained for 2 Years on Drinking
                 Water Containing Acrylonitrile 	      13-92

13-23          Tumor Incidences in Sprague-Dawley Rats Fed
                 Acrylonitrile in Drinking Water 	      13-96

13-24          Tumor Incidence in Fischer 344 Rats  Fed
                 Acrylonitrile in Drinking Water 	      13-98

13-25          Incidence of Tumors Observed in Rats During a
                 Three-Generation Reproductive Study 	     13-100

13-26          Tumor Incidence in Rats Fed Acrylonitrile Orally
                 by Intubation 	     13-103

13-27          Tumor Incidence by Rats Following Inhalation
                 of Acrylonitrile	     13-104

13-28          Stability and Trace Impurity Analysis  of  the
                 Acrylonitrile Liquid Test 	     13-106

13-29          Chamber Concentrations of Acrylonitrile Vapors  	     13-107

13-30          Cumulative Mortality Data of Male Rats Exposed
                 by Inhalation for 2 Years to Acrylonitrile Vapors ...     13-108

13-31          Cumulative Mortality Data of Female  Rats  Exposed
                 by Inhalation for 2 Years to Acrylonitrile Vapors ...     13-109

13-32          Tumor Incidence in Sprague-Dawley Rats Exposed
                 to Acrylonitrile by Inhalation 	     13-110

-------
                            LIST OF TABLES (cont.)

Table
 No.                                                                        Page

13-33          Enhancement of SA7 Transformation by Treatment of
                 HEC with ACN 	      13-113

13-34          Transformation of HEC by ACN 	      13-115

13-35          Number (Percentage) of Rats Developing Tumors in
                 At Least One of the Following Target Organs:
                 Zymbal Gland, Tongue, Stomach, Brain, and
                 Spinal Cord by Dose and by Sex 	      13-155

13-36          Relative Carcinogenic Potencies Among Suspect
                Carcinogens Evaluated by the Carcinogen
                Assessment Group 	      13-161

13-37          Carcinogenicity of Acrylonitrile in  Rats 	      13-165

13-38          Epidemiologic Studies Reviewed in Acrylonitrile
                 Risk Assessment 	      13-168
                                      XI

-------
                                LIST OF FIGURES

Figure
  No.                                                                      Page

 5-1      Flow Diagram for Acrylonitrile 	       5-5

 6-1      Biological Oxidation of Acrylonitrile in Aqueous Systems ...       6-5

13-1      Proposed Pathways for Acrylonitrile Biotransformation 	     13-10

13-2      Proposed Scheme for Metabolism of Acrylonitrile by
            the Rat 	     13-12

13-3      Distribution of Acrylonitrile, Cyanide, and Thiocyanate
            in the Blood after a Single Injection of
            Acrylonitrile 	     13-31

13-4      Effect of Sodium Thiosulfate on the Distribution of
            Acrylonitrile, Cyanide, and Thiosulfate 	     13-32

13-5      Distribution of Acrylonitrile, Cyanide, and Thiocyanate
            in the Blood after a Single Injection of
            of Acrylonitrile 	     13-35

13-6      Effect of E-Cysteine on the Blood Concentrations of
            Acrylonitrile, Cyanide, and Thiocyanate 	     13-36

13-7      Histogram Representing Frequency Distribution of the
            Potency Indices of 54 Suspect Carcinogens Evaluated
            by the Carcinogen Assessment Group 	    13-160
                                      XII

-------
     The EPA office of  Health and Environmental Assessment (OHEA) is  respon-

sible for the preparation  of  this health assessment document.   The OHEA Envi-

ronmental Criteria and Assessment Office (ECAO/RTP)  had overall  responsibility

for  coordination  and direction  of the  document,  preparation and production

effort (Dr.  Robert M.  Bruce,  Project Manager).   The  chapters addressing physi-

cal and chemical  properties,  sampling  and analysis, air quality  and  toxicity

data were written by  Syracuse Research Corporation.  The principal authors  of

these chapters are listed below.
     Dr. Dipak K.  Basu
     Life and Environmental Sciences Division
     Syracuse Research Corporation
     Syracuse, New York

     Dr. Robert S. Hsu
     Life and Environmental Sciences Division
     Syracuse Research Corporation
     Syracuse, New York

     Dr. Michael W. Neal
     Life and Environmental Sciences Division
     Syracuse Research Corporation
     Syracuse, New York

     Mr. Joseph Santodonato
     Life and Environmental Sciences Division
     Syracuse Research Corporation
     Syracuse, New York

     Dr. Richard H. Sugatt
     Life and Environmental Sciences Division
     Syracuse Research Corporation
     Syracuse, New York
     The OHEA Carcinogen Asssessment  Group (CAG) was responsible for prepara-

tion of the sections on carcinogenicity.   Participating members of the CAG are

listed  below  (principal  authors  of  present  carcinogenicity materials  are

designated by *).
                                  Xlll

-------
     Roy Albert,  M.D.  (Chairman)
     Elizabeth L.  Anderson,  Ph.D.
     Larry D.  Anderson,  Ph.D.
     Steven Bayard, Ph.D.*
     David L.  Bayliss,  M.S.*
     Chao W.  Chen, Ph.D.
     Herman J. Gibb, M.S.,  M.P.H.
     Bernard H.  Haberman,  D.V.M.,  M.S.
     Charalingayya B.  Hiremath,  Ph.D.*
     Robert McGaughy,  Ph.D.
     Dharm V.  Singh, D.V.M., Ph.D.
     Todd W.  Thorslund,  Sc.D


     The OHEA Reproductive Effects Assessment Group (REAG)  was responsible for

the preparation  of sections on mutagenicity.   Participating members  of  REAG

are  listed below  (principal  authors of  present  mutagenicity sections  are

indicated by *).


     Vicki Vaughn-Dellarco,  Ph.D.*
     Jack R.  Fowle III,  Ph.D.
     K.S. Lavappa, Ph.D.
     Sheila Rosenthal,  Ph.D.
     Carol Sakai,  Ph.D.
     Peter Voytek, Ph.D.
                                  xiv

-------
     The following individuals  provided  peer-review of  this  draft  or  earlier
drafts of this document:

     U.  S.  Environmental  Protection Agency

          Donald Barnes,  Ph.D.
          Office of Toxic Substances

          David Berg
          Office of Environmental
          Engineering and Technology

          H.  Matthew Bills
          Acting Director for Monitoring Systems
          and Quality Assurance
          Office of Research and Development

          Robert M. Bruce, Ph.D.
          Office of Health and  Environmental Assessment
          Environmental Criteria and Assessment Office

          Lester D. Grant, Ph.D.
          Office of Health and  Environmental Assessment
          Environmental Criteria and Assessment Office

          Gary E.  Hatch,  Ph.D.
          Office of Health Research
          U.S. EPA

          Richard N. Hill, M.D., Ph.D.
          Office of Toxic Substances
          U.S. EPA

          Raelyn Janssen
          Office of Health Research
          U.S. EPA

          Steven Nesnow,  Ph.D.
          Office of Health Research
          U.S. EPA

          Joseph Padgett
          Office of Air Quality Planning and Standards
          U.S. EPA

          Nancy Pate, D.V.M.
          Office of Air Quality Planning and Standards
          U.S. EPA

          Shabeg Sandhu,  Ph.D.
          Office of Health Research
          U.S. EPA

          Jerry F. Stara, D.V.M.
          Office of Health and  Environmental Assessment
          U.S. EPA
                                   xv

-------
     John Todhunter, Ph.D.
     Assistant Administrator
     Office of Toxic Substances
     U.S. EPA

     Herbert Wiser, Ph.D.
     Acting Director
     Office of Environmental
     Engineering and Technology
     U.S. EPA

Other Agencies

     William C. Brumley, Ph.D.
     Bureau of Foods
     Food and Drug Administration
     Health and Human Services

     Michael T. Flood, Ph.D.
     Bureau of Foods
     Food and Drug Administration
     Health and Human Services

     Chiu A. Linn, Ph.D.
     Toxicology Division
     Food and Drug Administration
     Human Health Services

     T. P. McNeal, Ph.D.
     Bureau of Foods
     Food and Drug Administration
     Health and Human Services

     Terry Troxell, Ph.D.
     Division of Food and Color Additives
     Food and Drug Administration
     Health and Human Services

Consultants and/or Reviewers

     Julian B. Andelman, Ph.D.
     Professor of Chemistry
     Graduate School of Public  Health
     University of Pittsburgh
     Pittsburgh, Pennsylvania

     Rudolph J. Jaeger, Ph.D.
     Consulting Toxicologist
     7 Bogert Place
     Westwood, New Jersey

     George Hoffman, Ph.D.
     Professor, Biology Department
     College of Holy Cross
     Worchester, Maine

                             xvi

-------
Edmond J. LaVoie, Ph.D.
Head, Metabolic Biochemistry Section
Naylor Dana Institute for Disease Prevention
American Health Foundation
Valhalla, New York

Richard R. Monson, M.D.,  Sc.D.
Associate Professor of Epidemiology
School of Public Health
Harvard University

Daniel Straus, Ph.D.
Associate Professor
Biomedical Science and Biology
University of California
Riverside, California

Gary Williams, M.D.
Associate Director
Naylor Dana Institute for Disease Prevention
American Health Foundation
Valhalla, New York
                        xvil

-------
                          1.  SUMMARY AND CONCLUSIONS



     Acrylonitrile is a clear, colorless, and highly  flammable liquid that has an



unpleasant and  irritating  odor  (Fassett,  1963).   The boiling  point  of acrylo-



nitrile is 77.3CC; the melting point is -83.55°C,  and the density of the liquid



at 20°C is 0.8060 (Groet, 1978).  Acrylonitrile is soluble in water between 7.2



and 9.1 weight %  at  temperatures  of  0°C and 60°C, respectively.   The  open cup



flash point of acrylonitrile is  0°C,  and the explosive limits are between 3.0 and



17$ by volume in air  at 25°C (Steere, 1968).  Synonyms for acrylonitrile include



2-propenenitrile,  cyanoethylene,   and  vinyl  cyanide.    Acrylonitrile  has  a



molecular weight  of  53-06  and  a  molecular formula of  C-,H,N.   The  structural



formula is given below.




                                 HxC=C-C=N
                                 H'     L~n .
                                     H





     Acrylonitrile monomer production capacity in the United States is approxi-




mately 1,128,000 million grams.  Of  the  862,000  million grams of acrylonitrile




produced in 1980, approximately 77$ (661,000 Mg)  will be used domestically;  the




remainder is  exported  (CMR,  1980).   Acrylonitrile is used primarily  as  a  raw




material  in  the synthesis  of  acrylic  and  modacrylic  fibers,  acrylonitrile-




butadiene-styrene (ABS)  and sytrene-acrylonitrile (SAN)  resins, nitrile rubbers,




adiponitrile, acrylamide, and barrier resins (Suta, 1979) • A small percentage of




the acrylonitrile produced is used as a chemical intermediate.




     Acrylonitrile is emitted to the atmosphere during monomer production,  poly-




mer production, transport,  and end-product usage; however, the major sources of




acrylonitrile emissions are monomer and  polymer  production  facilities.  Of the




estimated total of 3,856 Mg emission  per  year, monomer ABS-SAN resin, and arylic




fiber production facilities emit 802 Mg,  1H2H Mg,  and  1276  Mg,  respectively of
                                     1-1

-------
acrylonitrile  in the  atmosphere  per  year.    The  atmospheric  half-life  of




acrylonitrile has been estimated to be between 9 and 10 hours (Suta, 1979),  which




is  long  enough to  allow  transport of  acrylonitrile from emission sources  to




nearby populations.   In natural waters, acrylonitrile may  be  removed  by  either




chemical decomposition or microbial degradation (Going  et  al.,  1979;  Mills and




Stack, 1955).  Evaporation may also lower the  concentration of acrylonitrile in




water;  the  calculated half-life  for acrylonitrile  by  evaporation alone from




water 1 m deep  is 795 minutes (Dilling,  1977).   The percent  of  acrylonitrile




removal through each of the processes will depend  on the characteristics  of the




aquatic system.  For example, in biologically  active treatment ponds,  less than




0.1$ of the acrylonitrile was found to be removed  from the aquatic media  due to




volatilization  while  most  of  the  acrylonitrile  removal   took place  via




biodegradation.




     The acrylonitrile levels  in  the vicinity of  acrylonitrile production and




polymer manufacturing plants were investigated by  Hughes  and Horn (1977)  and




Going et al. (1979).  In general,  arylonitrile levels around user  plants  may be



greater than around  producing plants.   Acrylonitrile was detected  in the  air at




distances up to  5  km from  a user  facility;  however,  the  concentrations  of




acrylonitrile were  dependent on meteorological conditions and the  production




stage within the plant at  the time of sampling.  No acrylonitrile was detected in




the soil near these  plants.   Variable low levels of acrylonitrile were  generally



detected in the water downstream from the plants, except for high levels of 35 to




1300 ug/1  detected   in   some  samples   near   wastewater  discharge   points.




Acrylonitrile has also been detected  in  drinking water, although the levels were




not quantified (Kopfler £t al.,  1976).  The inhalation exposure of acrylonitrile




in the vicinity of a plant site  estimated  by dispersion modelling does not agree




well with the experimental monitoring data obtained from the same site  (Going et^
                                   1-2

-------
al., 1979; Suta,  1979).   There  are insufficient data with which to determine the


human intake of acrylonitrile through food and drinking water.


     Limited data  suggest that both  aerobic  and anaerobic microorganisms  are


capable  of  degrading  acrylonitrile,  especially  acclimated   microorganisms.


Certain isolated bacteria can tolerate 10,000  ppm acrylonitrile and use it as a


sole source of nitrogen.  In natural water, a concentration of 50 ppm may inhibit


aerobic microbial  degradation  of  acrylonitrile.   The  breakdown products  of


aerobic microorganisms  may include  ammonia  and organics,  followed by  nitri-


fication of ammonia.


     Acrylonitrile has been shown to affect some terrestrial and aquatic plants


at exposure concentrations of 9 to  100 mg/1,   Acrylonitrile is  toxic to aquatic


animals at exposure concentrations in  the  low milligrams per  liter range.  The


reported acute  LC50 values for  fish ranged between 10.1 and 70  mg/1.  Subchronic


exposure of fish  for 30  to 100  days resulted in LC50 values of about 2 mg/1, with


no evidence that a threshold concentration had been reached.  Although the only


tested invertebrate, Daphnia magna,  had the lowest acute LC50  value (7.6 mg/1),


this species was  not adversely  affected by chronic exposure to 3.6 mg/1 through-


out its whole life cycle.


     Use of acrylonitrile as a fumigant has shown that the vapor concentrations


required to kill 95% or more of many species  of pest insects is between about 1


and 10 mg/1.  No other  information was  found  concerning the effects of acrylo-
                         •

nitrile on wildlife.


     Acrylonitrile is  readily absorbed in animals following ingestion or inhala-


tion, while dermal  absorption  is poor and occurs at  about  1f of that  of the


lungs.  Following  absorption of radiolabeled  acrylonitrile, the radioactivity


disappears in a biphasic manner, with a half-life for the first phase of 3.5 to


3.8 hours  and  the  second phase of  50 to 77  hours  (Young et al_.,  1977).   The
                                       1-3

-------
predominant route of elimination is through the urine.   The routes of elimination


are dose-related  with the percent  eliminated through  the urine  less  for small


doses  as  compared  to  larger doses, while  the relative  amount  retained  by  the


carcass is  greater for  the  small dose  as  compared to  a larger  dose.   Acrylo-


nitrile is  metabolized to cyanide,  which is  transformed to  thiocyanate  and by


cyanoethylation  of sulfhydryl  groups  to  S-(2-cyanoethyl)cysteine,   followed by


elimination  of these  metabolites in  the urine.   Other  minor  metabolites  are


formed from  acrylonitrile.   The  toxicity of  acrylonitrile is  caused  by both the


acrylonitrile molecule itself and its metabolites.


     Acrylonitrile  intoxication  in humans  results  in irritation of  the  eyes  and


nose,  weakness,  labored  breathing,  dizziness,  impaired  judgement,  cyanosis,

                                                                   3
nausea,  and  convulsions.   The  TLV  of  acrylonitrile   is  45  mg/m   for  humans.


Acrylonitrile  also  causes  severe burns  to the  skin.   In experimental  animals,


there is considerable species variation in susceptibility to  acrylonitrile intox-


ication; the  guniea  pig  is the most resistant and the  dog is the most sensitive.


In animals, effects of intoxication include respiratory changes,  cyanosis, convul-


sions, and  death.   In rats,  the LD50  for  acrylonitrile  is  between 80  and  113


mg/kg  (Knoblock et  al.,  1971;  Smyth et  al.,  1969).  There is some evidence that


acrylonitrile  produces  abnormal function  of both  the  peripheral   and  central


nervous systems  and  that  acrylonitrile  causes  damage  to  the  adrenals.   With


subchronic  exposure  of animals  to  acrylonitrile,  some signs  of functional dis-


orders of  the  liver and  kidney  are  observed.  Chronic  exposure  of dogs and rats


results in  unthrifty appearance, weight loss, and  early death.  Some  of these


signs  may  be  related  to  low  food  and water consumption  resulting  from the  un-


pleasant taste of  acrylonitrile  in  the water.  Pathological  changes  in the rats


believed to  be treatment  related included hyperplasia and hyperkeratosis of the
                                       1-4

-------
squamous epithelium of  the  nonglandular portion of the stomach,  proliferation




of glial cells in the brain, and mammary gland hyperplasia in females.




     Acrylonitrile  adversely  affected  pup   survival  following  exposure  of




pregnant rats  and,  in  one  study, produced  teratogenic  events.    In a  three-




generation  study  in which  rats  were  exposed to 500 ppm  acrylonitrile  in the




drinking water, there was reduced pup survival in the first generation  (Beliles




et al.,  1980).   This  was a maternal effect  inasmuch  as  fostering the  pups on




untreated dams  eliminated the  poor  survival.  Reproductive  capacity was un-




changed in  the  other  generations, and the offspring showed no adverse  effects




on development.   Similarly, rats  exposed by  inhalation  to  40  or 80 ppm  of




acrylonitrile for 6 hours  a day on days 6 to 15 of gestation had no statisti-




cally significant changes in reproductive success or fetal development  (Murray




et al. ,  1978).   Only the pups  of rats administered acrylonitrile  per  ££ (65




mg/kg) for  days  6  to  15 of gestation had an  increase in malformations  (Murray




et al., 1976).  This increase was in total malformations,  with no statistically




significant  increase  occurring  in any single malformation.   It  was concluded




that   these  fetal abnormalities  were  the result of acrylonitrile  and not the




result of toxicity  in the dams.  Although several studies have been conducted




to evaluate the ability of acrylonitrile to cause adverse tetratogenic,  embryo-




toxic, and  reproductive  effects,  the  limitations of the available data  do not




allow  for  a full assessment of these  effects.  However, there  are data that




indicate that acrylonitrile  has the  potential to cause fetal malformations in




rats, in a  non dose-related manner.




     There   is  evidence  that acrylonitrile  as  an epoxide  metabolite  causes




point mutations  in  bacterial test systems (Vemitt et  al.  1977,  De Meester et




al.  1978, and  Kier  1982) and there is  suggestive evidence that acrylonitrile




produces a  positive response in the sex-linked recessive lethal mutation assay







                                       1-5

-------
in Drosophila Melanogaster  (Benes  and Sram 1969).   Chromosomal damage was  not




detected in plants or whole animals when treated with acrylonitrile.   In vitro




DNA binding studies indicate that acrylonitrile in the presence of a  rat liver




activation system  has  the  potential to react with DNA  as  well as its epoxide




metabolite,  2-cyanoethylene oxide  (Guengerich  et  al.,  1981).   Acrylonitrile




also  induces  sister chromated  exchange  in CHO  cells but  requires  metabolic




activation  (Ved  Brat and  Williams,  1982).  From  the data base  available it




appears that  acrylonitrile may  have  the  potential  to  bind to DNA  and cause




genetic damage.   If  the  pharmacokinetics of this chemical substance  in humans




results in  metabolic products  that can  intereact  with DNA, it may  cause  so-




matic mutations in humans.




     Acrylonitrile is not  a direct acting carcinogen and  hence  the  localiza-




tion and nature  of the effects depend on its metabolism.  It is probable that




the proximal  carcinogen  is  2-cyanoethylene  oxide,  since  it  has  been demon-




strated as  a reaction product  with calf  thymus DNA. However,  the metabolite




has not been tested directly for its carcinogenicity.  It has been shown to be




produced  in the  liver  and possibly  circulates  to  other organs.   However,




studies have not been done to determine where else in the body this metabolite




is  produced.   There  appears to  be  a  clear  difference between  animals  and




humans  in  their tumorigenic  response to  acrylonitrile:   no  lung tumors have




been  produced  in  animals  and no  brain  tumors have  been  observed in humans.




There are no human studies on the metabolism of acrylonitrile and there are no




pharmacokinetic  studies  that  would  be  relevant  to the  characterization of




dose-response relationships at low levels of exposure.




     The  carcinogenicity of  acrylonitrile has been studied  in  seven cancer




bioassays in rats:  four in drinking water, one by gastric intubation, and two




by  inhalation.   In  addition, ten  epidemiologic  studies  of  cancer  incidence




have been  reported.   A short description  of these studies is presented below.






                                        1-6

-------
     Quast  et  al.  (1980a)  administered  acrylonitrile  in  drinking water  to




Sprague-Dawley rats  for 2  years  at dose  levels  of  35,  100, and  300  ppm.   A




statistically  significant   incidence  of tumors  was  observed  in  the  central




nervous  system,  Zymbal gland,  stomach,  tongue,  and small  intestine  in  both




male and female  rats,  as  well as  in  the  mammary gland of  female  rats.   The




occurrence of central nervous  system and Zymbal gland tumors in Sprague-Dawley




rats was  further  confirmed  in four other  studies:   a  three-generation repro-




duction  study  performed at Litton-Bionetics  by  Beliles et  al.  (1980);  three




studies by Biodynamics, Inc. (1980a, b, c)  in which acrylonitrile was adminis-




tered in drinking water and via gastric intubation; and an inhalation study by




Quast et al. (1980b).




     A second inhalation study by Maltoni et al.  (1977) exposed rats to atmos-




phere containing 5,  10,  20, and 40 ppm  acrylonitrile  4  hours per day, 5 days




per week, for 12  months.  Marginal increases in tumors  of the mammary gland in




females  and  the  forestomach in males were observed, although  the sensitivity




of  this  test  was  limited  by the  relatively  low dose  levels  and  the  short




duration of exposure.




     Ten epidemiologic  studies  of  the  association between acrylonitrile expo-




sure and cancer  incidence  have been reported:   five  published [Monson (1981),




O'Berg  (1980), Thiess et al.  (1980),  Werner  and  Carter  (1981),  Delzell and




Monson (1982] and  five unpublished [Gaffey and Strauss (1981), Herman (1981),




Kiesselbach  et al.   (1980), Stallard  (1982),  and  Zack (1980].   Six of these




studies  present  no evidence  of carcinogenic  risk from  exposure  to acryloni-




trile.    However,  all  suffer from problems  in the design or methodology, i.e.,




small cohort size, insufficient characterization  of exposure, short follow-up,




and  relatively  youthful  cohort.    Because  of these probelms  none of  these




studies  can  be cited as adequate evidence  that  acrylonitrile is not carcino-




genic.






                                        1-7

-------
     Data presented  in the remaining four epidemiologic  studies  consistently




demonstrate a  statistically significant  risk  of lung cancer  in  various  sub-




groups of  the populations  studied.   All  four have problems with  the  method-




ology, definition, and/or  size  of the population, whether or  not exposure to




other  carcinogens  occurred, and  short  follow-up intervals.    In  three  of the




four  studies   [Delzell  and Monson  (1982),  Thiess et  al.  (1981), Werner and




Carter (1981)], the  problems  were sufficient to  cast  doubt  on the finding of




significantly  elevated  risks  of  lung  cancer reported in each study.   In the




fourth study by O'Berg,  the problems were insufficient to obscure the  signif-




icant finding of lung cancer.   After adjusting for latent factors  and evaluat-




ing the  contribution  due  to smoking, the finding of  a statistically signifi-




cantly elevated risk of lung cancer remained.  Thus,  one study  appears  adquate




and three  are suggestive,  while  the remaining six are  inadequate to  address




the issue of a risk of lung cancer.




     In addition to  lung  cancer,  two other findings  of concern are the signi-




ficantly elevated  risk of lymph system cancer found  in the Thiess  et  al. study




(4 observed versus 1.38 expected,  P <0.05)and the significantly elevated risk




of stomach cancer  found  in the Werner and Carter study (5 observed versus 1.9




expected, P < 0.05).   These findings provide  additional suggestive evidence of




the carcinogenicity of acrylonitrile.




     This level of animal  evidence would  be  regarded as "sufficient" evidence




of  carcinogenicity  according  to  the  International  Agency for  Research  on




Cancer (IARC)  classification  scheme.   The human evidence for the  carcinogeni-




city of  acrylonitrile  would be  regarded as somewhere between "sufficient" and




"limited",  using  the  IARC classification.  Therefore, in  combining  the human




and animal  evidence,  acrylonitrile  would be placed  in group  2A, which IARC




characterizes  as  "probably  carcinogenic  in  humans,  where  the  evidence for




human carcinogenicity is almost sufficient".





                                        1-8

-------
     To provide a  rough  estimate  of the potency of  acrylonitrile  relative to


other  chemicals  and a crude  indication of  population  risks associated  with


known exposures, unit  risks  estimates  have been calculated.  The  unit  cancer


risk for air is defined as the lifetime cancer risks  occurring if an individual

                                    3
is exposed  to  air  containing 1 (Jg/m  continuously for a lifetime.   The  linear


non-threshold  dose extrapolation  model has  been used  to  give a  rough  but


plausible upper-bound  of risk; that  is,  the  true  risk  is  not likely  to be


higher  than the estimate but  could be  lower.   For  a discussion of  the  limi-


tations and uncertainties of  the  procedure the reader  is  referred to Section


13.5.3.1.


     Three  unit  risk  estimates for air are  calculated; one based on a  human


occupational  study (O'Berg  190a,  b) and  two based  on rat  cancer  bioassays


(Quast et al.  1980a,  b).  The upper-bound lifetime  risk  of cancer associated


with a  lifetime inhalation  exposure of  ±  (Jg/rn  is  6.8 x  10    from  the  human


study  and  1.5  x 10   from the  rat  inhalation study.  The value based  on the


rat drinking water study is  1.2 x  10    (or  2 x 10   if  the  equivalent  human


dose is  assumed to be mg/kg/day rather than surface area) but  this  study is


less reliable because  of  the inappropriate route of  exposure.


     The estimate  based  on  the human study  is  uncertain  because of the rela-


tively weak documentation of  the  available exposure  estimates of the acrylon-


itrile workers.  The  air concentration  had not been  measured when the workers


experienced their  heaviest exposure and was  estimated  12  years  after the end


of the  exposure period.   However,  in  spite  of these difficulties,  the  esti-


mates are consistent with those of the  animal studies.


     The upper-bound  risk estimate  for 1 (Jg/1  of  acrylonitrile  in drinking


water  is 1.2  x 10   ,  based on  the  Quast,  et al.  (1980a) drinking water study


in rats.
                                        1-9

-------
     There is  evidence  that acrylonitrile  is a human carcinogen.   This  con-




clusion is based on:  1)  findings  of three positive drinking water  rat  bio-




assays and one positive  rat gastric intubation study; 2)  statistically signif-




icant positive  findings  of respiratory cancer in  four epidemiologic studies;




3) the positive  mutagenic  evidence in bacteria and  sister  chromated exchange




tests; 4)  .in vitro  evidence  of interaction acrylonitrile  and/or  its metabo-




lites with DNA;  and  5)  acrylonitrile's structural similarity to vinyl  chlor-




ide,  a known  animal and  human carcinogen.




     The carcinogenic potency of  acrylonitrile is in the third  quartile among




54 suspected  carcinogens evaluated by the  Carcinogen Assessment  Group.




     Using the  International Agency for Research on Cancer (IARC)  classifica-




tion scheme,   this  level  of evidence in animals and humans would be considered




sufficient for  concluding  that  acrylonitrile  is likely to be a  human carcino-




gen with rank of 2A.
                                        1-10

-------
                               2.   INTRODUCTION







     EPA's Office of Research and Development has prepared this health assess-



ment to serve  as  a "source document" for Agency use.   This health assessment



was originally developed  for use  by the Office  of  Air Quality Planning and



Standards to support decision-making regarding possible regulations of acrylo-



nitrile under  Section  112 of the Clean Air  Act.   However,  based on the ex-



pressed interest  of other  agency  offices,  the scope  of  this  document was



expanded to address  acrylonitrile  in relation to sectors of the environment



outside of air.  It is fully expected that this document will  serve the infor-



mation needs of many government agencies and private  groups that may  be in-



volved in decision-making activities related to acrylonitrile.



     In the development of the assessment document, existing scientific liter-



ature has  been surveyed  in detail.   Key studies have been evaluated and sum-



mary and conclusions  have been prepared so  that the chemical's toxicity and



related characteristics are qualitatively identified.



     The present document represents an up-to-date data  base.  The document



considers all  sources  of  acrylonitrile  in the environment,  the likelihood  for



its exposure to  humans,  and the possible affect  on man and lower organisms



from absorption.    The  information  found in the document is integrated into a



format designed as  the basis for performing risk  assessments.   When appro-



priate, the authors of the document have attempted to identify gaps in current



knowledge that limit risk evaluation capabilities.
                                     2-1

-------
                    3.   PHYSICAL AND CHEMICAL PROPERTIES

3.1  SYNONYMS AND TRADE NAMES

          Chemical Abstracts Name:   2-propenenitrile

          CAS No.:  107-13-1

          EPA Toxic Substances List No.:   R037-2101

          RTECS No.:  AT52500

          Standard Industrial Code:  2822;  2821

     The compound is also known as  acrylonitrile  (AN),  cyanoethylene,  propene-

nitrile, and  vinyl cyanide  (VCN).   Fumigant formulations containing  acrylo-

nitrile with  names Acrylon,  Carbacryl,  Fumigrain, Ventox, and  ENT-51 are  no

longer manufactured in the United States.

3.2  STRUCTURAL AND MOLECULAR FORMULAS AND MOLECULAR WEIGHT

          Hvt=C-GEN
          H'  H

          C H3N     Molecular Weight:  53.06

3.3  BOND ANGLES AND BOND DISTANCES

     A molecule  of acrylonitrile is  planar;  all  the bond angles are  close  to

120°.  Estimated  bond  distances are  as  follows  (Wilcox and  Goldstein,  195*0:

C-H:  1.09 A;  C-C:  1.H6 A;  C=C: 1.38 A;  and C N: 1.16 A.

3.1  PHYSICAL PROPERTIES

3.1.1  Description

     Acrylonitrile is a clear, colorless, and highly flammable  liquid that has an

unpleasant and irritating characteristic  odor (Fassett, 1963).

3.1.2  Boiling Point

     77.3CC at 1 atmosphere (Groet, 1978).

3.1.3  Melting Point

     -83.55°C (Groet, 1978).
                                        3-1

-------
3.1.1  Density




     djj°( liquid)  :  0.8060 g/ml;  vapor density:  1.83 (air=l)  (Greet,  1978).



3.1.5  Refractive Index


      on

     n£u :  1.3911 (Weast, 1976).



3-1.6  Spectroscopic Data




     The A     is  at  203 nm  with a  molar extinction  coefficient  of 6100;  a
          ZDoX



compilation of infrared,  Raman, NMR,  and  mass spectral  data  is available  in




Grasselli and Ritchey (1975).




3.^.7  Solubility




     Acrylonitrile is  soluble  in water, acetone,  and  benzene (Weast,   1976);




miscible with ethanol,  carbon  tetrachloride,  ethyl acetate, ethylene  cyanohy-




drin, liquid carbon dioxide, ether, toluene, petroleum ether, and xylene (Miller




and Villaume,  1978).




     The solubility in water is given below (Groet, 1978):




               0°C:      7.2 weight?




               20°C:     7.35  weight?



               10°C:     7.9 weight?




               60°C:     9.1 weight?




3.1.8  Volatility in Water




     Henry's law  constant:  0.063 at  25°C (Bocek, 1976).  Partial vapor pressure




(water azeotrope) :   log P = 7.518 -  j^'L   i-e.,  80 mm at 20CC  (Miller and




Villaume,  1978).   The half-life of evaporation of acrylonitrile  from water with




an assumed  1 meter depth can be calculated, using the method of  Dilling (1977),




to be 795 minutes.  The method of Dilling (1977) is applicable for slightly soluble
                                        3-2

-------
organic compounds present In a  static aquatic system.  Therefore, the calculated



evaporative  half-life  for  moderately  soluble  acrylonitrile  present  in  a



naturally flowing aquatic system may not be accurate.



3.4.9  Volatility



     Values for the vapor pressure of acrylonitrile (mm of mercury)  at different



temperatures are given below (Groet, 1978):



                8.7°C:   50.3



               23-6°C:   99.8



               H5.5°C:  2U9.8



               64.7°C:  500.3



               77.3°C:  759.8



3.4.10  Stability



     Flash-point (open cup): 0°C (Steere, 1968); Flash-point (closed cup):  -1CC



( Patterson et al., 1976), -U.1»°C (Miller and Villaume,  1978).



     Explosive limits:  3.0 to 17$ by  volume in air at  25°C (Steere,  1968).



     Ignition temperature:  48l°C (Steere,  1968).



3.4.11  Octanol-water Partition Coefficient



     k = 0.12 (Leo et al., 1971).



3.^.12  Conversion Factor



     1 ppm in air =2.17 mg/m3 at 25°C.



3.5  CHEMICAL PROPERTIES



3.5.1  Reactivity



     Undergoes reactions at both the nitrile group and  the double bond (Maltoni



£t al.,  1977).   Some  of these  reactions  are  used for  the quantification of



acrylonitrile and have been  discussed in Section 1.  Acrylonitrile also undergoes



the following reactions (Miller and Villaume, 1978), many of which are important



commercially.
                                        3-3

-------
3.5.2 . Polymerization


     Polymerization,  forming  high  molecular  weight  products,  is  the  most


important commercial  reaction of  acrylonitrile.    The polymerization  usually

requires the presence of free radical initiators, such as peroxydisulfate.  Heat

                    o
or light (X  <  2900 A) can also  initiate  polymerization reactions.   Oxygen and

methylhydroquinone are powerful inhibitors of the  polymerization reaction.


     Pure  polyacrylonitrile  cannot be   dyed  using  conventional  techniques.


Copolymerization of acrylonitrile with small amounts of methyl  methacrylate or

vinyl pyridine introduces reactive  dyeing  sites.   Acrylonitrile can be copoly-


merized with other monomers as well; examples of other acrylonitrile copolymers


include  nitrile rubber,  acrylonitrile-butadiene-styrene  (ABS), and  styrene-

acrylonitrile (SAN) resins.   Terpolymers  of acrylonitrile  or  methacrylontrile

are the so-called barrier resins.


3.5.3  Reaction at the Nitrile Groups


     Acrylonitrile when  reacted  with 8H.5% H2son at 100°C  produces acrylamide

sulfate, which yields acrylamide upon neutralization:

          CH2=CHCN •»• H20 + H^O^ 	)  CH2=CHCONH2 •  H^O^


     CH2=CHCONH2 • H2SOU + 2NaOH 	)  CH2=CHCONH2 +  Na2SOu + 2H20

     Until recently, the above process was used exclusively for the commercial

production of acrylamide.  The catalytic  hydration of  acrylonitrile is the cur-

rently used method  for the production of acrylamide. When acrylontrile is heated


with less concentrated H^O^  or with water, acrylic acid (CH2=CHCOOH) is formed.


In aqueous NaOH, NH., is the reaction product.  The hydrolysis constant for this

reaction was  measured as  3-8 x 10~^  min"  at 60°C  and 1.39  min"   at  100°C

(Linetskii and  Serebryakov,  1965).   Violent polymerization, however,  has been

reported  to  occur  with concentrated  alkali  (Steere, 1968).    Acrylonitrile

allowed  to  react  with  alcohols   in  the  presence   of  concentrated  H-SOj.
                                        3-4

-------
produces esters of  acrylic acid.  With  olefins,  it forms N-substituted acryl-



amides in the presence of concentrated H-SOj..



3.5.1)  Reactions at the Double Bond



     The double bond  in acrylonitrile acts  as  a  dienophile in the Diels-Alder



reaction.  Cyclic products are produced when acrylonitrile is treated with  ali-



phatic  or  alicyclic  compounds  containing  conjugated  carbon-to-carbon double



bonds.  An example is the reaction with butadiene:
     CH2=CHCN + CH2=CH-CH=CH2 -    HC           CHCN
                                       HC
                                        A-3-tetrahydrobenzonitrile



     In the presence of  catalyst,  acrylonitrile can be hydrogenated to  propio-



nitrile, which can be further hydrogenated to n-propylamine:





               CH -CHCN  "2     CH CH CN  "2






     It can also be reduced,  in the presence of magnesium and methanol,  in the



following manner to produce adiponitrile:



               or»u _ftJPM _i OPU f\U  • \Jtfr 	  PU PU /"*M
               £Un_ = L/nUN + cLH—UM + rlK ~~^"^"~  ^n_t/H^LN
                  2           3                 22.







This is  then further  reduced to  hexamethylenediamine,  which  is used  in  the



production of nylon.



3.5.5  Cyanoethylation Reactions



     These reactions  involve the  interaction  of  acrylonitrile with  compounds



containing active hydrogen.   Examples of compounds  containing active hydrogen



are water, alcohols, ammonia, amines,  mercaptans,  aldehydes,  and inorganic acids
                                        3-5

-------
and their salts (Miller  and  Villaume,  1978).   The generalized  reaction  can  be




written:




                    CH2=CHCN  + AH 	>  ACH2CH2CN



     The cyanoethylation  of pseudouridine, inosine, and ^-thiouridine by acrylo-




nitrile  has  been studied as a  model  for the  cyanoethylation  of  intact  tRNA




(Miller and Villaume, 1978).




3.6  CHARACTERISTICS OF THE  CHEMICAL PRODUCT




     Technical grade  acrylonitrile  is  a  highly  purified  product with  greater




than 99$  purity.   The major impurity  is  water,  which is usually present  at a




maximum of about 0.5$.  The water improves the stability of the  product.   Other




possible trace contaminants  include  acetone,  acetonitrile,  acetaldehyde,  iron,




peroxides, and hydrogen cyanide.  Highly pure acrylonitrile may polymerize spon-




taneously.   To prevent  this,  methylhydroquinone (35-50  ppm)  is added  to the




commercial product.   Yellowing upon  exposure to light indicates  photoalteration



to saturated derivatives.




3.7  CONCLUSION



     Acrylonitrile is moderately soluble  in water.   Because acrylonitrile does




not appreciably dissociate in water, hydrogen  cyanide is not expected to  be a




product of hydrolysis.   Because  the vapor pressure  of  acrylonitrile is appre-




ciably high,  most atmospheric emissions from its manufacture  and use should occur




as vapor.   The evaporation  rate of acrylonitrile  from water  is appreciable.




Therefore, evaporation from  contaminated water  surfaces   can  be  expected  to




occur.
                                        3-6

-------
                     H.   SAMPLING AND ANALYTICAL METHODS


     The level of acrylonitrile has been determined in a number of environmental


media of interest.  These include:  (1) air;  (2)  water;  (3) waste water;  (1)  soil


and sediment; (5)  residue  in polymers and the  extent  of monomer migration  in

food-simulating  solvents;  and (6) various  other media.    All these media  can


directly or indirectly affect the environmental level or human intake of acrylo-


nitrile.


     The sampling method  is generally dependent  on  the  medium  intended to be


monitored.    The  analysis  of samples can be  divided  into  two steps,  namely,

pretreatment or  clean-up procedure( s), when necessary, and  quantification  pro-


cedures. The selection  of  a  particular identification and quantification method

is dictated by the accuracy, reproducibility,  detection limit, and the  possible

interference(s)  of the method.  The sampling and analysis of acrylonitrile in the

individual  media are discussed below.


M.I  AIR


4.1.1  Sampling

     The collection of acrylonitrile from air  has  been  done by two methods.   The

first method consists of direct collection of  samples without preconcentration.

The second  method employs  the concentration  of acrylonitrile in a collection


medium during sampling.


     In direct sample collection, the air  is drawn into plastic bags (Keresztesy
                         •
et al., 1977) via a two-way valve by means of  a sampling  pump.  The samples are

then transported to the  laboratory for analysis.

     The disadvantage with direct collection is that it does  not allow  any

concentration of  the  sample  and  may  cause additional problems during transport

and storage  of  samples.   Therefore,  the  detection limit  of the method  is  not

satisfactory for ambient  air samples, even with the  most  sensitive method of
                                        4-1

-------
detection available presently.  The advantage of  direct  collection is that the



samples can be analyzed without any pretreatment, thereby reducing analysis time



and avoiding any sample losses.   It also provides a method for continuous area



monitoring  of  acrylonitrile and  for  monitoring acrylonitrile  from stack  and



other high concentration emission sources.



     In the  preconcentration  method  for sample collection, the  air containing



acrylonitrile is passed  through  either  a  solid sorbent  or a trapping  liquid



medium.   Table H-l  lists  the different sorbents  and trapping media used  for



collection  of acrylonitrile  in  air.    Activated  carbon,  silica,  and  porous



polymers have been  used as  solid sorbents.  The sampling unit usually consists of



a battery  operated pump and  a rotometer that indicates  the  sample  flow rate



through the sorbent.  A tube that contains the solid sorbent is held vertically



at a height of 1.5  m from the ground and is connected  to  the  rotometer  and the



pump unit.  Air is  drawn through  the  sorbent  tube at a certain flow rate for a



specified length of time.   At the end of sample collection, the  tube  is  closed




with caps and shipped to the laboratory for analysis.



     In addition to  the  error that could be made  in measuring the air  volume



(calibrated pumps  may not  maintain constant air flow  over the entire sampling



period), another disadvantage of  the  solid sorbent method  is  that  sample loss



will occur if the breakthrough capacity  for the sorbent is exceeded.  Therefore,



the breakthrough capacity of acrylonitrile through the sorbent should  be deter-



mined  experimentally.  The breakthrough capacity  is dependent on  the relative



humidity of air sampled and the presence of other interferences in the  air.  The



concentration of the compound  being collected, however, affects the breakthrough



volume only slightly.  A pollutant present at 20 ppm will break through at only a



slightly smaller volume than when present at 1 ppm (Russell, 1975).
                                        4-2

-------
                Table 4-1.   Various Sorbents  and Trapping  Media  for Collection of Acrylonitrile  in Mr
i
CO
Sample
Plastic producing
plant
Acrylic fiber
factory
Simulated air
Plastic processing
plant
Eleven production
plants
Simulated air
Simulated air
Settling pond In
acrylonltrlle plant
Occupational air
Occupational air
Craln fumlgant air

Occupational air
Acrylonltrlle plant
—
	
Rubber plant
SAN plant
— — —
Sorbent
Sllochrom-2 at
-50'C
Silica gel
Activated carbon
Activated carbon
Activated carbon
Porapak N
Porapak N
Porous polymer
Tenax GC(T)
Activated carbon
Carbosleve B
95Z ethanol
Chilled water
Water
Water
Chilled water
Chilled water
Water
H.SOg & glasa beads
1Z H2S04
1Z H.SO.
I n
DMF
25Z methanol at
-15'C
Sampling
Condition
0.5 1pm for
4-6 1
Personnel
sampler for
100 nln.
15 1 in gas
bag
0.2 1pm for
8 hr.
1.0 1pm for
=24 hr.
0.1 1pm for
3-5 1
0.1-0. IS 1pm
for 3-8 1
=0.07 m3/hr.
for 5-30 mln.
0.2 1pm for
20 1 max
1 1pm for
10-20 1
1 Ipa
0.03-0.33 1pm
0.015 1pm for
1.5-3 1
1 1pm for
20-60 1
1 1pm
	
0.4 1pm
	
0.1-0.3 1pm
— — —
Desorptlon Analysis
Method Technique
Water Colorlmetrlc
Thermal at 200'C CC-FIDh
Water ' CC-FID
2T acetone in CS_ CC-PID
2Z acetone In CS. CC-PID
CS? CC-FID
Thermal at 200*C CC-FID
Thermal CC-FID
Thermal CC-FID
CHjOB CC-FID
CHjOH CC-FID
	 Polarographlc
	 Polarographlc
	 Colorlmetric
	 Colorlmetrlc
	 u.v.
CC-FID
	 Colorlnetric
	 Tltrimetrlc
	 Colorimetrlc
	 Colorlmetrlc
	 CC-FID
GC-FIO
Conf Irmat Ion
Technique
None
None
None
None
None
Rone
None
None
None
None
Rone
Rone
None
None
None
None
None
None
None
None
None
None
None
Detection
Limit
	
0.1 ppm
0.25-0.5 ppm
0.03-0.31 ppm
< 0.3 ug/a3
< 1 ppb
< 0.26 ppm
10 ng
Range 5-135
mg/m3
Range 40-750
0.5-5 mg/50 ml
1.08 x 10"5 H
0.5 mg/m3
Range 10-100
ppm
50 ppb
1.0 ug/nl
20-300 ug/ml
< 0.85 ug/1
0.5 mg/m3
...
Reference
Rrynska. 1970
Korthova et^ al. , 1974
Sakural e^ a^. , 1978
CagnoTi & Pogner,
1979
Marrs ft. al. , 1978
Going et a_l. , 1979
Russell. 1975
Campbell & Moore.
1979
Hughes ft Horn, 1977
N10SH, 1977
Barrett. 1974
Rogaczewnka, 1964
Berck, 1962
Najiarova & Nakrap
1978
RogacKewnka, 1976
Brleger ct^ •!. . , 1952
Hughes ft Horn. 1977
Russklkh. 1973
Gunther & Bllnn. 1955
Babanov, 1960
Arato & Blttera, 1972
Bablna, 1979
Fasltta a Llcclardello,
1977
                    1pm - liters per minute
                    CC-FID - gas chromatography-flame lonlzatlon detectors

-------
      For air nearly saturated with moisture,  the breakthrough volume for acrylo-



nitrile  in  Porapax N (H"  x 1/1"  tube)  was determined  to be  3  to 5  liters.



Breakthrough volumes  can  be increased by  using  longer sampling tubes.   These



tubes, however, will result in increasing the back pressure in £he  tube.



     Activated carbon is  the most widely used sorbent  for the collection  of



acrylonitrile in air.   The adsorption parameters  for activated carbon will vary



depending on the nature of the carbon and the treatment it received.  With 6-10



mesh  BC-AC  granular  carbon, Sansone et al.  (1979)  determined the  adsorption



capacity and adsorption rate constant for acrylonitrile to be O.IOH g/g and 116.0



min.  , respectively.   With acrylonitrile that  contained 50$ relative humidity,



Nelson and Harder (197*0 determined the adsorption capacity to be 0.17** g/g.  The



breakthrough times  for 1$,  10$,  and 99$  passage from a  58 g  charcoal in  a



respirator cartridge  were determined to be U8.5 min.,  61.1 min.,  and 168 min.,



respectively (Nelson and Harder, 197*0.  The types  of  activated carbon used in



the above experiments were different.



     The effect  of humidity on sample  recovery  by 1.5 g  activated  carbon was



studied  in  detail by Going £t al.  (1979).    He  observed  that introducing  a



Drierite drying tube to absorb moisture did not significantly improve recoveries



at relative humidities equal to or exceeding 70$;  however, the average  recovery



for acrylonitrile with air volumes of 750 liters/g  or greater remained acceptable



at 75 i 9.1$ at relative humidities up to 60$, with or without the  drying tube.



     NIOSH (1977b)  determined  that  the  maximum  amounts of  acrylonitrile that



could be collected with a sorbent tube (8 cm x 5 mm) from 100 ppm acrylonitrile in



air were 6 mg, 15 mg, and  22 mg when the relative humidities greater than 95$,



equal to 50$, and less than 5$,  respectively.



     Segmentation of the carbon tubes into a front and back section remains the



most  widely  used method for the  determination of  breakthrough capacity limit
                                         4-4

-------
during  sample  collection.   If the  second tube  retains  more  than a  certain



predetermined percentage  of acrylonitrile,  it  indicates that the  breakthrough



limit has been exceeded.



     It is important to determine the stability of  acrylonitrile  in the sample



tubes during  transportation and storage.   Marrs e_t al^ (1978)  showed that acrylo-



nitrile would  remain stable on  the carbon tubes  at room   temperature  for  a



minimum of 5 days.   Going et  al.  (1979) demonstrated that  acrylonitrile would



remain stable up to 8 days on charcoal tubes stored at -17°C  and  up  to 24 days on



tubes stored at -78°C.



     The collection  of  acrylonitrile  in air by liquid  trapping media  has been



used rarely in recent years because of the inconvenience in  handling and trans-



porting the impingers and the need for cooling the trapping  media  during sample



collection.  The collection efficiencies and the storage stabilities of acrylo-



nitrile in these media  are  not  always known.    The collection efficiency  of



chilled water as a trapping medium was  determined to  be over  98$  (Berck, 1962;



Brieger, 1952).  Virtually no data are available regarding the storage stability



of acrylonitrile in  this  trapping medium,  but  it is expected to  be reasonably



stable in water and 1$ H-SCL, particularly since  Going et_ al.  (1979)  demonstrated



that acrylonitrile was stable for about 23 days in neutral distilled water and at



a pH of 4.  Acrylonitrile  collected in aqueous methanol and ethanol  and stored at



3°C can be expected to be fairly stable, based on the evidence  of  Going e_t al.



(1979), who demonstrated that acrylonitrile in carbon disulfide remained stable



for over 12 days when stored at 3°C.



4.1.2  Analysis



1.1.2.1  Pretreatment



     Acrylonitrile collected on solid sorbents  requires a desorption procedure



before  identification  and  quantification.   Thermal   desorption  and  solvent
                                        4-5

-------
desorption are two commonly used methods.   Thermal desorption normally uses  gas




chromatographic methods for identification  and  quantification.    In  this pro-




cedure, the sorbent tube is heated  between 100°C and 200°C in the injection port




of a gas chromatograph.  To avoid  peak  broadening,  the position of the  sample




tube in the gas chromatograph injection port  should  be  in the reverse order as




that used during sampling.  The thermal desorption method is appropriate for  use




with porous polymers, particularly with Porapak N.   Several  investigators have




used this technique (Campbell and Moore, 1979; Russell,  1975; Hughes  and Horn,




1977).  The advantages  of  thermal  desorption are that  it avoids manual  sample




pretreatment and the recovery of acrylonitrile is  almost quantitative  (Campbell




and Moore, 1979; Russell,  1975). The method also  has a higher sensitivity than



the solvent desorption method, where only a fraction of the eluted acrylonitrile




can be injected into  the  gas chromatograph.  The disadvantages  of  the thermal




desorption method  include its inability to afford replicate analysis of the same




sorbent tube  and   its  tendency   to  cause  other gas  chromatography separation




problems such  as shortening retention time and broadening eluted peak due to  the




presence of adsorbed moisture on the sorbent column (Russell, 1975).




     The solvents  that have been  used for desorption of acrylonitrile from acti-




vated carbon are acetone, methanol, carbon  disulfide, and 2% acetone  in carbon




disulfide. The selection of the  desorption sorbent is dictated by two considera-




tions, namely, the sorbent  desorption efficiency and  its compatibility with  gas




chromatography.  For example, if column separation  is not adequate, methanol that




has a high response on flame ionization detectors (FID) will produce a large peak




shadowing the  acrylonitrile peak; therefore, it is  not very compatible with flame




ionization detectors.  Although  carbon disulfide  appears  to  be  compatible with




flame ionization detectors,  it is a poor solvent for nitrogen/phosphorous detec-




tors.  Acetone is  the solvent of choice  in the latter case.
                                        4-6

-------
     The  recovery  efficiencies  of acrylonitrile  from activated  carbon with



various solvents are given in Table 1-2.








         Table 1-2.  Recovery of Acrylonitrile from Various Solvents
Solvent
% Recovery
Reference

Methanol
Acetone
2% acetone in CS
2% acetone in CS
CS2 (2 ml)
CS2 (U ml)
ca. 50$
73.5 ± 5.3$
95.5 ± 7.9$
91*
58$
75$
Going et al., 1979
Marano et al., 1978
Gagnon and Posner, 1979
Silverstein, 1977
Silverstein, 1977
Silverstein, 1977
     It should be recognized that the recovery of acrylonitrile is dependent  on



the nature  of the activated  carbon  and the  extent of  loading.   Using  three




different activated carbons  and a variable loading  of  2 ug to 200  |ig  acrylo-




nitrile, Going eta^. (1979)  determined that the CSp desorption efficiency varied




from 53$ to almost 100$.  It would appear from Table 4-2  that  a  2$ acetone in CS




is the best solvent for elution of acrylonitrile from activated carbon in terms




of both recovery and GC-FID compatibility.




     When the acrylonitrile is collected in liquid  trapping  media,  the  samples




from high concentration sources usually do not require any pretreatment prior to




the detection and quantification procedures.




1.1.2.2  Identification and Quantification




     The methods utilized for the  identification and quantification of  acrylo-




nitrile collected by sorbent or trapping techniques are  shown in Table 1-1 (see




1.1.1).  The methods for the analysis  of acrylonitrile collected without precon-



centration appear in Table 1-3.  The infrared techniques are used exclusively for




samples that  need no preconcentration.  Although a  number of methods including
                                        4-7

-------
                                      Table  4-3.  Direct Analysis  of  Acrylonitrile
Sample

Simulated occupa-
tional air
Occupational air
Acrylonitrile in N,,
Simulated air
Gases generated by
heating ABS plastic
Occupational air
Simulated air
Occupational air
Sample Introduction
into Detector Unit

Sample probe & pump
Introduction of air
from gas bag into
evacuated cell
Introduction to a
vacuum cell at 8 torr
Pump & rotometer
Forced air
Passive diffusion
Direct injection
Direct injection
Analysis Technique

IR-microcomputer
IR
IR lasers
Detector tube
Detector tube
TM
Abcor GASBADGE
GC-NPD
GC-FID
Comf irmation
Technique

None
None
None
None
GC-MS
None
MS
None
Detection
Limit

0.2 ppm
0.5 ppm
0.03 ppm
Range 1-120
ppm
Range 0.8-19
ppm
100 ppb
0.5 ppm
References

Jacobs & Syr Jala,
1978
AIHA, 1970
Sweger & Travis, 1979
Kobayashi, 1956
Grote et^ al. , 1978
Silverstein, 1977
Marano et al. , 1978
AIHA, 1970
-p.
I
CO
          GC-MS =  gas  chromatography-mass  spectrometry
          GC-NPD = gas chromatography-nitrogen/phosphorus  detectors
          MS  = mass spectrometry
          GC-FID = gas chromatography-flame  ionization  detectors

-------
colorimetric,  titrimetric,  and  polarographic  were  used for  the  analysis  of



acrylonitrile in the past, the GC method is most extensively  used at  the  present



time.  The reason for this  is  the  lower  interference and higher sensitivity  of



detection  obtained   from  GC-flame   ionization  detection   (FID)   and  GC-



nitrogen/phosphorous detection (NPD).  The chromatographic  columns  found  to  be



most appropriate for use  with  thermally  desorbed acrylonitrile were Porapak N



(Russell, 1975) and  Porapak  Q (Campbell and Moore, 1979) .   In the case of solvent



desorption, the columns  that were found to be suitable were Durapak OPN/  Poracil



C  (Going et al.,  1979),  SP-2100  (Grote  et  al.,  1979)  for carbon  disulfide;



SP-1000 (Marano et al^.,  1978) for acetone; and SP-1000 (Gagnon and Posner,  1979)



and TCEP (Marrs et al..,  1978) for methanol. The details of other chromatographic



columns are given in subsection M.7.



M.I.3  Conclusions



     The detectors available at  present often do not have  sufficient  sensitivity



to allow the  detection of acrylonitrile in ambient atmospheric samples,  if these



samples have been collected without preconcentration.  The low level of acrylo-



nitrile in atmospheric samples  virtually  mandates the use of a preconcentration



device during sample collection.  Of the  two  preconcentration methods presently



available, namely,  solid sorbents  and trapping  media,  the  former method  is



preferable to  the  latter.   The  collection of acrylonitrile by  solid  sorbents



affords convenience in handling, shipping, and storage of the samples.



     Activated carbon and  porous polymer Porapak  N  are  the two best  sorbents



available for the collection of acrylonitrile in air.  The breakthrough capacity



for Porapak  N  (3 to 5  liters)  is  lower  than  that for  activated carbon (over



1000 liters) .  Consequently, Porapak N cannot  be used  for 2U hour sampling.  Even



with much lower sample volume, Porapak N will  afford a detection limit comparable



to that obtained from a larger  volume of air  collected by activated  carbon, for
                                         4-9

-------
the following reasons.  The efficiency of  thermal desorption  for  acrylonitrile




from Porapak N is quantitative, while it is poor with activated carbon;  there-




fore, activated carbon requires solvent desorption.  Whereas  the  entire  sample




can be injected into the quantitative gas  chromatography column in the case  of




thermal desorption, in  the  case of solvent desorption, only  a fraction  of the




total sample can be injected.




     With activated carbon,  2% acetone in carbon disulfide is the best desorption




solvent, since it gives  the maximum recovery.  This solvent  system  is  compatible




with flame ionization detectors.  Carbon disulfide,  however,  is unsuitable when




the more sensitive nitrogen/phosphorus detectors are used.   Acetone  is the most




suitable solvent in this case.  A number of columns including SP-2100 (Gagnon and




Posner, 1979) and Durapak  OPN/Poracil C (Going e_t al.,  1979)  have been used  as




the quantitative column when carbon disulfide  or 2%  acetone in carbon disulfide



was  the  solvent.   For  acetone solvent,  SP-1000 was  found  to  be  a  suitable




quantitative column (Marano et^ al..  1978) .




4.2  WATER




     In this section, only water samples obtained from  either surface water or




treated drinking  water  will be discussed.   Wastewaters will be discussed  in




subsection 4.3.




4.2.1  Sampling



     Water samples have been collected almost  exclusively by the  grab technique




(Kopfler et al.,  1976; Wronski and Zbigniew, 1974; Going  et  al.., 1979).  In a few




instances, multiple grab  samples  were composited  for  analysis (Going e£ al.,




1979).  Proposed EPA methods require that the samples be adjusted to pH 7.0 + 0.5




and  be  colected  in screw-cap  vials with Teflon* septa  without headspace.  The




collected samples should be  stored in a refrigerator (U.S. EPA, 1979a;  U.S. EPA,




1979b;  U.S.    EPA,  1980c).   The  ASTM (1980)   procedure allows headspace,  but
                                        4-10

-------
recommends immediate freezing of the  samples.  However a better method suggested



that the samples should  be  collected in brown glass bottles  with Teflon-lined



caps and acidified  at  the site to a pH <_ 14 (Going  tit al_.,  1979).   The samples



should be maintained at 0 to iJ°C by ice or an ice substitute during transporta-



tion (Going et al.,  1979; Kopfler et al.,  1976) .



     In one instance (Going et^ al^., 1979) ,  an attempt was made to collect acrylo-



nitrile  from  water by  the use  of solid sorbents.   Four  different sorbents,



namely, activated carbon,  Porapak N, Chromosorb 101, and  Chromosorb 104,  were



tried.  Water  spiked with acrylonitrile was passed through the sorbents  at a rate



of 4 ml/min.   The sorbents were then eluted  with  25 ml methanol at a rate of



5 ml/min.  The recoveries of acrylonitrile were poor; with activated  carbon, the



recovery was 30-35 percent and  the  porous polymers showed zero percent recovery.



1.2.2  Analysis



lJ.2.2.1  Sample Treatment



     Some water samples were analyzed without any  pretreatment.   Water samples



containing low levels of  acrylonitrile were pretreated  in order to concentrate



the acrylonitrile.   Three  available methods  for concentration  are purge-trap




(Kopfler et al., 1976; Going etal., 1979),  distillation  (Going et al,.,  1979;



Peters, 1979), and  headspace (McNeil and Brader,  1981;  Markelor et al.,  1981)



techniques.



     Acrylonitrile can be purged from water at elevated temperatures by passing



an inert gas through it.   The acrylonitrile contained in the purged gas is then



trapped in  chromatographic  media for subsequent analysis.   The  details of the



purging system were  described by Going et^ al.  (1979) and Kopfler e_t al. (1976) .



The system used by Kopfler et^ al. (1976)  appears to be preferable to the system



used by Going  e£ al.  (1979), because the former allows purging of  1HO  ml of water
                                         4-11

-------
compared to 10 ml water in the  latter  case;  however,  the purge-trap efficiency



for acrylonitrile was  not studied in detail in the Kopfler et aK (1976) system.



     The purging conditions were studied in  detail by  Going et  al.  (1979),  who



found an  almost  quantitative  recovery when  helium  was  passed at  a rate  of



20 ml/min through water heated to 85°C for  30 minutes.   The effect  of different



trapping systems were also studied by Going et al. (1979).   It  was  established



that, under thermal  desorption conditions, both Porapak N and Chromosorb 104 gave



quantitative desorption.  The recovery from Tenax GC,  however, was  found  to be



poor (Going et al.,  1979).



     Two  different   distillation  techniques,  namely   steam  distillation  and



azeotrophic distillation have been used  to concentrate acrylonitrile  from  the



aquatic phases.   In the steam  distillation  technique, the  sample  is directly



distilled for an appropriate  interval and the organics of interest  is collected



in the distillate.  Peters  (1980) has used this technique to concentrate acrylo-



nitrile by a factor  of 300 in about 15 minutes of steam distillation.  The major



disadvantage of this method and  the  azeotropic  distillation is  that relatively



large sample volumes (>300 ml)  are required for these methods.



     In the azeotropic distillation technique, water containing acrylonitrile is



distilled with methanol, and a small volume of the azeotropic distillate contain-



ing acrylonitrile in methanol  is collected for further analysis.  This technique



serves as a simultaneous clean-up and concentration device for acrylonitrile in



water samples.  The  description  of  the  distillation apparatus was given by Going



et al_. (1979), who obtained maximum recovery under the following conditions.



     A 500 ml water  sample was added to the distillation flask along with 25 ml



methanol and 5 ml  18 N H-SOj. and  the content was  distilled at a rate of 1 ml/min.



The first  10 ml  of  the distillate  were collected for subsequent analysis.  The
                                        4-12

-------
percent recovery was  about 90J for the combined first and second 10 ml aliquot of


the distillate.


     In the headspace method, a  known volume of the aqueous sample is Introduced


into a  specially designed  enclosed glass  apparatus and  ths  sytem is  thermo-


statically maintained  at a  constant  temperature.   After  the  system  attains


equilibrium,  a known volume of the headspace vapor is withdrawn for acrylonitrile


determination.   This  method  in the  pas  thad  faced  problems  owing  to  the


difficulty in establishing a calibration procedure.  The  partition coefficient


of  a  component between gas and liquid  phase is dependent  on the  total  ionic


strength of  the  solution.   Therefore,  the same  concentrations  of  a  component


present  in  two  aqueous  solutions  of  different  ionic  strengths  but  otherwise


identical conditions, will not produce the same equilibrium vapor pressure.   This


problem of a  calibration curve has been  large obviated through the development of


a standard addition method.   The headspace method has been utilized by Markelor


et al. (1981) and McNeil and Brader (1981)  for the  analysis of acrylonitrile in


aqueous solutions.


U.2.2.2  Detection and Quantification


     With  the exception of  one case in  which a  titrimetric  method was  used


(Wronski and  Zbigniew,  1971*) ,   the  rest of  the  studies  reviewed  utilized  GC


separation and GC  retention  data for  the  identification  of acrylonitrile  from


water samples.  For direct aqueous injection, both Chromosorb 101 (Going  et^ al.,
                         •

1979)  and Chromosorb 102 (Marano et al.,  1978) were used,  although Going Q al.


(1979)   reported  better  separability  with Chromosorb 101.   The  same  column


(Chromosorb 101)  was used for acrylonitrile determination by the purge-trap and


azeotropic distillation techniques  (Going  e_t al.,  1979;   Kopfler  e_t al.,  1976;


Federal Register, 1979).  Other Ge columns including carbowax 1500 and Tenax GC


have been  used  for  the separation of  acrylonitrile (Markelov  e_t  al.,  1981;
                                        4-13

-------
Peters, 1980).   An  aliquot of the distillate  from  azeotropic  distillation was




injected  directly into  the  GC  column.    In the  purge-trap  technique,   the




acrylonitrile  from  the  trapping  column  was  thermally  desorbed  onto   the




separating column. A summary of water analysis  techniques is given in Table  1-1.




Although nitrogen/phosphorus  detectors (including the Hall  detector)  and flame




ionization  detectors  were  used  for  quantification,   the  former  detectors




antrogen/phosphorous)  have  better  selectivity and  sensibility than  the  latter




detector.  However,  the  best available method for the unambiguous identification




of acrylonitrile is provided by the more expensive mass  spectrometric method.




1.2.3  Conclusions




     The two best available methods  for the determination  of  acrylonitrile in




water  are  the  purge-trap  and the  azeotropic distillation techniques.    Both




methods gave almost quantitative  recovery  of acrylonitrile in water.  The detec-



tion limit for acrylonitrile by the purge-trap method was lower than that for the




azeotropic distillation method; however,  this  advantage  of the purge-trap tech-




nique is somewhat offset by the  experimental  complexity of the method  and its




inability to perform replicate analysis on the same water sample.




1.3  WASTEWATER




1.3-1  Sampling



     No details regarding the  sampling of wastewaters is  available.  Grab samples



may be suitable in certain cases.  To monitor  the discharges that  are dependent




on process operation stages, a 21-hour composite sample  is preferable.




1.3.2  Analysis




1.3.2.1  Pretreatment




     Azeotropic  distillation  with methanol is  a  method  used frequently for




wastewater.  This technique allows concentration of acrylonitrile  and  reduces



the possibility of interference.  In one case,  solvent extractions using benzene,
                                        4-14

-------
                           Table U-*4.  Analyses of Acrylonitrile in Water
Sample
Collection Analysis Technique Confirmatory Detection
Method Technique Limit
Reference


Spiked water
Drinking water
Surface water
Surface water
Surface water
Surface water
Spiked water
Spiked water
Spiked water
	 Direct injection, MS
GC-NPD
Grab Purge-trap ( Tenax GC) , None
GC-MS
Grab Titrimetric with None
mercaptoethanol > Na_SO,
Grab Direct injection, MS
GC-FID
Grab Purge-trap, GC-FID, > MS
GC-Hall
Grab Azeotropic distillation, MS
GC-Hall
Sorbent Methanol desorption, None
adsorption GC-Hall
	 Steam distillation, None
GC-FID
— Headspace, GC-Hall None
10 ppb
Qualitative
0.5-0.05 ppm
1 ppm
0.1 ppb
0.1-1.3 ppb
Showed poor
desorption
efficiency >
method
abandoned
NR
NR
Marano e^ al.
Kopfler et al.
, 1978
, 1976
Wronski > Zbigniew,
197U
Going et al. , 1979
Going ejt al. ,
Going et^ al. ,
Going et al. ,
Peters, 1980
Markelovetal.
1979
1979
1979

,, 1981
MS = mass spectroraetry
GC-NPD = gas chromatography-nitrogen/phosphorus detectors
GC-MS = gas chromatography-mass spectrometry
GC-FID = gas chromatography-flame ionization detectors
GC-Hall = gas chromatography-Hall detector
NR = Not Reported

-------
ether, and isobutyl  acetate were used (Ponomarev et al.,  197M) .   The recovery of



acrylonitrile after  three extractions was not quantitative but was reproducible.



     The various  pretreatment  methods  used for acrylonitrile  determination  in



waste water are shown in Table M-5.  Table M-5 also lists the different detection



and quantification  methods  and their  detection limits where  available.   The



principles of the  detection methods are  discussed in subsections 1.2.2.2 and 1.7.



U.3.3  Conclusions



     Although the analysis  of  acrylonitrile in  wastewater by the  purge-trap



technique has not been reported,   it  is a  potentially  appropriate method  for



wastewater analysis.  The usefulness  of the method can be enhanced by fractional



purging at different temperatures  (Kopfler et al., 1976).   A better  method for



acrylonitrile analysis in wastewater is the  azeotropic  distillation  technique.



The detection and  quantification can be  best achieved by either the GC-FID or the



GC-NPD technique.  The chromatographic  columns used for water analysis should be



suitable for wastewater analysis.



1.1  SOIL AND SEDIMENT



     Only one reference  could  be  found in  the literature  for  the analysis  of



acrylonitrile in these media.  The  following discussion in this section is based



on the work of Going et al.  (1979).



1.1.1  Sampling



     Soil from the top 12 cm was removed and placed in glass bottles and stored



over dry ice until analyzed.  Sediment  samples were collected with a dredge and



the samples were kept on dry ice until  analyzed.



1.1.2  Analysis



1.1.2.1  Pretreatment



     The pretreatment of the sediment samples after the removal of excess water



was the  same  as  that used for soil  samples.   The samples  were extracted with
                                        4-16

-------
                        Table 4-5.   Analysis  of Acrylonitrile in  Wastewaters
Sample
Pretreatment
Analysis Technique
Detection Limit
Reference
From polymeri-
zation process
Solvent extraction
Azeotropic distillation
Azeotropic distillation
None
Azeotropic distillation
None
Polarographic
Polarographic
Polarographic
Titrimetric
Colorimetric
Titrimetric
GC-FID
10 ppm
3-5 ppm
< 0.1 ppm
< 2 ppm
< 0.5 ppm

0.1 ppm
Sevest'ysmova et^ al^., 1966
Ponomarev et al., 1974
Daues & Hamner, 1957
Stefanescu & Ursu, 1973
Lawniczak, 1977
Covic-Horvat £t_ al., 1970
Deur-Siftar & Svob, 1976
     GC-FID = gas chromatography-flame  ionization  detectors

-------
water using ultrasonic agitation and the mixtures were centrifuged.   The super-

natants were withdrawn and  filtered, and the aqueous extracts were directly used

for analysis.   A  few samples  were concentrated  by  the purge-trap  technique

already discussed in subsection U.2.2.1.  It should be mentioned that the overall

recovery of  acrylonitrile  from  the soil and sediment samples  either by direct

extraction or by the subsequent purge-trap technique was not determined.

4.1.2.2  Detection and Quantification

     Gas chromatography with a Chromosorb 101 column and a Hall detector was used

for detection and quantification.  The detection limit for direct aqueous injec-

tion ranged  from 50 to MOO  ug/kg.   For the  purge-trap technique, the detection

limit was ca. 0.5  ug/kg.

4.U.3  Conclusions

     The desorption efficiency of acrylonitrile  from  soil and  sediment samples

by ultrasonic agitation has yet to be determined.  Although azeotropic distilla-

tion was  not used,  both this method  and the  purge-trap  technique  should  be

suitable for the determination of low levels of acrylonitrile in these samples.

U.5  RESIDUE IN POLYMERS  AND THE EXTENT OF MONOMER MIGRATION IN FOOD-
     SIMULATING SOLVENTS

     Several polymers of  acrylonitrile are presently used in  the United States as

food packaging materials.  Until recently, some containers for carbonated bever-

ages had  nitrile  barrier  resins as  components; however,  the use  of nitrile

barrier resins in beverage containers in the United States is presently banned by

FDA.   The ABS  resins (terpolymer  of  acrylonitrile,  butadiene,  and styrene),

however, are currently used  for such food packages as margarine tubs, fruit juice

containers, and vegetable oil bottles.   In this section, the analysis of acrylo-

nitrile residue in both polymers used for food packaging and their migration to

food-simulating solvents  kept in contact with the polymer will be discussed.
                                        4-18

-------
H.5.1  Analysis



14.5.1.1  Pretreatment



1.5.1.1.1  Pretreatment for Polymers



     Three methods are available for the pretreatment of polymers prior to quan-



tification.  In one method, the sample is dissolved in a suitable solvent  and  an



aliquot of this solution is used for analysis.   The solvents used for different



polymers are shown in Table ^-6.   The  disadvantage of direct injection  of the



solution is that it accelerates the deterioration of the separating column when



GC is used for  identification.   The polymer build-up in the  gas chromatograph



injection port can be prevented by addition of water or methanol to precipitate



the polymer and the supernatant can  be injected into the gas chromatograph. The



detection limit of this method, however,  is not satisfactory and was determined



to be about 10 ppm (Steinchen, 1976).



     To avoid column contamination, reduce  the  interference arising from large



amounts of solvent, and increase the sensitivity  of detection, the second method,



known as  the  head-space analysis,  is  presently used for  the determination  of



residual monomer in polymers.



     Two approaches to the head-space  analysis have been used:  solid and solu-



tion. The solid approach involves the equilibration of a solid polymer sample  in



the sealed tube at a constant  elevated temperature.  The advantage of the solid



approach is that  it  has  tenfold  more sensitivity  than the  solution  approach



(Steinchen, 1976).  The disadvantages of  this approach  are:  (a) equilibration



with the head-space may  take  a long time,  and  (b)  since  polymer  standards  of



known monomer content are not  readily  available, the head-space  monomer concen-



tration must be related to the original concentration in the  polymer either  by



assuming 100$ diffusion of the monomer into the head-space  or through determina
                                        4 -19

-------
                 Table  4-6.   Analysis of  Acrylonitrile  Residue in Polymers  and Food-Simulating Solvents
Sample
Plastic
Plastic
Residue In
food-simulating
solvent
Polymer
Residue In
food-simulating
solvent
Residue In
food-simulating
solvent
Residue in
food-simulating
solvent
Polymer
Residue In
food-simulating
solvent
ABS resin
Residue In
food-simulating
solvent
Polymer
Plastic
Residue In
beverages
Pretreatment
Pyrolysla & solvent
trapping
Dissolved in benzene
Direct
(a) Pyrolysls CC
(b) Dissolution in
dimethyl formamlde
(c) Dissolution In ether
Head-space equilibration
Azeotroplc distillation
Direct
Dissolved In
N,N-dlme thy lace t amide
Azeotropic distillation
Dissolved in dimethyl
formanlde or dimethyl
sulfoxide and head-space
equilibration
Head-space equilibration
Dissolved In o-dichloro-
benzene & head-space
equilibration
Dissolved in propylene
carbonate and head-space
equilibration
Head-space equilibration
Analysis
Technique
Polarographic
Tltrlmetric
GC-FID
CC-FID
CC-FID
CC-NPD
GC-NPD
CC-NPD
GC-NTD
GC-FID &
GC-NPD
CC-NPD
CC-FID
GC-NTD
CC-NTD
Confirmation Detection
Technique Limit
None 100 ppm
None < 6 ppm
None 	
None 10-100 ppm
None 	
MS < 0.01 ppm
None 0.04 ppm
None 0-5 ppm
None 50 ppb
None 0. 3 ppm
None 30 ppb
None 0.5 ppm
None °-1 PP™
None 5 PPb
Reference
Ulide & Koehler, 1967
Roy, 1977
Harkelov & Semenenko, 1976
Relchle & Tengler. 1968
Chudy & Crosby, 1977
KcNeal e£ a_l. , 1979
Brown £t al. , 1978
Brown e_t al. , 1978
Hartshorn, 1975
DIPasquale et al., 1978
DIPasquale et aJU , 1978
Stelnchen, 1976
Cawell, 1979
Cawell, 1979
-e»

CD
              CC-FID - gas cliroiaatography-flarae lonlzatlon detectors
              CC-NPD • gas chromatography-nltrogen/phosphorus detectors

-------
tion of equilibrium concentration using Henry's law and  the  appropriate  parti-

tion coefficient.

     The solution head-space approach has been used with a much wider range  of

samples (Steinchen,  1976;  Gawell,  1979; DiPasquale et al_., 1978).  In this case,

the polymer has been dissolved in a suitable solvent and allowed to equilibrate

in a sealed vial at a constant  temperature  (60  to 90°C).  Sometimes  water  has

been added  to  the  solvent to  enhance the  equilibrium monomer  concentration

(Steinchen,   1976).    The  advantages  of this  approach  are:    (a)  head-space

equilibration is rapid, (b)  calibration procedure  is  simplified, and (c)  the

head-space gas/solution  partitioning of  the constituents  is  not  appreciably

affected by changes in the solvent phase.

     The third  method,  which  is  rarely used at the present  time,  consists  of

passing nitrogen gas through the heated polymers and trapping the volatile com-

ponents in dimethyl formamide  (Uhde and Koehler,  1967).  The  dimethyl formamide

solution is  subsequently analyzed  for acrylonitrile.  A variation of this  method

called the pyrolysis GC technique has also been used (Reichle and Tengler,  1968).

M.5.1.1.2  Pretreatment  of Food-Simulating Solvent Containing Monomer
           Residue

     The  pretreatment  of  food-simulating  solvents  for the determination  of

monomer content as  a result of migration from the polymer has  been done in three

ways.  The first method  that has been used is the direct analysis of the solvent

left in contact with  the polymer  for a  predetermined time  and  temperature

(Markelov and Semenenko,  1976;  Brown et^ al., 1978).   The  disadvantage of this

method  was  that it had  a poor sensitivity and  the  large  amount  of solvent

interfered with the small  amount of  acrylonitrile during the detection and quan-

tification stage (Brown  et aU,  1978;  Hartshorn,  1975).
                                        4-21

-------
     To avoid the solvent interference and to enhance the sensitivity of detec-



tion, either the second or the third method,  namely,  azeotropic distillation or



head-space analysis,  is presently employed.



     In the  azeotropic distillation  technique,  the  solvent is distilled  with



methanol and the  azeotrope  containing methanol and acrylonitrile is collected



for further analysis.  When water or 8$ ethanol was used as the food-simulating



solvent, the distillation was  done  by adding methanol directly to  the  solvent



(McNeal et^ al.,  1979).   When  3$ acetic  acid  was used as  the food-simulating



solvent, the solution was neutralized with sodium hydroxide before distillation



with  methanol  (McNeal  £t  al.,  1979).   When  heptane  was used  as  the  food-



simulating solvent, it was extracted with  water  and  the  water  extract was dis-



tilled with  methanol (McNeal _et^  al.,  1979).  The average  recovery of  acrylo-



nitrile from the azeotropic  distillation of  all  the extracts varied  between



29.7 + 2.6$ and  32.6 + 2.8$ (McNeal et al.,  1979).



     The third  method employs  the  principle  of the head-space  equilibration



technique.  The solvent was  introduced  in a  sealed vial and allowed to equili-



brate at a predetermined  temperature for  a  certain  length of time (Chudy and



Crosby, 1977; DiPasquale et al., 1978; Gawell, 1979).



4.5.1.2  Detection and Quantification



     Although a polarographic (Uhde and Koehler, 1967) and a titrimetric method



(Roy, 1977) have been used  for  the  determination of  monomer in polymers, these



methods lacked  adequate  sensitivity  for  the  determination of  low  levels of



acrylonitrile.  The method  used  almost  exclusively at the  present  time is gas



chromatography,  either  with flame  ionization  detectors  or with nitrogen/phos-



phorus detectors.  The sensitivity and selectivity of NPD makes it a preferable



method over FID.
                                        4-22

-------
4.5.2  Conclusions




     The  best  available method  for the  determination of  monomer residue  in




polymers seems  to  be the solution head-space and GC analysis.  For the determina-




tion of the extent of monomer migration in food-simulating solvents, both azeo-




tropic distillation  and  solution head-space  equilibration  with subsequent  GC




analysis are the  preferred methods.  The  preferred  stationary  phases that have




been used for  non-aqueous injections are  Carbowax-20M  and Carbowax-1500 (Chudy




and Crosby, 1977;  DiPasquale et al., 1978; Steinchen,  1976;  Gawell, 1979).  For




aqueous injections, gas-solid chromatography with porous  polymer packings such




as Chromosorb-102 (DiPasquale et al.,  1978),  Chromosorb-101 and  -108  (McNeal




e£ al., 1979)  and Porapak-QS and -S (Brown et al.,  1978)  have been  used.




U.6  OTHER MEDIA




     Acrylonitrile has  been determined in many other media such  as  tobacco,




production streams,  and  foods and  grains consumed  by humans.   Since  acrylo-




nitrile-containing pesticides  have  been voluntarily withdrawn  from the market




and the  other  media  have  no  direct  bearing on  general population  exposure,




acrylonitrile analysis in these media will not be discussed.




1J.7  GENERAL METHODS FOR THE ANALYSIS OF ACRYLONITRILE




     This section presents  the general methods available for detection and quan-




tification of acrylonitrile  without  regard to the medium in which it is present.




Many of the methods  have been  applied  for analysis  of acrylonitrile present in




more than one  medium.   This approach has  been  adopted so  that  replication of




discussion of the same method from medium to medium  is avoided.




     The identification and  quantification methods can be  divided into two cate-




gories:   one based upon the chemical reactivities  of  the functional groups in




acrylonitrile   and    the    other   based   upon   instrumental   techniques.
                                         4-23

-------
     The chemical techniques that  have proved useful are based on (a)  hydrolysis



of the nitrile group and (b) additions  to  the  double  bond.  In  the  hydrolysis



method, acrylonitrile is hydrolyzed to ammonia and acrylate  ion by a strong base.



The resulting ammonia can be determined colorimetrically by the Nessler's method



(Aarato and  Bittera, 1972; AIHA,  1970)  or by  NaOCl  and Na-salicylate  in  the



presence  of   Na-nitroferricyanide  (Rogaczewska,  1976).    Alternatively,  the



liberated ammonia can be determined  by titrimetric method (Gunther  and Blinn,




1955).



     Colorimetric, titrimetric, and  thin layer chromatographic  procedures have



been developed based  upon the addition reaction  of acrylonitrile.  In one colori-




metric procedure,  acrylonitrile is  brominated  with (Krynska,  I960;  Russkikh,



1971;  Lawniczak,  1977) or  without (Russkikh,  1973; Nazarova and Nakrap,  1978)



u.v. light.  The excess bromine is neutralized  and the cyanogen bromide that is



formed is allowed to react  with a benzidine-pyridine solution to form a colored



complex (Nazarova and Nakrap,  1978).   In another  method,  the  acrylonitrile



complex that  is  formed with pyridine in the presence of a  basic  hypochlorite



solution at 60-65°C is measured at Mil nm (Hall and Stevens,  1977).



     The titrimetric procedures have  been used following the reaction of acrylo-



nitrile with  excess Na^SO,  (Burkart £t al., 1961; Terent'ev and Obtemperanskaya,



1956;  Taubinger,  1969),  thioglycolic acid (Stefonescu  and Ursu,  1973;  Covic-



Horvat Q  a^.,  1970),  NaHSO,  (Kostin and Vidanova, 1957),  dodecanethiol (Roy,



1977), or  lauryl mercaptan (Berck,  1975).   The excess  reagent is  then back-



titrated with the appropriate titrant in the presence of an indicator or by the




potentiometric method.



     A method  based on the interaction of  acrylonitrile  with  alkaline KMnO^



solution,  which  produces a  change  in the  permanganate color, has  been used to



determine the concentration of  acrylonitrile.   In this method, the concentration
                                         4-24

-------
was determined  by  comparison of the  color  from a calibration curve  (Gisclard

et al., 1958).

     In the  thin  layer  ehromatography (TLC)  procedure,  the mercuric  acetate

adduct of acrylonitrile was separated by TLC on silica gel.   The  solvent system

used  was   described  by  Braun  and  Vorendohre (1963).   In a  recent  method

(Plieninger and Shartna, 1978),  the indole adduct of acrylonitrile was  separated

by TLC procedure.

     The instrumental techniques for  identification and quantification have used

gas  ehromatography,  polarography,  infrared and  u.v. spectroscopy,  and  mass

spectrometry.    Of these  methods,  mass spectrometric  is  the most  unambiguous

method for the identification of acrylonitrile..

     The vast  majority of  GC procedures have  relied on  flame ionization detection

(FID) .  Electron capture detection resulted  in one-fifth  the sensitivity of FID

(Barrett,  197^, cited  in  Going  eit  al..  1978).   The use of nitrogen/phosphorous

detectors (NPD)  has resulted  in  a dramatic increase in the sensitivity  of detec-

tion.  The insensitivity  of  this  detector toward  compounds  that  do not contain

nitrogen   and   phosphorous  eliminates  interferences  from  many   compounds

(DiPasquale et al., 1978).   Neither  carbon  disulfide  nor  N-containing solvents

are suitable for NPD.

     The selection of  column packing materials  for the separation of acrylo-

nitrile from interferences depends on the source of the sample.  Numerous packing
                         •
materials have been used in the past.  Based on their chemical characteristics,

the stationary liquid  or  solid  phases  that  have been used for the  analysis of

acrylonitrile can be summarized as follows:

     Polyglycol:   (a)  Carbowax-1500  (Babina,  1979;  DiPasquale  et al.,  1978;
     Steinchen, 1976; Gawell, 1979) (b) Carbowax-1510 (Chudy and Crosby,  1977)
     (c) Carbowax-400 (LysyJ, I960; DiLorenzo and Russo,  1969) (d) Carbowax-20M
     (Chudy and Crosby, 1977; DiPasquale et  al., 1978; Gawell, 1979)  (e) SP-1000
     (Marano  et  al..  1978;  Gagnon  and Posner,   1979)  and  (f)  Di-glycerol
     (Ustinovskaya et al., 1977)
                                         4-25

-------
     Hydrocarbon:  (a)  Apiezon (Babina, 1979) (b) Tween-80 (Chopra et al.,  1978)

     Esters and  Polyesters:   (a) Polyethylene  glycol adipate (Panova et al.,
     1969; Kleshcheva et al.., 1971;  Markelov and Semenenko, 1976)  (b) Neopentyl
     glycol succinate (Kleshcheva et al.,  1971;  Korzhova et al.,  197*0  (c) Tri-
     ethylene glycol butyrate (Pokrovskaya  and  Frolova,  1969)  (d)  Polyethelene
     glycol succinate (Lysyj,  I960) (e) Dioctyl phthalate  (Nestler and Berger,
     1965)

     beta-beta'-oxydipoprionitrile (Pokrovskaya  and  Frolova,  1969; Reichle and
     Tengler, 1968; DiLorenzo and Russo,  1969)

     Silica gel ASK .(Ivanenko and Lukashevaskaya, 1976)

     Porous Polymers:  (a)  Porapak  Q  (Balak et  al.,  1977;  Campbell and  Moore,
     1979) (b)  Porapak QS (Brown et  al.,  1978) (c)  Porapak  N (Russell,  1975;
     Tanaka e_t al.., 1975) (d) Porapak S (Brown et al_., 1978) (e)  Chromosorb 101
     (McNeal et al.,  1979;  Brown et al., 1978) (f) Chromosorb 102 (Marano et al.,
     1978; DiPasquale et al., 1978)  (g) Chromosorb 101 (Going et al., 1979)  (h)
     Chromosorb 108 (McNeal et al.,  1979)

     Methylsilicone:   (a) SE 30 (Berck, 1965) (b) DC-200 (Beaumont and Garrido,
     1979)

     Aminoalcohol:  THEED (Hughes and Horn, 1977)

     Tetracyanoethyl:  Pentaerytritol  (Ustinovskaya  e£  al..  1977; Deur-Siftar
     and Svob, 1976;  Marrs et al., 1978)

     The  polarographic  method was  used  most extensively  in the  past  for the

determination of acrylonitrile.  Tetramethyl ammonium iodide (Sevest'yanova and

Tomilov, 1963; Gorokhovskaya and Geller,  1962;  Uhde and Koehler,  1967;  Chao and

Ch'en,  1966;  Deus and Hamner,  1957; Lezovic and Singliar, 1977; Mekhtiev et al.,

1968; Klyaev  et^ al., 1966; Rogaczewska,  196H),  tetramethyl ammonium hydroxide

(Berck,  1962; Sevast'yanova et al_.,  1966),  and LiCl (Sevest'yanova and Tomilov,

1963; Bogaczek and Joworski, 1970) have been used as  the supporting  electrolyte

for the dropping mercury electrode.   Both standard calomel and silver were used

as the reference electrode.  A continuous polarographic method  has  been used for

the determination of  acrylonitrile in  industrial streams  (Bogaczek  and Jaworski,

1970).

     Spectroscopic methods  using both infrared and ultraviolet techniques have

been used for the detection and quantification of acrylonitrile.  The  sensitivity
                                         4-26

-------
of  acrylonitrile determination  by  the  earlier  IR  methods  was  rather  poor



(Scheddel, 1958;  Karaenev et al.,  197 *0 .   In recent years, however, the  use  of



multiple reflection, which,  in essence,  has the effect of increasing  the  cell



path-length, has increased the sensitivity significantly (Kurapov £t al^., 1977;



AIHA, 1970;  Beaumont and Garrido, 1979).  The same principle has been applied for



the continuous monitoring of acrylonitrile at 10.5  run with a  portable  infrared



analyzer (Jacobs and SyrJala, 1978).



     An infrared laser technique,  called  laser stark spectroscopy, that applies



electric fields to perturb the molecular  rotational energy levels has been used



to enhance and modulate  the  absorption of  acrylonitrile.   Using the P(28)  line



from a CO. laser  and a MO cm IR cell, the  method was shown to  detect 0.03 ppm of



acrylonitrile in air (Sweger and  Travis,  1979).



     Acrylonitrile in solution has been  determined by u.v. absorption at 195 run



(Petrova e£ al.,  1978)  and at 210 nm (Brieger £t al.,  1952).



     Mass spectrometry  alone  is  rarely  employed  for  the  quantification  of



acrylonitrile.   It  is  usually used as a confirmatory technique.  In one study,



however,  it  was used  for monitoring  acrylonitrile  concentration in  process



streams (Thomson, 197*0 .  In combination with GC, mass spectrometry was  used for



confirmatory identification  (Grote  et al.. 1978;  Marano  et  al..  1978;  Tanaka



et al., 1975; Going et al.,  1979;  McNeal  et al.,  1979).  The use of multiple ion



monitoring mode increases the detection limit five-fold over full mass scan mode



(McNeal et al., 1979).



      Finally,  detector tubes for  area  monitoring  (Kobayashi,  1956)  and gas



badges for personnel monitoring (Silverstein, 1977) of  acrylonitrile have been



proposed.
                                         4-27

-------
                        5.  SOURCES IN THE ENVIRONMENT

5.1  PRODUCTION PROCESSES

     Acrylonitrile can be produced by the following methods:

     (a)  Oxidation of  propylene in the  presence  of ammonia (ammoxidation  of
          propylene)   using  either a bismuth  phosphomolybdate or  uranium-base
          catalyst;

     (b)  Addition of hydrogen cyanide to  acetylene using a cuprous  chloride
          catalyst;

     (c)  Catalytic reaction of propylene with nitrous oxide;

     (d)  Reaction of ethylene oxide with hydrogen  cyanide,  followed  by  cata-
          lytic dehydrogenation of ethylene cyanohydrin;  and

     (e)  Ammoxidation of propane.

     Processes (a) through (d) have been  used  for  the commercial production of

acrylonitrile, and process (e)  has  been studied on  a pilot scale (Hughes  and

Horn, 1977).  Since 1971, however, process (a)  (the  ammoxidation of propylene) is

the  only  process  that has been  used commercially in  the  United States.   The

process is patented by the Standard Oil Company (SOHIO) and is  known as the SOHIO

process.

5.2  ACPYLONITRILE PRODUCERS

     The producers of acrylonitrile  monomer in the  United States  are  given in

Table  5-1,  while  distribution for  1980  and  growth projections through  1981

figures are-given in Table 5-2.
                                         5-1

-------
   Table 5-1.  Producers of Acrylonitrile in the United States  (Grume,  1982)

Producer                                             Capacity,  Mg x 10^


American Cyanimid
  Westwego, LA                                          123
DuPont
  Beaumont, TX                                          159

DuPont
  Memphis, TNa                                          136

Monsanto
  Alvin, TX                                             113

Monsanto
  Texas City, TX                                        191
Vistron
  Lima, OH                                              136
Vistron
  Victoria, TX                                          270
TOTAL                                                 1,128

aDupont has announced plans to close this plant (C and EN Oct.  11, 1982)
    Table 5-2.  Distribution of Acrylonitrile in 1980 and Projected Growth
                Through 1984 (Chemical Marketing Reporter,  1980)

                         Acrylonitrile                 Projected Annual
Distribution             ( Mg x 103)a                        Growth
Consumption
Imports                    Negligible
Exports                       198                           Decline

Production at Vistron, Victoria, Texas, is not included


5.3  ACRYLONITRILE USES

     Acrylonitrile  is  used primarily  as  a  raw material  in the  synthesis  of

acrylic and  modacrylic fibers,  ABS  and SAN  resins,  nitrile  rubbers,  adipon-

itrile, acrylamide, and  barrier resins.  Other miscellaneous  uses include the

production of  fatty amines and  their  derivatives,  cyanoethylation  of  various
                                         5-2

-------
Table 5-3.  Primary Uses for Compounds Synthesized  from  Arylonitrile-Containing
            Compounds (Suta, 1979)
Compound
Uses
Acrylic and Modacrylic Fibers
ABS Resin
SAN Resin
Nitrile Elastomers
Adiponitrile


Acrylamide
Nitrile Barrier Resins
More than 60$ of these fibers is used
in apparel.  Carpeting is the second
largest use.  Home furnishing uses
include blankets, draperies, and upholstery.
Industrial uses include sandbags, filter
cloths, tents, and tarpaulins.  The
fibers are also used in synthetic hair
wigs.

Its major markets are pipes and pipe
fittings, and automotive components.
Other important markets are large appliances,
housing for business machines and telephones,
recreational vehicle components, toys,
sporting goods, sheeting material
for luggage, and food containers.

Its primary uses are for drinking tumblers
and other houseware items, for automobile
instrument panels, instrument lenses,
and food containers.

Its major uses are in rubber hose,
seals, gaskets, latex, adhesives, polyvinyl
chloride blending, paper coatings,
and pigment binders.

It is hydrogenated to hexamethylenediamine,
which is used to produce nylon.

Its largest use is in the production
of polyacrylamides for waste and water
treatment flocculants.  Other acrylamide
products are used to aid sewage dewatering,
and for papermaking strengtheners and
retention aids.

They are used in the manufacture of
non-beverage containers for glue, nail
polish, correction fluid, air freshener,
contact lenses, tooth brushes, and
combs (Miller and Villaume, 1978).
                                        5-3

-------
alcohols and amines, fumigant formulations, as an absorbent, and as an anti-stall

automotive additive (Miller  and Villaume, 1978).   A flow diagram  summarizing

direct and indirect uses of acrylonitrile  is  given  in Figure  5-1.   The primary

uses  of  the compounds  that  are  synthesized  from acrylonitrile-containing

compounds are presented in Table 5-3-

5.H  CONSUMPTION OF ACRYLONITRILE BY PRODUCT

     A breakdown of acrylonitrile consumption and projected growth  of products

using acrylonitrile in their manufacture is given in Table 5-H.
       Table 5-U.  Acrylonitrile Consumption and Project Growth of Products
                   Using Acrylonitrile
Product
Acrylic and
modacrylic fibers
ABS and SAN resins
Nitrile elastomers
Adiponitrile
Acrylamide
Nitrile barrier
resins
Other
1977
Consumption
(Mg x 103)a
331
142°
24
73
24
9
33
Projected
Annual
Growth 1 980
Through Consumption
1982($)a (Mg x 103)b
4.5-5.5 345
7.5-9.5 172
2.0-3.0
10.5-12.5
8.0-10.0
12.0
4.0-6.0 147
Projected
Annual
Growth
Through
1984
None
Some Growth
—
—
—
__
Decline
aSuta, 1979
bCMR, 1980
C126 x 103 Mg for ABS and 16 x 103 Mg for SAN resins
     The U.S. manufacturing plants that use acrylonitrile in production (except

for the plant producing adiponitrile)  are listed  in Tables 5-5 through 5-8.  The

only  U.S.  producer  of adiponitrile  from  acrylonitrile is  Monsanto.    Their
                                         5-4

-------
ACRYLONITRILE-
                    FIBERS•
                    NITRILE RUBBER
                      & LATEXES	
                    ACRYLAMIDE
                    ADIPONITRILE-
                    GLUTAMIC ACID	

                    ABS & SAN RESINS
                    BARRIER RESINS
                    CYANOETHYLATION
                    FATTY AMINES
 ACRYLIC
 MODACRYLIC
  POLYACRYLAMIDE -
  ADHESIVES  &  FILMS
  DYES
  PHOTOGRAPHIC EMULSIONS
  INTERNAL PLASTICIZERS
  NYLON
FLOCCULENT
SIZING PAPER
PLASTICS
THICKENING AGENT
-IMONOSODIUM GLUTAMATE
                Figure 5-1.  Flow Diagram for Acrylonitrile Usage
                             (NIOSH, 1977c)
                                          5-5

-------
        Table 5-5.  Producers of SAN and ABS Resins (Crume, 1982)


     Producer                                     Capacity (Mg x

ABTEC (Mobay)
     Louisville, KY                                         31.8
Borg-Warner
     Ottawa, IL                                            105
     Washington, WV                                        136
     Port Bienville, MS                                     NA
Dow
     Allyns Point, CT                                       29.5
     Midland, MI                                            65.9
     Pevley, MO                                             29.5
     Torrance, CA                                           3*4.1
     Irontown, OH                                           29.5

Monsanto
     Addyston, OH                                          159
     Muscatine, IA                                          56.8
     Springfield, MA                                        13.6
USS Chemical
     Scotts Bluff, LA                                       90.9
NA = Not Available
  Table  5-6.   Producers  of Acrylic  and  Modacrylic  Fibers (Crume,  1982)

     Producer                                     Capacity (Mg x 10^)

American Cyanamid
     Milton, FL                                             59.1
Badische
     Williamsburg, VA                                       36.8
DuPont
     Camden, SC                                             77.3
     Waynesboro, VA                                         61.1*
Tennessee, Eastman
     Kingsport, TN                                          16.U
Monsanto
     Decatur, AL                                           1H5
                                         5-6

-------
  Table 5-7.  Producers of Nitrile Rubbers and Elastomers (Crume, 1982)


     Producer                                Capacity (Mg x l(r)
Copolymer Rubber
     Baton Rouge, LA                                6.2

B.F. Goodrich
     Akron, OH                                     1H.1
     Louisville, KY                                28.6
Goodyear
     Akron, OH                                      7.5
     Houston, TX                                   18.2

Uniroyal
     Painesville, OH                               16.4

Reichold
     Cheswold, DE                                  10.14
             Table 5-8.  Producers of Acrylamide (SRI, 1978)
     Producers                               Capacity (Mg x 10 )
American Cyanamid
     Linden, NJ                                     36a
     Westwego, LA
Dow
     Midland, MI                                    23
Nalco
     Garysville, LA                                  U.5

aCombined total for both plants
                                         5-7

-------
facility at Decatur,  Alabama,  produced 67 x 10^ Mg of adiponitrile in 1977 (SRI,




1978).




5.5  SOURCES OF EMISSIONS




     Acrylonitrile emissions occur during: (1) monomer  and  polymer  production;




(2) transportation;  and  (3)  end  product usage.   These  emission sources  are




discussed below.




5.5.1  Monomer and Polymer Production




     Acrylonitrile emission estimates for monomer, ABS-SAN resin, acrylic fiber,




nitrile elastomer, and adiponitrile  production are given  in Tables  5-9  through




5-13i  respectively.  (Other acrylonitrile emission sources are relatively minor




and are not included).  These estimates are based on information provided by  both




industry and  EPA,  and are included  in a  report  prepared by Suta (1982).   The




estimates  assume  full capacity;  however,  in  recent years actual  production




levels have been somewhat below full capacity.  Actual production levels were not




used in the estimates because production levels vary considerably from  year to




year.     Also,  some  plants  consider   their  production  levels  proprietary




information.  The  estimates of public exposure potential presented elsewhere in




this  report  are based on  the  emission  estimates contained in the  Suta (1982)




report.



     EPA has  independently calculated  the emission  estimates contained in the




Suta (1982) report (Crume, 1982).   The  current EPA estimates agree closely  with




the Suta estimates  for most acrylonitrile production plants.   The  current EPA




estimate for total acrylonitrile emissions from the entire industry differs from




the Suta total by  less than +2$.
                                        5-8

-------
       Table 5-9.  Estimated Atmospheric Emissions of Acrylonitrile from
                   Monomer Production Facilities (Suta,  1982)

                                                Acrylonitrile  Emissions
Producer                                              (Mg/year)
American Cyanamid
  Westwego, LA                                             9H.7

DuPont
  Beaumont, TX                                             78.8
  Memphis, TN                                             216

Monsanto
  Alvin, TX                                                95.1
  Texas City, TX                                          162

Vistron
  Lima, OH                                                132
  Victoria, TX                                             23-3

  TOTAL                                                   802
                                         5-9

-------
     Table 5-10.  Estimated Acrylonitrile Emission Rates  from ABS-SAN
                 Resin Production (Suta, 1982)
                                             Acrylonitrile Emissions
Producer                                                 (Mg/year)


ABTEC (Mobay)
     Louisville, KY                                      5.0
Borg Warner
     Ottawa, IL                                         96.6
     Washington, WV                                    657
     Port Bienville, MS                                  1.3
Dow
     Allyns Point, CT                                    7.9
     Midland, MI                                        17.3
     Pevley, MO                                          1.1
     Torrance, CA                                        9.1
     Irontown, OH                                        6.2
Monsanto
     Addyston, OH                                       89.6
     Muscatine, IA                                     365
     Springfield, MA                                    10.0
USS Chemical
     Scotts Bluff, LA                                  158
     TOTAL                                           1,l|2U
     Table  5-11.   Estimated Acrylonitrile Emission  Rates  from  Acrylic
                 Fiber Production (Suta, 1982)

Producer                                     Acrylonitrile Emissions
                                                    (Mg/year)

American Cyanamid
     Milton, FL                                        1H2
Badisohe
     Williamsburg, VA                                  357
DuPont
     Camden, SC                                        352
     Waynesboro, VA                                    309
Tennessee Eastman
     Kingsport, TN                                      25.0
Monsanto
     Decatur, AL                                        91.0
     TOTAL                                           1,276
                                         5-10

-------
    Table 5-12.  Estimated Acrylonitrile Emission Rates from Nitrile
                    Elastomer Production (Suta,  1982)


Producer                                     Acrylonitrile Emissions
                                                    (Mg/year)


Copolymer Rubber
     Baton Rouge,  LA                                     H.3
B.F. Goodrich
     Akron,  OH                                         123
     Louisville, KY                                     63.H
Goodyear
     Akron,  OH                                          55.2
     Houston, TX                                         0.2
Uniroyal
     Plainsville,  OH                                    HO.O
     Reichold Cheswold,  DE                               8.1

     TOTAL                                             295
  Table 5-13.  Estimated Acrylonitrile Emission Rate from Adiponitrile
                 Production (Suta,  1982)


Producer                                     Acrylonitrile Emissions
                                                    (Mg/year)

Monsanto
     Decatur, AL                                        59.0
                                        5-11

-------
5.5.2.    Emissions During Transportation


     Estimates  concerning the  importance  of potential  acrylonitrile  spills


during  transportation-related  accidents are  presented in  Table 5-11.    These


estimates were originally developed  to determine the relative costs of different


modes of  transportation and were  not  intended  to indicate  the magnitude  of


potential spills for the entire acrylonitrile industry.  The estimates do suggest


the magnitude of potential spills from several  modes of transportation, but under


highly  idealized   circumstances.     (See,    for  example,  the   footnotes   to


Table 5-14).


     Some data on actual acrylonitrile spills have been reported to the Oil  and


Hazardous Materials  Spill  Information  Retrieval  System (OHM-SIRS) of the EPA.


From August 1970 to July 1975, 12 acrylonitrile spills were reported to OHM-SIRS,


10 of which occurred during transport.  Of these 10 spills,  7 occurred from tank


cars, 2 from barges,  and 1  from a tank truck.  However,  EPA cautions that only 10


to 20$ of all spills are ever reported (Miller and Villaume, 1978).   The Inter-


governmental  Maritime  Consultative Organization  estimated  that 11  tons  of


acrylonitrile were discharged into  the  sea  from  transport and handling in 1970


(NAS, 1975, cited in Miller and Villaume, 1978).


5.5.3  Emissions from End-Product Usage


     Another  source  of  environmental  contamination is  from  residual  monomer


release during  end-product usage.   Monomer residues occurring in end-products
                         •

are  given  in Table 5-15.  The level of  acrylonitrile  in  fibers  is  so low that


handling of the fibers is not a likely  source of acrylonitrile exposure.  This


conclusion is  supported experimentally by  Finkel  jet^ al.  (1979).  Even  if  the


product were heated, this would not  result  in a  significant release  of acrylo-


nitrile (Federal Register, 1978a).   Acrylonitrile  may  possibly be leached from


fabrics during laundering;  however,  research has not been conducted in this area.
                                         5-12

-------
             Table 5-1^.   Hazards of Acrylonitrile Transportation
                       (A.D.  Little, Inc.,  197H as cited  in
                       Miller and Villaume,  1978)
Hazard Parameter
   Barge
Truck
   Rail
Spill Pool
Radius (meter)

Hazard Radius
(meter)

Hazard Area
(m2)

Relative Exposure ($)
Urban/Rural

Expected Number
of Annual Spills
Probability of Ignition
Following Spill

Expected Annual Number of
People Exposed
Urban/Rural

Expected Annual
Property Damage ($)r
     61.0
    122
 H6,700C
  5,U60C
   8/92
  17.1

  38. «l

U,U50C


23/77
    31.7
    68.3
      0.0117

      0.30
   0.063
   0.25
13,1*00°
   27/73
    0.17
    O.UO
0.008/0.OOU    0.010/0.002     0.16/0.016
Urban/Rural
f
Recurrence Interval
( Years)
129/55

85.5
160/20

15.8
2H23/252

5.8
  Calculations are based upon the assumption that each mode of transportation
handles 100 percent of the quantity shipped, and that a total of 73,000 Mg
per year of acrylonitrile are shipped between two points.

  Area affected by spills into water which ignite.  Assumes entire spill
quantity contributes to burning pool.

c Area affected by spills on land which ignite.  If no ignition occurs the
exposed land area is equivalent to the pool spill area (irR  spill).

  For spills into water which do not ignite.  The water toxicity hazard dis
tance (meters) measured downstream from spill location for a 152 meter wide,
3.05 meter deep river flowing at 0.70 meters per second.   Assumes vertical
dispersion rate at 0.30 meters per minute until uniform mixing is achieved
throughout the entire depth of the river.  Thereafter, plug flow is assumed
with no synergistic or antagonistic reaction between the pollutant and the
receiving body of water.  For this situation the entire spill quantity contributes
to water.
  Expected number of people exposed annually and property damage is based
upon ignition of the flammable pool for both land and water based spills.
f
  Average number of years between accidents.
                                        5-13

-------
Table 5-15.  Monomer Residue in End-Products of Acrylonitrile
Product Name
Usage
Monomer Residue
(ppm)
Reference

Acrylic and Modacrylic
Fiber
Hycar
Kralastic and Paracril
UCAR-380
UCAR-1358
Acrylamide Monomer
Polyacrylamide
ABS Resin
SAN Resin
SAN Resin
Fabric
Rubber
Resin
Latex
Latex
See Figure 5-1
See Figure 5-1
Packaging
Containers
Containers
<1
0-100
50
250
750
50-100
1
2*
3-7
2-5
Miller and Villaume,
1978
Miller and Villaume,
1978
Miller and Villaume,
1978
Miller and Villaume,
1978
Miller and Villaume,
1978
Miller and Villaume,
1978
Kearney, 1978
Brown e_t al. , 1978
McNeal et al . , 1979
Gawell, 1979
                            5-14

-------
     Although no study has been made, it is possible that some acrylonitrile will

be released from automobile tires, since the rubber  polymers in the tires contain

large amounts  of  monomer residue.   Research by A.T. Kearney, Inc.   (Kearney,

1978) indicates that in non-food contact, ABS/SAN containers do not release any

acrylonitrile under normal use;  in contact with foods, however,  these ABS/SAN

containers  may release  acrylonitrile into  the foods.    Brown  et al..  (1978)

determined that SAN bottles (7 ppm residual monomer content)  in contact with 3%

acetic acid at H9°C for 1 month releases 0.013 ppm of acrylonitrile to the acetic

acid.   ABS resin (2H  ppm residual monomer content) under  the  same conditions

released 0.283 ppm of acrylonitrile (Brown et al..  1978) .

     Acrylonitrile may be present as an impurity in  products made from acrylamide

(Miller and Villaume,  1978).  However, the extent to which these products release

acrylonitrile  to   the environment  is  not  reported.    Fumigant  formulations

containing  acrylonitrile  were  once   used  as  pest  control  for  residential

buildings,  tobacco,  grains,   and  nuts (Davis  et al.,  1973).   However,  today

fumigants containing acrylonitrile are no longer in use.

5.5.JJ  Conclusions

     The major  sources of acrylonitrile  emissions  in the  U.S.  are monomer and

polymer production facilities. The estimated acrylonitrile emissions from these

facilities are shown below:
                                                  Estimated Acrylonitrile
                                                  Emissions (Suta, 1982)
          Production facility                     	(Mg/yr)	
          Monomer                                           802
          ABS-SAN resin                                    1U2H
          Acrylic and modacrylic fiber                     1276
          Nitrile elastomer                                 295
          Adiponitrile                                       59.0
                                                           3856
                                        5-15

-------
Although the  relative importance of  other potential sources  of  acrylonitrile



emissions is  difficult to assess,  these  emissions  are believed  to be  small



relative to monomer and polymer production.
                                         5-16

-------
              6.  ENVIRONMENTAL FATE,  TRANSPORT,  AND DISTRIBUTION



     the environmental fate of acrylonitrile  in  air,  water, and soil, is  dis-



cussed in the following sections.  The discussion is based on only a few studies



that have been conducted in the field.



6.1  ATMOSPHERIC FATE, PERSISTENCE, AND TRANSPORT



     Very few studies have been conducted to  investigate the  fate  of acryloni-



trile under atmospheric conditions. Based on  the similarity of the physical and



chemical properties of acrylonitrile and the olefins, however, it is possible to



predict the atmospheric fate of acrylonitrile from what is known about that class



of compounds.   Like  other olefins, acrylonitrile  is  expected to  undergo  both



chemical and  photochemical  reactions  in the  atmosphere.   These reactions  are



discussed individually in the  following sections.



6.1.1  Atmospheric Chemical Reactions



     Although no specific references are available, atmospheric oxidation reac-



tions typical of olefins  may take place with acrylonitrile.   For example,  oxygen



atoms formed as a result of the photolysis of  nitrogen dioxide in the atmosphere



usually add to the olefinic double bond.  Oxygen atoms  react  with  olefins  more



rapidly than with other unsaturated aromatic and  acetylinic hydrocarbons.  This



addition reaction forms an excited  epoxide that subsequently decomposes to allcyl



and acyl radicals (U.S. EPA,  1979).



     Hydroxyl radicals, formed as a result of atmospheric  photolysis of nitrous



acid and degradation  of  other  free radicals, add to  the  double  bond  of  the



olefins.  The rate constant for  this addition  reaction is about 10 times greater



than for the atomic oxygen olefin reaction (Morris and Niki, 1971).



     Atmospheric ozone is formed  in significant quantities when nitrogen dioxide



levels  in  the atmosphere  are  about  25  times greater  than  nitrogen  monoxide
                                       6-1

-------
levels.  Ozone, while  not  as strong an oxidizing agent as 0» or  »OH  radicals,


reacts with  olefins at appreciable  rates when  ozone  concentrations reach  or


exceed 0.25 ppm.  Ozone adds to the olefinic double bond forming an aldehyde and


a diradical.  The diradical may further decompose or may participate in reactions


with 02,  N02, and NO.


6.1.2  Photochemical Reactions


     The photochemistry of  acrylonitrile vapor at 213.9 nm was studied by Gandini


and  Hackett  (1978).   The  photolysis was  shown  to  proceed  via two  molecular


elimination pathways, one yielding acetylene and  hydrogen  cyanide  and the other


yielding cyanoacetylene and hydrogen.  The quantum yields  for the  two processes


were  determined  to  be 0.50  and 0.31, respectively.   In  the presence  of  such


photosensitizers as  xanthene, triphenylene, benzophenone,  acetophenone,  fluore-


none, and dibromoanthracene, the major product of photolysis of acrylonitrile in


solution  was  shown  to  be  1,2-dicyanocyclobutane  (Gale,  1971;  Hosaka  and


Wakamatsu, 1968) . The dicyanocyclobutane  is not  very stable, however, and it is


unlikely that the reaction will proceed in the gas phase.


6.1.3  Atmospheric Persistence and Transport


     Only one study  that experimentally investigated  the atmospheric persistence


of acrylonitrile is  available.  Joshi (1977,  cited in Suta, 1979)  estimated the


atmospheric half-life of acrylonitrile to be 9  to  10 hours.  An atmospheric half-


life of 9-10 hours is sufficiently long for aerial transport to play a signifi-
                        •

cant role in the distribution  of acrylonitrile in  the  neighborhood of emission


sources.   It has been calculated by Suta (1979) that  when the average wind speed


is  H meters/second, 86$ of the emitted  acrylonitrile will  survive  at 30  km


downwind from the source,  and 78$ will survive at 50 km downwind.
                                         6-2

-------
6.2  FATE, PERSISTENCE,  TRANSPORT,  AND BIOACCUMULATION IN AQUEOUS  MEDIA




6.2.1  Chemical Reactivity in Water




     The  chemical  stability of acrylonitrile  in water  at  different pH's  was




studied by Going e_t al. (1979) .  He spiked distilled water and  Mississippi  River




water with 10 ppm acrylonitrile.  Prior to spiking, the pH was adjusted to 4 or 10




or left unadjusted.  All samples were stored at room temperature in  Teflon-capped




vials for 1, 6, and 23 days.  There was no indication of sample decomposition in




distilled water even after 23 days at any  of the tested pH  values. The samples in




river water showed  decomposition on storage.  The sample with unaltered pH showed




complete decomposition after 6 days.   The sample stored  at  pH  10  showed  little




decomposition after 6 days  but completely decomposed after 23 days.  The pH *J




sample showed even  less decomposition  after  6 days and  only  23$  decomposition




after 23 days.  It is not  certain whether the decomposition of acrylonitrile is



at least partly due to microbial effect.   If  so, the extremes  of pH may  have an



inhibitory effect on the microorganisms in the river water.  Spiking sterilized




water with  acrylonitrile  and  monitoring sample decomposition  may provide  an




answer.




     Acrylonitrile, if  present in  surface waters  that are used  as  sources of




drinking water, may react  with chlorine or hypochlorite during the chlorination




step of the treatment process.  It  has been suggested that the reaction products




could be  a  mixture of OHCH2CHC1CN  and C1CH2CHOHCN  (Kondratenko et al.,  1971).




The presence of detectable levels of acrylonitrile in previously aerated surface




water (used as a source of drinking water), however, is not very likely  because




of rapid volatilization.
                                         6-3

-------
6.2.2  Photochemical Reaction in Water




     Another mode of acrylonitrile degradation may be photochemical  reaction in



water;  however,  little is known  regarding the photochemistry of acrylonitrile in



water in the concentration region likely to be present in natural  water bodies.



6.2.3  Degradation of Acrylonitrile by Microorganisms



     Limited data  suggest  that  loss  of acrylonitrile  from  water systems  via



biological  degradation  can be  expected.   Both  aerobic and anaerobic  microo-



rganisms are capable  of degrading acrylonitrile,  especially acclimated  micro-



organisms.  The  breakdown products  of  aerobic microorganisms may include ammonia



and acrylic acid (Mills and Stack,  1955),  followed by nitrification of ammonia



(Chekhovskaya et al.,  1966).   The  latter  authors found that  acrylonitrile at



concentrations  of 50 ppm or higher may inhibit nitrification.



     Mills and Stack (1955) suggested a mechanism  for the biological oxidation of



acrylonitrile.   Using microorganisms  from  the  Kanawha  River (WV)  that  had been



acclimated with  acrylonitrile for 27 days,  the Biological Oxygen Demand  (BOD) of



acrylonitrile was measured.  The rate of aerobic oxidation was quite rapid and



reached completion  in  five days.   As shown in  Figure  6-1A,  about 70$  of the



acrylonitrile was degraded.   From the nitrogen  balance  data,  the authors sug-



gested that  the biological oxidation  of acrylonitrile proceeds  by  an enzyme-



catalyzed hydrolysis of the nitrile group to acrylic acid and ammonia.



     The microbial fate of acrylonitrile in natural water was studied by Cherry



et al.  (1956).   Acrylonitrile (10 ppm)  was added to filtered aerated water from



the Hackensack  River (NJ).   Nitrogen and phosphorus nutrients were added to the



water.   The complete disappearance  of acrylonitrile from  water  took  about



20 days.  Subsequent redosing with  acrylonitrile reduced  the  degradation time.



These results are shown in Figure 6-1B. Similar  results were obtained at 25 and
                                         6-4

-------
                          496
                          DAYS OF INCUBATION
                                            9      10
- SO
a.
a.
2*°
o
o
x
o
o
   10
                Figure 6-IB
                             REDOSE
    REDOSE    REDOSE
                     REDOSE
              10
     20       30        40
       ELAPSED TIME (DAYS)
                        50
                            60
       80
                    2Q°C
                        5°C
      o
     u uj
     W.P  4O
      •*

     w 5  20

     ^5  o


     ° <"20i


Figure 6-1.
              10
                                  I
                          Figure  6-1C
                              I     I
20   30
60
70   60
DAYS
90   100   110
           Biological  Oxidation of Acrylonitrile in Aqueous Systems
           A (Mills  and Stack,  1955)
           B (Cherry £t al.,  1956)
           C (Ludzack  £t al.,  1958)
                                6-5

-------
50  ppm acrylonitrile;  that  is,  the  acclimated  microorganisms  degraded  the



acrylonitrile more rapidly than did the unacclimated microorganisms.



     Ludzack et al.  (1958)  also found similar results. These authors spiked Ohio



River water with 10 ppm acrylonitrile at 20°C.  As  shown in Figure 6-1C,  there



was a lag period of about a week,  followed by several days  of  rapid degradation



after  which  a plateau was reached.   By day  22,   another  period of  activity



occurred.    Redosing this water with  acrylonitrile produced no lag  period and



plateau but  produced  rapid degradation of acrylonitrile by the already accli-



mated microorganisms.   Evidently the degradation rate was temperature  dependent;



when a sample was redosed  at  5°C,  the  degradation  rate was found to  be slower



than that seen in the sample redosed at 20°C.



     Ludzack e_t al.  (1958)  noted that acrylonitrile was degraded more  rapidly by



microorganisms in Ohio River water than by microorganisms in aged  sewage.   They



also found that acrylonitrile was  more resistant to biological degradation than



aceto-, adipo-, benzo-, and lacto-nitriles.



     The  aerobic  degradation of  acrylonitrile  in   water can  also proceed via



activated sludge.   Experiments conducted by Dow  Chemical Company (NAS,  1975,



cited  in  Miller  and Villaume, 1978)  indicated  almost complete degradation of



acrylonitrile to ammonia in 20 days.  The effectiveness of acclimated activated



sludge for  the  rapid degradation of  acrylonitrile in water was  also  shown by



Ludzack et al. (1961) .



     The  effectiveness  of   activated   sludge  systems   for   almost  complete



biodegradation of  acrylonitrile  in industrial wastewater  was also shown  by



Kincannon and  Stover, 1981  and  Freeman e_t  al. (1981) .   The latter  authors



demonstrated that the residue level of  acrylonitrile in wastewater  treatment



basins after aerated  activated sludge treatment was  below  the  detection limit



(<0.1 ppm).
                                        6-6

-------
     Kato and  Yamamura (1976)  discovered that  aerobic  microorganisms of  the



genus Nocardia were capable of degradation of cyanides and  nitriles.   More  than



90J of the acrylonitrile was degraded by these microorganisms.



     The proceeding studies show that acrylonitrile can be degraded aerobically.



Ludzack  et  al.  (1961), Lank  (1969,  cited in  Miller  and Villaume, 1978),  and



Hovious et al. (1973,  cited in Miller and Villaume,  1978)  studied acrylonitrile



degradation  under  anaerobic  conditions.   Lank  (1969,   cited   in  Miller  and



Villaume, 1978)  found that acrylonitrile at a  concentration of 10  ppm could be



treated  by  anaerobic  digestion.    Hovious  et al.   (1973,  cited  in Miller and



Villaume, 1978),  however,  determined  that, even  at a concentration of  50  ppm,



acrylonitrile was inhibitory to some anaerobes. The  inhibition was not complete,



so some  residual  activity remained.   Acrylonitrile1s inhibition  of  anaerobic



digestion by microorganisms was also confirmed by Ludzack e_t al.  (1961).  These



authors recommended that the anaerobic digestion should not be used for treatment



of acrylonitrile-containing water.



6.2.1!  Bioaccumulation in Water



     A bioconcentration factor (BCF)  relates  the  concentration of a chemical in



water to the concentration in aquatic organisms.  It  is  important to determine



the BCF for acrylonitrile  in aquatic organisms  in order to evaluate  the levels of



human intake of  acrylonitrile from this source  and also to  assess  ecological



effects.  There are a number of theoretical correlation equations that have been



established to relate BCF  to  either  the octanol-water partition  coefficient or



the water solubility of the chemical.  These  equations are given  below:



          log BCF = 0.76 log KQW - 0.23 (Veith et al., 1979)



          log BCF = 0.512 log KQW + 0.12H (Neely  et al.,  197U)



          log BCF = -0.508 log S + 3.^1 (Chiou et al., 1977)
                                        6-7

-------
where K   = partition coefficient of  the chemical between octanol and water,  and



S = water solubility of the chemical  expressed in unol/1.



     The above equation of Veith et al. (1979) is applicable for the whole fish;



whereas the equations of Neely et al. (1974)  and Chiou et  al.  (1977)  are appli-



cable for fish muscle only.




     If the values  for KQW and  S for acrylonitrile are assumed to be 0.12 (Leo



et al., 1971)  and 1.385 x 10  umol/1 (Groet, 1978), respectively, the theoretical



values for BCF can be calculated to be 0.1 (equation of Veith  et al.. 1979)  for



the whole  fish  and  O.U (equation of Neely e_t al.,  1974)  and 2.0  (equation of



Chiou et al.,  1977)  for fish muscle.



     U.S. EPA (1978a) experimentally measured the  steady-state  BCF for acrylo-



nitrile in bluegills containing about 4.8$ lipids.   The experimental value was



48.  The BCF for lipid-soluble compounds is proportional to percent lipids (U.S.



EPA, 1979). The weighted average lipid content in all the aquatic foods consumed



by an individual in  the United States was calculated to be  3.0$ (U.S. EPA, 1979) .



An adjustment  factor of 3.0/4.8  = 0.625 was used to adjust  the  measured BCF from



the 4.8$ lipid  of the bluegill  to  the  3.0$  lipids that is the weighted average



for consumed  fish and  shellfish.  Thus, the  weighted  average bioconcentration



factor for acrylonitrile  in the  edible portion of all aquatic organisms consumed



by an individual in  the United States was calculated to be  48 x 0.625 = 30 (U.S.



EPA, 1980).



     It can be concluded  from  the above discussions that the experimental BCF for



acrylonitrile in consumable aquatic foods is about two orders of magnitude higher



than the calculated value.



6.2.5  Transport in Water



     Few studies are available  that  investigate the transport of acrylonitrile



in water other than the river water  studies  discussed  previously.   The partial
                                         6-8

-------
vapor pressure  of acrylonitrile  in  its water  azeotrope  is 80  mm Hg at  20°C



(Miller  and  Villaume,  1978);  this  pressure is  significant enough  to  cause



evaporation of acrylonitrile from  water.  Using the method of Billing (1977) , the



half-life of evaporation of  acrylonitrile  from  water with an assumed depth of



1 meter can be  calculated to be 795  minutes.   It  should be mentioned, however,



that no  experimental  data  are available  to demonstrate this transport  possi-



bility and as discussed  in Section 3.1.8 the calculated evaporative half-life may



not accurate for acrylonitrile in natural aquatic  systems.



     The removal of acrylonitrile  from  an aquatic system will take place  through



various  chemical  (photochemical,   oxidative   etc.),   biologicaland  physical



(evaporative, adsorptive  etc.)  processes  as discussed  in  subsections  6.2.1,



6.2.2,  6.2.3 and this subsection.   The  percent of  acrylonitrile  removal  through



each of the processes will  depend on the characteristics of the  aquatic  system.



For example,  in  biologically  active systems  <0.1$ of the acrylonitrile was found



to be removed from the aquatic media due to volatilization (Freeman et al., 1980;



Kincannon and Stover,  1981).  The  calculated half-life value given above  is only



for the fraction of acrylonitrile  that  is  removed from the water by the evapora-



tive route.



     The observations from the two studies of accidental spills of acrylonitrile



can be used to provide further insight  into the transport of acrylonitrile from



other medium into  water (see Section 7.3.1).   Both  these  spill incidents show



that it is possible to transport acrylonitrile from contaminated land to  surface



waters (since the medium of land to water transfer was percolated water).  In the



absence of evaporative  effect,  acrylonitrile can  be expected  to have  a  long



persistence in water.  The bacterial  decomposition  of acrylonitrile in soil will



probably be of little importance in cases of spills because the toxic effect of
                                        6-9

-------
the large spills on  the  bacteria.   If the spill occurs during winter, the  low



temperature of soil will  further decrease the influence of biodegradation.



6.3  FATE,  PERSISTENCE, AND TRANSPORT  IN  SOIL



     Few data are available on this subject.  Acrylonitrile can be  degraded by



soil fungi  (Giacin  et al.,  1973).  Fungi capable of acrylonitrile biodegradation



included Penicillium,  Aspergillus.  and  Cladosporium species (Giacin  et al..



1973)•  The products of decomposition  were probably carbon dioxide and  ammonia.



Although other microorganisms slowly degraded acrylonitrile, best  results were



obtained with  soil fungi.   The microbe  Nocardia  rhodochrous LL100-21  slowly



degraded acrylonitrile, but the  rate of degradation increased with  added acetate



(DiGeronimo and Antoine,  1976).
                                        6-10

-------
                     7.  ENVIRONMENTAL LEVELS AND EXPOSURE

7.1  ENVIRONMENTAL LEVELS

     Acrylonitrile  levels  in  occupational  atmospheres  (Marrs et  al.,   1978;

Sakurai et aK,  1978;  Martin,  1978)  and in industrial  point  sources  (Hughes  and

Horn, 1977; Hollingsed,  1978,  cited in Miller and Villaume, 1978;  Suta,  1979)

have  been determined.   One  published  study  conducted by  Midwest Research

Institute  (Going  et  al..  1979)  has   determined   environmental   levels   of

acrylonitrile.  This study  was designed to determine the levels  of  acrylonitrile

in  ambient  air  samples,  surface  waters,  and  soils  and  sediments   around

acrylonitrile and acrylonitrile polymer manufacturing  facilities.

     PEDCo Environmental has  recently completed a study monitoring levels of AN

near  U   acrylonitrile  producing  or  consuming  plants.    This  study  is  yet

unpublished.   Preliminary  results indicate acrylonitrile to have  been measured

at all four locations and, in  general,  levels around  user  plants were  greater

than around producing plants.

7.1.1 Atmospheric Levels of Acrylonitrile Around Its Major Production and
       Usage  Facilities

     The atmospheric levels of acrylonitrile  around manufacturing facilities are

given in  Table  7-1.   The  samples  were  collected by  passing  the  air  through

activated carbon over a period of approximately 21 hours for  all stations,  except

Monsanto in Decatur, Alabama,  where two  21-hour  samples were collected.   The

acrylonitrile from  the  activated carbon  was  desorbed by carbon  disulfide  and

analyzed by gas chromatography with a flame  ionization  detector.   The  average

recovery of acrylonitrile determined with spiked samples was 63^  (Going  et al.,

1979) .

     The values for maximal average  concentrations given in  Table 7-1 have been

derived in the following manner:  The sum of all  the  determined  concentrations
                                        7-1

-------
                              Table 7-1.  Atmospheric Monitoring Data for Acrylonltrile (Going et al_.,  1979)
Site
American Cyanamld,
New Orleans, LA
American Cyanamld,
Linden, NJ
Monsanto,
Texas City, TX
Monsanto,
Decatur, AL
DuPont,
Lugoff, SC
DuPont,
Waynesboro, VA
Borg -Warner,
Washington, WV
Ooodrich,
Louisville, KT
Monsanto,
Addyston, OH
Unlroyal,
Paneavllle, OH
Vlstron,
Lima. OH
Product
Acrylonitrlle
Acrylonitrlle
Acrylonltrlle
Aoryllo and
Modacryllc Fibers
Acrylic Fibers
Acrylic and
Modacryllc Fibers
ABS/SAN Resins
Nltrlle Elastomers
> ABS/SAN Resin
ABS/SAN Resins
Nltrlle Elastomers
Acrylonltrlle
> Acrylamlde
Distance (km)
of Collection
from Plant3
0.6-1.8
0.5-1.8
0.7-2.6
1.3-5.0
0.7-2.2
0.3-0.9
0.5-1.3
0.4-3.0
0.3-1.0
0.3-0.9
0.2-0.6
No. of Samples
Collected0
5
6
7
5
8
6
4
4
5
5
5

High
13.6
15.9
8.9
4.2
1.1
7.0
325.0
4.3
1.1
3.1
148.0
Concentration of Acrylonltrlle( iig/n^)
Low Maximal Average6
<0.1 2.6
<0.1 3.2
<0.3 3.2
<0.1 1.2
<0.1 0.3
<0.2 1.3
<0.2 84.0
<0.2 1.2
<0.2 0.5
<0.1 1.0
<0.2 3«.«»
*Tne distances of sample collection points are obtained from Suta, 1979.
 Soaetlnes replicate samples were collected from the same sampling point.
C3ee text for explanation.

-------
has been divided by  the number of determinations.  When the determined concentra-



tions were less than the detection limit,  the  data used for averaging are  the



values at the detection limit.



     The  highest  individual  concentration  from  this  monitoring  data   was



325 ug/nr and  the  lowest was <0.1 ug/m^  (Going et al.,  1979).   The  recorded



concentrations depend greatly on  the meteorological conditions,  the production



stage within the plant  at  the  time of sampling,  and  the presence  of  emission



control  devices  in  the plant.   This  is  reflected in the high maximal  average



acrylonitrile  level  (8H  ug/nr)  in one plant  and a low  level  (0.5 ug/nr)  in



another  plant,  even though  both  produced ABS/SAN resins.   Such high and  low



atmospheric levels  of acrylonitrile  are  also reflected  in  the  data for other



plant emissions, all manufacturing the same acrylonitrile (see Table 7-1).   A



comparison of  the experimental  monitoring data (Going  e_t  al.,  1979) with  the



dispersion modeling data of Suta (1979)  is given in Table 7-2.



     The chemical analysis technique used for generation  of the data presented in



Table 7-2  have  since  been  improved.     The  extraction   of   the  collected



acrylonitrile was performed  for  the monitoring data using carbon  disulfide.   The



present  accepted extraction procedure  uses a  mixture  of carbon  disulfide  and



acetone (19:1)  which improves recovery (95$) (Gagmon,  1979)  over that of carbon



disulfide alone (60-75$)  (Silverstein,  1977).



     A comparison of the dispersion modeling data  (Suta,  1979) with the  actual



monitoring data (Going  et aU, 1979) shows  that, although the difference between



the experimental concentrations and the concentrations  derived  from dispersion



modeling (Suta,  1979),  on  the  average,  was about  20$,  90$ of  the individual



values had much higher variations.  In many instances,  the agreement between the



two were poor.   Therefore, the need for more experimental monitoring data cannot



be overemphasized.
                                         7-3

-------
Table 7-2.   Comparison of Monitoring and Dispersion Modeling Data
            (Suta,  1979)
Plant/Location Distances
(km)
Acrylonitrile Concentration (mg/m^)
Monitoring
Dispersion
Modeling0

American Cyanamid
New Orleans, LA

American Cyanamid
Linden, NJ

Monsanto
Texas City, TX



Monsanto
Decatur, AL



DuPont-May
Cam den, NJ


DuPont
Waynesboro, VA
Borg-Warner
Washington, WV
B£FJl Goodrich
Louisville, KY



Monsanto
Addyston, OH

Uni royal
Plainsville, OH
Vistron
Lima, OH
0.50-0.99
1.00-1. 49
1.50-1.99
0.50-0.99
1.00-1.49
1.50-1.99
0.50-0.99
1.00-1.49
1.50-1.99
2.00-2.49
2.50-2.99
1.00-1.49
1.50-1.99
2.00-2.50
2.50-4.99
5.00-5.49
0.50-0.99
1.00-1.49
1.50-1.99
2.00-1.50
0.30-0.49
0.50-0.99
0.50-0.99
1.00-1.49
0.30-0.49
0.50-1.99
2.00-2.49
2.50-2.99
3.00-3.49
0.30-0.49
0.50-0.99
1.00-1.49
0.30-0.49
0.50-0.99
0.30-0.49
0.50-0.99
1.3
0.1
0.1
0.5
6.0
0.7
2.4
nd
3.8
0.9
5.2
1.2
2.3
0.8
nd
0.2
0.7
0.3
0.1
0.2
3.6
0.2
157.6
0.3
2.3
nd
0.2
nd
0.2
0.2
O.U
1.1
1.3
0.7
43.4
0.2
H.3
2.5
1.9
d
d
d
6.5
-
2.5
1.8
1.3
21.0
9.3
7.2
-
3.1
7.4
2.3
1.3
1.1
5.2
3.4
42.1
30.4
7.2
-
1.7
-
1.0
5.9
3.9
2.8
1.2
0.9
10.6
5.2
 nd = no data
 Estimated distance from the acrylonitrile production within the plant.
  Average of all monitoring stations within the indicated distances.
 cEstimated concentrations at the midpoint of the distances.
 TJispersion modeling estimates were not made for acrylamide plants.
                                     7-4

-------
     It is interesting to note that the atmospheric acrylonitrile level near the



American Cyanamid Plant in Linden, New Jersey, which produces only acrylamide,  is



comparable  to  that near  plants  that manufacture  other  products derived  from



acrylonitrile.   In deriving the sources of emissions (see  Section 5.5.2),  Suta



(1979)  made the assumption that  acrylamide production is  a  negligible source of



acrylonitrile emission.



7.1.2  Acrylonitrile Levels in Surface Waters



     The acrylonitrile monitoring data for surface waters is given in Table 7-3.



These data were obtained  by Going £t al. (1979).  Whenever possible, grab samples



were collected upstream and downstream of the plant  discharge  points.   In some



instances, discharged wastewater from the plants were collected for  analysis.



     Two complementary techniques, azeotropic distillation  and  purge-trap,  were



used for  reducing  interference and concentrating acrylonitrile in the samples.



The method used for quantification was GC with a Hall nitrogen-selective detec-



tor.   Samples  that appeared to  contain  acrylonitrile were confirmed  by GC-MS



analysis.  Analytical quality assurance was done by spiking and analyzing field



samples (Going et^ al., 1979)•



     The  two highest  levels  of  acrylonitrile shown in Table 7-3 were obtained



from  the  Monsanto plant  in  Decatur,  Alabama,   and  the  Uniroyal  plant  in



Painesville, Ohio. These values were high,  however, because these samples repre-



sent discharged wastewater  prior to adequate dilution  in surface water.  The high



acrylonitrile content  in the discharged wastewaters  is  an indication  that  an



effective control of  these  wastewaters is necessary to minimize the pollution of



surface waters.
                                         7-5

-------
                                         Table 7-3-  Acrylonitrlle Monitoring Data  for Surface  Waters  (Going e£ al.,  1979)
~o

CTi
                         Site
                                                            Product
  City, TX

Tennessee River near Monsanto,  Decatur,  AL

Wateree River near DuPont,  Lugoff,  SC

South River and Discharge Point at  DuPont,
  Vaynesboro, VA

Ohio River near Borg-Warner,  Washington, WV

Ohio River near BIPl Ooodrlch,  Louisville, KY


Ohio River near Monsanto, Addyaton, OH

Influent and effluent from treatment facilities
  from wastewater at Onlroyal,  Palnesvllle, OH

Discharge point at Ottawa River at  Tlstron,
  Lima, OH
Acrylic and modacryllc fibers

Acrylic fibers

Acrylic and modacyllc fibers


ABS/SAN resins

Hltrlle elastomers and
ABS/SAN resins

ABS/SAN resins

Nltrlle elastomers


Acrylonitrile and acrylaolde
                                                                                                            Concentration in water,  ug/1
                                                                                                              High
 3500

19.7

 n.d.


 1.9

 2.0


 8.0

 1300


 n.d.
                                                                                                                       Low

Mississippi River near American Cyanamld,
New Orleans, LA
Arthur Hill near American Cyananld,
Linden, NJ
Texaa City Ship Channel near Mom ant 0, Texas
Acrylonitrile

Acrylamide

Acrylonitrile
n.d . <0.1

O.B n.d.

0.1 <0.1
                                                                                                                                  0.1
                                                                                                                                   9.3
               *n.d. t not determined

-------
7.2  ACRYLONITRILE LEVELS IN SOILS AND SEDIMENTS




     The environmental levels of acrylonitrile in a few soil and sediment samples




are shown in Tables 7-1* and 7-5.  These data were  obtained  from  the  investiga-




tions of Going et al. (1979).




     The  soil  samples were  collected from  the air sampling locations.   The




collection of sediment samples  was restricted by the accessibility of the sedi-




ments from the waterbody.  The methods of  analyses were the same for both soil




and sediment samples.  The sediments free from excess water and the soil samples




were ultrasonically agitated with water.   In most  analyses,  the  water extracts




were directly injected into  the gas chromatograph equipped with a Hall nitrogen-




selective detector.  One sample each from soil and  sediment was further purified




and concentrated by the purge-trap technique before GC injection.  The recoveries



of acrylonitrile from the samples were not determined.




     It is obvious from Tables  7-^ and 7-5 that, with the exception of one soil



sample, the level of acrylonitrile in  all other  samples was  below the detection




limit of the method used.  This may be expected  in view  of the relatively high




water solubility  and high volatility of  acrylonitrile.   The  detection limit,




however, could have been lowered either with the  purge-trap or azeotropic distil-




lation of the water extract.




7.3  ENVIRONMENTAL EXPOSURE




     Population  exposure  from  environmental  acrylonitrile emissions  can take




place  through  four   principal  sources:     (1)   industrial  emissions   in  air;




(2) drinking water; (3) consumed foods;  and (1)  spillage during transportation.



The exposure from each of these sources is discussed below.
                                         7-7

-------
  Table 7-4.   Acrylonitrile Monitoring Data for Sediments (Going .et al.,  1979)
Site                                              Concentration (ug/kg)'
Mississippi River near American Cyanamid,                   <0.5
  New Orleans, LA
Tennessee River near Monsanto, Decatur,  AL                <50
Wateree River near DuPont, Lugoff,  SC                     <50
South River near DuPont, Waynesboro, WV                    <50
       figures were the lowest detection limit
   Table 7-5.  Acrylonitrile Monitoring Data for Soils (Going et al., 1979)
Site                                              Concentration (ug/kg)a
American Cyanamid,  New Orleans,  LA                          =0.5
American Cyanamid,  Linden,  NJ                              <50
Monsanto, Texas City,  TX                                  <100
Monsanto, Decatur,  AL                                       <50
DuPont, Lugoff, SC                                         <50
DuPont, Waynesboro, VA                                     <50
Borg-Werner, Washington, WV                                <50
B£F£ Goodrich, Louisville,  KY                             
-------
7.3.1  Exposure From Air Polluted by Industrial Sources



     The total number  of people expected to be exposed to  different  levels  of



acrylonitrile concentrations from different industrial sources was calculated  by



Suta (1979) and is given in Table 7-6.



     The estimated values were  derived on the basis of a dispersion modeling from



the emitted acrylonitrile concentration values and estimated population density



around  the  plants.   The  estimated  exposure  values were determined  for people



residing within 10 concentric rings (of various radii ranging  from 0  to 30 km)



about each plant.



     The estimated exposures in Table 7-6 are somewhat underestimated  for two



reasons. First,  the exposures beyond 30  km were not included in all the calcula-



tions.  Second,  the exposure from acrylamide  plant  emissions  was ignored, even



though the experimental data from Going e_t al. (1979)  indicated  such exposures



might be significant.



7.3.2  Exposure From Drinking Water



     Trace  amounts of acrylonitrile  have  been   detected  in drinking  water



(Kopfler et a^.,  1976),  although the amount was not quantitated.  In the absence



of such data, it is impossible to evaluate the human intake  from this source.



7.3-3  Exposure From Foods



     The  three   possible  sources  of acrylonitrile  exposure  from foods  are:



(a) fish and shellfish;  (b)  food  containers and  packaging materials;  and



(c) foods fumigated with acrylonitrile-containing  fumigants.



     Edible aquatic organisms may  bioconcentrate acrylonitrile from contaminated



waters.  The weighted average  bioconcentration factor  for acrylonitrile in the



edible portion of all aquatic organisms consumed by Americans has been calculated



to be 30 (U.S. EPA, 1980).
                                         7-9

-------
           Table 7-6.  Estimates  of  Population  Exposures  to Atmospheric Acrylonitrile from Specific
                       Emission Source Categories (Suta, 1982)
Average Annual AN
Concentration
( ug/m)
15.0
10.0
5.0 -
1.0
0.50 -
0.10 -
0.050 -
0.010 -
0.005 -
0.001 -
19.9
114.9
9.9
U.9
0.99
O.U9
0.099
0.0149
0.009
0.004
AN
Monomer


U.900
36,000
260,000
720,000
1,602,000
8,000b
28,000b
ABS/SAN
Resins

2,700
T4.000
53,000
238,000
382,000
1,1407,000
1456, 000b
1,701,000b
Acrylic
and
Modacrylic
Fibers


390
5,000
mo, ooo
UUO.OOO
320,000
26,000b
15,000b
Nitrile
Elastomer


18,000
146,000
1421,000
1425,000
1,25*4,000
265,000b
292,000b
Adiponitrile




22,000
20,000
78,000
ob
ob
Total People
  Exposed
2,659,000'
14,254,OOO1
9146,OOO1
2,721,000
aTo convert for ng/nr to ppb, multiply by O.i<6.
 Exposures in these ranges are underestimates because calculations were only made for
 exposures within 30 km of each plant.
120,OOO1

-------
     The polymers and  copolymers containing residual acrylonitrile monomer could



migrate from the food-contact items to the food itself.   The  amount of migration



depends on the residual monomer content  in the polymer or copolymer, the time  of



storage, and the temperature of storage.  The effects of these factors on acrylo-



nitrile migration are shown in Table 7-7.



     Since the monomer migration  is substantial  from ABS/SAN resins to ethanol,



the FDA currently does not permit the use  of these containers for  alcohol and



carbonated beverages.  Gawell  (1979), using SAN bottles containing  3 to 5 ppm



residual monomer,  showed, however, that the migration in some samples of beer and



soft drinks amounted  to < 0.005 mg/kg.  The author did  not  specify the storage



conditions.  FDA  has  determined  that the migration  of  monomer from  SAN resin



containers (3-3 ppm residual  acrylonitrile)  to the  beverages could be as high  as



11 ppb (Flood,  1980)  after 96 days contact  at 120°F.



     Under FDA regulations,  copolymers of acrylonitrile  listed in Table 7-8 are



permitted  in food-contact  applications  including  food  packaging,  such  as for



luncheon meats,  peanut butter, margarines,  fruit  Juices,  and vegetable oils.



Some acrylonitrile exposure may result from these sources, although it is diffi-



cult to  quantify the  amount.   FDA  has determined  that  the migration  of the



monomer from the containers to vegetable oil and margarine  could  be  as high  as



37 ppb (Flood,  1980).



     Foods that have been  fumigated with acrylonitrile-containing fumigants pre-



sent a risk of  acrylonitrile exposure.  Fumigants  containing acrylonitrile for



grain pest control have been voluntarily withdrawn from the market.  Other foods,



such as  walnuts,  are  no longer  fumigated  with acrylonitrile-containing fumi-



gants.   The residue level of acrylonitrile  in walnuts that  have  been fumigated



may range  from  7.5-17.5  ppm after 2  days  to 0-8.5  ppm after 38  days (Berck,



I960).   Finally, acrylonitrile has been used as  a  fumigant  for stored tobacco.
                                        7-11

-------
Table 7-7.   Acrylonitrile Migration Under Different Storage Conditions
Migrating Nature
Solvent of
Plastic
Residual
Monomer
Content
of
Plastic
(ppm)
Time of
Storage
(months)
Temp, of
Storage
(°C)
Acrylo-
nitrile
Found
(ppm)
Reference


3$ acetic acid SAN




3% acetic acid SAN

3$ acetic acid ABS



8% ethanol SAN

7




3

24



10
25
1
2
3
4
5
2
2
1
2
3
5
?
?
49
49
49
49
49
49
66
49
49
49
49
49
49
0.013
0.017
0.022
0.028
0.038
0.012
0.023
0.283
0.557
0.723
0.920
0.078
0.126
Brown _et al. ,
Brown ^t al. ,
Brown jjt al. ,
McNeal et al.
Brown et al. ,
McNeal et al.
McNeal et al.
Brown e^ al. ,
Brown e^ al. ,
Brown et^ al. ,
Brown e^ al. ,
McNeal et al.
McNeal et al.
1978
1978
1978
, 1979
1978
, 1979
, 1979
1978
1978
1978
1978
, 1979
, 1979
                                    7-12

-------
       Table 7-8.  Amounts of Various Acrylonitrile Copolymers Used in
                    Food-Contact Applications  (Troxell,  1980)
Copolymer
ABS Resins
ABS Resins


ABS Resins

ABS Resins
SAN Resins

Nitrile Elastomers/
  Latexes
High Nitrile Resins
Polyvinylidine
  chloride -
  Acrylonitrile Resin
          Use£
     Amount used in 1977
(millions of pounds)
Refrigerator/freezer linings
Small appliances (motor housings,
  bases for blenders and can
  openers, etc.)
Packaging (sheet and film for
  blister packs)
Margarine tubs
Drinking tumblers, blenders, Jars,
  components of appliances, etc.
Hoses and paper coating
Mostly vegetable oil bottles
Cellophane > paperboard
            125
             10-15

             15

             <1
            =50
            =20
       Some of the applications may result in incidental  food  contact.
                                         7-13

-------
Guerin e_t al^ (1971*) have qualitatively determined the presence of acrylonitrile



in tobacco  smoke,  and,  the amount  has  been determined to  be 1-2 mg per  U.S.



cigarette smoked (IARC, 1979).   It is not clear whether the  source  of  acrylo-



nitrile is from fumigation  of the tobacco or from the combustion process  itself.



7.3.1*  Exposure From Spillage during Transportation



     The expected numbers of people in urban/rural  areas exposed annually  from



spills during  transportation of acrylonitrile by three  modes  of transportation



were calculated as  follows:  barge,  0.008/0.004;  truck,  0.010/0.002;  and  rail,



0.16/0.016 (Miller  and Villaume,  1978).   These figures represent  the cases  in



which the entire commodity  is transported by one mode. It  can  be concluded  that



transportation of acrylonitrile by  rail  (10.4$  of overall  shipment) poses the



greatest hazard and by barge (1.7$ of the overall shipment)  the least hazard.



     The preceding estimate was based on  exposure due to ignition of the  spilled



area and toxic exposure  from contaminated  surface water  due  to spill  (see



Section 5.5.3)•  However,  contamination of the ground water and the  subsequent



exposure from usage  of the water as a source of drinking water was not considered



in the calculation.



     Two recent cases  of accidental  spills leading to ground water contamination



show the possibility of  population  exposure  from this  source.   A spill  of



36,000 gallons of acrylonitrile  onto farmland (Gilford, Inc.,  2/22/77)  caused



contamination  of the nearby groundwater and a creek (Miller and Villaume, 1978).



Evidently, the  acrylonitrile  percolated  through the soil into the groundwater



and the creek.   For  several months after the spill,  the concentration of  acrylo-



nitrile in the  groundwater increased after it rained  (Miller and Villaume, 1978).



     Another spill  of  20,000 gallons  of acrylonitrile  (near  Mapleton,  IL,



12/23/77) caused  similar  contamination  of  the  groundwater and  creeks  located



near the spill  (Miller and Villaume,  1978) .  Monitoring data  from  five wells, all
                                        7-14

-------
located within 100 feet from  the spill site, showed high levels of acrylonitrile



in the water (16 mg/1 to 3520 mg/1) at the end of 108 days (Miller and Villaume,



1978).  Acrylonitrile levels  started decreasing about 170 days after the spill,



but acrylonitrile  had not completely disappeared from the well water even after



351 days.  Nine additional wells, located  at  an average distance of about 1000



feet from the spill site, showed  no  trace  of  acrylonitrile.   Tap water  at six



nearby residences (an  average  distance  of 1150 feet  from  the spill site)  con-



tained no acrylonitrile.  A little Marsh Creek  located  about  750 feet from the



site of the spill  showed 32 mg/1 of acrylonitrile at the end of 58 days after the



spill, but acrylonitrile finally disappeared after about 108  days.



7.3.5  Exposure From Thermal  Degradation



     Thermal  decomposition   of  polymers  containing  acrylonitrile  is  another



source of acrylonitrile in the atmosphere.  Pyrolysis  of the following polymers—



polyacrylonitrile  (Tsuchiya  and Sumi, 1977; Guyot et al^., 1978),  ABS/SAN resins



(Chaigneau  and LeMoan,  1974),  acrylonitrile-methacrylate  copolymers  (Guyot



et_ a^., 1978), vinyl chloride-acrylonitrile copolymers  (Tanaka et al.,  1975)—



produces hydrogen  cyanide and acrylonitrile.



     The  composition of  the  gases evolved during pyrolysis   and combustion of



polymers and copolymers of acrylonitrile  depends on a number of factors,  includ-



ing the nature of the polymer,  gas composition, gas flow rate, and heating rate



of the flame  (Sarofim e_t al.,  1973) •   Pyrolysis  prevails at lower temperature and



decomposition requires  higher temperature (Sarofim et^ al.,  1973) •  The pyrolysis



of SAN bottles with He and air at a flow rate of 1 to 3 1/min and heated at a rate



of  5°C/min  to  18000°C/min   produced  the  following  major nitrogen components



(Sarofim et al.,  1973) .
                                         7-15

-------
               HNC:  5-1* to U.3JE of total nitrogen in SAN
               Acrylonitrile:   7.5 to 38.0$ of total nitrogen in SAN

     The combustion of SAN bottles under a variety of conditions was studied  in

detail by Kaiser and Bergman (1973).  Their results are summarized below:

     The combustion of  SAN  bottles alone with  B3% excess air  produced  smoke,

odor, and toxic gases in  appreciable  amounts.   The concentrations of two major

nitrogen components and hydrogen cyanide in the flue gases were the following:

               Acrylonitrile:   270 ppm of total gases
               Methacrylonitrile:  34 ppn of total gases
               HCN:  1 ppm of total gases

From these results, the authors concluded that the  burning  of SAN  bottles  in

campfires, fireplaces, and outdoor trash burners could be hazardous.

     Mixed with household refuse or  the  usual commercial or  industrial  plant

wastes, a few percent  (2 to  4$) of SAN containers will burn in large incinerators

(about 1 ton/hr) without producing  any detectable level of HCN or acrylonitrile.

As a  matter  of fact, other than a slight  increase in NO production,  the SAN

addition to household  and commercial refuse will produce less harmful gases than

the burning of refuse without SAN.  This is probably due  to the fact that SAN  in

refuse aids in the completion of combustion of the refuse.

     As  the  size  of  the incinerator decreases,  combustion  of SAN  with  other

refuse will  increase  production of smoke,  particulate matter,  acrolein,  acid

gases, HCN, NO , and total hydrocarbons.

     Preliminary  estimates  of  ambient  concentrations and exposure  levels  of

acrylonitrile may  be  obtained from the modeling  of the  dispersion  of acrylo-

nitrile as a result of pyrolysis and combustion.   However,  in the absence of any

such dispersion modeling  data,  it  is  difficult to  estimate  the exposure level

from this source.
                                         7-16

-------
7.4  CONCLUSIONS



     Most acrylonitrile  exposures  result from ABS/SAN resin and  nitrile  elas-



tomer  production.   The  risk for  population exposure  from the  acrylonitrile



sources is dependent on a number of factors,  including the height of the release



point source;  that is, the higher release point will result in greater dilution



of the pollutants  at  ground level and also  spreads it over a  larger  area.   In



general,  for the same amount of emission and topography,  acrylonitrile emission



points with lower  elevation will result in  higher  ground  level concentrations



than elevated acrylonitrile emission points.  On the basis of the elevation and



other  relevant  factors,  monomer  and  ABS/SAN resin production results in  the



highest estimated total risk in  terms of exposed population and exposure concen-



tration.
                                         7-17

-------
                  8.  BIOLOGICAL EFFECTS ON MICROORGANISMS



     Loveless et al. (195*0 studied  the effect of acrylonitrile on growth  and



cell  division  of yeast  (Saccharomyces cerevisiae)  and bacteria  (Escherichia



coli), as measured by dry weights and cell counts during the logarithmic growth



phase.  Treatment with 1000 mg/1 reduced the growth of E. coli but had no effect



on cell size;  this concentration inhibited growth and division in S. cerevisiae.



Treated cells were 170$ larger than control cells and weighed 52$ more.



     Acrylonitrile was  not toxic  to the bacterium  Nocardia rhodochrous at  a



concentration  of 10,000 mg/1  in  both  (DiGeronimo and  Antoine,  1976).   This



concentration  of acrylonitrile supported  growth of these  bacteria as  a  sole



source of nitrogen,  but not as a source of carbon.



     Acrylonitrile  has  been  shown  to  be  inhibitory  to  anaerobic  bacteria.



Ludzack et al.  (1961)  reported that acrylonitrile inhibited  gas  production by



anaerobic digester  cultures which were dosed repeatedly  with 10  to  10  mg/1



acrylonitrile.  Hovious et al. (1973, cited in Miller and Villaume, 1979) found



that 50 to 100 mg/1  acrylonitrile inhibited gas production by anaerobic methano-



genic bacteria in proportion to dose.



     Acrylonitrile was  used as a  fumigant  to control mold  growth on packaged



papads, an  Indian bread (Narasimhan ejt  al.,  1972).   Papads with  18  or  20$



moisture content were sealed in polyethylene bags, fumigated for ^8  hours with 32



or 6M  mg/1  acrylonitrile,  and  checked  for mold  growth  after one month.   The



higher dose prevented mold  growth at both moisture levels, whereas the lower dose



prevented mold growth only in the 18$  moisture papads.  The species of molds were



not identified.



     Some limited information  concerning the effects  of acrylonitrile on aquatic



microorganisms  was  provided by  Cherry  et  al.  (1956).    Nutrient-enriched  and



aerated river water  was dosed  with 10,  25,  or 50 mg/1 acrylonitrile.  Balanced
                                         8-1

-------
populations of bacteria,  diatoms,  algae,  protozoa, and rotifers developed at 10



and 25 mg/1, whereas fungal species predominated at 50 mg/1.
                                         8-2

-------
                      9.  BIOLOGICAL EFFECTS ON PLANTS



     There is  limited  information concerning the  effects  of acrylonitrile  on



plants.



     Garrison (1978) studied the  effects of  acrylonitrile  on  cultured  seagrass



(Ruppia maritima).   Acrylonitrile was added to the water column  to give concen-



trations ranging from 10 ng/1 to  10  g/1.   Concentrations greater than  100 mg/1



totally  inhibited  photosynthesis  and respiration,  as  measured  by dissolved



oxygen changes. Lower concentrations had no effect on these processes.  Although




all concentrations  reduced the growth rate of shoots,  the  growth  rate  of roots



was stimulated at  concentrations below 1 mg/1 acrylonitrile.



     The effect of  acrylonitrile on pea seedlings (Pisum sativum) was studied by



Burg and Burg  (1967), who  reported that 0.17 mM  (ca.  9 ppm)  acrylonitrile was



"toxic" (undefined  effect)  and  that  lower  levels  showed no effect  on  seedling



elongation.



     Fumigant mixtures of 1:1 acrylonitrile:carbon tetrachloride had no adverse



effects  on seed germination of  beans,  beets,  corn,  peas,  lettuce,   onions,



tomatoes,  wheat, and oats  when  seeds  were  fumigated  for 2^  to M8  hours  at



concentrations  ranging  from 1  to 25 pounds per  1000  cubic  feet (Glass  and



Crosier, 19^9, cited in  Miller and Villaume,  1978).



     When  acrylonitrile  was  added to aerated,  nutrient-enriched  river  water,



balanced growth of  bacteria,  diatoms, algae,  protozoa, and rotifers occurred at



10 to 25 mg/1 acrylonitrile (Cherry ejt al., 1956).  At 50 mg/1,  however,  fungal



growth predominated.



     Kihlman (1961) reported that 1 mM (53  PPm)  acrylonitrile was not mutagenic



to broad bean  root tips  (Vicia  faba).  The details of this study  are  given in



Section 13.4.
                                         9-1

-------
                  10.   BIOLOGICAL EFFECTS ON DOMESTIC ANIMALS



     No information was found concerning the effects  of acrylonitrile  on  domes-



tic animals  other  than dogs and  cats.   The effects  of acrylonitrile on these



animals are discussed  in Section 13.
                                         10-1

-------
                      11.   BIOLOGICAL EFFECTS ON  WILDLIFE




     No information was found on the toxicity of acrylonitrile to wildlife  other




than insects.




11.1  INSECTS




     Judson et, al.  (1962) studied the ovicidal effects of acrylonitrile and  other




chemicals on the eggs  of the yellow-fever mosquito (Aedes aegypti).   Mature eggs




were exposed for 24 hours to the vapor of 5 or 10 ul added acrylonitrile in sealed




one quart jars (21  to  32°C, 100$ humidity)  and then placed  in deoxygenated  water




to determine  hatchability.  The  percent mortality at the  two treatment levels




(4.2 or 8.4 mg/1, by calculation)  was 60 and 92$, respectively.




     Bond  (1963)   exposed  adult  granary  weevils  (Sitophilus  granarius)  and



cadelle larvae (Tenebroides mauritanicus)  to a series  of acrylonitrile fumigant




concentrations for an unspecified  period.  The exposed insects were then divided




into three groups, which were kept for  48  hours  in an atmosphere  of  nitrogen,




oxygen, or air and then placed  in air for 5 days.  The dosage (expressed as the




product of concentration and exposure time)  required  to kill 50$ of the insects




kept in air was 23-0  for T. mauritanicus and 5.4 for  S.  granarius. The median




lethal concentration cannot be calculated from these values because Bond did not




specify the  duration  of exposure to  the  fumigant.   The results did  indicate,




however, that oxygen enhanced the toxicity of acrylonitrile and most of  the  other




chemicals tested.




     Lindgren et^ al. (1954) fumigated eight species of insects with a  series of




acrylonitrile concentrations for  2 or 6  hours. Mortalities were counted 4 days




after fumigation.  The LD50 and LD95  values are given  in Table 11-1.




     Similar  toxicity  studies  were  conducted  with acrylonitrile  by  Bond  and




Buckland (1976, 1978)  with several  insect  species.  The  duration  of  exposure,
                                         11-1

-------
      Table  11-1.   Lethal Dose  Values  for  Insects  Exposed  to  Acrylonitrlle  Fumigation
     Species
Length of
Exposure
 (hours)
                                       Length of
                                         Test
                                        (days)
 LD50
(mg/1)
 LD95
(n.g/1)
 LD99
(mg/1)
Reference
Granary Weevil
(Sltophllua granarlua)
Rice Weevil
(Sitophllua oryra)
Mexican Bean Weevil
(Zabrotea pectoralla)
   8
   6
   5
   2

   6
   2
  24

   6
   2
Bean Weevil                 6
(Acanthosceelldea obtectus)  2
Drug-store Beetle
(Stegoblum panlceum)

Confused Flour Beetle
(Trlbollum confusum)
Pulse Beetle
(Calloaobruchus chlnenala)

Red Flour Beetle
(Trlbollun caataneura)
                            5
                          24
Saw-toothed Grain Beetle     6
(Oryzacphllus surlnamenala)  2
                          24
Lesser Grain Borer
(Rhyzopertha domlnlca)
Cadelle
(Tenebroldes maurltanlcus)
   6
   2
  24

   8
   5
                                                      0.7
                                                      2.0
                                                      1.4
                                                      4.5

                                                      1.0
                                                      2.5
                                                      0.40

                                                      1.4
                                                      2.0

                                                      1.1
                                                      3.0

                                                      1.7
                                                      3.0

                                                      1.9
                                                      3.0
                                                      6.5

                                                      0.43
                              2.6
                              1.05

                              1.1
                              3.5
                              0.76
                                                      0.8
                                                      2.5
                                                      0.76

                                                      2.8
                                                      3.8
                2.9

                8.0

                1.8
                6.5
                0.78

                2.1
                4.0

                2.0
                5.5

                2.5
                7.0
                                            4.9
                                           11.0

                                            0.56
                1.32

                2.0
                6.5
                1.27

                1.4
                4.0
                1.08
                1.2        Bond  6  Buckland,  1976
                          Llndgren  et^ aj^.,  1954
                2.2        Bond  &  Buckland,  1978
                          Llndgren  e^ aJL.,  1954

                          Llndgren  et_ al.,  1954
                          Llndgren  et^ al. ,  1954
                          Rajendran & Mutlm,  1976

                          Llndgren  et_ a_l.,  1954
                          Llndgren  et^ al..,  1954

                          Llndgren  ejc al.,  1954
                          Llndgren  et^ nl.,  1954

                          Llndgren  et^ a^.,  1954
                          Llndgren  et_ a_l.,  1954

                2.5        Bond  &  Buckland,  1976
                          Llndgren  et^ al.,  1954
                          Llndgren  e_t a_l.,  1954

                          Rajendran & Huthu.  1976
                4.3       Bond & Buckland, 1978
                         Rajendran & Hulhu, 1976

                         Llndgren et^ tU., 1954
                         Llndgren et^ al., 1954
                         Rajendran & Huthu, 1976

                         Llndgren et^ al., 1954
                         Llndgren et_ al_., 1954
                         Raj end ran & Huthu, 1976

                6.0       Bond & Buckland, 1976
                8.4       nond & Buckland. 1976

-------
time at which mortalities were counted, and the LD50 and LD99 values are given in



Table 11-1.  Bond and  Buckland  (1976)  found that acrylonitrile alone was  more



toxic than methyl bromide  alone  or mixtures of both compounds.   Although acrylo-



nitrile  is  too flammable  for  use as  a fumigant  alone,  it  does enhance  the



toxicity of non-flammable  methyl bromide, especially at low temperatures.   Bond



and  Buckland (1978)  showed  that  fumigation  with  acrylonitrile  and  methyl



bromide:acrylonitrile mixtures was more  effective  in atmospheres of 20  to 50$



carbon dioxide than in air.



     Rajendran and  Muthu  (1976)  also  conducted  fumigation bioassays with six



species of stored product  insects.  As shown in Table 11-1,  the longer  exposure



period (24 hours) resulted  in LD50  and LD95 values  that  were lower  than those



reported for the same species by the previously cited workers.



     The results  presented in  Table  11-1 indicate  that  the  concentration of



acrylonitrile required to  kill 95$ or more of test  groups  of insects  is between



about 0.5 and 10 mg/1, depending on species and exposure time.  This  concentra-



tion is between 1.3 and 2.6 times the LD50 concentration.



     The only  other information  found  concerning  effects of acrylonitrile in



insects was  by Benes  and  Sram (1969) ,  who found  that acrylonitrile  was not



mutagenic in fruit flies (Drosophila melanogaster).
                                         11-3

-------
               .  12.  BIOLOGICAL EFFECTS ON AQUATIC ORGANISMS




12.1  ACUTE TOXICITY




     The acute toxicity  of acrylonitrile has been determined for several species




of marine and freshwater fish and invertebrates.




12.1.1  Freshwater Fish




     The majority of information concerning the acute toxicity of acrylonitrile




to aquatic organisms has been developed with freshwater fish.   The most compre-




hensive study  is  that of Henderson jjt al. (1961), who reported  median lethal




concentration (LC50) values  for fathead minnows (Pimephales promelas),  bluegill




sunfish  (Lepomis   macrochirus)  and  guppies  (Poecilia reticulata  =  Lebistes




reticulatus).  Each bioassay utilized  5 acrylonitrile  concentrations in a geo-




metric series and 10 fish per concentration.  The test solutions were not renewed




during the 96-hour  exposure  period.   LC50 values were  calculated  by graphical



interpolation from  mortality data at 21!, U8, and 96 hours of exposure.  These and




other values are presented  in Table 12-1.




     The 96-hour LC50  values ranged between  11.8  mg/1 for the  most sensitive




species (bluegills)  and 33.5 mg/1 for the least sensitive species (guppies).  The




LC50 value decreased by a factor of about 2  between 2H and 96 hours in most tests,




indicating that toxicity increased  with  exposure time.  In contrast,  toxicity




tests  with  other organic nitriles (lactonitrile,  benzonitrile,  acetonitrile,




adiponitrile, oxydipropionitrile) showed relatively  little  or no increase in the




toxicity of these compounds with longer exposure time.  Acrylonitrile LC50 values




for fathead minnows  were slightly lower in hard water (320 mg/1 hardness) than in




soft water (20 mg/1 hardness), which  indicates  that acrylonitrile toxicity may




increase with higher water hardness. Because confidence intervals for these LC50




values cannot  be calculated from the   reported data, it is unknown whether the
                                          12-1

-------
    Table  12-1.   Median  Lethal  Concentration (LC50)  Values
                    for  Fish Exposed  to  Acrylonitrile
                                                LC50  (mg/1)
Species
Temp.  Type  Dilution  24 hr  48 hr  72 hr  96 hr   Reference
(1*C)  Teat   Water                       	
Fathead Minnows*  25     S   FV(hard)   32.7   16.7
(Piaephales
promelas)

Fathead Minnows*  25     S   FW(soft)8  34.3   21.5    20.5
(Pimephales
promelas)

Fathead Minnows*  25     F   FW(soft)g  33.5   14.8    11.1
(Pinephale3
promelas)

Bluegill Sunfishb  25     S   Fw(soft)8  25.5   14.3
(LepomlB
macrochirus)
Cuppies
(Poecilia
reticulata)
 25     S   FW(soft)8  44.6   33.5
Goldfish
(Carasslua
auratus)
Zebrafish
(Brachydanlo
rerio)
Minnow
(Phoxinus
phoxinus)
                  NR     S   FW
 20     F   FW
                  NR   NX   FW
                  KR   NR   FW
                                                    40
Carp
(Cyprinus
carp io)

Rainbow Trout     NR    NR   FW(hard)
(Salmo
gairdnerl)

         d
Golden Ide
(Leueiscus idus
melanotus)
Pinfish
 Lagodon
rhomboides)
                  20
13.7-   S   SW
20.4
                             15
                       38.2   17.6
                       37.4   24.0
                             70
                            13.28
                                       24.5
                                           14.3   Henderaon «t al..
                                                  1961
                                           18.1   Henderson et al..
                                                  1961
                                           10.1   Henderson et al.,
                                                  1961
                                           11.8   Henderson et al.,
                                                  1961
33.5   Henderson et al.,
       1961
       Paulet  and Vidal,
       1975

       Slooff,  1978
       Marconi and
       lonescu, 1974


       Marconi and
       lonescu, 1974


       Jackson and
       Brown.  1970

       Juhnke  and
       Luedemann, 1978

       Daugherty and
       Garrett. 1951
     "length 50.8-63.5 am  (2-2.5 inches); weight "1.5 g
     blength 38.1-50.8 mm  (1.5-2 inches); weight »2 g

     Clength 25.4 mm (1  inch); weight «0.1 g

      length 50-70 mm; weight 1.5 ± 0.3 g

     estandard length 57-113 mm
     fhardwater - pU 8.2,  alkalinity 320 mg/1; acidity 0 mg/1;  hardness 380 mg/1

     8soft water - pH 7.4, alkalinity 16 mg/1; acidity 2 mg/1;  hardness 20 mg/1
     hpH 7-8; degree of  water hardness 15 t 3' (German "Hartegrad")

     S • static exposure
     F • flow through exposure
     FW- freshwater
     SW- seawater
     NR- not reported
                                    12-2

-------
difference is  statistically  significant.   It can  be concluded, however,  that



water hardness has little effect on acrylonitrile toxicity.



     Henderson et al. (1961) also  conducted  continuous-flow  acrylonitrile  bio-



assays with fathead minnows in soft  water.  The exposure conditions are described



in Section 12.2 and  the 21,  18,  72,  and  96  hour LC50 values  are given  in



Table 12-1.   Comparison  of  static and  continuous-flow  LC50  values  shows  that



toxicity is equal at 2U hours,  but  is  greater under continuous-flow conditions



after M8 hours. Lower toxicity under static conditions may indicate that acrylo-



nitrile was  lost  from water  through adsorption, volatility, chemical  change,



fish uptake,  or biodegradation.



     Renn (1955) exposed  bluegill sunfish under static and continuous-flow expo-



sure conditions to 0.38-3.79 mg/1 acrylonitrile  and found no  mortality during a



214-hour exposure period.  White crappies (Pomoxis annularis) exposed to H acrylo-



nitrile concentrations under continuous-flow conditions began dying after about



2 hours in 90.9 mg/1 and after about 8 hours in 68.2 mg/1.  No mortality occurred



during 2U-hour exposure to 37-9 or 22.7  mg/1.  The concentrations given here were



calculated from Renn's concentration data,  which were reported in mg/1 nitrogen




as acrylonitrile.



     The  US-hour  LC50 of  acrylonitrile to  zebrafish (Brachydanio  rerio)  was



determined to be 15 mg/1  by  Slooff (1978) .   This bioassay was conducted with 10



fish per  concentration in  closed 10-liter aquaria  under  flow-through (6 1/hr)



conditions.



     Paulet and Vidal (1975) determined a  72-hour  LC50  of MO mg/1 for goldfish



(Carassius auratus).   This  bioassay was conducted under static  conditions in



12-liter aquaria.   No other information was provided.



     Bandt (1953)  provided  some  limited   information  concerning  the  static



toxicity  of   acrylonitrile  to  two  freshwater   fish  species,  bleak  (Ablurnus
                                         12-3

-------
alburnus)  and roach (Rutilus rutilus).  Bleak and roach were exposed in groups of

one or two fish  to 20  to 100 mg/1 acrylonitrile  for  up to 20 days.   Although

concentrations of  25 mg/1 and greater were  eventually lethal, the  only  bleak

exposed to 20 mg/1 showed no effect after 20 days of  exposure.  Renn concluded

from  this  limited information that  the  threshold concentration  for prolonged

exposure would be about 20 to 25  mg/1.  As seen in Section 12.2,  this  value is

much too high.

     Juhnke and Luedemann (1978)  reported acute toxicity levels of acrylonitrile

to golden ide (Leuciscus idus melanotus)  that had been  determined by using the

identical protocol  in  two  different  laboratories.  The 48-hour LCD,  LC50,  and

LC100 (0, 50,  100$ mortality)  values determined  at  each laboratory are  given

below:

                               48 Hour Lethal Concentration Values (mg/1)
                                   LCO	LC50	LC100
          Laboratory 1              16              28              48
          Laboratory 2               8              13              20

Although these tests presumably were  conducted under  identical conditions (see

Table 12-1),  the values reported  by laboratory 1 are about twice as high as those

reported by  laboratory 2.  The  authors did  not discuss  the reasons for  the

different results.

12.1.2  Marine Fish

     The only report of acrylonitrile  toxicity to marine fish is by Daugherty and

Garrett (1951).  Groups of eight pinfish (Lagodon rhomboides) were acclimated for

22 to 24 hours in 30  liters of aerated seawater at 13.7 to 20.4°C.  Acrylonitrile

was then added to give 16 concentrations ranging  from 0.25 to 60 mg/1.  No deaths

occurred at 20 mg/1 or  less, or in controls.  All fish died at  30 mg/1, which was

the  next  higher  concentration  tested.    The 24-hour LC50,   as  determined  by
                                         12-4

-------
graphical interpolation,  was 21.5 mg/1.  This value is about the same as the 24-



hour LC50 values reported for freshwater fish (Table  12-1).



12.1.3  Freshwater Invertebrates



     Bandt (1953)  provided some limited  information concerning  the  static  acute



toxicity of  acrylonitrile to  scuds  (Gammarua  sp.,   a  freshwater  crustacean).



Groups of ten scuds were exposed to 25, 50, or 100 mg/1 acrylonitrile for up to  72



hours.  All animals exposed  to 50 and 100 mg/1 were dead  by 48 hours.   There was



no mortality in the group exposed to 25  mg/1 for 3 days.



     Acrylonitrile may be more  toxic to Daphnia magna, another species of fresh-



water Crustacea (LeBlanc, 1980).  The  24 and 48-hour LC50 values (and 95? confi-



dence interval values) for this species tested under static conditions  were  13



(11-15)  and 7.6 (6.2-9.2) mg/1, respectively. The "no discernible effect concen-



tration" was 0.78 mg/1.



     Randall and Knopp (1980)  determined the static 48-hour EC50 (median effec-



tive concentration) of acrylonitrile  for D. magna.   This  value  (and  its 95$



confidence interval)  was  10.95  (9.54-12.56) mg/1, which is in  reasonably good



agreement with the 48-hour LC50 value reported by LeBlanc (1980). Both of these



studies utilized  young (< 24  hours  old)  D. magna.   and  water  of  similar mean



hardness (155-173 mg/1 as CaCOj  and temperature (22°C).



12.1.4  Marine Invertebrates



     Portmann  and  Wilson (1971,  cited  in Miller and  Villaume, 1978)  exposed



groups of 8 to 25 brown shrimp (Crangon  crangon)  to serial dilutions of acrylo-



nitrile  in aquaria containing  10 gallons  (37.85 1)  seawater  at  15°C.   The



reported LC50 was 10-33 mg/1.
                                         12-5

-------
12.2  SUBCHRONIC TOXICITY
     There is information concerning the subchronic toxicity of acrylonitrile to
several species of freshwater fish  and one  species  of freshwater invertebrate.
No information was found for marine fish or invertebrates.
12.2.1  Freshwater Fish
     Henderson £t al. (1961)  conducted five replicate subchronic bioassays with
fathead minnows under continuous-flow exposure conditions.  Groups  of  50 fish
were  exposed to  each  of  7  acrylonitrile  concentrations  or to  control  water
(20 mg/1 hardness, 25°C)  for up to 30 days.   Controlled amounts of test solutions
were pumped into glass bottles containing  10 fish and  10 liters of water so that
a renewal time of 100 minutes was obtained.  The fish were fed daily.  The mean
LC50 values determined for the five replicate bioassays are given below:

Exposure Time (Days)     1     2     3     1     5    10    15    20    25   30
LC50 (mg/1)             33.5  1H.8  11.1  10.1  8.1   6.9   5.2   1.2   3.5  2.6

The LC50 decreased linearly with exposure time between  15 and 30 days, indicating
that the  lethal  threshold concentration  had not been  determined at the 30 day
mark and that mortalities  would have  continued to occur at concentrations well
below 2.6 mg/1 after 30 days.
     Cumulative subchronic toxicity was also  found  by Jackson and Brown (1970,
cited in Miller and Villaume, 1978), who exposed rainbow trout to 2 to 200 mg/1
acrylonitrile for  time  periods as  long as  100 days.   Although  the LC50  after
48-hour exposure was 70 mg/1, exposure to  2.2 mg/1 for 100 days resulted in 50$
mortality.
     Very little is known about the mechanisms of acrylonitrile toxicity in fish.
Henderson et al. (1961)  noted that  the first  sign of acrylonitrile toxicity in
fathead minnows was  extreme  darkening of  the  skin, followed  in  1 to 3  days by
death.  They found no cyanide, formed  from acrylonitrile, in the exposure water.
                                         12-6

-------
In contrast, the more rapid,  non-cumulative toxicity of lactonitrile was attri-



buted to the formation of cyanide, which was measured in the exposure water.   It



is not known whether metabolic  formation of cyanide from acrylonitrile occurs in



fish, although  cyanide  or thiocyanate formation  has been reported  in  mammals



(Section 13).  Toxic action in  mammals is attributed primarily to direct  effects



of acrylonitrile and secondarily  to cyanide toxicity.  Slooff (1978)  studied  the



effect of acrylonitrile and other chemicals on  respiratory activity of  rainbow



trout.  The frequency of respiratory movements was significantly increased after



24-hour exposure to  5 mg/1 acrylonitrile under flow-through conditions.  This



effect was preceded by a temporary, but significant, decrease in breathing rate.



The  concentration  at which this sublethal  effect occurred (5 mg/1)  was three



times lower  than the 48-hour  LC50  for zebrafish  (15  mg/1),  determined  under



similar flow-through exposure conditions.



12.2.2  Freshwater Invertebrates



     The only  information concerning subchronic  toxicity of  acrylonitrile to



aquatic invertebrates was  developed  with Daphnia magna  (U.S.  EPA,   1978).   No



adverse effects  were  found when  this invertebrate was  exposed over its entire



life  cycle  to  as  much  as 3.6  mg/1  acrylonitrile.   As  mentioned  previously



(12.1.3),  the 48-hour EC50  reported for this species was 7.55 mg/1, which is only



about  twice as  high as   the  chronic  no-observed-effect  concentration.    In



contrast,  the 48-hour LC50 was about 32  times  higher than the 100-day LC50  for



rainbow trout (Jackson and Brown, 1970,  cited  in Miller and Villaume, 1978)  and



about six  times higher than  the  30-day LC50  for fathead minnows  (Henderson



et al.,  1961).   Although  the  chronic no-observed-effect concentration  has  not



been determined with any fish species,  it is probable that  it would be consider-



ably lower than 2 mg/1.
                                         12-7

-------
12.3  SUMMARY AND CONCLUSIONS




     Acrylonitrile has been shown to affect some terrestrial and aquatic plants




at exposure concentrations of 9 to 100 mg/1.   Acrylonitrile is toxic to aquatic




animals at exposure concentrations in the low milligrams per liter range.  The




reported acute LC50 values for fish ranged between 10.1 and 70 mg/1.  Subchronic




exposure of  fish  for  30 to 100 days resulted in LC50 values of about 2 mg/1,




with no evidence  that a threshold concentration had been  reached.  Although




the only tested  invertebrate,  Daphnia magna, had the lowest acute LC50 value




(7.6 mg/1), this species was not adversely affected by chronic exposure to 3.6




mg/1 throughout its whole life cycle.
                                        12-8

-------
           13.  BIOLOGICAL EFFECTS IN MAN AND EXPERIMENTAL  ANIMALS



13.1  PHARMACOKINETICS



13.1.1  Absorption and Distribution



     Young et al. (1977) studied  extensively  the pharmacokinetic and metabolic


                                                              14
fate of acrylonitrile  in  male  Sprague-Dawley rats  by giving   C-acrylonitrile



with different  routes  and dose levels.   When acrylonitrile was orally  admin-



istered to rats, essentially all (95$)  of  the dose was absorbed. After 72  hours,



the percentage (^ the  standard  deviation) of  recovered  radioactivity accounted



for 82 + 9.61$ and  101  +  14.4$  of the administered dose at  0.1  and 10  mg/kg,



respectively (Table 13-D•  The present  recovery of  administered radioactivity



in the urine was much higher in the animals  administered the  higher dose,  while



both doses resulted in a 5% recovery of radioactivity in the  feces.



     The absorption of acrylonitrile through  inhalation was also determined  by



Young et  al.  (1977).   Animals  were  exposed  to acrylonitrile vapor in a  "nose



only" chamber at concentrations of 5 ppm and  100 ppm  acrylonitrile.  Following



exposure at the higher dose, a significantly higher recovery was found in  urine,



but a smaller percentage was recovered in the expired air as CO- and in  the body



(Table 13-2).



     Rogaczowska (1975)  studied  dermal absorption of acrylonitrile vapor in rab-



bits and found that  the  penetration rate of acrylonitrile vapor through  the skin



is about 1$ in relation to the  quantity absorbed through the  lungs.



     The plasma concentration of acrylonitrile (ug Eq/ml  plasma) as a function of



time, routes of administration (per os and intravenous),  and dose levels (0.1, 1,



and 30 mg/kg) has been studied  (Young e± al.,  1977).   Following oral and  intra-


                         14
venous administration of   C-acrylonitrile,  typical plasma  concentration  versus



time curves were observed.   The biphasic disappearance of radioactivity indi-



cated a pharmacokinetic two-compartment open  model for  elimination.  The half-
                                        13-1

-------
         Table 13-1.   Recovery of Radioactivity from Rats
                      Given Single Oral Doses of 0.1 or
                      10 mg/kg IuC-Acrylonitrilea-(Young etal., 1977)


Radioactivity in Urine
Radioactivity in Feces
Expired Air
Organics in carbon
Organics in solvent
H1ACNe
14c02
Body
Carcass
Skin
Cage Wash8
h
Total Recovery
Percentage of
0.1 mg/kg
Dose, Mean + S.D.
10 mg/kg

34.22 ± 6.26b
5.36 + 1.43

0.09 + 0.09
0.19 + 0.19
0.07 + 0.05
4.56 + 1.82
37.02 + 6.09b
24.24 + 5.02b
12.78 + 1.17
0.86 + 0.37
V,
82.37 + 9.64°
66.68 + 10.60b
5.22 + 1.17

0.11 + 0.06
0.20 ± 0.16
0.08 ± 0.03
3.93 + 1.79
26.61 + 5.91b
16.04 + 1.87b
10.57 + 4.55
1.22 + 0.41
K
104.04 + 14.40°
 Two groups of rats housed individually in glass metabolism cages were
 given 0.1 mg/kg (4 rats) or 10 mg/kg (5 rats); excreta were collected
 at 8 hour intervals for 72 hours.

 The mean values for the two dose levels are different at the P=0.05
 level of significance.
Q
 Pittsburg activated coconut charcoal, 12 x 30 mesh.

 trap « 2-methoxyethanol
"trap =• 0.02 M Ag^O^, in 0.1 N ^SO^

'trap - 5 M ethanolamlne in 2-methoxyethanol
8,
 'water-acetone
     mean and S.D. of the recovery were calculated from the total
 recovery of the dose In individual rats.
                                       13-2

-------
Table 13-2.  Recovery of Radioactivity from Rats Exposed by Inhalation to
             5 or 100 ppm ^C-Acrylonitrile for 6 hours3 (Young e£ al., 1977)


Urine
Feces
14co2
Body
Cage Wash
Total dose
in pg Equivalents
Percentage of
5 ppm
Recovered Dose
100 ppm

68.50 + 9.38C
3.94 + 0.97
6.07 + 1.58°
18.53 ± 4.68C
2.95 + 3.95
172.92 + 28.35
82.17 + 4.21C
3.15 + 0.82
2.60 + 0.83°
11.24 + 2.85C
0.85 + 0.58
2556.65 + 672.10
    Rats were exposed in a "nose only" chamber under dynamic air flow
    conditions for 6 hours.  After exposure, rats were housed in glass
    metabolism cages and excreta were collected for 220 hours.

    Values are the mean ;+ S.D. for 4 rats per exposure level.

   CThe mean values for the two exposure levels are different at the
    P-0.05 level of significance.
                                       13-3

-------
life values of alpha and beta  phases, calculated by linear regression  analysis,


ranged from 3-5 to 5.8 hours and 50 to 77 hours,  respectively.


     Freshour et al. (I960) investigated the phannacokinetic profile  of  intact


acrylonitrile  after  intravenous or  per  os administration of acrylonitrile  to


male Fischer F3W rats.  The  plasma  concentration  of  acrylonitrile  versus  time


obtained  after  intravenous  administration  was   characteristic   of a  one-


compartment model with first-order elimination, but a Diphasic  elimination was


observed following a 30 mg/kg  oral  administration.  The half-life of first-order


elimination ranged  from 7.8 to 13.9 minutes  after 30 mg/kg intravenous and  per os


doses, and the half-life for  terminal  phase after  30  mg/kg oral administration


was 85 to 120 minutes.


     Hashimoto and  Kanai (1965)  have  studied the blood concentrations of acrylo-


nitrile and cyanide as a function of  time in relation to toxicity.  When acrylo-


nitrile was given to rabbits intravenously  at a sublethal dose (30 mg/kg,  LD50 =


50 mg/kg),  acrylonitrile was  biphasically  eliminated with  1  ug/ml of acrylo-


nitrile remaining  four hours  after dosing.  The cyanide  concentration rose to


about 1.5  ng/ml at  1.5 hours after dosing, and then gradually returned to zero at


four hours.  After injection  of a  lethal dose  of acrylonitrile (75 mg/kg), the


hydrogen cyanide concentration rose steadily until  the death of the animal.


     Tissue distribution of radioactivity  (acrylonitrile and  its  metabolites)

                                                                           in
was  determined  in  rats   given  a  single  oral  or intravenous  dose  of    C-


acrylonitrile at doses of  0.1 and 10 mg/kg  (Young et al.,  1977).  Acrylonitrile


and  its metabolites  were  distributed  to  all  tissues examined  (lung, kidney,


liver, stomach, skin, blood,  etc.);  notably, high  levels  of radioactivity  were


observed in stomach, skin, and red blood cells regardless of route and dose level


(Table  13-3).    The  high  accumulation of radioactivity in  the stomach  wall


following  intravenous acrylonitrile administration  indicated   that  the  high
                                        13-4

-------
        Table 13-3.   Distribution of Radioactivity in Selected
                     Tissues of Rats Given 1<4C-Acrylonitrile
                     (Young e£ al.,  1977)






Lungs
Liver
Kidneys
Stomach Wall and Contents
Intestines
Duodenum
Je j unum
Ileum
Cecum
Colon
Skeletal Muscle
Heart
Spleen
Brain
Thymus
Testes
Skin
Carcass
Packed Blood Cells
Plasma Concentration
(ug Eq/ml)
Tissue

i.
2 hours
1 mg/kg
14
to Plasma Ratios of C-Activity
a
V.
24 hours
1 mg/kg

0.96
0.68
1.33
8.11
0.72
0.85
0.92
1.01
0.37
0.51
-
0.64
0.98
0.65
0.74
0.67
1.91
0.54
3.33

0.727
1.07
1.20
3.15
7.36
0.31
0.68
0.57
0.44
0.07
0.07
0.50
0.78
-
0.76
-
-
2.85
0.73
7.16

0.220
b
p.o.
72 hours
0.1 mg/kg 10




mg/kg

0.87
0.57
0.83
14.28
1.06
c
-
-
-
-
0.35
-
-
-
-
-
2.10
0.71,
2.26d

0.022
0.96
0.84
1.08
11.26
0.99

-
-
-
-
0.41
-
-
-
-
-
2.70
0.55
5.73e

1.437
 One rat per time.   (i.v.  = intravenous.)
 Average of 2 rats  per dose.   (p.o.  = per  os.)
"A dash indicates no analysis was performed.
 For all 4 rats in  this group, the ratio was 2.18 + 0.20 (Mean + S.D.)
EFor all 5 rats in  this group, the ratio was 5.16 + 1.19 (Mean + S.D.)
                               13-5

-------
concentration in the stomach wall following acrylonitrile oral dosing was not due


to poor absorption.  Analysis of  the  stomach  following  intravenous  administra-


tion showed that the radioactivity had increased  from 30.33 ug Eq at 5 minutes to


68.64 pig  Eq at  21 hours.   The particular  retention  of acrylonitrile  and  its


metabolites in the stomach  seems in part due to enterogastric circulation (Young


et al.,  1977) .   The accumulation of  radioactivity in blood was mainly due to


covalent binding of acrylonitrile to macromolecules and lipids in the red blood


cells (Ahmed and Patel, 1979;  Ahmed  e£ al.,  1982).  In other tissues, the amount


of radioactivity declined rapidly with time because of excretion (Young et al.,


1977; Ahmed and Patel,  1979;  Ahmed e_t al.,  1982).


     Similar distribution patterns were obtained by Sandberg and Slanina (I960)


using the  technique  of whole-body autoradiography following  administration of

                                                                           TU
labeled acrylonitrile to rats and monkeys.  Male and  female rats received   C-


acrylonitrile by intravenous injection  (13mg/kg) and were  killed  1  minute,


20 minutes, 1 hour, H hours, and 7 days after treatment. Two monkeys were given


1 or 6 mg/kg acrylonitrile orally and killed 1 and  6 hours later, respectively.


There was no difference in the distribution  patterns of acrylonitrile between the


two  routes  of administration in  rats, except  for the slower  absorption time


following oral dosing.   High  levels of activity were present in the  blood and


excretory  pathways  (bile,  intestinal  contents,   and  urine)   with  the liver,


kidney,   lung,  and adrenal  cortex also  accumulating appreciable  label.   The


stomach and hair follicles showed constant uptake of label throughout this study.


Radioactivity was still present in animals  killed  seven days  after  administra-


tion of a_.'ylonitrile.   In fetuses  exposed  in utero,  only the eye lens accumu-


lated label at a higher concentration  than that observed in maternal  blood.  The


labeling  pattern in monkeys  was similar  to that in rats  except  for  a more


pronounced activity in  the liver.
                                         13-6

-------
13.1.2  Metabolism



     The  metabolism  of  acrylonitrile has  been studied  by many  researchers.



Dudley and Neal (1912)  and Brieger e£ al. (1952) suggested that  acrylonitrile is



metabolized to  cyanide,  which  is transformed to thiocyanate and eliminated in



urine; however, less than one third of administered dose has been accounted for



by this metabolic route (Gut et al., 1975) .   Acrylonitrile reacts with sulfhydryl



groups through cyanoethylation, which  prevents  further metabolism to cyanide and



thiocyanate (Hashimoto and  Kanai,  1965).   Other metabolic pathways  have  been



suggested including formation of  mercapturic acid (U.S. EPA, 1979),  oxidation to



CO  via cyanate ion (Boxer and Richards, 1952) ,  and conjugation with  D-glucuronic



acid (Hoffman e£ al.., 1976) .  Evidence  from Gut  et al. (1975) , Wright (1977) , and



Young ^t al.  (1977) suggested that the metabolism of acrylonitrile in mammals was



related to dose level,  route of administration, and species.



13.1.2.1  Metabolism to Cyanide and Thiocyanate



     Acrylonitrile is metabolized in  laboratory animals to cyanide (CN~),  and



then converted  to  thiocyanate  (SCN~).  Evidence of  this  pathway has  been pre-



sented by  Dahm (1977).  The  excreted urine  collected  from orally-dosed    C-



acrylonitrile rats was analyzed  by high pressure liquid chromatography (HPLC).



The suspected thiocyanate peak matched  the retention time of a thiocyanate stan-



dard.  Another experiment by Dahm (1977) showed direct evidence that thiocyanate



comes  from  the cyano  group  carbon of  acrylonitrile.   When  rats  were  orally


             m
administered   C-acrylonitrile,  labeled on either the olefin or cyano group, the



thiocyanate peak by HPLC was  observed only  when  the cyano group  was labeled.



Ahmed and Alreu (1982)  have shown that  acrylonitrile is converted  to cynide in



the brain and liver of rats by enzymes associated with the microsomal fraction.



     Urinary excretion of thiocyanate following acrylonitrile administration at



different dose levels and by various routes  accounted for from 2 to 33$ of dose



(Gut et a^.,  1975;  Hashimato and  Kanai, 1965).  Gut et al. (1975) emphasized that
                                         13-7

-------
the extent of conversion  of acrylonitrile to cyanide  and thiocyanate is  both



route- and species-dependent.



     Using female Wistar  rats,  albino mice, and  Chinese hamsters, Gut et  al.



(1975) studied the extent  of the conversion of acrylonitrile to thiocyanate.   In



rats, there was higher thiocyanate excretion after oral administration (I1*.6 to



33.1$) than after intraperitoneal  (2.2 to 5.7$),  subcutaneous  (1.6$),  or intra-



venous (1.2$)  injection.    After  oral dosing, thiocyanate excretion  showed  a



distinct lag period  (l|  hours), which suggested that acrylonitrile was not appre-



ciably metabolized shortly after oral  dosing (Gut  e_t al^., 1975).  In mice,  there



was also a higher percentage of thiocyanate excreted after oral (35$)  than  after



intraperitoneal (8  to  10$)  and intravenous  (11$)  injections.  The  portion of



acrylonitrile converted to thiocyanate was approximately the  same  for mice and



for  rats.   The higher total  recovery of thiocyanate in mice  compared to rats



indicated that mice  may  have a higher metabolic  capacity  and a lower binding



capacity for acrylonitrile (Gut  et al., 1975).  Hamsters were similar to rats and



mice in that more thiocyanate was excreted in urine after oral  than  after intra-



peritoneal administration of acrylonitrile.   In contrast to rats, elimination of



thiocyanate in mice and hamsters after oral  dosing showed no  lag  period.   Gut



et a^. (1975) indicated that body  size may be an  influencing factor.



     Pretreatment of rats  with phenobarbital  (a microsomal inducer), SKF525-A (a



microsomal inhibitor),  cysteine, or dimercaprol did not significantly influence



the excretion of  thiocyanate  in the  urine after  the administration  of acrylo-




nitrile.  Simultaneous  intraperitoneal administration of acrylonitrile and  thio-



sulfate to rats and mice  increased thiocyanate elimination twofold in rats and



threefold in mice, but no effect of thiosulfate was observed after oral adminis-



tration of acrylonitrile (Gut et al.. 1975).  The lack of effect of microsomal



enzyme induction  or inhibition on acrylonitrile-thiocyanate  balance indicated
                                        13-8

-------
that the route-dependent differences in thiocyanate elimination observed within

one strain of animals are caused  by  factors affecting the tissue distribution of


acrylonitrile (Gut et al., 1975) .

13.1.2.2  Reaction with Sulfhydryl Groups


     Acrylonitrile  reacts  with   sulfhydryl  compounds  in  laboratory  animals


through cyanoethylation (Hasimoto and Kanai,  1965;  Gut et al.,  1975;  Dahm,  1977

and Langvardt et al.,  1979); the reaction products are  not further metabolized by


animals but rather are excreted unchanged (Gut et al., 1975).   It has also  been


demonstrated in hamsters following interperitoneal  injection, that acrylonitrile


treatment  resulted in  a rapid  decrease (within  1 hours)   in  the  levels  of

glutathione in  both  the liver and brain (Zitting  et  al., 1981).   Hasimoto and

Kanai (1965) showed that acrylonitrile rapidly forms stable  conjugates in vitro

with  materials  that  have active  hydrogen atoms,  such as  L-cysteine  and  L-

glutathione.  That acrylonitrile  forms  conjugates with cysteine in vivo was  also

demonstrated and the  chemical  structures of the metabolites were identified as S-


(2-cyanoethyl)cysteine  (Gut  et   al.,  1975;   Dahm,  1977)   and  N-acetyl-S-( 2-


cyanoethyl) cysteine(cyanoethylated  mercapturic acid)  (Dahm,  1977;  Langvardt


et al.,  1979).

     It has been suggested that  acrylonitrile  reacts  in vitro  with glutathione

(GSH)  via  GSH transferase, but  the conjugation was  determined  indirectly  by

measuring  the  disappearance of  GSH  substrates (Boyland and Classeaud,  1967).

Recently, the presence of cyanoethylated mercapturic acid in rat urine was  con-
                         •

firmed (U.S. EPA,  1979).  A proposed scheme for the  various metabolic pathways of


acrylonitrile biotransformation is presented  in Figure 13-1.

     Langvardt et al. (1980) detected 7 urinary metabolites of acrylonitrile, by


high pressure liquid  chromatography,  in rats following administration of acrylo-

nitrile (30 mg/kg)  by oral intubation.  Of the 7 metabolites, the  3 major metabo-

lites (metabolites 1, 6, and  7)  were further  investigated to  identify chemical
                                          13-9

-------
  I OIIDATIVE PATHWAYS
                            KM CKZWIC CMJUEAUON
                       i.  KKUIC KIM
                       z.  pwniw
                       1.  IIIXOCICAi MUMTMKSxmOU:  l) MRCWU.1XC *«D ITS
                          «MLOCS; 6) SEROTONIH; c) T-MIKMinT'IC ACID;
                          D) NISTMIKE
                       4.  
-------
structures.  Metabolite 7 was confirmed to be N-acetyl-S-( 2-cyanoethyl)  cystein,



and metabolite 1 was thiocyanate,  as  indicated by i.n vivo labeling patterns when


                       m                     iu
animals were given 2,3-  C acrylonitrile or  1-  C acrylonitrile,  and  retention



times and fragmentation patterns on GC-MS.  Metabolite 6 was tentatively identi-



fied  as  N-acetyl-3-carboxy-5-cyanotetrahydro-l,H-2H-thiazine since  it had  an



apparent molecular weight 2  units less than metabolite 7  and formed  artifacts




similar to metabolite 7 on being reduced to dryness  under N~ gas.   No  authentic



standard to  metabolite 6 could be synthesized.  By  using analogues  to  known



metabolic pathways for other halogenated vinyl compounds,  Langvardt £t al. pro-



posed a possible metabolic pathway of  acrylonitrile which  proceeds through the



formation of an epoxide intermediate (Figure 13-2) .  The authors suggest that the



epoxide metabolite may be an important factor in the toxicity of acrylonitrile.



     Abreu and Ahmed (I960) also demonstrated that an in vitro system,  employing



a rat liver microsomal fraction and  NADPH, was  capable  of forming cyanide from



acrylonitrile.  Using microsomal fractions from rats induced with phenobarbital



or aroclor 1251* increased the production of cyanide,  while addition of SKF525-A



to the reaction mixture inhibited the formation of cyanide,  These results were



consistent with a metabolic  pathway employing an epoxide as an intermediate.



     In a similar study using [1-  C] - or [2,3-  C] -acrylonitrile incubated with



rat liver microsomes, Guengerich et^ al,. (1981)  isolated labeled compounds which



co-chromatographed on high performance liquid chromatography (HPLC) with authen-



tic 2-cyanoethylene  oxide.   The  formation  of the epoxide was  enhanced  by the



inhibition of epoxide hydrolase.  Addition of epoxide hydrolase to a solution of



0.5 mM 2-cyanoethylene oxide  resulted in the destruction of the epoxide  at a rate



of  1.7 mmol/min,   and   the   formation  of   HCN.     Both   acrylonitrile  and



2-cyanoethylene oxide  when  mixed with  GSH (reduced glutathione)  reacted non-



enzymatically, and  HCN  was  released  in the  reaction with  the epoxide.   The



epoxide reacted less rapidly than did acrylonitrile in  this system, however, when
                                         13-11

-------
                             H/
  [GS-CH2-CHj-CN]
  +Ac
         -Glu,  Gly
   0       0
   II       II
HO-C-CH-NH-C-CH,
     I          3
      ^*"
-------
a cytosol preparation from rat liver was used,  the rate of GSH conjugate forma-



tion was  309 and  31 mmol/mg/min,  respectively,  for 2-cyanoethylene oxide  and



aer»ylonitrile.   The cytosol  from  rat brain and  human liver was inactive  in



forming GSH conjugates with acrylonitrile.  These preparations were active with



2-cyanoethylene oxide,  although the activity was diminished by more than an order



of magnitude when  compared to that  in the rat  liver preparation.   A metabolite



with a mass spectrum similar to that  which Langvardt e_t al. (1980)  ascribed to



l|-acetyl-3-carboxy-5-cyanotetrahydro-1,i»-2H-thiazine   was   obtained  by   non-



enzymatic  reaction  of 2-cyanoethylene   oxide  with  N-acetylcysteine.    This



evidence supports  a metabolic pathway in  which  acrylonitrile  is metabolized by



microsomal enzymes through an epoxide intermediate to HCN.  It was demonstrated



that 2-cyanoethylene oxide non-enzymatically bound to both protein and DNA,  and



that acrylonitrile did not, indicating that the 2-cyanoethylene oxide may be an



important metabolite in the toxicity and carcinogenicity of acrylonitrile.



     Wright (1977)  studied the dose-related and  species  strain-related  meta-



bolism of acrylonitrile.  Spartan rats, a Charles River rat,  and rhesus monkeys



were orally dosed  with  cyano-labeled  acrylonitrile  and the urinary metabolites



were determined (Table 13-1*) .  In rats, a lower  dose (0.1 mg/kg) of acrylonitrile



resulted in about  90$  of the administered  dose being metabolized  to  cysteine



conjugates.  At a higher dose (30 mg/kg), a great portion of administered acrylo-



nitrile was  excreted  as thiocyanate  and  unidentified  metabolite "C."   Wright



(1977)  suggested that conjugation with cysteine is the preferred metabolic path-



way  for  acrylonitrile  but  has limited  capacity.   No significant  difference



between the two strains of rats given a 30 mg/kg oral dose of acrylonitrile was



observed.  As  shown in Table 13-^, there is no significant  difference  in  the



proportion of metabolites excreted in urine by  rhesus monkeys following 0.1 and



30 mg/kg  of  acrylonitrile.   This  may  indicate  that  the  rhesus  monkey  has  a
                                         13-13

-------
                                 Table 13-4.   Urinary Metabolites Following the Oral Administration

                                              of C-l (Cyano) Labeled Acrylonltrlle (Wright, 1977)
Animal

Spartan Rat
Spartan Rat
Charles River Rat
Rhesus Monkey
Rhesus Monkey
Number

1
1
1
1
1
Dose
(rag/kg)

0.1
30.0
30.0
0.1
30.0
Percent of Radioactivity In
Urinary Metabolites
Cystelne-
Conjugates

88.6
37.3
47.5
91.
84.
"C"
(unidentified)

3.5
20.5
13.8
6.
14.
Thio-
cyanate

8.0
42.1
38.8
3.
2.
CO
I
                0-20 hours In rats and 0-24 hours In monkeys.

-------
greater metabolic capacity in conversion of acrylonitrile to cysteine-conjugates



than does the rat.



13.1.2.3  Minor Metabolites



     In addition to the cyanide and thiocyanate formed by acrylonitrile's meta-



bolism, acrylonitrile-glucuronide was  formed and excreted in rat urine.  Hoffman



et al.  (1976)  suggested that  acrylonitrile is initially hydrolyzed,  and  then



undergoes a condensation reaction with UDP-glucuronic acid.




     Young et^ al. (1977)  identified C02 as a metabolite of acrylonitrile.  Carbon



dioxide was expired in the breath and  accounted for 5 to 6% of the administered



dose of acrylonitrile.  Boxer and Richards (1952)  demonstrated that cyanide and



thiocyanate are in a dynamic equilibrium in the  rats and that, to a large extent,



cyanide and thiocyanate carbon  are oxidized directly to CO.. Thus, C0? may arise



as a product of cyanate metabolism.



     The majority of the administered  dose of  acrylonitrile was metabolized and



excreted;  however,  Hashimoto and Kanai (1965)   and Hoffmann et al.  (1976)  esti-



mated that 15$ of an administered acrylonitrile dose is eliminated unchanged in



the urine and breath.



13.1.2.^  Route and Dose Dependence of Acrylonitrile Metabolism



     The metabolic fate  of  acrylonitrile  and  its dependence on dose  level and



route of administration have been studied  by many researchers (Gut Q al., 1975;



Wright, 1977; Young et al.,  1977).   Gut et al,.  (1975) indicated that the forma-



tion of both cyanide and thiocyanate are dose-  and route-related.  Wright (1977)



found  that  the  reaction  of  acrylonitrile  with  sulfhydryl   groups   is  dose-



dependent.  Recently, Young et  a],.  (1977)  quantitatively isolated three metabo-



lites designated as "A," "C," and "E" as well as C02 in male Sprague-Dawley rats
                                          13-15

-------
                14
given doses  of   C-acrylonitrile by several routes.  Metabolites A, C, and E,



which accounted  for  more  than 95% of the total radioactivity, were excreted



primarily in  the urine,  while CCL  was excreted in the breath.  The chemical



structures of metabolites A,  C, and E were cited by Keresztesty et al. (1977)



as N-acetylated  conjugates of cysteine,  thiocyanate,  and thiocyanate,  respectively



(apparently  there  was an error in  the  paper; thus,  the  reader  should not



assume that  the  structure of  C and E are the  same), but there was no evidence



for  structure  identification in the original  report  (Young et al.,  1977).



     The proportion  of  the  metabolites  A,  C,  and  E in various rat  fluids was



route- and dose-dependent (see Table 13-5).   The 0-72 hour urine sample showed



the  highest  amount of metabolite C (73%) after an oral dose of 0.1 mg/kg, but



the maximum amount of metabolite A (61%)  was  observed after a higher  oral dose



of 10 mg/kg (Young et al., 1977).



     When rats were exposed to acrylonitrile  vapor at a concentration  of  5 ppm



for  6 hours,  61% of  metabolite  E  was excreted in the urine, while a dose of



100 ppm resulted in equal proportions  of the  three metabolites.



     The predominant metabolite excreted  in  bile  after a  1 mg/kg intravenous



dose was metabolite C (91%).  For  stomach,  red blood  cell, and plasma,  metabolite



E accounted for 93, 83,  and 42% of  total  radioactivity,  respectively.



13.1.2.3.   Summary and  Conclusions—Acrylonitrile  is  readily absorbed  in



animals following ingestion or inhalation,  while dermal absorption  is  poor and



occurs at about  1% of that of the  lungs.   Following absorption of radiolabeled



acrylonitrile, the  radioactivity  disappears   in  a biphasic manner, with a



half-life for the  first  phase of  3.5  to 3.8  hours and the second phase of 50



to 77 hours  (Young et al.,  1977).   The predominant  route of elimination is



through the urine.   The routes of  elimination are  dose-related with the percent



eliminated through the urine less  for small doses  as  compared to larger doses,
                                        13-16

-------
while the relative  amount  retained by the carcass  is  greater  for the small




dose as  compared  to a larger dose.  Acrylonitrile is metabolized to cyanide,




which is transformed to thiocyanic and by cyanoethylation of sulfhydryl groups




to S-(2-cyanoethyl)cysteine, followed  by  elimination of these  metabolites in




the urine.   Other minor metabolites are formed from acrylonitrile.  The toxicity




of acrylonitrile  is  caused  by  both the acrylonitrile molecule  itself and its




metabolites.




13.2  ACUTE,  SUBCHRONIC, AND CHRONIC TOXICITY




13.2.1.   Acute Toxicity




     The initial  investigation  of the acute  toxic  effects  of  acrylonitrile




began in the 1940's when industrial use of acrylonitrile increased dramatically




due to the production of oil-resistant synthetic rubber.  These studies included




both laboratory investigations  in animals and compilations of  case histories




of human intoxication  that  had  occurred in the workplace.   Since these early




cases of intoxication, occupational exposure to acrylonitrile has been greatly




limited by
                                        13-17

-------
                            14
Table 13-5.   Metabolites of   C-Acrylonitrlle Separated from Various
             Fluids of Rats by High Pressure Liquid Chromatography (UPLC)
             (Young £t_ al., 1977)
Sample

Drine, 0-72 hr.
0.1 mg/kg, p.o.
10.0 mg/kg, p.o.
5 ppm
100 ppm
Bileb, 1 hr.
In vitro0
Scomachb>d, 24 hr.
RJBCb>dt, 24 hr.
Plasma , 24 hr.
14 a
Percentage of Total C
A


12
61
9
32
2
17
4
7
28
C


73
8
30
33
91
78
3
10
28
E


15
32
61
35
1
1
93
83
42
 aMetabolites A, C, and E were the major metabolites separated by HPLC.
  Values are expressed as percentages of the total radioactivity in
  each sample applied to the column.  The total percentages of A, C,
  and E are less than 100% when minor metabolites B and D are present.

  A dose of 1 mg   C-acrylonitrile/kg was given intravenously.

  Metabolites extracted from reaction mixtures with ethanol.  The source
  of enzyme was rat liver 9000 x g supernatant fluid.

  Ethanol extracts of stomach or red blood cells (RBC).  About 80% of
  the radioactivity was extracted.
                                       13-18

-------
engineering controls,  and no  recent cases  of acute  human intoxication  have



occurred.



13.2.1.1  Acute Systemic Human Toxicity



     Non-fatal intoxication by  acrylonitrile has  been reported in workers who



clean polymerizers in  plants  that  manufacture synthetic rubber (Wilson,  194H;



Wilson e_£ jil., 1948).  During  these operations, workers were exposed  to between



16 and  100 ppm of acrylonitrile for  periods  of 20 to 45 minutes.  All workers



complained of  nasal irritation and an oppressive feeling in the upper respiratory



passages.  Dull headaches, nausea, and subjective  feelings  of  apprehension and



nervous  irritability were also common complaints  following mild  intoxication.



In more  severe cases  of  intoxication,  a  low grade anemia and mild  jaundice



occurred.  All exposed workers appeared to recover fully,  with  the exception of



one who had severe Jaundice.   This worker complained of lassitude and fatigue a



year later, although  no pathological signs could be found.   Similar symptoms of



headache, vertigo, nausea, and  vomiting  were reported by a  chemist  engaged in



distilling acrylonitrile  (Sartorelli,  1966).   Tremors,  uncoordinated movement,



and convulsions were  also  experienced.  As with the other cases  of intoxication,



complete recovery occurred. Baxter (1979)  has recently summarized the symptoms



of acrylonitrile poisoning in  man,  in order of occurrence,  as follows:   irrita-



tion of eyes and nose,  limb weakness, labored breathing,  dizziness and impaired



Judgement, cyanosis and nausea,  collapse, irregular breathing,  and convulsions,



possibly followed by  cardiac arrest.  A few additional case histories  of occupa-



tional exposure to acrylonitrile were presented in an IARC Monograph  (1979).



     Five cases  of acute lethal human  intoxication have been  reported (Davis



et al.,  1973)   following  fumigation  of  domiciles with acrylonitrile  mixed  with



ether 66$ carbon tetrachloride or 70$ methylene chloride (v/v).  It is not clear



whether acrylonitrile or the other component  of the fumigant was  the causitive
                                          13-19

-------
agent in these deaths;  however,  one patient had measurable levels  (0.05  mgt)  of



cyanide in his blood at autopsy, and,  in  another  case,  the odor of cyanide  was



noted in the tissue.   The  presence  of cyanide only indicates that  acrylonitrile



was metabolized prior to death.  The only consistent symptom in these victims  was



nausea as  an initial indication of intoxication.  On autopsy, no gross abnormali-



ties that could be attributed to the fumigant were noted.



     In a report of four case histories of human disease following fumigation of



domiciles with acrylonitrile and carbon tetrachloride mixtures, Radimer et^  al.



(1971) described the  appearance  of  toxic epidermal necrolysis.  The first  symp-



toms appeared between  11  and 21 days following rehabitation of the house (all



houses were declared safe by the  exterminators).  Symptoms observed in victims on



entering the  hospital resembled  those of extensive  second-degree burns.  Three of



the patients died 2 to 9  days after hospitalization and one recovered  and  was



discharged in  10 days.  Biopsy  samples showed blisters at the  junction of  the



dermis with a necrotic epidermis. It was believed that exposure occurred through



inhalation rather than skin absorption.  Although a participating role for carbon



tetrachloride could not be overruled, the  lack  of reports  of epidermal  disease



following carbon tetrachloride  exposure made the authors  suspect that  acrylo-



nitrile was the toxic agent.



13.2.1.2  Acute Systemic  Animal  Toxicity



     Acrylonitrile is a highly toxic compound to laboratory animals, regardless



of the route  of  administration.  In rats,  the  oral LD50   of acrylonitrile  was



determined to be 113  mg/kg by Smyth et al. ( 1969) and the minimal fatal oral dose



in rats was reported  to be 150 mg/kg (Wilson et al., 1918).  The LD50 in Wistar



rats  following  subcutaneous  or  intraperitoneal injection  of acrylonitrile  was



80 mg/kg and 100 mg/kg, respectively (Knoblock et  al.,  1971).   Benes and Cerna



( 1959) had  reported  the  oral LD50 of acrylonitrile in both rats  and  mice as
                                         13-20

-------
78 mg/kg and 27 mg/kg,  respectively.   The  approximate  lethal  doses  in mice  have



been determined for acrylonitrile administered  intragastrically and intraperi-



toneally, and  were  greater than 20 mg/kg and  15 mg/kg,  respectively (McOmie,



19^3).    The  symptoms of intoxication in rats were briefly described by Wilson



et al^  (19^8)  as  respiratory changes,  cyanosis,  and  convulsions  followed  by



death.   For the investigation of acute toxic effects of acrylonitrile, exposure



by inhalation  has been studied  at  length since this  is  the  route of greatest



concern in industrial exposure.



     Dudley and Neal (19^2) exposed  Osborne-Mendel rats to atmospheres of  90  to



2^5 ppm acrylonitrile for  30 minutes  to 8 hours.   The  subjective  signs  of



toxicity and the lethality  of acrylonitrile were strictly dependent on the length



of exposure and the concentration of  acrylonitrile.   Short periods of exposure



(665 ppm for 30 minutes) produced only mild  effects, while a  similar dose  given



over a long exposure period (635 ppm for H hours)  was  fatal to all  the animals.



     Smyth and Carpenter (19^8) observed fatalities in all 6 Sherman strain rats



following a  4-hour  exposure to  1000  ppm  acrylonitrile,  and  no  deaths  from  a



500 ppm exposure.   In a second study,  however,  Carpenter  et^ al. (19^9) reported



that 2 to H rats (Sherman strain) of a group of 6 animals succumbed following a



500 ppm  exposure  for  1 hours.   Brieger et  al.  (1952)  exposed Wistar  rats  to



100 ppm acrylonitrile for 7 hours,  with lJ of 20 animals dying.  A  similar dose



(130 ppm for  H hours)  was a  non-lethal  dose in  the  study of Dudley and  Neal



(1912)  (Table 13-6).  Due to insufficient data, it is  not possible  to  determine



whether the observed differences in  susceptibility of  rats in these experiments



were a result of mere experimental uncertainty, or were related to  variations in



strains  of  rats used,  sex of the animals (this was not  reported in any of the



studies), or age of the animals (all animals were between 100 and  150  g, except



the Osborne-Mendel rats, which were  295 g).
                                          13-21

-------
Table 13-6.  Summary of Results of Exposures of Rats to Acrylonitrile*
             (adapted from Dudley and Neal, 1942)
Length of
Exposure
(hours)

1/2




1






2


4



8





Concentration
(ppm)


2445

1490
1270
665
2445

1490

1270

665
1260
595
305
635
315

130
320
270

210
135
90
Dying
During
Exposure

0

0
0
0
0

0

0

0
0
0
0
50
25

0
94
44

6
0
0
Total
Deaths
(%)

0

0
0
0
81

25

0

0
100
6
0
100
31

0
94
44

6
0
0
Severity of Effects and Remarks



Marked. Slight residual effects to
24 hrs.
Marked. No residual effects in 24 hrs.
Marked. No residual effects in 24 hrs.
Moderate, transitory.
Deaths in 4 hrs. Show some slight
effects 24 hrs after exposure.
Deaths in 4 hrs. Show some slight
effects 24 hrs after exposure.
Marked. Showed some effects 24 hrs
after exposure. Normal in 48 hrs.
Marked, transitory.
Fatal. Dead in 4 hrs after exposure.
Marked, transitory.
Slight, transitory.
Fatal. Dead in 4 hrs after exposure
Marked. Survivors show no effects
in 24 hrs.
Slight transitory effects.
Fatal
Marked. Survivors show no effects
in 24 hrs.
Marked, transitory.
Moderate, transitory.
Slight discomfort only.
   16 rats were exposed at each concentration for length of time shown.

  Rats used for these studies were adult albinos, pure Osborne-Mendel;
  average weight, 295 grams.
                                        13-22

-------
     There was considerable species variation in the  susceptibility  of  animals



to  the  lethal effects  of  acrylonitrile,  with  the  guinea pig  being the most



resistant species  and the dog the  most sensitive (Table  13-7).   Many of  the



animals did not die during the exposure period but rather succumbed to the  toxic



effects later.  The number of fatalities given in the  table indicates the  total



number of animals  that  died  during the exposure and  observation period.    The



concentrations given  in Table 13-7 should  not  be  interpreted  as the  minimum



lethal doses in the cases where a substantial number of challenged animals  died;



these were  the lowest concentrations reported  but  are presumably higher  than



would be the minimum lethal doses.



     The symptoms  of  intoxication  were  similar  in all  species  except dogs  and



guinea pigs.   As reported by Dudley  and Neal  (1912),  the symptoms  following



lethal exposures included gasping (particularly of  the  abdominal type), spasm-



like convulsive movements of the abdominal wall, general  convulsions, and  coma



followed by death.  With less severe exposure, hyperactivity and flushing (parti-



cularly in the Rhesus monkeys)  were  noted along with indication of slight  irrita-



tion of the mucous membranes.   Animals that  were removed from  the acrylonitrile



atmosphere prior to the gasping stage recovered completely with no therapeutic



attention (McOmie, 19^3).  Dogs  showed  all the same  symptoms as other  species



upon exposure to a lethal concentration  of acrylonitrile;  however, for the dogs



that did recover,  the  residual effects of poisoning lasted for a longer period of



time (Dudley and Neal,  19^2).  The  slow recovery included refusing  to  eat  for



10 days  following  exposure  in 1 animal,  and,  for  another,  3 days  of  partial



paralysis of  the  hind legs.   In  the most resistant  species,  the guinea  pig,



symptoms included  watering of the eyes,  nasal  discharge, and  coughing.   When



death did occur, it followed exposure by  3 to  6 days.  Dudley  and Neal (19^2)
                                         13-23

-------
   Table 13-7.   Minimal Lethal Concentration of Acrylonltrile
                During Four-Hour Exposure
Animal Species

Guinea Pig
Cat
Rat
Micea
Rabbit
a
Rhesus Monkey
Dog
PPM of Acrylonitrile

575
600
315
415
260
75
65
Fatalities/No.
Challenged

10/16
16/16
5/16
3/16
16/16
1/3
1/2
Reference

Dudley & Neal, 1942
Dudley & Neal, 1942
Dudley & Neal, 1942
McOmie, 1943
Dudley & Neal, 1942
Brieger et al., 1952
Dudley & Neal, 1942
Exposure time was three hours instead of four hours.

The great sensitivity of dogs was confirmed in the study of Brieger et al.,
1952.
                                     13-24

-------
concluded that death  of  guinea  pigs was due  to  mucous membrane and  pulmonary




irritation and lung edema.




13.2.1.3  Mechanisms of Acute Toxicity




     Early observation of animal experiments and human  intoxication  suggested




that acrylonitrile toxicity was  due  to the liberation of cyanide following meta-




bolism.  Dudley el^ al. (19^2)  compiled data comparing the toxic levels of hydro-




cyanic acid and of acrylonitrile in a number of species of experimental animals




(Table 13-8).   Both in the quantities needed for intoxication and in the symptoms




observed, acrylonitrile appeared to act  in  a  manner similar to that  of hydro-




cyanic acid.   Furthermore,  in humans,  the gross signs and symptoms  of acrylo-




nitrile and cyanide poisoning were similar (Brieger ^t  al.,  1952).




     With human exposure, the main biochemical evidence in support of the theory




that acrylonitrile toxicity occurred through the action of cyanide  was the pre-




sence  of  thiocyanate  in  the  blood  and  urine of acrylonitrile-exposed indivi-




duals.  Thiocyanate is the end product of the body's detoxification of cyanide.




Following  occupational  exposure  to acrylonitrile,  maximum  serum  thiocyanate




levels occurred immediately  after exposure,  while  urinary maxima were reached




between 24 and 48 hours (Mallette,  19*43). The persistence of  thiocyanate in the




serum was proportional to the level of  exposure, with blood levels  normal  at




2.5 hours after  an exposure of  22  ppm  acrylonitrile for 30 minutes  and  blood




levels not returning to normal values by  12 hours after an exposure of 50 ppm for




30 minutes (Wilson £t al.,  1948).   Due  to  the quantitative correlation between




acrylonitrile  exposure and  urinary  and  blood thiocyanate concentrations,  mea-




surement of levels of  thiocyanate was suggested as a  method of monitoring acrylo-




nitrile absorption in individuals (Mallette,  1943).



     To determine whether acrylonitrile  acts  by the same mechanism as cyanide,




animal models  have been  used to identify and  compare cyanide metabolites  in
                                         13-25

-------
              Table 13-8.   Comparison of the Effects  of  Acrylonitrile and of Hydrocyanic Acid on Various Species

                           of Animals (Dudley and Neal,  1942)




Species
Rats
n
Guinea Pigs

n n

Rabbits

It
M
Cats

M

Hogs
tt

M

N
Monkey •
Acrylonitrllc
Concentration


•g/i
0.26
0.21
1.25

0.58

0.56

0.29
0.21
0.60

0.33

0.24
0.213

0.12

0.063
0.33
Calculated
as HCN
Bg/1
0.14
0.10
0.62

0.29

0.28

0.15
0.10
0.30

0.17

0.12
0.11

0.06

0.03
0.17
pp»
130
100
575

265

260

135
100
275

153

110
100

56

30
153



SyaptOM
Slight transitory effect
Slight transitory effect
Fatal during or after
exposure
Slight transitory effect

Fatal during or after
exposure
Harked transitory effect
Slight transitory effect
Markedly toxic

Markedly toxic, sovetlaes
fatal
Fatal to 3/4 of the doga
Convulsions + coma, 0 death

Transitory paralysis.
1 dog died
Very alight effects
Definite toxic effects
Hydrocyanic Acid
Concentration


•8/1
0.12

0.35

0.23

0.35

0.13

0.35

0.14

0.125
0.1

0.07-0.04

0.035
0.14


PP"
110

315

200

315

120

315

125

115
90

63-35

30
125



Svxptosu
Fatal In 1 1/2 hours

Fatal concentration

Tolerated for 1 1/2 hours without
ayoptosja
Fatal concentration

No narked toxic sysipto«a

lo 2 Bin respiratory paralysis, In
5 to 10 »ln death
Markedly toxic in 6 to 7 "In

Fatal to dogs
May be tolerated for houra, death
after exposure
Voaltlng, convulsions, recovery.
•ay be fatal
May be tolerated
Distinctly toxic after 12 mln
CO
I
ro

-------
animals exposed to acrylonitrile.  Cyanide acts by complexing with heavy metal




ions, alone or as  part of organic molecules, and the ultimate toxic effect is due



to the complex of cyanide with the ferric ion in the cytochromes.  This causes




disruption in cytochrome function and cellular anoxia.  During cyanide intoxica-




tion, two less toxic metabolites—cyanmethemoglobin and thiocyanate—are formed




which competitively inhibit the toxic action of cyanide.  The ability of cyanide




antidotes, which  facilitate the formation  of competitive  metabolites of  the




cyanide ion, to protect against  acrylonitrile poisoning has also  been investi-



gated.




     Cyanide  has  been  found  at various  levels in  animals  after exposure  to




acrylonitrile.  Dudley and Neal (19^2)  examined the  liver,  kidney,  and gluteus




muscle of rats and  guinea  pigs immediately after death  due to  inhalation  of




acrylonitrile and could  find  no detectable levels of  cyanide.    Hashimoto  and




Kanai (1965), however,  detected cyanide at 2.5 to  3.5 ug/ml in blood of rabbits




following intravenous administration of acrylonitrile (75 mg/kg)  and in rats and




guinea pigs after intraperitoneal administrations  of  100 and 125 mg/kg, respec-




tively.  Cyanide was also detectable in  the blood of both dogs and Rhesus monkeys




following inhalation exposure  to acrylonitrile, but not  in the  blood of  rats




(Table 13-9).  In dogs,  the levels of cyanide in the  blood were  proportional to




the extent of exposure  to acrylonitrile.  This  was the only study that found a




positive correlation between the sensitivity of the  species exposed  to acrylo-




nitrile (Table 13-7)  and the level of cyanide in the  blood.




     Both cyanide metabolites,  thiocyanate  and  cyanmethemoglobin,   have  been




detected in the blood (and  urine,  in the  case of  thiocyanate) of animals  after




exposure to acrylonitrile.   Breiger ejt aU (1952) compared the  levels of  thio-



cyanate in the blood  of dogs with the levels in the blood of rats, a species which




was  less  sensitive to  the  toxic  effects  of acrylonitrile (Table  13-9).    An
                                         13-27

-------
                            Table 13-9.  Cyanide and Thlocyanate in Blood of Animals Exposed

                                         to Acrylonitrile (adapted from Breiger et al., 1952)
Species

Rats
Dogs
Dogs
Rats
Dogs
Dogs
Rats
Dogs
Dogs
Rhesus monkey
Rhesus monkey
Acrylonitrile
in Air (ppm)

100
100
100
75
75
75
50
50
50
75
75
Length of
Exposure

7
1
7
7
1
7
7
1
7
3
7
Cyanide in
Blood
(vimol/ 100ml)

1
2.93
10.10
0
1.85
6.65
0
trace
0.51
3.70
8.72
Thiocyanate mg/100 ml
Blood

6.63b
0.45
0.99
4.21b
0
0.39
1.74b
0.55
0.80
1.67
7.98
Urine

2.2
	
3.6
1.1
	
1.93
0.25
	
1.47
	 i
4.12
CO
I
ro
00
                Zero exposure levels are subtracted from the values presented.

                Determined in plasma.

-------
inverse relationship was demonstrated between the levels  of thiocyanate,  levels

of cyanide,  and degree of sensitivity, with the species less sensitive to acrylo-

nitrile poisoning having lower blood cyanide and higher thiocyanate levels.   The

authors were surprised at  the high thiocyanate levels as compared to the cyanide

levels in the blood  from monkeys.   Cyanmethemoglobin was also detected  in  the

blood of all exposed animals, with the highest concentrations found in the  Rhesus

monkey (H.8 umol/100 ml) and the dog (2.7 umol/100 ml)  following a 7-hour expo-

sure to 75 ppm acrylonitrile.   There was no detectable cyanmethemoglobin  in  the

blood  of  rats at  this  exposure level;  however,  traces were  detected at  the

100 ppm exposure level.  Breiger et al. (1952) concluded that acrylonitrile must

be metabolized  to  cyanide  and  they suggested  the possibility that  a  species'

ability to  detoxify cyanide was related  to its  resistance  to  acrylonitrile

poisoning.

     Several antidotes to  cyanide poisoning have been evaluated for their  effec-

tiveness  following acrylonitrile exposure,  and  as  an  indication of whether

acrylonitrile has a mechanism of action similar to that of cyanide.  Two types of

antidotes have been used.  The first,  sodium thiosulfate, facilitates the detoxi-

fication of cyanide to thiocyanate by the following reaction:

      Thiosulfate + cyanide       rhodanese         sulfite +  thiocyanate
                               thiocyanate oxidase

The second type consists of chemicals (methylene  blue  and  nitrites)  that cause

the  formation of methemoglobin, which  competes with  the  cytochromes for  the

cyanide ion.  The use of both types of antidotes has indicated that acrylonitrile

toxicity cannot be attributed solely to the release of cyanide,  and that  either

acrylonitrile or  another  metabolite of  acrylonitrile has  significantly  toxic

properties.

     Hashimoto  and  Kanai  (1965)  examined  blood  levels of  acrylonitrile,  thio-

cyanate,  and cyanide  in  rats, guinea  pigs,  and  rabbits  after  intravenous
                                         13-29

-------
(rabbits) or intraperitoneal (rats and guinea  pigs)  injection  of acrylonitrile


or of sodium thiosulfate followed in three minutes by acrylonitrile.   As illus-


trated in Figures 13-3 and  13-1*, in the rabbit, thiosulfate pretreatment did not

affect the  levels of acrylonitrile in  the blood, but thiosulfate did  cause  a


dramatic decrease in the cyanide levels and an increase  in thiocyanate levels.


The use of thiosulfate prevented the death of three of the four rabbits exposed


to 75 mg/kg acrylonitrile and reduced the severity of the symptoms of intoxica-

tion.  At 100 mg/kg,  death  was delayed by thiosulfate,  but all animals eventually


succumbed to the toxic effects of acrylonitrile, even though the cyanide levels


present  in the blood were  lower  than  those in animals that survived a 30 mg/kg


exposure to acrylonitrile  alone.  Similar  results were obtained using rats and


guinea pigs, with the animals dying  in spite of  reduced  blood cyanide levels

(2.62 to 0.18 ug/ml  in  guinea pigs and 3-01  to 0.51 ng/ml in  rats)  following


treatment with thiosulfate.  It was apparent that even  with extensive detoxifica-


tion of liberated cyanide,  acrylonitrile lost little of its toxicity.


     Magos  (1962)  attempted  to prevent  the  toxic  effects of subcutaneously

injected acrylonitrile by pretreating rats with an intraperitoneal injection of

sodium nitrite.   The results are shown in Table 13-10.   Sodium nitrite effec-

tively protected rats from the lethal effects  of cyanide,  but  had no effect on

either survival rate or  survival  time  of animals  treated with acrylonitrile.  An


estimation of the percentage inhibition of cytochrome oxidase in rats receiving
                         •
cyanide and in rats receiving acrylonitrile revealed greater inhibition (8H$) in


rats that survived cyanide poisoning  than  in rats (71.5$) that were killed by

acrylonitrile.   From these  results,  it appeared  that cyanide did  not  play  a

significant role in the  toxic effects  of acrylonitrile in rats unless the forma-

tion  of  cyanide  from  acrylonitrile  in  the  tissue  of exposed  animals allows

toxicity at a lower cyanide concentration.
                                          13-30

-------
    £30-
                       OACRYLONITRILE
                       • THIOCYANATE
                       ACYANIDE
           0      60     120     180      240
         TIME AFTER INJECTION OF ACRYLONITRILE (min.)

                 a. ACRYLONITRILE 30mg/kg
                                               | I I I I I I I  I Tl I

                                       0      60     120     160
                                 TIME AFTER INJECTION OF ACRVLONITRILE (min.)

                                        b. ACRYLONITRILE 7Sm«/kg
Figure  13-3.
Distribution of Acrylonitrile,  Cyanide,  and Thiocyanate
in the Blood after a  Single Injection  of Acrylonitrile
(Rabbit)  (Hashimoto and  Kanai,  1965)
                                           13-31

-------
                            TOXIC ACTION OF ACRYLONITRILE
                     O ACRVLONITRILE
                     • THIOCYANATE
                     ACYANIDE
                                              BO-
                                                   DEATH
        0       60     120     180     240
     TIME AFTER INJECTION OF ACRVLONITRILE (min.)

        • . ACRVLONITRILE 75mg/kg 3 MINUTES
          AFTER SODIUM THIOSULFATE 320mg/kj
                                           1	I "'" I
                                    0      60      120     180
                             TIME AFTER INJECTION OF ACRYLONITRILE (min I
                                b. ACRYLONITRILE lOOmg/kg 3 MINUTES
                                  AFTER SODIUM THIOSULFATE 430m«/kg
Figure 13-4.
Effect of Sodium Thiosulfate  on the  Distribution
of Acrylonitrile, Cyanide,  and  Thiosulfate
(Rabbit)  (Hashimoto  and Kanai,  1965)
                                         13-32

-------
Table 13-10.  Effect of Methemoglobinemia on Mortality
              Ratios In Albino Rats Poisoned with
              Aerylonitrile, Potassium Cyanide, and
              Acetone Cyanohydrin (Magos, 1962)
Compound

Acrylonitrile
Potassium cyanide
Potassium cyanide
Acetone cyanohydrin
Dose
(millimole/kg)

2.8
2.4 x 10"1
3.7 x ICf1
1.6 x 10"1
No. That Died /No. Given
Cyanide Compound
Without Sodium
Nitrite

5/5
5/5
—
5/5
With Sodium
Nitrite

5/5
1/5
4/4
0/5
                             13-33

-------
     Benes and Cerna (1959) treated mice and rats orally with both sodium nitrite



and thiosulfate and observed  changes in the lethality of a subsequent oral dose



of acrylonitrile.  In rats, the cyanide  antidotes  protected  30f  of the animals



from a LD100 dose of acrylonitrile.  In mice, the effect  was more  dramatic, with



protection afforded at up to three  times  the  LD50  dose.   It  was  concluded that



toxicity in mice might  be due to  cyanide, while  in  rats the toxicity  is  due



mainly to the acrylonitrile molecule  itself.   It was shown that  by 3 days post



exposure, rats had only metabolized 19.'I?  of  the acrylonitrile to  cyanide (the



data for mice were not  given).  Benes and Cerna (1959)  suggested  that species



variation in susceptibility to acrylonitrile  intoxication might  be  due  to  the



ability of animals to metabolize  acrylonitrile to cyanide.  Other investigators



(Magos, 1962;  Gut et al_.,  1975),  however, suggested that the  difference was due



to differences in ability to detoxify the liberated cyanide.   The exact role of



liberated cyanide both in the toxicity of acrylonitrile and in the almost tenfold



difference  in susceptibility  of  experimental animals to acrylonitrile1s toxic



effects is still unclear.



     In further studies  of the mechanism of acrylonitrile's toxicity, Hashimoto



and Kanai  (1965)  pretreated  rabbits  intravenously  with L-cysteine 3 minutes



prior to dosing  with acrylonitrile.   It  was shown  in a  test  tube reaction that



L-cysteine forms an addition  product with acrylonitrile  by cyanoethylation,  and



Gut e_t al.   (1975)  demonstrated that S-(2-cyanoethyl)-cysteine was non-toxic and



not metabolized to thiocyanate when administered orally  to rats.   Hashimoto and



Kanai (1965) analyzed the blood of  rabbits treated with acrylonitrile and with



acrylonitrile and cysteine,  and found a large decrease in the levels of acrylo-



nitrile and cyanide  in  animals treated with  cysteine (Figures  13-5 and 13-6).



Unlike the antidotes to cyanide poisoning, cysteine not  only relieved the symp-



toms of  poisoning,  but  rabbits receiving 100  mg/kg acrylonitrile also survived
                                        13-34

-------
                     OACRYLONITRILE
                     • THIOCYANATE
                     A CYANIDE
                       120
                              180
                                     240
      TIME AFTER INJECTION OF ACRYLONITRILE (men.)

              I. ACRVLONITRILE 30mfl/kg
                                     0      60      120     180
                              TIME AFTER INJECTION OF ACRYLONITRILE (min.l

                                      b. ACRYLONITRILE TSmg/kg
Figure  13-5-
Distribution  of Acrylonitrile,  Cyanide,  and Thiocyanate
in  the Blood  after a  Single Injection  of Acrylonitrile
(Rabbit) (Hashimoto and Kanai,  1965)
                                       13-35

-------
                 OACRYLONITRILE
                 • THIOCYANATE
                 ACYANlOE
                           r
    0      60     120     180     240
  TIME AFTER INJECTION OF ACRYLONITRILE (mm.I
      ». ACRYLONITRILE 75ing/kt 3 MINUTES
        AFTER L-CYSTEINE-HCI 220mg/kg
                                           10-
                        o
                        8
                        a
                        ut
                        Z
                        »-
                        Z
                                         8
                                            s-
                                0      80     120      180     240
                              TIME AFTER INJECTION OF ACRYLONITRILE I mm. I

                                 b. ACRYLONITRILE lOOmg/kg 3 MINUTES
                                    AFTER L-CYSTEINE-HCl300m«/kg
Figure 13-6.
Effect of L-Cysteine  on  the Blood Concentrations
of-Acrylonltrile, Cyanide,  and  Thiocyanate
(Rabbit)  (Hashimoto and  Kanai,  1965)
                                       13-36

-------
and completely recovered from  the  treatment.   Similar results were  noted  with




guinea  pigs,  rats,  and mice (Table  13-11).   Positive effects were  noted  with




other sulfur-containing compounds.   Because cysteine could protect animals  from




the toxic effects of acrylonitrile by  reacting  directly  with  the acrylonitrile




molecule,  and  cyanide antidotes  were  not  effective  in  preventing  death,




Hashimoto and Kanai  (1965) speculated that acrylonitrile itself was a toxic agent




and may act in conjunction with cyanide to produce the ultimately toxic effects.




Pilotti (1975) found no acrylonitrile toxicity in vitro, however, as indicated by




inhibition of growth of ascites sarcoma BP8 cells  in  culture  (1.0 mraol acrylo-




nitrile  was  assayed)   and  attributed  this to  the  lack  of formation  of  toxic




metabolites.   It  is uncertain  whether the  toxic activity ascribed  to acrylo-




nitrile is due to the acrylonitrile molecule per se or to a non-cyanide metabo-



lite of acrylonitrile.




     Acrylonitrile also reacts  with endogenous glutathione  by either one or both




of two possible mechanisms.  Direct  interaction may occur by S-cyanoethylation of




glutathione, or alternatively  by enzymatic conjugation  with  glutathione during




the  detoxification  of acrylonitrile.    Dinu  (1975b)   demonstrated  dramatic




decreases in  non-protein tissue thiol  concentration (protein  thiol levels  were




unchanged) in rats following subcutaneous injection of  two times the LD50 dose of




acrylonitrile  (Table  13-12).   Similar results were  observed by Szabo et^ al.




(1977a, 1977b), who used direct measurements of tissue glutathione.  The possi-




bility  that  glutathione levels  were  important  in mediating  the toxicity  of




acrylonitrile was indicated  by the greater  sensitivity of  rats to  the  toxic




effects of acrylonitrile at night when liver glutathione levels were low due to




normal diurnal variation (Jaeger,  1979).   Also, Szabo ejt  al.   (1977a)  showed a




correlation between  the doses of  acrylonitrile that caused  a precipitous decline




in glutathione concentrations  in the  brain of  rats and  the doses  that caused
                                         13-37

-------
                       Table 13-11.   Therapeutic Effect of SH .and S-S Compounds on Acute
                                     Acrylonltrlle Poisoning3 (Hashimoto and Kanai,  1965)
Animal
Amount of
Acrylonitrile
Injection
Site of
Injection


Rabbit (5)b
100 rag/kg
iv
Guinea-pig (5)
125 rag/kg
IP
Rat (5)
100 mg/kg
IP
Mouse (5)
100 mg/kg
IP
Control
survival survival
rate time
(min)

0/5 98(5)
0/5 86(5)
0/5 85(5)
0/5 101(5)
L-Cysteine HC1
300=370 mg/kg
survival survival
rate time
(min)

5/5 -C(5)
2/5 132(3)
2/5 254(3)
5/5 -C(5)
L-Glutathione
570 mg/kg
survival survival
rate time
(min)

5/5 -c(5)
d
— — — _— _
... ...
a-Mercapto-
propionylglyclne
350 mg/kg
survival survival
rate time
(min)

4/5 100(1)
	 	
	 	
"
L-Methionine
425 mg/kg
survival survival
rate time
(min)

5/5 (5)
	 	
— — — — —
i
CO
I
OJ
00
           Amount of AN = 2 x LD50, amount of SH and S-S compounds = AN x 1 = 1.5 equivalent
           All numbers in parentheses represent number of animals.
          CAnimals survived treatment.
           	•» Animals Not Treated.

-------
Table 13-12.  Concentration of Protein (PBSH) and Nonprotein (NPSU)
              SH Groups In Normal and Acrylonitrile-Intoxicated
              Animals (ymoles SH/100 g vet tissue)  (Dlnu,  1975b)


Liver


Kidney


Brains


Lung


Testis


Adrenal



PBSH
NPSH

PBSH
NPSH

PBSH
NPSH

PBSH
NPSH

PBSH
NPSH

PBSH
NPSH
Normal Animals
Mean + Standard Error


1604 +28.3
412 + 16.8

952 + 40
250 + 9.63

968 + 27.4
103 + 3.55

989 + 30
82 + 3

1070 + 41
237 + 3.6

528 + 20.7
127 + 4
Intoxicated Animals
Mean + Standard Error


1524 +46.3
42.4 + 1

900 + 26.3
24.6 + 0.6

836 + 26
34 + 1.48

864 + 24.4
28 + 1.48

1000 + 39
86 + 4.9

538 + 11
0 + 	
                                     13-39

-------
mortality;  however,  they noted no tissue damage by light microscopy in the liver,


lung, or  kidney of  rats  60 minutes  after an intravenously administered  dose


(15 mg/100 g body weight) of acrylonitrile, even though  the  glutathione levels


were reduced by 80 to 90$.


     Dinu  (1975a)  presented  evidence that  low glutathione  levels caused  by


acrylonitrile lead to excess formation of lipid peroxides and subsequent disrup-


tion of membrane function.  Following treatment of rats with acrylonitrile (two


times the  LD50  dose administered  subcutaneously),  increases in  malonaldehyde


were observed in the liver.   According to Dahle et  al.  (1962),  the formation of


malonaldehyde parallels the  formation of lipid peroxides.  The increase in lipid


peroxides occurred even though the  treatment with  acrylonitrile  stimulated the


activity of glutathione peroxidase.  The low levels of glutathione in intoxicated


animals  made the  protective  effects of  glutathione  peroxidase  ineffective.


Although comment was made on  the  disruptive  properties  of lipid  peroxides  on


membrane function, no measurements of membrane function were made.


     Hashimoto and  Kanai  (1965)  examined  respiration  in guinea  pig  brain and


liver slices incubated with acrylonitrile  (10"  M) .  The oxygen consumption of


the brain slices was measured under  both potassium stimulating and non-potassium


stimulating conditions.  Acrylonitrile caused a 20$ decrease in the oxygen usage


of the potassium stimulated preparation,  while no inhibition was  noted in the


non-potassium stimulated  preparation.  This inhibition of respiration appeared


to be specific to the brain slice.   Inhibition of respiration in liver slices was


not observed until 5 x 10~  M acrylonitrile was used,  at which point precipita-


tion of protein occurred in  the liver.  There was  no protection  offered by the


simultaneous administration of  sodium thiosulfate with  acrylonitrile.   It was


also apparent that inhibition of  respiration  was  not due to the  liberation of

                             _H
cyanide, since  cyanide at 10    M inhibited both potassium  stimulated and non-
                                        •13-40

-------
stimulated brain slices, and sodium thiosulfate protected the brain slices from



the  detrimental  effects  of cyanide.   Effects on  the peripheral  nerves  were



demonstrated by applying acrylonitrile to the nerve trunk in an isolated Sciatic-



gas trocnemius  preparation from frog and  observing  a  rapid  anesthetic action.



This anesthetic effect  rapidly disappeared  following  removal of acrylonitrile.



It was concluded that acrylonitrile produces abnormal  function of both the peri-



pheral and central nervous systems.



     Acute adrenal apoplexy has been described  in  rats  following the administra-



tion of acrylonitrile (Szabo et al., 1976b).  Following intravenous administra-



tion of 15 mg acrylonitrile/100 g body weight,  the lesions appeared bilaterally



between 1 and 2 hours after treatment.  The intravenous  administration of this



dose produced hemorrhage  in 90 to  100$  of the  animals, while administration of



the same dose of acrylonitrile by gavage produced hemorrhage in only 20$ of the



animals.   As reported  by Szabo and Selye (1971), mortality in  both  cases was



100$.  Szabo e_t a^. (1977b) observed early hematological changes consistent with



a longer blood clotting  time in rats  following  the intravenous administration of



acrylonitrile, but it was not  clear  if  these hematological  changes were suffi-



cient in themselves  to  cause  the hemorrhage.   Protection from adrenal apoplexy



and,  to some degree,  mortality was  offered by pretreating the animals for 1 days



with phenobarbital and some steroids.  It was believed that  this protection was



offered by the ability of these xenobiotics to induce drug-metabolizing enzymes



and hence alter the metabolism of acrylonitrile (Szabo and Selye, 1972).  Since



mortality was not always associated with  adrenal  apoplexy,  it  was  not apparent



how this hemorrhage effect related to the systemic toxicity  of acrylonitrile.



13.2.1.1  Acute Topical  Irritation and Toxicity



     McOmie (1913) applied  acrylonitrile  to the shaved abdomens of rabbits at



doses of 1.0, 2.0, and 3.0 ml/kg to  areas of  100, 200, and 200 cm2, respectively.
                                          13-41

-------
No attempt was made to prevent evaporation of the test compound.   Slight  vaso-

dilatation was observed at  the low dose, while the two higher doses caused slight

erythema.  Only at  the  highest dose was systemic toxicity  noted,  as indicated by

depression  and increased  respiration  rate.    The animal  showing toxic  signs

recovered without  any  therapeutic  treatment.   When  acrylonitrile was  kept in

contact with the abraded skin of  rabbits by absorption into a cellulose pad,  the

LD50 was 0.25 ml/kg (Roudabush et^ al.,  1965).   When  guinea pigs were  used,  the

LD50 for  animals with  abraded  skin was 0.86 ml/kg,  while  the LD50 for animals

with intact skin was O.H6 ml/kg.   It is interesting  to note that  in  the guinea

pig, toxicity was greater for acrylonitrile when applied  to intact skin.

     Acrylonitrile  caused  only  mild irritation to   the  eye  of  a rabbit when

applied in a single 0.05 ml drop  (McOmie,  19^3) •  One hour following  treatment,

mild conjunctivitis was  observed with  no clouding of  the cornea  or  pupillary

damage.  The animal had completely  recovered by 2H hours.

     Skin irritation as  a  result of contact  with acrylonitrile has  also been

observed in humans  (Wilson £t al_.,  19^8).   When the  skin  was  exposed  to acrylo-

nitrile  vapors only,  workers complained of  itching although  there  was   no

observable dermatitis.   When the  skin was  in contact  with the  liquid,  the symp-

toms included irritation, erythema  followed by the formation of blebs,  desquama-

tion, and slow healing.  Dudley and Neal (19^2)  reported on a laboratory accident

in  which  a small  quantity of acrylonitrile  was spilled  on  the  hands of an
                         •
individual.  The symptoms of diffuse erythema occurred by  24 hours, with blisters

on  the finger  tips  and slight  swelling.   Erythematous  and  painful  itching

occurred  by  the third  day.  Ten days after  exposure, the  skin  was cracked  and

peeled and  large  areas  of  tender  new  skin  were  present.  No  cases  of  human

fatality have been directly associated with absorption of acrylonitrile through

the skin.
                                         13-42

-------
13.2.2  Subchronic Toxicity in Non-Human Mammals



     In a preliminary study, Dudley et  al. (1912)  exposed  4  rhesus monkeys and



2 dogs to an atmosphere containing 56 ppm acrylonitrile for M hours a day, 5 days



a week,  over a  1-week  period.   The  rhesus monkeys showed  no overt  signs  of



toxicity. Of the two dogs treated,  one died during the first exposure period and



the second  had  intermittent  paralysis  of the hind legs  following  three of the



exposure periods (the fifth,  thirteenth,  and fourteenth).  In subsequent experi-




ments, 16 rats,  16  guinea pigs, 3 rabbits, and 1 cats were  exposed to 100 ppm



acrylonitrile,  and  16  rats,  16 guinea  pigs,  1 rabbits, I*  cats,  and  2 rhesus



monkeys were exposed to 153 ppn  acrylonitrile.   The exposure  conditions were



similar to those described  above except the period was increased to  8 weeks.  At



the 100  ppm level,   rats and guinea pigs showed  only  slight lethargy.   Three



female rats  gave birth to normal pups during the study.  The rabbits did not gain



weight during the experiment, and the  cats suffered the  most severe symptoms of



vomiting and loss of weight.  At the highest exposure levels, severe toxicity was



noted  in  all  animals,  with  many  dying  prior to  the  completion of  the study;



symptoms were similar to those noted after acute exposure to acrylonitrile.  In



general,  species sensitivity to subchronic acrylonitrile intoxication was simi-



lar to that observed following acute  exposure,  with  the dog  being  the  most



sensitive and rats and guinea pigs the  least  sensitive  to acrylonitrile.



     Dudley e_t  al.  (1912)  examined the  spleen,  kidneys, liver, lungs, heart,



pancreas, lymph  nodes,  stomach, and small and  large intestines of  18 rats,  6



rabbits,  6 cats, 16 guinea pigs, and 1 monkey for pathological changes (there was



no explanation of how the animals were selected for examination).  The spleen of



rats showed  hemosiderosis, which is indicative of blood destruction.  The kidneys



of most animals  showed renal  irritation, with hyaline casts in animals in the two



higher exposure groups (the  exceptions were  1  cat and  1 rabbit exposed at the
                                          13-43

-------
100 ppm level and 1 monkey exposed at the 153 ppm level).   Interstitial  nephri-



tis, although not very extensive, was also noted in many animals (the exception



being the  monkey and rats exposed  at  the 100  ppm level).    Subacute  broncho-



pneumonia was present in at  least some members of all the species studied except



the cat; the pneumonia appeared  as  congestion and  edema in the  alveoli.   Cats



were  the  only species  that  showed  signs of  liver damage,  although  the  exact



lesions were not described.



     Hematological determinations in rats  and rabbits were made  by  Dudley and



coworkers (19^2)  using animals that had been  exposed  to  153 ppm acrylonitrile.



These determinations revealed no changes in red and white blood cell counts or



hemoglobin levels.   At  the  end  of  the  first  week  of exposure, an  increase in



eosinophils  was  noted  using  differential cell counts,  with the maximum for



rabbits and rats being U2$ and 21$  above control,  respectively.   Minami e_t al.



(1973) claimed that  changes  in the partial pressures of oxygen  and carbon dioxide



in the blood occurred,  along with increases in hemoglobin and hematocrit values



following exposure of 8 rabbits once a week for 8 hours to 20 ppm acrylonitrile



for a period of  8 weeks.  Due  to the small differences observed  and  the  rela-



tively large standard deviations, however, the statistical significance of these



conclusions is in doubt.



     Some cumulative effects were observed in rats that received 50 mg/kg acrylo-



nitrile by daily intraperitoneal  injection for a 3-week period (Knoblock e^ aj..,



1971)•  The animals  lost body weight, but the liver, kidneys, and heart increased



in weight.  Signs of functional disorders of both liver and kidneys were noted,



and microscopic examination showed slight damage in neuronal cells of the brain



cortex, brain stem,  and  in the parenchymal cells of the liver.  The only hemato-



logical change noted was leukocytosis.
                                          13-44

-------
     In an abstract, Szabo  and coworkers (1976a)  reported subacute  effects  of



acrylonitrile in female Sprague-Dawley rats (200 g) exposed to  the  compound  at



2000 ppm in  the  drinking water for 21 days.  At this  concentration of acrylo-



nitrile, water intake was reduced  in  the experimental animals  to one-third  of



that of the control animals.  Limited  pathological  studies indicated atrophy  of



the zona fasciculata  and increases in  the  zona glomerulosa  of the  adrenals.



Elevated plasma Na* concentration  of  9 mEq/1 (K* was  not changed)  and reduced



corticosterone concentrations of 2H ug (the  denominator was not  indicated)  were



noted.  The authors concluded that  acrylonitrile affected the  mineral-corticoid



and glucocorticoid cells of the adrenal cortex.  Using the same strain of rats



(Sprague-Dawley),  Szabo £it al.  (1977b) demonstrated a dose-dependent increase in



glutathione levels following a 21-day exposure to water  containing  20,  100,  or



500 ppm acrylonitrile.   It  was  suggested  that  this  increase  in  glutathione was



due to a rebound effect following the utilization of glutathione in the detoxifi-



cation of acrylonitrile.



     Because of the limited number  of  studies on subchronic exposure to acrylo-



nitrile, it is difficult to  formulate a unified pattern of toxic  action, both  in



terms of the mechanism of toxicity  and the target organs.  Although  a number  of



biological  parameters have  been reported  to change  following  subacute exposure



to acrylonitrile,  it is not clear how  these  effects relate to  the primary toxic



action of acrylonitrile.



13«2.3  Chronic Toxicity in Non-Human  Mammals



     Two chronic toxicity studies have been  conducted  by the Dow Chemical Toxi-



cology Research Laboratory for the  Chemical  Manufacturers Association.  In both



studies, the animals were exposed by ingesting drinking water containing acrylo-



nitrile.  The animals were frequently  evaluated during the studies for food and



water consumption, weight gain, and blood and urine chemistry.   On  death or  at
                                         13-45

-------
necropsy at the end of the study,  autopsies  were performed in order  to  detect



gross or microscopic pathological conditions.



     In the first  study (Quast et al., 1975), beagle dogs (four dogs of each sex



and exposure level) were  exposed to 300,  200, and  100  ppm acrylonitrile for a



6-month period  starting when the dogs were 8 months old.  By using water consump-



tion data, the mean doses  of acrylonitrile ingested at the three exposure levels



were for male dogs 17 + 4, 16  + 2, and 10  +  1  mg/kg/day,  respectively,  and for



female dogs 18 + 5, 17 +. 2,  and 8+1 mg/kg/day,  respectively.  The two highest



exposure levels (300 and 200 ppm) were highly toxic to the  dogs; 3 of the *4 males



and 2 of the 4 females at the  300 ppm level and 2 of the H males and 3 of the 4



females at  the 200 ppm  levels  either  died spontaneously  or were  euthanized



because of their debilitated state  prior to  three months  (the ingested dose of



acrylonitrile at these exposure levels was nearly identical).   Two male dogs at



the 300 ppm  exposure  level  were removed from  the  study at 57 days  because of



signs of overt  toxicity and were placed on  water containing  no  acrylonitrile.



These dogs appeared to recover  completely (they  regained  their pretest weight)



and, after removal  from acrylonitrile exposure for  approximately 1 month, they



were  returned  to  the  study.    Subsequently,  the  dogs again showed  signs  of



toxicity and  died within 24 and 15 days.   All  dogs that died showed the following



progression of signs:   terminal depression, lethargy,  weakness, emaciation, and



respiratory distress.  Dogs that ingested water containing 100 ppm acrylonitrile



and dogs that survived treatment at the higher doses appeared normal during the



period of study.



     The presence  of  acrylonitrile in the drinking water  caused a significant



decrease  in  water  and food consumption  in  both male  and  female dogs  at the



300 ppm level.   At the other exposure levels, male dogs showed normal water and



food consumption,  while  female  dogs had sporadically lower water consumption and
                                          13-46

-------
consistently lower food consumption.  A supplemental study,  which used 8 female



dogs that ingested water containing  100 ppm acrylonitrile, failed to demonstrate



any decrease in either food or water consumption.  With regard to general weight



gains (only  surviving animals were considered  since emaciation was  a symptom



ensued by morbidity),  all male dogs gained weight, although less than that gained



by control animals, while female dogs  lost weight, except for the animals at the



100 ppm exposure levels.  It was not  clear whether  the unpalatability of water



containing acrylonitrile may have affected these results.



     Chemical studies on blood and urine samples revealed no significant abnor-



malities that were associated directly with acrylonitrile treatment.  Some hema-



tologic changes, however, did occur at the higher exposure levels.  These changes



were  noted  and  were  consistent with bronchopneumonia, which  was  chronically



present (and the reason for death)  in the high  dosed  animals.   Analysis of the



liver and  kidneys for  non-protein  sulfhydryl content  revealed  no  statistical



differences in the animals at the 100  ppm exposure levels and no apparent effect



at other doses, although these data could not be analyzed statistically because



of the small group size.  There was also no difference in the immunologic status



of treated animals as  assessed by percent of albumin, alpha 1, alpha 2, beta, and



gamma globulin in the serum.



     On autopsy,  some  minor  or indirectly-caused  pathological findings  were



noted.  In male dogs at the 100 and 200 ppm exposure levels, there was an increase



in the relative kidney to body weight  ratio (there were no changes in the single



male  dog  exposed to  300  ppm acrylonitrile).   The  kidney  appeared  normal  on



histological examination and the weight gain may have been compensatory.  There



was also a decrease in the relative brain weight of the 2 surviving male dogs at



the 200 ppm exposure level  and an increase in the relative brain weight of the 2



surviving female dogs  at the 300 ppm exposure  level.   No mention was  made  of
                                          13-47

-------
results from histological  examination of these brains.  The lungs of dogs at the



higher  exposure  levels  showed typical  pathologies of  foreign body  broncho-


pneumonia, the cause of which appeared to be aspirated food particles.  The only



treatment-related  pathology  was  dilation,  thinning  of  the  walls, and  focal


erosions or ulcerations of the esophagus in male and female dogs in the 300 and


200 ppm exposure groups.   In some animals, although not consistently,  there was a



thickening of the epithelium lining of the dorsal surface of the tongue.  It was



believed that  acrylonitrile may  have  caused irritation of  the  membrane of the


throat, resulting in the aspiration of food particles.



     In the second study conducted by Dow Chemical  Company, Quast et. aK (1980a)



exposed Sprague-Dawley rats to drinking water  containing  35,  85,  and  210 ppm


acrylonitrile.  After the  first 21 days the concentrations of acrylonitrile were


changed to 35,  100, and 300 ppm. The calculated  amount of acrylonitrile ingested


was  3.12,  8.53,  and 21.18 mg/kg/day  for  male  rats  and H.36,   10.76,  and



24.97 mg/kg/day for female rats.  By 9 months of  treatment,  the animals at the



two highest dose levels had rough  hair coats and an unthrifty appearance.  There


was a treatment-related early mortality in female rats at all exposure levels and


in male rats at  the  highest exposure levels (Table 13-13).   The differences in


males and  females  may be  partly related to the larger doses  of acrylonitrile


ingested by female rats.


     There was a significant dose-related decrease  in daily water consumption of
                        •

both male  and  female rats exposed  to  acrylonitrile.   This decrease  in  water


consumption was probably due to voluntary restriction of intake as a result  of an


unpalatable taste  imparted by acrylonitrile.    The lower water consumption in


treated  rats  was   the probably  cause   of  the  concomitant  decrease   in  food


consumption and body weight.  Chronic renal  disease, which was common  in older of


this strain, occurred less frequently in  treated rats probably as a  result of the
                                        13-48

-------
       Table 13-13.   Cumulative Mortality of Male and Female Rats Maintained for Two Tears on Drinking Water Containing Acrylonltrlle  (Quast  e_t al.. 1980a)
Percent Dead
Days on Test
0-30
31-60
61-90
91-120
121-150
151-180
181-210
211-240
211-270
271-300
301-330
331-360
361-390
391-120
M21-U50
151-180
181-510
511-510
5H-570
571-600
601-630
631-660
661-690
691-720
721-715
Total Number
of Rats
Control
Male Female
0
0
0
0
0
1.3
1.3
1-3
2.5
3-8
7.5
8.8
8.8
10.0
13.8
18.8
26.3
36.3
«1.3
50.0
60.0
67.5
83-8
87.5
91.3
80

0
0
0
0
0
0
1.3
1.3
1.3
1.3
1.3
1.3
1.3
3-8
3.8
7.5
11.3
13.8
22.5
?7.5
12.5
16.3
56-3
67.5
75.0
80

35 ppra
Male Female
0
0
0
0
0
0
2.1
2.1
2.1
2.1
4.3
".3
6.1
10.6
1U.9
17.0
27.7
29.8
12.6
18.9
57.1
70.2
83.0
87.2
89.1
17

0
0
0
0
0
0
0
0
0
0
2.1
2.1
6.3
6.3
10.1
11.6
20.8
25.0
11.7°
50.0
56.3.
68.8"
79. 2a
87. 5a
91.7
18

100 ppm
Male Female
0
0
0
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
0.2
10.4
11.6
22.9
33-3
37.5
17.9
51.2
68.8
72.9
83.3
89.6
18

0
0
0
0
0
0
2.1
2.1
1.2
6.3
6.3
6.3
10.1s
12.5
12.5
22.9s
27. 1B
35.1"
56.3"
70.8s
79.2s
87. 5a
91.7s
95.8s
97.9s
IB

300 ppa
Male Female
0
0
0
0
0
0
0
0
4.2
4.2
4.2
8.3
11.6
16.7
22.9
31.3
15.8s
60.1s
66.7S
70. 8a
77.1s
83.3s
100.0
100.0S
100.0S
48

0
0
0
0
2.1
2.1
2.1
2.1
6.3
8.3.
18.8s
29.2s
33.3s
41.7s
50.0°
64.6s
72.9S
77.1s
87.5S
93.8s
95.8s
95. 8a
100. 0B
100.0s
100.0°
48

u>
 I
               aSlgnlflcantly different  from controls by Fisher's Eitact Probability  Test,  p  <  0.05.

-------
lower consumption  of  food and water.  With  this  decrease  in renal disease

there was a  decrease  in both gross and microscopic pathologies and clinical

findings associated with  renal failure.  Decreased food and water consumption


resulted in  emaciation  of the animals, which caused  other  clinical signs and

pathologies to change.

     The majority of clinical chemistry and pathological changes that occurred

at a  statistically  significant  level (P < 0.05) were attributed to secondary

effects of treatment and were not considered to be caused by a toxic action of

acrylonitrile.  A  summary of both  the  clinical chemistry  and pathological

changes observed is presented in Table 13-14.   Although it was proposed that

these changes were, for the  most part, a result of restricted food and water

consumption, it was impossible to state this unequivocally  because of the lack

of pair-fed  controls.   There were no severe or irreversible non-tumorigenic

pathological  findings.  The  only  non-tumorigenic  changes that were  considered

to be treatment-related were  of a possible pre-neoplastic  type; these  include

hyperplasia and hyperkeratosis of the squamous  epithelium  of the nonglandular

portion of  the  stomach  (100  and 300 ppm males  and all treatment levels in

females), proliferation of the glial cells in the brain  (300 ppm males  and 35

and  100 ppm  females),  and mammary gland hyperplasia  (300  ppm  females).  To

fully understand the chronic  toxicity of acrylonitrile, studies must be performed

with pair-fed controls that are in the same nutritional state as the experimental

animals, and evaluation of clinical signs must be made in the absence of gross

neoplasisas.

13.2.3.A  Summary and Conclusions

     Acrylonitrile intoxication  in  humans  results in  irritation of the eyes

and nose, weakness, labored breathing, dizziness, impaired judgement, cyanosis,
                                                               3
nausea, and  convulsions.   The TLV of acrylonitrile  is 45  mg/m  for humans.
                                         13-50

-------
Acrylonitrile also causes severe burns to the skin.  In experimental animals,




there is  considerable species variation in  susceptibility  to acrylonitrile




intoxication; the guinea pig  is  the most resistant  and  the dog is the most




sensitive.  In animals, effects  of intoxication include respiratory changes,




cyanosis,  convulsions,  and  death.   In rats, the  LD,.n  for acrylonitrile is




between 80 and 113 mg/kg  (Knoblock et al.,  1971;  Smyth et al., 1969).   There




is some evidence  that acrylonitrile produces abnormal function  of both the




peripheral and central nervous systems and that acrylonitrile causes damage to •




the adrenals.  With   subchronic exposure of  animals  to acrylonitrile,  some




signs of  functional  disorders  of the liver  and kidney are observed.  Chronic




exposure of dogs  and rats  results  in unthrifty appearance, weight  loss, and




early death.   Some of these  signs may be related to low food and water consumption




resulting from the unpleasant taste of acrylonitrile in the water.   Pathological




changes in the rats  believed to be treatment-related included  hyperplasia and




hyperkeratosis of the squamous epithelium  of the  nonglandular portion of the




stomach, proliferation of  glial cells in the brain, and mammary gland hyperplasia




in females.




13.3  TERATOGENICITY  AND REPRODUCTIVE TOXICITY




     The teratogenicity of acrylonitrile has been  investigated using Sprague-




Dawley rats  exposed to the compound by inhalation  or by gavage on days 6 to 15 of




gestation (the results of both the inhalation and  gavage  treatments have been




reported by  Murray et al.,  1978,  while a detailed  report of the results of the
                                       13-51

-------
               Table 13-1U.  Significant Changes Considered to be Secondary to Ingestion of Acrylonitrile
                             (Quast et al.,  1980a)
  Type of Analyses
          Significant Findings
     Suspected Cause of Findings
Hematology



Urinalysis

Clinical Chemistry




Pathology
Decreased white blood cell count
     Decreased packed cell volume
     Decreased hemoglobin

Increased urine specific gravity

     Increased blood urea nitrogen

     Decreased serum glutamic pyruvic
       transaminase

     Emaciation
     Decreased adipose tissue

     Dental malocclusion

     Increased blood and bile-like fluid


     Decreased chronic renal disease


     Increased hepatocellular atrophy

     Decreased signs of altered nervous
       system function

     Changes in organ weight
Blood loss from ulcerated tumors



Restricted water consumption

Restricted water and food consumption

Within normal limits


Restricted water and food consumption


Ear canal tumors deviating the lower Jaw

Restricted food consumption and ulcerating
  tumors

Restricted food and water consumption and
  early death

Poor nutritional state

Decrease in chronic renal disease


Poor nutrition

-------
gavage treatment  was presented by  Murray e_t  al.,  1976).   Acrylonitrile  was




administered at  doses  of 10, 25,  and 65 mg/kg/day per  os or by  exposing  the




animals to an atmosphere of 10 or 80 ppm acrylonitrile for 6 hours  per day.   It




was estimated by measuring blood levels of acrylonitrile and its metabolites that




an exposure  to 80 ppm  acrylonitrile for 6  hours  was equivalent  to  a  single




23 mg/kg dose of acrylonitrile.  On day 21 of gestation, the dams were sacrificed




and examined for implants and fetal abnormalities.




     Rats  exposed to  65  mg/kg/day  showed  moderate  signs of  toxicity  with




decreased body weight, thickening of  the  non-glandular portion of  the stomach,




and  increased  liver weight.   The  only  adverse effect  on animals  exposed to




25 mg/kg/day was a slight thickening of the non-glandular portion of the stomach,




while animals exposed to the lowest  dose or animals exposed by inhalation showed




no major  treatment-related  toxic effects.   In all cases,  except  the animals



treated with 10 mg/kg/day,  food consumption initially  declined  during the first




3-day interval following treatment, but returned to normal during the two subse-




quent 3-day  intervals.   Neither the  stress  on  the  dams  due to  the mild toxic




effects of acrylonitrile nor  any  direct  toxic effects of  acrylonitrile  on  the




fetus significantly  altered the number of  litters,  implants per  dam,  or live




fetuses per litter.   A slight but significant decrease in fetal body weight  and




crown-rump length was  noted in the 65 mg/kg/day group,  but not in  any  of  the




other groups.




     Fetal malformations were observed primarily in  the offspring  of animals




that received acrylonitrile  by  gavage (Table 13-15).   Following inhalation of




acrylonitrile, there  was no  significant single major malformation observed; how-



ever, when malformations were considered  collectively, there was a significant




(P = 0.06) but slight increase in the 80  ppm exposure group.   At the high dose




level in  the  gavage-treated animals, there was a significant increase in acaudate
                                        13-53

-------
                                        Table 13-15.  Incidence or Petal Malformations Among Litters of Rats Given Acrylonltrlle
                                                      (Murray et a_l.,  1978)
Type of
Malformation

External and skeletal malformations
Visceral malformations

External ma 1 Populations
Short tall
Short trunk
Iraperforate anus
Omphalocele
Visceral malformations
Right-sided aortic arch
Missing kidney, unilateral
Anteriorly-displaced ovaries
Skeletal malformations
Missing vertebrae (associated with
short tall)"
Missing two vertebrae and a pair
of ribs
Healvertebra
Total malformed
Acrylonltrlle
0

1113/38
151/38


1(1)
0(0)
0(0)
0(0)

0(0)
1(1)
0(0)


1(1)

7(1)
0(0)
8(2)
given by gavage (mg/kg body weight/day) Acrylonltrlle
10

388/35
135/35


0(0)
0(0)
0(0)
0(0)

0(0)
0(0)
0(0)


0(0)

0(0)
0(0)
0(0)
25
No.
312/29
111/29
No.

2(2)
0(0)
0(0)
0(0)

1(1)
0(0)
1(1)


2(2)

7(2)
0(0)
10(1)
65
fetuses/No.
212/17
71/17
of fetuses

8(6)a
3(3)a
2(2)
0(0)

1(1)
1(1)
1(1)


8(6)a

0(0)
0(0)
8(6)a
0
litters examined
121/33
110/33
(litters) affected

0(0)
0(0)
0(0)
0(0)

0(0)
0(0)
0(0)


0(0)

8(1)
0(0)
8(1)
given by inhalation (ppn!
10

111/36
1H8/36


0(0)
0(0)
0(0)
1(1)

0(0)
0(0)
0(0)


0(0)

2(1)
0(0)
3(2)
80

106/35
136/35


2(2)
1(1)
0(0)
1(1)

0(0)
0(0)
1(1)


2(2)

7(2)
1(1)
11(6)°
00
 I
in
              Values differ significantly from the control value (P < 0.05)  using the Wllcoxon test as modified by Baseman and Hoel (1971).

              This ralfomatlon was observed only In fetuses with short tall.

             cValue differs significantly from the control value (P - 0.06)  using the Wllcoxon test as modified by Haseman and Hoel (1971).

-------
or short-tailed  fetuses.   The majority of other  abnormalities  including short



trunk,  anteriorly  displaced ovaries,  missing  ribs,  and imperforate  anus were



observed in the acaudate and short-tailed fetuses whether these animals were from



the  control  or experimental group.   The only anomaly that  occurred  solely in



treated animals was a right-sided aortic arch,  which appeared in one fetus from



the 25 mg/kg/day group and one fetus from the 65 mg/kg/day group.  This anomaly



had never been noted in over 1000 litters examined by this laboratory.



     The  authors  concluded  that  the  malformations  that  occurred  in  the



65 mg/kg/day group were caused by effects of acrylonitrile directly on the fetus,



and  not  by effects on the  dam.   This conclusion was based  on  historical data



indicating that mild stress, as caused by acrylonitrile treatment,  did not lead



to these  types of malformations  and  on the lack of  correlation between gross



signs of  toxicity  in the  dams  (e.g., poor weight  gain,  and decrease  in food



consumption)  and litters  with  malformed fetuses.   There also appeared  to be a



dose response relationship because the 25 mg/kg/day group and the 80 ppm exposure



group also  showed increased  incidences of malformed  fetuses.   This was only



suggestive, however, since  these  increases were  not  statistically  significant



when compared with control animals.



     The  teratogenicity  of acrylonitrile  was  also  evaluated in chick  embryos



(Kankaanpaa e_t al^, 1979).  Acrylonitrile was injected into either the air space



or yolk sac of 3-day old White Leghorn eggs.  The  acrylonitrile was dissolved in



saline at concentrations of 0.001, 0.01, 0.1, 1.0, and 10  umol,  and a total of



25 ul was used for injection.   The  only observable  adverse effect of treatment



was embryo toxicity (dead embryos and empty eggs)  of approximately 100$ and Wf>



at the two highest concentrations when acrylonitrile had been injected into the



air space.  When acrylonitrile was injected into the yolk sac,  only the 10 uraol
                                        13-55

-------
dose was toxic (60$) .   There was no evidence of a teratogenic effect of acrylo-




nitrile treatment in this study.




     Beliles et al. (1980)  performed an extensive three-generation reproductive




study on Charles River  rats  that ingested either 100 or 500 ppm acrylonitrile in




drinking water.  The  rats  were initially exposed  following  weaning,  and mated




100 days after the  start  of  the experiment.  Two litters were  produced from each




female and members of the second litter were used to produce the next generation




(rats from the first litter were used  only  when  an insufficient  number of pups




was available  from the  second  litter).  The  parents of the  first generation




showed some adverse effects of treatment in the  500  ppm group,  with  food  and




water consumption  and  body  weights significantly  lower than those  of control




animals (other generations were not monitored for these parameters).  Reproduc-




tive toxicity was observed  in the  two  matings  of the  first generation, with an




increased number of deaths during the lactation  period  among pups of animals that




had been treated at the  500 ppm  level.  These  deaths  may have been a result of




acrylonitrile's effect on the  dams,  since  pups fostered  by  untreated  dams  had




normal survival.  In the other  generations,  reproductive capacity and pup sur-




vival were within the  normal range.  The only adverse  effect  observed in pups




that survived  treatment  was a  decrease in body weight  in  the 500  ppm group




(Table 13-16).




     Poor weight gain in the pups may have been caused by poor lactation in the




dams due to the decreased water consumption.  Ten weanlings of each sex from the




control and high dose groups of the F3b litter were sacrificed and examined for




histopathological changes.   No  adverse  findings were noted in the tissues listed




in Table 13-17.  It appeared that acrylonitrile had little direct effect on the




development of the embryo and pup up to the time of weaning.
                                         13-56

-------
13.3.1  Summary and Conclusion




     Acrylonitrile  adversely  affected  pup  survival  following  exposure  of




pregnant rats  and, in  one  study, produced  teratogenic  events.   In  a  three-




generation study  in which  rats  were exposed to 500 ppm acrylonitrile  in the




drinking water, there was reduced pup survival in the first generation (Beliles




et al.,  1980).  This  was a  maternal effect  inasmuch  as  fostering the pups on




untreated dams eliminated the poor survival.  Reproductive  capacity was  unchanged




in  the  other  generations,  and  the offspring  showed no  adverse  effects  on




development.   Similarly, rats exposed by inhalation to 40 or 80 ppm of acrylo-




nitrile  for  6 hours a  day  on days  6  to 15 of gestation  had  no  satistically




significant  changes  in reproductive success  or fetal development  (Murray et




al., 1978).  Only the pups of rats administered acrylonitrile per pjs (65 mg/kg)




for days 6 to 15 of gestation had an increase in malformations (Murray et al.,




1976).  This  increase was in total malformations,  with no statistically  signi-




ficant increase occurring in  any single malformation.  It was  concluded that




these fetal  abnormalities were  the result of acrylonitrile and not the  result




of  toxicity  in  the dams.  Although several studies  have been  conducted  to




evaluate the  ability of  acrylonitrile  to  cause adverse teratogenic,  embryo-




toxic, and reproductive  effects,  the limitations of the  available data  do not




allow for  a  full  assessment  of  these  effects.   However,  there are data that




indicate that  acrylonitrile has  the  potential to cause fetal malformations in




rats, in a non dose-related  manner.
                                        13-57

-------
      Table 13-16.   Pup Weight on Days H  and  21  of Lactation
                    (Beliles et al.,  1980)
Generation
Dose Level
(ppm)
Mean Pup
Day H
Weight (g)
Day 21

Fla


Fib


F2a


F2b


F3a


F3b


0
100
500
0
100
500
0
100
500
0
100
500
0
100
500
0
100
500
11
10
9a
10
9
10
11
10
9a
11
10
9
10
9a
8a
10
10
8a
i»2
J40
28a
38
35
3Ua
39
39a
30a
51
16
30a
13
43a
30a
U9
46
32a
ap < 0.05 using Students t-test.
                                    13-58

-------
Table 13-17.  Tissues Examined for Histopathologic  Changes  in
              the F3b Litter (Beliles et al.,  1980)

Adrenal gland
Aorta
Bladder, urinary
Bone
Bone marrow
Cerebellum
Cerebrum
Colon
Esophagus
Eye
Heart
Ileum
Jejunum
Kidney
Liver
Lung
Lymph node
Mammary gland
Muscle, skeletal
Nerve, sciatic
Ovary
Pancreas
Parathyroid
Pituitary gland
Prostate
Salivary gland
Seminal vesicle
Skin
Spinal cord
Spleen
Stomach
Testis
Thymus
Thyroid
Tongue
Trachea
                               13-59

-------
13.4  MUTAGENICITY






     The objective of this mutagenicity evaluation  is  to  determine  whether  or



not acrylonitrile has the potential  to cause mutations in humans.   This



qualitative assessment is based on  information  derived from several  tests



measuring different types of genetic damage (e.g.,  gene mutation, chromosome



damage, DNA strand breakage).  These tests  include  bacterial,  Drosophila,



plant, cultured mammalian cell, and whole rodent  assays.   The  results  of these



studies on acrylonitrile are discussed below and  each  study is summarized  in



Table 13-18.







13.4.1  gene mutation studies



bacteria



    Acrylonitrile has been shown to cause point mutations in bacteria.



Venitt et al. (1977) evaluated the  mutagenicity of  acrylonitrile  (purity >



99%, impurities not reported) in a  plate incorporation assay with the



tryptophan-dependent Escherichia col i  strains WP2 (repair proficient),  WP2



uvrA (lacks excision repair), WP2 uvrApolA  (lacks excision repair and  DNA



polymerase I), and WP2 lexA (deficient in an error-prone  pathway).   In  the



absence of rat liver enzyme activation, acrylonitrile  produced a weak



dose-related mutagenic response in  all strains  except  WP2 lexA when examined



at concentrations of 0, 75, and 150 umoles/plate.  It  should be pointed out



that acrylonitrile is volatile and  in  this  test no  precautions to prevent



evaporation were reported.  If evaporation  occurred under the  standard  test



conditions, the weak responses observed would not reflect the  mutagenic



potency of acrylonitrile in this system.  These reported  weak  responses also



are difficult to interpret because  no positive  control data were given  and



thus, it is unclear how well the test system was  responding to mutagen
                                         13-60

-------
treatment.  The responses reported were as follows:   WP2  was  slightly more
sensitive to the mutagenic effect of acrylonitrile than the other  tester
strains; a fourfold increase in the spontaneous level  was produced by 150
umole/plate compared with a threefold increase for WP2 urvA and  a  twofold
increase for WP2 urvApolA.  The authors reported that doses above  150 umole
per plate caused a decline in the mutagenic response,  which was  explained by
increasing toxicity (reduction in bacterial lawn observed).
    Because of the toxic effects and weak mutagenicity of acrylonitrile in the
plate test, Venitt et al. (1977) conducted a fluctuation  test (which is a
sensitive assay for detecting low levels of mutagens)  at  concentrations that
ranged from 4 x 10"4 M to 2 x 10"3 M.  The results of the fluctuation
tests confirmed the mutagenicity detected in WP2, WP2 uvrA (results based on
one experiment and data were not presented), and WP2 urvApolA by the plate
assay.  In contrast to the results derived from the  plate tests, WP2 uvrApolA
was more sensitive to the mutagenic effects of acrylonitrile  than  WP2; the
responses of WP2 and WP2 uvrA were reported as similar.   Acrylonitrile again
tested negative in WP2 lexA.  The lack of a mutagenic effect  in  WP2 lexA
suggests that acrylonitrile may produce mutations by mis-repair  DNA damage,
which is believed to be associated with the generation of DNA strand breaks
(Green and Muriel  1976).  When the resistance factor pKM  101  was transferred
to WP2 (strain designated WP2P) to increase the sensitivity of this strain,
WP2P (pKM 101) was reported to be more sensitive to  the mutagenic  effects of
acrylonitrile than WP2.  A dose-related increase in  mutagenic activity was
observed at a concentration range as low as 4 to 40  umoles per plate in the
plate test.  These results found in £. coli strain WP2P further  support the
notion that acrylonitrile is producing mutations via mis-repair  DNA damage
because Salmonella strains containing the pKM 101 plasmid have been shown to
                                         13-61

-------
be more sensitive to the mutagenic effects of chemical  mutagens which cause
mutations by a error-prone DNA repair-dependent mechanism (McCann et al.
1975).
    Venitt (1977) indicated in his report that activation by an exogenous S-9
system (S-9 mix prepared from livers of Aroclor 1254-induced rats)  did not
enhance the mutagenic activity of acrylonitrile in £.  coli;  thus
acrylonitrile, according to the authors, is primarly detected as a
direct-acting mutagen in these bacteria.  It should be  pointed out  that
because data were not presented in this report, it is  uncertain whether Venitt
included an activation-dependent mutagen in his study  to demonstrate that the
S-9 fraction used was functional.  Thus, his negative  results with  the
activation system may be an experimental error.
     Venitt et al. (1977) also studied acrylonitrile in Salmonella  typhimurium
and reported that acrylonitrile was not detected as mutagenic for tester
strains his D3052, TA 1535, TA 100, TA 1538, and TA 98  using either plate
incorporation assays or fluctuation tests.  Although mutagenic activity was
reported to be detected in strain his G46 using a fluctuation test, the
results were reported as "erratic and statistically nonsignificant".  The
authors did not present the protocol; thus it is not known  if mammalian liver
activation was employed in the Salmonella assays or if  precautions  to prevent
evaporation were taken.  In addition, they did not present  data to  support
their negative conclusions for Salmonella.  Therefore,  Venitt's conclusion of
nonmutagenicity in Salmonella cannot be evaluated.
    McMahon et al. (1979) reported that acrylonitrile  (purity not reported)
was mutagenic in both Escherichia coli and Salmonella  typhimurium strains when
screened in a qualitative gradient plate assay, but they do  not indicate  if
acrylonitrile required metabolic activation for activity, the concentration(s)
                                         13-62

-------
at which activity was observed, and which bacterial  strains were reverted.

Therefore, the conclusions of the authors cannot be  evaluated.

    Milvey and Wolff (1977) reported positive results for acrylonitrile (99%

purity, impurities not reported) in a modified Salmonella/microsome assay when

tested as a vapor.  The authors emphasized that the  mutagenic effects of the

acrylonitrile in Salmonella were observed only in the presence  of metabolic

activation (S-9 mix prepared from livers of Aroclor  1254-induced male

Swiss-Webster mice).  Data were not shown in the absence  of S-9 activation.

There are major deficiencies in the reporting of the data that  make

interpretation of this study difficult.   In addition, Venitt (1978) pointed

out errors in calculation found in this  report.  The studies by DeMeester and

coworkers (discussed below), however, provide more convincing evidence for the

mutagenicity of acrylonitrile in Salmonella typhimurium.

    DeMeester et al. (1978) and other investigators  (Venitt 1977) have noted

that acrylonitrile is not detected as mutagenic in Salmonella when tested in a

"classical" plate incorporation assay.t   DeMeester et al. (1978) were able to

detect mutagenic activity in Salmonella  when acrylonitrile exposure was as a

vapor in a desiccator.   Unlike the results for Escherichia coli, DeMeester

et al. (1978) found that the mutagenicity of acrylonitrile in Salmonella was

detected only in the presence of an in vitro liver microsome-activation

system.  This difference is not necessarily contradictory because these are

two different bacterial  test systems, which may have different  metabolizing

capabilities and sensitivities.  Also, as stated earlier, Venitt does not
    tin the plate incorporation test,  the chemical  is  mixed  with  the  bacteria,
S-9 mix, and then melted top agar is immediately added.   This  mixture is
poured on petri  plates,  which are then incubated 2  to  3  days at 37°C.
                                         13-63

-------
provide data in his report to support his  conclusion  that  the  mutagenic
activity of acrylonitrile 1s not enhanced  by a S-9 activation  system.
    In the study by DeMeester et al.  (1978), 0.15 liters of  gaseous
acrylonitrile (99% purity, impurities not  reported) was injected  into  a
desiccator and the cells were exposed for  one hour.   Using a gas
chromatograph, they measured the concentration of acrylonitrile to be  about
0.2% in the atmosphere.  The concentration of acrylonitrile  on the test  plates
was also measured by freezing the plate agar and  analyzing the concentration
of acrylonitrile by gas chromatography. This was found to be  about  200
ug/plate.   In Salmonella, the toxicity of  0.15 1  of gaseous  acrylonitrile  was
reported as weak (cell  survival  80% to 100%) after a  one hour  exposure,  and
the toxicity was reported to be  high  when  the exposure was either to 0.15  1 of
gaseous acrylonitrile for 2 hours or  to 0.24 1 for one hour  (the  method  used
to measure toxicity was not described). Reversion to histidine prototrophy
was reported for the base-pair substitution-sensitive strains  TA  1538, TA
1535,  TA 1530, TA 100,  TA 1950,  and for the frameshift-sensitive  strains TA
1978,  TA 98, in the presence of  microsomal  activation (Aroclor 1254-induced
rats,  300 ul S-9/ml mix).  Acrylonitrile had the  most pronounced  effect  on the
base-pair substitution-sensitive strains TA 1530, TA  1535, and TA 1950,  with
TA 1530 exhibiting the  highest number of revertants over the spontaneous
number (up to approximately a 15-fold increase over the negative  control).
Activity in base-pair substitution sensitive strains  would be  consistent with
the properties of an alkylating  agent.  The frameshift-sensitive  strains TA 98
and TA 1978 and the base-pair substitution-sensitive  strain  TA 100 were  only
weakly reverted (approximately a twofold or less  increase  over spontaneous
level).  Negative results were found  with  the strains TA  1975, TA 1532,  TA
1537 and his G46.  A fluctuation test was  conducted to confirm the sensitivity
                                         13-64

-------
of TA 1530 to the mutagenic effects of acrylonitrile.   Mutagenic  activity was
detected at a concentration as low as 2.5 ug/ml  (P  < 0.001).   Although
DeMeester and coworkers only evaluated one dose  and thus  do not demonstrate  a
dose-related mutagenic response, their studies do indicate  a  clear  increase  in
the number of revertants over the background number which appears to  be  a
reproducible increase.
    When the amount of acrylonitrile needed to induce  a mutagenic response and
the magnitude of the response are considered, acrylonitrile appears to be a
weak/moderate mutagen in Salmonella.  However, it is difficult to
unequivocally classify the mutagenic activity of acrylonitrile from DeMeester
et al. (1978) studies because no concurrent positive control  chemicals were
evaluated and only one dose was tested.    Because of this,  Monsanto  (Dr. L.
Kier 1982) conducted a modified liquid suspension assay using Salmonella
exposed to acrylonitrile and several standard mutagens
(4-nitroquinoline-N-oxide, 2-acetylaminofluorene, 2-aminoanthracene,
benzo[a]pyrene) to determine the relative activity  of  acrylonitrile to these
known mutagens.  In this test, the bacteria, chemical, and  S-9 mix  (Aroclor
1254-induced rat liver) were mixed and incubated for two  hours at 30°C.  After
a two hour exposure to concentrations of acrylonitrile that ranged  from  0.02
mg/ml to 20 mg/ml, the cells were washed, mixed  with top  agar, plated in petri
dishes and incubate at 37°C for 2 days before scoring. Viability was not
quantitatively determined in these experiments (only the  observation  of
bacterial lawn growth or presence of small colonies).   It was concluded  that
acrylonitrile was weakly mutagenic in tester strain TA 100  with S-9 activation
but not towards TA 98 and TA 1535.  No increased responses  were observed
without S-9 mix.  In TA 100, average revertant counts  at  the highest  dose  (6
mg/ml) tested was about 1.5-fold over the solvent control.   Therefore, under
                                          13-65

-------
the liquid suspension treatment conditions used by Monsanto,  acrylonitrile
caused only a marginal  Increase in the number of revertants.
    The bacterial  studies by DeMeester et al. (1978)  and Monsanto (1982)
indicate that acrylonitrile was detected as mutagenic only when metabolized  j_n
vitro by using homogenates of mammalian liver.  In addition to acrylonitrile
being activated by in vitro S-9 systems, Lambotte-Vandepaer et al.  (1980)
examined the ability of the whole mammal to metabolize acrylonitrile.  These
authors reported that urine (0.1 ml per plate) from both rats (adult  male
Wistar) and mice (NMRI) treated with a single intraperitoneal dose  of
acrylonitrile (30 mg/kg; purity 99%, impurities not reported) was mutagenic  in
Salmonella strain TA 1530 (9-fold increase in the spontaneous level  of
revertants when exposed to mouse urine and 13-fold increase for exposure to
rat urine).  S-9 mix (from Arcolor-1254 treated mice) added to the  test plates
caused a fivefold reduction in the mutagenicity of urine from
acrylonitrile-treated rats.  Only a slight reduction (1.3-fold) was seen with
the urine of acrylonitrile-treated mice.  When animals were pretreated with
phenobarbital and exposed to acrylonitrile, the urine from rats produced no
detectable mutagenicity and the mutagenicity of urine from mice was reduced.
The addition of B-glucuronidase to the test plates only slightly enhanced
(approximately 1.4-fold) the mutagenic effects of urine collected from
acrylonitrile-treated rats and mice (B-glucuronidase is added to cleave
possible conjugates).  However,  3-glucuronidase had a marked effect  on the
mutagenicity of urine from rats treated with phenobarbital and exposed to
acrylonitrile (an eightfold increase in mutagenic activity observed).  The
authors indicated that XAD-2 resin concentrates of urine from all treatment
groups were not detected as mutagenic (data not presented), suggesting that
the mutagenic metabolite(s) was very hydrophilic.  Therefore, the urine from
                                          13-66

-------
acrylonitrile-exposed rats and mice was mutagenic and glucouronoconjugation
appeared to play a minor role in the deactivation of the  acrylonitrile
derivative(s) (except in the case of phenobarbital/acrylonitrile-treated  rats,
where a greater effect was found).
    Conner et al. (1979) also examined whole mammal  activation  of
acrylonitrile and found that bile obtained from rats treated  with
acrylonitrile (i.p. injection of 45 mg/kg body weight)  was  not  mutagenic  in
Salmonella.  However, these negative findings cannot be interpreted  because
data were not presented in the report and several  known mutagens  (e.g.,
dimethylnitrosamine, 3-methyl-cholanthrene), which require  metabolic
activation, and the direct-acting mutagen methyl  methanesulfonate  were also
found as negative in this study.
    The Salmonella studies indicate that metabolic activation of  acrylonitrile
is necessary to detect mutagenic activity.  A potential  reactive metabolite  of
acrylonitrile is an epoxide.  Monsanto (Dr.  L. Kier, 1982)) evaluated  the
acrylonitrile epoxide, 2,3-epoxypropionitrile, (purity  97.4%) in  a modified
liquid suspension Salmonella assay  and found it to be mutagenic in tester
strains TA 100, TA 1535, and TA 98  in the absence or presence of  Aroclor-1254
induced rat S-9 mix.  This epoxide  was most  active in the base-pair
substitution sensitive strains.  For example, in  TA 100 with  S-9 mix,  a
fourfold increase in revertant counts over background counts  was  seen  at  0.4
mg/ml.  This epoxide does not appear to be a strong-acting  mutagen under  the
liquid suspension treatment conditions used  in this study compared to  the
positive control mutagens (4-nitroquinoline-N-oxide, 2-acetylaminofluorene,
2-aminoanthracene, benzo[a]pyrene).
                                         13-67

-------
Drosophila
    Although acrylonitrile has been shown to cause point mutations in
bacteria, its ability to act at the level of the gene in a eucaryotic organism
has not been adequately examined.  Only one eucaryotic gene mutation study was
available.  This was a study by Benes and Sram (1969) in which the occurrence
of sex-linked recessive lethal mutations in Drosophila  melanogaster was
examined.  A 0.1% solution of acrylonitrile (purity not reported) was injected
into the abdomen of male flies.  (The authors do not give the percent survival
but indicate that chemical concentrations given to male flies were the highest
possible being slightly below a lethal  or sterile dose.)  Although the authors
report the results as insignificant, the data can be considered suggestive of
an increase in the mutation frequency over that of the spontaneous frequency
by a factor of three.  The compound, however, has not been properly evaluated
in this system and retesting is necessary to permit a final  judgment on the
mutagenic activity of acrylonitrile in  Drosophila.  Several  deficiencies in
this report were found:
    1.   A small  sample size of flies was tested (1297 chromosomes-in order to
        detect a doubling in mutation frequency 7000 chromosomes would have to
        be tested, Lee et al. in press).
    2.   No information was given on whether clustering of mutations occurred.
    3.   The purity of the test material  was not described.
    4.   Post-mating days were not given.
    5.   No information was reported for concurrent positive or historical
        negative controls.

13.4.2  chromosomal aberration studies
    The ability  of acrylonitrile to induce chromosomal aberrations in vivo has
                                         13-68

-------
been investigated in both rats and mice.  Acrylonitrile was not detected as



clastogenic in two independent bone marrow assays (Rabello-Gay and Ahmed 1980,



Leonard et al. 1981).  Rabello-Gay and Ahmed (1980)  tested acrylonitrile



(purity reported as 99.5%, impurities not identified)  for chromosomal  effects



in mice (Swiss albino) and rats (Sprague-Dawley).  No  significant increase (at



a confidence level of 0.05) in the incidence of chromosomal  aberrations (gaps,



breaks, fragments, Robertsonian translocations) were found in mouse bone



marrow cells.   At least 1200 cells were examined per treatment group and male



mice were given acrylonitrile by gavage for 4,  15,  and 30 days each at doses



of 7, 14, and  21 mg/kg/day.  In a second experiment  mice received 10,  15, and



20 mg/kg/day for the same time intervals.  The  toxicity of 7, 14, and  20



mg/kg/day given orally was reported to correspond to 0.25, 0.50,  and 0.75 of



the LDsQ, respectively.  Mortality for the acrylonitrile-treated  animals was



reported (3 of the 72 animals died).  The chemical  also was found to be



negative when  rats were exposed orally to 16 doses  at  40 mg/kg/day  (reported



to represent one-half of the LO^g).  Because current evidence (see Abreu and



Ahmed 1980, and Guengerich et al. 1981) indicates that acrylonitrile can be



biotransformed to cyanide these authors also evaluated potassium  cyanide (KCN)



and found no increase of chromosomal aberrations in  the bone marrow cells of



rats treated with 16 oral  doses of KCN 5 mg/kg/day  (0.5 LD5g).



    Leonard et al. (1981)  evaluated the clastogenicity of acrylonitrile



(purity not given) in mouse bone marrow cells in vivo.  These authors



conducted both chromosomal  aberration and micronuclei  analyses.  The



percentage  of chromosomal  aberrations or micronuclei  formation did not differ



between acrylonitrile-treated animals and negative  control animals.  Male mice



(NMRI) were injected intraperitoneally with a single acute dose of 20  or 30 mg



acrylonitrile/kg of body weight.  The authors indicated that 30 mg/kg  was the
                                         13-69

-------
maximum dose that allows survival of mice for several  weeks.  For chromosomal
aberration analysis (gaps, breaks, fragments, rearrangements), 200 cells (4
animals and 50 cells/animal) were examined 6, 18, 24,  48, and 72 hours after
each treatment.  The background control frequency was  0.5% cells with
chromosomal anomalies (1 gap/200 cells) and the acrylonitrile-treated animals
did not exceed 1.5% cells with chromosomal anomalies (3 gaps/200 cells).  It
should be noted that gaps are not regarded as true chromosome aberrations at
this time.  For micronucleus analysis, polychromatic erythrocytes were sampled
24, 48, or 72 hours after initial treatment.  The background frequency was
1.8% cells with micronuclei  and acrylonitrile-treated  cells did not exceed
2.5% with micronuclei.  No dose-related effects were seen in these studies.
    Leonard et al. (1981) also conducted a mouse dominant lethal  assay with
acrylonitrile.  This test evaluates the DMA damaging effects of chemicals on
germinal  tissue (i.e., is indicative of the potential  of a chemical  to
contribute to the genetic disease burden).  Male mice  (NMRI) were given a
single injection of 30 mg of acrylonitrile/kg of body  weight followed by a
weekly mating (1 to 5 weeks).  The authors concluded that there was no
difference in the percent of postimplantation loss between the background
control group and the acrylonitrile-treated groups.  The findings in this
test, however, are considered inconclusive because of  the high frequency of
dominant  lethals observed in the negative control  group (33.6%
postimplantation loss and only 41% pregnant females).
   Thiess and Fleig (1978) examined the clastogenic  effects in lymphocytes of
18 workers who had been exposed to acrylonitrile.   The workers had been
exposed for an average of 15.3 years, and were compared with 18 age-matched
workers who had no known exposure to acrylonitrile or  any other compounds that
were suspected of causing chromosomal damage.  Although no information is
                                         13-70

-------
available concerning the exact levels of exposure in the past, atmospheric
monitoring data betweeen 1963 and 1974 indicated a typical  exposure level  of 5
ppm for acrylonitrile, and atmospheric monitoring data between 1975 and 1977
indicated an exposure level  on an average of 1.5 ppm.   The  authors point out
that individual workers may have been exposed to higher concentrations during
specific operations.  Also,  the workers could have been exposed on the job to
styrene, butadiene, ethylbenzene, butylacrylate, and diphyl  (a mixture of
diphenyl and diphenyl oxide  ether).   For each subject, 100  metaphases were
examined.  The type of chromosomal  aberrations scored  was not indicated in the
report except for gaps and isogaps.   One hundred cells were analyzed per
person.  Chromosomal aberrations in  the exposed group  were  1.8 +_ 1.3%
(excluding gaps) and 5.5 +_ 2.5% (including gaps and isogaps), while chromosome
aberrations in the control group were 2.0 _+ 1.6% (excluding gaps)  and 5.1 +_
2.4% (including gaps and isogaps).   Therefore, apparent chromosomal  damage was
not detected in these workers.
    Loveless (1951) and Kihlman (1961) reported negative findings  for
chromosomal  effects in plants (Vivia faba).  However,  data  were not reported
in these articles and details of the protocol were not presented.   Thus, the
negative conclusions of the  authors  cannot be evaluated.

13.4.3  other tests indicative of genetic damage
    Other tests have been conducted  which do not measure mutation  pe_f_ If. ^ut
may indicate that acrylonitrile has  the potential to cause  mutations.
    Chemical-adduct formation in DNA is a critical event in certain types of
mutagenesis.  DNA binding studies have been conducted  using radiolabeled
acrylonitrile to determine its ability to react with DNA.  Guengerich et al.
                                          13-71

-------
 (1981) provided evidence that acrylonitrile can alkylate DNA ui_ vitro.t  These
 authors found a low level of ^C-labeled acrylonitrile binding to calf
 thymus DNA when a metabolizing system (uninduced-rat liver microsomes plus
 NADPH) was included in the reaction mixture.   A twofold increase in  DNA
 binding over the response observed with uninduced rat microsomes was seen when
 liver microsomes were derived from rats pretreated with phenobarbital, an
 inducer of cytochrome P-450.  There was a much higher level  of binding to
 microsomal protein than to DNA (protein to DNA ratio was approximately 20 to
 1).  It should be pointed out that several known mutagens (e.g.  ethylene
 oxide, methyl methanesulfonate) have been shown to alkylate  protein  to a
 greater extent than DNA (Ehrenberg and Ostermar-Golkar 1980, Ehrenberg 1979).
 When Guengerich and coworkers (1981) used rat brain or human liver microsomes
 plus NADPH, the binding to DNA was much lower than when rat  liver microsomes
were used for metabolic activation.  Apparently, the in vitro rat brain and
 human liver microsomes activation systems do  not metabolize  acrylonitrile as
effectively.   These authors showed that the rat liver microsomes (or a
 reconstituted cytochrome P-450 system) metabolize acrylonitrile  to
2-cyanoethylene oxide, an epoxide.  This epoxide was shown to be stable and  to
also bind calf thymus DNA and protein.
    Glutathione S-transferase appeared to play a role in the deactivation of
acrylonitrile (and/or its metabolites) in these in vitro studies by  Guengerich
et al.  (1981).  A rat liver cytosol preparation of glutathione conjugated
 2-cyanoethylene oxide at a greater rate than  it conjugated acrylonitrile.
Human liver and rat brain cytosol  preparations reacted with  2-cyanoethylene
oxide at a much lower rate than did rat liver, and not at a  detectable rate
     tAcrylonitrile has been shown to react  with certain  minor tRNA
nucleosides, and at a slower rate, ribothymidine and  thymidine (Ofengard  1967
and 1971.
                                        13-72

-------
with acrylonitrile.  Therefore, if acrylonitrile was activated to



2-cyanoethylene oxide and reached the brain via blood circulation,  this



epoxide may not be effectively inactivated.  It should be cautioned,  however,



that these are in vitro studies and whole mammal  tests are needed to  elucidate



the events in vivo.  These results, however, are consistent with the  results



obtained mutagenicity studies in Salmonella and in Chinese hamster  ovary cells



(discussed below) in that metabolic activation was required for a detectable



response.



    Monsanto has sponsored preliminary experiments to study alkylation of DNA



in various tissues of rats (Dr. Paul  Wright, personal communication).   During



a EPA Science Advisory Board Meeting (August 2, 1982) Monsanto indicated that



no binding of acrylonitrile to DNA was detected in these experiments.



Monsanto made the preliminary data available to EPA, however,  without  details



of the protocol  and additional data,  a definitive judgment cannot be made



regarding these whole mammal  tests.  It appears,  however, that the  specific



activity of the label was too low to detect binding in these studies.   In



addition, only a nontoxic dose (approximately 10 mg/kg)  was evaluated.



Therefore, further studies are needed using radiolabeled acrylonitrile with a



higher specific activity and using several  doses  (including a  toxic dose).



    If a chemical  does react with DNA, it may ultimately cause DNA  damage.



Parent and Casto (1979) reported that acrylonitrile caused single strand DNA



breaks in primary Syrian golden hamster embryo cells in  vitro  as detected by



alkaline sucrose gradient sedimentation.  When cells were treated for  18



hours at concentrations of 200 or 400 ug of acrylonitrile/ml,  a shift  in



sedimentation patterns was observed.   This did not occur at 50 ug/ml  or  100



ug/ml.  The authors stated that these shifts are  comparable to those  produced



by known carcinogens and are not seen with non-carcinogenic chemicals.  They
                                         13-73

-------
concluded that the  results are suggestive of carcinogenicity.   However,
because the toxicity of the concentrations tested was not given, it is
uncertain if the DNA damage observed at 200 or 400 ug/ml  is simply a
reflection of nonspecific toxicity.  It should be noted  that  these data were
generated in the absence of an exogenous metabolic activation system.
    Dr. G. Williams (personal  communication, 1982) evaluated  the DNA damaging
potential of acrylonitrile by measuring DNA repair synthesis  in  a rat
hepatocyte primary culture.  A major advantage of the intact  hepatocyte test
system is that it provides a better approximation than an exogenous S-9 system
of the activation/deactivation of a chemical substance as it  would occur  in
the rat liver in vivo.  Dr. Williams indicated that acrylonitrile was negative
in this system; however, he emphasized that the test agent was intensely  toxic
to the hepatocytes.  He found that the hepatocytes were  biotransforming
acrylonitrile to cyanide, which is very toxic to the liver.   Thus, these
negative findings do not necessarily indicate the inability of acrylonitrile
to damage DNA.  This conclusion is supported by other studies by Dr. S. Ved
Brat and Dr. G. Williams (unpublished, 1982) that have shown  a significant
increase in sister chromatid exchange formation in Chinese hamster ovary  (CHO)
cells when metabolic activation is mediated by rat hepatocytes.
    S. Ved Brat and G. Williams (1982, unpublished) studied the  induction of
sister chromatid exchange (SCE) in CHO cells by acrylonitrile.   The evaluation
of SCE formation, which is thought to involve DNA breakage, is a very
sensitive method that can detect genotoxic effects at lower concentrations
than those needed to detect an increase in chromosome aberrations (Wolff  1977,
Latt 1974).  In two separate experiments, exposure to acrylonitrile did not
produce SCE in CHO cells without a metabolic activating  system when evaluated
at five concentrations from 10~7M up to IQ-^M, but did cause  a progressive
                                         13-74

-------
increase in the SCE incidence when CHO cells were cocultured  with  rat
hepatocytes for metabolic activation.   The highest response  (approximately
twofold increase over background levels)  was observed  at  10"4M
acrylonitrile.  It should be pointed out  that coculturing hepatocytes  with  the
CHO cells resulted in a elevated background level  of  SCE  in  CHO (twice than
what is normally observed).  Nevertheless, in two separate experiments an
dose-related response (with a twofold increase at the  highest dose)  was
demonstrated.  These results indicate that the rat liver  cells can mediate  the
activation of acrylonitrile and that the  reactive metabolite(s) apparently  is
reaching the CHO cells to cause genetic damage.

13.4.4  summary and conclusions
    Acrylonitrile has been evaluated for  its ability  to cause gene mutations
in bacteria and Drosophila and for its ability to cause chromosome damage  in
plants and rodents (acrylonitrile-exposed workers have also  been screened  for
induced chromosome damage).  Other endpoints indicative of mutagenicity have
been examined, such as DNA repair synthesis and  sister chromatid exchange
formation in cultured mammalian cells  and chemical-DNA adduct formation j_n_
vitro.
    There is evidence that acrylonitrile  can cause point  mutations in  both
Salmonella tyhimurium and Escherichia  coli.  A difference in  the requirement
of acrylonitrile for exogenous hepatic metabolic activation  was reported for
the Salmonella and Escherichia strains, i.e., liver S-9 activation was
required for acrylonitrile mutagenicity in Salmonella  but not in Escherichia.
Further investigation would be necessary  to determine  the reason for this
reported difference.   The epoxide of acrylonitrile (a  potential metabolite)
has been shown to be mutagenic in Salmonella.  Because acrylonitrile and the
                                        13-75

-------
metabolite can cause point mutations in bacteria it  may cause  point/gene
mutations in other organisms as well.   Results  from  a  sex-linked  recessive
lethal test in Drosophila, in which the data are considered  suggestive  of an
increase in mutation frequency over that of  the spontaneous  frequency by a
factor of three, provide some support  for this  expectation.  Further testing,
however, is needed in other organisms  to confirm the ability of acrylonitrile
to cause point/gene mutations.
    Acrylonitrile has not been detected as causing chromosome  damage.   Studies
in plants have been reported to be negative.  In rats  and mice, acrylonitrile,
(when evaluated up to toxic doses) was not found to  be clastogenic  in the bone
marrow cells of these mammals.  In addition, no apparent chromosomal damage
was detected in peripheral blood lymphocytes from workers exposed to
acrylonitrile.  However, it did induce an increase in  sister chromatid
exchange formation in Chinese hamster  ovary  (CHO) cells in vitro when
metabolic activation was mediated by intact  rat hepatocytes.   The induction of
SCE is thought to involve DNA breakage and this type of genetic damage  can be
detected at concentrations lower than  those  needed to  detect an increase in
chromosome aberrations.   Acrylonitrile was negative  in a DNA repair test using
primary rat hepatocytes.  In this test, acrylonitrile  was intensely toxic to
the liver cells because  of the production of cyanide.   Thus, the compound
cannot be adequately evaluated in the  rat hepatocyte-DNA repair test system.
    DNA binding studies  indicate the potential  of acrylonitrile to  react with
DNA, a critical event in certain types of mutagenesis.  A low  level of  binding
of radiolabeled acrylonitrile to calf  thymus DNA in  vitro was  detected  in the
presence of exogenous rat liver microsomal activation.  This requirement of
metabolism is consistent with the Salmonella mutagenicity studies and SCE
studies in CHO cells.  A metabolite of acrylonitrile,  the epoxide
                                         13-76

-------
2-cyanoethylene oxide, also binds calf thymus  DMA  in  vitro.   However,
additional studies are necessary to determine  if DNA  binding  occurs  in  intact
cells In vitro and in whole mammals after acrylonitrile  exposure.
    The ability of acrylonitrile and/or an active  metabolite  to  reach
mammalian gonads and cause germ cell  mutations has not been adequately
studied.  Only one study was available for review.  This was  a mouse dominant
lethal assay in which the test results were not significant.  However,  the
high background of dominant lethals in this study  precluded a negative
interpretation of the results.
     In conclusion, the weight-of-evidence indicates  that  acrylonitrile  has
the potential  to bind to DNA (as demonstrated  by the  calf  thymus DNA binding
studies in vitro), and cause genetic  damage (as shown in the  point mutation
studies in bacteria and in sister chromatid exchange  studies  in  mammalian
cells in vitro and as suggested in the Drosophila  studies).   The available
evidence indicates that conversion of acrylonitrile to a metabolite(s),  such
as an epoxide, may be required to produce the  ultimate mutagenic form(s).
Thus, if the metabolism and pharmacokinetics of this  chemical substance  in
humans results in metabolic products  that can  interact with DNA, as  is  the
case for the test systems employed in the reported studies, it may cause
somatic mutations in humans as well.   An assessment of genetic risk with
respect to germ cell mutagenicity cannot be made because of the  lack of
appropriate data.  Additional  studies are needed,  particularly in eukaryotic
organisms to confirm that acrylonitrile is mutagenic  as  indicated by the
positive responses observed in bacteria.  It would also  be appropriate  to
conduct tests using radiolabeled acrylonitrile to  measure  alkylation of  DNA in
various tissues of whole mammals.
                                         13-77

-------
       A.  GENE MUTATION TESTS:  BACTERIA
                                                     fable  13-18.   MUTAGEN1CITY TESTS OF ACRYLONITRILE
—i
00

Reference
Connor et al .
1979








DeMeester
et al. 1978














Test System
Body fluid
analysis: bile
from rats.
Salmonella
typhlmurium as
indicator organ-
ism; plate incor-
poration assay


Salmonel la
test: plate
test, fluctu-
ation assay












Strain
TA 1535




Activation
System
In vivo male Sprague-
DTwley rats




Concentration
45 mg/kg acrylo-
nitrl le in sal ine
Injected i.v. Into
femoral vein, 200 ul
of bile tested/plate
Reported
Result*
Negative 1.

2.

3.
Source: Aldrich Chemical




TA 1530
TA 1535
TA 1537
TA 1538
TA 98
TA 100
TA 1950
his G46
TA 1978
TA 1532
TA 1975








S-9 mix prepared from
1 Ivers of styrene,
Aroclor 1254, acrylo-
nitrlle, methyl -3-chol -
anthrene, phenobarbital
-induced male rats
(Wlstar); Aroclor
1254- and phenobarbital
-Induced female mice
(NMRI); uninduced male
beagle dog




Company



Tested in atmos-
phere at + 0.2%
for 1 hour, purity
reported as 99%
(Impurities not
described).
Source: Aldrich,
Europe







4.

5.

Positive 1n the 1.
presence of meta-
bolic activation
in both framesnlft
-sensitive and
base-pair sub-
stitution sen-
sitive strains, 2.
with most pro-
nounced effects In 3.
base-pair substi-
tution sensitive 4.
strains. Negative
in the absence of
metabolic activation

Comment
Negative findings are
inconclusive.
One bacterial strain
tested.
Number of animals used not
reported.
Number of revertants not
reported.
Purity of chemical not
given.
Responses are considered to
Indicate mutagenic activity
because there were clear
Increases in revertant colonies
over the background frequencies
and these increases were
reproducible.
No concurrent positive controls
reported.
Single dose tested; thus, no
dose-response curve.
Method to determine toxlcity
not described.


            •Authors conclusions.
(continued on following page)

-------
                                                                Table 13-1B.    (continued)
      A.  GENE MUTATION TEST:  BACTERIA
GO
 I
in

Reference
Kier
1982
Monsanto
Report












K1er
1982
Monsanto
Report






Test System
Salmonella/
micros ome
assay:
modi fied
1 iquld
suspension
test









Salmonella/
micros ome
assay:
modi fied
liquid
suspension
test


Activation
Strain System
TA 98 S-9 mix from
TA 100 livers of
TA 1535 Aroclor 1254-
induced rats












TA 98 S-9 mix from
TA 100 livers of
TA 1535 Aroclor 1254-
induced rats






Concentration
Acrylonitrile
0.02 mg/ml to
20 mg/ml . 1 iquld
material diluted
in distilled H20
Purity: > 991










2,3-epoxypropionitri le
(acrylonitrile epoxlde)
0.2 mg/ml to 20 mg/ml
diluted in distilled
H20
Purity: 97.4%



Reported
Result
Weak mutageniclty 1.
in strain TA 100
only after metabolic 2.
activation












Mutagenic in strains 1.
TA 100 and TA 1535,
TA 100 as the most
sensitive strain.
Mutagenic activity
expressed in absence
of S-9 mix and a si ight
Increase was found in
the presence of S-9 mix.

Comments
Results considered
border! ine at best.
A much higher concen-
tration (43 ug/ml )
was used for the
TA 1535 positive control
2-amlnoanthracene
than is generally '
employed (2.5
ug/plate). The
response of this
control was weak.
Therefore, results In
tester strain TA 1535
are difficult to
Interpret.
These data demonstrate
the mutageniclty of
acrylonitrile epoxlde.






                                                                                                                             (continued  on the next page)

-------
       A.  GENE MUTATION TESTS:  BACTERIA
                                                                 Table 13-18.   (continued)
00
 i
oo
o
Reference
Lambotte-
Vandepaer
et al. 1980
McMahon
et al. 1979
Test System Strain
Body fluid TA 1530
analysis:
urine from rats
and mice,
Salroonel la
typhimurium
as indicator
organism; plate
incorporation
assay
Reversion In
Salmonel la
typhimurium
and Escherichia
coll; gradient
plate test
Activation
System
In vivo adult male
FTstar rats and
adult male NMRI
mice; in vitro,
S9 mixTrom livers
of Aroclor 1254-
induced mice
(strain not
reported)

Concentration
30 mg/kg (purity >
99%, impurities
not described; in
saline injected
1 . p . ; urine
collected 24 hours
Source: Aldrich,
Europe

Reported
Result Comments
Positive 1. Single dose tested, thus
no dose-response curve
determined, but the data
are still considered to
Indicate mutagenic activity
2. No concurrent positive
control data reported.
Reported as 1. Positive conclusions cannot be
positive for both evaluated because of lack of
Salmonella and data.
Escherichia 2. Parity and source of chemical
not given.
3. Authors did not Indicate 1f
mutagenic activity was
dependent on exogenous
metabolic activation.
4. Concentratlon(s) where activity
was observed not reported.
5. Bacteria strains, which were
reverted, were not given.
6. Information on toxlcity not
reported.
(continued on following page)

-------
       A.  GENE MUTATION TESTS:  BACTERIA
                                                                  Table 13-18.   (continued)
CO
oo
Reference
HI Ivy and
Wolff 1977










Test System Strain
Salmonella TA 1535
test: spot TA 1538
assay, liquid TA 1978
suspension.
plate test (In
vapor phase)






Activation
System
S9 mix from
1 i vers of
Aroclor 1254-
induced male
Swiss-Webster
nice






Concentration
Spot test: 5 and
10, ul , liquid
suspension: 5, 10,
20 ul for 30 min.
to 4 hrs

Source: Aldrich
Chemical
Company
Purity: Reported
as 99+J

Reported
Result
Liquid suspension 1.
and vapor: positive
only In the pre-
sence of metabol 1c
activation in both 2.
frameshlft and base-
pair substitution 3.
sensitive strains;.
negative results using
spot test and testing
In the absence of
metabolic activation
Comments
Positive conclusions are
difficult to interpret
because of study design
problems and lack of data
No clear dose-response observed
with vapor exposure.
Neither the actual revertant
counts per plate not the
standard deviation per dose
was given; thus the extent of
variation Is not known.

                                                                                                                             (continued on following  page)

-------
                                                                  Table 13-18.  (continued)
        A.   GENE MUTATION TESTS:  BACTERIA
u>
i
oo
ro
Reference
Venltt
et al. 1977
Test System
Escherichia
col i test:
plate incor-
poration ,
fluctuation
assay
Salmonella
test: plate
Incorporation,
fluctuation
assay
Strain
WP2
WP2 urvA
WP2 urvApolA
WP2 lexA
WP2P (pKM 101)
TA 100
TA 1538
TA 98
TA 1535
his G46
his 03052
Activation
System
S9 mix from 1 ivers
of Aroclor 1254-
induced CB hooded
male rats, 0.5 ml/
plate
Concentration
0, 75, 150 umole/
plate for plate
incorporation
assay 4xlO'4M to
2xlO-3M for
fluctuation assay.
Dissolved in DMSO,
(purity > 99t as
determined by GLC) ;
Source: Kochlight
Lab
Reported
Result
Weak positive in
the absence of
metabol ic activa-
tion for al 1
strains except
UP2 lexA
Reported as
negative
Comments
1. Clear dose-related response.
2. No concurrent positive control
reported.
3. Data in the presence of
metabolic activation not
reported.
4. No precautions to prevent
evaporation were reported.
1. Negative conclusion cannot
be evaluated because data
were not presented and
protocol was not described.
                                                                                                                             (continued on following page)

-------
                                                                Table  13-18.   (continued)
       B.  GENE MUTATION TESTS:   DROSOPHILA
       Reference
Test System
                                            Dose
                              Result
                    Brood I               Brood II
                 (postmeiotic)    (melotlc and premeiotic)
                                                                                                       Total
                                                                                                     Comments
       Benes and
       Sram 1969
oo
 I
oo
CO
Sex-linked
recessive
lethal  test
In Drosophlla
melanogaster
Injection of a
0.1% solution
Into abdomen

spontaneous
frequency
2/572 (0.351)
4/725 (0.55%)
0.46X
                                   58/41.921  (0.14%)
1.  Results considered suggestive of
    three-fold Increase in mutation
    frequency.
2.  Toxlcity not given.
3.  Sufficiently large sample size not
    analyzed (7000 chromosomes are
    needed to detect a doubling 1n
    mutation frequency).
4.  Purity and source of chemical
    was not given.
5.  No concurrent positive controls
    and information concerning
    negative controls not given;
    appears to be historical.
6.  Post-mating times not given.
                                                                                                                            (continued on following page)

-------
                                                                   Table  13-18.   (continued)
         C.   HAMMALIAN IN VIVO CYTOGENETICS
Reference
Test System Strain
Reported
Concentration Result
Comments
         Leonard.
         et al.
         1981
                Bone  marrow
                assay
                (micronucleus
                and chromosomal
                aberration
                analysis)

                Dominant  lethal
                assay
                                          NMRI male mice
                                          NMRI male mice
                                         Mice give single i.p. Injection       Negative
                                         of 20 or 30 mg/kg and cells
                                         examined 6, 18, 24, 48, and 72
                                         hours later for chromosome analysis
                                         and 24, 30, and 48 hours later for
                                         micronucleus analysis.

                                         Mice given I.p. treatment of          Negative
                                         30 mg/kg followed by weekly
                                         matings (1 to 5 weeks).
                                         Source:  AldMch Europe
                                                  (purity not given)
                                                                            1.  Aberrations scored were gaps,
                                                                                breaks, and rearrangements.
                                                                            1.  Results considered Inconclusive
                                                                                because background control
                                                                                frequency was very high  (33.6%
                                                                                postimplantation  loss).
OJ
i
CO
Rabello-
Gay and
Ahmed 1980
Bone marrow
assay
Albino Swiss
mice. Sprague-
Dawley rats
Mice given at 7, 14, and 20
mg/kg/day by gavage
for 4, 15, 30 days ;
rats exposed orally
to 16 da1ly doses  of
40 mg/kg/day (purity
99.5i).
Source:  AldMch Chem.  Co.
Negative      1.   Aberrations scored were gaps,
                  Isogaps. breaks,  fragments, and
                  translocatlons.
         Thelss and
         Flelg 1978
                Examination of
                lymphocytes from
                workers with
                known  exposure
                to  acrylonltrlle
                                         Atmospheric monitoring
                                         data indicated a
                                         typical exposure level
                                         of 5 ppm between
                                         1963 and 1974, and a
                                         level of 1.5 ppm
                                         between 1975 and 1977
                                                             Negative
                                                        Cohort  was  exposed  to
                                                        acrylonltrlle for an  average of
                                                        15.3  years.
                                                        Aberrations scored  included  gaps
                                                        and Isogaps;  other  aberrations
                                                        scored  not  defined  in the  report.
                                                                                                                             (continued on following page)

-------
                                                                 Table 13-18.   (continued)
        0.  OTHER TESTS  INDICATIVE GENETIC DAMAGE
Reference
Guengerlch
et a). 1981
Test System Strain
DNA adduct
formation
in vltro-
cFlf thymus
DNA
Activation
System
Rat and
human liver
microsomes.
and puri fied
rat liver
P-450
Concentration
ImM 14C labeled
acrylonltrile (purity
> 99%) for 1 hour.
14C labeled
2-cyanoethylene oxide,
0.25 mM at different
times up to 4 hours
Reported
Result Connents
Low level of DNA binding
detected with rat liver
activation for acrylonitrlle,
2-cyanoethylene also
binds DNA In vitro.

 I
CO
Parent and
Casto 1979
Detection of
single strand
DNA breaks by
alkaline sucrose
gradient
Primary Syrian     None        Cells  Incubated for
Golden hamster                 18 hours  at  50, 100,
embryo cells                   200, and  400 ug of
In vitro                       acrylonitrile/ml
                               (>_ 99% pure);  and
                               contains  30-45 ppm
                               l-hydroxy-4-methoxybenzene
                               0.3% water,  <  0.5% acetonitrile.
                               Source: Aldrlch Chemical  Co.
DNA breakage at
200 and 400 ug/ml
but not at 50 or
100 ug/ml.
1.   Toxlclty data not given;
    thus, it is unclear
    whether the results are
    due to nonspecific toxlclty.
                                                                                                                              (continued  on  the  next  page}

-------
                                                          Table 13-18.    (continued)
D.  OTHER TESTS INDICATIVE OF GENETIC DAMAGE
Activation
Reference Test System Strain System Concentration
Ved Brat and Sister chromatid Intact rat Cells were exposed
Williams exchange (SCE) hepatocytes 3 hours to
1982 in CHO eel Is jji acrylonitri le
Unpublished vitro coculture dissolved in DMSO.
with rat hepatocytes Dose range from
10-7 H to 10-3M.
Purity: Not given
OJ
i
co
CT>
Concentration
Positive Control DMBA 5 x lO"5
Acrylonitri le 10"3
lo-^
10-5
ID"6
10-7
DMSO (1%)
Reported
Result
Negative results
for CHO cells
without hepatocytes.
Positive results for
CHO with hepatocytes
(M) Experiment
33.5 + 1.8
	
55.7 +_ 3.6
28.2 +_ 1.2
24.9 +; 1.4
23.7 jf 1.2
21.7 _* 1.7
Comments
1. Response was dose-
related In two
separate
experiments.
2. Background level of
SCE was very high.
SCE/CELL
1 Experiment 2
54.5 + 2
Toxic
45.6 +_ 1.5
36.4 + 1.0
30.9 ± 1.9
	
24.9 + 1.9

-------
13.5  CARCINOGENICITY






    The purpose of this section is to provide an evaluation of the likelihood



that acrylonitrile is a human carcinogen and, on the assumption that it is a



human carcinogen, to provide a basis for estimating its public health impact



including a potency evaluation in relation to other carcinogens.   The evaluation



of carcinogenicity depends heavily on animal  bioassays and epidemiologic



evidence.  However, information on mutagenicity and metabolism, particularly in



relation to interaction with DNA, as well  as  the pharmacokinetic  behavior, have



an important bearing on both the qualitative  and quantitative assessment of



carcinogenicity.  The available information on these subjects is  reviewed in



other sections of this document.  This section presents an evaluation of the



animal bioassays, the human epidemiologic evidence, the quantitative aspects of



assessment, and finally, a summary and conclusions dealing with all  of the



relevant aspects of carcinogenicity.







13.5.1  ANIMAL  STUDIES



    Seven studies in which acrylonitrile was  administered to rats will  be



discussed.  In four studies (three cancer bioassays and one three-generation



reproductive study), the route of administration was via drinking water, one



study was via gavage, and two (cancer bioassays) were via inhalation.







Drinking Water Studies



Dow Chemical Company (Quast et al. 1980a)--



    Quast et al. (1980a) of the Dow Chemical  Company performed a  2-year chronic



study under the auspices of the Chemical Manufacturers Association.   The



acrylonitrile used in this study was produced by E.I. du Pont de  Nemours and
                                         13-87

-------
Company, Inc.  Its purity was greater than 99fc.   In this study 6-to 8-week-old



male and female Sprague-Dawley rats (48 animals  of  each sex  at each exposure



level and 80 animals of each sex in the control  group)  were  given  acrylonitrile



in drinking water.  For the first 21 days, the concentrations  given were  35,  85,



and 210 ppm; however, the two higher concentrations were subsequently  raised  to



100 and 300 ppm.   After 9 months of treatment, the  animals at  the  two  higher



doses showed signs of toxicity as indicated by decreased self-grooming and  an



unthrifty appearance.  There was also a dose-related decrease  in food



consumption (except in male rats at the lowest exposure level), and concomitant



dose-related decrease in water consumption.  Using  water consumption data and



the weight of the rats, the calculated mean amount  of acrylonitrile ingested  for



males was 3.42, 8.53, and 21.18 mg/kg/day, and for  females,  4.36,  10.76,  and



24.97 mg/kg/day for the 35, 100, and 300 ppm exposure levels,  respectively.



    Cumulative mortality data for male and female rats  at all  exposure levels



are presented in Tables  13-19 and 13-20.  The results  indicate that early



mortality was observed at all dose levels in females.



    On death or at necropsy at 24 months, histopathologic examination  was



performed on complete sets of tissues from the control  and high dose animals  and



on selected target tissues and tissues with grossly recognized tumorous changes



from the other exposure groups.



    Statistically significant increases in tumor incidence at  multiple sites  in



male and female rats exposed to acrylonitrile were  observed  as described  in  Tables



13-21 and 13-22.  Statistically significant increases in the  incidence  of



specific tumor types with respect to individual  treatment groups are summarized



as follows:  central nervous system tumors in males and females  in all treatment



groups; Zymbal gland tumors in females in all treatment groups and in  males  in



the high-dose group; tumors in the nonglandular  portion of the stomach in males






                                          13-88

-------
TABLE 13-19.
CUMULATIVE MORTALITY DATA OF MALE RATS MAINTAINED FOR 2 YEARS ON
      DRINKING WATER CONTAINING ACRYLONITRILE
                (Quast et al. 1980a)
Days on test
0-30
31-60
61-90
91-120
121-150
151-180
181-210
211-240
241-270
271-300
301-330
331-360
361-390
391-420
421-450
451-480
481-510
511-540
541-570
571-600
601-630
631-660
661-690
691-720
721-745
Total number
of rats
Control
No. Dead
(% dead)
0
0
0
0
0
1(1.3)
1(1.3)
1(1.3)
2(2.5)
3(3.8)
6(7.5)
7(8.8)
7(8.8)
8(10.0)
11(13.8)
15(18.8)
21(26.3)
29(36.3)
33(41.3)
40(50.0)
48(60.0)
54(67.5)
67(83.8)
70(87.5)
73(91.3)

80(10056)
35 ppm
No. Dead
(% dead)
0
0
0
0
0
0
1(2.1)
1(2.1)
1(2.1)
1(2.1)
2(4.3)
2(4.3)
3(6.4)
5(10.6)
7(14.9
8(17.0)
13(27.7)
14 29.8
20(42.6)
23(48.9)
27(57.4)
33(70.2)
39(83.0)
41(87.2)
42(89.4)

47(100%)
100 ppm
No. Dead
(% dead)
0
0
0
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
1(2.1)
2(4.2)
5 10.4)
7(14.6)
11(22.9)
16(33.3)
18(37.5)
23(47.9)
26(54.2)
33(68.8)
35(72.9)
40(83.3
43(89.6)

48(100%)
300 ppm
No. Dead
(% dead)
0
0
0
0
0
0
0
0
2(4.2)
2(4.2)
2(4.2)
4(8.3)
7(14.6)
8(16.7)
11(22.9
15(31.3)
22(45.8)*
29(60.4)*
32(66.7)*
34(70.8)*
37(77.1)*
40(83.3)*
48(100)*
48(100)*
48(100)*

48(100%)
    *Sign1f1cantly different from controls by Fisher s Exact Probability Test,
P < 0.05.
                                          13-89

-------
TABLE 13-20.
CUMULATIVE MORTALITY DATA OF FEMALE RATS MAINTAINED FOR 2 YEARS ON
        DRINKING WATER CONTAINING ACRYLONITRILE
                   (Quast et al.  1980a)
Days on test
0-30
31-60
61-90
91-120
121-150
151-180
181-210
211-240
241-270
271-300
301-330
331-360
361-390
391-420
421-450
451-480
481-510
511-540
541-570
571-600
601-630
631-660
661-690
691-720
721-745
Total number
of rats
Control
No. dead
(% dead)
0
0
0
0
0
0
1(1.3)
1(1.3)
1(1.3)
1(1.3)
1(1.3)
1(1.3)
1(1.3)
3(3.8)
3(3.8)
6(7.5)
9(11.3)
11(13.8)
18(22.5)
22(27.5)
34(42.5)
37(46.3)
45(56.3)
54(67.5)
60(75.0)

80(100%)
35 ppm
No. dead
(% dead)
0
0
0
0
0
0
0
0
0
0
1(2.1)
1(2.1)
3(6.3)
3(6.3)
5(10.4)
7(14.6
10(20.8)
12(25.0)
20(41.7)*
24(50.0)*
27(56.3)
33(68.8)*
38(79.2)*
42(87.5)*
44(91.7)*

48(100%)
100 ppm
No. dead
(% dead)
0
0
0
0
0
0
1(2.1)
1(2.1)
2(4.2)
3(6.3)
3(6.3)
3(6.3)
5(10.4)*
6(12.5)
6(12.5)
11(22.9)*
13(27.1)*
17(35.4)*
27(56.3)*
34(70.8)*
38(79.2)*
42(87.5)*
44(91.7)*
46(95.8)*
47(97.9)*

48(100%)
300 ppm
No. dead
(% dead)
0
0
0
0
1(2.1)
1(2.1)
1(2.1
1(2.1)
3(6.3)
4(8.3)
9(18.8)*
14(29.2)*
16(33.3)*
20(41.7)*
24(50.0)*
31(64.6)*
35(72.9)*
37(77.1)*
42(87.5)*
45(93.8)*
46(95.8)*
46(95.8)*
48(100)*
48(100)*
48(100)*

48(100%)
  P < 0.05.
                                           13-90

-------
                                    TABLE 13-21.  HISTOPATHOLOCIC DIAGNOSES AND TUMOR INCIDENCES IN MALE RATS
                                          MAINTAINED FOR 2 YEARS ON DRINKING WATER CONTAINING ACRYLONITRILE
                                                                 (Quast et al. 1980a)
OJ
 i
Diagnoses
Ear Canal Gland (Zymbal gland)
carcinoma
Tongue
squamous cell tumors
(papHloma and/or carcinoma)
Stomach - Nonglandular portion
papllloma, squamous cell
Stomach - Nonglandular portion
squamous cell tumors
papllloma and/or carcinoma
Brain and/or Spinal Cord
focal or multlfocal gllal
cell proliferation suggestive
of early tumor
Brain and/or Spinal Cord
gllal cell tumor
(astrocytoma)
Brain and/or Spinal Cord
gllal cell tumor
(benign and/or malignant)
0 ppm (%)
3/80 (3.8)
1/75 (1.3)
0/80 (0.0)
0/80 (0.0)
0/80 (0.0)
1/80 (1.3)
1/80 (1.3)
Tumor Incidences
35 ppm (%)
4/47 (8.5)
2/7§
2/46 (4.3)
3/46 (6.5)
P = 0.047t
4/47 (8.5)
P = 0.017t
8/47 (17.0)
P = 1.5 x 10~3t
12/47 (25.5)
P = 2.27 x 10'5t
at Various Dose Levels*
100 ppm (%)
3/48 (6.3)
4/9§
P = 3.10 x 10"4t
20/48 (41.7)
P = 1.40 x 10-10t
23/48 (47.9)
P = 2.24 x 10'12t
4/48 (8.3)
P = 0.0181
19/48 (39.6)
P = 7.86 x 10"9t
22/48 (45.8)
P = 1.61 x 10-lot
300 ppm (X)
15/48 (31.3)
P = 2.57 x 10-5t
5/40§
P = 0.019f
34/47 (72.3)
P = 1.59 x 10-20t
39/47 (83.0)
P = 3.95 x 10-25t
8/48 (16.7)
P = Z.64 x 10-"
23/48 (47.9)
P = 4.16 x lO"111
30/48 (62.5)
P = 1.21 x 10-15t
                   •Tumor Incidence expressed as number of rats bearing certain tumor types over the number  of  rats  examined
               microscopically for that particular organ.
                   tp-values calculated using the one-tailed Fisher Exact Test.
                   §A11 samples were not taken from all animals.

-------
                                   TABLE  13-22. HISTOPATHOLOGIC DIAGNOSES AND TUMOR  INCIDENCES  IN FEMALE RATS
                                        MAINTAINED FOR 2 YEARS ON DRINKING WATER CONTAINING ACRYLONITR1LE
                                                               (Quast et al. 1980a)
<£>
ro
Diagnoses
Mammary Gland
f Ibroadenoma/
adenof Ibroma
Mammary Gland
benign and/or malignant
mammary gland tumors
Ear Canal Gland (Zymbal gland)
adenoma
Ear Canal Gland (Zymbal gland)
carcinoma
Ear Canal Gland (Zymbal gland)
benign and/or malignant tumors
Tongue
carcinoma,
squamous cell
Tongue
papllloma and/or carcinoma,
squamous cell
Stomach - Nonglandular
papllloma, squamous cell
0 ppm (%)
56/80 (70.0)
57/80 (71.3)
0/80 (0.0)
1/80 (1.3)
1/80 (1.3)
0/78 (0.0)
0/78 (0.0)
1/80 (1.3)
Tumor Incidences
35 ppm (%)
41/48 (85.4)
P = 0.037t
42/48 (87.5)
P = 0.026f
1/48 (2.0)
4/48 (8.3)
5/48 (10.4)
P = 0.028t
1/5
1/5
1/47 (2.1)
at Various Dose Levels*
100 ppm (%)
39/48 (81.3)
42/48 (87.5)
P = 0.026t
3/48 (6.3)
P = 0.051*
6/48 (12.5)
P = 0.011*
8/48 (16.7)
P = 1.67 x 10-3t
0/3 (0.0)
2/3 (66.7)
P = 9.26 x 10~4t
12/46 (25.0)
P = 2.72 x 10-5t
300 ppm (%)
25/48 (52.1)
35/48 (72.9)
4/48 (8.3)
P = 0.018f
14/48 (29.2)
P = 3.00 x 10-6t
18/48 (37.5)
P = 2.72 x 10'8t
10/45 (22.2)
P = 2.13 x 10"5t
12/45 (26.7)
P = 2.00 x 10~6T
29/48 (60.4)
P = 6.03 x 10'15t
                  •Tumor  Incidence expressed as number or rats bearing certain
              microscopically  for that particular organ.
                  tp-values  calculated using the one-tailed Fisher Exact Test.
                  §A11  samples were not taken from all animals.
tumor types over the number of rats examined
                                                                                                  (continued on the following page)

-------
                                            TABLE  13-22.   (continued)


h-»
OJ
1
10
CO


Diagnoses
Stomach - Nonglandular portion
carcinoma, squamous cell
Stomach - Nonglandular portion
papllloma and/or
carlnoma, squamous cell
Small Intestine
cystadenocardnoma, muclnous
Brain and/or Spinal Cord
focal or multlfocal glial
cell proliferation suggestive
of early tumor
Brain and/or Spinal Cord
focal or multifocal glial
cell tumor (astrocytoma)
Brain and/or Spinal Cord
focal or multlfocal glial
eel 1 tumors
(benign and/or malignant)
0 ppm (%)
0/80 (0.0)
1/80 (1.3)
0/80 (0.0)
1/80 (1.3)
0/80 (0.0)
1/80 (1.3)
Tumor Incidences at
35 ppm (%)
0/47 (0.0)
1/47 (2.1)
l/7§
5/48 (10.4)
P = 0.028f
17/48 (35.4)
P = 6.90 x 10-9t
20/48 (41.7)
P = 2.21 x 10-9*
Various Dose Levels*
100 ppm (%)
0/48 (0.0)
12/48 (25.0)
P = 2.72 x 10'5t
4/ll§
P = 1.23 x 10-4t
4/48 (8.3)
22/48 (45.8)
P = 9.15 x 10-12t
25/48 (52.1)
P = 2.52 x 10'12t
300 ppm (%)
12/48 (25.0)
P = 2.94 x 10~6t
30/48 (62.5)
P = 1.21 x 10-15t
4/48 (8.3)
P = 0.018t
7/48 (14.6)
P = 4.38 x 10'3t
24/48 (50.0)
P = 5.34 x lO'13*
31/48 (64.6)
P = 2.31 x 10'16t
    *Tumor Incidence expressed as number of  rats  bearing  certain  tumor  types over the number of rats examined
microscopically for that particular organ.
    tp-values calculated using the one-tailed  Fisher  Exact  Test.
    §A11 samples were not taken from all  animals.

-------
in all treatment groups and in females in the mid- and high-dose groups;  tongue
tumors in males in all  treatment groups and in females in  the mid-  and  high-dose
groups; mammary gland tumors in females in the low- and mid-dose groups;  tumors
in the small intestine in females in the mid- and high-dose  groups.
    The central nervous system tumors included a  statistically significant
increase in the incidence of astrocytomas.  There was  also a statistically
significant increase in the incidence of glial cell proliferations,  suggestive
of early tumors, which were observed most frequently in the  cerebral  cortex.
Zymbal gland tumors were observed in the ear canal, and these tumors  were
usually ulcerated and,  in some animals, caused displacement  of the  lower  jaw
that resulted in the inhibition of food consumption.  In the nonglandular
portion of the stomach, both papillomas and carcinomas were  found with  a
progression from hyperplasia and hyperkeratosis to papilloma and, finally,  to
carcinoma.  Tongue tumors were diagnosed as squamous cell  papillomas  and
carcinomas, and tumors in the small  intestine were identified as
cystadenocarcinomas.

Biodynamics, Inc. Study in Sprague-Dawley Rats (1980a)--
    A report entitled "A Twenty-four Month Oral Toxicity/Carcinogenicity  Study
of Acrylonitrile Administered to Spartan Sprague-Dawley Rats in the  Drinking
Water," dated June 30,  1980, was submitted to the U.S. Environmental  Protection
Agency by the Monsanto Company, St.  Louis, Missouri.  This study, conducted by
Biodynamics, Inc. for the Monsanto Company, was designed to  evaluate  the
toxicity/carcinogenicity of acrylonitrile.  Acrylonitrile  (100% pure, supplied
by the Monsanto Company) was administered in the  drinking  water to  100
Sprague-Dawley rats of each sex at dose levels of 0, 1, and  100 ppm.   Interim
necropsies were performed at 6, 12,  and 18 months (10/sex/group).  The  study was
                                         13-94

-------
terminated early due to low survival  rates;  females  were  sacrificed  at  19 months
and males were sacrificed at 22 months.
    Body weights for the high-dose males  and females were consistently  lower
than the weights of controls; body weight differences  between controls  and
treated male rats were less than 10%,  while  the  body weight  differences between
controls and treated female rats were  less than  8%.   Body weights  for the
low-dose males and females were generally comparable to the  controls throughout
the study.  Reduced water intake (test substance)  was  consistently noted in the
high-dose group.  Water intake for the low-dose  group  was generally  comparable
to the controls throughout the study.   Slight, but consistent decreases in
hemoglobin concentration, hematocrit,  and erythrocyte  counts were  noted for the
high-dose males and females.
    Histopathology evaluation revealed an increased  incidence of astrocytomas of
the brain and spinal  cord, carcinomas  and adenomas of  the Zymbal gland  or ear
canal, and squamous cell carcinomas and papillomas of  the forestomach in the
high-dose males and females  (Table 13-23).   Increased incidences  of the
aforementioned tumor types were observed  predominantly in animals  dying,
sacrificed in a moribund condition, or at sacrifice  intervals after  the 12th
month of the study, although the increased incidences  of  astrocytomas of the
brain and carcinomas of the Zymbal  gland  or  ear  canal  were noted earlier in the
high-dose female after the sixth month of the study.
    In conclusion, the carcinogenic effect of acrylonitrile  administered to rats
in drinking water further reconfirmed  the earlier  findings of Dow  and Litton-
Bionetics study discussed latter in this  section.
                                        13-95

-------
        TABLE 13-23.
TUMOR INCIDENCES IN SPRAGUE-DAWLEY RATS FED
 ACRYLONITRILE IN DRINKING WATER*
     (Biodynamics, Inc.  1980a)
Dose
Level
(ppm)
0
1
100
Sex
M
F
M
F
M
F
Brain
Astrocytoma
2/98 (2%)
0/99 (0%)
3/95 (3%)
1/100(1%)
23/97(24%)t
32/97(33%)t
Spinal Cord
Astrocytoma
— §
0/96(0%)
— §
0/99(0%)
— §
7/98(7%)t
Zymbal Gland/
Ear Canal
Carcinomas
1/100(1%)
0/99 (0%)
0/91 (0%)
0/95 (0%)
14/93(15%)t
7/98(7%)t
Stomach
Papil loma/
Carcinoma
3/98 (3%)
1/100(1%)
3/98 (3%)
4/99 (4%)
12/97(12%)t
7/99(7%)t
    *Animals intentionally sacrificed  at  6,  12,  and  18 months  after
acrylonitrile administration are included in  the denominator.

    tStatistically significant at P <  0.05.

    §Tissues not analyzed.
                                         13-96

-------
Biodynamics, Inc. Study in Fischer 344 Rats  (19806)--



    A report entitled "A Twenty-four month Oral  Toxicity/Carcinogenicity  Study



of Acrylonitrile Administered to Fischer 344 Rats  in Drinking  Water,"  dated



December 12, 1980, was submitted to the U.S. Environmental  Protection  Agency  by



the Monsanto Company, St.  Louis, Missouri.  This study,  conducted  by Biodynamics



Inc. for the Monsanto Company, was designed  to evaluate  the toxicity/



carcinogenicity of acrylonitrile.  Acrylonitrile (100% pure supplied by the



Monsanto Company) was administered in the drinking water to 100  Fischer 344  rats



of each sex at dose levels of 1, 3, 10, 30,  and 100 ppm, and the control  group



contained 200 animals/sex.  Interim necropsies were performed  at 6,  12, and  18



months (20/sex from the control  group and 10/sex from  each  treatment group each



time interval).  This study was originally designed to be 24 months  in duration;



however,  to ensure at least 10 animals/sex/group for histopathological



evaluation at termination, all females were  sacrificed at 23 months due to low



survival.  The males were  continued on test  until  the  26th  month when  similar



survival  levels were reached.  Mortality in  the males  and females  receiving  100



ppm was markedly greater than controls, while mortality  in  the 10  ppm  males  and



females receiving 3 and 30 ppm was also somewhat greater than  controls.   Food



consumption was comparable for all  groups on a g/kg/day  basis.  Liquid



consumption for the females receiving 100 ppm was  slightly  lower than  controls



on a ml/kg/day basis, while values for the males in this group were comparable



to or greater than controls.



    Slight, but generally  consistent reductions in hemogoblin, hematocrit, and



erythrocyte counts were noted for the females receiving  100 ppm  throughout the



study.



    Histopathological evaluation presented in Table 13-24 revealed an  increased



incidence of malignant tumor-bearing animals in the groups  receiving 10,  30,  and
                                          13-97

-------
        TABLE 13-24.   TUMOR INCIDENCE IN FISCHER 344 RATS FED ACRrLONITRlLE
                                     IN DRINKING WATER*
                                 (Biodynamics, Inc.  19805)
        Dose Level
Groups    (ppm)    Sex
   Brain      Spinal  Cord
Astrocytomat  Astrocytomat
Ear Canal
(Zymbal  Gland)      Stomach-Squamous
Papil Ionia/Adenoma/  Cell  Papilloma/
Carcinoma            Carcinoma of
(Squamous  Cell )t    Forestomacht
IA & IB Control H
F
II
III
IV
V
VI
1 M
F
3 M
F
10 M
F
30 M
F
100 M
F
2/200 (1.0)
1/199 (0.5)
2/100 (2.0)
1/100 (1.0)
1/100 (1.0)
2/101 (2.0)
2/100 (2.0)
4/95 (4.2)§
10/99(10. 1)§
6/100 (6.0)§
21/99(21. 2)§
23/98(23. 4)§
1/196
1/197
0/99
0/97
0/92
0/99
0/98
l/92§
0/99
0/96
4/93
1/91
(0.5)
(0.5)
(0.0)
(0.0)
(0.0)
(0.0)
(0.0)
(1.0)
(0.0)
(0.0)
(4.3)§1I
(1.1)
2/189(1.0)
0/193(0.0)
1/97 (1.0)
0/94 (0.0)
0/93 (0.0)
2/92 (2.2)
2/88 (2.3)
4/90 (4.4)§
7/94 (7.4)§
5/94 (5.3)§
16/93(17. 2)§#
10/86(11.6)§
0/199(0.0)
1/199(0.5)
1/100(1.0)
1/100(1.0)
4/97 (4.1)§
2/100(2.0)
4/100(4. 0)§
2/97 (2.1)
4/100(4. 0)§
4/100(4. 0)§
1/100(1.0)
2/97 (2.1)
    *AnimaTs sacrificed intentionally ate, 12,  ana 18 months  after  acrylomtrile
administration are included in the denominator.

    tNumbers in the parentheses are percentages.

    §Statist1cally significant at P < 0.05.

    HThese rats had astrocytoma in both brain  and spinal  cord.

    fOne rat had unilateral papilloma/carcinoma.
                                            13-98

-------
100 ppm.  The observed tumors were:  astrocytomas of the central  nervous  system
(brain and/or spinal cord) and squamous cell  carcinomas  of  the  ear canal,  as
well as mammary gland carcinomas in the females receiving 100 ppm.
    In summary, the ingestion of acrylonitrile via drinking water at  doses of
10, 30, and 100 resulted in an increased incidence of certain tumors.   The
target organ specificity (central  nervous system, ear canals) confirms  the
similar earlier findings of the drinking water study in  rats by the Dow Chemical
Company.


Litton-Bionetics Study (Beliles et al.  1980)--
    Beliles et al.  (1980) of Litton-Bionetics, Inc., sponsored  by the Chemical
Manufacturers Association, performed a  three-generation  reproductive  study in
Charles River rats  [CRL:COBS CD (SD) BR].  These  rats and their offspring
ingested water containing 100 or 500 ppm acrylonitrile starting approximately  15
days post weaning  and were mated after  100 days.   Female rats were maintained  on
water containing acrylonitrile for 20 weeks;  following delivery of the  second
litter, the animals were exposed to acrylonitrile for approximately 45  weeks.
Following exposure, the animals in the  three  generations FQ, F^b, and p£b
were sacrificed and observable masses were evaluated histologically.  Results  of
histologic evaluations are presented in Table 13-25.  The tumor incidence  was.low;
only rats of the second generation at the high-dose level showed  a significant
increase in the number of tumors.   The  low tumor  incidence  was  probably due to
the relatively short exposure and  observation period (approximately 45  weeks).
This study was suggestive of tumorigenic action of acrylonitrile  since  the type
of tumors observed  were the same as the statistically highly significant tumor
incidence of Quast  et al. (1980).   This study provides further  confirmation of
the incidence of astrocytoma and Zymbal gland tumors observed by  Quast  et  al.
(1980).
                                         13-99

-------
    TABLE 13-25.  INCIDENCE OF TUMORS OBSERVED IN RATS DURING
                  A THREE-GENERATION REPRODUCTIVE STUDY
                         (Bellies et al.  1980)
Generation 0
F0 0/19 (0%)
Fjb 0/20 (0%)
F2b 0/20 (0%)
Total 0/59 (0%)
Generation 0
F0 0/19 (0%)
Fjb 0/20 (0%)
F2b 0/20 (0%)
Total 0/59 (0%)
Astrocytoma Incidence
Dose (ppm acrylonitrile in water)
100 500
1/20 (0%)
1/19 (5.2%)
1/20 (5%)
3/59 (5%)
Zymbal Gland Tumor
Dose (ppm acrylonitri
100
0/20 (0%)
2/19 (10.5%)
0/20 (0%)
2/59 (3.4%)
2/25 (8%)
4/17 (23.5%)
P = 0.036*
1/20 (5%)
7/62 (11.2%)
P = 7.8 x ID'3*
Incidence
le in water)
500
1/25 (4%)
4/17 (23.5%)
P = 0.036*
3/20 (15%)
8/62 (12.9%) o
P = 3.7 x ID'3*
*P-values calculated using the Fisher Exact Probabiity Test
                                     13-100

-------
Gavage Studies



Maltonl et al. 1977--



    Maltoni et al. (1977) performed a cancer bioassay of acrylonitrile in which



40 Sprague-Dawley  rats of each sex in both the treated and control  groups were



exposed to a single dose of 5 mg/kg acrylonitrile by gavage dissolved in olive



oi1, 3 times a week, for 52 weeks.  On spontaneous death,  a moderate increase in



tumors of the mammary gland region and forestomach of female rats was described.



Although this study was very limited, with only a single exposure level  and a



relatively short observation period (52 weeks), the results present further



evidence for the carcinogenicity of acrylonitrile.







Biodynamics, Inc.  Gavage Study in Sprague-Dawley Rats (1980c)--



    A report entitled "A Twenty-four Month Oral Toxicity/Carcinogenicity Study



of Acrylonitrile Administered by Intubation to Sprague-Dawley (Spartan)  Rats"



was submitted to the U.S. Environmental Protection Agency, June 30, 1980, by the



Monsanto Company,  St. Louis, Missouri.  This study was conducted by Biodynamics,



Inc. for the Monsanto Company.  In this study, acrylonitrile (100% pure*) was



administered by intubation to Sprague-Dawley (Spartan) rats (100/sex/group) at



three dose levels  of 0, 0.10, and 10.0 mg/kg/day, 5 days/week.   Interim



necropsies were performed at 6,  12, and 18 months (10/sex/group).  This  study



was originally designed to terminate at 24 months;  however, because only 10 and



13 high-dose males and females,  respectively, were alive by 20  months, all



surviving animals  in all groups  were terminated during the 20th month to ensure



at least 10 animals/sex for histopathological evaluation.   The  body weights of



high-dose group males were consistently slightly lower than control.
    *Supplied by Monsanto Company, Texas City, Texas.
                                         13-101

-------
    Histopathological evaluations presented in Table 13-26 show that there were
statistically significant increased incidences in tumors of the brain and ear
canal (Zymbal gland) in both high-dose males and females.   Stomach and
intestinal tumors were observed only in high-dose males and mammary gland tumors
were observed in high-dose females.  Statistically significant tumor incidences
were not observed in low-dose groups either in males or females.
    In summary, acrylonitrile administered orally via intubation  produces a
carcinogenic effect in Sprague-Dawley rats at the following tumor sites:  brain,
Zymbal gland, mammary gland, stomach, and intestine.

Inhalation Studies
Maltoni et al. (1977)--
    Maltoni et al. (1977) exposed Sprague-Dawley rats to atmospheres containing
5, 10, 20, and 40 ppm acrylonitrile 4 hours/day, 5 days/week,  for 12 months.
The rats were maintained for their entire lifetime.  Histological  examination of
the selected tissues were made.  The incidence of tumors observed in this  study is
shown in Table 13-27.  Slight increases-were observed in mammary gland tumors of
males and females, the forestomach of males, and the skin of females.   Maltoni
et al. (1977) claimed that these results  indicated a "border-line carcinogenic
effect."  The detection power (sensitivity) of this assay was  low because  of  the
low concentration of acrylonitrile used and the relatively short exposure  period
(12 months).

Dow Chemical Company (Quast et al. 1980b) —
    A second inhalation study of acrylonitrile was conducted by Dow Chemical
Company, sponsored by the Chemical Manufacturers Association.   In this study,
100 male and female Spraque Dawley rats (Spartan substrain) were exposed to 0,
                                          13-102

-------
                 TABLE 13-26.  TUMOR INCIDENCE  IN RATS FED ACRYLONITRILE  ORALLY  BY  INTUBATION
                                         (Biodynamlcs,  Inc.  1980c)
1— •
CO
1— '
CD


Dose Level
mg/kg/day Sex
0 M
F
0.10 H
F
10.0 M
F
Brain
Astro-
cytoma
2/100 (2%)
1/99 (1%)
0/97 (0%)
2/100 (2%)
16/98 (16%)*
17/100(17%)*
Spinal Cord
Astrocytotna
0/94 (0%)
0/100(0%)
0/93 (0%)
0/95 (0%)
1/97 (1%)
1/99 (1%)
Ear Canal
(Zymbal Gland)
Squamous Cel 1
Carcinomas
1/96 (1%)
0/85 (0%)
0/93 (0%)
0/94 (0%)
10/96(10%)*
9/94 (10%)*
Stomach
Carcinoma/
Papilloma
2/99 (2%)
2/99 (2%)
6/97 (6%)
4/99 (4%)
40/99(40%)*
17/99(17%)
Intestine
0/100 (0%)
— t
1/100 (1%)
— t
6/100 (6%)*
---t
Mammary Gland
---t
7/101 (7%)
— t
6/100 (6%)
---t
22/101(22%)*
*Statistically significant at  P <  0.05.

tTlssues not analyzed.

-------
 TABLE 13-27.  TUMOR INCIDENCE IN RATS FOLLOWING INHALATION  OF  ACRYLONITRILE
                             (Maltoni  et  al.  1977)
Tumor Type (sex)                   Number of Animals  with  Tumors  (%)

                                Exposure Concentration  of  Acrylonitrile  (ppm)

                               0          5         10        20          40
Number of animals of each
sex
Mammary tumors (female)
Mammary tumors (male)
Zymbal gland (female)
Zymbal gland (male)
Encephalic tumors (female)
Encephalic tumors (male)
Forestomach papillomas
(female)
Forestomach papillomas
(male)
Skin carcinomas (female)
Skin carcinomas (male)
Uterine carcinomas
(female)
Total tumors (female)
Total tumors (male)
30
5(16)
1(3)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
0(0)
1(3)
9(30)
4(13)
30
10(33)
0(0)
0(0)
0(0)
0(0)
0(0)
1(3)
1(3)
4(13)
P = 0.056*
0(0)
3(10)
17(57)
1(3)
30
7(23)
1(3)
1(3)
1(3)
0(0)
0(0)
2(7)
2(7)
1(3)
0(0)
1(3)
11(36)
10(33)
30
10(33)
4(13)
P = 0.17*
1(3)
0(0)
0(0)
1(3)
1(3)
0(0)
1(3)
0(0)
2(7)
14(47)
13(43)
30
7(23)
4(13)
P = 0.17*
0(0)
0(0)
0(0)
2(7)
0(0)
3(10)
1(3)
1(3)
1(3)
8(27)
12(40)
    *P values calculated using the one-tailed Fisher Exact Test.
                                         13-104

-------
20, or 80 ppm of acrylonitrile 6 hours per day, 5 days per week,  for 2 years,
except during weekends or holidays.  The acrylonitrile used for this study was
produced by E.I. du Pont de Nemours and Company, Inc.  Its stability and
impurities are summarized in Table 13-28.  The nominal concentration of
acrylonitrile in the exposure chamber was calculated from the rate at which the
liquid test material was dispensed and the rate of airflow through the chamber.
The target concentrations and chamber concentrations are shown in Table 13-29.
    During the course of the 2-year study, hematology, urinalysis, and clinical
chemistry determinations were performed at periodic intervals.  The results of
these determinations indicated that acrylonitrile did not have an adverse effect
on bone marrow, kidney, or liver functions in either male or female rats.
During the first 6 months of the study, the exposed rats drank more water and
appeared to excrete a lower specific gravity urine than control rats.
    In lifetime observations of male and female rats exposed to acrylonitrile
vapors, toxic effects, characterized by decrease in body weight and early
mortality, were observed.  The cumulative mortality data are shown in Tables 13-30
and 13-31.
    Microscopic examination of tissues revealed a treatment-related
statistically significant incidence of tumors in the central nervous system, ear
canal gland (Zymbal gland), tongue, small intestine, and mammary gland.  In male
and female rats, statistically significant increased incidences were observed
only at the 80 ppm dose levels with the exception of glial cell tumors of the
central nervous system which were also increased in female rats at 20 ppm. Detailed
results are presented in Table 13-32.   however, these results are based on
data presented in the Quast et al. (1980b) inhalation study and do not reflect
separation of individual animals having one or more tumor types.   The CAG has
requested these individual data from Dr. Quast and the Chemical Manufacturers
Association.
                                          13-105

-------
              TABLE  13-28.  STABILITY AND TRACE IMPURITY ANALYSIS OF THE ACRYLONITRILE  LIQUID TEST MATERIAL
                                                    (Quasi et al. 1980b)

Refractive Index (0), 25.00°C
P-Methoxy phenol (MEHQ) Inhibitor, +1 ppm
Methacrylonltrile by GLC, +2 ppm
Other Compounds by GLC
Acetone by GLC, ppm
Volatiles by Mass Spec.
Peroxides as ppm \\2®2 (Polargraphic Data)
Metals by Atomic Emission with their Lower
Detection Limit in Parentheses, ppmll
Sample #1
Dated
1/27/76
1.3908
34
74.5
None
None
Water Onlyt
5
Sample 12
Dated
5/24/76
1 . 3908
32
75
None
None
Water Onlyt
1
Sample 13
Dated
11/19/76
1 . 3908
35.5
74.8
None
None
Hater Onlyt
4.5
Sample |4
Dated
3/21/77
1.3901
34.5
47
None
88+5
Water Onlyt
3.8
Sample, 15
Dated
9/7/77
1.3904
38.8
33-
None
195+5
Water Onlyt
0.2§
Sample 16
Dated
1/13/78
1 . 3909
41.7
56.5+1.5
	 •
131+7
AN Unly
N.O.
                Metal
Fe
Ni
Cu
Al
Ca
Si
Mg
Mn
Pb
Cr
Sn
Zn
Ti
Data 1s given In ppm on a
0.5a
N.D.(0.2)
O.la
N.D.(0.2)
0.7
b
N.D.(0.2)
N.D. (0.1)
N.D. (0.5)
N.D. (0.2)
N.O. (0.2)
N.D. (0.5)
N.D. (0.2)
volume/volume basis for the liquid
0.8a 0.8(0.2)
N.D.(0.2) N.D. (0.2)
O.la 0.1(0.1)
N.D.(0.2) 0.4(0.2)
1.0a 0.8(0.2)
b c(0.2)
N.D.(0.2) N.D.(0.2)
N.D (0.1
N.O. (0.5
N.D. (0.2
N.D. (0.2
N.D. (0.5
N.D. (0.2
0.1(0.1)
N.O. (0.5)
N.D. (0.1)
N.O. (0.2)
c(0.5)
N.D. (0.2)
test material.
la
N.D.(l)
N.D. (0.2)
b
b
b
N.D. (0.1)
N.D. (0.5)
N.O. (3)
N.D. (0.5)
b
b
N.D.(l)

N.D.(l)
N.D.(l)
N.D. (0.2)
N.D.(l)
N.D.(l)
b
N.D. (0.1)
N.D. (0.5)
N.D. (3)
N.D.(l)
N.D.(l)
N.D. (3)
N.D.(l)

KD
N.D.(l)
N.D. (0.2)
N.D.(l)
Kl)
b
N.D. (0.1)
N.D. (0.5)
N.O. (3)
N.D.(l)
N.D.(l)
N.D. (3)
N.D.(l)

•Small  unidentified peak ahead of acetone estimated  to  be  about  25  ppm.
tSens1t1v1ty 1s limited to about 100 ppm.
§Detection limit 1s 0.2.
HThe analysis of the sample analyzed 1s accurate to  +_ 10-15%  of  amount present.
*Lower detection limit was not specified.
bNo analysis was made for this metal.
cMetal  was detected, but unmeasurable.
N.D. - Not Detectable.

-------
         TABLE 13-29.  CHAMBER CONCENTRATIONS OF ACRYLONITRILE VAPORS
                              (Quast et al.  19806)
Target Concentration (ppm)
Analytical Concentration* (ppm)
X +_ S.D.
Range of Daily Analytical Concentration
(ppm)
Nominal Concentrationt (ppm)
X +_ S.D.
Range of Daily Nominal Concentration
Average Analytical Concentration/
Average Nominal Concentration
Number of Exposure Days Within +10%
of Target Concentration
Number of Exposure Days Within +_25%
of Target Concentration
Total Exposure Days
Total Number of Exposure Analyses
Exposure
20 ppm
20.0+1.9
7.3-35.0
20.7+2.6
11.00-37.8
0.97
433
64
507§
1484
Group
80 ppm
80.0+5.9
45.2-106.0
90.1+7.3
59.5-138.2
0.89
470
33
508
1494
    *Data represents mean (X) +_ standard deviation (S.D)  of daily 6-hour
time-weighted average for the total  number of exposure days.

    tData represents mean (X) _+ standard deviation (S.D.)  of daily 6-hour
nominal  concentration for the total  number of exposure days.

    §The 20 ppm exposure group had one less exposure day  than the 80 ppm
exposure group as a result of mechanical failure in the ventilation system of
the 20 ppm exposure chamber.
                                          13-107

-------
             TABLE 13-30.  CUMULATIVE MORTALITY DATA OF MALE RATS
           EXPOSED BY INHALATION FOR 2 YEARS TO ACRYLONITRILE VAPORS*
                              (Quast et al. 1980b)
                                            Exposure Concentration
Days on Test                         0 ppm             20 ppm         80 ppm
0-30
31-60
61-90
91-120
121-150
151-180
181=210
211-240
241-270
271-300
301-330
331-360
361-390
391-420
421-450
451-480
481-510
511-540
541-570
571-600
601-630
631-660
661-690
691-720
721-735
0
0
0
0
1
2
2
2
2
2
3
3
4
6
11
14
19
23
27
35
43
62
71
78
82
                                                         1              0
                                                         1              0
                                                         1              1
                                                         2              1
                                                         2              1
                                                         3              2
                                                         3              6
                                                         4             12t
                                                         5             13t
                                                         5             14t
                                                         6             16t
                                                         6             18t
                                                         8             19t
                                                         9             22t
                                                        12             24t
                                                        15             28t
                                                        26             39t
                                                        34             47t
                                                        38             56°
                                                        47             63t
                                                        59t            76t
                                                        68             83t
                                                        72             85t
                                                        81             94t
                                                        86             96t

Total Number of Rats                100                100            100
    *Data listed as number dead which is equal  to percent "cfead^

    tSignificantly different from control  by Fisher's Exact Probability Test,
P < 0.05.
                                          13-108

-------
            TABLE 13-31.  CUMULATIVE MORTALITY DATA OF FEMALE RATS
           EXPOSED BY INHALATION FOR 2 YEARS TO ACRYLONITRILE VAPORS*
                               (Quast et al. 1980b)
Days on Test
          Exposure Concentration
 0 ppm          20 ppm           80 ppm
0-30
31-60
61-90
91-120
121-150
151-180
181-210
211-240
241-270
271-300
301-330
331-360
361-390
391-420
421-450
451-480
481-510
511-540
541-570
571-600
601-630
631-660
661-690
691-720
721-735
1
1
1
1
1
1
1
1
3
5
5
7
9
11
14
14
19
26
34
36
50
63
66
71
78
0
0
0
0
0
0
1
1
1
1
2
2
3
5
10
14
22
31
36
43
54
70
81t
88t
91t
0
0
0
0
0
0
1
2
4
6
9
11
19t
27t
33t
41t
57t
71t
80t
88t
94t
98t
98t
99t
99t
Total Number of Rats
100
100
100
    *Data listed as number dead which is equal  to percent dead.

    tSignificantly different from control  by Fisher's Exact Probability Test,
P < 0.05.
                                          13-109

-------
            TABLE 13-32.  TUMOR INCIDENCE IN SPRAGUE-DAWLEY RATS
                     EXPOSED TO ACRYLONITRILE BY INHALATION
                              (Quast et al.  1980b)
Diagnoses
 0 ppm
     20  ppm
         80 ppm
Males

Zymbal  gland tumors
  of external  ear canal
  (benign and malignant)

Small intestine tumors
  (benign and malignant)

Brain and/or spinal cord
  glial  cell tumors
  (benign and malignant)

Tongue-squamous cell tumors
  (benign and malignant)

Females

Zymbal  gland tumors
  of external  ear canal
  (benign and malignant)

Mammary gland
  fi broadenoma/adenofi broma

Mammary gland
  adenocarcinoma

Mammary gland tumors
  (benign and malignant)

Brain and/or spinal cord
  glial  cell tumors
  (benign and malignant)
2/100
2/99
4/100
2/20
(P
(P
11/100
= 0.009)*
15/98
= 7.03 x.lO'4)*
 0/100
 1/96
 0/100


79/100



 9/100


88/100



 0/100
     4/99                22/99
                (P =  5.71  x lO'8)*
     0/14                7/89
                (P =  0.0251)*
     1/100               11/100
                (P = 3.65 xlO'4)*

     96/100              75/100
(P = 2.06 x ID'4)*
     8/100


     96/100
     8/100
(P = 0.003)
         20/100
(P = 0.022)*

         85/100
         21/100
(P = 3.32 x ID'7)*
    *P values calculated by Fistier" Exact P>obabiTfty~Test.
                                          13-110

-------
    In summary,  in the Dow study acrylonitrile  induced  a  statistically



Significant increased incidence  of  tumors  in  male  and female  rats  following



exposure by inhalation.
                                         13-111

-------
CELL TRANSFORMATION STUDY



    Parent and Casto (1979) reported the effect of acrylonitrile on Syrian



golden hamster embryo cells (HEC), and found that acrylonitrile (ACN)



transforms cells in culture and enhances the transformation of cells



previously affected with simian adenovirus SA7, a colony transforming



oncogenic virus.



    ACN from Aldrich Chemical  Co. (Milwaukee, Wis.)  was used in this



experiment.  The purity was stated to be greater than or equal  to 99%.   The



impurities included about 0.3% water, less than 0.5% acetonitrile,  and  30  to



45 ppm l-hydroxy-4-methoxybenzene.  In this experiment cultures of  primary



Syrian golden HEC were prepared by trypsinization of decapitated and



eviscerated embryos after 13-14 days of gestation.  Cells were plated  into



60-mm-diameter Lux plastic dishes at a density of 5  x 106 cells/dish with



modified Dulbecco's medium and 10% fetal  bovine serum (Reheis Chemical  Co.,



Kankakee, 111.) and incubated  at 37°C for 3 days in  5% C02.   ACN was



dissolved in 100 mg acetone/ml and diluted in complete medium to give the



final  concentrations.



    In this viral transformation enhancement assay,  HEC were exposed to ACN in



concentrations of 0, 25, 50,  100, and 200 ug/ml.   Treatment  of HEC  with ACN



for 18 hours before SA7 inoculation resulted in only slight  but significant



enhancement to 1.8-fold (Table 13-33).  When cells were treated with ACN 5 hours



after they were inoculated with virus (Table 13-33), a significant enhancement  of



8.9- and 8.4-fold was  observed at 200 and 100 ug ACN/ml,  respectively.



Treatment with 200 ug  ACN/ml  reduced the  cloning efficiency  to less than 10%,



but the number of SA7  foci  only decreased from 41 in control  to 26  in treated



cells.  The enhancement found  when cells  were chemically treated after  virus



inoculation was observed with  several  other chemicals.




                                         13-112

-------
TABLE 13-33.
ENHANCEMENT OF  SA7  TRANSFORMATION BY  TREATMENT  OF  HEC  WITH  ACN*
                (Parent  and Casto 1979)
   Time  of  ACN
    treatment
          ACN
          ug/ml
Surviving      SA7         Enhancement
fractiont      foci§      ration
18 hr before SA7




18 hr before SA7




5 hr after SA7




200
100
50
25
0
200
100
50
25
0
200
100
50
25
0
0.08
0.34
0.62
0.75
1.00
0.18
0.46
1.07
1.02
1.00
0.07
0.21
0.60
0.69
1.00
1
32
41
45
50
6
19
31
37
20
26
75
37
41
41
0.3
1.8
"O
1.2
1.0
1.6
2.1
1.5
1.8
TTfJ
8.9
F^f
TT5
1.4
1.0
      *Chemical  dilutions  were  added  to mass  cultures  of  HEC  18  hr  before  or  5  hr
   after  treatment  with  SA7.  Virus was absorbed  3 hr, and the cells were
   transferred  for  survival  (500-700  cells/dish)  and for  transformation assays
   (200,000-300,000 cells/dish).

      tDetermined from plates receiving 500-700 cells.  Number of colonies  from
   virus-treated and chemically-treated cells was divided by  the number of
   colonies  from virus-inoculated  control cells to give the surviving  fraction.
   Cloning  efficiency of control cells was  10-15%.
      §Number of  foci  from  106  plated  cells.
      ^Enhancement  ratio  was  determined  by  dividing  the  TF  of  untreated  cells  (TF
   = SA7  foci  x  reciprocal  of the  surviving fraction)  by that  obtained from
   control  cells.   Underlined values  are statistically significant  at the  5%
   level.
                                            13-113

-------
    When HEC were treated for 6 days with ACN (chemical  transformation)
without added virus, foci of morphologically transformed cells were observed
that were similar to those described previously with known chemical
carcinogens.  At 100 ug ACN/ml, three foci  were observed on nine dishes and
two foci on six dishes at 50 ug ACN/ml; BP treatment resulted in three foci on
four dishes at 1.25 ug/ml and two foci  on ten dishes at  0.62 ug/ml  (Table 13-34)
No foci were observed on medium or solvent control  dishes.
    The observation that acrylonitrile transforms cells  adds support to the
animal and human evidence that acrylonitrile may be carcinogenic.
                                          13-114

-------
                 TABLE 13-34.  TRANSFORMATION OF  HEC  BY  ACN
                           (Parent and Casto 1979)
Treatment*
ACN



BP

Control
ACN
ug/ml
100
50
25
12
1.25
0.62
—
Surviving
fractiont
0.06
0.76
0.84
1.06
0.78
0.94
1.00
Foci /dishes
3/9
2/6
0/6
0/5
3/4
2/10
0/7
     *Chemicals were added to tertiary HETT plated 24 hr earlier with 50,000
(transformation) or 1,000 (survival) cells/dish.   Fresh medium with chemical
was added after 3 days and removed after 6 days.   Colonies  were fixed and
stained at 9 days for survival  assays; focus assays for transformation were
done 25 days after treatment was indicated.

     tDetermined from dishes receiving 1,000 cells.  Number of colonies from
treated plates was divided by the number from control  dishes.
                                         13-115

-------
13.5.2  EPIDEMJOLOGIC STUDIES

O'Berg (1980)

   There are no community epidemiologic studies available that  attempt  to

demonstrate an association between exposure to ambient  levels of  acrylonitrile

and the development of disease.   However,  an occupational  epidemiologic study

that involved workers exposed to acrylonitrile at  a Du  Pont  textile  fibers

plant in Camden, South Carolina  was conducted by Maureen T.  O'Berg  of E.I.  du

Pont de Nemours and Co., Inc. (O'Berg 1980).  The  study cohort  was  1,345 male

employees "identified as having  had potential exposure  to acrylonitrile at

some time between start-up in 1950 and 1966."  The 1966 cutoff  date  allowed

for a minimum 10-year follow-up  through the end of 1976.  Expected  numbers

were stated to be based both on  company and national  rates.   The  analyses

presented in the paper, however, derive expected number of cases  based  only on

company rates, ignoring the possible effects of other chemicals on  this

"control" cohort.*  The analyses consider  calendar time, payroll

classification, occupation, duration of exposure,  latency, and  severity of

exposure.  The severity of exposure levels were designated as high,  medium,

and low.   Du Pont representatives agreed that 20 ppm, 10 ppm, and 5  ppm might

be used to represent the designated classification of high,  medium,  and low

exposure levels.  This was documented in a trip report  by Jane  Brown,

Industrial Hygienist, National  Institute for Occupational  Safety  and Health

(NIOSH 1978).

    Overall, 25 cases of cancer  occurred,  with 20.5 expected based  on company
    *While this paper also attempts to compare observed  and expected  deaths,
there is an inconsistent result.   According to O'Berg, 4.4 respiratory  cancer
cases were expected in this cohort using the Du Pont  controls.   However,  using
the same controls, O'Berg predicts 6.1 respiratory cancer deaths.   A  telephone
conversation with O'Berg confirmed that this inconsistency was  due to
different methods of following the cohort for cases and  deaths.   Cases  were
described only for active employees, while deaths included retirees.


                                          13-116

-------
rates.  Of these 25 cases, eight were respiratory cancer cases versus 4.4

expected.  Excesses were found primarily among wage roll  employees who had

worked during plant start up, 1950 to 1952, and had been exposed for at least

6 months.  For these employees there were 8 cases of respiratory cancer versus

2.6 expected (P < 0.01).  Furthermore, most of this above excess occurred

during the latest follow-up period, 1970 to 1976, when there were six cases of

respiratory cancer versus 1.5 expected (P < 0.01).  Total cancer cases in this

latest follow-up period for this group were also significant,  17 versus 5.6

(P < 0.01).

    A trend toward increased risks was seen not only with increased follow-up

time but also with severity of exposure.  In wage roll workers with at least

moderate exposure and probable latent period of at least 15 years, the

observed and expected numbers of cancer cases was 13 and 5.5,  respectively.

Furthermore, half of this excess cancer was due to respiratory cancer, 5

versus 1.4 (P < 0.05).  Thus, this study provides some evidence that acryloni-

trile is carcinogenic to humans.


    However, because of the known relationship between smoking and lung

cancer, further analysis is attempted concerning the role of smoking behavior

in these findings.  (Seven of the eight lung cancer cases were reported to be

smokers by their supervisors or associates; the eighth was unknown.)  Dr.

Bruce Karrh of Du Pont stated that there were pathology slides for five of the

eight respiratory cancer cases.  He further stated that these  were identified

as four squamous cell  carcinomas and one oat cell  carcinoma.   Of the three

remaining respiratory cancers cases for whom slides were unavailable, he

stated that two were bronchogenic and the other unknown.   These cell types are

generally believed to be associated with both chemicals and smoking by most

pathologists.
                                         13-117

-------
    In an attempt to investigate the impact of smoking  on  the  risk  of

developing lung cancer in this population,  the Du  Pont  Company  provided

additional data (letter from Sidney Pell, Du Pont, July 23,  1980) to the

Carcinogen Assessment Group (CAG) regarding the smoking habits  of 32 of the  36

cancer cases reported on in this plant (some of these were not  in the  study

cohort), as well as data on the smoking habits of  a matched  group of non-cases

in the plant.  Of the 32 cancer cases, 22 were cancers  other than lung, and  16

or 73% of the nonlung cancer cases were smokers.   The smoking habits of the

matched group of 36 noncancer controls from the same plant were also provided.

They were matched on a three to one basis to certain selected cancers*

occurring in the study group on the basis of age,  payroll  classification,  date

of first exposure, and date of termination.  It was found  that  25 or 69% of

the 36 cases were smokers in this group.

    Based on this information, we can estimate that 70% of the  plant

population were smokers, and 30% nonsmokers.  Of the 70% who were smokers, we

will assume that 50% were "moderate" smokers while 20%  were  "heavy" smokers

based on figures by Axelson (1978).  We will assume also that the relative

risk of lung cancer is 1, 10, and 20, respectively, for the  nonsmokers,

moderate smokers, and heavy smokers (Doll and Hill 1952, Hammond 1975).  Using

the method of Axelson (1978) to adjust for smoking differences  in cohorts, the
    *TFe author selected only cancer cases that have a  high  correlation  with
cigarette smoking for his case control  study.   The purpose was  to test whether
cigarette smoking was different in the  matched non-cases  selected from the
same plant compared to the cases.  These cases were eight lung  cancers,  one
esophageal, one nasopharygeal, and two  bladder cancers.   The resulting test
statistic (X^) was nonsignificant, indicating  that there  was no reason to
suspect that there was an uneven distribution  of cigarette smoking in the
cases compared to the matched non-cases.

                                         13-118

-------
relative incidence of lung cancer in the plant population is








    (1)       Ip = 1(0.30)I0 + 10(0.50)I0 + 20(0.20)I0 = 9.3 I0








where I0 equals the risk of lung cancer in a nonsmoking population.



    Based on a nationwide survey (U.S.  Health, Education, and  Welfare  1973)



which found that 62% of the male blue collar workers smoked in 1966, we assume



that the overall company population had 40% nonsmokers.   In addition,  we



assume that the distribution of moderate and heavy smokers is  the same as



above, so that moderate and heavy smokers constitute 43% and 17% of the



population respectively.  The computed  incidence of lung cancer in this



comparison population is then







    (2)        Ig = 1(0.40)I0 + 10(0.43)I0 + 20(0.17)I0 = 8.1  IQ







where Iq is the incidence in the company population.  Hence, the relative



contribution of smoking to the risk of  lung cancer in the study cohort is  the



ratio of the two incidence rates:







    (3)                Ip/lg = 9.3 I0/8.1 I0 = 1.15







Therefore, because of the slightly higher proportion of  smokers in the study



cohort relative to the reference company population, the number of respiratory



cancer cases would be about 15% higher  than the 1.4 cases expected without



considering smoking differences, or 1.4 x 1.15 = 1.61 cases.  Assuming a



Poisson distribution of cases, the probability of  seeing 5 cases or more when



only 1.61 are expected is only 0.024.   Therefore,  after the adjustment for




                                         13-119

-------
smoking differences, the respiratory cancer rate in the study cohort exposed



to acrylonitrile for over 6 months is significantly higher  than  that of  the



reference population.



    In an attempt to assess whether the excess respiratory  cancer rates  could



be ascribed to a higher ratio of heavy to moderate smokers  in the



acrylonitrile-exposed cohort than that in the reference population,  the



following calculation was done.  In equation (1) for Ip the ratio of heavy



to moderate smokers was increased from 0.2/0.5 = 1/2.5, keeping  the  total



fraction of smokers at 0.7, until the point was reached where the probability



of observing five cases or more giving the expected number, 1.4  x (Ip/Ig),



was increased from 0.024, as above, to 0.05.  The result is that in  order for



a higher ratio of heavy to moderate smokers in the cohort to account (at the



0.05 level  of significance) for the observed excess respiratory  cancer rate,



that ratio would have to be 1.6/1.  Our judgment is that such a  marked excess



of heavy versus moderate smokers would not likely occur in  the acrylonitrile



workers, since for blue collar workers generally there  are  about 2.5 times



more moderate than heavy smokers.



    In conclusion, the observations by O'Berg of a statistically significant



excess of respiratory cancer in workers exposed to acrylonitrile and followed



up for more than 10 years constitute significant evidence that acrylonitrile



is likely to be a human carcinogen, although smoking, at least as a



contributing factor, cannot be completely ruled out at  this time.







Delzell and Monson (1982)



    Delzell and Monson (1982) conducted a cohort mortality  study of  327  white



male employees of a rubber manufacturing plant in Akron, Ohio who were



employed for at least 2 or more years between January  1, 1940 and July 7, 1971





                                         13-120

-------
in two areas with potential  exposure to acrylonitrile.   The  first  area was  a
nitMle rubber manufacturing operation.  Butadiene  and  styrene  were  present in
this area between 1938 and 1949 while vinyl  pyridine  has been present since
1949.  The latex that was produced in closed vessels  was "blown down" into
tanks after excess monomers   were removed.   In  the  second area, latex was
coagulated and dried into solid rubber.  From 1938  to 1960,  the nitrile  rubber
was washed on an open mill and dried in vacuum  ovens.  After 1960, this  job
was done by extruder drying.  Delzell and Monson  did  not report the  levels  of
acrylonitrile that these workers were exposed to  while  on the job.
    Follow-up was continued  until  July 1, 1978, according to the authors, but
the extent of success is not indicated.  The authors  state that the  follow-up
was "identical" to that of a recently published study by the same  authors.  The
authors do indicate that anyone without a death record  prior to July 1,  1978
was assumed alive on that date.  Expected deaths  were generated based on U.S.
age and calendar time specific white male mortality.   Overall,  74  deaths
occurred in this group compared to 89.5 expected, a deficit  readily
attributable to the healthy  worker effect.   On  the  other hand,  the risk  of
death due to cancer was somewhat elevated with  22 observed versus  17.9
expected.  Most of this excess was due to lung  cancer deaths which were
nonsignificantly higher with 9 observed versus  5.9  expected. Only in workers
employed for 5 to 14 years and followed for  at  least  15 years,  did a
significantly elevated risk  of lung cancer appear  (4  observed versus 0.8
expected P <0.01).  However, if workers who  were  employed for more than  14
years and followed for at least 15 years are added  to the latter group,  not an
unreasonable consideration since presumably  they  are  most exposed, the  risk of
lung cancer was actually reduced but still significant  (4 observed versus  1.4
expected, P <0.05).
                                         13-121

-------
    The excessive risk of lung cancer in this group may not be entirely  due to



exposure to acrylonitrile, but may be partially reflective of exposures  to



other substances within the same environment, i.e., the author reports that



one lung cancer victim had worked for 17 years in the curing division  of the



company prior to employment in the acrylonitrile area, but the authors did not



elaborate further.  Additionally, no discussion of the effects of  smoking on



lung cancer is included.



    However, the possibility that the excess  risk of lung  cancer is  due  to



acrylonitrile cannot be dismissed.







Thiess et al. (1980)



    A cohort mortality study of 1,469 workers from 12 factories located  in the



Federal Republic of Germany was conducted by  Thiess et al. (1980).   These 12



factories are part of the BASF company.   The  BASF company  produces no



acrylonitrile but buys acrylonitrile in  order to produce styrene-acrylonitrile



and acrylonitrile butadiene-styrene polymers  as well  as organic intermediate



products.  The processing methods differ from factory to factory.  No  historic



exposure data for acrylonitrile exists according to the authors.



    The population at risk was defined as all workers who  were employed  for



over 6 months in acrylonitrile processing presumably from  time of  first  use of



acrylonitrile (around 1956) until the cut-off date of May  15, 1978.   Included



were 1,081 German workers and 338 "foreigners" (nationality not given).



Follow-up was 98% complete on the German workers but only  56% complete on the



foreign segment.  This left about 170 lost to follow-up or about 12% of  the



work force.



    Expected deaths were generated based upon mortality in three areas of



Germany, the city of Ludwigshafen, the state  of Kheinhessen-Pfalz, and the






                                          13-122

-------
Federal Republic of Germany (FRG).   A total  of 89 deaths  were  observed



compared to 92.3, 96.4, and 99.0 expected deaths, respectively,  in  the  three



geographic entities cited.   Twenty-seven deaths from malignant tumors were



observed compared to 20.5 expected  based upon  rates  of  the  FRG.   The most



striking finding was a statistically significant higher risk of cancer  of the



lung (11 observed versus 5.65 expected based on FRG  rates,  P < 0.05; or 5.92



expected based on rates of  Rheinhessen-Pfalz,  P < 0.05).   The  authors



recalculated their results  for the  FRG only, with the 78  members of one



factory excluded from the group of  12 factories whom they said had  "contact



with substances since proven to be  carcinogenic." Not only  did an excessive



significant risk remain with respect to lung cancer  (9  observed versus  4.37



expected based on FRG rates, P < 0.05), but  also a significant excess risk of



cancer of the lymphatic system was  seen (4 observed  versus  1.38 expected,



P < 0.05).



    These results are questionable.  First,  the members of  this cohort  were



apparently exposed to a number of different  carcinogens,  i.e., vinyl chloride,



distillation residues containing polycyclic  hydrocarbons, cadmium,



B-napthylamine, dimethylsulfate, and epichlorohydrin.  Several of these have



been associated with a higher risk  of lung cancer.  Second, all  lung cancer



victims were found to be smokers, while not  all  lung cancer victims were



proven to be exposed to acrylonitrile, according to  the author.



    On the other hand, the  lung cancer risk  estimated from  this study may



actually be an underestimate of the true risk  for the following reasons:



    1.  Lumping together workers from 12 different factories who may be



subjected to differing levels of exposures to  acrylonitrile could have  led to



an underestimate of risk by the inclusion of minimally  exposed or unexposed



members.






                                         13-123

-------
    2.  The healthy worker effect would contribute to an underestimate of



risk.



    3.  Insufficient follow-up on a relatively youthful  cohort did not allow



enough latency in the segments of the cohort most likely to exhibit a  true



estimate of risk.  Only 447.1 person-years were accumulated in members over 64



years of age.



    4.  Underascertainment of vital status may have resulted in an undercount



of observed deaths.  Twelve percent were lost to follow-up.  Person-years have



been accumulated for each of the lost individuals until  the last date  known to



be alive.



    Although, there are many limitations to this study,  it is possible that



exposure to acrylonitrile may indeed be related to an excessive risk of lung



cancer and cancer of the lymph system as well as to other carcinogens  present



in the workplace.







Werner and Carter (1981)



    Werner and Carter (1981) conducted a cohort mortality study of 934 men who



worked on the polymerization of acrylonitrile and spinning of acrylic  fiber at



six different factories sometime between 1950 and the end of 1968 for  a



minimum of one year. Two of the six factories were located in Scotland and



Northern Ireland; the remainder were located in England  and Wales.  The cohort



was followed through the end of 1978, which resulted in  a vital status



ascertainment near 100%.  Expected deaths were calculated based upon standard



age-specific mortality rates for England and Wales combined.



    Overall, only 68 deaths had occurred through 1978 where 72.4 were



expected. Twenty-one were due to malignant neoplasms versus 18.6 expected.



Deaths from cancer of the stomach were statistically significantly elevated
                                         13-124

-------
over all age groups (5 observed versus 1.9 expected P < 0.05)  with  deaths  to



persons in the 55-64 age group contributing the largest portion  (3  observed



versus 0.7 expected, P < 0.05). On the other hand,  a statistically  significant



elevated risk of cancer of the lung, trachea,  and bronchus  appeared only  in



the age group 15-44 (3 observed versus 0.7 expected, P < 0.05) but  not  in  any



other age group.



    In a nested (within the larger study) small case control  study, the three



lung cancer cases were compared with controls  (born the same  year and  living



beyond the date of death of the case) with respect  to duration of employment.



In one case the duration of exposure was  significantly greater than the



average for the matched controls.   In the other two cases,  the differences



were not significant.   The result  are equivocal at  best.



    There are many difficulties with this study, not the least of which is the



lack of data regarding actual  measured levels  of exposure to  acrylonitrile



during the period 1950 thru 1968,  according to the  author.   This  lack  of



quantitative data makes it difficult to distinguish between plants  in  terms of



their relative exposure levels.  Implicit in the assumption of lumping  the



employed populations of six different factories together is that they  are  all



relatively similar with respect to levels of exposure.  This  is  usually not



the case, however, for some factories tend to  be better controlled  than



others.



    Other problems with this study include the relatively short  follow-up  in



the small subgroup of  the cohort where one would expect to  have  the greatest



risk,  i.e., the 158 men who had the earliest exposure in 1950-58.   This group



was followed for a minimum of  20 years accumulating a total  of 3,241



person-years of which  only 780 person-years were to individuals  age 55  and
                                         13-125

-------
over.  Like several other studies, this is a relatively youthful  cohort and



the number of expected deaths are too small  to provide confidence in the



results.  The findings of this study are only suggestive due to small  numbers,



youthfulness of the cohort, insufficient follow-up, and lack of consideration



of effects of smoking.  However, this study should be followed several



additional years to see if the significant excesses of lung and stomach cancer



found in this study remain high.








Monson (1978)



    An epidemiologic study was conducted by Richard R. Monson for the  B.F.



Goodrich Company and the United Rubber, Cork, Linoleum, and Plastic Workers of



America (Federal Register 1978).  This retrospective cancer morbidity  and



mortality study included some workers with exposure to acrylonitrile.   Among



workers with potential exposure to acrylonitrile (extent of exposure not



known), there was a slight overall excess in the number of deaths from cancer



observed compared to the number expected (the data used for determining the



expected number were not specified):  lung cancer deaths (7 observed versus



4.4 expected); genitourinary cancer deaths (2 observed versus 1.6 expected



deaths, and 6 observed versus 3.1 expected incidences); and Hodgkin's  disease



deaths (2 observed versus 0.3 expected).  Deaths from all  causes  (cancer and



noncancer) were not significantly elevated in the worker study group.   Because



the workers in this study had potential for exposure to other carcinogens,  the



Occupational  Safety and Health Administration concluded that the  study could



not be used to support an association between acrylonitrile exposure and human



cancer development (Federal  Register 1978).
                                         13-126

-------
Zack (1980. unpublished)



    Judith A. Zack (unpublished report 1980)  of the Monsanto  Chemical  Company



investigated the mortality experience of 76 white males  at  the Texas  City



facility and 276 white males at the Decatur facility who were exposed  to



acrylonitrile for a minimum of 6 months prior to January 1, 1968 and  followed



until December 31, 1977.  A total  of 15 deaths, all  causes, were found to  have



occurred during that interval  of time, 9 at the Texas City  facility and 6  at



the Decatur facility.  Of the  15 deaths, 3 were attributable  to cancer, at all



sites.  Only one was due to respiratory cancer.  The author expected  18.11



deaths, all causes, based upon the Monson life table technique with 2.80



malignant neoplasms expected.   The author concluded that although the  number



of deaths is small, the observed numbers do not differ significantly  from  the



expected numbers.



    This study is relatively insensitive in its ability  to  assess a cancer



risk in this cohort of acrylonitrile workers  for two reasons.  First,  because



of the limited cohort size, the small  numbers of expected deaths (18.11 due to



all causes, 2.8 due to malignant neoplasms, 0.84 due to  respiratory



malignancy) do not provide enough  information to place any  confidence  in the



statement of nonsignificance made  by the author.  Second, the usual latent



period for cancer of most sites is in  excess  of 15 years.  Although the study



design guaranteed at least 10  years of lapsed time since onset of employment



to the cutoff date, still 52%  or 183 of 352 members  of the  cohort were



observed for no longer than 15 years since initial  employment, a time  frame



which is probably insufficient for a carcinogenic potential to manifest



itself, if, in fact, one is present.  This report cannot be considered



supportive of a "negative" risk assessment of acrylonitrile with respect to



its cancer-causing potential in humans.




                                         13-127

-------
Gaffey and Strauss (1981, unpublished)



    The Gaffey and Strauss study (1981, unpublished) is a cohort mortality



study of 326 white males who were potentially exposed to acrylonitrile in a



chemical plant in Decatur, Alabama.   This plant was also included in the



earlier study by Zack (unpublished).  The cohort was defined to be all white



males who had achieved 6 months of employment at the plant between April 1952



and December 1953.  Follow-up for vital status ascertainment was continued to



the end of 1977 and was 95% complete according to the author.  However, it was



noted that 73 who were found "alive" by retail credit check were subsequently



assumed to be "alive" by virtue of the fact that their names appeared in a



current telephone directory.  Such an assumption may be unwarranted without



actual confirmation with the study member.  Many widows keep their telephone



listing in their husbands' names for security reasons.  Another 15 remained



with an unknown vital status.



    Only 26 deaths were observed during this time frame resulting in an



unusually low standard mortality ratio (SMR) overall of 47, while for



malignant neoplasms the SMR was only 37.  Only 4 of the 26 observed deaths



were attributable to cancer.  And of these four only two were cancer of the



lung (expected = 3.74), while the remaining two were cancer of the kidney



(expected = 0.3).  Workers in this study could not be classified by level of



exposure according to the author.  However, the cohort was first dichotomized



into hourly versus salaried employees and then into maintenance versus



everyone else.  No significant excessive risks were evident from the data



presented in any of the groups mentioned not even after a 20-year follow-up.



The author notes that the small number of malignancies observed make any



conclusions at best tentative.
                                         13-128

-------
    Furthermore, the definition of the cohort was  not  restrictive  enough  to
Identify only an "exposed" population for study.   The  author  points  out that
21 of the members of his present cohort were also  members  of  the earlier
cohort study by Zack.  Zack was careful  to include only  white male employees
who "worked in an acrylonitrile manufacturing or  processing area"  for  a
minimum of 6 months prior to January 1, 1968).  In fact, during the  period
1952-53 Zack noted that only 17 employees of the  Decatur plant could qualify
for inclusion into her cohort of 352 males,  while  Gaffey counted everyone who
was there at start-up (326 white males) including  95  salaried employees.
Hence, there is a distinct likelihood that Gaffey's cohort consists  of large
numbers of unexposed employees.  Thus, in addition to  the  healthy  worker
effect and incomplete follow-up, it is likely that the relative insensitivity
of this analysis to detect a significant risk is  due  to  1) the relatively
youthful nature of the study group (average  age at the end of follow-up was
only 54.9 years), 2) the inclusion of large  numbers of unexposed or  minimally
exposed employees in the cohort, and 3) the  possibility  that  large numbers of
workers were classified erroneously as "live" via  the  telephone directory
check.  These problems make this study questionable with respect to  the
detection of a cancer risk.

Kiesselbach et al. (1980. unpublished)
    This study appears to be a cohort mortality  study  of some 884  male workers
who were presumably exposed to acrylonitrile for  a minimum of 1 year during
the period from 1950 to August 1, 1977 in 16 different plants (pilot,
laboratory, and production) of Bayer's Leverkusen  Division in West Germany.
Workers selected for inclusion into the cohort were involved  in the  production
and processing of acrylonitrile.  Follow-up was  complete on  93.2%  of the
                                          13-129

-------
cohort.  Only 60 persons were without a vital  status  ascertainment  by  the
close of the study.  The male population of the German  state  of
Nordrhein-Westfalen was used as a reference population,  although  no indication
is given that any of the 16 plants are located in  that  state.  A  total  of  58
persons died in this cohort compared to 104.3  expected.   By contrast,  deaths
due to malignant neoplasms ("Tumors" as they are referred to  in the
translation) were very close to expected (20 observed versus  20.4 expected).
Similarly, 6 respiratory "tumors" were observed versus  6.9 expected.   No
excessive risks appeared even considering "risk time" after 10 years and 15
years.  With respect to lung tumors after a 15 years  "risk time," there were  2
observed versus 5.6 expected (assuming "risk time" is the translators  word for
follow-up).   The authors noted that there were no  existing data on
acrylonitrile concentration in the air in any  work areas over the period of
the study.
    There are several  problems with this study.  One  of  the more  important
concerns was the selection criteria for inclusion  of  male workers in the
cohort.  Originally the authors had in their possession  lists containing 1973
names of former and active employees of the company.  These lists were
submitted to present and former plant managers, supervisors,  and  foremen to
determine the degree to which each person on the list was exposed to
acrylonitrile and the periods of time covered  by this exposure.   To be
included each worker must have worked more than 12 months in  one  of the 16
plants and either worked at least 30% of his time  with  acrylonitrile,  or if
less than 30%, he had to have been under a heavy acrylonitrile burden  such as
would be found in test plants, according to the authors.  Over a  thousand
persons were eventually excluded from the cohort through this review process.
                                         13-130

-------
    Additionally, a justification is needed for the choice of  the
Nordrheim-Westfalen state as a comparison population.   The location  of  the  16
factories with respect to the comparison population is  not given.
Furthermore, without exposure data it is impossible to  tell  how  similar each
of these plants are with respect to the extent of  exposure to  acrylonitrile.
Also, the study cohort was a relatively youthful group  even  after  the close of
the study.  Only 13% of the total person-years accumulated were  in the  age
category 55 and over, the category in which the greatest  increase  in lung
cancer mortality could be expected to occur.  In individuals under age  55,
lung cancer is a disease which appears infrequently and a  rather large  number
of person-years accumulating to such persons would be necessary  to allow the
detection of a significant risk if in fact one exists,  and only  after a
sufficiently lengthy latent period has passed.
    Furthermore, inconsistencies in the tabular data presented by  the authors
cast doubt on the validity of the presented findings.   Two examples  are :   1)
the calculation of expected deaths in calendar-time periods  where  none  should
appear based on the tabular heading description, and 2) dates  of onset  of
cancer in individuals who exhibit more exposure to acrylonitrile than their
ages warrant (one individual  was exposed 12 years  before  he  was  born).   In
short, it appears that this study suffers from a general  lack  of critical
review by its author.
    Therefore, because of the problems and inconsistencies in  this study, the
results do not provide evidence to support a lack  of a  cancer  risk to
acrylonitrile-exposed employees and cannot be used to refute the results of
the O'Berg study as was suggested by the author.
                                         13-132

-------
    This selection can be affected by recall  bias on  the part  of  the
reviewers.  They are more likely to better remember details  regarding  the
working status and working conditions of actively employed persons  than  of
inactive and former employees.   Managers,  supervisors,  and foremen  frequently
cannot recall even the names of former employees.  Active employees, who have
better survivorship, are likely to be overrepresented as a result.  The
combined total effect of such a biased selection, the influence of  the healthy
worker effect, and the lack of  a complete  follow-up,  such as occurred  in this
study, could have led to a substantial underestimation  of risk.
    Another major concern involves the computation of expected deaths  in this
study, i.e., the choice of an appropriate  comparison  population as  well  as  the
proper allocation of person-years into the appropriate  risk categories as the
members of the cohort pass through those categories in  time.   The greatest
deficit of lung cancer deaths compared to  expected appeared in members of the
cohort who attained an age of 60 and over.  With  respect to all causes
combined,  the greatest deficit  occurred in persons 45 and over.   Only  937
person-years were generated in  the age group  60+  in this cohort (out of  13,375
person-years altogether), but the authors  calculate that 3.708 lung cancer
deaths could have been expected to occur versus only  1  observed death  in this
age group.  This deficit of deaths in persons  age 60  and above was  reversed in
persons under age 60 (5 observed versus 3.23  expected).   Such  a sudden
reversal  of risk ratios leads one to suspect  that a methodologic  problem
exists in  the analysis.  Since  the greatest risk  of deaths occurs to older
persons and is reflected in the death rates based on  that age  group, either an
improper allocation of person-years to older  age  categories or overestimates
of the death rates in the comparison population would have the effect  of
adding additional  expected deaths to that  cause and would result  in an
underestimate of the true risk.
                                         13-131

-------
compared to 92 expected based on the above method.   Eleven of these were



"neoplasms" versus 16.13 expected.   Only 3 lung cancers  were  found  versus  5.34



expected.  This rather striking shortfall of deaths compared  to expected can



be attributed only partially to the healthy worker  effect  and to the fact  that



no cause of death was found for 4 of the 59 deaths.  More  importantly,  at  the



close of the study on December 31,  1977, few members of  the cohort  had  reached



age 60 and beyond (less than 9%), the age category  in which the greatest



mortality is likely to occur.  In addition, if latent factors are also



considered, then it is unlikely that a carcinogenic effect would be seen in



this relatively youthful group.  Time since onset of employment was not



considered, although the author himself suggested another  10  years  of



surveillance would be required before any exposure-related deaths would occur.



No exposure data are provided.  This study provides little evidence to  support



the premise that exposure to acrylonitrile does not increase  the risk of



cancer.







Stallard (1982, unpublished)



    Stallard (1982, unpublished) conducted a cohort mortality study of  419



white male employees of an oil company facility in  Lima, Ohio.  The production



of acrylonitrile began at this plant March 14, 1960.  Members of the cohort



had to have been employed at the facility at some time during the period March



14, 1960 to March 14, 1980 in a job that was assessed by plant industrial



hygiene personnel as having potential exposure to acrylonitrile.  Follow-up



through March 14, 1980 was complete on 92% of the cohort (33  remained lost to



follow-up) expected deaths were generated through the utilization of 1970



age-cause specific death rates in white males for the United  States and



separately for the State of Ohio.






                                          13-134

-------
Herman (1981, unpublished)
    A cohort mortality study of 989 male employees in two rubber plants at
Baton Rouge and Scotts Bluff, Louisiana was conducted by D.  R.  Herman (1981,
unpublished).  There were 799 wage and 190 salaried employees  having  a minimum
of 1 year service in the plants during  the period 1951  through 1977.  The
author reported that race was known on 85.5% of this cohort  of 989 males;  162
(24.8%) were black, while 654 (60.7%) were white.   Vital status ascertainment
was accomplished on a cohort that included 88 females.   In this larger cohort
97% were successfully followed to the end of 1977.  Only 28  remained  with  an
unknown vital status.  The 88 females were excluded from subsequent analyses
because of the small number.
    Expected deaths were generated separately for  wage  versus  salaried
employees as follows:  for salaried workers, because almost  all  were  white and
a few were unknown with respect to their race, expected  deaths  were generated
utilizing U.S. white male age and calendar-time-specific mortality rates only.
However,  race was not known for 40% of the wage workers.  A  separate
calculation of expected deaths was done for known  whites and blacks by the
method described above.  In the remaining 40% with unknown race, two
calculations of expected deaths were accomplished, first, assuming they were
all  white, and second, assuming they were all  black.  The author indicated
that these represented high and low estimates.  Since the author also knew
that about 75% of the total  cohort was white, a weighted average of these  two
separate  estimates was obtained for this group as  follows:

          Expected deaths (weighted) = 3/4 x (white expected deaths)
                                     + 1/4 x (black expected deaths)
No further explanation is given.
    Overall, 59 observed deaths occurred to the 989 males of the cohort

                                          13-133

-------
    Only 7 deaths occurred during this time span  versus  25.5  expected  based  on
U.S. rates, another surprising shortage of deaths overall.    Of  the  seven
deaths, four were due to malignant neoplasms versus  4.66 expected  based  on
U.S. rates and 4.99 expected based on Ohio rates. Of  these four malignant
neoplasms, two were due to lung cancer versus 1.56 expected based  on U.S.
rates.  Unfortunately, the expected deaths calculated  for lung cancer  based  on
the rates for Ohio are in error (in the age group 45-64, the  author  shows  a
figure 15.4 expected deaths; the correct figure is 1.54  and the  resulting
corrected total number of expected lung cancer deaths  all ages combined  based
on Ohio death rates is 1.78).
    Aside from the biases introduced by the incomplete follow-up of  all
employees (former and present) and the healthy worker  effect, a  major  problem
with this study is the relative youthfulness of the  cohort even  at the end  of
the observation period (only one person in the successfully followed portion
of the cohort of some 386 persons reached age 65).  In fact the  author himself
reported that 87% of the cohort was under age 45  in  1970.  Thus, the relative
youthfulness of the cohort and a follow-up of insufficient duration  did  not
allow an accurate assessment of the latent effects of  the risk of  cancer in
this cohort.  The author himself states the study is "inconclusive"  and  cannot
be used to "support or refute whether or not acrylonitrile is carcinogenic  in
humans."
                                          13-135

-------
13.5.3  QUANTITATIVE ESTIMATION



Introduction



    This quantitative section deals with the unit risk for acrylonitrile in  air



and water and the potency of acrylonltrile relative to other carcinogens that



the CAG has evaluated.  The unit risk estimate for an air pollutant  is  defined



as the lifetime cancer risk occurring in a hypothetical  population  in which  all



individuals are exposed continuously from birth throughout their  lifetimes to a



concentration of 1 ug/m^ of the agent in the air which they breathe.   This



calculation is done to estimate in quantitative terms the impact  of  the agent as



a carcinogen.  Unit risk estimates are used for two purposes:   1) to compare the



carcinogenic potency of several  agents with each other,  and 2)  to give  a crude



indication of the population risk which might be associated with  air or water



exposure to these agents, if the actual exposures are known.







13.5.3.1  Procedures for Determination of Unit Risk



    The data used for the quantitative estimate is one or both  of two types:



1) lifetime animal  studies, and 2) human studies where excess cancer risk has



been associated with exposure to the agent.  In animal  studies  it is assumed,



unless evidence exists to the contrary, that if a carcinogenic  response occurs



at the dose levels used in the study, then responses will  also  occur at all



lower doses with an incidence determined by the extrapolation model.



    There is no solid scientific basis for any mathematical  extrapolation model



that relates carcinogen exposure to cancer risks at the  extremely low



concentrations that must be dealt with in evaluating environmental hazards.  For



practical  reasons such low levels of risk cannot be measured directly either by



animal experiments  or by epidemiologic studies.  We must,  therefore, depend  on



our current understanding of the mechanisms of carcinogenesis for guidance as to
                                        13-136

-------
which risk model to use.  At the present time the dominant  view of the
carcinogenic process involves the concept that most agents  that cause cancer
also cause irreversible damage to DNA.   This position  is  reflected by the  fact
that a very large proportion of agents  that cause cancer  are also  mutagenic.
There is reason to expect the quanta!  type of biological  response  that  is
characteristic of mutagenesis is associated with a linear non-threshold
dose-response relationship.  Indeed, there is substantial evidence from
mutagenicity studies with both ionizing radiation and  a wide variety of
chemicals that this type of dose-response model  is the appropriate one  to  use.
This is particularly true at the lower  end of the dose-response curve;  at  higher
doses, there can be an upward curvature probably reflecting the effects of
multistage processes on the mutagenic  response.   The linear non-threshold
dose-response relationship is also consistent with the relatively  few
epidemiologic studies of cancer responses to specific  agents that  contain  enough
information to make the evaluation possible (e.g., radiation-induced leukemia,
breast and thyroid cancer, skin cancer  induced by arsenic in drinking water,
liver cancer induced by aflatoxin in the diet).   There is also  some evidence
from animal experiments that is consistent with  the linear  non-threshold model
(e.g., liver tumors induced in  mice by  2-acetylaminofluorene in the large  scale
EDgi study at the National Center for  Toxicological  Research and the
initiation stage of the two-stage carcinogenesis model in rat liver and mouse
skin).
    Because it has the best, albeit limited,  scientific basis of any of the
current mathematical  extrapolation models, the linear  non-threshold model  has
been adopted as the primary basis for  risk extrapolation  to low levels  of  the
dose-response relationship.  The risk estimates  made with this  model should be
regarded as conservative, representing  the most  plausible upper-limit for  the
                                        13-137

-------
risk, i.e., the true risk is not likely to be higher than  the  estimate,  but  it



could be lower.



    The mathematical formulation chosen to describe the linear non-threshold



dose-response relationship at low doses is the linearized  multistage model.



This model employs enough arbitrary constants to be able to fit almost any



monotonically increasing dose-response data and it  incorporates a  procedure  for



estimating the largest possible linear slope (in the 95% confidence limit sense)



at low extrapolated doses that is consistent with the data at  all  dose levels of



the experiment.







Description of the Low Dose Extrapolation Model--



    Let P(d) represent the lifetime risk (probability)  of  cancer at dose d.  The



multistage model  has the form





          P(d) =  1 - exp [-(q0 + q^ 4 q2d2 + ... 4 q






where





                        q.  > 0, i = 0, 1, 2, .... k
                         I  ™"™'




Equivalently,





          A(d) =  1 - exp [(qjd + q2d2 + ... + qRdk)]





where
                            A(d)  =
                                 _  P(d)  -  P(o)
                                     -  P(o)
is the extra risk  over background  rate  at  dose  d.



    The point estimate of the coefficents  qi,  i  =  0,  1,  2	k,  and



consequently the extra risk  function A(d)  at  any given dose  d,  is calculated  by



maximizing the likelihood function of the  data.




                                        13-138

-------
    The point estimate and the 95% upper confidence limit of the extra  risk  A(d)



are calculated by using the computer program GLOBAL 79 developed by  Crump  and



Watson (1979).  At low doses, upper 95% confidence limits on the extra  risk  and



lower 95% confidence limits on the dose producing a given risk  are determined



from a 95% upper confidence limit, qt, on parameter q..  Whenever



q, > 0, at low doses the extra risk A(d) has approximately the  form  A(d) =



q. x d.  Therefore, q. x d is a 95% upper confidence limit on the extra



risk and R/qt is a 95% lower confidence limit on the dose producing  an



extra risk of R.  Let LQ be the maximum value of the log-likelihood  function.



The upper limit q1 is calculated by increasing q. to a value qt



such that when the log-likelihood is remaximized subject to this fixed  value



qf for the linear coefficient, the resulting maximum value of the



log-likelihood LI satisfies the equation





                           2 (L0 - L\) = 2.70554





where 2.70554 is the cumulative 90% point of the chi-square distribution with



one degree of freedom, which corresponds to a 95% upper-limit (one-sided).  This



approach of computing the upper confidence limit for the extra  risk  A(d) is  an



improvement on the Crump et al. (1977) model.  The upper confidence  limit  for



the extra risk calculated at low doses is always linear.  This  is conceptually



consistent with the linear non-threshold concept discussed earlier.   The slope,



q|, is taken as an upper bound of the potency of the chemical in inducing



cancer at low doses.



    In fitting the dose-response model, the number of terms in  the polynomial is



chosen equal  to (h-1), where h is the number of dose groups in  the experiment



including the control  group.



    Whenever the multistage model  does not fit the data sufficiently well, data
                                         13-139

-------
at the highest dose is deleted and the model  is refitted to the rest  of the


data.  This is continued until an acceptable  fit to the data is obtained.  To


determine whether or not a fit is acceptable, the chi-square statistic
                            h

                        2      (Xi  - N-iP-^2
                       A  ~    I. p*   7 ^  "i
                           1 = 1



is calculated where Ni is the number of animals in the ith dose group,  XT


is the number of animals in the i^1 dose group with a  tumor response,  P^  is


the probability of a response in the itn dose group estimated  by fitting  the


multistage model to the data, and h is the number of remaining groups.  The fit


is determined to be unacceptable whenever X^ is larger than the cumulative 99%


point of the chi-square distribution with f degrees of freedom, where  f equals


the number of dose groups minus the number of non-zero multistage coefficients.




Selection of Data--


    For some chemicals, several  studies in different animals species,  strains,


and sexes, each run at several  doses and different routes  of exposure,  are


available.  A choice must be  made as to which of the data  sets from  several


studies to use in the model.   It may also be appropriate to correct  for


metabolism differences between  species and absorption  factors  via different


routes of administration.  The  procedures used in evaluating these data are


consistent with the approach  of making a maximum-likely risk estimate.  They  are


1isted below.


    1.  The tumor incidence data are separated according to organ sites or tumor


types.  The set of data (i.e.,  dose and tumor incidence) used  in the model is


the set where the incidence is  statistically significantly higher than  the
                                       13-140

-------
control for at least one test dose level  and/or  where  the  tumor  incidence  rate
shows a statistically significant trend with respect to  dose  level.   The data
set which gives the highest estimate of the lifetime carcinogenic  risk, q*,
is selected in most cases.   However, efforts are made  to exclude data  sets which
produce spuriously high risk estimates because of a small  number of  animals.
That is, if two sets of data show a similar dose-response  relationship, and one
has a very small  sample size, the set of data which has  larger sample  size is
selected for calculating the carcinogenic potency.
    2.  If there are two or more data sets of comparable size which  are
identical  with respect to species, strain, sex,  and tumor  sites, the geometric
mean of q*  estimated from each of these data sets, is used  for  risk
assessment.  The geometric  mean of numbers A}, A2,  ...,  A,,, is defined  as
                                              m
                            x A2 x  ...  x

    3.  If two or more significant  tumor  sites  are  observed  in  the  same  study,
and if the data are available,  the  number of  animals  with  at least  one of  the
specific tumor sites under consideration  is used  as incidence data  in the  model.
    4.  Following the suggestion of Mantel and  Schneiderman  (1977),  we assume
that mg/surface area/day is an  equivalent dose  between  species.   Since,  to a
close approximation, the surface area  is  proportional to the 2/3rds  power  of the
weight as would be the case for a perfect sphere, the exposure  in mg/day per
2/3rds power of the weight is also  considered to  be equivalent  exposure.   In an
animal experiment this equivalent dose  is computed  in the  following  manner.
Let
    Le = duration of experiment
    le = duration of exposure
    m  = average dose per day in mg during administration  of the  agent (i.e.,
         during le), and
    W  = average weight of the  experimental animal
                                        13-141

-------
Then, the lifetime average exposure is





                                    1    x m
                               d=
                                    I. x W2/3
     Often exposures are not given in units of ing/day and it becomes necessary



to  convert  the  given exposures  into  nig/day.  For  example,  in  most  feeding



studies exposure is in terms of ppm in the diet.   In this case the exposure in



ing/day is





                               m = ppm x F x r





where ppm is parts per million in the diet of the carcinogenic agent, F is the



weight of the  food consumed per day in  kg,  and  r is the absorption fraction.



In the absence  of  any data to the  contrary,  r is assumed to be equal  to one.



For  a  uniform diet,  the  weight of  the food consumed  is proportional  to the



calories required,  which in turn is proportional  to the surface area or 2/3rds



power of the weight, so that




                                         2/3
                              m a ppm x W    x r
or
                                         a ppm
                                 rW2/3
     As a result, ppm in the diet is often assumed to be an equivalent exposure



between  species.    However,  we  feel   that  this  is  not  justified  since  the



calories/kg of food is very different in the diet of man compared to laboratory



animals  primarily due  to moisture  content differences.   Instead we  use an



empirically-derived food  factor,  f  = F/W, which  is  the fraction of a species
                                         13-142

-------
body weight that is consumed per day as  food.   We  use the  following  rates:

                          Species       W         f
                            Man       70        0.028
                            Rats      0.35      0.15
                            Mice      0.03      0.13

Thus, when the exposure is given as  a certain  dietary concentration  in  ppm,  the

exposure in mg/w2/3 is


               m    =  £E1xF = ppm  x f  x U  =      x  f
                        w2/3
When exposure is given in terms  of mg/kg/day =  m/Wr  =  s,  the  conversion  is
simply
- s x
                                         W1/3
    When exposure is via inhalation,  the  calculation  of  dose  can  be  considered

for two cases where 1)  the carcinogenic agent  is  either  a  completely water-

soluble gas or an aerosol, and  is  absorbed  proportionally  to  the  amount  of air

breathed in, and 2) where the carcinogen  is a  poorly  water-soluble gas which

reaches an equilibrium  between  the air breathed and the  body  compartments.

After equilibrium is reached, the  rate of absorption  of  these agents is  expected

to be proportional  to the metabolic rate, which in turn  is  proportional  to the

rate of oxygen consumption,  which  in  turn is a  function  of  surface area.



Case 1—

    Agents that are in  the form of particulate matter or virtually completely

absorbed gases, such as S02, can reasonably be  expected  to  be absorbed

proportional to the breathing rate.  In this case the exposure in mg/day may be

expressed as


                                m =  I x  v  x r
                                         13-143

-------
where I = inhalation rate per day in m3, v = mg/m3 of the agent  1n  air,  and

r = the absorption fraction.

    The inhalation rates, I, for various species can be calculated  from  the

observations (FASEB 1974) that 25 g mice breathe 34.5 liters/day and  113 g  rats

breathe 105 liters/day.  For mice and rats of other weights,  W (in  kilograms),

the surface area proportionality can be used to find breathing rates  in  m3/day

as follows:
                  For mi
                  For rats
ice,  I  =  0.0345 (W/0.025)2/3 m3/day
ats,  I  =  0.105 (W/0.113)2/3 m3/day
For humans, the values of 20 m3/day* is adopted as a standard  breathing  rate

(ICRP 1977).

    The equivalent exposure in mg/W2/3 for these agents can be derived  from

the air intake data in a way analogous to the food intake data.   The  empirical

factors for the air intake per kg per day, i  = I/W, based upon the  previous

stated relationships are tabulated as follows:

                        Species        W       i  = I/W
                          Man        70         0.29
                          Rats        0.35      0.64
                          Mice        0.03      1.3

Therefore, for particulates or completely absorbed gases, the  equivalent

exposure in mg/W2/3 is
            .    	m_ _  Ivr
                w2/3 " W2/3
    *From "Recommendation  of the International  Commission  on  Radiological
Protection," page 9.   The  average breathing  rate  is  10?  cm3 per  8-hour
workday and 2 x 107 cm3 in 24 hours.
                                        13-144

-------
In the absence of experimental  information or a  sound theoretical  argument  to


the contrary, the fraction absorbed, r, is assumed to be the same  for all


species.





Case 2—


    The dose in mg/day of partially soluble vapors is proportional  to the Q£


consumption, which in turn is proportional to W^/3 and is also proportional to


the solubility of the gas in body fluids, which  can be expressed as an


absorption coefficient, r, for the gas.  Therefore, expressing the 02


consumption as 02 = k W^/3> where k is a constant independent of species, it


follows that


                             m = k W^/3 x v x r
or
                             d = —JJ-TT = kvr
                                 W2/3



As with Case 1, in the absence of experimental  information  or  a  sound


theoretical argument to the contrary, the absorption fraction, r,  is assumed  to


be the same for all species.  Therefore,  for these substances  a  certain


concentration in ppm or ug/m^ in experimental  animals is  equivalent to the


same concentration in humans.  This  is supported by the observation that  the


minimum alveolar concentration necessary  to produce a given "stage" of


anesthesia is similar in man and animals  (Dripps et al. 1977).  When the  animals


are exposed via the oral route and human  exposure is via  inhalation or


vice-versa, the assumption is made,  unless there is pharmacokinetic evidence  to


the contrary, that absorption is equal  by either exposure route.
                                         13-145

-------
    5.  If the duration of experiment (Le) is less than the natural  lifespan
of the test animal  (L), the slope q*, or more generally the exponent g(d),
is increased by multiplying a factor (L/Le)3.  We assume that  if the average
dose d, is continued, the age-specific rate of cancer will  continue  to increase
as a constant function of the background rate.  The age-specific rates for
humans increase at  least by the 2nd power of the age and often by a  considerably
higher power as demonstrated by Doll (1971).  Thus, we would expect  the
cumulative tumor rate to increase by at least the 3rd power of age.   Using  this
fact, we assume that the slope qt, or more generally the exponent g(d),
would also increase by at least the 3rd power of age.  As a result,  if the  slope
q* [or g(d)] is calculated at age Le> we would expect that  if  the
experiment had been continued for the full lifespan, L, at  the given average
exposure, the slope q| [or g(d)] would have been increased  by  at least
(L/Le)3.
    This adjustment is conceptually consistent with the proportional hazard
model  proposed by Cox (1972) and the time-to-tumor model  considered  by Crump
(1979) where the probability of cancer by age t and at dose d  is given by

                        P(d,t) = 1 - exp [-f(t) x g(d)]
Calculation of the Unit Risk
    The risk associated with d mg/kg2/3/day is  obtained  from GLOBAL  79  and,
for most cases of interest to risk assessment,  can be adequately  approximated  by
P(d) = 1 - exp (-q?d).  A "unit risk"  in units  X is simply the risk
corresponding to an exposure of X = 1.   To estimate this value we simply  find
the number of mg/kg^/3/day corresponding to one unit of  X and substitute  this
value into the above relationship.  Thus, for example, if X is in units of
                                         13-146

-------
 ug/m3 in the air, we have that for case (1) d = 0.29 x 701/3 x 10~3
 mg/kg2/3/day and for case (2) d = 1, when ug/m3 1s the unit used to compute
 parameters in animal experiments.
     If exposures are given in terms of ppm in air, we may simply use the fact
 that
                  1 ppm = 1.2 x molecular weight (gas) mg/m3
                                molecular weight
(gas)
TaTr)
Note, an equivalent method of calculating unit risk would be to use mg/kg  for
the animal exposures and then increase the jth polynomial coefficient  by an
amount
                    (wh/wa)J/3 ;    j  = i, 2,  ..., k

and use mg/kg equivalents for the unit risk values.

Estimation of Unit Risk Based on  Human Data
    If human epidemiologic studies and sufficiently valid exposure  information
are available for the compound, they  are always  used in  some way.   If  they show
a carcinogenic effect, the data are analyzed  to  give an  estimate of the linear
dependence of cancer rates on lifetime average dose, which is equivalent to  the
factor BH»  If they show no carcinogenic effect  when positive animal evidence
is available, then it is assumed  that  a  risk  does exist,  but it  is  smaller than
could have been observed in the epidemiologic study, and  an  upper-limit to the
cancer incidence is calculated assuming  hypothetically that  the true incidence
is just below the level  of detection  in  the cohort studied,  which is determined
largely by the cohort size.  Whenever  possible,  human data are  used  in
preference to animal  bioassay data.
    Very little information exists that  can be utilized to extrapolate from  high
exposure occupational  studies to  low  environmental  levels.   However, if a number
                                        13-147

-------
of simplifying assumptions are made, it is possible to construct a  crude  dose-

response model whose parameters can be estimated using vital  statistics,

epidemiologic studies, and estimates of worker exposures.

    In human studies, the response is measured in terms of the relative  risk  of

the exposed cohort of individuals compared to the control group.  The

mathematical model employed assumes that for low exposures the lifetime

probability of death from lung cancer (or any cancer), PQ, may be represented

by the 1inear equation


                                P0 = A + BHx


where A is the lifetime probability in the absence of the agent, and  x  is the

average lifetime exposure to environmental levels in some units, say  ppm.  The

factor, BH, is the increased probability of cancer associated with  each  unit

increase of r, the agent in air.

    If we make the assumption that R, the relative risk of lung cancer  for

exposed workers, compared to the  general population, is independent of  the

length or age of exposure but depends only upon the average lifetime  exposure,

it follows that
                    R - JL - A + _B_n (x,  + x?)
                      - P0 " A + BH   X!
or
                         RP0 = A + BH (K! + x2)
where Xj = lifetime average daily exposure to the agent for the  general

population, X£ = lifetime average daily exposure to the agent  in the

occupational  setting, and PQ = lifetime probability of dying of  cancer  with  no

or negligible acrylonitri le exposure.
                                         13-148

-------
Substituting P« = A + BH x,  and rearranging gives






                               BH = P0 (R - l)/x2





To use this  model,  estimates  of R and x?  must  be obtained from the epidemic-



logic  studies.   The  value  PQ  is derived  from the age-cause-specific  death



rates for combined  males  found in 1976 U.S.  Vital Statistics tables using the



life table  methodology.   For  lung cancer  the  estimate of PQ  is  0.036.   This



methodology is used in the section on unit risk based on human studies.



Interpretation of Quantitative Estimates



     For several reasons,  the  unit risk estimate is only an approximate indication



of the upper-bound on absolute risk in populations exposed to known carcinogen



concentrations.    First,  there  are  important  species  differences  in  uptake,



metabolism,  and organ distribution of carcinogens, as well as species differences



in  target  site  susceptibility,  immunological  responses,  hormone  function,



dietary factors,  and disease.   Second,  the concept  of equivalent  doses  for



humans compared  to animals on  a mg/surface area basis  is  virtually  without



experimental  verification  regarding carcinogenic  response.   Finally,  human



populations  are variable with  respect to  genetic constitution and diet,  living



environment, activity patterns, and other cultural factors.



     The unit risk estimate can give  a rough indication of the relative potency



of a given  agent  compared with other carcinogens.   The comparative potency of



different agents  is  more  reliable when the comparison  is  based  on studies in



the  same  test species, strain,  and   sex,  and  by the  same route  of exposure,



preferably by inhalation.



     The  quantitative  aspect  of the carcinogen  risk  assessment  is included



here because it may be of use  in the  regulatory decision-making process,  e.g.,





                                         13-149

-------
setting regulatory priorities, evaluating the adequacy of technology-based
controls, etc.  However, it should be recognized that the estimation of cancer
risks to humans at low levels of exposure is uncertain.  At best, the linear
extrapolation model used here provides a rough, but plausible estimate of the
upper-limit of risk; i.e., it is not likely that the true risk would be much
more than the estimated risk, but it could very well be considerably lower.   The
risk estimates presented in subsequent sections should not be regarded as an
accurate representation of the true cancer risks even when the exposures are
accurately defined.  The estimates presented may be factored into regulatory
decisions to the extent that the concept of upper risk limits is found to be
useful.

Alternative Methodological  Approaches
    The methods used by the CAG for quantitative assessment are consistently
conservative, i.e., tending toward high estimates of risk.  The most important
part of the methodology contributing to this conservatism in this respect is the
linear non-threshold extrapolation model.  There are a variety of other
extrapolation models that could be used, all of which would give lower risk
estimates.  These alternative models have not been used by the CAG because they
have no biological  basis, whereas the linear model has at least some biological
plausibility.
    The position is taken by the CAG that the risk estimates obtained by use of
the linear non-threshold model  are upper-limits and the true risk could be
1 owe r.
    Another alternative method involves the choice of animal bioassay as the
basis for extrapolation.  The present approach is to use the most sensitive
responder.  Alternatively,  the average responses of all  of the adequately tested
                                         13-150

-------
bioassay animals could be used.  In the case of acrylonitrile,  only  rats  were



tested in the various studies with little evidence for substantial  variation in



tumorigenic susceptibility.



    Extrapolations from animals to humans could also be done on the  basis of



relative weights rather than surface areas.  The latter approach,  used here, has



more basis in human pharmacological responses;  it is not clear  which of the two



approaches are more appropriate for carcinogens.  In the absence of  information



on this point, it seems appropriate to use the  most generally obtained method,



which also is more conservative.







13.5.3.2  Unit Risk Estimate Based on Human Studies




   Of the ten epidemiologic studies reviewed in this document,  the  study  of



workers at the Du Pont May Plant,  Camden, South Carolina (O'Berg 1980) .clearly



presented the most significant evidence of acrylonitrile as  a human  lung



carcinogen.  Furthermore, it is the only plant  for which we  have been able to



estimate exposure levels.  Therefore,  this study is used to  estimate a unit risk



for human inhalation.   Furthermore, the lack of exposure levels for  the other



studies precludes estimating an upper-limit risk based on negative  studies.



    The carcinogenic effectiveness (potency) of acrylonitrile,  BH,  is



calculated as follows:
where PQ = 0.036, the background  lifetime  probability of death  due  to



respiratory cancer.   This  factor  is  derived  from  the  1976 U.S.  Vital Statistics



tables via a life table calculation  (CAG 1978).
                                        13-151

-------
    The Du Pont Chemical  Company follow-up study by O'Berg (1980)  of 1,345

workers exposed to acrylonitrile between I960 and 1966 found the observed and

expected number of cancer cases to be 13 versus 5.5, respectively, in workers

with at least moderate exposure and with a probable latency period of at  least

15 years.  This excess incidence was found to be statistically significant

(P < 0.05).  Half of this excess was due to respiratory cancer, 5 versus  1.4,

which is also statistically significant (P < 0.05).  If we adjust  for the

effects of smoking as described earlier (p 34), we have 5 observed vs. 1.6

expected, which is statistically significant.  Therefore, the relative risk,  R,

is 5.0/1.6 = 3.1.  For this analysis we will equate respiratory cancer incidence

with mortality.

    Estimates of exposure at the plant are those suggested by J. Brown,  an

industrial hygienist at N10SH who visited the plant in 1978.  She states,  "While

no statistical  data were used to establish the system used (to rank the  included

cohort according to exposure), the Du Pont representatives agreed that 8-hour

time-weighted averages might be used to represent possible past exposure  levels,

based principally on past recall of process levels taken some years past"  (NIOSH

1978).  Also in the Brown report were some measurements of potential  ACN

exposure which averaged between 1 and 3 ppm for the first three quarters  of 1977

and less than 1 ppm for the last quarter.

    Since the workers were exposed to at least a moderate level (10 ppm  =

moderate level, 20 ppm = high level), we will assume 15 ppm to be the 8-hour

time-weighted average (TWA).  We estimate the uncertainty attached to this

estimate to be as high as a factor of 5.

    To convert the 8-hour TWA exposure to a lifetime average, we use


                                  8   240    9
                   Xl = 15 PPm x 24 X 365 X 60

                      = 0.5 ppm continuous equivalent lifetime exposure


                                         13-152

-------
where 9 years is estimated to be the average  exposure  duration,  and 60 years  is
estimated to be the maximum possible age  at the  end  of the  observation period.
    The value of BH [in (ppb)-1] is  derived using  the  above estimates as
follows:
                           R    0.036 (3.1  -  1)
                            H        560  ppb
                              =1.5  x 10-4  (ppb)-l

Therefore, the lifetime risk of cancer for  people  continuously exposed to  1 ppb
is 1.5 x 10-4.
    To express risk in terms of ug/m3 concentration, the  conversion factor for
acrylonitrile is
          1 ppb . 1.2 x * ffy""""-"'  .  M"-°6)  .  2.21  ug/m3
or
                              1  ug/m3  =  0.45  ppb

Therefore the upper-bound  risk associated  with  a  lifetime exposure of  1
in air is

                     P =  1.5  x 10'4  x  0.45 =  6.8  x  10-5
Unit Risk Estimate Based  on  Animal  Studies
    For comparison, a  unit  risk  assessment  is  calculated based on both
inhalation and drinking water  studies  (Quast et  al.  1980a, b).
                                         13-153

-------
Drinking Water Study--
    The three rat bioassays of acrylonitrile in drinking water have  all  shown
significant increases in brain and/or spinal cord astrocytomas, Zymbal  gland
carcinomas, and stomach papillomas/carcinomas (Tables 13-21  through  13-24).
    Of all three studies, the Dow Chemical  Company rat study (Quast  et  al.
1980a) showed the highest incidence of tumors per unit of administered  dose.   On
the assumption that humans are as sensitive as the most sensitive of the test
animals, this study was chosen as the basis of the risk estimate.  Since several
tumor sites are affected, the overall  risk  of tumors is determined from the
number of animals having tumors at any of the sites.  This is  shown  in  Table  13-35.
Also presented is the percentage of animals by sex with the  tumors discussed
above.  In addition, the small  intestine and mammary gland,  which are target
sites for rats, are included.  For the mammary gland, however, because  the
control  group incidence is so high, the percentages for the  females  are
tabulated separately.
    To convert the animal doses into a human equivalent dose,  the standard
approach is to equate the doses on the basis of mg per body  surface  area.
Estimating the weight of the rats to be about 350 grams, the human dosage
equivalent to 3.42 mg/kg/day for the rat (Table 13-35) is

                3.42 mg/kg/day x (70/0.350)1/3 = 0.6 mg/kg/day

When these human dose equivalents are used  with the animal  response  data in
fable 13-35,  the multistage model yields a  value of qf = 1.7 x 10
(mg/kg/day)~l for the males, qt = 1.6 x 10~1 (mg/kg/day)-l for the
females  excluding mammary gland tumors, and qf = 4.2 x 10"1
(mg/kg/day)-l for the females including mammary gland tumors.   The latter
value is chosen as the basis for the unit risk estimate, since it is the
                                         13-154

-------
TABLE 13-35.   NUMBER  (PERCENTAGE)  OF  RATS  DEVELOPING TUMORS IN AT LEAST  ONE  OF  THE
   FOLLOWING  TARGET ORGANS:   ZYMBAL GLAND,  TONGUE,  STOMACH, BRAIN,  AND SPINAL
                               CORD BY  DOSE AND  BY  SEX
                          (adapted from Quast  et al. 1980a)
              Males                                         Females*
Animal Dose
(mg/kg/day)
0
3.42
8.53
21.18
Number/Total
(Percent)
4/80 (5.0%)
18/47 (38.3%)t
36/48 (75.9%)t
45/48 (93.8»)t
Animal Dose
(mg/kg/day)
0
4.36
10.76
24.97
Including Mammary
Gland Tumors
Number/Total
(Percent)
59/80 (73.8%)
47/48 (97.9%)§
46/48 (95.8%)§
48/48 (100%)§
Excluding Mammary
Gland Tumors
Number/Total
(Percent)
3/80 (3.8%)
24/48 (50.0%)t
37/48 (77.1%)t
45/48 (93.8%)t
       *Includes  those  with  small  intestine  tumors.
       tP  <  0.00001.
       §P  <  0.001.
                                           13-155

-------
highest value, and since it includes all  of the organ sites  at  risk  to tumor



induction.  If the human dose had been assumed equivalent  on a  mg/kg/day  basis,



the value of these parameters would be smaller by a  factor of



3Nf(70/0.35f = 5.8



    In order to estimate the unit risk for 1 ug/1  of acrylonitrile  in  drinking



water, we assume that the average 70 kg person drinks 2 liters  of water per day.



This corresponds to a daily dose of 2 ug/day x 10-3  mg/ug  x  1/70 kg  =  2.86 x



10"5 mg/kg/day.  The upper limit unit risk corresponding to  1 ug/1



acrylonitrile concentration in water is then






                  P = 1 . e~4'2 x I0"1  x  2.86 x lO'5



                    = 1.2 x 10-5





    The drinking water study could also be used to estimate  the inhalation risk,



although such an estimate is expected to  be unreliable because  of the  difference



in exposure routes.  The dose rate, d(mg/kg/day),  resulting  from breathing 20



m3/day of air containing a concentration  of 1 ug/m3  can be determined  if  one



assumes that 100% of the inhaled acrylonitrile is  absorbed into the  body.  This



was shown to be the case by Young et al.  (1977).   With this  assumption, the dose



rate is





                 d = 1 ug/m3 x 20 m3/day  x 10~3 mg/ug x 1/70 kg



                   = 2.86 x 10-4 mg/kg/day





The upper-limit estimate of the air unit  risk, P,  can be found  using this value



of d and the value of q* estimated above  as follows:





            P = 1 _ e-4-2 x l°~l x 2*86 x 10~4



              = 1.2 x 10-4
                                         13-156

-------
This is a factor of 2 greater than  that  derived from the  human  epidemiologic



studies and one order of magnitude  greater than that derived  from  the  animal



inhalation studies (see below).








Inhalation Study--



    Of the two inhalation bioassays previously discussed, the Quast  et al.



(1980b) study produced a clear carcinogenic effect,  while the Maltoni  et  al.



(1977) study produced no statistically significant  increases  in tumors in rats



exposed to low dosages for 12 months.   Therefore,  the Quast  study  is used for



determining a unit risk estimate via the inhalation  route.



    The tumor incidence data on individual tumor types have  been presented  pre-



viously (Table 13-32).  For the males  the response  occurs at  the following  four



target organs:  Zymbal gland, small intestine, brain, and/or  spinal  cord.  For



males the percentages of rats with  tumors at one or  more  of  the target organ



sites are 5% (5/100), .9% (9/100), and  47% (47/100)  for the 0  ppm,  20 ppm, and 80



ppm groups, respectively.  For the  females the corresponding  percentages  are 0%,



(0/100), 9% (9/100), and 31% (31/100).  In analyzing these data, both  sets  of



responses were fit by the model. The  estimate of  carcinogenic  potency for  the



females was slightly higher and is  presented below.



    In this study, animals were exposed  to 0, 20,  or 80 ppm  of  acrylonitrile 6



hours per day, 5 days per week, for 2  years.  Therefore,  the  lifetime  average



concentration for the 20 ppm group  is







                       20 x "2T * 7  = 3*57 ppri





Similarly, the lifetime equivalent  dosage for the  80 ppm  group  is  14.29 ppm.



Using the multistage model with incidence data for  females,  the carcinogenic



potency is qt = 3.35 x 10~2 (ppm)-l, and the upper-bound  estimate  of




the lifetime risk of cancer associated with 1 ug/m^  = 4.53 x  10~4  ppm  of




                                          13-157

-------
acrylonitrile is

                  p = 1 - exp (-3.35 x ID'2 x 4.53 x 10'4)
                    = 1.5 x 10-5

    In summary, the upper-limit unit risk estimates for humans breathing
1 ug/m3 of acrylonitrile in ambient air (equivalent to 0.45 ppb)  are
6.8 x 10-5 based on the occupational study, 1.2 x 10-4 based on the rat
drinking water study, and 1.5 x 10~5 based on the rat inhalation  study.
Parenthetically, it should be noted that if the human equivalent  dose assumption
were changed to dose per body weight, the unit risk for inhalation based on the
rat drinking water study would be 1.2 x 10~4 x 1/5.8 = 2.1 x 10~5, a value
which is close to the other two estimates.  Although this estimate is considered
unreliable because of the inappropriate route of administration,  it is included
here as a matter of interest.  The upper-limit unit risk for 1 ug/1 of
acrylonitrile in drinking water is estimated to be 1.2 x 10"^.

Relative Potency
    One of the uses of unit risk is to compare the potency of carcinogens.  To
estimate the relative potency, the unit risk slope factor is divided by the
molecular weights and the resulting number expressed in terms of
(mMol/kg/day)-l.  This is called the relative potency index.
    Figure 13-7 is a histogram representing the frequency distribution of potency
indices of 54 suspect carcinogens evaluated by the CAG.  The actual data summar-
ized by the histogram is presented in Table 13-36.  When human data are
available for a compound, they have been used to calculate the index.  When no
human data are available, animal oral studies and animal inhalation studies have
been used in that order.
                                         13-158

-------
    The potency index for  acrylonitrile  based  on  the O'Berg  study of Dupont


workers is 4.5 x 10~3 (mMol /kg/day)'1.   This  is derived as follows:  the


slope estimate from the O'Berg  study  [6.8  x  10~5(ug/m3)-l],  is  first


converted to units  of (mg/kg/day)"*,  assuming  a breathing rate  of ZOm3 of


air per day and a 70 kg person.
     6.8 x 10-5(ug/ra3)-l  x 1^- x — l^L  x  70  kg
               v  3   '      20  m3     10-3 mg
                                                                   -1
                                                   =  0.24(mg/kg/day)




    Dividing by the molecular weight  of 53.1  gives  a  potency  index  of  4.5  x


10~3.  Rounding off to the nearest  order of magnitude gives a value of 10~3


which is the scale presented on the horizontal  axis of Figure 13-7.  The index of


4.5 x ID"3 lies in about the middle of the  third  quartile  of  the  54 suspect


carcinogens.
                                       13-159

-------
   12-
   10-
u
c
5xlO"4

  and ^6x10-3



Index>6xlO"3


and <6xlO'2
                                           1st quartile:   Index>6xlO
           -6    -5   -4    -3    -2    -1    0     +1    +2     +3     +4
        10    10    10   10    10    10    10    10    10    10    10

                            Potency Index
Figure 13-7.   Histogram representing frequency distribution of the potency
              indices of 54  suspect  carcinogens evaluated by the Carcinogen
              Assessment Group.
                                    13-160

-------
TABLE 13-36.  RELATIVE CARCINOGENIC POTENCIES AMONG SUSPECT  CARCINOGENS
              EVALUATED BY THE CARCINOGEN ASSESSMENT GROUP

Compounds
Acrylonitrile
Aflatoxin Bj
Allyl Chloride
Aldrin
Arsenic
B[a]P
Benzene
Benzidine
Beryllium
Cadmium
Carbon Tetrachloride
Chlordane
Hexachlorobenzene
1 ,2-dichloroethane
1 ,1 ,2-trichl oroethane
1 ,1 ,2,2-tetrachloroethane
Hexachl oroethane
2,4 ,6-trichlorophenol
Bis(2-chloroethyl )ether
Bis(chloromethyl )ether
Chloroform
Chromium
Dichlorobenzidine
DDT


Slope
(mg/kg/day)
0.24
2924
1.19x10-2
11.4
14(H)
11.5
5.2xlO-2
234(W)
4.86
6.65(1)
8.28xlO-2
1.61
1.67
3.70x10-2
5.73x10-2
0.20
1.42x10-2
1.99x10-2
1.14
9300(1)
0.11
63(W)
1.69
8.42



Molecular Potency
Weight Index
53.1
312.3
76.5
369.4
149.8
252.3
78
184.2
9
112.4
153.8
409.8
284.4
98.9
133.4
167.9
236.7
197.4
143
115
119.4
104
253.1
354.5

4.5xlO-3
9xlO+0
2xlO-4
3x10-2
9x10-2
5x10-2
7xlO-4
lxlO+0
SxlO'1
6x10-2
5x10-4
4xlO-3
6xlO-3
4xlO-4
4x10-4
IxlO-3
6x10-5
1x10-4
8x10-3
8xlO+l
9x10-4
6X10-1
7xlO-3
2x10-2
(continued on the
Order of
Magnitude
(exponent of
base 10)
-3
0
-4
-2
-2
-2
-4
0
-1
-2
-4
-3
-3
-4
-4
-3
-5
-4
-3
1
-4
-1
-3
-2
following page)
                               13-161

-------
TABLE 13-36.  (continued)
Compounds
1,1-dichloroethylene
Dieldrin
Dinitrotoluene
Tetrachlorodioxin
Diphenylhydrazine
Epichlorohydrin
Ethyl ene Di bromide (EOB)
Ethylene Dichloride (EDC)
Ethyl ene Oxide
Formaldehyde
Heptachlor
Hexachlorobutadiene
HexacM orocycl ohexane
technical grade
alpha Isomer
beta isomer
gamma isomer
Nickel
Nitrosamines
Di methyl nitrosamine
Diethylnitrosamine
Oibutylnitrosamine
N-n1trosopyrrol Idine
N-nitroso-N-ethylurea
N-nitroso-N-methyl urea
N-ni troso-di phenyl ami ne
PCBs
Slope
(mg/kg/day)
1.04
30.4
0.31
4.25x10$
0.77
7.69xlO-4(I)
8.51
1.44x10-2
1.86x10-2(1)
2.14x10-2(1)
3.37
7.75x10-2

4.75
11.12
1.84
1.33
6.30(W)
25.9(not by q*)
43.5(not by q{)
5.43 '
2.13
32.9
302.6
4.92xlO-3
4.34
Molecular
Weight
97
380.9
182
322
180
92.5
187.9
99.0
44.0
30
373.3
261

290.9
290.9
290.9
290.9
58.7
74.1
102.1
158.2
100.2
117.1
103.1
198
324
Potency
Index
1x10-2
8x10-2
2xlO-3
lxlO+3
4xlO-3
8xlO"6
5x10-2
1x10-1
4xlO-4
7xlO-4
9x10-3
3xlO'4

2x10-2
4x10-2
6xlO-3
5x10-3
IxlO'1
4x10-*
4x10-1
3x10-2
2x10-2
3x10-1
3x1 0+0
2x10-5 -
IxlO-2
Order of
Magnitude
(exponent of
base 10)
-2
-2
-3
3
-3
-6
-2
-4
-4
-4
-3
-4

-2
-2
-3
-3
-1
-1
-1
-2
-2
-1
0
-5
-2
                          (continued on the following  page]
         13-162

-------
                                TABLE  13-36.  (continued)
Compounds
Tetrachloroethylene
Toxaphene
Trichloroethylene
Vinyl Chloride
Vinyl idene Chloride
Slope
(mg/kg/day)
5.31x10-2
1.13
1.26x10-2
1.75x10-2(1)
0.13(1)
Molecular
Weight
165.8
414
131.4
62.5
97
Potency
Index
3xlO-4
3xlO-3
IxlO-4
3xlO-4
ixin-3
Order of
Magnitude
(exponent of
base 10)
-4
-3
-4
-4
-3
Remarks:
    1.   Slopes (q?)  in  (mg/kg/day)-1 are calculated based on animal oral studies,
        except for those  indicated by  I (Animal inhalation), W (human occupational  exposure),
        and H (human drinking water exposure).

    2.   The potency  index  Is a rounded-off slope 1n (mMol/kg/day)'1 and is calculated by
        dividing  the slopes in (mg/kg/day)-l by the molecular weight of the compound.

    3.   Not all the  carcinogenic potencies presented in this table are final.  Some are
        subject to change  as the CAG is getting the individual risk assessment documents
        approved.
                                           13-163

-------
13.5.4  SUMMARY



Qualitative Assessment



    Acrylonitrile is not a direct acting carcinogen and hence the localization



and nature of the effects depend on its metabolism.  It is probable that the



proximal carcinogen is 2-cyanoethylene oxide, since it has been demonstrated



as a reaction product with calf thymus DNA.  However, the metabolite has not



been tested directly for its carcinogenicity.  It has been shown to be



produced in the liver and possibly circulates to other organs.   However,



studies have not been done to determine where else in the body  this metabolite



is produced.  There appears to be a clear difference between animals and



humans in their tumorigenic response to acrylonitrile:  no lung tumors have



been produced in animals and no brain tumors have been observed in humans.



There are no human studies on the metabolism of acrylonitrile and there are no



pharmacokinetic studies that would be relevant to the characterization of



dose-response relationships at low levels of exposure.



    The carcinogenicity of acrylonitrile has been studied in seven cancer



bioassays in rats:  four in drinking water, one by gastric intubation, and  two



by inhalation.  These are summarized in Table 13-37.  In addition, ten



epidemiologic studies of cancer incidence have been reported.  These are



summarized in Table 20.  A short description of these studies is presented



below.



    Quast et al. (1980a) administered acrylonitrile in drinking water to



Sprague-Dawley rats for 2 years at dose levels of 35, 100, and  300 ppm.  A



statistically significant incidence of tumors was observed in the central



nervous system,  Zymbal  gland, stomach, tongue, and small  intestine in both



male and female  rats, as well as in the mammary gland of female rats.  The



occurrence of central nervous system and Zymbal  gland tumors in Sprague-Dawley
                                         13-164

-------
                                            TABLE  13-37.   CARCINOGEN1CITY  OF  ACRYLONITRILE  IN RATS
             Reference
Strain of Rat
   Route of
Administration
Dose Level*
Treatment
Schedule
Site of Tumors
CO
 i
             Quast et al.  (1980a)
Sprague-Dawley   Ingestlon 1n      35-300 ppm     24  mo.
                 drinking water
             Bellies et al. (1980)        Charles  River     Ingestlon  1n       500  ppm
                                                          drinking water
             Blodynamlcs, Inc. (1980a)    Spartan
             Btodynamics, Inc. (1980c)    Fischer  341
             Blodynamlcs, Inc. (1980b)    Spartan
                 Ingestlon 1n      100 ppm
                 drinking water
                                 45 wk.t


                                 19-20 mo.
                 Ingestlon 1n      10-100 ppm     23-26 mo.
                 drinking water
                 gavage
                  10 mg/kg/day   19-20 mo.
                             Brain (astrocytomas)
                             Ear canal  (Zymbal  gland)
                             Stomach (nonglandular
                               portion)
                             Mammary gland (females only)
                             Tongue
                             Pituitary gland
                             Pancreas (males only)
                             Uterus (females only)

                             Brain (astrocytomas)
                             Ear canal  (Zymbal  gland)

                             Brain (astrocytomas)
                             Ear canal  (Zymbal  gland)
                             Stomach

                             Brain (astrocytomas)
                             Ear canal  (Zymbal  gland)
                             Stomach
                             Mammary gland (females only)

                             Brain (astrocytomas)
                             Ear canal  (Zymbal  gland)
                             Stomach
                             Mammary gland (females only)
                                                                                                    (continued on  the  following page)
                 *Not all the tumors Indicated  for  each  study  were  observed  at  the  lower dose.
                 tThis was a three-generation  reproductive  study; the  tumors were observed in the second generation.

-------
                                                        TABLE 13-37.   (continued)
oo
 i
en
CTl

Reference
Maltonl et al. (1977)
Quast et al. (1980b)

Route of Treatment
Strain of Rat Administration Dose Level* Schedule
Sprague-Dawley Inhalation 5-40 ppm 4 hr/day
5 days/wk.
for 12 mo.
Sprague-Dawley Inhalation 20-80 ppm 6 hr/day
5 days/wk
for 24 mo.


Site of Tumors
Stomach (males only)
Mammary gland (females only)
Skin (females only)
Central nervous system
Ear canal (Zymbal gland)
Gastrointestinal tract
(males only); mammary gland
(females only); tongue
(males only)
              "Not  all the tumors IndTcatea for eacn study were observed at the 1 ower dose.

              tThls was a three-generation reproductive study; the tumors were observed in the second  generation.

-------
rats was  further  confirmed in four other  studies:   a  three-generation repro-



duction  study  performed at  Litton-Bionetics  by Beliles et  al.  (1980);  three



studies  by  Biodynamics,  Inc.  (1980a,  b, c) in  which acrylonitrile was admin-



istered  in  drinking  water  and via gastric intubation;  and an inhalation study



by Quast et al. (1980b).



     A second  inhalation study by Maltoni et al. (1977) exposed rats to atmos-



phere containing  5,  10,  20,  and 40 ppm  acrylonitrile  4  hours  per day, 5 days



per week, for  12 months.  Marginal increases in tumors of the mammary gland in



females  and the  forestomach  in males were observed, although  the sensitivity



of  this  test  was limited  by the  relatively  low dose  levels and  the  short



duration of exposure.



     Ten epidemiologic studies of acrylonitrile and cancer have been reported:



five published [Monson (1981), O'Berg (1980),  Thiess et al.  (1980), Werner and



Carter (1981),  Delzell  and  Monson  (1982)] and  five unpublished  [Gaffey  and



Strauss (1981), Herman (1981), Kiesselbach et  al.  (1980), Stallard (1982),  and



Zack (1980)].  These are  summarized  in  Table 13-38.  Six  of these studies present



no evidence of carcinogenic risk from exposure to acrylonitrile.   However,  all



suffer from  problems in the design or methodology, i.e.,  small  cohort size,



insufficient  characterization of  exposure,  short  follow-up,   and relatively



youthful  cohort.   Because of these problems none of these studies can be cited



as adequate evidence  that acrylonitrile is not carcinogenic.



     Data presented  in  the remaining four epidemiologic studies consistently



demonstrate a  statistically  significant  risk  of lung cancer  in various sub-



groups of  the populations  studied.   All four have  problems with the method-



ology, definition, and/or  size of the population, whether  or  not exposure to



other carcinogens  occurred,   and  short  follow-up intervals.   In  three of  the



four  studies   [Delzell  and Monson  (1982),  Thiess et  al.  (1981),  Werner  and



Carter (1981)], the  problems were sufficient  to cast  doubt  on the finding of






                                         13-167

-------
                     TABLE  13-38.   EPIDEMIOLOGIC  STUDIES  REVIEWED  IN  ACRYLONITRILE  RISK  ASSESSMENT
       Study
  Results
Observed/Expected    Highlights and Deficiencies
       O'Berg  (1980)
cr>
oo
       Delzell  and Monson
         (1982)

       Thiess et al.
         (1980)
       Werner and  Carter
         (1981)
       Monson (1978)

       Zack (1980,
       unpubli shed)

       Gaffey and  Strauss
       (1981, unpublished)
Positive
  lung cancer
Positive
  lung cancer

  Positive
  lung cancer
  Lymph cancer
Positive
  lung cancer
  stomach cancer
Inconclusive

Inconclusive


Inconclusive
                                                     5/1.4  (P <0.05)
 4/1.4 (P <0.05)
 9/4.4 (P <0.05)
 4/1.4 (P <0.05)
 3/0.7 (P <0.05)
 5/1.9 (P <0.05)
                     Significant excess occurred in group
                     with highest exposure followed for the
                     longest time.  Excess remains after
                     adjustment for smoking.
Potential exposure to other carcinogens,


Incomplete follow-up on foreign workers,
Mixed exposures to beta-naphthalamine,
polycyclic hydrocarbons, and vinyl
chloride.
Relatively short follow-up.  Combined
Population from 6 polymerization
factories.

Potential exposure to other carcinogens.

Small cohort size.  Short follow-up on
most members of cohort.

Small cohort size.  Includes large numbers
of minimally or unexposed employees.
                                                                                  (continued on the following  page)

-------
oo
I
                                            TABLE  13-38.   (continued)
    Study
 Results
Observed/Expected    Highlights and Deficiencies
    Kiesselbach et al.
    (1981, unpublished)
    Herman (1981,
    unpublished)

    Stallard (1982,
    unpublished)
Inconclusive
Inconclusive


Inconclusive
                     Markedly reduced mortality due to  1)
                     healthy worker effect, and 2) possible
                     selection bias in determination of study
                     population.  Lack of critical editorial
                     review.

                     Relatively short duration of employment.
                     Insufficient follow-up.

                     Small cohort.  Short follow-up period.
VO

-------
significantly elevated  risks  of  lung  cancer  reported in each study.   In  the



fourth study by O'Berg,  the  proglems were insufficient to obscure the signif-



icant finding of lung cancer.   After adjusting for latent factors and evaluat-



ing the contribution  due  to  smoking, the finding of  a statistically signifi-



cantly elevated risk of lung  cancer remained.   Thus,  one study appears adequate



and three  are  suggestive, while  the remaining six are  inadequate  to address



the issue of a risk of lung cancer.



     In addition  to  lung cancer,  two  other findings of concern  are  the  sig-



nificantly elevated  risk of  lymph  system cancer  found in the  Thiess  et  al.



study (4 observed  versus  1.38 expected,  P <  0.05) and the significantly  ele-



vated risk of  stomach cancer  found in the Werner and Carter study (5 observed



versus 1.9 expected,  P  <  0.05).   These findings provide additional  suggestive



evidence of the carcinogenicity of acrylonitrile.



     This level of  animal  evidence would be regarded as "sufficient" evidence



of  carcinogenicity  according to  the International  Agency  for Research  on



Cancer  (IARC)  classification  scheme.   The  human evidence  for  the  carcino-



genicity of acrylonitrile would  be regarded as somewhere between "sufficient"



and  "limited,"  using  the  IARC classification.   Therefore,  in  combining  the



human and  animal  evidence, acrylonitrile  would be .placed in  group  2A, which



IARC characterizes as "probably carcinogenic in humans, where the evidence for



human carcinogenicity is almost sufficient."



Quantitative Assessment



     Three unit  risk estimates for  air  are calculated; one based  on a human



occupational  study  (O'Berg,   1980a,  b) and two based  on  rat  cancer bioassays



(Quast et al.,  1980a,b).   The upper-bound lifetime  risk  of  cancer  associated



with a  lifetime  inhalation   exposure of  1 ug/m   is  6.8 x 10   from the human



study and  1.5  x 10   from the  rat  inhalation study.  The value  based  on the
                                         13-170

-------
                                     -4            -5
rat drinking water  study  is  1.2 x 10    (or  2  x 10   if the  equivalent  human



dose  is  assumed to be mg/kg/day  rather  than surface area) but this  study  is



less reliable because of the  inappropriate route of exposure.



     The estimate based on the human study is uncertain  because of the relati-



vely weak documentation of the available exposure estimates of the acrylonitrile



workers.   The air concentration had not been measured when  the workers experienced



their heaviest exposure and was estimated 12 years after the end of the exposure



period.   However, in spite of these difficulties, the estimates are consistent



with those of the animal  studies.



     The  upper-bound  risk estimate  for  1 ug/1  of acrylonitrile  in  drinking



water is 1.2 x 10  , based on the Quast et al.  (1980a) drinking water study  in



rats.



     These values are regarded as rough but plausible estimates of an upper-bound



of  risk;  i.e.  it  is  not  likely  that  the true  risk would be  much  more than



these values,  but it could  very  well  be considerably  lower.   In  using  these



quantitative  estimates,   the assumptions  made  in their  derivation and  the



limitations of their  interpretation  should be  kept clearly  in  mind.   This  is



discussed on pages 13-136 through 13-151 of the document.



13.5.5  CONCLUSIONS



     There is evidence that acrylonitrile is a  human carcinogen.  This conclu-



sion is  based  on:   1) findings of three positive drinking  water rat bioassays



and one  positive rat gastric  intubation  study;  2)  statistically significant



positive  findings of  respiratory  cancer in four epidemiologic studies; 3) the



positive  mutagenic  evidence  in bacteria and sister  chromated exchange tests;



4)  jjn  vitro evidence of  interaction of acrylonitrile  and/or its metabolites



with  DNA;  and 5)  acrylonitrile's  structural similarity to  vinyl  chloride,  a



known animal and human carcinogen.
                                        13-171

-------
     The carcinogenic potency  of  acrylonitrile is in the third quartile among



54 suspected carcinogens evaluated by the Carcinogen Assessment Group.



     Using the  International  Agency  for Research on Cancer  (IARC)  classifi-



cation scheme, this level of evidence in animals and humans would be  considered



sufficient for concluding that acrylonitrile is likely to be a human  carcinogen



with rank of 2A.
                                         13-172

-------
                                   REFERENCES


 Aarato,  S.E., and Bittera, E. (1972).  Determination of acrylonitrile in the air
      in  the  presence  of dimethylformamide  and  ammonia.   Hunkavedelen  18:50-51.
      (In Hung.) Taken from:  Chem. Abst. l£:57280b, 1973.

 Abreu, M., and Ahmed, A.E. (1980).  Metabolism of acrylonitrile to cyanide:  In
      vitro  studies.   In:   Abstracts of Papers:   Society of Toxicology.   19th
      Annual  Meeting,  Washington,  DC,  March  9-13,  1980.    Abstract  No.  393,
      p.  A131.

 A.D.  Little,  Inc.  (1971).  A  Model  Economic and Safety  Analysis of the Trans-
      portation of Hazardous Substances in Bulk, Final Report (NTIS COM-71-11271)
      261 pp.  (Cited  in Miller and Villaume, 1978.)

 Ahmed, A.E., and Patel,  K. (1979).  Pharmacokinetics, distribution and binding of
      l-1lC-acrylonitrile  in rats.  Pharmacologist 21( 3):A131.  (Abstract.)

 Ahmed, A.E., and Abreu, M.E. (1982).  Microsomal metabolism of acrylonitrile in
      liver and brain.   Adv. Exp. Med. Biol. 1366:1229-1238.  Taken from:  Chem.
      Abst.  57:l8598p.

 Ahmed,   A.E.;  Farooqui,  M.Y.H.;   Upreti,  R.K.;   and El-Shabrawy,  0.  (1982).
      Distribution and covalent interactions of  [1-11C] acrylonitrile in the rat.
      Toxicology 23^:159-175.  Taken from:  Chem. Abst. 27:67136t.

 AIHA  (American  Industrial Hygiene Association)  (1970).    Acrylonitrile.  Amer.
      Ind. Hyg. Assoc. J., July-August 1970, pp. 529-531.

 ASTM  (American Society  for  Testing and Materials)  (1980).   Nitriles in aqueous
      solution  by  gas-liquid  chromatography,   1980.    Annual  Book  of  ASTM
      Standards, Part  31,  Water,  Method No. ASTM D3371-79,  ASTM, Philadelphia,
      PA.

Axel son,  0.   1978.   Aspects  on  confounding  In occupational health and
    ep1den1ology.   Scan. J.  Work  Environ, and Health 4:85-89.


 Babanov, G.P. (I960).    Problems  of  industrial  hygiene in the production of
      synthetic butadiene-acrylonitrile  rubbers.   Gigiena Truda i Professional.
      Zabolevaniya  JK12):7-12.   (In Russ.)  Taken  from:   Chem.  Abst. 55:23883a,
      1961.

 Babina,   M.D. (1979).   Gas-chromatographic  determination of  trace impurities
      released into the  air  in  the  production of  some styrene copolymers.  Gig.
      Tr. Prof. Zabal. 2:55-7.  (In Russ.)  Taken  from:   Chem. Abst. 90;19l833q.
      1979.

 Balak,  J.;  Cavojcova, E.; and Polievka, M.  (1977).   Problems of acrylonitrile
      analysis  in  cyanoethylene  reaction  mixtures  using gas  chromatography.
      Petrochemia lJ(l-2):37-16.  (InSlo.)  Taken  from:  Chem. Abst. 88:1l531v,
      1978.

 Bandt, H.J.  (1953).   [Acrylonitrile  as  an  effluent.]   Z. Fisch.  2:157-161.  (In
      Ger.)
                                         R-l

-------
Barrett,  V.J. (1971).   Sampling and Analysis of  Four Organic Compounds Using
     Solid Sorbents.  Southern Research Institute Report, SORI-EAS-71-301, U.S.
     Department of Health, Education and Welfare.  (Cited in Going et al., 1978.)

Baxter,  R.A. (1979)*   Evaluation and control of Industrial exposure to acrylo-
     nltrile. Ann. Occup.  Hyg.  22:129-135.

Baxter,  R.A. (1980).   Written communication from R.A. Baxter, Monsanto Europe,
     Brussels, Belgium,  to  W.R.  Gaffey, Monsanto World Headquarters, St. Louis,
     MO,  dated June 21,  1980, including  minutes  of the meeting held to Review
     Recently Completed Studies on the Epidemiology  of Acrylonitrlle,  held in
     Brussels, April  23 and 21,  1980.

Beaumont, G.P.,  and Garrido,  C.H.  (1979).  Respirator cartridge test system and
     test-results for  benzene  and acrylonitrile.   Aoer.  Ind. Hyg.  Assoc. J.
     JKK 10): 883-87.

Bellies,   R.P.;  Paulin,  R.J.;  Makris, N.G.;  and Weir,  R.J. (1980).   Three-
     Generation  Reproductive Study of Rats Receiving Acrylonitrile In Drinking
     Water.   Prepared  by  Litton  Bionetlcs,  Inc.,  Kensington,  MD,  under LBI
     Project No. 2660 for  the Chemical  Manufacturers Association, Washington,
     DC.   Available from:  CMA,  Washington,  DC.

Benes, V., and Cerna, V.  (1959).  [Acrylonitrile:  Acute toxicity and mechanism
     of action.]   J.  Hyg. Epidemlol. Mlcrobiol. Immunol. 3_: 106-116.  (In Ger.)

Benes, V.,  and  Sram,  R. (1969).   Mutagenic activity  of some  pesticides in
     Drosophila  melanogaster. Ind.  Med.  Surg. 38(12) :112-111.

Berck, B. (I960).  Fumlgant  residues:  Retention of acrylonitrile and carbon
     tetrachloride by shelled walnuts fumigated with Acrylon.  J. Agr.  Food Chem.
     8(2): 128-31.

Berck, B.  (1962).   Polarographlc  determination  of methyl  bromide, ethylene
     dibromide,  acrylonitrile,  chloropicrln, and  carbon tetrachloride in air.
     J.  Agr. Food Chem. 10:158-62.

Berck, B.  (1965).   Determination  of fumlgant  gases  by  gas chromatography.
     J.  Agr. Food Chem.  113( D;373-77.

Berck, B. (1975). Analysis of fumlgants  and fumigant  residues.  J. Chromatogr.
     Sci. 13.(6): 256-67.  .

Bio/dynamics Inc. (1980a).   A Twenty-Four Month Oral Toxlcity/Carcinogenicity
     Study of Acrylonitrile Administered  to  Spartan Rats in the Drinking Water.
     Vols. 1 and 2 of Final Report.  Prepared by Bio/dynamics  Inc., Division of
     Biology and Safety  Evaluation, East  Millstone,  NJ,  under Project No. 77-
     17U5 for Monsanto  Company,  St.  Louis, MO.

Bio/dynamics Inc. (1980b).   A Twenty-Four Month Oral Toxlcity/Carcinogenicity
     Study of Acrylonitrlle Administered by Intubation to Spartan Rats.  Vols. 1
     and 2 of Final Report.  Prepared  by  Bio/dynamics Inc., Division  of Biology
     and Safety Evaluation,  East  Millstone,  NJ,  under Project No. 77-1716 for
     Monsanto Company,  St.  Louis,  MO.
                                              R-2

-------
Bio/dynamics Inc. (1980c).   A Twenty-Four Month Oral Toxicity/Carclnogenicity
     Study of Acrylonitrile Administered In  the  Drinking Water to Fischer 311
     Rats, Vols. 1-1  of Final Report.  Prepared by Bio/dynamics Inc., Division of
     Biology and  Safety  Evaluation,  East  Millstone,  NJ,  under Project  No.
     77-1744 (BDN-77-27)  for Monsanto Company, St.  Louis, MO.

Bocek, K. (1976).   Relationships  among activity coefficients, partition coeffi-
     cients and solubilities.   Experientia,  Suppl.  23:231-10.

Bogaczek, J., and Jaworski, M. (1970).  Continuous polarographic determination
     of acrylonltrile.  Fresenius1 Z. Anal. Chem. 251(3): 171-76.  (InGer.) Taken
     from:  Chem.  Abst. £l:91406u,  1970.

Bond, E.J. (1963).   The  action of fumigants  on insects.   IV.   The effects of
     oxygen on the toxicity of fumigants to  insects.  Can. J. Biochem. Physiol.
     11:993-1004.

Bond, E.J., and Buckland, C.T.  (1976).  Control of Insects with fumigants at low
     temperatures: Toxicity of mixtures of methyl bromide and acrylonitrile to  3
     species of insects.   J.  Econ.  Entomol.  69(6);72527»

Bond, E.J., and  Buckland,  C.T. (1978).  Control of Insects with  fumigants at low
     temperatures:   Toxicity of  fumigants  in atmospheres  of carbon dioxide.
     J. Econ. Entomol.  71(2);307-9*

Boxer, G.E., and Rickards, J.C. (1952).  Studies on the metabolism of  the carbon
     of cyanide and thlocyanate.   Arch. Biochem. 39:7-26.

Boyland,  E.,  and  Chasseaud,  L.F.  (1967).    Enzyme-catalyzed  conjugations of
     glutathione with unsaturated compounds.   Biochem. J. 101:95-102.

Braun,  D.,  and  Vorendohre,  G.  (1963)•   Thin-layer  chromatography  of  some
     unsaturated polymerizable compounds.   Z. Anal. Chem.  199(1);37-11.   (In
     Ger.) Taken from:  Chem.  Abst. 60:l253d,  1964.

Brieger, H.;  Rleders, F.;  and  Modes,  W.A. (1952).   Acrylonitrile:  Spectophoto-
     metric determination, acute toxicity,  and mechanism of action.  Ind. Hyg.
     Occup. Med. 6:128-40.

Brown, M.E.; Breder, C.V.; and McNeal, T.P. (1978).  Gas-solid chromatographic
     procedures  for   determining   acrylonitrile   monomer   in  acrylonitrile-
     containlng polymers  and  food  simulating solvents.  J.  Assoc.  Off. Anal.
     Chem. 6K 6): 1383-388.

Burg, S.P., and Burg,  E.A.  (1967).  Molecular requirements for the  biological
     activity of ethylene.  Plant Physiol. 12_:111-152.

Burkhart, P.;  Bartels,  D.; and Hoyme,  H. (1961).  Analytical  chemistry of poly-
     acrylonltrlle fibers.  I.   Analytical  studies on  the separation of vinyl
     monomer mixtures.   Faaerforsch. Textiltech.  l£:581-87..   (In Ger.)  Taken
     from:  Chem.  Abst. 5_7_:1107i, 1962.
                                            R-3

-------
Campbell,  D.N.,  and Moore,  R.H.  (1979).   The  quantitative determination  of
     acrylonitrile,  acrolein,  acetonitrile  and  acetone  in  workplace  air.
     Amer. Ind. Hyg.  Assoc.  J. 40(10):904-09.

Carcinogen  Assessment   Group,  U.S.  Environmental  Protection  Agency.   1978.
     Carcinogen Assessment Group's  final  report on population  risk to ambient
     benzene exposures.   41  pp.

Carpenter,  C.P.;  Smyth, H.F., Jr.;  and Pozzani,  U.C.  (1949).  The  assay  of
     acute  vapor  toxicity,  and  the  grading and interpretation of  results  on
     ninety-six chemical compounds.  J.  Ind. Hyg. Toxicol.  3_1:343-346.

Chaigneau, M., and LeMoan,  G. (1974).   Pyrolysis of plastic materials.   VIII.
     Polyacrylonitrile  and copolymers.  Ann.  Pharm.  Fr.  3^(9-10):485-90.  (In
     Fr.) Taken from:  Chem.  Abst. 82:73614f,  1975.

Chao, L.-S., and  Ch'en,  S.-S. (1966).   Polarographic determination of acrylo-
     nitrile-vinyl acetate  copolymer systems.   Hua  Hsuen Tung Pao (l):47-9.
     (In Ch.) Taken from:  Chem.  Abst.  65:10669g, 1966.

Chekhovskaya, E.V.; Gorban,  N.S.; Yakovleva, L.I.;  Nazarenko, G.P.;  and Necha-
     eva,  A.V.  (1966).    Data for  experimental studies  of toxicity of wash
     waters  from   acrylonitrile  production.  vodosnabzh.  Kanaliz.  Gidrotekh.
     Sooruzh. Mezhued.   Resp.  Nauch.  Sb.  No. l_:83-88.  (In  Russ.)  Taken from:
     Chem. Abst.  66:88487k,  1967.

Chemical Marketing Reporter  (1980).  Chemical profile:  Acrylonitrile.  Febru-
     ary 4, 1980,  p.  9.

Cherry,  A.  B.; Gabaccia, A.J.;  and Senn, H.W.  (1956).  The assimilation be-
     havior of certain  toxic  organic compounds in natural water.   Sewage Ind.
     Wastes 28(9):1137-46.

Chiou, C.T.; Freed,  V.H.; Schmedding,  D.W.; and Kohnert, R.L.  (1977).  Parti-
     tion  coefficient   and  bioaccumulation  of selected  organic  chemicals.
     Environ. Sci. Technol.  ll.:475-478.

Chopra,  S.K.;  Vishnoi,   S.C.; Kapoor, V.B.;  and  Gupta,  P.L.  (1978).   A gas
     chromatographic method  for   the estimation of  residual monomers in sty-
     reneacrylonitrile  copolymer.  Res.  Ind.  23(1):27-30.   Taken from:   Chem.
     Abst. 89-.164097a,  1978.

Chudy, J.C.,  and  Crosby, N.T.(1977).  Some observations  on the  determination
     of monomer residues in foods.  Food Cosmet. Toxicol. 15(6):S47-51.

Conner,  T.H.;  Forti, G.C.;  Fitra,  P.;  and Legator, M.S.  (1979).   Bile as a
     source  of mutagenic metabolites produced  in vivo  and detected by Sal-
     monella typhimurium.  Environ. Mutag. 1^:269-276.

Covic-Horvat, S.;  Vrzina, J.; and Sateva, M. (1970).  Determination of acrylo-
     nitrile in waste waters.  Nafta 21_(2) :79-81.  (In Croatian.)  Taken from:
     Chem. Abst. 73:28605g,  1970.

Cox,  C.R.  1972.   Regression  model  and  life  tables.   J.  Roy.  Stat.  Soc.  B
     34:187-220.

                                        R-4

-------
Crume, R.V.  (1982).  Summary  of acrylonitrile producers,  emissions,  and plant
     capacities.  Referenced  memo  available from the Chemicals  and  Petroleum
     Branch, Emission  Standards and Engineering Division, U.S.  Environmental
     Protection Agency,  Research Triangle Park, North Carolina.  October  29,
     1982.

Crump, K.S., H.A. Guess,  and  L.L.  Deal.   1977.   Confidence intervals and test
     of  hypotheses  concerning  dose-response  relations  inferred from  animal
     carcinogenicity data.  Biometrics 33:437-451.

Crump,  K.S.  1979.   Dose-response problems  in  carcinogenesis.  Biometrics
     35:157-167.

Crump, K.S., and W.W.  Watson.   1979.   GLOBAL 79.   A fortran  program  to  extra-
     polate dichotomous animal  carcinogenicity  data to  low dose. Natl. Inst.
     of Environ. Health Science.  Contract No. l-ES-2123.

Dahle, L.,- Hill, and Holman,  R. (1962).   The thiobarbituric  acid reaction and
     the autoxidations  of polyunsaturated fatty methyl esters.   Arch. Biochem.
     Biophys. 98:253.

Dahm, K.J.  (1977).   Indentification of metabolites  of  acrylonitrile.   Testi-
     mony before FDA.   (Cited  in Miller and Villaume, 1978.)

Daues,  G.W.,  and Hamner, W.F.  (1957).    Determination  of  small  amounts  of
     acrylonitrile   in  aqueous  industrial  streams.   Anal.  Chem.  29:1035-37.

Daugherty, F.M., Jr.,  and Garrett,  J.T.   (1951).   Toxicity  levels  of  hydro-
     cyanic  acid and  some industrial  by-products.   Texas J.  Sci. 3:391-396.

Davis, J.H.; Davies, J.E.; Raffonelli,  A.; and Reich, G.  (1973).   The investi-
     gation  of  fatal  acrylonitrile intoxications.   In-.   Pesticides and  the
     Environment:    A  Continuing  Controversy.   Intercontinental Medical  Book
     Corp., Vol. 2,  pp. 547-556.

Delzell, E., and R.R.  Monson.   1982.  Mortality among rubber workers. VI.  Men
     with exposure  to acrylonitrile.  J.  Occup.  Med. 24(10):767-769.

DeMeester,  C.;  Duverger-VanBogaert,  M. ,-  Lambotte-Vandepaer,  M. ;  Roberfroid,
     M.; Poncelet,   F.;  and Mercier, M. (1979).  Liver extract mediated muta-
     genicity of acrylonitrile.   Toxicology 13(1):7-15.

DeMeester, C.;  Poncelet,  F.;  Roberfroid,  M.; and  Mercier,  M.  (1978).   Muta-
     genicity of acrylonitrile.   Toxicology 11:19-27.

Deur-Siftar,  D.,  and   Svob,  V.  (1976).   Procedure  for   gas  chromatographic
     determination   of  acrylonitrile as a residual  monomer  in polymerization
     process waste  waters  and   in  polymers.   Kern.  Ind.  2_5( 12)-.703-08.   (In
     Croatian.) Taken from:   Chem.  Abst.  87:58072n,  1977.

DiGeronimo,  M.J.,  and  Antoine,   A.D.  (1976).  Metabolism  of acetonitrile and
     propionitrile   by  Nocardia  rhodochrous  LL100-21.   Appl.  Environ.  Micro-
     biol. 31(6):900-906.
                                         R-5

-------
Dilling, W.L. (1,977).   Interphase  transfer processes.   II.   Evaporation rates
     of  chloromethanes, ethanes,  ethylenes,  propanes,  and propylenes  from
     dilute  aqueous  solutions.    Comparison  with  theoretical  predictions.
     Environ. Sci. Technol.  n(4);405-09.

DiLorenzo, A., and Russo, G. (1969).   Gas  chromatographic analysis of mixtures
     of acrylonitrile  with  acetonitrile and water.  Chim.  Ind.  51^(2): 170-71.
     (In Ital.)  Taken from:   Chem.  Abst. 70:120896p,  1969.

Dinu, V.  (1975a).  Activity of  glutathione peroxidase  and catalase and  the
     concentration  of   lipid peroxides  in  acute  intoxication with  acryloni-
     trile.  Rev.  Roum. Biochim.  12(1):11-14.

Dinu, V.  (1975b).   Intracellular thiol  concentration in acrylonitrile intoxi-
     cation.  Rev. Roum. Biochim.  12(3);155-58.

DiPasquale,  G.;  Dilorio, G.; Capaccioli,  T.;  Gagliardi, F.; and Verga,  G.R.
     (1978).  Gas chromatographic head-space determination of residual acrylo-
     nitrile  in acrylonitrile-butadiene-styrene resins  and migration  into  a
     simulated fatty foodstuffs  liquid.   J. Chromatogr.  160(1):133-40.

Doll, R.,  and A.B.  Hill.  1952.   A study  of the aetiology of carcinoma of the
     lung.  Br.  Med. J. 2:1271-1286.

Doll, R.  1971.   Weibull distribution of  cancer.  Implications  for  models  of
     carcinogenesis. J. Roy. Statistical So. A13:133-166.

Dripps, Robert  D.,  J.E. Eckenhoff,  and L.D. Vandam.   1977.   Introduction  to
     anesthesia,  the  principles   of  safe  practice.   5th  Ed.  W.B.  Saunders
     Company, Phil.  Pa. pp.  121-123.

Dudley, H.C., and  Neal, P.A.  (1942).   Toxicology  of  acrylonitrile.   I.   A
     study of the acute toxicity.   J.  Ind. Hyg. Toxicol. 24:27-36.

Dudley, H.C.; Sweeney,  T.R.;  and Miller,  J.W.  (1942).   Toxicology of acrylo-
     nitrile  (vinyl  cyanide).   II.   Studies of effects  of daily inhalation.
     J. Ind. Hyg.  Toxicol.  24:255-258.

Durkin, P.;  Sugatt,  R. ; Colman,  J.;  and Ryan,  E. (1981).  Ambient Water Qual-
     ity  Criteria Document  for  Ammonia.  Prepared for  the  Environmental Cri-
     teria  and  Assessment  Office,  U.S. Environmental Protection Agency,  Cin-
     cinnati, OH.

Ehrenberg, L. and S. Ostermar-Golkar.   1980.  Alkylation of marcomolecules for
     detecting  mutagenic agents.  Terato.  Carcinog.  and  Mutag.  1:105-127.

Ehrenberg,  L. 1979.  Risk  assessment of ethylene oxide  and other compounds.
     Banbury  Report 1:   Assessing Chemical Mutagenesis-The Risk  to Humans.
     (Eds.,  V.K.  McElheny  and S.  Abrahamson),  Cold  Spring  Harbor Laboratory.

FASEB.  1974.   Biological Data Books.  2nd ed. Vol. III.  Edited by Philip L.
     Altroan  and  Dorothy  S. DHtmen.  Federation of American Societies for
     Experimental  Biology.   Bethesda, MD.  Library of Congress No. 72-87738.
                                       R-6

-------
 Fasitta,  V.A.,  and Licciardello, G. (1977).  Determination of acetonitrile and
      acrylonitrile in  the air by gas chronatography.  Ann. 1st. Super. Sanita.
      13(1-2) :245-48.   (In  Ital.)  Taken  from-.   Chem. Abst.  89:11151p,  1978.

ITassett, D.W. (1963).  Cyanides  and nitriles.  In:  Toxicology.  Passett, D.W.,
     and Irish, P.O., editors.  Vol. 2 of Industrial Hygiene and Toxicology, 2nd
     ed.  Mew York:   John Viley  and Sons,  Inc.,  pp. 2009*2012.

Federal Register (197Ba).  Emergency temporary standard for occupational expo-
     sure   to   acrylonitrile   (vinyl    cyanide):      Notice   of   bearing.
        11) :2586-2607.
Federal  Register  (1978b).    Occupational  exposure  to  acrylonitrile  (vinyl
     cyanide) .  l£( 192) :U5762-U58l9.
Federal   Register  (1979).      Method   603.   acroleln   and  acrylonitrile.
     jyj.{233):69l79-81.

Finkel, J.M.; James, R.H.; and Miller, B.C. (1979).  Residual Monomers in Acrylic
     and  Modacrylic Fibers  and  Fabrics.    Draft  Final  Report.    Prepared by
     Southern Research Institute, Birmingham, AL, under Contract No.  68-01-4716.
     U.S.  Environmental  Protection   Agency,   Office  of  Toxic   Substances,
     Washington, DC.

Flood, M. (1980). Memorandum from M. Flood, Div. of Chemistry and Physics.,  Food
     and  Drug Administration,  Department  of Health,  Education,  and  Welfare,
     Washington,  DC,   to  R.M.  Bruce,  U.S. Environmental  Protection Agency,
     Research Triangle Park,  NC,  dated July 17,  1980.

Freeman, R.A.; Schroy, J.M.; Klieve, J.R.;  and Archer, S.A. (1980). Experimental
     studies  on  the rate  of air  stripping of  hazardous  chemicals  from waste
     treatment systems.  APCA Meeting, Montreal, Canada,  June 1980.

Freshour, K.L.;  Ramsey, J.C.; and Braun, W.H. (1980).  Acrylonitrile:  Pharmaco-
     kinetic profile  in rat  plasma after single  oral or intravenous  administra-
     tion.   In:   Abstracts  of Papers:   Society  of Toxicology.    19th  Annual
     Meeting, Washington,  DC, March 9-13,  1980.  Abst.  No. 179, p.  A60.

Caffey, W.R.. and M.E. Strauss.  1981.  A  mortality study of workers potentially
    exposed to acrylonltHle  during startup. Monsanto Decatur plant.
    Unpublished report.                               *

Gagnon, Y.T. , and Posner,  J.C. (1979).  Recovery of acrylonitrile from charcoal
     tubes at low levels.   Amer. Ind. Hyg. Assoc.  J. JKX 10) :923-5.

Gale,  D.M.  (1970).   Mechanistic aspects  of the photochemistry  of  unsaturated
     nitriles.  J. Org. Chem. 3JK H) : 970-72.
                                       R-7

-------
Orote, A. A.; Kim,  V.S.;  and Kupel,  E.R. (1978).  Establishing • protocol from
     laboratory studies to be used in field sampling operations.  Amer. Ind. Hyg.
     Assoc. J. ££( 11): 680-6)4.

Guengerich, F.P.;  Oeiger, L.E.; Hogy, L.L.; and Wright,  P.L.  (1981).   In vitro
     metabolism of acrylonitrile to 2-cyanoethylcne oxide,  reaction with gluta-
     thione, and Irreversible binding to proteins and nucleic acida.  Cancer Res.
Guerin, M.R.; Olerich, G.;  and Norton, A.D. (197*0.  Routine gas chromatographic
     component profiling of cigarette smoke for the  identification of  biologi-
     cally significant constituents.  J. Chromatogr.  Sci.  1_2( 7) : 385-91.

Gunther, F., and Blinn, R. (1955).  Analysis of Insecticides and Acarlcldes.  New
     York:  Interscienoe Publishers.

Gut,  I.;  Nerudova,  J.;  Kopecky,  J.;  and Holecek,  V.  (1975).   Acrylonitrile
     biotransformation in rats, nice, and Chinese hamsters as influenced  by the
     route  of  administration  and by phenobarbital, SKF  525 -A,  cystelne,  dimer-
     caprol, or thlosulfate.  Arch. Toxicol.  33(2) : 151-161.

Guyot, A.; Bert,  M.;  Hamoudi,  A.;  McNeill, I.; and Grassie,  N. (1978). Pyrolysis
     of acrylonitrile-methacrylic and copolymers and  derivatives.  Eur. Polymer
     J. 1^(2): 101-07.

Hall, M.E., and Stevens, J.W., Jr. (1977).  Spectrophotometric determination  of
     acrylonitrile.  Anal. Chem. «<9(U) :2277-80.

Harmon,  E.G.  1975.  Tobacco,  1n:  V.F.  Fraumenl  (ed.).  Persons at high risk of
    cancer:  an approach to cancer etiology and control.  Academic Press, New
    York,  San Francisco,  and London.

Hartshorn,  J.H.  (1975).   Detection of  acrylonitrile In FDA  "food  simulating
     solvents."   In:   Abstracts of  Papers  of  the  American  Chemical  Society
     1975(169): 121.

Baseman,  J.K., and Hoel, D.G.  (197^) .  Table of Gehan's generalized Wilcoxon test
     with fixed point sensoring.  J. Statist. Comput. Simul. 3_:117.

Hashimoto, K., and Kanai, R. (1965). Studies on the toxicology of acrylonitrile:
     Metabolism,  mode of action,  and  therapy.    Ind. Health (Kawasaki,  Japan)
 Henderson,  C.;  Pickering,  Q.H.;  and Lemke,  A.E.  (1961).  The  effect  of some
      organic  cyanides  (nitriles)  on  fish.    In:    Proceedings  of the  15th
      Industrial Waste Conference, Purdue University J<5( 2) : 120-130.

 Herman, D.R.  1981.  Cohort mortality study of the Scotts Bluff/Baton Rouge
    Uniroyal Plant.  Unpublished report.

 Hoffman, P.; Kleine, D.; and Franzen, E. (1976).  [On  the  proof of  acrylonitrile
      exposure:   Detoxification of  acrylonitrile by coupling with D-glucuronic
      acid.]  A Gesampte Hyg.  Ihre Grenzgeb. 22( 5) : 310-312.  (In  Ger.)

 Hollingseed,  D.W.  (1978).    Letter  from D.W.  Hollingsed  of  Vitron  to  F.
      Letkiewicz  of  OTS/EPA  dated  5/3/78;  personal  communication  from D.W.
      Hollingsed  to P.  Hilgard of  OTS/EPA on  5/10/78.   (Cited  in Miller and
      Villaurae, 1978.)

 Hosaka,  S.,  and Wakamatsu,  S. (1968).    Photodimerization  of acrylonitrile.
      Tetrahedron Lett.  2:219-220.
                                         R-8

-------
Hovious, J.C.; Haggy, G.T.;  and Conway, R.A. (1973).  Identification and control
     of petrochemical pollutants inhibitory to anaerobic processes.   Available
     from:  National Technical  Informa tion Service, Springfield, ?A (RT1S PB-
     222-287).  (Cited in Miller and Villaume,  1976.)

Hughes, T.W., and Horn,  D.A. (1977).  Source Assessment:  Acrylonitrile Manufac-
     ture  (Air Emissions),  Publication  No. EPA-600/2-77-107J.   Prepared  by
     Monsanto Research  Co.,  Dayton,  OH,  under  Contract  No.  68-02-1871.   U.S.
     Environmental Protection Agency, Office of Research and Development, Indus-
     trial  Environmental  Research  Laboratory,  Research  Triangle  Park,  NC.
     Available from:  National Technical Information Service, Springfield, VA
     (NTIS PB-271-969).

IAFC  (International  Agency  for Research  on Cancer)  (1979).   Acrylonitrile,
     acrylic  and  modacrylic  fibers, and  acrylonitrile-butadiene-styrene  and
     styrene-acrylonitrile copolymers.  In:  Some Monomers, Plastics  and Synthe-
     tic Elastomers,  and Acrolein.  IARC Monographs on the Evaluation  of Carcino-
     genic Risk of Chemicals to Man.   Lyon,  France:   World Health Organization,
     IAPC, Vol. 19, PP.  73-113.

International  Commission on  Radiological  Protection.  1977.  Recommendation of
    the International Commission on Radiological Protection, Pub. No. 26.
    adopted Jan.  17,  1977.   Per gammon press,  Oxford, England.

Ivanenko,  C.S.,  and Lukashevskaya ,  A.V.  (1976).    Chromatographic method  of
     determining bound acrylonitrile  in butadiene-nitrile rubbers. Metody Anal.
     Kontrolya Proizvod. Khlm. Prom-sti.  2:19-20.  (InRuss.)  Taken from:  Chem.
     Abst. 8j4_:l8m03», 1976.

Jackson, S.,  and Brown, V.M,  (1970).  Effect of toxic wastes on treatment  pro-
     cesses and watercourses.   Water Pollut. Contr. 6£( 3) :292-313.  (Cited  in
     Miller and Villaume, 1978.)

Jacobs,  H.W., and Syrjala,  R.H. (1978).   The use  of infrared analyzers  for
     monitoring acrylonitrile.  Amer. Ind. Hyg. Assoc. J. 3JK 2) ; 161-65.

Jaeger, R.J.  (1979). Time-related variation of non-protein sulfhydryl  concen-
     trations in rat tissues and human blood.  Int. Arch. Occup.  Environ. Health
Joshi,  S.B.,  Northrop  Services Corp. (1977).  Letter  to J.  Cervello, U.S. EPA
     regarding photochemical reactivity of acrylonitrile,  dated  July 10, 1977.
     (Cited in Suta, 1979.)

Judson, C.L.; Hokama, Y.; and Bray, A.D. (1962).  The effects of various chemi-
     cals  on eggs  of  the  yellow-fever  mosquito,  Aedes  acgypti.   J.  Econ.
     Entomol. 5_5( 5) : 805-07.

Junke,  I.,  and  Ludemann,  D.  (1978).   (Results of  the study of  200 chemical
     compounds on acute  fish toxicity  using  the Golden  Orfe  test.]   Z. Wasser
     Abwasser Forsch. IK 5) : 161-61.  (In Ger.)

Kaiser, E.R., and Bergman, F. (1973).  Combustion Characteristics of Lopac Con-
     tainers.    In:   Proceedings of  the  Symposium on  Environmental Impact of
     Nitrile Barrier Containers, Hartford, CT, July 19,  1973.  Sponsored by and
     available from Monsanto Co., St Louis, MO.
                                        R-9

-------
Kankaanpaa, J.;  Elovaara,  E.; Hemminki,  K.;  and Vainio, H.  (1979).   Embryo-
     toxicity of  acrolein,  acrylonitrile and  acrylamide in  developing  chick
     embryos.   Toxicol.  Lett.  4(2):93-96.

Karaenev, S.S.; Kostov, G.; Milina,  R.;  and Mikhailov, M.Kh.  (1974).  Deter-
     mination of the composition  of  ternary copolymers by IR spectroscopy  and
     pyrolytic gas chromatography.  Vysokomol.  Soedin.,  Ser.  A.,  1£(9):  2162-
     165.  (In Russ.)  Taken from-.   Chem.  Abst.  82-.32109b,  1975.

Kato, A., and Yamamura, K.  (1976).   Treating waste water  containing  nitriles
     and cyanides.  U.S.  Patent  3,940,332.   February 24, 1976.  Taken  from:
     Chem.  Abst. 85:67783f,  1976.

Kearney, A.T. (1978).  Assessment of Acrylonitrile Contained in Consumer Pro-
     ducts.  Final  Report.   Prepared by  A.T.  Kearney, Inc., Washington,  DC,
     under Contract No.  CPSC-C77-0009.   U.S. Consumer Product  Safety Commis-
     sion,  Washington,  DC.

Keresztesy, J.;  Taylor,  T.J.; Katz,  A.;  Haddock,  J.T.; Byer, W.L.;  and Mai-
     seed,   R.T.  (1977).   Acrylonitrile.   Prepared  by Auerbach Associates,
     Inc.,  for  the U.S. Consumer Product Safety  Commission, Washington,  DC.
     Available from:  National Technical  Information  Service, Springfield,  VA
     (NTIS PB-280-478).

Kier, L.D.  1982.   Ames/Salmonella mutagenicity assay  of  acrylonitrile.  Mon-
     santo Company, St. Louis.  Report  No:  MSL-2063,  Job/Project No:   800282-
     ML-80-441.

Kier, L.D.  1982.  Ames/Salmonella  mutagenicity assay of acrylonitrile  epoxide.
     Monsanto  Company,   St.   Louis.   Report  No:  MSL-2064,  Job/Project  No:
     800283-ML-80-442.

Kiesselbach,  N.,  U. Korallis,  H.J.  Lange,  A.  Neiss,  and T. Zwingers.  1980.
     Bayer-ACN-Studie  1977.   Zentralblatt  fur  arbeitsmedizin,  arbeitsschutz,
     prophylaxe und ergonome, Band 7.

 Kihlman,  B.A.  (1961).  Cytological effects  of  phenylnitrosaraines.   I.   The
      production of structural chromosome changes  in the presence of  light and
      acridine orange.   Radiat.  Bot.  1^35-42.

 Kincannon, D.F., and Stover,  E.L. (1981).  Stripping characteristics  of  prior-
      ity pollutants  during  biological  treatment.   74th  Annual  AICHE Meeting,
      New Orleans, LA,  November 1981.

 Kleshcheva,  M.S.; Usacheva,  V.T.;  and  Pozharova.  V.N.  (1971).   Determination
      of residual  monomers   and  non-polymerizing  impurities  in  polystyrene
      plastics  by  a gas chromatographic  method.   Plast.  Massy 7:57-58.   (In
      Russ.)   Taken from:   Chem. Abst. 75:130411a,  1971.

 Klyaev, V.I.;  Slisarenko,  F.A.; and  Finkel'shtein, A.V.  (1966).   Polarographic
      determination of  acrylonitrile.  Fiz.-Khim. Issled.  Prir.  Sorbentov Ryada
      Anal.  Sist.  1^27-33.    (In  Russ.)  Taken  from:   Chem.  Abst.  69:64465n,
      1968.

 Knoblock,  K.;  Szendzikowski,  S.;  Czajkowska, T.; and  Krysiak, B.  (1971).  Acute
      and subacute toxicty  of  acrylonitrile.  Med.  Pracy 22(3):257-69.

                                        R-10

-------
Kobayashi, Y.  (1956).   Determination of  acrylonitrile vapor by means  of  de-
     tector tubes.   Yuki Gosei Kagaku Kyokai Shi JL4:673-75,  1956.   (In Japan.)
     Taken from:   Chem.  Abst.  5J.:7240i,  1957.

Kondratenko,  V.I.;  Geller,  B.A.; and Khaskin,  I.G.  (1971).   Chloro derivatives
     of aliphatic nitriles.   IV.   Hypochlorination  of acrylic and methacrylic
     acid nitriles.  Zh.   Orgn.  Khim.  7(7):1343-47.   Taken from:   Chem. Abst.
     7_5:129033r,  1971.

Kopfler,  F.C.; Melton,  R.G.;   Lingg,  R.E.; and  Coleman,  W.E.  (1976).   GC/MS
     determination  of  volatiles  for   the National  Organics  Reconnaissance
     Survey  (NORS)  of  drinking water.   In:   Identification and Analysis  of
     Organic  Pollutants in Water.   Keith,  L.H.,  ed.  Ann Arbor,  MI:   Ann Arbor
     Science  Publishers, Inc., p. 87.

Korzhova,  I.T.;  Kleshcheva, M.S.; Porodina, M.N.; Korkhova,  R.D.;  Balandina,
     V.A.; and Khokhlov,  V.A.  (1974).   Gas chromatographic  determination of
     toxic concentrations  of  styrene  and  acrylonitrile   vapors  during  the
     synthesis of  polystyrene  plastics.    Plast.  Massy 5:67-9.   (In  Russ.)
     Taken from-.   Chem.  Abst.  81.: 110759J ,  1974.

Kostin, L.D., and Vidanova, V.A. (1957).  Quantitative determination of acrylic
     acid  nitrile.   USSR  Patent  104,643.  January  25,  1957.   Taken  from:
     Chem. Abst.  5_l:7246e,  1957.

Krynska,  A.  (1970).   Method  of determining acrylonitrile  in air.   Pr. Cent.
     Inst. Ochr. Pr.  20(67):303-10.   (In  Pol.) Taken  from:  Chem.  Abst.  74:
     90798w,  1971.

Kuparov,  P.B.; Klyuev,  N.A. ,-  Mal'tseva, T.A.; and Palagushkina,  L.A.  (1977).
     Quantitative determination  of  acrylonitrile  in  SKN vulcanizates by an IR
     spectroscopic  method.   Izv.  Vyssh.  Uchebn.  Zaved., Khim. Khim.  Techknol.
     20(11):1638-640.    (In Russ.)  Taken from:   Chem. Abst.  88: 63091t, 1978.

Lambotte-Vandepaer, M. ;  Duverger-Bogaert,  M.;  deMeester, C.; Poncelet,  F. ,-  and
     Mercier, M.  (1980).   Mutagenicity of urine from  rats and mice  treated
     with acrylonitrile.  Toxicology 16:67-71.

Langvardt, P.W.;  Putzing,  C.L.;  Young,  J.D.;  and Braun, W.  H. (9179).   Isola-
     tion and  identification  of urinary metabolites  of vinyl-type  compounds:
     Application  to  metabolites  of acrylonitrile  and  acrylamide.   Toxicol.
     Appl. Pharmacol.  48(1):161.

Lank,  J.C.,  Jr.  (1970).   Effect of Acrylonitrile  on  Anaerobic  Digestion of
     Domestic  Sludge.   Thesis,  University of Idaho  Graduate  School,  March
     1970, 76 pp. (Cited in Miller and Villaume, 1978).

Latt,  S.A.   1974.  Sister  chromatid exchanges,  indices of  human  chromosomes
     damage  and  repair:  detection of fluorescence  and induction by mitomycin
     C. Proc. Natl. Acad.  Sci. (U.S.A.) 71:3162-3166.

Lawniczak, H.  (1977).   Quantitative determination of vinyl cynaide  in indus-
     trial effluents.   Barwniki-Srodki Promocnicze  2_1(4) :87-91.    (In Pol.)
     Taken from:   Chem.  Abst.  89:220318n,  1978.
                                        R-ll

-------
 LeBlanc,  G.A.  (1980).   Acute toxicity of  priority pollutants to  water flea
      (Daphnia aagna).  Bull.  Environ. Contain. Toxicol. 24:684-91.

 Lee,  W.R., S.  Abrahamson,  R. Valencia, E.S. Von  Halle,  F.E.  Wurgler,  and S.
      Zimmering.   In  press.   The  sex-linked recessive lethal  test for Buta-
      genesis  in Drosophila melanoqaster.  Mutat. Res.

 Leo,  A.;  Hansch, C. ,- and Elkins,  D.  (1971).  Partition coefficients and their
      uses.  Chem. Rev. 71_(6) -.525-616.

 Leonard,  A.,  V.  Garny,  F. Poncelet, and M.  Mercier.   1981.  Hutagenicity of
      Acrylonitrile  in Mouse.  Toxicol.  Lett.  7:329-334.

 Lezovic,  A.,  and Singliar, M.  (1977).  Polarographic  determination of acryla-
      nide and  acrylonitrile.  Petrochemia  17(5-6) :128-32.    (In  Slo.) Taken
      front:  Chen. Abst.  8916338q,  1978.

Lindgren,  D.L.4,  fincent,  L.E.; and Krohne, B.E.  (1954). Relative effectiveness
     of ten fuaigants to adults  of tight apeoiaa of etored-produet inaeets.
     J. Econ.  Entoool. *7_: 923-26.

Linetskii, V.A.,  and Serebryakov,  B.R. (1965).   The alkaline hydrolysia of
     nltriles.  Khia.  Proo. JM(7) :H98-99.

Loveless,  A. (1951).  Qualitative  aspects of the chemistry and biology of radio-
     mimetic (mutagenic)  aubstances.  Mature (London)  16?:33B-I>2.

Loveless,  L.E.; Spoerl,  E.;  and Weisman,  T.H. (195^).  A aurvey of effects of
     chemicals  on  division  and   growth   of  yeast  and  Eseherichia  coli.
     J. Bacteriol. 68:637-W.

Ludzack,  F.J.;  Schaffer,  R.B.;  Bloomhuff,  R.N.;   and Ettinger,  M.B.  (1958).
     Biochemical oxidation of some commercially important  organic cyanides.  I.
     River  water  oxidations.    In:    Proceedings  of  the  Industrial Vaste
     Conference, pp. 297-312.

Ludzack, F.J.; Schaffer,  R.B.; and Bloomhuff, R.N.  (1961).  Experimental treat-
     Bent  of  organic cyanides  by  conventional  processes.    J. Water  Pollut.
     Contr. Fed.
        I.  (I960).   Gas-chromatographic  analysis  of nltriles.   Anal.  Chem.
 Magos,  L.  (1962)..  A etutfy of acrylonitrile poisoning in relation to »ethaeao-
      globin-CK oomplex  formation.  Brit. J. Ind. Med. I9_:283-286.

 Hal let te,  F.S. (19^3) .  Industrial hygiene  in synthetic rubber manufacture.  Ind.
      Med.
 Maltonl, C.;  Cillberti,  A.;  and DlWaio, V. (1977). Carclnogeniclty bloassays on
      rats of acrylonitrile administered by  inhalation  and  by Ingestion.   Med.
      Lavoro 68(6) :U01-«11.
                                         R-12

-------
 Mtntel, *.. and M.A. Schntldenun.   1977.   CfttlMtlng •§•!«• Uvtll. • Mzardous
     undertaking.  Cancer fees.  35:1379-1586.
Marino,  R.S.;  levine,  8. P.;  and Banrey, T.M.  (1976).   Tract determination of
     aubnanograa amounts of acrylonitrile in complex aatrioas by gas ehromato-
     graphy with a nitrogen-ael active detector.  Anal. Cheo. 50(13) : 194 8-50.

Marooni, S., and lonescu, M. (1974).  Toxioation of acrylonitrile on fiah.  Stud.
     Prot. Calitatii Apelor 16:7-22.  (Cited in Miller and Villaume, 1976.)

Markelov, M.A.,  and  Semenenko,  E.I. (1976).  Determination of •icroeoncentra-
     tions  of  substances migrating  froo  polymer materials to  water.   Plast.
     Massy 1:57-59.  (In Buss.)  Taken from:  Chea.  Abst. 84:155747g, 1976.

Markelov, M.E.; McMaster, A.; and Meyer, N. (1981).  Analysis and Identification
     of  organic  impurities  in  water  by  a headspace  gas   chrooatographic
     technique,  presented at  the  6th Annual  Meeting of  the Federation of
     Analytical Chemistry and Applied Spectroscopy Societies, Philadelphia, PA,
     Septemer 20-25, 1981.

Marrs, D.B.; Campbell,  R.L.; Boyko,  J.M.; Parrish, A.L., Jr.;  and Standish,  N.W.
     ( 1976) .  Measurement of acrylonitrile in ambient air in the area of plastic
     processing equipment.  Paper presented at the 1976  National Technical  Con-
     ference of the Society of  Plastics Engineers,  Inc., Chicago,  IL, November
     14, 1978.  Available from:  The Standard Oil Company, Cleveland, OH.

Martin, G.M. (1976).  Acrylonitrile.  Job Safety  Health 6(3) :*-13.

McCann,  J.;  Choi, £.;  Yamasaki, £.;  and Ames,  B.N.  (1975a).   Detection of
     carcinogens as  mutagens  in the Salmonel la/micros ome test:   Assay of 300
     chemicals.  Proc.  Nat. Acad. Sci. USA 21:5135-5139.

McCann, J.; Splngarn, N.E.; Kobori,  J.;  and Ames, B.N. (1975b).   Detection of
     carcinogens as mutagens:   Bacterial tester strains with  R  factor  plasmids.
     Proc. Nat. Acad. Sci. J2( 3) :979-83.
McMahon, B.E.; Cllne, J.C.;  and Thompson, C.Z. (1979).  Assay of 855 test chemi-
     cals in ten tester strains using a new modification of the Ames test  for
     bacterial mutagens.  Cancer Res. 39( 3) ; 682-93.

McNeal, T.; Brumley, V.C.; Breder,  C.; and Sphon, J.A. (1979). Gas-solid chrooa-
     tographic-mass spectrooetrlc confirmation of  low levels of acrylonitrile
     after distillation from  food-simulating solvents.   J.  Assoc.  Off. Anal.
     Chem. 62(l):4l-46.

McNeil, T., and Broader, C.V. (1981).  The determination of sub-ppb quantities of
     acrylonitrile  in  3?  acetic  acid  using  manual  headspace  gas  solid
     chromatography.  U.S. Food and  Drug Administration,  draft procedure, June,
     1981.

McOmie, W.A.  (19^3).   Comparative  toxlcity of  methacrylonitrile  and  acrylo-
     nitrile.  J.  Ind. Hyg.  Toxicol. 3J.:113-H6.


                                        R-13

-------
 Mekhtiev, S.I.; Mamedova, V.M.;  Guseinov,  A.G.;  Abaahln, V.C.; and Agaeva, I.E.
      (1968).   Polarographic determination of acrylonitrile  and Bethacryloni-
      trile.   Azerb.  Heft. Khoz. ^[(6):33-3^.   (In  Ruas.)   Taken froo:  Chem.
      Ab3t. 68:113370p, 196B.

 Miller, L.M.,  and Villaume,  J.E.  (1978).   Investigation of Selected Potential
      Environmental Contaminants:  Acrylonitrile.  Publication Ho. EPA-560/2-78-
      003.     Prepared  by   the  Franklin  Institute  Research  Laboratories,
     .Philadelphia, PA,  under  Contract No.  68-01-3893.    U.S.  Environmental
      Protection Agency, Office  of Toxic Substances, Washington, DC.  Available
      froo:  National  Technical  Information Service, Springfield,  VA (NTIS PB-
      285-881).

 Mills, E.J., Jr.,  and Stack, V.T., Jr. (1955). Acclimation of microorganisms for
      the oxidation of pure  organic  chemicals.   In:   Proceedings  of the Ninth
      Industrial Waste Conference,  Purdue University, IN, pp. 4U9-6II.

 Milvy, P., and Wolff, M. (1977).  Mutagenlc studies  with aorylonitrile. Mutat.
      Res. jlSCS-1*): 271-278.

 Minami, M.;  Sato, M.;  Suzuki, A.; Tanamura, K.; and lahizu, 5.  (1973)*   Studies
      on acrylonitrile  poisoning.  1.   On the result of exposure to acrylonitrile
      vapor.    Tokyo Joshi Ika Diagaku  Zasshi j£(10):8U9-S55.   (In Jap.; Eng.
      aummary.)

 Monson, R.R.   1978.   Mortality  and cancer morbidity among  chemical  workers with
     potential  exposure  to acrylonltrlle.


Morris, E.D.,  Jr., and Niki, B. (1971).   Reactivity of nydroryl radicals with
     olefins.  J. Phys. Chem. 7£(23):3640-41.

Murray, F.J.; Nltschke, J.A.; John, P.A.; Smith, F.A.; Quast, J.F.;  Blogg, C.D.;
     and Sehwetz, B.A. (1976).  Teratologlc Evaluation of Acrylonitrile  Monomer
     Given to Rats  by Gavage.  Prepared by Toxicology  Research Laboratory, Health
     and Environmental Research, Dow Chemical USA,  Midland,  MI, for Chemical
     Manufacturers  Association,  Washington,  DC.     Available  from:    CHA,
     Washington, DC.

Murray, F.J.;  Schwetz,  B.A.;  Nltschke, K.D.;  John,  J.A.;  Norris,  J.M.;  and
     Gehring,  P.J. (1978).  Teratogenicity of acrylonitrile  given to  rats by
     gavage or by inhalation.  Food Cosmet. Toxicol.  16( 6) :5*<7-552.

Narasiahan, K.S.; Shurpalekar, S.R.;  and Venkatesh, K.V.L. (1972).  Use of fumi-
     gants  in  prevention  of  insect  infestation  and  Bold  growth  in  papads.
     J. Food Scl. Technol. 9_( 3): 13^-37.

National Academy of Sciences (1975).   Acrylonitrile Assessing Potential Ocean
     Pollutants, A Report  of the Study  Panel  on  Assessing  Potential  Ocean Pol-
     lutants to the Ocean Affairs Board  Committee on Natural Resources, National
     Research Council, pp. 209-227.  (Cited in Miller and Villaume,  1978..)

National Institute  for Occupational  Safety and Health.   1978.   Trip  report to
    review environmental monitoring and personal  work history records at Dupont
    May plant. Camden, South  Carolina, by Jane Brown. N10SH (April  17).
    Submitted to Branch Chief of Industrial Workers Safety  Board, MOSH.
                                          R-14

-------
 Nazarova, V.I., and Nakrap, L.I. (1978).  Determination of acrylonitrile in air.
      Melody  Anal.  Kontralya  Kach.  Prod. Khim. Prom-sti.  £:28-29.   (In Buss.)
      Taken from:   Chem. Abst. 2p_:2826lt, 1979.

 Neely,  V.B.; Branson, D.R.;  and Blau,  G.E.  (1971).  Partition  ooeffioient  to
      •easure bioconcentration potential of organic chemicals in fish.  Environ.
      Sci. Technol. 8(13) : 1113-15.

 Nelson, C.O., and Harder,  C.A. (197M).  Respirator cartridge efficiency studies.
      V.  Effect of solvent vapor.  Aaer. Xnd. Byg. Aasoc. J. 35J 7) : 391-410.

 Nestler,  H., and  Berger,  V.  (1965).   Gas  chromatographic analysis  of water
      containing  solutions of acrylonitrile  and various  esters in dimethyl-
      fonnamide.  Chem. Tech.  17( 3) :169-70.  (InGer.)  Taken from:  Chem. Abst.
      63_:1213d, 1965.

 NIOSH (National Institute for Occupational Safety and Health) (1977a).  Acrylo-
      nitrile Method No. S156.   NIOSH  Manual of  Analytical  Methods,  2nd ed.
      Cincinnati, OH:  U.S.  Dept. of Health, Education, and Welfare, Public Health
      Service, Center  for Disease Control, NIOSH, Vol. 3.

 NIOSH (National Institute  for Occupational Safety and Health)  (1977b).  Method
      No. P and CAM 202.  NIOSH Manual of Analytical Methods, 2nd ed.  Cincinnati,
      OH:   U.S. Dept.  of Health, Education,  and Welfare, Public Health Service,
      Center  for Disease Control, NIOSH, Vol. 1.

NIOSH (National Institute  for Occupational Safety and Health) (1977c).  Current
     Intelligence  Bulletin 18:  Acrylonitrile.  Current Intelligence Bulletin:
     Acrylonitrile. July 1, 1977.  U.S. Dept. of Health, Education, and Welfare,
     Public Health Service, Center for Disease Control, NIOSH.  Available from:
     NIOSH, Rockville, MD.

Norris,  M.B.  (1962).  Acrylonitrile.   Encycl.  Ind. Chem.  Anal. £:368-8l.

O'Berg,  M. (1980).   Epldemlologic study  of workers exposed to  acrylonitrile.
     J.  Occup. Med. 22:2^5-252.

Ofengand,  J.  (1967).  The function of pseudouridylic acid in transfer ribonucleic
     acid.   I.  The specific cyanoethylatlon of  pseudourldine, inosine, and 4-
     thiouridine  by acrylonitrile.   J. Biol. Chem.  2H2(21) t
Ofengand, J. (1971).  Cyanoethylation of nucleotides and tRNA  by  acrylonitrile.
     In:  Methods of Enzymology.  New York:  Academic Press, Vol. 20, pp. 150-68.

Panova, R.V.;  Belova,  R.A.; and Lavrushina, T.S. (1969).  Chromatographic deter-
     mination of residual monomers in latexes.   Prom.  Sin. Kauch.,  Nauch.-Tekh.
     Sb. .1:21-23.   (In Russ.)   Taken from:   Chem.  Abst. 7JUl3238d,  1971.

Parent, R.A.,  and Casto,  B.C.  (1979). Effect of acrylonitrile on  primary Syrian
     golden hamster embryo cells in culture:  Transformation and  DNA  fragmenta-
     tion.  J. Nat. Cancer Inst. 62( «4) : 1025-29.

Patterson, R.M.; Bornstein, M.I.; and Carshick,  E. (1976).  Assessment  of Acrylo-
     nitrile as a Potential Air  Pollution Problem,  Publication No.   GCA-TR-75-
     32-0(6).   Prepared  by CCA Corporation, CCA/Technology Division,  Bedford,
     MA, under Contract  No. 68-02-1337.  U.S.  Environmental Protection Agency,
     Research Triangle Park,  NC.

                                        R-15

-------
 Paulet, C.t and  Vidal,  Mrs.  (1975).   [Toxicity of aone acrylic and •ethaerylie
      esters of acrylamide and polyacrylaaides .]   Arch. Mai. Prof.  Med.  Trav.
      Secur. Soc. 26(1-2) : 58-60.   (In Fr.)   Taken from:  Chem. Abst. 83j9l832a,
      1975 •

 Pell, $., C.I. duPont de Nemours and  Co.  1980.  Letter to the Carcinogen
     Assessment Group. U.S. Environmental Protection Agency (July 23).
 Perry,  F.H.,  and Chi It on,  C.R.,  editors (1973).  Chemical Engineers* Handbook,
      5th ed.  New York:  McGraw-Hill Book Co., pp. 3-61.

 Peters, T.L.  (1980).   Steam distillation apparatus  for concentration of trace
      water soluble organics.  Anal. Chem. J52_:21 1-213.

 Petrova,  L.I.;  Maksimova, N.S.;  and Boikova,  Z.K.  (1978).    Determination of
      traces of atyrene, acrylonltrile, and nethacrylate monomers in binary mix-
      tures.    Gig.  Sanit.  14:814-86.    (In  Buss.)    Taken from:    Chem.  Abst.
      2P_:6935P, 1979.

 Pilotti,  A.;  Ancker, K.;  Arrhenius, E.;  and Enzell,  C.  (1975).   Effects of
      tobacco  and tobacco smoke  constituents  on cell  multiplication  in vitro.
      Toxicology ^(1): 19-62.
Plieninger, H., and Sharma,  S.  (1978).   Mote on a simple detection of acrylo-
     nitrile in solvents.  Justus Liebigs Ann.  Chem.  3j535-36.  (In Cer.) Taken
     from:  Chem. Abst. 8_9_:52913g, 1978.

Pokrovskaya, L.A., and Prolova,  C.S.  (1969)*  Chromatographic method in produc-
     tion  control  of  butadiene-acrylonitrile  rubbers.    Piz.  Khim.  Khim.
     Tekhnol., pp. 201-06.  (In Puss.)  Taken from:  Chem. Abst. 15_:9913**e, 1971.

Ponomarev, Yu P.;  Anufrieva,  T.L.; and Shkorbatova, T.L. (19710.  Polarographic
     determination  of  acrylonltrile  In  waste   waters  using  extraction.
     Vodosnabzh.,  Kanallz.,  Cldrotekh. Sooruzh.  n.:67-70.  (In  Russ.)   Taken
     from:  Chem.  Abst. tte^SOSx, 1975.
Portmann, J.E., and Wilson,  K.E. (1971).  The toxiclty of 1*40 substances  to the
     brown shrimp and other marine animals. Shellfish Inf. Leafl. No. 22,  12 pp.
     Minist.  Agric.,  Fish,  and Food,  Fish.  Lab.,  Burnham-on-Crouch,  Essex,
     England.

Quast,  J.F.;  Enriquez,  R.M.;  Wade, C.E.;  Humiston,  C.G.;  and Schwetz,  B.A.
     (1977).  Toxicity of Drinking Water Containing Acrylonitrile (AN)  in Rats:
     Results after 12 Months.   Prepared  by the Toxicology Research  Laboratory,
     Health and Environmental Research, Dow Chemical USA, Midland,  MI, for the
     Chemical Manufacturers Association, Washington, DC.   Available  from:   CMA,
     Washington, DC.

Quast,  J.F.;  Humiston,  C.G.;  Schwetz, B.A.;  Frauson,  L.A.;  Wade, C.E.;  and
     Norris, J.M. (1975) .  A Six-Month Oral Toxicity Study Incorporating Acrylo-
     nitrile in  the  Drinking Water of  Purebred Beagle Dogs.   Prepared  by the
     Toxicology  Research  Laboratory,  Health  and  Environmental Research,  Dow
     Chemical, USA,  Midland,  MI,  for  the Chemical Manufacturers Association,
     Washington, DC.  Available from:   CMA,  Washington,  DC.
                                       R-16

-------
  Quasi,  J.F.;  Wade, C.E.; Humiston,  C.G.;  Carreon,  R.M.; Hermann, E.A.;  Park,
       C.N.; and Schwetz, B.A. (1980a).  A Two-Year Toxicity and Oneogenlcity Study
       with Acrylonitrile Incorporated in the Drinking Water of Rats.  Prepared by
       the Toxicology Research Laboratory, Health and Environmental Sciences,  Dow
       Chemical  USA, Midland,  MI,  for  the  Chemical Manufacturers  Association,
       Washington, DC.  Available from:  CMA, Washington, DC.

  Quasi,  J.F.; Schuetz, D.J.;  Balmer,  M.F.; Gushow, T.S.; Park, C.N.; and McKenna,
       M.J.  (198Ob).   A  Two-Year Toxicity  and  Oncogenicity Study  with Acrylo-
       nitrile Following Inhalation Exposure of flats.  "repared by the Toxicology
       Research Laboratory, Health and Environmental Sciences, Dow Chemical, USA,
       Midland,  MI,  for the Chemical  Manufacturing  Associalion,  Washinglon,  DC.
       Available from:  CMA, Washington, DC.

  Rabello-Gay, M.N., and Ahmed, A.E.  (1980).  Acrylonitrile:   In vivo cytogenic
       studies in Bice  and rats.  Mutat. Res. 79_:2l9-255«

  Radimer, G.F.; Davis, J.H.; and Ackernan,  A.B. (1971).  Fumigant-induced toxic
       epidermal necrolysis.  Arch. Dennatol. 110(1);103-01*

Rajendran,  S.,  and Muthu,  M. (1976).  Toxicity of acrylonitrile to the adulta of
     five  apecies  of  atored  producls  insects.      Bull.   Grain  Technol.
     11( 3): 179-181.

Randall,  T.L.,  and Knopp,  P.V.  (1980).  Detoxification  of  specific organic  aub-
     slances by wel oxidation.   J.  Water Pollut. Contr.  Fed. £2:2117-2130.

Reichle,  A., and Tengler,  H. (1968).   Gas-chrooatographic  procedures for deler-
     nining residual monomers in the polymer.  Delerminalion of acrylonitrile in
     polyacrylonitrlle.  Plastverarbeiter ^9_( 12):921-21.  (InGer.)  Taken from:
     Chen.  Abst. £0:97206c, 1969.

Renn,  C.E. (1955).  Biological  properties  and behaviors of cyanogenic  wastes.
     Sewage Ind. Wasles 27( 3):297-310.  (Cited in Miller and Villaume,  1978.)

Roberfroid,  M.; Poncelet, F.; Lambotte-Vandepaer, M.;  Duverger-VanBogaert, M.;
     DeMeester,  C.;  and Mercier, M. (1978).  Acute biotoxic effect of atyrene on
     rat  liver.   Correlation  with enzyme-mediated nutagenicily  of  benzypyrene
     and  acrylonitrile.  Scand.  J.  Work Environ. Heallh 1(suppl.  2):163-68.

Rogaczewska, T. (1961).   Polarographic determinallon of acrylonitrile  in  air.
     Chem. Anal. $ 3):117-23.  (In Pol.)  Taken from:   Chem. Abst.  62:995d, 1965.

Rogaczewska, T. (1975).   [Percutaneous  absorption  of  acrylonitrile vapors  in
     animals.l   Med. Pracy 26:159-65.  (In Pol.)

Rogaczewska, T.  (1976).   Acrylonitrile determination  in air.   Med.  Pracy
     27(2):115-26.  (In Pol.)   Taken  from:   Chem. Absl.  86:21137r, 1977.

Roudabush,  R.L.; Terhaar,  C.J.; Fassett, D.W.; and Dziuba, S.P. (1965).  Compara-
     tive acute  effects of some chemicals on  the  skin of rabbits and guinea  pigs.
     Toxicol. Appl. Phannacol.  l(1):559-65.

Roy, S.S. (1977).   Titrimetric  delerminalion of residual monomers  in  olyrene-
     acrylonitrile copolymers.   Analyst 102( 1213);302-05.
                                          R-17

-------
 Russell,  J.V.  (1975).   Analysis of air pollutants using sampling tubes and gas
      chromatography.  Environ. Sci. Technol. £(13) :1175-78.

 Russkikh,  A. A. (1971).   Determination of acetone oyanohydrin and acrylonltrile
      in air.   Piz.-Rhim. Metody Anal. Kontr. Proizvod., Mater. Konf. Rab. Vuzov
      Zavod.  Lab.  Yugo-Vostoka SSSR, 4th 1:100 (Pub.  1972).   (In Russ.)   Taken
      from:  Chen. Abst. tU:l6301g,  1974.

 Russkikh,  A. A. (1973).   Photometric determination of  acrylonitrile and neth-
      acrylonitrile  in  air.   Zavod.  Lab. 29_(l):5-6.  (In Russ.)   Taken from:
      Chem. Abst. J_9_:101566e,  1973.

 Sakurai, H.; Onodera, M.;  Utsunoaiya,  T.; Minakuchi, H.;  Iwai,  H.; and Matsumura,
      H. (1978).  Health  effects  of acrylonitrile  in acrylic fiber factories.
      Brit. J.  Ind.  Med. 25:219-22.

 Sandberg, E. Ch., and Slanina, P.  (1980). Distribution  of [I-^CJ -acrylonitrile
      in rat and monkey.  Toxicology Lett. 6:187-191.

Sana one, E.B.;  Tewari,  I.E.; and Jonas, L.A. (1979).  Prediction of removal of
     vapors from air by  adsorption on  activated carbon.  Environ. Sci. Technol.
     12(12): 1511-13.

Santodonato,  J.; Basu,  D.; and Howard, P.H. [19791 .  Multimedia human exposure
     and carcinogenic risk  assessment for environmental PAH.  In:  Chemistry and
     Biology of Pol/nuclear Aromatic  Hydrocarbons.  (Proceedings  of  the lith
     International  Battelle  Symposium on  Polynuclear  Aromatic  Hydrocarbons,
     Columbus,   OH,  October 2-1,  1979.)   Ann  Arbor, MI:   Ann  Arbor Science
     Publishers, Inc. (Book in press.)

Sarofim, A. P.;  Wagner,  J.P.;  and Tewarson, A. (1973).  Gas evolution  during the
     pyrolysis  and combustion of acrylonitrile/styrene polymers.  In:  Proceed-
     ings of the  Symposium on  Environmental  Impact of Nitrile  Barrier  Cont-
     ainers,  Hartford,  CT, July  19,  1973.   Sponsored  by and  available from
     Monsanto Co., St.  Louis,  MO.

Sartorelli, E.  (1966).   [Acute poisoning due  to acrylonitrile.)   Med.  Lavoro
     52:184-87.  (In Ital.)

Scheddel,  R.T.  (1958).  Infrared quantitative analysis data:  Analysis of acrylic
     acid,  ethylbenzene,  methyl  methacrylate,   °-methylstyrene,   styrene, and
     acrylonitrile mixtures.  Anal. Chem. 2
Sevast'yanova, I.G.,  and  Tomilov, A. P.  (1963).    Reduction of  nitriles  on a
     dropping mercury cathode.   Zh.  Obshch. Khim.  ^1( 9) : 2815-19.   (In Russ.)
     Taken from:   Chem.  Abst. 60:1583e,  19&U.

Sevast'yanova, I.G.; Tomilov, A. P.; and  Yanaleev,  I.Ya. (1966).   Polarographic
     determination  of  acrylonitrile  in  industrial solutions.    Zavod.   Lab.
     22(10): 1210.  (In Russ.)  Taken from:  Chem. Abst. 66:31651r,  1967.

Silverstein,  L.G. (1977).   Validation of abcor GASBADGE for  acrylonitrile and
     improved desorption efficiency.   Amer. Ind.  Hyg. Assoc. J. 58(8) : 112-13.

Sloof, W. (1978).  Biological monitoring based on fish respiration for continuous
     water quality control.   In:   Aquatic Pollutants:   Transformation and Bio-
     logical  Effects.   Hutzlnger,  0.; VanLelyveld,  L.H.; and  Zoeteman,  B.C.J.,
     editors.  Elmsford, NY:  Pergamon Press, pp. 501-06.


                                        R-18

-------
 Smith, C.J. (1975).   Solvent vapour phototoxicity.  Lab. Pract. 2^:9-11.

 Smyth, H.F., Jr., and Carpenter, C.*.  (1918) .  Further experience with the range
      finding  test  in  the  industrial  toxicology  laboratory.    J.  Ind.  Hyg.
      Toxicol. 20:63-68.

 Smyth, H.F.,  Jr.;  Well,  C.S.;  West,  J.S.;  and Carpenter,  C.P.  (1969).   An
      exploration of  joint toxic  action:   Twenty-seven industrial  chemicals intu
      bated  in  rats  In  all  possible  pairs.    Toxicol.  Appl.   Pharmacol.
      1JU 2): 3^0-07.

Stallard,  C.  1982.   AcrylonHMle  epidemiology.  Standard 011 Company of Ohio
    Unpublished report.


Stanford Research Institute  (1978).  1978 Chemical Economics Handbook.   Menlo
     Park, California, 1978.  p. 607.5032J.

Steere,  N.V.  (1968).   Safety in  the chemical laboratory.   XLV1.   Hazardous
     chemicals data.  J. Chem. Ed. ££( 3) :A199-A205.
 Stefanescu, T., and Ursu, Oh. (1973)'  Method for acrylonitrile and acetonitrile
     determination in  residual waters.   Mater.  Plast.  10(6) : 330 -3^.   (In Rom.)
     Taken from:  Chen. Abst. 60: 10770 J, 1971.

 Steichen, R.J.  (1976).   Modified solution  approach for the gas ehrooatographic
     determination  of residual  monomers by head-space analysis.   Anal.  Chem.
     18(9): 1398-1102.

 Suta, 6.E. (1979).  Assessment of Human  Exposures to Atmospheric Acrylonitrile.
     Prepared by Stanford Research Institute, Menlo Park, CA, under Contract No.
     68-02-2835.   U.S. Environmental Protection Agency, Office  of Air Duality
     Planning and Standards, Research Triangle Park, NC.

 Suta,  B.E.  (1982).   Letter  to J.E. Hadley,  Jr.  (LaRoe,  Vinn,  and  Moennan,
     Attorneys  at Lav, Washington, D.C.) transmitting enclosed report entitled:
     "Revised Assessment of Human Exposure  to Atmospheric.  Acrylonitrile Using
     Industry Supplied Emission  Estimates".  June 21, 1982.

 Sweger, D.M., and Travis, J.C. (1979).  An  application  of infrared  lasers to the
     selective  detection of trace organic gases.  Appl. Spectrosc. jg( 1) ;16-51.
 Szabo,  S., and  Selye,  H. (1971).   Adrenal apoplexy and  necrosis  produced by
     acrylonitrile.  Endokrinologie 5J( 3) : 105-08.

 Szabo,  S., and Selye,  H. (1972).   Effect of phenobarbital and steroids on the
     adrenal  apoplexy  produced  by acrylonitrile  in rats.    Endocrinol.  Exp.
     6(3): 111-16.

 Szabo,  S.; Reynolds, E.S.; Komanicky, P.; Moslen,  M.T.;  and Melby, J.C. ( 1976a) .
     Effect of chronic  acrylonitrile ingestion on rat adrenal.  Toxicol. Appl.
     Pharmacol.  J7_:133.   (Abstract.)

 Szabo,  S.; Reynolds,  E.S.;  and  Kovacs, K.  (1976b).   Animal model  of human
     disease.   Waterhouse-Friderichsen syndrome.  Animal model:  Acrylonitrile-
     induced adrenal apoplexy.  Amer. J.  Pathol.  82( 3) :653-56.

 Szabo,  S.; Bailey, K.A.;  Boor, P.J.; and Jaeger, R.J. ( 1977a) .   Acrylonitrile and
     tissue glutathione:  Differential effect of acute and  chronic interactions.
     Biochem.  Biophys. Res. Commun. 7_9_( 1) : 32-37.
                                          R-19

-------
 Szabo,  S.:  Boor,  P.J.; Oelwang, M.C.; and Bailey,  K.A.  (1977b).  Pathogenesis of
      acrylonitrile poisoning:  Role of glutathione and  coagulation changes.  In:
      International Congress  on Toxicology.   Abstracts.  Burford,  P.O., editor.
      Toronto, Ontario, Canada:  International Congress on Toxicology.

 Tanaka, M.; Nishinura, P.; and Shono, T. (1975). Pyrolysis-gas ohromatography of
      vinyl  chloride-methyl nethacrylate and  vinyl chloride-acrylonitrile copoly-
      mers.  Anal. Chim. Acta ]±( 1) : 119-21.
Taubinger,  R.P.  (1969).   Direct determination of free aerylonitrile in aqueous
     copolyaer latexes.  Analyst 9JK 1121) : 628-33.

Terent'ev,  A. P., and Obtenperanskaya ,  S.I. (1956).  Quantitative determination
      of acrylonitrile with sodium sulfite.   Zhur. Anal. Khim 11: 638-39.   (In
      Buss.)  Taken from:   Chem.  Abst. £l:11935h, 1957.

Thiess,  A.M., and Plelg,  I. (1978).  Analysis of chromosomes of workers exposed
      to aerylonitrile.  Arch. Toxiool. |U(2) :ll9-52.

Thiess,  A.M., R.  Frentzel-Beyme, R. Link, and H. Wild.  1980.
     Mortal itatsstudle Be1  ChemlefacharbeUern VerscMedener Produktlonsbetriebe
     Mit  Exposition Auch Gegenuber Acrylonitrile Zentralblatt Arbeltsmedizln
     30:259-267.

Thompson,  K.R.   (1971).    Process  BBSS  ipeetrometry .    DuPont  Innovation
     £(3): 18 -20.

Troxell, T.C. (1980).  Letter  from T.C. Troxell,  Division  of Pood and Color
     Additives,  Food and Drug Administration, Department of Health, Education,
     and Welfare, Washington,  DC, to J. Santodonato,  Syracuse Research  Corpora-
     tion, Syracuse, NY,  dated August 20, 1980.

Tsuchiya, T., and Sumi,  K.  (1977).   Thermal decomposition products  of poly-
     acrylonitrile.  J. Appl.  Polymer Sci.  21(1) : 975-80.

Uhde, W.J., and Koehler,  0. (1967).  Testing of plastic articles.  Determination
     of monomerlc  compounds in  the  volatile fraction of etyrene polymers.  Z.
     Lebensm.-Unters. Forsch 135(3) : 135-10. (InGer.)  Taken from:  Chem. Abst.
     67:117599b,  1967.

U.S. EPA (Environmental Protection Agency)  (1976b).  EPA Method 621,  Purgeables,
     11 Fed. Feg.  69532-69593, Dec.  3, 1979.

U.S. EPA (U.S. Environmental Protection Agency) ( 1978a) .  Data submitted to U.S.
     EPA, Washington, DC,  under  Contract No. 68-01-1616 by E.G. and G. Bionomics,
     Wareham, MA.  Contract title:   In-Depth Studies on Health and Environmental
     Impacts of Selected Water Pollutants.

U.S. EPA (Environmental Protection Agency) (1979a).  EPA Method 603, Acrolein and
     Acrylonitrile, 11 Fed. Reg. 69179-69181, Dec. 3, 1979.

U.S.  EPA  (U.S.   Environmental   Protection  Agency)  (1980).   OAQPS,  Emission
     Standards  and  Engineering Division,  "Organic  Chemical Manufacturing  -
     Vol. 10:        Selected    Processes,    Chapter 2:       Acrylonitrile, "
     EPA-l50/3-80-028e, Research Triangle Park, NC, December 1980.

                                       R-20

-------
U.S.  EPA (U.S. Environmental Protection Agency)  (1980).  Ambient Water Quality
     Criteria Document:  Acrylonitrile.  U.S. EPA.  EPA U10/5-80-017, Oct.,  1980.
     Available from:  National Technical Information Service, Springfield, VA.

U.S. EPA (Environmental Protection Agency)  (1980c).   EPA  Method 1621.  Volatile
     Organic  Compounds  by  Purge and  Trap Isotope  Dilution  GC-MS,  U.S. EPA
     Effluent Guidelines Division, June, 1980.

U.S. EPA (U.S. Environmental Protection Agency)  (1981).  The Carcinogen Assess-
     ment Group's Carcinogen Assessment of Acrylonitrile.  Available from:  U.S.
     EPA, Office of Research and Development,  Washington, DC.

U.S. Environmental Protection Agency.  Science Advisory Board Environmental
    Health Committee.  Transcripts from August 2. 1982.

U.S. Health, Education, and Welfare, National  Clearinghouse  for Smoking and
    Health.  1973.  Adult use of tobacco, 1970.   U.S. HEW 1973.

 Ustinovskaya,   I.A.;  Klzllova,  L.I.;   Oavrllina,  L.Ta.;  Malakhov,  V.V.;  and
      Yashin,  Ya.I. (1977)*  Gas-chromatographlo analysis of products  from  the
      oxidative amaonolysis of propylene.  Izv.  Sib. Otd. Akad. Hauk SSSR, Ser.
      Khim.  Nauk 2:115-18.  (In Fuss.)  Taken from:  Chem. Abst. 87_:77980y,  1977.


 Ved Brat, S. and G.M. Williams.  1982.  Hepatocyte-mediated production of sister
     chromatid exchange 1n cocultured cells by acrylonitrile:   evidence for
     extra-cellular transport of a stable reactive Intermediate.  Unpublished.

 Veith, G.D.; DeFoe, D.L.;  and Bergstedt, B.V. (1979). Measuring and estimating
      the bioconcentration  factor of chemicals In fish.  J.  Pish. Res. Board Can.
      26:1010-18.

 Venitt,  S.  (1978).    Mutagenlc  studies  with  acrylonitrile.   Mutat. Res.
      5_7_(1): 107-09.

 Venitt, S.; Bushell,  C.T.; and Osborne, M. (1977).  Mutagenicity of acrylonitrile
      (cyanoethylene)  In Eseherichia coli.  Mutat. Res. 45_(2) :283-88.

 Werner, J.B., and Carter,  J.T. (1981).  Mortality of United Kindom acrylonitrile
      polymersation workers.  British Journal  of Industrial  Medicine 38:217-253.

 Wilcox, W.S.,  and Goldstein,  J.H. (1951).   The microwave  spectrum  of vinyl
      cyanide.  J. Chem. Phys.  22(3):516-18.

 Wilson, R.H. (1911).  Health hazards encountered in the manufacture of synthetic
      rubber. J. Amer. Med. Assoc. 121:701-03.

 Wilson, R.H.;  Hough,  G.V.;  and McCormlck, W.E. (1918).  Medical problems encoun-
      tered in the manufacture of American-made rubber.  Ind. Med. 17(6):199-207.

 Wolff, S.  1977.  Sister  chromatid exchange.   Annu. Rev. Genet. 11:183-201.
                                        R-21

-------
Wright, P.L.  (1977).   Studies on  the metabolism of acrylonitrile.   Testimony
     before the FDA.  (Cited in Miller and Villaume,  1978.)

Wronski,  M.,  and  Zbigniew,  S.  (1971).    Thlotnercurinetric  determination  of
     acrylonltrile In water.  Chem. Anal. 1£(3):633-38.   (In Pol.)   Taken from:
     Chem. Abst. 8ll75988c, 197H.

Young, J.D.;  Slauter,  R.V.;  and Karbovstcl, R.J. (1977).  The Phannacoklnetic and
     Metabolic Profile of UlC-Acrylonitrile  Given to Rats by Three Routes.  Pre-
     pared by  the  Toxicology Research  Laboratory,  Health and  Environmental
     Research, Dow  Chemical USA, Midland,  MI,  for  the  Chemical  Manufacturers
     Association, Washington, DC.  Available from:  CMA, Washington, DC.

Zack, J.A. (I960).  Written communication from J.A. Zack, Monsanto,  St.  Louis,
     MO, to W.R. Gaffey,  Monsanto,  St. Louis, MO, dated June 26, 1980, regarding
     Acrylonitrile Epidemiology Study:  Latent Analysis.

Zitting,  A.;   Tenhanen,  R.;   and  Savolainen,   H.   (1981).      Effects   of
     intraperitoneally injected acrylonltrile on liver,  kidney and brain.  Acta
     Pharmacol. Toxicol.  Jr9:Ul2-ll15.  Taken from :  Chem. Abst. 56:117119u.
                                         R-22

-------