United States
             Environmental Protection
             Agency
            Office of Health and
            Environmental Assessment
            Washington DC 20460
EPA/600/8-84/009F
June 1985
Final Report
             Research and Development
&EPA
Health Assessment
Document for
Ethylene Oxide
 Final
 Report

-------
                            EPA-600/8-84-009F
Health Assessment Document
        for Ethylene  Oxide
        U S. ENVIRONMENTAL PROTECTION AGENCY
          Office of Research and Development
        Office of Health and Environmental Assessment
        Environmental Criteria and Assessment Office
           Research Triangle Park, NC 2771 1

-------
                                  DISCLAIMER




     This document has been reviewed in accordance with  the  U.S.  Environmental



Protection Agency's  peer and administrative  review  policies and approved  for




presentation and  publication.   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 ethylene oxide 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.
                                      iii

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

AUTHORS;

     The  EPA's  Office  of  Health  and  Environmental  Assessment  (OHEA)  is

responsible for  the  preparation  of the  health  assessment document.   The OHEA

Environmental   Criteria   and  Assessment   Office   (ECAO-RTP)   had   overall

responsibility for  coordination  and direction  of  the document  (Dr.  Robert M.

Bruce,  Project  Manager).     The   chapters  addressing  physical  and  chemical

properties,  sampling and  analysis,  air quality,  and biological  effects  in

animals  and  man were  originally written  and  revised  by  Syracuse  Research

Corporation   with   the   exception   of  chapters    or   sections   addressing

mutagenicity,  teratogenicity  and  reproductive effects  and  carcinogenicity.

The air  quality  chapters  (5,  6,  7) were reviewed  by Radian Corporation under

contract  to the  Office  of Air  Quality Planning  and Standards  and  recommen-

dations proposed.

     The  principal  authors of the chapters or  sections  prepared  by  Syracuse

Research Corporation are:
        D. Anthony Gray, Ph.D.
        Life and Environmental Sciences Division
        Syracuse Research Corporation
        Syracuse, NY

        Bruce Harris, Ph.D
        Life and Environmental Sciences Division
        Syracuse Research Corporation
        Syracuse, NY

        Stephen Bosch
        Life and Environmental Sciences Division
        Syracuse Research Corporation
        Syracuse, NY
                                       IV

-------
        Joseph Santodonato,  Ph.D,  CIH
        Life and Environmental Sciences Division
        Syracuse Research Corporation
        Syracuse, NY


     The   OHEA   Carcinogen   Assessment   Group   (CAG)   was   responsible  for

preparation of  the  sections  on carcinogenicity.   Participating members of the

CAG are  listed  below  (principal  authors  of present  carcinogenicity materials

are designated by an asterisk(*).


        Roy E. Albert,  M.D.  (Chairman)
        Elizabeth L. Anderson, Ph.D.
        Larry D. Anderson,  Ph.D.
        Steven Bayard,  Ph.D.*
        David L. Bayliss,  M.S.
        Robert P. Beliles,  Ph.D.
        Chao W.  Chen,  Ph.D.
        Margaret M.L.  Chu,  Ph.D.
        Herman J. Gibb, B.S., M.P.H.*
        Bernard  H. Haberman,  D.V.M.,  M.S.
        Charalingayya B. Hiremath,  Ph.D.
        Robert E. McGaughy,  Ph.D.
        Jean C.  Parker, Ph.D
        Dharrn 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 rnutagenicity,  teratogenicity  and  reproductive

effects.   Participating members of REAG are listed below  (principal authors of

present sections  are   indicated by an  asterisk).   The  Environmental  Mutagen

Information Center (EMIC) ,  in  Oak  Ridge,  TN,  identified  literature  bearing on

the nutagenicity of ethylene  oxide.


        Eric Clegg,  Ph.D.*
        John R.  Fowle,  III, Ph.D.*
        Ernest R. Jackson,  M.S.
        Casey Jason,  M.D.
        David Jacobson-Kram,  Ph.D.
        K.S. Lavappa, Ph.D.
        Sheila L. Rosenthal,  Ph.D.

-------
        Carol N. Sakai,  Ph.D.*
        Carmella Tellone,  B.S.*
        Vicki-Vaughan Dellarco, Ph.D.
        Peter E. Voytek, Ph.D (Director)
REVIEWERS

     The following  individuals  provided peer review  of this draft  or  earlier

drafts of this document:

U.S. Environmental Protection Agency

        Karen Blanchard
        Office of Air, Noise and Radiation
        Office of Air Quality Planning and Standards
        Research Triangle Park,  NC

        Robert M. Bruce, Ph.D.
        Office of Health and Environmental Assessment
        Environmental Criteria and Assessment Office
        Research Triangle Park,  NC

        James W. Falco, Ph.D.
        Office of Health and Environmental Assessment
        Exposure Assessment Group
        Washington, D.C.

        Lester D. Grant, Ph.D.
        Office of Health and Environmental Assessment
        Environmental Criteria and Assessment Office
        Research Triangle Park,  NC

        Joseph Padgett
        Office of Air, Noise and Radiation
        Office of Air Quality Planning and Standards
        Research Triangle Park,  NC

        William E. Pepelko
        Office of Health and Environmental Assessment
        Environmental Criteria and Assessment Office
        Cincinnati, OH

        Jerry F. Stara, D.V.M.
        Office of Health and Environmental Assessment
        Environmental Criteria and Assessment Office
        Cincinnati, OH
                                       VI

-------
Consultants and Reviewers

        I.W.F. Davidson,  Ph.D.
        Bowman Gray School of Medicine
        Wake Forest University
        Winston Salem, NC

        Larry Fishbein, Ph.D.
        National Center for Toxicological Research
        Jefferson, AR

        Richard N. Hill,  M.D., Ph.D.
        Office of Toxic Substances
        U.S. Environmental Protection Agency
        Washington, DC

        Derek Hodgson, Ph.D.
        University of North Carolina
        Chapel Hill, NC

        George R. Hoffman, Ph.D.
        Holy Cross College
        Worcester, MA

        Rudolph J. Jaeger, Ph.D.
        Consultant Toxicologist
        7 Bogert Place
        Westwood, NJ

        Marshall Johnson, Ph.D.
        Thomas Jefferson Medical College
        Philadelphia, PA

        Edmond J. LaVoie, Ph.D.
        American Health Foundation
        Valhalla, NY

        P.O. Lotilaker, Ph.D.
        Fels Research Institute
        Temple University Medical Center
        Philadelphia, PA

        Sam Shibko, Ph.D.
        Health and Human Services
        Division of Toxicology
        Washington, DC

        Charles M. Sparacino, Ph.D.
        Research Triangle Institute
        Research Triangle Park, NC
                                      Vll

-------
       Danial  S. Straus, Ph.D.
       University of  California
       Riverside, CA

       Robert  Tardiff, Ph.D.
        1423  Trapline  Court
       Vienna,  VA

       Norman  M. Trieff, Ph.D.
       University of  Texas  Medical Branch
       Department of  Pathology,  UTMB
       Galveston, TX

       Benjamin Van Duuren, Ph.D.
       New York University  Medical Center
       550 First Avenue
       New York, NY

       Janes R. Withey, Ph.D.
       Department of  National  Health and Welfare
       Tunney's Pasture
        Ottawa,  Ontario
       Canada,  KIA  01Z


                            SCIENCE ADVISORY BOARD
                        ENVIRONMENTAL HEALTH COMMITTEE

     The  content  of  this  health assessment  document  on  ethylene  oxide  was

independently  peer-reviewed  in  public  session  by  the Environmental  Health

Committee of the Environmental  Protection Agency's Science Advisory Board.


                    CHAIRMAN, ENVIRONMENTAL  HEALTH COMMITTEE

     Dr.  Herschel E.  Griffin,  Professor of Epidemiology,  Graduate  School of
     Public Health,  6505  Alvarado Road, San Diego State University, San Diego,
     California  92182-0405


                  EXECUTIVE  SECRETARY.  SCIENCE ADVISORY  BOARD

     Dr.   Daniel  Byrd  III,  Executive  Secretary,   Science  Advisory  Board,
     A-101 F, U.S. Environmental  Protection  Agency, Washington, D. C.  20460
                                     Vlll

-------
                               MEMBERS

Dr.  Seymour  Abrahamson,  Professor of Zoology  and  Genetics,  Department of
Zoology, University of Wisconsin, Madison, Wisconsin  53706

Dr.  Morton Corn,  Professor and Director,  Division of Environmental Health
Engineering,  School  of  Hygiene  and  Public  Health,  The  Johns  Hopkins
University, 615 N. Wolfe Street, Baltimore, Maryland  21205

Dr.  John  Doull,  Professor  of  Pharmacology and Toxicology,  University of
Kansas Medical Center, Kansas City, Kansas  66103

Dr.  Jack  D.  Hackney,  Chief, Environmental Health  Laboratories,  Professor
of  Medicine,   Rancho  Los  Amigos  Hospital Campus  of  the  University  of
Southern California, 7601 Imperial Highway, Downey, California  90242

Dr.  Marvin Kuschner,  Dean, School  of Medicine,  Health Science  Center,
Level 4, State University of New York, Stony Brook, New York  11794

Dr.  Daniel Menzel,  Director  and Professor,  Pharmacology and  Medicine,
Director,  Cancer Toxicology  and  Chemical Carcinogenesis  Program,  Duke
University Medical Center, Durham, North Carolina  27710.

Dr.  Steven M.   Rappaport,  Associate  Professor  of  Industrial  Hygiene,
School  of  Public  Health,  Department  of Biomedical  and  Environmental
Health Sciences,  University of California, Berkeley

Dr.  Michael J.  Symons,  Professor, Department  of Biostatistics,  School of
Public Health, University  of  North Carolina,  Chapel Hill, North Carolina
27711

Dr.  D.  Warner North,  Principal,  Decision Focus  Inc.,  Los Altos  Office
Center,  Suite 200, 4984 El Camino Real,  Los Altos,  California  94022

Dr.  Bernard   Weiss,  Professor,  Division  of  Toxicology,  P.O.  Box  RBB,
University of Rochester,  School of Medicine,  Rochester,  New York  14642

Dr.  Ronald  Wyzga,   Electric   Power  Research  Institute,   3412  Hillview
Avenue,  P.O.  Box  1041,  Palo Alto, California   94303

Dr.  Edward F. Ferrand, Assistant  Commissioner  for  Science  and  Technology,
New York City Department of Environmental  Protection, 51 Astor  Place,  New
York, New York  10003

Dr.  Ronald D.  Hood,  Professor, Developmental  Biology Section,  Department
of Biology, The University  of Alabama, and Principal Associate,  R.D.  Hood
and  Associates,   Consulting Toxicologists,  P.O.  Box  1927,   University,
Alabama   35486
                                 IX

-------
                         SPECIAL CONSULTANTS

Dr. Robert R. Maronpot, Head, Experimental  Pathology,  National  Toxicology
Program,  NIEHS,  P.O.  Box  12233, Research  Triangle Park, North  Carolina
27709

Dr. Walderico M.  Generoso,  Senior Scientist, Biology  Division, Oak  Ridge
National Laboratory, Oak Ridge,  Tennessee  37830

-------
                              TABLE OF CONTENTS
LIST OF TABLES	       xv

LIST OF FIGURES	      xix

LIST OF ABBREVIATIONS	       xx

1 .   SUMMARY AND CONCLUSIONS	      1-1

2.   INTRODUCTION	      2-1

3.   PHYSICAL AND CHEMICAL  PROPERTIES	      3-1

     3.1  SYNONYMS AND  CAS  NUMBER	      3-1
     3.2  STRUCTURAL AND MOLECULAR  FORMULAS	      3-1
     3.3  TORSIONAL ANGLES  AND  BOND DISTANCES	      3-1
     3.4  PHYSICAL PROPERTIES OF  PURE ETHYLENE OXIDE	      3-2

          3.4.1    Description	      3-2
          3.4.2    Molecular Weight	      3-2
          3.4.3    Melting Point	      3-2
          3 .4.4    Boiling Point	      3-2
          3.4.5    Boiling Point Change with Pressure Change	      3-2
          3.4.6    Density	      3-2
          3.4.7    Coefficient of  Cubical Expansion	      3-2
          3.4.8    Refractive Index	      3-3
          3.4.9    Vapor Pressure	      3-3
          3.4.10  Aqueous Solubility	      3-3
          3-4.11  Freezing  Point of Aqueous Solutions	      3-4
          3.4.12  Boiling Point of  Aqueous Solutions	      3-4
          3.4.13  Flash Point	      3-4
          3.4.14  Flash Point of Aqueous Solutions	      3-5
          3.4.15  Explosive Limits  in Air, Voluras %	      3-5
          3.4.16  Heat  of Combustion at 25°C	      3-5
          3.4.17  Log Octanol/Water Partition Coefficient	      3-5
          3.4.18  Ultraviolet Spectroscopic Data	      3-5

     3.5  PHYSICAL  PROPERTIES AND DESCRIPTION OF COMMERCIAL
           ETHYLENE OXIDE	      3-5
     3.6  CHEMICAL  PROPERTIES	      3-7

          3.6.1    Reduction	      3-7
          3.6.2    Clathrate  Formation	      3-7
                                      XI

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

                                                                          Page

          3.6.3   Polymerization	      3-7
          3.6.4   Other Reactions	      3-8
          3.6.5   Hydrolysis  and Related  Reactions	      3-8
          3.6.6   Free Radical Reactions	     3-17

4.   SAMPLING AND ANALYTICAL  METHODS	      4-1

     4 .1  SAMPLING	      4-1
     4 .2  ANALYSIS	      4-6

5.   SOURCES IN THE ENVIRONMENT	       5-1

     5 .1  PRODUCTION	       5-1

          5.1.1   Quantities  Produced	       5-1
          5.1.2   Producers,  Production Sites, and  Distribution	       5-1
          5.1.3   Production  Methods and  Processes	       5-1

     5.2  USES OF ETHYLENE OXIDE	      5-13

          5.2.1   Ethylene Glycol.	      5-13
          5.2.2   Nonionic Surface-Active Agents	      5-15
          5.2.3   Di-,  Tri-,  and Polyethylene Glycols	      5-15
          5.2.4   Glycol Ethers	      5-16
          5.2.5   Ethano la mines	      5-16
          5.2.6   Miscellaneous Applications..,	      5-16
          5.2.7   Discontinued Uses  of Epoxides	      5-18
          5.2.8   Projected or Proposed Uses	      5-18
          5.2.9   Alternatives to Uses for Ethylene Oxide.	      5-18

     5.3  POTENTIAL FOR ENVIRONMENTAL  CONTAMINATION	      5-19

          5.3.1   Air Emissions from Production.	<.	      5-19
          5.3.2   Handling, Transport, and Storage	      5-22
          5.3.3   Potential Environmental Formation	      5-23

     5 .4  SUMMARY	      5-25

6.   ENVIRONMENTAL FATE, TRANSPORT,  AND DISTRIBUTION	       6-1

     6 .1  INTRODUCTION	       6-1
     6.2  ETHYLENE OXIDE FATE IN WATER	       6-1
     6.3  ETHYLENE OXIDE FATE IN SOIL	       6-3
                                     xn

-------
                          TABLE OF CONTENTS  (cont. )
     6.4  ETHYLENE OXIDE FATE IN  THE ATMOSPHERE	       6-4
     6.5  DEGRADATION IN COMMODITIES AND MANUFACTURED PRODUCTS	       6-6
     6.6  BIO ACCUMULATION IN AQUATIC ORGANISMS	       6-9
     6.7  SUMMARY	       6-9

7.   ENVIRONMENTAL LEVELS AND EXPOSURE	       7-1

     7 .1  INTRODUCTION	       7-1
     7.2  ENVIRONMENTAL  LEVELS	       7-1
     7.3  EXPOSURE	       7-4
     7 .4  SUMMARY	       7-4

8.   ECOLOGICAL EFFECTS	       8-1

     8.1  MICROORGANISMS AND INSECTS	       8-1
     8 .2  PLANTS	       8-2
     8 .3  AQUATIC ORGANISMS	       8-2

9.   BIOLOGICAL EFFECTS  IN ANIMALS AND MAN	       9-1

     9 .1  PHARMACOKINETICS	       9-1

          9.1.1   Absorption	       9-1
          9.1.2   Distribution	       9-1
          9.1.3   Metabolism	       9-3
          9.1.4   Elimination	       9-4

     9.2  ACUTE,  SUBCHRONIC, AND CHRONIC TOXICITY	       9-6

          9.2.1    Effects  in Humns	       9-6
          9.2.2   Effects  in Animals	     9-14
          9.2.3   Summary  of Toxicity	     9-24

     9.3   TERATOGENICITY AND REPRODUCTIVE TOXICITY	     9-26

          9.3.1.   Teratogenic Effects	     9-26
          9.3.2.   Reproductive Effects	     9-31
          9.3.3.   Testicular Effects	     9-33
          9.3.4.   Adverse  Reproductive Outcome  in Humans	     9-35
          9.3.5.   Summary  of Teratogenic ity and  Reproductive
                  Tox ic ity	     9-38
                                    Xlll

-------
                          TABLE OF CONTENTS (cont. )
     9.4  MUTAGENICITY	      9-H6

          9.4.1.   Gene  Mutation Studies	      9-i»9
          9.4.2.   ChromosonB Aberration Studies	      9-64
          9.4.3.   Chromosome Mutations in Human Populations	      9-77
          9.4.4.   Other Studies Indicative of Genetic Damage	      9-81
          9.4.5.   Summary and Conclusions of the Mutagenicity of
                  Ethylene Oxide	      9-92

     9 .5  CARCINOGENICITY	      9-94

          9.5.1    Animal Studies	      9-94
          9.5.2    Epidemic logic Studies	    9-118
          9.5.3    Quantitative Estimation	    9-136
          9.5.4    Summary	    9-162
          9.5.5    Conclusions	    9-168

     Appendix  9A:  Comparison of Results by Various Extrapolation
                  Models	       A1
     Appendix  9B:  International Agency for Research on Cancer
                  Classification System for the Evaluation  of the
                  Carcinogenic Risk of Chemicals to Humans	       B1

10.   REFERENCES	      10-1
                                     xiv

-------
LIST OF TABLES
Table
3-1
3-2
3-3
3-4

4-1

5-1
5-2

5-3

5-4

5-5
5-6

8-1
9-1
9-2
9-3

9-4

9-5

9-6




Hydrolysis Kinetics of Ethylene Oxide 	
Specific Rates of Reaction of Anions and Lewis Bases with
Ethylene Oxide 	
Breakthrough and Safe Sampling Volumes for Propylene Oxide

Ethylene Oxide Production 	
Ethylene Oxide Producers, Plant Sites, Capacities, Processes,
and Tec hnology 	
Ranges of Reaction Systems Variables in the Direct
Air -Oxidation of Ethylene Oxide 	
Ranges of Reaction Systems Variables in the Direct
Oxygen-Oxidation of Ethylene Oxide 	

Typical Vent Gas Composition for Both Air- and Oxygen-Based
Ethylene Oxide Plants 	
Acute Aquatic Toxicity of Ethylene Oxide 	
Acute Toxicity of Ethylene Oxide 	
Subchronic Toxicity of Ethylene Oxide 	
Summary of Studies on Teratogenicity and Reproductive
Tox ic ity 	
Summary of Mutagenicity Testing of Ethylene Oxide: Gene
Mutations in Bacteria 	
Summary of Mutagenicity Testing of Ethylene Oxide: Gene
Mutations Tests in Lower Plants. (Yeast) 	
Summary of Mutagenicity Testing of Ethylene Oxide: Mutation
Te st s in Higher Plants 	
Page
3-6
3-9
3-12

3-13

4-2
5-2

5-3

5-9

5-11
5-14

5-21
8-3
9-2
9-17

9-39

9-47

9-53

9-55
      XV

-------
                            LIST  OF TABLES (cont. )
Table
9-7

9-8

9-9

9-10

9-11

9-12

9-13

9-14

9-15

9-16

9-17

9-18

9-19



Summary of Mutagenicity Testing of Ethylene Oxide: Gene

Summary of Mutagenicity Testing of Ethylene Oxide: Mammalian
Cells in Culture 	
Summary of Mutagenicity Testing of Ethylene Oxide: Dominant
Lethal Tests 	
Summary of Mutagenicity Testing of Ethylene Oxide: Heritable

Summary of Mutagenicity Testing of Ethylene Oxide: Chromosome
Aberra tion Te st s 	 	 	
Summary of Mutagenicity Testing of Ethylene Oxide:
Micronucleus Tests 	
Summary of Mutagenicity Testing of Ethylene Oxide: Chromosome

Summary of Mutagenicity Testing of Ethylene Oxide: SCE
Formation in Human Populations 	 	
Summary of Mutagenicity Testing of Ethylene Oxide: SCE
Formation in Experimental Studies 	
Summary of Mutagenicity Testing of Ethylene Oxide: Unscheduled
DNA Synth esis 	
Design Summary for Carcinogenicity Testing of Ethylene Oxide

Tumor Induction by Intragastric Administration of Ethylene
Oxide in Female Sprague-Dawley Rats 	
Cumulative Percentages of Male Fischer 344 Rats that Died
or were Sacrificed in a Moribund Condition After Exposure
to Ethylene Oxide Vapor 	
Page

9-58

9-62

9-65

9-67

9-68

9-69

9-70

9-82

9-84

9-85

9-97

9-97


9-101
9-20    Cumulative Percentages of Female Fischer 344 Rats that Died
        or were Sacrificed  in a Moribund Condition After Exposure
        to Ethylene Oxide Vapor	    9-102
                                     xvi

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

Table

9-21    Cumulative Percentages of Male  Fischer 344  Rats  that were
        Alive at the Beginning of Month 17, but Died or  ware
        Sacrificed in a Moribund Condition  After Subsequent
        Exposure to Ethylene Oxide Vapor	    9-103

9-22    Cumulative Percentages of Female Fischer 344 Rats that were
        Alive at the Beginning of Month 17, but Died or  were
        Sacrificed in a Moribund Condition  After Subsequent
        Exposure to Ethylene Oxide Vapor	    9-104

9-23    Summary of Selected Tumor Incidence Comparisons  for Male
        and Female Fischer 344 Rats Exposed to Ethylene  Oxide  for
        Two Years	    9-106

9-24    Ethylene Oxide  2-Year Vapor  Inhalation Study:   24-Month Final
        Sacrificed Frequency of Expo sure-Re la ted Neoplasms
        for 110- to 116-Week-Old Fischer 344 Rats	    9-107

9-25    Ethylene Oxide 2-Year Vapor Inhalation Study:  Frequency of
        Exposure Related Neoplasms at 24-Month Final Sacrificed and
        in Fischer 344 Rats Dying Spontaneously or  Euthanized
        Wh en Mo rib und	    9-108

9-26    Ethylene Oxide 2-Year Vapor Inhalation Study:  Frequency of
        Brain Neoplasms in Fischer 344  Rats	    9-112

9-27    Ethylene Oxide 2-Year Vapor Inhalation Study:  Frequency of
        Primary Brain Neoplasms Types in Fischer 344 Rats	    9-113

9-28    Leukemia Incidence in  Male  Fischer 344 Rats Exposed
        to Ethylene Oxide for 2 years	    9-115

9-29    Incidence of Neoplastic Lesions in Male Fischer  344
        Rats Exposed to Ethylene Oxide  for 2 years	    9-117

9-30    Observed and Expected  Number of Deceased  Among 153 Women
        and 50 Men with Continuous  or Intermittent  Exposure to
        Ethylene Oxide	    9-127

9-31    Comparison of Observed Numbers  of Cancer Deaths  in
        BASF-Aktiengesellschaft,  Ludwigshafen  Plants 1928-80
        for Persons Having 10  Years of  Observation  Following
        Exposure to Alkylene Oxide  With  That Expected Based
        on Mortality Statistics for Rhinehessia-Palatinate
        1970-75, Ludwigshafen 1970-75,  and  the Federal
        Republic of Germany 1971-74, By ICD Code and Cause
        of Death	    9-130

                                     xvii

-------
                            LIST OF TABLES (cont.)
Table
9-32    Related Risks of Death  From Cancer  in  the Alkylene Oxide
        Cohort as Compared With the Styrene  Cohort, By Age	    9-132

9-33    Bushy Run Ethylene Oxide Inhalation  Study in Fischer 344 Rats.
        Incidence of Peritoneal Mssothelioma and Brain Glioma in Males,
        and MDnonuclear Cell Leukemia  and Brain Glioma in Females by
        Dose Among Survivors to First  Tumor.   Maximum Liklihood
        Estimates of Linear Term and 95$ Upper-Limit q-|*	   9-150

9-34    NIOSH Ethylene Oxide Inhalation Study  In Nfele Fischer 344 Rats.
        Incidence of Peritoneal Mesothelioma and Brain Glioma by Dose
        Among Total Examined.   Estimates of 95% Upper-Limit Risk Based
        on Human Equivalent Dose	   9-154

9-35    Leukemia (ICD 204-207)  Incidence and Nbrtality:  Ethylene Oxide
        Epidemiology.  Included are Relative Risks, 95%  Confidence
        Limits, Nominal Exposure Estimates,  and 95/8 Confidence Limits
        on Unit Risk	   9-156

9-36    Relative Carcinogenic Potencies Among  54 Chemicals Evaluated
        by the Carcinogen Assessment Group as  Suspect Human
        Carcinogens	    9-163
                                     xviii

-------
                                LIST OF FIGURES


Figure                                                                    Page

5-1     Schematic for Air-Based Ethylene Oxidation	      5-6

9-1     Mutagenic Response of Salmonella typhimurium Strain
          TA1535 Exposed to Ethylene Oxide	     9-50

9-2     Mutagenic Response of CHO  Cells to  EtO	     9-63

9-3     Percentages of Male and Female  Fischer 344 Rats with
          Histologically Confirmed Mononuclear Cell Leukemia
          at 24-Month Sacrifice	    9-111

9-4     Histogram Representing the Frequency  Distribution of the
          Potency Indices of 54 Suspect Carcinogens Evaluated
          by the Carcinogen Assessment  Group	    9-161
                                     xix

-------
                             LIST OF ABBREVIATIONS



BCF                              Bio concentration  factor



BOD                              Biochemical oxygen demand



CNS                              Central nervous system



DMSO                             Dimethyl su If oxide



DNA                              Deoxyribonucleic  acid



GC                               Gas chromatography



GSH                              Gluthathione



LC,.,.                             Concentration lethal to 50%  of recipients
  50


LDp.,.                             Dose lethal to 50% of  recipients



MS                               Mass spec tome try



ppb                              Parts per billion



ppm                              Parts per million



PVC                              Poly vinyl chloride



SCE                              Sister chromatid  exchange



TWA                              Time-weighted average



v/v                              VoluriE per volume
                                      xx

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









     Ethylene oxide is a  colorless,  flammable gas at ambient  temperature.   It




is soluble  in  water (195 m£  vapor  in  1  mSL  water  at  20°C)  and is  a  highly




reactive chemical.  The  chemical  reaction of  ethylene  oxide that  is  environ-




mentally significant is the  hydrolytic reaction in aqueous media.




     The most  suitable chemical method available for the  analysis  of  ethylene




oxide in the atmosphere is its collection on a sorbent cartridge and subsequent




analysis by  gas  chroraatography  with  flame  ionization  or mass  spectrometric




detector.  Biological  samples are  analyzed  by  the  purge  and  trap  method,




followed by  gas  chromatography  with  flame  ionization  or mass  spectrometric




detection.




     Ethylene oxide  is produced  almost  exclusively  by  direct  oxidation  of




ethylene.  Its  1981  production volume  was 4937  million pounds.  The  largest




single use  of  ethylene oxide  is  in the  synthesis  of ethylene glycol.   Small




amounts of  ethylene  oxide are used  as  a  sterilant  and in the manufacture  of




pesticides, Pharmaceuticals, and medicinal devices.




     The major  emission  sources  from  production  facilities  of ethylene  oxide




are the  main process  vents  and  purge  gas  vents.   Total  air emission  from




production in  1978 has  been estimated to  be  2 million  pounds.  Although  only




small amounts of ethylene oxide are used for consumer products,  this represents




a considerable potential for human exposure.




     In aquatic media, ethylene oxide  will degrade by hydrolysis with  a  half-




life of =  12-14 days.  Evaporation from aquatic media will also be a significant




loss process.  There  is  no   conclusive  evidence that microbial degradation  is




significant in aquatic media.  The fate of ethylene oxide in soil will probably




                                      1-1

-------
be similar to that in water (its fate in the atmosphere is not obvious from




the available data).   Available rate constants for its reaction with hydroxyl




radicals and oxygen atom (^P) and its reaction in smog chambers predict that




ethylene oxide will persist in the atmosphere.  No reports measuring ambient




levels of ethylene oxide have been found.  Only one report exists for its




detection in ambient aquatic media.  Several studies have detected the presence




of ethylene oxide in commodities and commercial goods including food, medical




supplies and drugs.




     The pharmacokinetics of ethylene oxide have not been studied extensively.




Only one study was found about the absorption of this chemical; it concerned




the inhalation exposure of rats.  The toxicity data suggest that absorption




occurs via the respiratory and gastrointestinal tracts.  During inhalation




exposure, the highest concentration of ethylene oxide was associated with the




protein fraction of the lungs, while ethylene oxide that reaches the systemic




circulation is distributed widely to various tissues (liver, kidney, lung,




testes, brain, spleen and intestinal mucosa).  Ethylene oxide is eliminated




primarily by the kidneys with the metabolite, ethylene glycol, as well as




glutathione conjugates identified in the urine.  Ethylene oxide also reacts




with cellular macromolecules, and reaction with DNA results in small quantities




of 7-hydroxyethylguanine in the urine.  The half-life of ethylene oxide has




been estimated to be between  =  10 and 30 minutes, indicating rapid  removal of




absorbed compound.  Macromolecular-bound products and metabolites such as




ethylene glycol are removed more slowly.




     The effects of acute exposure of humans  to ethylene oxide have  been




described in case reports and,  to a more limited extent, in control  studies.




Case reports have indicated  that headaches, nausea, vomiting, dyspnea and




                                      1-2

-------
respiratory irritation  occur from  exposure to  the  vapors  of  ethylene oxide.




Dermal contact  with  the  liquid,   aqueous   solutions   or  clothing  containing




absorbed ethylene  oxide  results  in  skin  burns  and  possibly  sensitization.




Control studies in humans  indicate  that  exposure to  2200 pptn of ethylene oxide




was slightly  irritating,  while  exposure  to  22,000  ppm  adversely  affected




membranes in  the  nose.   Studies of dermal  contact indicate  that 50% solutions




are optimum for producing  chemical  burns.   Some effects may be delayed several




hours subsequent to exposure.




     In acute  studies in  laboratory  animals,  the LC5Q values  for  inhalation




exposure (rats, mice  and dogs)  range from  835-5000  ppm for  a 4-hour exposure,




and the  oral  LD5Q  value  (rabbits,  guinea  pigs  and rats)  range  from 100-631




mg/kg body  weight.    Gross  symptoms  of  toxicity were  respiratory irritation,




salivation, nausea,   vomiting,   diarrhea,  convulsions,  and  death.   Pathologic




findings included  lung,   liver,  and  kidney  damage.   Ethylene  oxide  was  an




irritant in dermal  studies,  but  failed  to  result  in  sensitization  in guinea




pigs.  Acute  effects  in  experimental animals  appear  to  be similar  to  those




reported in case reports of human exposure to ethylene oxide.




     The subacute and chronic  effects of ethylene oxide in man are  not  well-




documented nor readily available  from clinical  case  reports.   Case  reports of




workers exposed  repeatedly  to  ethylene  oxide  indicate  that  neurotoxicity




occurred consistent  with  sensorimotor  neuropathy.   Some  signs of  neuropathy




appeared to persist  after  cessation  of  exposure.  In occupational  epidemiology




studies, the  non-neoplastic  results  were  reported  to  be  an  increase  in




lymphocyte count,  a decrease  in hemoglobin  value (levels  of  exposure  were not




reported) and an  increased incidence  of  deaths  resulting  from  diseases  of the




circulatory system  (exposure  level <50  mg/m-*);  however, differences  in  study




                                      1-3

-------
 design  may  account  for  these  apparent  inconsistencies.   From  the available




 data, few  conclusions  can be  drawn  regarding the  systemic toxic  effects  of




 chronic exposure  to  ethylene  oxide.




     In animal  studies, repeated exposures to high  concentrations of ethylene




 oxide,  1533-643 mg/m3,  resulted in the  death  of  rats and  mice.   At the lower




 exposure, guinea  pigs  and  monkeys showed  signs   of  neurotoxicity  and  growth




 depression.  Decreased  body weight gain was  observed in  rats  at exposures as




 low as  60 mg/m3,  6 hours/day,  5 days/week  for  2  years,  while  signs of neuro-




 toxicity, hunched posture  and reduced  locomotion  were observed at doses as low




 as 90 mg/m3,  5  hours/day  for 10 weeks.   In rats  and mice, the no effect level




 appeared to be  18 mg/m3.    Similar  signs of toxicity were  observed after oral




 exposure to ethylene  oxide,  with weight loss  observed at 100 mg/kg body weight




 and no  effects  observed  at <10 mg/kg  body weight (15-20  doses).   Studies  in




 animals support  the  observation  of   neurotoxicity   described  in  human  case




 reports.




     The ability of ethylene  oxide to cause teratogenic or adverse reproductive




 effects has been  examined in a number  of species (mouse,  rat,  rabbit,  monkey




 and human) by two routes of administration (inhalation and intravenous).  Rats,




 but not rabbits,  exposed  to 150 ppm ethylene oxide  administered  by inhalation




 displayed signs of maternal toxicity  and toxicity to the developing conceptus.




Ethylene oxide  (150  mg/kg) administered intravenously to mice  caused maternal




 toxicity and developmental toxicity.    Intravenous  administration  of  ethylene




oxide in rabbits  (9,  18 and 36 mg/kg)  produced embryotoxicity  associated with




maternal toxicity.  Ethylene chlorohydrin (ECH),  a reaction product of ethylene




oxide,  produced adverse effects on maternal and  fetal  well-being in mice  but




not in  rabbits  at  120  mg/kg administered  intravenously and produced  adverse




                                      1-4

-------
developmental effects without  significant  toxicity  when administered ECH at 60




mg/kg intravenously.  Ethylene  oxide  (100  ppm)  administered by inhalation in a




one-generation study caused severe adverse effects including a higher incidence




of infertility, longer  gestational periods,  a  decrease  in the number  of pups




born, and a decrease in the number of implantation  sites.   The same laboratory




observed lowered fetal weights, but not a substantial level of malformations in




response to  100 ppm  ethylene  oxide  administered  to  rats  by  inhalation  on




gestation days 6-15.  Testicular degeneration was observed in hamsters and rats




inhaling 204  to  357 ppm  ethylene  oxide.   More  recently,  adverse  effects  on




sperm concentration  and  motility  but not  morphology  in   Cynomologus  monkeys




exposed to  50 and  100 ppm  ethylene  oxide  by  inhalation  were reported.   An




epidemiologic study  of  nursing personnel  exposed to  ethylene  oxide found  an




association between ethylene oxide exposure and spontaneous abortion.




     In conclusion, ethylene oxide  produces adverse reproductive and teratogenic




effects in  both  females  (maternal toxicity, depression  of  fetal  weight  gain,




fetal death,  fetal  malformation)   and  males  (reduced  sperm numbers 'and  sperm




motility) if  the  concentration of  the chemical  reaching  the target organ  is




sufficiently high or if exposure  at lower levels is  sufficiently  long.   Thus,




the experiments in which ethylene oxide was injected intravenously  have produced




more detrimental effects than  the  short-term inhalation experiments.  However,




even short-term inhalation  experiments have  resulted  in suggestive evidence of




detrimental effects.




     Ethylene oxide has been shown to induce gene mutations in bacteria,  fungi,




higher plants,  Drosophila,  and  cultured  mammalian  cells  in tests  conducted




without the  use  of exogenous  hepatic metabolic  activation  systems.   It  is




therefore a direct-acting  rautagen.   Ethylene   oxide  has  also  been shown  to




                                      1-5

-------
induce dominant  lethal  effects in  mice  and  rats;  chromosomal  aberrations  in




higher plants, Drosophila,  mice,  and rats;  and micronuclei in  mice and  rats.




Based on these positive  findings  in different test systems, ethylene  oxide  is




judged to  be capable  of causing  chromosomal aberrations.  It  has also  been




shown to induce  sister  chromatid  exchange (SCE) in rabbits, rats,  and humans.




     Tissue distribution  studies  have shown that ethylene  oxide  reaches  the




gonads.  This result  is  consistent  with evidence  that ethylene  oxide  causes




unscheduled DNA  synthesis  (UDS)  in  germ  cells  of  male  mice  and  heritable




mutations in  insects  and  rodents   (i.e.,   sex-linked  recessive  lethals  and




heritable translocations  in Drosophila,   dominant lethals  in   rats and  mice,




and heritable  translocations  in  mice).   Ethylene  oxide  can  therefore  be




regarded as mutagenic both in somatic cells and in germ cells.




     Based on the available data, there  is  overwhelming evidence that  ethylene




oxide is a  direct-acting mutagen  that has the  potential to  cause  mutations  in




the cells  of  exposed  human   tissue.   The  observations  that   ethylene  oxide




reaches and  reacts  with mammalian gonadal DNA,  and causes heritable mutations




in intact  mammals,   indicates  that  it   may  be  capable  of  causing  heritable




mutations in man provided that the pharmacokinetics of ethylene oxide in humans




also results in  its distribution to the DNA of germ cells.




     Both human  and  experimental  animal data  are  available  to  assess  the




carcinogenicity  of  ethylene  oxide.   The  human evidence suggests an association




between exposure and  cancer incidence,  while the  animal  evidence  is more sub-




stantial.




     Three  epidemiologic studies   showed a  significant  association  between




ethylene oxide  exposure  and   the  occurrence of  cancer.   Two   of  the studies




found  an  excess risk  of  leukemia  associated   with  ethylene   oxide exposure.




                                      1-6

-------
While  these  studies  have shortcomings and are  not  definitive,  they do, never-




theless, constitute  "limited,"  bordering on inadequate, epideraiologic evidence




for human  carcinogenicity  using the U.S. Environmental Protection Agency (EPA)




Proposed Guidelines  for Carcinogen  Risk Assessment.   Two long-term inhalation




studies in  rats show  statistically  significant responses  for  leukemia,  brain




tumors and  peritoneal  mesothelioma.   In  addition,  positive  results  for  the




carcinogenicity of ethylene  oxide have  been  obtained by subcutaneous injection




in mice and  by intragastric administration  in rats.   The  animal  evidence  for




the carcinogenicity  of ethylene oxide  is  judged to be  "sufficient" using the




EPA weight-of-evidence  classification system.




     On the  basis  of  the  human,   animal,   and mutagenic  data cited  herein,




ethylene oxide  is  classified as  being  "probably  carcinogenic  to  humans"  and




belonging in EPA  Group Bl.  This  classification is qualified as  bordering on




Group B2,  however, because  of limitations  in the human evidence.   According to




the International Agency for Research on Cancer  (IARC) guidelines for evaluating




carcinogen evidence,  ethylene oxide would be  classified as Group  2A,  meaning




that ethylene oxide is  a "probable human carcinogen," but bordering on Group 2B




because of limitations  in the human  evidence.   (See  Appendix 9B for a descrip-




tion of the IARC classification system.)




     Presuming that ethylene  oxide  is carcinogenic  in  humans,  upper-limit  in-




cremental  unit  risk  and potency estimates have  been extrapolated  from the 2-




year rat inhalation studies.  These estimates are upper limit in the sense that




a true risk  level cannot  be pinpointed because of uncertainties   in  low-dose




extrapolation,  and, therefore,  a modelling technique is employed which produces




a statistical upper-bound estimate  of risk while  retaining  biological  plausi-




bility.  Upper  limit  means  that  the true  risk is  not  likely  to  exceed  the




                                      1-7

-------
calculated value and  may be  lower.   These estimates are  within the  range  of




uncertainty of those  derived  from the human  studies.   Based on  leukemias  and




brain gliomas in rats, extrapolation to humans yields an upper-limit incremental




unit risk  estimate  of  1.0 x  10~^,  for  lifetime  cancer  risk  resulting  from




continuous exposure to  air that  contains an ethylene  oxide concentration  of




1 y g/rn-^.  The  relative potency index for ethylene oxide,  which is  based  on




both the upper-limit  unit  risk value  and the  molecular  weight,  is in the lower




part of  the  third  quartile  of  54 chemicals  that  the CAG  has  evaluated  as




potential or known human carcinogens.
                                      1-8

-------
                                 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  regulation of ethylene




oxide under Section 112  of  the Clean Air Act.   However,  based on the expressed




interest of other  Agency offices, the  scope of this document  was expanded to




address ethylene oxide  in  relation  to sectors  of  the  environment  other than




air.  It is fully  expected  that  this document  will serve the information needs




of many  government agencies and  private groups with health-related interests




in ethylene oxide.




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




ature has been  surveyed,  key studies have been  evaluated and conclusions have




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




identified.




     The document  considers  all  sources  of ethylene oxide  in the  environment,




the likelihood  of  human  exposure,  and  the  possible  effect  on man and  lower




organisms from absorption.  The information found in the document is integrated




into a format  designed  for  risk  assessment  use.  However,  the  information in




this document  on environmental levels and exposure is not intended,  nor  should




it be used,  to support any  conclusions regarding risks  to public health.   When




appropriate,  the authors  of the  document  have attempted to identify gaps  in




current knowledge that limit risk evaluation  capabilities.




     The literature searches that support  this document  vary  somewhat.   The




document is  current   to  February  1984  with  the  following  exceptions:   the




                                      2-1

-------
mutagenicity section  is  current  to  January   1984,  and  the  carcinogenicity




section is current to January 1985.
                                      2-2

-------
                      3.   PHYSICAL AND  CHEMICAL  PROPERTIES
3.1.  SYNONYMS AND CAS NUMBER

     Synonyms:  1,2-epoxyethane
                ethylene oxide
                oxirane

     CAS Number:  75-21-8
3.2.  STRUCTURAL AND MOLECULAR FORMULAS

     Structural formula:
     Molecular formula:
3.3.  TORSIONAL ANGLES AND BOND DISTANCES (Hirose, 197*0a
Torsional
Angles

-------
3.4.  PHYSICAL PROPERTIES OF PURE ETHYLENE OXIDE








3.^.1.   Description.    Ethylene  oxide  is  a  colorless,  flammable gas  which




condenses  at low  temperatures to  a  colorless,  clear,  mobile  liquid  (Cawse




et al., 1980; Hawley, 1981).








3.4.2.  Molecular Weight.




     44.05 (Weast, 1972)








3.4.3.  Melting Point.




     -111°C  (Weast, 1972)








3.4.4.  Boiling Point.




     10.4°C  (at 101.3 kPa = 1 atm) (Cawse et al., 1980)








3.4.5.  Boiling Point Change with Pressure Change (Abp/pressure at  100 kPa).




     0.25 K/kPa (Cawse et al., 1980)




     0.033 K/torr  (Cawse et al., 1980)








3.4.6.  Density.



     dj°:  0.8824  (Weast, 1972)








3.4.7.  Coefficient of Cubical Expansion  (at 20°C, per °C).




     0.00161  (Cawse et al., 1980)
                                      3-2

-------
3.4.8.  Refractive Index (at 7°C).

     1 .3597 (Weast, 1972)



3.4.9.  Vapor Pressure (Cawse et al. ,  1980).

      Temperature               Vapor Pressure

          °C                  kPa           Torr
-40
-30
-20
-10
0
10
20
30
40
50
60
70
80
90
100
8.35
15.05
25.73
42.00
65.82
99.54
145.8
207.7
288.4
391.7
521.2
681.0
875.4
1108.7
1385.4
62.6
112.9
193.0
315.0
493.7
746.6
1093
1558
2163
2938
3909
5108
6566
8315
10390
3.4.10.   Aqueous Solubility3 (Cawse et al.,  1980).

      Pressure              	  Temperature	

     kPa   torr             5°C       10°C       20°C
20
27
40
53
67
80
93
101
150
202.5
300.0
397.5
502.5
600.0
697.5
757.5
45
60
105
162
240
NT '
NT
NT
33
46
76
120
178
294
NT
NT
20
29
49
74
101
134
170
195
      Solubility in  m£  vapor/mJ,  water, vapor  volume
      at  0°C  and 1 atm
     NT = Not tested


                                      3-3

-------
3.4.11.  Freezing Point of Aqueous Solutions (Cawse et al.,  1980)
     Ethylene Oxide
Freezing Point
Weight %
0
2.5
5
10
15
20
30
40
50
60
70
80
90
100
Mole
0
1.
2.
4.
6.
9.
14.
21.
29.
38.
48.
62.
78.
100
*

0
1
4
7
3
9
4
0
0
8
1
6

o
0.0
-0.9
-1.6
5.6
8.9
10.4
11.1
10.4
9.3
7.8
6.0
3.7
0.0
-112.5
C


(eutectic)



(max)







3.4.12.  Boiling Point of Aqueous Solutions (Cawse et al.,  1980)
     Ethylene Oxide
Boiling Point
Weight %
0
2.5
5
10
15
20
30
40
50
60
70
80
90
100
Mole %
0
1.0
2.1
4.4
6.7
9.3
14.9
21.4
29.0
38.0
48.8
62.1
78.6
100
°C
100
70
58
42.5
38
32
27
21
19
16
15
13
12
10.4
3.4.13.  Flash Point (tag open cup).



     <-l8°C (Cawse et al., 1980)
                                      3-4

-------
       .  Flash Point of Aqueous Solutions (Cawse et al., 1980),



      Ethylene Oxide                  Flash Point

         Weight %                   Closed Cup (°C)
            1                             31

            3                              3

            5                             -2
3.4.15.  Explosive Limits in Air, Volume % (Cawse et al., 1980).



     Upper Limit   100?



     Lower Limit     3%








3.4.16.  Heat of Combustion at 25°C (Cawse et al.,  1980).



     5.17 kJ/mol



     1.24 IcCal/mol







3.4.17.  Log Octanol/Water Partition Coefficient.



     -0.30 (Hansch and Leo,  1979)







3.4.18.  Ultraviolet Spectroscopic Data (Weast,  1972).



     Xgas = 169 nm
      max


     log e = 3.58



     Xgas = 171 nm
      max


     log e = 3.57







3.5.  PHYSICAL PROPERTIES AND DESCRIPTION OF COMMERCIAL ETHYLENE OXIDE



     The physical properties and specifications for commercial  ethylene  oxide



are described in Table 3-1.






                                      3-5

-------
                                                  TABLE 3-1
                             Manufacturers'  Specifications for Ethylene Oxide
                                                                             a,b

Purity, wt % rain
Water, wt % max
Aldehydes, as acetaldehyde, wt % max
Acidity, as acetic acid, wt % max
C0_, wt % max
uo Total Cl as Cl~, wt % max
CTi
Nonvolatile residue, g/100 m£, max
Color, APHA, max
Residual Odor
Appearance
Acetylene, max
BASF
99.95
0.005
0.005
0.002
0.005
0.005
0.010
10
NA
NA
NA
Celanese
99.95
0.02
0.01
0.002
NA
NA
0.01
10
none
clear
NA
Dow
NA
0.03
0.005
0.002
0.002
0.005
0.01°
5
NA
NA
0.0005
Jefferson
NA
0.03
0.025
0.005
NA
nil
0.01
NA
none
clear
nil
Shell
NA
0.03
0.010
0.0020
NA
NA
0.010
10
none
clear
NA
Wyandotte
NA
NA
0.003
0.002
0.005
0.0005
0.01
10
mild
NA
NA
 Source:  U.S. EPA, 1980
 This information was obtained from the respective manufacturer's product data sheets, available from each
manufacturer on request.
Presently, 0.005 g/100 m£ in Dow ethylene oxide (Kurginski,  1979)
NA = Not available; wt = weight;  max = maximum;  min = minimum

-------
     Commercial grade ethylene  oxide  has  a purity >99.9/f.  Specific impurities




include trace quantities  of  water,  aldehydes (specified as acetaldehyde), acid




(specified  as  acetic  acid),  chloride,  and an  unspecified  residue.   Since




commercial grade ethylene  oxide  is  virtually pure,  its physical properties are




the same as those described  above for pure ethylene oxide.









3.6.  CHEMICAL PROPERTIES




     The  majority  of  information  contained in  this  section  was  taken  from




Cawse et al. (1980).




     Ethylene oxide  is a  highly reactive  epoxide.    Industrially,  it is used




principally as  an  intermediate  for a wide  variety  of compounds.   Most of its




reactions  involve  opening the epoxide  ring.  An  exception is the formation of




oxonium salts with strong anhydrous mineral acids.









3.6.1.  Reduction.   Catalytic hydrogenation  or chemical  reduction of  ethylene




oxide results in the formation of ethanol.









3.6.2.  Clathrate Formation.  Ethylene  oxide and  water form a stable  clathrate




containing 6.38-6.80 molecules  of ethylene oxide to  46 units of  water in the




unit  cell.   The maximum  observed melting point  for  these compounds  is 11.1°C




(Section 3.4.11).









3-6.3.   Polymerization.    Low molecular weight polymers  can be  formed by the




reaction  of  ethylene  oxide  and  water  or  alcohols.    The average   molecular




weight  of  these  polymers   (polyethylene  glycols)   ranges  from  200-14,000,




depending upon the reaction  conditions.   High polymers, with molecular weights




                                      3-7

-------
ranging   from   90,000  to   U   x   10  ,   are  formed  by   coordinate  anionic

polymerization.    This  reaction  involves  the  coordination  of  a  metallic

compound   with   ethylene   oxide   to   initiate   the   reaction.      Numerous

organometallic  and   alkaline   earth   compounds  and  mixtures  are   used  as

catalysts.    This  process  is  important   in  the  formation  of  non-volatile

residues  during ethylene oxide storage  (Section  3.5).   The main  catalyst for

this process is rust, and no inhibitor has been found.



3.6.4.  Other  Reactions.   Table 3-2 lists  some other representative reactions

of ethylene oxide.



3.6.5.    Hydrolysis  and  Related  Reactions.   Epoxides  degrade  in water  by

hydrolysis and  related ionic  reactions  and,  possibly,  by  radical oxidations.

The hydrolysis  chemistry involves  cleaving a carbon-oxygen  bond  of the cyclic

ether to  form  ethylene  glycol.   Bronsted et al.  (1929) noted  the  pathways for

ethylene  oxide  hydrolysis  in  aqueous  hydrochloric  acid,  describing hydrolysis

as  a  combination  of  a  noncatalytic  reaction  (herein  referred to  as  the

spontaneous  hydrolysis)  and  an  acid-catalyzed  hydrolysis.    Reaction  with

chloride  ion  was  similar  to hydrolysis in  that  chloride and  epoxide reacted

without catalysis and with acid catalysis.

     Long and  Pritchard (1956)  demonstrated that epoxide  hydrolysis was also

base catalyzed.   For any epoxide,  the degradation pathways  are  as follows for

the neutral (I), acid-catalyzed (II), and alkali-catalyzed hydrolyses (III):

                  k
(I)  C HaO + HpO ——>HOCH CH  OH          (II)  C^H.O + HO k.
                                                            	>•   HOCH0CH.OH
                                                            H 0+      ^  *
                                      3-8

-------
                                   TABLE 3-2



                      Typical Reactions of Ethylene Oxide
1.   Crown Ethers



     n H-C - CH0 Catalyst> cyclic {CH-CH n^


        2\/2                     2  2  "
           0
2.   Hydrolysis



     H C - CH0 + H00 	> HOCH0CH0OH

      2\/ 2    2           2  2

         0
3.   Reaction with Alcohols



     H0C - CH~ + ROH 	> R-0-CH.CH_OH      -> R_04CH.CH00^ H
      £ \   /  d.                  22    oxide          2  2  n


         0
4.   Reaction with Organic Acids and Acid Anhydrides



     RCOOCOR + H_C - CH0 	> RCOOCH.CH0OH ethyj-ene> RC004CH0CH00^ H
                d \   t d            22    oxide           2  2  n


                   0
5.   Reaction with Ammonia and Primary and Secondary Amines
     R-NH  -.- H_C - CH_ - > R-NH-CH0CH0OH      -> R_NH4CH0CH00^ H
         d    d. \.   /  d             22    oxide           2  2  n


                 0
6.   With Hydrogen Sulfide and Mercaptans (e.g., glutathione, cystine)



     H2C - CH2 + RSH 	>



         0
                                      3-9

-------
                                 TABLE 3-2 (cont.)
7.   Reaction with Pyridine (and possibly other nitrogen heterocycles)

           + H0C - CH0 —	>    /(5V-CH0-CH9OH + OH~ 	>    /C^N + HOCH?CH?OH
                \ /             ^—/                           ^	f
                 0
8.   With Phenols
                  - CH2
                  0
9.   With Hydrogen Cyanide


     H C - CH  + HCN 	> HOCH CH CN 	> CH =CH-CN
        \ /                                  acrylonitrile
         0
                                        3-10

-------
   (III)  C2H40 + H20 —*-+ HOCH2CH2OH

                      OH~
Table  3-3  summarizes  hydrolysis  data  for  ethylene oxide.   The  temperature

coefficients for the rate constants are the following:



          log kA = 10.753 + log T -0.0255/R - 79.5/RT (Long et al., 1957)
          log kN = 7.726 - 79.5/RT (Lichtenstein and Twigg, 1948)

          log kg = 9.312 - 75.3/RT (Lichtenstein and Twigg, 1948)



     Epoxides can  also  react  with nucleophiles  (anions or Lewis  bases).   The

chemistry,  although  similar  to hydrolysis, is more  complex.   The epoxide ring

can be cleaved by spontaneous reaction or by acid-catalyzed reaction:


                                         I
               -C                 k    -COH
                '^0 + X~ + HO —2-+   I    + OH~
               — U             d.        — L/A
                                         I

                                         I
               ~C \               kx  ~COH

                                       "l


Table  3-4  summarizes specific rate  constants  for reactions  of  ethylene oxide

with various  anions.  The  consensus is that the  spontaneous reaction is S.,2,

but disagreement  exists  over whether  the  acid-catalyzed  epoxide  ring opening

is A1-like  or A2-like (Long  et  al., 1957; Lamaty et al.,  1975;  Pritchard and

Long,  1956;  Pritchard and Siddiqui,  1973;  Virtanen  and Kuokkanen,   1973).   A

discussion of the mechanism is beyond the scope of this  review.
                                      3-11

-------
                                   TABLE 3-3
                     Hydrolysis Kinetics of Ethylene Oxide

Temperature
(K)
293
293.2
298
298
298
298
298
298
NR
303.2

kA x 103
(M-IS-I)
5.34a
NR
9.3°
NR
NR
NR
9
NR
10. Od
16. 9e
Specific Rate Constants
kN x 107
(S-1)
3.6ia
4.2b
6.75°
5.62f.S
6.17f'h
6.6lf,i
5.56J
5.8^
NR
NR

kB x 104
(M-1S-1)
NR
0.65b
1.0d
NR
NR
NR
1.1
NR
NR
NR
aBronsted et al., 1929
 Lichtenstein and Twigg, 1948
GEastham and Latreraouille,  1952
 Pritchard and Long, 1956
eLong et al., 1957
fConway et al.,  1983
sRiver water pH 7.4
h
i
Sterile river water pH 7.4
 Sterile distilled water
JLong and Pritchard, 1956
kKoskikallio and Whalley, 1959
NR = Not reported                     3_12

-------
                                   TABLE  3-4



    Specific Rates of Reaction of Anions and Lewis Bases with Ethylene Oxide
Lewis Base
or Anion
C1~



Br~

Pyridine
Temperature
K
293
298
298
300
293
298
291
I06ky
U/mol-sec,)a
NR
NR
0.3056
NR
NR
NR
200 (water)d
102kx
U2/mol2-sec)a
2.17 (water)5
3.67 (water)b
NR
8.23 (50% aqueous
ethanol)0
8.67 (water)5
14.5 (water)5
NR
 k  = neutral reaction; k  = acid catalyzed
  y                      x



 3Bronsted et al., 1929
 Lamaty et al.,  1975




 Pritchard and Siddiqui, 1973
 Conway et al.,  1983



NR = Not reported
                                     3-13

-------
     Some products of epoxide reaction with Lewis bases or with anions are not



stable.    For  example,  tertiary  amines,  such  as  pyridine,   are  capable  of



catalyzing epoxide hydrolysis to glycol:
                            , HOCH2CH2 NC5 H5  -^ HOCH2CH2OH

   CH
     Aqueous chemical degradation in the environment can be estimated from the



contributions of hydrolysis (Equation  1)  and anion reactions (Equation 2):




          -dC
             dt      - V*N - *A"H3Cf   "  *B°OH~'"epox                           (1)
          -dC
                     = (k .C,,  + k  .C-.C,,  0+)C	                             (2)
epox

dt
where C..,  k .,  and k .  refer to the concentration and specific rate constants
       AI   y^-        xi


for each anion  or  Lewis  base.   The overall degradation rate is the sum of all



contributions,  as given in Equation 3:
       dC
                                             Z(ky.  +  KxiCH0+)CA.]C          (3)
     The relative importance of chemical hydrolysis vs. reaction with  chloride



ion  was assessed  for  ethylene  oxide.    Degradation  half-lives  and  product



distributions  (chlorohydrin-to-glycol  ratios)  were  estimated  for  freshwater
                                     3-14

-------
and  marine  water  (NaCl  concentration  of  3%  or  0.513 M).    The  following

specific  rate  constants  from  Tables  3-3 and  3-4 were  used:



                          kN    0.661  x  10~6 s~1

                          kA    9 x  10~3  M"1 s"1

                          kg    1 x  10~4  M"1 s"1

                          k P1  0.305  x  10~6 M~1 s~1
                          y             ?    ?   ?
                               3.67 x 10~^ M   s"


Estimates were calculated for pH  5.0,  7.0,  and  9.0, which is approximately the

pH  range  of  natural  waters.   Half-lives  for  chemical degradation  and the

chlorohydrin/glycol ratios (for sea  water reactions)  are  summarized  below:


                               Calculated Ethylene Oxide
                                  Half-Life  at 298 K
                                       (hours)

                            pH       5      7      9

  Freshwater                       256    291    291
  Saline Solution
    0.85? (physiological)                 273
    1/K                             240    270    270
    3% (marine)                    212    236    236

Conway et al.  (1983)  used buffered  (pH 7)  sterile  solutions of 0,  1,  and 3%

NaCl to  hydrolyze  ethylene  oxide and  reported half-lives of 314,  265, and 224

hours, respectively.   The half-lives  for  river  water  (pH 7.4),  sterile river

water (pH 7.4), and sterile distilled  water  were  341, 310, and 293 hours.  The

chlorohydrin/glycol ratios experimentally determined  by  Conway et  al.  (1983)

were 0.11 and 0.23 for 1  and 3% saline  solutions.

     These data provide  some  understanding  of  the  fate  of  ethylene  oxide in

biological fluids.   The hydrolysis half-life  in physiological  saline  (0.85?

NaCl) is  273  hours  or 11.4 days.    This  long a  half-life would  clearly allow


                                     3-15

-------
for  other reactions  to  take  place.   As  an  example,  the  half-life  for  the
ethylene  oxide  reaction  with  pyridine  in  water  is  58  minutes.    Other
nucleophiles present  in  biological  systems (e.g., RS~,  PhNH_)  are  known to be
more nucleophilic  than pyridine and  may react with  ethylene oxide  much more
rapidly than water or chloride.
     Hydrolysis   or   hydrolysis-type   reactions   are   the   most  significant
industrial reactions  of  ethylene  oxide.   Ethylene  glycol  is  the  hydrolysis
product;  higher glycols  (diethylene,  triethylene,  and  polyethylene glycols)
and glycol ethers result  from  the  reaction of ethylene  oxide  with  glycols and
alcohols,  respectively.   Glycol esters of carboxylic  acids and phenols,  and
ethers  of cellulose,  starch,  and  other  polyols  are  also prepared  through
hydrolysis-like  reactions.     For   example,  reaction  of ethylene   oxide  and
nonylphenol  yields  nonylphenoxypolyethoxyethanol, a  non-ionic,  surface-active
agent (Blackford, 1976a).


          n CH2 	 CH2 + C9HigOH 	> CgH^OCCH^O^H


     Ethylene  oxide  reacts  with amines  by pathways  similar  to  reactions with
hydroxyl  compounds.   Reaction  of ethylene  oxide  and ammonia yields the commer-
cially important ethanolamines:

                     0
          NH3 + nH2C - CH2 	> H£N  4 CH2CH20 ^ H


where n  is typically  1 to 4.   Choline is prepared  by reacting trimethylamine

with ethylene oxide (Jukes,  1964):
                                      3-16

-------
          (CH ) N + H2C - C
Some ionic reactions of ethylene oxide are listed in Table 3-2.






3.6.6.    Free  Radical  Reactions.    The  free-radical  chemistry  of  ethylene



oxide is  of  particular importance in  determining its fate  in the atmosphere.



The  most  important  free-radical  reaction  is  the  reaction  with  hydroxyl



radical.



     Only  one  study  was  found  in  the available  literature.   Fritz  et al.



(1982) reported  the  results of a study  utilizing a laser photolysis/resonance



fluorescence  unit  designed  to  study  the  reactions  of  OH  radicals  with



anthropogenic pollutants.   They  generated hydroxyl radical  by HNO_ photolysis



and  studied   the  reaction  at  297,  377, and  435  K,  at  10  torr  (Ar).   The



following relations were reported (mean  + 36 ):
and
           v\ = (8.0 + 1.6) x 10    cm /molec-sec
           K )
     k(T) = (1-1 ± O-1*) x 10~11 exp (-1460/T) cm3/molec-sec
The mechanism  involved hydrogen  abstraction  followed by ring opening, reaction



with oxygen,  nitrogen oxide and  finally decomposition  to  carbon  monoxide and



formaldehyde.   Ring opening may  take place  either  before  oxygen  addition or



after NO reaction.
                                      3-17

-------
                      4.  SAMPLING AND ANALYTICAL METHODS









4.1.  SAMPLING




     The  state-of-the-art  in  air  sampling utilizes  solid sorbents.   Samples




can subsequently desorb by solvent  or thermal means.   Critical  factors in the




method  are  the  capacity  of  the  sorbent  to  retain  the  epoxide  during the




collection and the complete desorption of the epoxide.




     Brown  and  Purnell  (1979)  evaluated  Tenax GC  sampling  tubes  for  use in




ambient air monitoring  studies and  found  them to be inappropriate for ethylene




oxide.   Although  most  of the 71  compounds  tested  were  adequately retained,




ethylene oxide was not,  having the third poorest retention.




     Pellizzari  et  al.  (1976)  evaluated  Tenax  GC  and  other  sorbents for




sampling  atmospheric  propylene oxide,  a  compound  chemically very  similar to




ethylene  oxide.   Table  4-1 compares  the  breakthrough volumes for several  sor-




bents.  The effect of humidity on the breakthrough volume was tested for  Tenax




GC: breakthrough  volume remained  unchanged  in  the range  of 4.0-4.5 &/g  when




humidity  was  increased  from  41-92$.    Storage  time  affected the  recovery of




diepoxybutane  (300  ng)  from  Tenax   GC   cartridges  (desorbed   thermally and




analyzed  by GC.    When  analysis  was  immediate,  recovery was 100/&.   After the




loaded  cartridge   was stored  for 1 week, recovery  dropped  to  76/6.   Combined




transport  (6 days) and  storage yielded recoveries of 75 and  6^% after  1  and  2




weeks,  respectively.    Since  Brown and  Purnell  (1979) and  Pellizzari  et al.




(1976)  used comparable  methods  for  determining  the  breakthrough  volume, it




appears  that  propylene  oxide  and  ethylene oxide  behave similarly.   Brown and




Purnell  (1979)  have  noted  that,   under   the conditions  of  the  test   (5-600




                                       4-1

-------
                                   TABLE 4-1

          Breakthrough and Safe Sampling Volumes for Propylene Oxide
                             with  Several Sorbents
Sorbent Breakthrough Volume
H/g (sorbent)a
PEL Carbon
PCB Carbon
SAL9190
MI808
Tenax GC (35/60)°
Porapak Q (100/120)
Chromosorb 101 (60/80)
Chromosorb 102 (60/80)
Chromosorb 101 (60/80)
36
40
40
24
4
4
4
8
>36
Safe Sampling Volume
a/g)b
9
10
10
6
1
1
1
2
9
 Pellizzari et al.,  1976

DBrown and Purnell,  1979

'Mesh size
                                      4-2

-------
m£/minute  flow rate,  <100  ppra vapor  concentration,  <20°C,  and  <95% relative




humidity)  the  breakthrough volume  is  not  less  than  50%  of  the  retention




volume,  and  a safe  sampling  volume   is  50%  of  the  retention  volume.   If




propylene  oxide behavior is analogous  to  ethylene oxide,  the reported  levels




of ethylene  oxide  present  in air  could not be determined accurately, since the




great  majority of monitoring  studies  use  air  samples  larger  than  the  break-




through volume  for ethylene  oxide.




     The  National Institute for  Occupational  Safety  and  Health  (NIOSH)  has




published  standard procedures  for ethylene  oxide collection in  air   (NIOSH,




1977).   Their  procedure  calls for  the sampling  of  5  £  of  air  through glass




tubes  packed with activated coconut-shell  charcoal.   For  ethylene  oxide,  two




tubes  mounted  in  series are used,  containing 400 and 200  mg  of  charcoal.   If




the back-up  tube  (200 mg)  contains  >25% of  the  epoxide,   the  analysis  is  not




considered  valid.   Ethylene oxide should  be desorbed  from  the  charcoal with




2.0 m£ of  carbon  disulfide;  aliquots are then analyzed by  GC with flame-ioni-




zation  detection.    In  NIOSH  (1977)   tests on  the  analytical  parameters,




ethylene  oxide was sampled  at concentrations  from  41-176 mg/m   (23-98 ppm);




precision  (CVT) was  0.103  (or  standard deviation of 9.3 mg/m ),  and accuracy




was 0.9$  lower than  the  "true"  value.   NIOSH  (1977) recommended  this  method




for industrial hygiene monitoring at sample concentrations of 20-270 mg/m .




     The monitoring  of occupational exposure to  ethylene   oxide  by  adsorption




through activated  carbon and  subsequent desorption with carbon  disulfide  was




reported by  Qazi  and  Ketcham  (1977).    These investigators  evaluated  several




carbon  and  noncarbon  adsorbants  and   concluded   that  Columbia  JXC  activated




carbon was  most suitable for  the collection of  ethylene  oxide  in air.   The




breakthrough volume  for ethylene  oxide with this adsorbant  was  dependent  on




                                      4-3

-------
both  the  flow  rate  and   the  moisture  content  of  the  air.    At  relative




humidities >60% and  sampling rates  of  20-25  nd/minute,  the breakthrough volume




was  <10  H.    The  quantification   of  ethylene oxide  was  done  by  GC  with  a




TERGITOL  TMN  or  UCON LB550X column interfaced with  flame-ionization detector.




At concentration levels of  0.5-5.0  ppm,  the  average  recovery of ethylene oxide




by this  method was  97%, with  a relative standard deviation and error  of 3.8




and 2.956,  respectively.   The lower detection  limit  of  the method  was 0.15 ppm




with a sample volume of 10 9,.




     The  quantification of  ethylene oxide and  its  two  volatile metabolites, 2-




chloroethanol  and  ethylene glycol, at a concentration level of  1-10  ppb in




biological  samples was attempted   by  EOIC  (1984).   A  purge and  trap method




consisting  of nitrogen gas  bubbling  through  purge  cells,  and  charcoal traps




for  collecting   the  transferred  volatiles  was   found   most   suitable  for




biological  samples.   The subsequent quantification  of ethylene oxide  and its




metabolites was  done by thermal  desorption  of the  charcoal trap  and analysis




on interfaced  GC/MS operated on a  selected  ion-monitoring  mode.   Both packed




carbowax  20  M  and  Earbopack/THEED  GC  columns  were  used;  however,   this




technique  provided non-reproducible data due  to deterioration  (peak tailing)




of the GC columns  with time.



     Romano and  Renner  (1975)  described  the  results  of a six-laboratory inter-




comparison  of three methods  for  sampling  ethylene  oxide  concentrations in




surgical  equipment.   The  study  was  administered  through a Subcommittee on




Ethylene  Oxide Sterilization  of  the  Association  for  Advancement  of  Medical




Instrumentation.    The  three   sampling  methods  were  vacuum extraction   with




sample  freezeout,   headspace  analysis, and  acetone  extraction.    The  vacuum-




freezeout technique requires  distillation  of volatiles from the  sample, and




                                       4-4

-------
freezing  them  in a cold  trap.   The  sample is then vaporized and an aliquot  is




removed with  a vacuum syringe for GC  analysis.    This  method requires greater




time  and  equipment  than the  other  techniques and  is subject  to  errors  from




equipment  leaks; however,  it is  the  most  sensitive  method,  and,  since the




sample  injected into  the GC  is  a vapor,  column  life is  prolonged.   Acetone




extraction  consists of  partitioning the  epoxide  between  the sample  and the




acetone solvent.   Its advantage is  its  simplicity.   Its disadvantages include




its  inability  to quantitatively extract  epoxide,  impurities  from  the solvent




and  the plastics,  the  reduced lifetime  of GC columns, and  low sensitivity.   In




headspace analysis,  the  sample  is  placed into a  vial equipped with  a septum




for  gas withdrawal  by syringe.  The epoxide partitions between the sample and




headspace gases.   Romano et  al.  (1973)  reported  that  the  headspace technique




has  a lower limit  of 0.1 ppm and  that  the technique can be  automated.   The




advantages  of  this   technique   include  its  ease   of   performance,   speed,




sensitivity, and  relatively  long column life.  Its  disadvantage is that  leaks




in  septa,   vial  caps,   etc.,  can  yield  low measurements.    Among the   three




overall methods,  Romano  and Renner (1975)  found  no  significant  differences,




though they did find slight differences between laboratories.




     Ben-Yehoshua et al.  (1971)  analyzed fruit pulp  by blending it with  50  m£




of  analytical  grade acetone for  30 seconds and  filtering the homogenate  to




clarity.  The  samples were  then stored at  -10°C  in  bottles  with  self-sealing




stoppers.   Measurements  (by  GC)  of added ethylene  oxide and  its residues were




accurate to +5%.   Scudamore  and Heuser  (1971) extracted  wheat flour and  other




commodities, including  coconut,  sultanas,  lentils,  and ground  nuts with 5:1




(v/v) analytical  grade  acetone:water.   The  extraction used  as  little  as 3 m£




solvent/g  sample.  A contact  time of 24 hours was  sufficient  to yield ethylene

-------
oxide  recoveries  (by GC)  of  >95%.   Pfeilsticker et  al.  (1975) extracted  10 g




of grain  (not crushed) with 5  mJl  of methanol using  continuous agitation for 24




hours.  Recovery of ethylene oxide was 73% (25+1.7).




     Brown  (1970)  sampled  and  analyzed  surgical materials (plastic and rubber)




for ethylene  oxide residues by means of  a  three column chromatography system.




This   system  could  separate  ethylene  oxide  and  its  degradation  product,




ethylene  chlorohydrin.     Epoxides  were  extracted  with  £-xylene  (3 days of




contact)  or _co-sweep distillation.   The  £-xylene solution  was passed through




one column  of Florisil;  ethylene  chlorohydrin remained fixed in the column and




ethylene  oxide  passed  through.    The  solution was  then  passed  through  the




second  acid-celite  column,  which  converted  any  ethylene  oxide  to  ethylene




chlorohydrin.   A  third  Florisil  column  retained  the ethylene  chlorohydrin,




which  was   subsequently   eluted   with  petroleum  ether.     The  sample  was




concentrated and analyzed  by  GC.   Brown  (1970)  reported  values as  low  as 1.8




ppm,  but  the  accuracy,   precision,  and  minimum  detection   limit  were  not




described.








4.2.  ANALYSIS




     To date, GC analysis  for  ethylene  oxide has used only flame-ionization or




thermal-conductivity detection.   Neither detection  system is selective,  so the




epoxides must be separated from all  interfering components,  and the analytical




column  must  be chosen  around  potential  interferences.    Columns  used  for




epoxide analysis have  included uncoated  Poropak Q,  QS,  and  R,  and Chromosorb




102 (Taylor,  1977a,b; Ben-Yehoshua  and  Krinsky, 1968; Steinberg,  1977),  and a




variety of  coated  columns.  The  most common liquid phases  appear to be SE-30,




Carbowax 20M,  and  polypropylene  glycol  (Ben-Yehoshua  and  Krinsky,   1968;




                                      4-6

-------
Casteignau and  Halary,  1972; Steinberg,  1977;  Hughes et  al.,  1959).   Bertsch




et al.  (1974)  used a  100m x  0.5mm capillary  column  coated with  Emulphor ON



870.  The  GC methods  in current use appear  capable  of epoxide analysis at the




ppm level.



     The  EOIC  (1984)   used  a  GC/MS/SIM system  for  the  quantification  of




ethylene  oxide and  its  volatile  metabolites  in  biological  samples.    More




recent  techniques  involving  a  negative-ion  atmospheric  pressure  chemical




ionization MS/MS  system developed  by  Sciex  of Canada were  used by  U.S.  EPA




(1984)  for  the  on-site  monitoring of  pollutants  in the  atmosphere.    This




system  was used for the  identification,  but not for the  quantification,  of




ethylene oxide in a synthetic gas mixture.




     Other   analytical  methods   include  various   wet   chemical  techniques.




Epoxides can be analyzed by  ring  opening with specific reagents and subsequent




analysis for that reagent  or one  of its products (Dobinson et al., 1969).  For




example, Mishmash and  Meloan (1972) reported what may be the  most  recent use




of this approach.  Butylene  oxide  was  hydrolyzed to  its glycol, and the glycol




was oxidized with periodic acid.   Residual oxidant was analyzed by adding Cdl?




-starch and  measuring  the starch-I.,  complex concentration  at 590 nm.   They




claimed a detection limit in the nmole range.
                                      4-7

-------
                         5.   SOURCES IN THE ENVIRONMENT








5.1.  PRODUCTION









5.1.1.    Quantities  Produced.   Production  volumes  and  sales  quantities for




ethylene  oxide are listed in Table  5-1 for the years  1972 to  1982.









5.1.2.  Producers, Production  Sites,  and Distribution.  The  producers, produc-




tion  sites,  and  annual capacities  of  ethylene  oxide are listed  in Table 5-2.




ICI  Americas is  building a new  ethylene oxide  plant in Bayport,  Texas; the




naraeplate  capacity is  520  million pounds/year  (Anonymous,   198la).   Dow will




add  400 million pounds/year capacity  onto  its  Plaquemine,  Louisiana, facility




during the fourth quarter  of  1983.   Union  Carbide is building  a  400 million




pounds/year  unit in  Alberta,  Canada,  slated to  be on  stream in  1985.   PPG




Industries and DuPont  are conducting  a  feasibility study to determine whether




to  move   the  former's  idle Guayanilla,  Puerto  Rico,  facility  (rated  at 300




million pounds/year) to Beaumont, Texas, to be operated by both companies.








5.1.3.  Production Methods and Processes.









     5.1.3-1.  INTRODUCTION ~ The  majority  of  the information in this section




was obtained from Cawse et al. (1980).




     Ethylene oxide  is produced  almost  exclusively by direct oxidation,  using




either air or oxygen.   Other processes cannot compete with  the lower operating




costs of  direct  oxidation.   Only one  plant  in  the United States currently has




                                      5-1

-------
                                   TABLE 5-1

                          Ethylene Oxide Production3)b
Year
1982d
1981
1980
1979
1978
1977
1976
1975
1974
1973
1972
Production
5200 (2359)
4937 (2240)
5220 (2368)
5665 (2570)
5012 (2273)
4364 (1980)
4184 (1898)
4467 (2026)
3893 (1766)
4167 (1890)
3962 (1797)
Sales0
NA
NA
531 (241)
560 (254)
525 (238)
549 (249)
439 (199)
409 (186)
457 (207)
501 (227)
454 (206)
aSource:  USTC, 1974, 1975; USITC, 1976, 1977a, 19775, 1978, 1979, 1980, 1981

 All quantities are expressed in millions of pounds; SI units in millions of
kilograms are given in parentheses.
Q
 The difference between production and sales does not enter the merchant
marketplace.

 Projected (Source:  Anonymous, 1982)

NA = Not available
                                      5-2

-------
                                                     TABLE 5-2
                   Ethylene Oxide Producers, Plant Sites, Capacities, Processes, and Technology^
I
uo
Company
BASF Wyandotte, Indust. Chem. Group
Basic Chems. Div.
Calcasieu Chem. Corp.0
Celanese Corp.
Celanese Chem. Co., Inc.
Dow Chemical U.S.A.
Eastman Kodak Co.
Eastman Chemical Prod., Inc.
Subsid. Texas Eastman Co.
ICI Americas, Inc., Petrochems. Div.
Location
Geismar, LA
Lake Charles, LA
Clear Lake, TX
Freeport, TX
Plaquemine, LA

Longview, TX
Bayport, TX
Annual
Capacity^
481 (216)
225 (101)
425 (191)
26Qd (117)
450e (203)

195 (88)
520 (234)f
Process
Oxidant
oxygen
oxygen
oxygen
air
air

oxygen
NA
Technology
Shell
Shell
Shell
Dow
Dow

Shell
NA
  Inter-North, Inc.
    Northern Petrochem. Co.,
    Subsid. Petrochems. Div.

  Olin Corp., Olin Chems. Group

  PPG Industries, Inc.
    Chems. Group, Chem. Div.-U.S.
Joliet, IL          230 (104)     oxygen     Scientific Design

Brandenburg, KY     110 (50)      oxygen     Shell
Beaumont, TX
155 (70)
air
Scientific Design

-------
                                               TABLE 5-2 (cont.)
Company
Shell Chemical Co.
Sun Olin Chemical Co.
Texaco, Inc.
Texaco Chemical Co., Div.
Union Carbide Corp.
Chems. and Plastics Div.
Union Carbide Carbie, Inc., Subsid.
Location
Geismar, LA
Claymont, DE
Port Neches, TX
Seadrift, TX
Taft, LA
Ponce, PR
Annual
Capacity13
700
100
700
1000
1250
640
(315)
(45)
(315)
(450)
(563)
(288)
Process
Oxidant
oxygen
oxygen
air
air
air
air
Technology
Shell
Shell


Scientific Design
Union
Union
Union
Carbide
Carbide
Carbide
aSources:  Anonymous, 198la;  SRI International,  198la,b;  Cawse,  1980
 Capacities are expressed in millions of pounds;  capacities in millions of kilograms are in parentheses.
GPlant is on indefinite standby as of January 31,  1981  (Anonymous,  198la).
Approximately 200 million pounds/year (90 million kg/year) additional capacity can be obtained from a
chlorohydrin unit used for propylene oxide production.
Expansion of 400 million pounds/year (180 million kg/year) is due  in the fourth quarter of 1983.
f
 Under construction
NA = Not available

-------
chlorohydrin  capacity  (Dow  at Freeport,  Texas;  see  Table  5-2).    The  major




drawback  of  the  direct  oxidation  process  is  the  loss  of  =25-30?  of  the



ethyl ene to carbon dioxide and water.








     5.1.3.2.   DIRECT OXIDATION —  The  overall reaction  for  direct oxidation




can be represented as follows:
                                      CH2
                                   0
     5.1.3.2.1.   Air-Based Oxidation  —  The schematic  for  air-based ethylene




oxidation is  presented  in Figure 5-1 .   Little detailed  information  is  avail-




able  concerning   process  technology;  however,   the   salient   features  are




presented below.




     In  the  first  section,  air  and  ethylene  are  fed  into  the  recycle  gas




stream  (the recycle  gas  contains  unreacted  starting  material  from  the main




absorber).   The recycle  stream  is  fed into  a  bank of  tubular  main  reactors,




the  number  of reactors  depending  chiefly  on  the  capacity  of  the  plant,




activity of  the catalyst,  and  size of the  reactors.  In the main reactor,  the




ethylene is oxidized  to ethylene  oxide,  carbon dioxide, and water, as well as




to minor components such as formaldehyde and acetaldehyde.




     Ethylene  conversion  to  ethylene oxide  per pass in the main  reactors is




20-50?.    Oxidation inhibitors  (e.g., vinyl chloride, ethylene  dichloride)  are




added to  retard  carbon  dioxide  formation.    The process  stream  leaving  the




reactor  may  contain  1-2  mole %  ethylene  oxide.    This hot  effluent  gas is




cooled to around 35-40° C and fed to the main absorber.
                                      5-5

-------
         tor
                     •btorbar      raactor
 Puff*
absorber
Dvtorbor
Stripper      Refiner
 Steam
Coolant
                                          FIGURE  5-1
                      Schematic  for Air-Based Ethylene  Oxidation
                                 Source:  Schultze, 1965

-------
     The  main  absorber uses  cold water  to  dissolve  the  ethylene oxide,  some




carbon dioxide, and  traces  of hydrocarbons and aldehydes.   The unabsorbed gas




is split  overhead.   The  largest portion is  used  as  recycle  gas,  and to  cool




the  effluent stream  from  the   main  reactor;  the  gas  then  enters  the  main




reactor.   A much smaller portion of the absorber effluent gas  is  fed as the




main stream to  the secondary or purge reactor.   The effluent  from the purge




reactor is  heat  exchanged with   the main  stream and sent to the purge absorber




which operates in the same manner as the main absorber.




     The  purge  reactor system reacts  a large  portion of  the ethylene present




in the  purge gas  from the main reactor which must  be vented from  the  main




reactor so  that  inert gases  (principally nitrogen  and  carbon dioxide)  do not




accumulate.   Although Figure  5-1  shows  a  two stage  air-based plant  with  a




single purge  reactor,  some  large  plants  have  three  or  more  stages to improve




the  overall yield.    These  plants  merely  place  another purge  reactor  and




absorber in series.




     In some plants, the  ethylene  content of the  vent gas is  sufficiently  high




to make  energy recovery  economical.    This  not only produces  valuable power




from  the  vent  gas,   but  also  reduces  the  hydrocarbon   emissions  from  the




process.




     The  remainder  of  the process involves purification.   The ethylene oxide-




water solution from the absorbers is heat-exchanged  and sent  to the desorber,




where  the  ethylene oxide  is  steam  stripped  under reduced  pressure.    The




ethylene  oxide  is  collected at  the  top  and  compressed for  further purifica-




tion,  while  the  stripped  water  is  recirculated   to  the  main  and  purge




absorbers.
                                      5-7

-------
     The ethylene  oxide  from the desorber still contains  some  carbon dioxide,




nitrogen, aldehydes,  and traces of  ethylene  and ethane,  and  must be  sent to




the stripper.   Here,  the light gases are  separated  overhead and vented, while




the partially purified ethylene oxide is taken  from  the  bottom  of the stripper




and sent  to the mid-section  of a final refining  column.    The  ethylene oxide




from the refining section should have a >99.5 moIe-% purity.




     The  specific  conditions  used  to  operate  ethylene  oxide  plants  are




proprietary  information;  however,   the   general   ranges   suggested   by  the




literature and  patent  reviews have been summarized  by Cawse et al. (1980) and




are presented in Table 5-3.








     5.1.3.2.2.  Oxygen-Based Oxidation — The  differences in oxygen-based and




air-based oxidation processes are  almost  entirely  the result of the  change in




oxidants.   The  main  difference  is  that  the  purge  reactor is absent  in the




oxygen-based process and a carbon dioxide removal unit  and an  argon  vent are




added.  In  the air-based cycle, the  low per-pass  conversion,  the necessity of




complete  ethylene  oxide removal  in the  absorber,  and  the  accumulation of




nitrogen necessitates  a  substantial purge system.   Because of  this,  a staged




reaction-absorption system  is required.   Since the  oxygen-based  process  uses




essentially  pure  oxygen,  the   recycle  gas  is  almost   entirely  unconverted




ethylene;  hence,  there  is  no   need  for  a  purge  system.    Carbon  dioxide,




however, is still  produced  in the  oxygen  system,  and because it has a negative




effect  on   catalyst  selectivity,  the carbon   dioxide  must  be  removed.    In




addition to the  carbon dioxide  removal unit,  an argon vent  is required.  Argon




is a major  impurity in oxygen and  can build  up to the extent of 30-40 rnole-^.
                                      5-E

-------
                                   TABLE 5-3

               Ranges of Reaction System Variables in the Direct
                        Air-Oxidation  of Ethylene  Oxidea
          Variable                                           Range


ethylene, mole %                                        2-10

oxygen, mole %                                          4-8

carbon dioxide, mole %                                  5-10

ethane, mole %                                          0-1.0

temperature, °C                                         220-277

pressure, MPa (psi)                                     1-3 (145-435)

space velocity , h~                                     2000-4500

pressure drop, kPa (torr)                               41-152 (308-1140)

conversion, %                                           20-65

selectivity or yield (mole basis, %)                    63-75
aSource:  Cawse et al., 1980
 The space velocity is the standard volume of the reactant stream fed per unit
time divided by the volume of reactor space filled with catalyst.
h = hour
                                      5-9

-------
In spite  of  this additional purge, the  total  vent  stream from an oxygen-based




plant is much smaller than that of an air-based plant.




     As  is  the  case with  an  air-based unit,  the  main  process  vent  stream




usually contains  a high concentration  of  hydrocarbons.   In  such a  case,  the




purge  stream can  be  used  for  energy  recovery.  The operating  ranges  for an




oxygen-based process are summarized in Table 5-4.




     The  choice   of  oxygen versus  air  as  the  oxidant   is  based  strictly on




economics; in general,  for  small-to-raedium  capacity units (<50,000 tons/year),




oxygen-based plants have lower  capital  cost even with the necessary air separ-




ation  facility.    For medium-to-large  plants  (75,000-150,000  tons/year),  the




air  process  investment  is  smaller  unless  oxygen can be  purchased  from  a very




large  air  separation  facility.   Operating  costs of the  facilities  can  differ




significantly  and are  based  on  the  cost  of  ethylene,  oxygen,  catalyst,  and




energy.








     5.1.3.2.3.   Chlorohydrin  Processes — The chlorohydrin  process was  the




main  method  of  ethylene  oxide manufacture  until   1957.    In  1972,  the  Dow




Chemical  Company converted the remaining  chlorohydrin capacity  plants  to  the




production of propylene oxide,  and  the  process was  not  used again for ethylene




oxide   production  until   1975.    The  Dow   Chemical   Company  has  built-in




flexibility  for   using  the chlorohydrin  process to  produce  either propylene




oxide  or  ethylene oxide.   Since  1975,  part of this capacity has been used for




ethylene  oxide.   During 1975,  the Dow Chemical  Company made  between 25  and 50




million  pounds  of ethylene  oxide  via the  chlorohydrin  process  (Blackford,




1976b).   The chlorohydrin  process is attractive  commercially  only when  a good




supply  of captive  low-cost chlorine  and   lime  or  caustic  soda  is available.




                                      5-10

-------
                                   TABLE 5-4

                   Ranges  of Reaction  System Variables  in  the
                   Direct  Oxygen-Oxidation  of Ethylene  Oxide3
         Variable                                                Range


ethylene, mole %                                             15-40

oxygen, mole %                                               5-8.5

carbon dioxide, mole %                                       5-15

ethane, mole %                                               0-2

argon, mole %                                                5-15

nitrogen, mole %                                             2-60

methane, mole %                                              1-60

temperature, °C                                              220-275

pressure, MPa (psi)                                          1-2.2 (145-319)

space velocity , h~                                          2000-4000

conversion, %                                                7-15°

selectivity or yield (mole basis, $)                         70-77




aSource:  Cawse et al., 1980

 The space velocity is the standard volume of the reactant stream feed per
unit time divided by the volume of reactor space filled with catalyst.
°At 30 mole % ethene
h = hour
                                      5-11

-------
Also, satisfactory markets or disposal facilities are needed  for the major by-



products (Schultze,  1965).



     The chlorohydrin  process  starts with conversion of ethylene  to  ethylene



chlorohydrin  with  hypochlorous   acid.    The  chlorohydrin  is  converted  to



ethylene  oxide  by   dehydrochlorination  with  slaked  lime.    Two  major  by-



products, 1 ,2-dichloroethane ( = 100-150 pounds/1000 pounds  ethylene  oxide)  and



b is (2-chlo roe thyl) ether (=70-90 pounds/1000  pounds ethylene oxide),  are formed



during  the  chlorohydrin   formation;   acetaldehyde   (5-10  pounds/1000  pounds



ethylene oxide) is produced during the dehydrochlorination.



     The formation of  ethylene oxide  from ethylene chlorohydrin can be repre-



sented by the following equation:
          2 HOCH0CH0C1 + Ca(OH)_ 	>•  2  0HC-CH0 + CaCl_ + 2H00
                d  d           
-------
is  refined in two  columns;  the first  column  removes chlorinated  hydrocarbons

and the second column removes acetaldehyde.



5.2.  USES OF ETHYLENE OXIDE

     A description  of the various uses  of ethylene oxide  is given below:


                                        Billion Pounds3    Percent  of Total

     Ethylene glycol                         3.2                    62
     Nonionic surface-active agents          0.62                   12
     Glycol ethers                           0.31                    6
     Ethanolamines                           0.26                    5
     Miscellaneous  applications              0.78                   15
     (higher glycols, urethane
       polyols, sterilant, fumigant,
       export)

     Source:  Anonymous, 1981 a

      Based on 1982 production estimates of 5.2 billion pounds.


The major  users and use  sites  for  ethylene  oxide are listed  in  Table  5-5.   As

can be seen from  this  table,  a very large percentage of  production is captive-

ly  consumed  by  the  primary  manufacturers.    A  general description  of  the

various uses of ethylene oxide is presented below.



5.2.1.   Ethylene Glycol.  By  far,  the largest  single use of  ethylene  oxide is

its  use  captively  as  an  intermediate  in  the  synthesis of  ethylene  glycol,

which is currently  produced by hydration of ethylene oxide.   Current  industry

capacity   to   produce   ethylene   glycol   is   5815 million   pounds   annually

(Anonymous, 198lb).  The growth in  consumption  of ethylene  oxide  has  largely

depended  on  its  use  as  an  intermediate  for  ethylene  glycol  production.
                                      5-13

-------
                                                   TABLE 5-5
                                     Users  and  Use Sites of Ethylene Oxide*
     Company
                  Ethylene  Glycol  Diethylene  Ethanol-
   Location        Glycol   Ethers    Glycol     amine
Triethylene  Polyethylene
  Glycol        Glycol
BASF Wyandotte Corp.

Calcasieu Chem.
Celanese Chem.
Dow Chem.

Eastman Kodak
Northern Petrochem.
Olin Corp.
PPG Ind.

Shell Chem.
Texaco Jefferson Chem.
Union Carbide
Ashland Chem.
Hoadag Chem.
Geismar, LA
Wyandotte, MI
Lake Charles, LA
Clear Lake, TX
Freeport, TX
Plaquemine, LA
Midland, MI
Lcngview, TX
Morris, IL
Brandenburg, KY
Beaumont, TX
Guayanilla, PR
Geismar, LA
Port Neches, TX
Seadrift, TX
Taft, LA
Penuelas, PR
Texas City, TX
Institute and
S. Charleston, WV
Janesville, WI
Skokie, IL
•Source:  SRI International, 1977
+ indicates user of ethylene oxide;  - indicates non-users of ethylene oxide

-------
(Blackford,  19?6b).   Ethylene  glycol  is  used  mainly in  polyester production




and antifreeze formulations (Anonymous, 198lc).








5.2.2.  Nonionic  Surface-Active Agents.   Of the nonionic surface-active agents




synthesized  from  ethylene oxide, -25%  are of  the  cyclic variety,  while -75%




are of  the acyclic variety.   In the  cyclic  group,  ethylene oxide  is used to




make  ethoxylate   alkyl  phenols  and   alkylphenol-formaldehyde  condensates.




Production  of ethoxylated nonylphenol  is probably the  largest  volume product




of  the  cyclic  group;  another   large-volume  product  is  ethoxylated  dodecyl-




phenol.    These surface-active  agents  are  primarily  used in detergents.   The




acyclic surface-active  category includes ethylene oxide used in the synthesis




of  surface-active  polyethylene  glycol esters,  ethoxylated alcohols, polyether




polyols,  ethoxylated  fats and  oils,  and miscellaneous  ethoxylated products,




such as raercaptans, glycols,  and polyols (Cogswell,  1980).  Industry estimates




that ethylene  oxide  consumption for acyclic  surface-active  agents  is expected




to  increase.   The manufacture  of ethoxylated  linear  alcohols,  used in heavy-




duty liquid detergents, will account for most of this growth (Cogswell, 1980).








5.2.3-    Di-, Tri-,  and Polyethylene Glycols.    Ethylene oxide and  ethylene




glycol  react  to form diethylene  glycol,  triethylene  glycol, and polyethylene




glycol.    Diethylene and triethylene  glycols are obtained mainly as  by-products




of  ethylene glycol manufacture.   Diethylene  glycol  is  used to produce  poly-




ester resins,  as  a textile lubricant, and  in solvent extraction.   Triethylene




glycol  is used as a  humectant  and  In  natural  gas dehydration,  vinyl plasti-




cizers,  and polyesters.   Industry capacity  to make  diethylene glycol  is 472
                                      5-15

-------
million  pounds/year;   capacity  to  make  triethylene  glycol  is  =145  million




pounds/year (SRI International, 1977).








5.2.4.   Glycol Ethers.   Ethylene oxide  is combined with  alcohols to manufac-




ture glycol monoethers.   These include ethylene glycol  monomethyl,  monoethyl,




and rnonobutyl  ethers  and  diethylene  and  triethylene monoethyl, monornethyl, and




monobutyl ethers (Cogswell,  1980).  Glycol  ethers  are  used mainly as solvents.




Industry  capacity  to make  glycol ethers  is  865 million  pounds annually (SRI




International, 1977).








5.2.5.   Ethanolamines.   Ethylene oxide  reacts with ammonia  to  form a mixture




of mono-, di-, and triethanolamines.   The  proportion of  each of three ethanol-




amines  is dependent  upon  the ratio  of  reactants  used.   About  25-30/5  of all




ethanolamines are used in  soaps  and  detergents, 5-20% in  scrubbing  acid gases




(especially in the synthesis  of  ammonia),  10?  by the metal industry, 8% by the




textile industry and  5-15$  in toilet goods (Blackford,  1976b).   The remainder




is used in various other applications.








5.2.6.  Miscellaneous Applications.   Ethylene  oxide is consumed in the synthe-




sis of numerous commercial  chemicals.  The  largest amount  in the miscellaneous




group  described  above goes  into  the  production  of  polyether polyols  for




flexible  polyurethane foams.   In  1978,  =100 million pounds  (45  million  kg)  of




ethylene  oxide were consumed  in making these polyols (Cogswell,  1980).




     Approximately  17 million pounds of  ethylene oxide  is used annually  to




make  the  medicinals,  choline and choline  chloride; another 10 million pounds




of ethylene oxide  is  used annually in the  manufacture of  hydroxyethyl starch,




                                      5-16

-------
a semi-synthetic gum  used in textile sizing  and  in adhesives.  The production




of hydroxyethyl cellulose, another adhesive additive,  produced by the reaction




of cellulose  with  ethylene  oxide uses -25 million pounds  (11  million  kg) of




ethylene  oxide annually  (Cogswell,  1980).    Arylethanolamines  are  made  by




reacting  ethylene  oxide  with  either aniline or  aniline  derivatives.   It is




estimated that  -3  million pounds (1.1 million  kg)  of ethylene  oxide  are used




annually  to  make   arylethanolamines  (Cogswell,   1980).     They  are  used  as




intermediates in the production of monoazo dyestuffs.




     Acetal copolymer resins are  produced  by  the  catalytic copolymerization of




1,3,5-trioxane  with  a cyclic  ether  such  as  ethylene oxide.    Ethylene  oxide




consumed  in this  use  is believed  to have  been  -2 to 3  million pounds/year




(0.9-1.1 million kg) from 1977-1978 (Cogswell, 1980).




     Ethylene  oxide  is  used  to  produce  ethoxylated cationic  surface-active




agents  (non-ionic   surface-active agents  are  discussed   in  Section  5.2.2).




Several million pounds of ethylene oxide  are  used annually to produce cationic




agents  such as  ethoxylated   (coconut  oil  alkyl)  amine,   ethoxylated  (tallow




alkyl)  amine,   and various  ethoxylated   fatty  acid amino  amides (Blackford,




1976b).




     Small  amounts  of  ethylene  oxide  are  also  used  as  a  fumigant,  as  a




sterilant  for   food  and   cosmetics,  and  in  hospital  sterilization  (Gilmour,



1978).   In  1975,   an estimated  0.1  million pounds  of  ethylene oxide  were used




for  fumigant  purposes (Landels,  1976).    Dow  Chemical  (Kurginski,  1979)  has




estimated that  
-------
5.2.7.   Discontinued Uses of Epoxides.   The only significant discontinued  use




of  ethylene  oxide  known is the production of acrylonitrile.   Until 1953  (when



acetylene  was  first  used),  all  acrylonitrile  was  produced by  the  catalytic




dehydration  of ethylene cyanohydrin that was prepared  from ethylene oxide  and




hydrogen cyanide.  The  reaction may be represented as follows:
                 HCN - »- HOCH2CH2CN - >• CH2 = CHCN
         0
In  1956,  American  Cyanamid Company closed its  35  million pounds/year plant at




Warners,  New Jersey,  which was  based on this  process.   From then until  1966,




when  it was  discontinued, this  process was  used  only by  Union  Carbide at




Institute,  West  Virginia  (Blackford,  1974).   In  1965,  Union Carbide consumed




90 million  pounds of ethylene oxide to make acrylonitrile.








5.2.8.  Projected  or Proposed  Uses.   Wood treatment is a potentially important




market  for  epoxides (Anonymous,  1977).   The  USDA Forest Product Laboratory has




reported  that  treating southern yellow  pine  with  epoxides  (including ethylene




oxide,  propylene oxide,  and   butylene  oxide)  improves  its  durability.   The



treatment adds 20-30$ (by weight) of the epoxide to the wood.








5.2.9.  Alternatives  to Uses  for  Ethylene  Oxide.   More than  99$  of ethylene



oxide  produced in  the  United  States  is used  as  a  chemical  intermediate in




chemical syntheses of glycols  and  other  compounds.   Alternatives would require




production routes from raw materials other than ethylene oxide.




     Roughly  62% of  the  ethylene  oxide produced is  hydrolyzed  to  ethylene




glycol.   A  new  process  for making ethylene  glycol directly  from ethylene has




                                      5-18

-------
been developed  by Halcon,  Inc.  (Klapproth,  1976).   Ethylene  is reacted with




acetic acid in the presence  of a catalyst to form mono-  and diacetates, which



are then  hydrolyzed  to ethylene  glycol.   Oxirane Corporation  has constructed




an  800  million pounds/year  plant based  upon  this  technology  in Channelview,




Texas.   This  represents  -25%  of the  present  total industry  ethylene glycol




capacity.



     For  other  compounds synthesized  from ethylene oxide,  no  information was




available on synthesis from other raw materials.




     About  0.1  million  pounds  of  ethylene  oxide  are  used   as a   fumigant




annually  (Dow  Chemical  estimates that the volume  of ethylene  oxide  used  as a




fumigant  is  <0.2% of  total production,  which  in 1978  would  be  equal to <10




million  pounds;  Kurginski,   1979).     It  seems  possible  that  alternative




commercial  fumigants could  replace  ethylene  oxide  in  many of  its   fumigant




uses.








5-3.  POTENTIAL FOR ENVIRONMENTAL CONTAMINATION








5.3.1.   Air Emissions  from Production.   Air emissions  from  direct-oxidation




ethylene oxide plants of  all types  consist mainly of ethylene,  ethylene oxide,




and traces of ethane.  The main  process  vent  stream  is  responsible for most of



the  air emissions in  both air-  and  oxygen-based units.    In  air units,  this




vent  is located  on  the  last purge reactor  absorber and  is principally spent




air  (Np,  02,  and some inert gases),  carbon dioxide, traces  of ethylene oxide,




and  generally  <2  mole  %  hydrocarbons.    A catalytic  converter  is   sometimes




added to the main process vent in an air  system.
                                      5-19

-------
     The  analogous vent  stream  from an  oxygen-based  system  is  ^lOO  times




smaller and  contains  a much  higher hydrocarbon  concentration,  and  is  conse-



quently used as  a fuel.   Table  5-6 presents  approximate  concentrations  of




typical vent stream contaminants  for the main process  vent and the  purge  gas




vent.




     For  unburned  vent gas from  an oxygen-based  unit,  the  total  hydrocarbon




emissions have been estimated  to be = 12 g/kg product.  If methane  is  used as a




diluent and  the purge gas  incinerated, the  emissions can be reduced to -^ g/kg




product.  In an air-based  unit  without catalytic combustion  of  the purge gas,




hydrocarbon  emissions  are  estimated  to  be  >30  g/kg  product.  The  use of a




catalytic converter  can reduce  emissions to  -15 g/kg  product.   In a  study




conducted for  the U.S. EPA,  the total ethylene  oxide  emissions in  1978 were




estimated to be  -2 x  10   pounds (9.09 x  10   kg)  (Systems  Application,  Inc.,




1982).




     The  major  aqueous waste is  draw-off from  separator  bottoms  (Liepins  et




al.,  1977).  The  process  water  is  recycled  in ethylene oxide  manufacture  and




in  the  primary use  of ethylene  oxide as an  intermediate   in  ethylene  glycol




manufacture  (Sittig,  1962, 1965).   The  aqueous waste  from  direct  oxidation




plants  will  contain small amounts  of glycols,  aldehydes,  and heavy  glycols




(Cawse et al.,  1980).   No  information  was available  on how  much of the process




water  eventually  is   treated,   and  no  details  were  provided  on  treatment




methods.  The  wastewater  will  have a high BOD,  but inorganic  and  refractory




organics  appear  to be  minimal  problems  (Sittig,  1962,  1965;  Spencer,   1971).




Conventional water treatment (including  filtration and  flocculation) with  a
                                      5-20

-------
                                    TABLE  5-6

           Typical  Vent  Gas  Composition  for Both Air- and  Oxygen-Based
                             Ethylene Oxide Plants*
            Stream
Air-Based
                                                     Range, mole %
                                                           Oxygen-Based
Main Process Vent

  nitrogen
  oxygen
  methane
  ethane
  ethylene
  ethylene oxide
  carbon dioxide
  argon
  water
85-93
1.0-5
0-0.9
trace-0.2
trace-2.5
0-0.01
5-15
NP
0.1-1.5
                                                            2-35
                                                            5-7
                                                            1-35
                                                            trace-0.2
                                                            13-35
                                                            0-0.01
                                                            5-15
                                                            5-15
                                                            0.1-0.5
C02 Rich Purge Gas (water-free)

  nitrogen
  oxygen
  ethylene and hydrocarbons
  ethylene oxide
  carbon dioxide
  inert compounds
                                        13-25
                                        1-26
                                        2.5-8.0
                                        0-1.0
                                        62-80
                                        NP
                    NP
                    0.02
                    0.3-0.9
                    NP
                    99-99.7
                    0.005-0.015
*Source:  Cawse et al., 1980
NP = Not present
                                     5-21

-------
biological  treatment  appears sufficient  to  remove the  major contaminants of




the process water (Spencer, 1971;  Shenderova  et al.,  1972).  There is no solid



waste produced during ethylene oxide manufacture.








5.3-2.  Handling, Transport,  and Storage.  Ethylene oxide could be released as




a  result  of fugitive emissions or  venting during  its handling,  transport, or




storage.    No  specific   information  was available  to describe  these  losses;




information  on  current  practices,  procedures,  or environmental  controls  was




sparse and  no  monitoring information  was available.   The potential situations




of  release  of  epoxides  without  attempting  to  establish  their  relative




importance have been discussed in  the  following paragraphs.




     Bulk shipments of ethylene oxide are commonly made by railroad 10,000 and




20,000 gallon  freight tankers.   Shipments are also made  in  special 55-gallon




drums  and  by  highway   truck  tankers.    Ethylene  oxide  is  stored  in  bulk




containers,   as  well  as  in  smaller  quantities  in  55-gallon  drums.    No




information was  available  on  the usual emission controls  used  on storage and




transport containers.    "Padded"  containers,  if used,  would  conserve  vapors




which would otherwise be vented to  the atmosphere.   Emissions could also occur




during  equipment purging  in  routine  maintenance,  gauge  glass  blowdown,  or




leaks.




     Release is  also  possible during  transfer.   In  normal  practice,  railway




tankers are  loaded  and  unloaded  directly from  or into  storage  tanks.   The




chemical  is  transferred  under nitrogen pressurization (-50  psi)  for pumping.




Faulty equipment or over-pressurization could cause epoxide  emissions.   Small




amounts could be spilled  during  handling as well.
                                     5-22

-------
     One concern  in  addition  to normal working  and  handling losses is release




from a storage container or transport-related  accident.   This could range from



a relatively  minor  incident,  such  as  release through a  pressure safety valve




or a rupture disc, to a major accident  in  which  an entire storage container or




tanker ruptures.  No information was available  to predict how often the minor




release accidents occur or the amount of ethylene  oxide they annually release.




     Storage,  transport,  and  handling  methods have  been  extensively described




in  literature  supplied  by  manufacturers  (BASF  Wyandotte  Corp.,   1972;  Dow




Chemical Company,  1977;  Jefferson  Chemical  Company,  undated a  and b;  Oxirane




Corporation,  undated)  and  safety information sources  (NFPA,  1975;  MCA,  1971).




This literature chiefly concerns safety of humans and property.  Tank  cars for




ethylene oxide  and  propylene  oxide are  specified  as ICC-105A100W and  105A100.




These  are  equipped  with pressure  relief valves which vent excessive  pressure




into the atmosphere.   The epoxides should  be stored in  an  area detached from




the plant  site  and  storage tanks should be diked.  Ethylene  oxide  tanks  should




be  equipped  with  cooling  pipes.   Tanks must be equipped with pressure  relief




valves, but  specific  instructions  on  emission control of excess pressure was




not  included.    Vapor  recompression  systems  could be  applied   to  prevent




emissions  (Spencer,  1971).








5.3.3.   Potential  Environmental  Formation.    The  major source  of potential




inadvertent  production of  ethylene oxide  in  the environment  is probably the




combustion of  hydrocarbon  fuels.    Hughes  et  al.  (1959)  used  gas-liquid




partition  chromatography   to  separate  and  identify  oxygenated  derivatives  of




hydrocarbons  that were found in  the  combustion  products  of  hydrocarbon  fuels.




Among  the  oxygenated  combustion products  identified were  ethylene oxide and




                                      5-23

-------
propylene oxide.   Barnard  and  Lee  (1972)  also identified  these  compounds in




the  oxygenated  products of  n-pentane  combustion.   Seizinger  and  Dimitriades




(1972)  suggested  that  ethylene  oxide  is  a  component  of  automobile  exhaust.




They  tested  the   combustion of   simple  unleaded  hydrocarbon  components  of




gasoline.   Stationary  sources of hydrocarbon combustion also  might emit large




quantities of these compounds into the environment.




     Ethylene oxide has been identified in tobacco smoke  (Binder  and Lindner,




1972;  Binder,  1974).    It  is not uncommon  for  tobacco  to  be treated  with




ethylene oxide by  cigarette  manufacturers  for its  fumigant properties.  Binder




and  Lindner  (1972)  determined  that  the  ethylene  oxide  concentration  of




unfumigated tobacco smoke was 0.02 jig/m&,  while  fumigated  tobacco  smoke had a




concentration  of  0.05  ng/m£ and  extensively  fumigated  tobacco  smoke had  a




concentration of 0.30  jig/mjl.  Binder  (1974)  determined  that the ethylene oxide




content of smoke from unfumigated tobacco was 1 ng/g.




     Epoxides  are  formed  in the  photochemical  smog  cycle.   Olefins  can be




converted to  the  corresponding epoxides by  reaction  with  an  organic peroxide




(Altshuller and Bufalini,   1965).   Alkyl peroxides  can decompose  to  yield an




epoxide and oxy radical  (NAS, 1976).




     Water  disinfection procedures might convert olefins  to  epoxides.  Olefin




conversion  during  chlorination  of potable  water would  proceed  by  the  same




route as  for  chlorohydrination  production  of the  epoxide  (see Section 3-6.5).




However,  this  process  would  require  the  conversion  of  ethylene  to  the




chlorohydrin  (Morris,  1975;  Carlson and Caple,  1977).  Since ethylene is very




volatile, it  seems unlikely  that  ethylene  remains  in water  long  enough  for




this process to occur significantly.
                                      5-24

-------
5.4. SUMMARY



     This  section  discusses production, uses,  and  emission of ethylene  oxide.




Ethylene oxide  is  produced almost exclusively  by  direct oxidation of ethylene



using  either air  or  oxygen.    Its  1981  production  volume was  4937 million




pounds, down from 5220 million pounds in 1980.




     The  major  emission  sources  from  production  facilities  are   the main




process  vent for both  air and  oxygen  units and  the purge  gas  vent  for air




units;  fugitive  emissions  are  also  a  source  of  ethylene  oxide  in  the




atmosphere, although no emission  estimate  is  available for this source.  Total




air  emissions  from  production  have been estimated  to  be  around  2 million




pounds  based on  1978  production  volume.   Ethylene  oxide  also enters  the




atmosphere  from handling,  storage,  and  transfer  operations,  as well  as the




disposal of process wastes.    There is  no  solid waste  from  ethylene  oxide



manufacture.




     More  than  90%  of the  ethylene oxide   produced  is  used  captively  as  a




chemical  intermediate,  where  there   is  some  potential  for   environmental




contamination.  Up  to  10  million pounds are used  annually for fumigation and




sterilization;  ethylene oxide  emissions from such  uses might  be significant.




Ethylene  oxide  may  also  be   produced  by  hydrocarbon  combustion   (e.g.,



automobile  exhaust).
                                     5-25

-------
               6.  ENVIRONMENTAL  FATE, TRANSPORT, AND  DISTRIBUTION









6.1.  INTRODUCTION




     Epoxides  are not  persistent in the environment.  Available information  on




their  chemical and  biological  properties characterizes  them  as  highly  reac-




tive.   The  available  information was  not  sufficient  to  develop  a  definite




description  of  their  environmental   transport  characteristics.     Interphase




transport  from water  to  air seems  to be a  slow process,  but evaporation  of




ethylene  oxide applied as  a sterilant  or a  fumigant appears  to  be  a  rapid




process.   High water solubility and  high  vapor  pressure result in  significant




mobility within water or air.




     Epoxide  degradation  has been  fairly well  characterized,  and indicates




that ethylene  oxide is reactive  in all media.   Available  information on its




ionic  reactions  indicates   that  chemical  (see  Section   3)  and   biological




degradation produce  the  same degradation products.   Its  degradation in water,




soil, commodities,  and  manufactured products  proceeds through ionic  reactions.




Degradation in  the  atmosphere has not been well-characterized with respect  to




processes  or products.   Available information indicates  that it is very  reac-




tive in photochemical  smog cycle  reactions.   No information  was  available  on




whether ionic   reactions  (e.g.,  with water  vapor or  water  within aerosols)




significantly contributes to  its degradation in the atmosphere.








6.2. ETHYLENE OXIDE FATE IN WATER




     Ethylene  oxide  degrades in water  by hydrolysis  and  related nucleophilic




reactions;  aqueous   radical   reactions  are not  a  significant  process.    The




                                      5-1

-------
hydrolysis  chemistry  of ethylene oxide  has  been discussed  in  Section  3.6.5,



and  the information presented there  will  be used  in  the present discussion.



     Ethylene  oxide  has a  hydrolysis  half-life  of 12.2  days  in  pure  water,



 12.9 days in filtered (0.22 \m filtered)  Kanawha River water, and 14.2 days in



 unfiltered  Kanawha River water (Conway et al. , 1983).  The Kanawha River water



had  a  pH of 7.4  and  the  initial ethylene  oxide concentration was  =^70  mg/&.



 These  variations  in  hydrolysis  rates  are   well  within  the error  limits of



 hydrolysis  experiments discussed  by  Mabey and Mill (1978).



     It  is   interesting  to  note  that the  presumed  presence  of a  microbial



population  in the unfiltered river  water  did  not decrease  the half-life  of the



 ethylene  oxide.   Although  the microbial concentration  was  not  reported,  the



 lack of a significant change  in  degradation  rate  may indicate  that biological



reactions are not significant in  river water.  Also,  it should be noted that a



half-life of  12-14 days allows  for exposure of biota and possibly  humans to



ethylene  oxide,    although the  addition of hypochlorite  in  water  treatment



plants reduces  the  likelihood of human  exposure,   Conway  et al.  (1983)  also



reported  that  pH  variations would have less  of an  effect on  the rate  of



hydrolysis  than temperature over  a pH range of 5-10.



     Evaporation from water also appears to  be  a significant removal process.



Conway et al.  (1983)  reported the calculated relative  desorption coefficient



a,  (a, =  K. (ethylene oxide)/K,( 00),  K, is  the desorotion coefficient)  to be
 d   d     a                   a   2    a


0.31, 0.34,  and 0.36  for 10, 20,  and 30° C water.  Experimental values for 22° C



water of 0.36  for no  wind  and 0.39  for a 5  m/s  wind are reasonably consistent



with a calculated value of 0.34  and  may  be the  result of  increased  turbulence



and  wind  flow.   These  values  of  a,  indicate  that ethylene  oxide will be



desorbed from  a  body  of water with a  rate  dependent upon the  actual oxygen-



                                     6-2

-------
transfer rate  in  a specific system.  The  rate  of  desorption will be less than



that for  volatile low solubility  organic  compounds such  as  toluene,  benzene,



and chloroform, which have an a  of =0.65  (Rathburn and Tai, 1981).



     Conway  et al.  (1983)  also  measured the  BOD  using  2  mi of  domestic



sewage/BOD  bottle.   They found  biooxidation  was  5,  22,  40,  and  52%  (of



theoretical) on days 5,  10,  15,  and 20, respectively.  They  suggested that in



a  sewage  treatment  plant,  where  the  microbial  population  is  much  higher,



biodegradation might be very fast;  however, from their data it is not possible



to  determine  whether  the  chemical  degraded   is  actually  ethylene  oxide  or



whether  it is ethylene  glycol  (from hydrolysis),  since  the  hydrolysis  half-



life (=14  days) is similar to the BOD half-life of slightly less  than 20 days.



     Hendry  et al.  (1974)  reported the rate  constant  for the reaction  of one



epoxide with alkyloxy radical  proceeding  by a-hydrogen abstraction to be 8.5 x



10  M~   s~ /a-hydrogen,  or  3.4  x  10   M~  s~   for  ethylene oxide.   Given an


                                                       -14
alkyloxy radical  concentration in  ambient  water of 10     M,  the half-life for



this process  is  -6 years.   Hence,  hydrolysis  and  evaporation  appear to be the



dominant fate  processes  for ethylene oxide,  while no  definitive statement can



be made regarding its biodegradation.







6.3.  ETHYLENE OXIDE FATE IN SOIL



     Pertinent data  regarding chemical degradation  of ethylene  oxide in soil



were not located  in  the  available  literature.   It seems  reasonable,  given the



composition  of soil,  that  the half-life of ethylene  oxide would be shorter in



soil than  in water.
                                      6-3

-------
6.4.  ETHYLENE OXIDE FATE IN THE ATMOSPHERE

     Little  direct  information on  epoxide  behavior  in  the  atmosphere  was

available;  however,  some  characteristics  of  ethylene  oxide  behavior can  be

inferred from the data on free-radical chemistry.

     The  atmospheric  reactivity  of  volatile  organic   chemicals   has   been

characterized by  their relative reaction  rates  with hydroxyl  radicals in the

gas phase (Cupitt, 1980;  Darnall et al., 1976);  however,  there  are a number of

difficulties  in  determining an atmospheric half-life or  lifetime for ethylene

oxide based on  hydroxyl  radical  reactions.   One  important  difficulty  is  in

choosing the  appropriate  hydroxyl  radical  concentration:  a number of different

modeling and  direct  measurement efforts have  provided  a wide  range  of values

for both average  and  altitude  specific  concentrations of hydroxyl radicals.  A

reasonable  compromise for an  average  OH  concentration  is  1  x  10   molecules

cm   based  on more recent modeling efforts  (Cupitt,  1983).   For  ground level,

the concentration may be  somewhat higher, possibly  around 1.3  to  1.4  x  10

molecules  cm   during the summer  (Crutzen and Fishman,  1977;  Logan  et al.,

1981).   Using these  two  values,   a  temperature of 300  K,  and  the  Arrhenius

equation of Fritz et al.   (1982) (see Section  3.6.5),  the lifetime of ethylene

oxide   is  found   to  lie  somewhere  between  100 days (using the upper  limit of

the Arrhenius equation)  and 215 days (the  lower limit)  for a hydroxyl radical
                         r              _-j
concentration of  1 x  10   molecules  cm,  and  between  74  days  and  159 days

(using  the  upper  limit)  for  a hydroxyl radical  concentration of  1.35  x 10

molecules  cm.   This lifetime is  in  sharp  contrast  with  the  significantly

shorter  values  found for  other ethers.   For  example,  tetrahydrofuran,  a five

membered  cyclic   ether,  has  a lifetime  of   =1  day.    Fritz  et  al.   (1982)

suggested  that  this  disparity is  due to  the  distorted sp   bonds  in ethylene

                                      6-4

-------
oxide  that  give  rise  to  a  hydrogen-abstraction  activation  energy  of 5.8


kcal/raol, higher  than  the standard 2.8 kcal/mol.

     Bogan and  Hand  (1978)  found the absolute rate constant of the reaction of

                   —                                                          1 A
oxygen  atoms   [0(-*P)]   for  ethylene   oxide   to  be  (6.3  +  0.18)   x   10~


cra3/mcule-sec at  300 K.   This rate  is  several orders of magnitude slower than


that for  the  hydroxyl  radical  reaction,  and yields  a  half-life  of 1400 years,

                        -3                             h             o
given an  atmospheric 0(JP)  concentration of  2.5  x 10  molecules/cm   (Graedel,


1978).

                                                                              P
     Sickles  et al.  (1980)  measured the rate  of ozone production in  a Teflon


smog  chamber  to  rank  19 compounds relative  to propane.   The  chambers  were


irradiated with sunlight outdoors.   Purified air,  the  organic  compound to be


tested and N02  were  added before sunrise  to multiple chambers; ethylene oxide-


to-NC>2 ratio  at the onset of  the experiment was 4:0.067.   Ethylene  oxide was


much less  reactive than propane.   The  rank  order  of reactivities found  was:


Acrylonitrile   >perchloroethylene  >ethanol  >ethylacetate  >acetone  >methanol


>acetic acid  >propane >ethylene dichloride  >acetylene  >chloroforra  >dimethyl


formamide >benzaldehyde  >methylene  chloride >pyridine  >*ethylene  oxide >methyl


chloroform >phenol >acetonitrile >nitrobenzene.  The relative  ordering of  com-


pounds was  similar  in an indoor smog chamber  study (Dimitriades and  Joshi,


1977).  The smog  chamber study  of Joshi et al.  (1982) also concluded  the low


reactivity of  ethylene  oxide  (half-life  >53 hours).    All  of  these  results


indicate   that  ethylene   oxide   is  relatively  unreactive  in  the atmosphere

compared  to other ethers.


     With  the information currently available, no definitive  statement  can  be


made regarding  the atmospheric  fate or lifetime of ethylene oxide.
                                      6-5

-------
6.5.  DEGRADATION IN COMMODITIES AND MANUFACTURED PRODUCTS




     Ethylene oxide  is  registered  in the United  States  for use as  a fumigant




or  sterilant for  several  stored-food  commodities  and   manufactured products




(Goncarlovs,  1983).    These  include its use  as  a  fumigant  for   bulk  food




containers,  other   food   containers,   stored  grain,   stored  fruits,  stored




processed  foods,  tobacco  products,  garments, furs,  stored herbs and spices,




furniture, aircraft, buses,  railroad  cars,  and   laboratory animal bedding.  As




a sterilant, it  is used  principally on  hospital equipment and Pharmaceuticals.




Ethylene  oxide  is  used as a fumigant chiefly to  protect stored products from




insect or microbial destruction.  The fate  of this epoxide and its residue are




especially important in materials,  commodities,  and products coming  into close




contact  with humans,  such as  surgical   equipment,  Pharmaceuticals,  and  food




service  and  packaging  materials (Wesley  et al.,  1965; Alguire,  1973; Holmgren




et al., 1969; Gilmour,  1978).




     The  study  of  the   fate   of   ethylene  oxide  in   these  materials  has




established  that it  will  degrade to glycol and  halohydrin or evaporate.   The




degradation  results  from  chemical  and/or enzymatic activity.   The  halohydrin




route  requires  epoxide  reaction with inorganic  halide.    The halide  could  be



naturally  present,  added  or derived  from organic  halides.   Bromide  ion  is




often  supplied  by  degraded methyl  bromide, another fumigant  (Rowlands,  1971;



Lindgren et al., 1968).




     Scudamore  and  Heuser   (1971)   measured  the  apparent  degradation  and




evaporation  of  ethylene  oxide   and  its  residues,  ethylene  chlorohydrin  and




ethylene  bromohydrin,  over  a  1-year  period.   Apparent  first-order specific




rate  constants,  k,   were  calculated  for  epoxide  dissipation.     The  rate
                                      5-6

-------
constant,  k,  combined  losses  from  the  degradation  (chemical  and metabolic




pathways), k-, and evaporation, k..:
                    k = kD + kv
The glycols  (ethylene  and  diethylene)  were measured once at 6 months or  1 year




after  treatment.    The  parameters  considered   included   the  ethylene  oxide




treatment  (dose  and temperature  during application),  the  moisture content of




the  commodity,   storage  temperature  and  type  of  storage  (closed containers




versus  open  trays).   Ethylene oxide  residues  dissipated  rapidly.   While the




estimated  half-life  was  longest  at  10°C  in   sealed containers,  it  never




exceeded  2 weeks.    Increasing the ethylene  oxide  dose did not  have  a  simple




effect  on its  loss rate.   For  the  most  part,  small  increases  in  the dose




slightly  decreased   the  loss  rate, while  very  large  increases  caused  larger




decreases  in the rate  of loss and, sometimes, non-linear  changes.   The  effect




of  moisture  content appeared  varied  and relatively  small.    Scudamore  and




Heuser  (1971)  also  monitored  some commercially  treated  products and  found




ethylene  halohydrin  residues but  no  ethylene oxide residues.   They concluded




that  ethylene  oxide will  usually dissipate  from  treated commodities,  but,




under some circumstances, small quantities could persist for several months.




     Stijve  et  al.  (1976)   discussed  the fate of ethylene  oxide applied  as  a




fumigant  to  commodities.  They suggested that ethylene oxide could be retained




by physical  adsorption,  but that  it  would persist  not more than  a  few weeks




before volatilization or reaction with natural constituents of the commodity.




     Ben-Yehoshua et al. (1971)  examined  ethylene  oxide  residues  during  the




treatment  of dates.   They  reported a small  ethylene  oxide loss in  the empty




                                      6-7

-------
container  and ascribed  this to  apparent adsorption  to  the  container walls.




The  larger losses  found  with 2.1  kg  of dates  in the container  were  due to



ethylene oxide uptake  by  the fruit.  Ethylene oxide loss in treated dates  left




in  open containers  was  attributed to  degradation (to  the  chlorohydrin  and




glycol) and volatilization.




     The  available  information  on the  fate  of  ethylene oxide  applied to




manufactured goods was less  extensive as  that  on  its  fate in commodities.  All




available   information  suggests  behavior  similar   to   that  discovered  in




commodities.     Alguire   (1973)   described  losses  of   ethylene   oxide   from




polystyrene creamer  cups  and cream cheese wrappers at ambient temperature and




open  to  the  environment.    The  ethylene  oxide  did  not  degrade  on  the




polystyrene  cups,  and  was  lost solely  through  out-gassing.    More  than  90%




evaporated by the  first  day, and no  residual  ethylene oxide  remained  after 5




days.   Ethylene  oxide loss  from cream cheese wrappers consisted  primarily of




conversion  to  ethylene glycol;  no  ethylene chlorohydrin was  detected  at  any




time.  Ethylene oxide was completely gone by the tenth day.




     Some  studies  have identified  ethylene  chlorohydrin  residues  in  manufac-




tured goods  sterilized with  ethylene  oxide.  These  studies did not  seek  any



information on volatilization  losses.   Brown  (1970) identified  ethylene oxide



and  its derivatives  on  treated  equipment  made  of rubber, Dacron,  and poly-




vinylchloride, but  did  not  detect  chlorohydrin  on  polyethylene  equipment.




Holmgren et al.  (1969) measured 0-1500 ppm  chlorohydrin  on 21 ethylene oxide-




treated drugs.
                                      6-8

-------
6.6.  BIOACCUMULATION IN AQUATIC ORGANISMS



     Specific   experimental   information   regarding  the   b ioac cumulation   of



ethylene  oxide  in aquatic organisms  is not  available.   Veith et al.  (1979)



have suggested the calculation of BCF from the following  equation:
               log BCF = 0.76 log K   -0.23
                                   \J W
where K    is  the partition coefficient between octanol and  water.   Using  this
       o w


equation and  the  log K   of  -0.30,  reported  by Hansch and Leo  (1979),  the BCF



for whole fish was calculated to be 0.34.
6.7. SUMMARY



     This  section  discusses  the results  of studies  related to  the fate  of



ethylene oxide  in  the  environment.   In water,  ethylene  oxide will degrade  by



hydrolysis and  related  nucleophilic  reactions with a half-life of  =12-14  days



at 298 K.  Lower  temperatures  lengthen the half-life; pH changes have minimal



effects.   Volatilization  will also  be a  significant process although less  so



than  for  sparingly  soluble compounds like  toluene,  chloroform  or  benzene.



There  is  no  conclusive evidence  that microbial  degradation is  significant;



however, the  biological components  of  sewage  sludge  might  react rapidly  with



ethylene oxide.   The fate  of  ethylene oxide in soil will probably be similar



to that in water;  its half-life will probably be shorter.



     The  fate  of  ethylene  oxide in  the  atmosphere  is  not  obvious  from  the



information  present  in  the  literature.    Rate  constants  are  available  for



hydroxyl radical  and oxygen  atom  [0(  P)]  reactions as well as smog chamber
                                      6-9

-------
studies.  All  predict  that ethylene  oxide will persist in  the atmosphere,  but




the actual lifetime cannot be predicted.




     In commodities,  food containers,  and  manufactured  goods,  ethylene  oxide




appears to volatilize or to hydrolyze to glycol or halohydrin with a half-life




of -2 weeks.
                                     6-10

-------
                     7.  ENVIRONMENTAL LEVELS AND EXPOSURE








7.1. INTRODUCTION




     The purpose of this document  is  to  present available information relevant




to human health effects that could be caused by this substance.




     Any information  regarding  sources,  emissions,  ambient air concentrations,




and public  exposure  has been  included  only to give  the reader  a preliminary




indication  of the potential  presence of  this substance  in  the  ambient  air.




While the available information  is presented as accurately  as possible,  it is




acknowledged  to be  limited  and  dependent in  many  instances on  assumption




rather than specific  data.   This information is not  intended,  nor  should  it be




used to support any conclusions regarding risks to public health.




     If a  review of  the health  information indicates  that the  Agency  should




consider regulatory action  for this  substance,  a considerable  effort will be




undertaken to obtain appropriate information regarding  sources,  emissions,  and




ambient air concentrations.  Such  data will provide  additional information for




drawing regulatory conclusions regarding the extent  and  significance  of  public




exposure to this substance.








7.2.  ENVIRONMENTAL LEVELS




     Ambient monitoring has portrayed ethylene  oxide as an almost non-existent




contaminant of  environmental  or biological  samples.   Although  ethylene  oxide




is rarely identified  in monitoring studies,  its  principal degradation products




(glycols and halohydrins) have been found.
                                      7-1

-------
     No  monitoring  data  were   available   for   ethylene  oxide  in  biological


tissues,  except  for  some  tissue-distribution   studies.    Since  epoxides  are


reactive  alkylating  agents,  it  is  reasonable  to  expect that  ring-opening


reactions  will  occur  rapidly  in biological  systems  so  that  the   finding  of


detectable levels in environmental biota is unlikely (Anderson, 1971).


     Only  one ambient  air monitoring study reporting  the  presence  of ethylene


oxide  in  air was  found in the  available   literature.   Bertsch et  al.  (1974)


tentatively  identified  ethylene  oxide in  the  ambient air  near  the  University


of Houston.  However,  the  authors used  Tenax  as the adsorbant for trapping air


contaminants and its  use casts  doubt on their  tentative  identification, since


Tenax does not adequately retain ethylene oxide  (see Section 4.1).


     U.S.  EPA  (1976)  listed one monitoring observation  of ethylene  oxide  in


water.  It was  observed in the  effluent from  a chemical  plant in Brandenburg,


Kentucky.   No  other epoxide observation was  reported.   U.S.  EPA  (1976)  also


noted  observations  of ethylene  halohydrin, which  might have  been  released  in


industrial wastes as such, rather than occurring as residues from epoxide.


     No other  reports  of  ethylene  oxide  in ambient  air  or water  were found,


yet  Systems  Application,   Inc.   (1982)  reported   that  the  maximum  possible


exposure  concentration level of ethylene  oxide,  based on  dispersion models,

            •3
was  5  |ig/m   (2.77  ppb).    The  justification  for this  value  could  be  the


reactivity  of  ethylene oxide or  the lack  of an adequate  sampling  method (see


Section 4).  Most sampling  methods  either  lose  significant amounts  of ethylene


oxide  on  even  short-term storage,  or use  adsorbants  with  a very poor affinity


for ethylene oxide  (e.g.,  Tenax GC); however,  this is certainly not the case


for   all   studies,  especially   those  using   freeze-out   techniques.    Since


environmental  sample  concentrations  are  rarely as high  as  workplace  sample


                                      7-2

-------
concentrations  (particularly  in  the  case  of  a  reactive  molecule  such  as




ethylene oxide),  the well documented  NIOSH  (1977) method  becomes inadequate.



The  problem  is  compounded  by  the  fact  that   few  monitoring  studies  are




undertaken  to identify  only  a single  compound   in  the  environment.    These




studies must  assume  some compromise between completeness  and speed,  making it




impossible to optimize conditions for the detection of any one compound.




     Several  studies have  examined the  residues of  ethylene oxide  that  has




been  applied  to  commodities  and manufactured  goods  as  a  fumigant  and




disinfectant.    Another  portion   of   this   report  (Section  6.5)  describes




investigations on  the fate  of this epoxide.   The  information here concerns




residues in actual commercial  products.




     Scudamore and Heuser (1971) evaluated  ethylene oxide  and its metabolites




in commercially treated  products.   While  they  never  detected ethylene oxide in




commercial products,  they did  find  ethylene  chlorohydrin residues  ranging from




10-70 ppm.  Lindgren  et  al.  (1968)  reviewed studies on  residues  from ethylene




oxide  treatment,  most   of  which   were   fate  studies   rather  than  ambient




monitoring studies.   They suggested that residual epoxide  could be  present in




commercial products.




     Ethylene oxide  is a common sterilant for  surgical equipment.   Its fate in




plastic and  rubber surgical equipment parallels  its behavior  in  commodities.




Brown (1970)  monitored  residues on various  hospital equipment sterilized with




ethylene oxide.   Ethylene oxide was  observed  in  three  samples,  one  of which




had  received  treatment  =80  days  previously.    Ethylene  chlorohydrin  was




detected in 10 samples.
                                      7-3

-------
7.3.  EXPOSURE




     The  available  data concerning the  environmental  levels of ethylene  oxide




are  insufficient to  properly  estimate  exposure;  however,   a  general overview




can  be made.    Over  5 billion  pounds  (>2  billion kg)  of ethylene  oxide  is




produced  yearly.    The  vast  majority  is  used  captively  as  a  synthetic




intermediate.    Perhaps  10  million  pounds  (4.5 million kg)   is  used  for




fumigation/sterilization  for  products that  include food  commodities,  medical




devices,  Pharmaceuticals,  and  cosmetics.     This use  constitutes  the  only




documented  potential  exposure  to ethylene  oxide,  though  the extent  of this




exposure  must  be determined.    Ethylene  oxide also appears  to  be  a product  of




incomplete  combustion,  and  has   been  identified  in  automobile  and  diesel




exhaust and  in  tobacco smoke.   It  can be formed during the photochemical smog



cycle, but appears to be rapidly destroyed.








7.4. SUMMARY




     This  section  discusses  the  results of  monitoring  studies  conducted   to




measure the  levels  of  pollutants,  including  ethylene  oxide,  in  the environ-




ment.    Very  little  information  is  available  on  ambient  monitoring,   no



confirmed detection of  ethylene oxide in air  has  been reported,  and  only one




report exists  for  water.    The  lack of more monitoring reports  may be because




most, but not all,  sampling methods would not  detect  ethylene  oxide even if  it




were  present.    Several  studies  have examined  the  persistence  and  fate   of




ethylene  oxide  in commodities  and  commercial  goods  including food,  medical




supplies,  and drugs.
                                      7-4

-------
                             8.  ECOLOGICAL EFFECTS








 8.1. MICROORGANISMS  AND  INSECTS




     Ethylene  oxide  is used as  a  fumigant for foods  (particularly grains)  and




 spices,  and  shows  major  microbial,   insecticidal,   and  acaricidal  activity




 (Sykes,  1964;  Lindgren and  Vincent,  1966).




     Fumigation  with ethylene  oxide has been used to  control a  wide variety of




 bacteria, fungi,  rickettsiae and viruses.   Sykes (1964), for example, reported




 that exposures to gaseous  ethylene  oxide  at  concentrations of  1-10$ will kill




 Bacillus  globigii,  Staphylococcus  aureus,  Escherichia  coli,   Chromobacterium




 prodigiosum,  and  Mycobacterium  phlei  within  a few  hours.    Roberts  et  al.




 (19^3)  found  that 10$ gaseous ethylene oxide  will  kill Bacillus anthracoides




 in  8  hours.   Ethylene  oxide   also  has  significant  sporicidal activity against




 dry  bacterial  spores  (Bruch  and  Koesterer,  1961).    Exposure of  Bacillus




 subtilis spores  to 1-2$  vapor concentrations  of  ethylene oxide killed j>95$  of




 the  spores  within  4 hours.   A  5$  gaseous  concentration of  ethylene  oxide




 produced 90$  kill of airborne B.  globigii  spores in <2 hours (Roberts et al.,




 19^3).   Treatment of agar  slants  containing  yeasts  and  fungi  with  8$ gaseous




 ethylene oxide for 3 hours  was lethal  to these microorganisms (Whelton et al.,




 1946).    Skeehan  (1959)  indicated  that herpes  simplex,  vaccinia, and  bovine




 respiratory  viruses  are  susceptible  to   saturated  ethylene  oxide  vapor




 treatment.




     Susceptible  insects  common  to  stored products  include  the flour beetle,




 rice weevil,  and grain weevil (Lindgren  et al., 1954).    Ethylene oxide  will




kill one-half  of  the  stored-product  insect population  at a concentration range




                                      8-1

-------
of  6-18 mg/&  (Ong,   1948)..   Lindgren  and  Vincent  (1966)  reported  a  major




reduction in available tissue  glutathione  content  of  Calliphora larvae exposed




to ethylene oxide.  Decrease  in tissue glutathione via depletion of reduced-SH




groups  may  be  the  mechanism  of toxicity.   The  insect  toxicity  of ethylene




oxide  has  been  ranked  by  these  authors as  intermediate  between  those  of




ethylene dibromide and' ethylene dichloride.   A  bibliography  of ethylene oxide




insecticidal properties,  citing 185 references, has  been  published (Young and




Busbey, 1935).








8.2. PLANTS




     Pertinent  data  regarding  the  effects of  exposure  to ambient  levels  of




ethylene  oxide  on plants  were  not  found in  the available  literature.   As




detailed in  Section  9.4, ethylene  oxide is  capable  of  inducing mutations and




chromosomal aberrations in plants.








8.3. AQUATIC ORGANISMS




     Limited  information is  available on the  toxicity  of ethylene  oxide  to




aquatic  organisms.    The  acute  toxicity  of  ethylene   oxide   appears   to  be




moderate, as  indicated  by LC^s  in the range  of  84-90  mg/Jl for  fish, a mean




48-hour LC5Q  of  212  mg/&  for  Daphnia and 745 mg/Jl  for  brine  shrimp  (Table




8-1).   kCj-Q  values  for  the  hydrolysis product  ethylene  glycol  were >10,000




mg/& for the above species  except goldfish (which were not tested) (Conway et




al.,  1983).   If  reacted to  form  ethylene chlorohydrin,  the 96-hour LC,-0 for




fathead minnows was =90 mg/Jl (Conway et al.,  1983).
                                      8-2

-------
                                                      TABLE 8-1
                                      Acute Aquatic Toxicity of Ethylene Oxidec
LCc;n (95? Confidence limits), mg/£
Test Procedure
range-finding , static, aerated

range-finding , static, sealed
under oxygen
definitive static acute (no
(X> aeration)
static acute

static acute


static acute


Test Organism 24 hr
fathead minnow 274 (150-500)

fathead minnow 86 (50-150)

fathead minnow 90 (63-125)

goldfish 90

Daphnia magna >300
270
260
brine shrimp >500
350
570
48 hr
NA

NA

89 (63-125)

NA

300
137 (83-179)
200 (150-243)
>500
1000
490
96 hr
NA

NA

84 (73-96)

NA

NA
NA
NA
NA
NA
NA
Reference
Conway
et al. ,
Conway
et al. ,
Conway
et al. ,
Bridie
et al. ,
Conway
et al. ,

Conway
et al . ,


1983

1983

1983

1979

1983


1983

 Source:  Conway et al., 1983
 Range-finding tests used 2 fish/test concentration
Definitive tests used 10 fish/test concentration
NA = Not applicable

-------
                   9.  BIOLOGICAL EFFECTS IN ANIMALS AND MAN








9.1.  PHARMACOKINETICS








9.1.1.   Absorption.    Only limited  data  regarding the  absorption  of ethylene



oxide were  found  in the  available  literature.   However,  acute  toxicity data



suggest  that  absorption  occurs  readily  via  the respiratory  and gastrointes-




tinal tracts (Table 9-1).








9.1.2.    Distribution.   Information  concerning  the distribution  of ethylene



oxide in the body is limited.  Two studies have shown that  it  is found in many



tissues following inhalation exposure or intravenous administration.



     Ehrenberg et al.  (1974)  conducted inhalation  studies  with radioactively-



labeled  [1,2- H]  ethylene oxide.   Following  exposure  of mice  to  1.15 ppm of



the  labeled  chemical  in  air  for 75 minutes,  the  highest  levels  of radioac-



tivity  (in  unidentified chemical form) were associated  with  proteins isolated



from  the lungs,  kidneys,  and  liver.    Lower  levels   of  radioactivity  were



measured  in the  testes,  brain,  and  spleen,  but  additional  organs  were  not



analyzed.



     Appelgren  et  al.  (1977)  carried  out  whole-body autoradiography  on mice


                                           14
injected  intravenously with radioactive  [  C]-ethylene  oxide  (label position



unspecified).    Preliminary inhalation  studies  with  labeled  ethylene  oxide



showed  a similar tissue distribution  of the  compound  similar  to  that  seen



following intravenous  injection, except  for  a high initial  labeling  of  the




respiratory mucosa  (data not  shown).   Two  minutes after the  injections,  con-



centrations of  radioactivity  2-3 times  those  seen in the  blood were observed




                                      9-1

-------
ru
                                                        TABLE 9-1




                                             Acute Toxicity of Ethylene Oxide
Route
oral
oral
oral
oral
oral
ihl.
ihl.
ihl.
ihl.
ihl.
i .v.
i .v.
i.p.
i.p.
i.p.
s.c.
Species
rat
rat
rat
guinea pig
rabbit
rat
rat
guinea pig
mouse
dog
rabbit
rat
rat
mouse
rabbit
rabbit
Sex
M
M
M
M,F
M,F
M
M,F
NR
F
M
M,F
M
M,F
M,F
M,F
M,F
Strain
Wistar
NR
NR
NR
NR
white
Sherman
NR
white
beagle
NR
NR
NR
NR
NR
NR
Dose
330 mg/kg
100 mg/kg
200 mg/kg
270 mg/kg
631 mg/kg
1460 ppm/4 hours
4000 ppm/4 hours
7000 ppm/2.5 hours
835 ppm/4 hours
960 ppm/4 hours
178 mg/kg
355 mg/kg
178 mg/kg
178 mg/kg
251 mg/kg
200 mg/kg
Response
LD50
0/5 died
5/5 died
LD50
LD50
LC50
LC50
LClow
LC50
LC50
LD50
LD50
LD50
LD50
LD50
LD50
Reference
Smyth et al., 1941
Hollingsworth et al.,
Hollingsworth et al.,
Smyth et al., 1941
Woodward and Woodward,
Jacobson et al., 1956
Carpenter et al., 1949
Waite et al., 1930
Jacobson et al., 1956
Jacobson et al., 1956
Woodward and Woodward,
Bruch, 1973
Bruch, 1973
Bruch, 1973
Woodward and Woodward,
Woodward and Woodward,


1956
1956

1971





1971



1971
1971
   Ihl. = inhalation; i.v. = intravenous; i.p. = intraperitoneal; s.c. = subcutaneous; NR = not reported

-------
in  the  liver,  kidneys,  and  pancreas.   Tissue  labeling 20  minutes  to  4 hours



after  exposure  showed high  levels of  radioactivity  in  the  liver,  kidneys,



lungs,  intestinal  mucosa,  epididymis,  cerebellum,   and  testes.    Twenty-four



hours after  injection, radioactivity was  still  found  in  the liver,  intestinal



mucosa,  epididymis,   cerebellum,   bronchi,   and  bone  marrow.    Since  these



observations  were  made   on  auto radiographs, quantitative   results  were  not



reported.   The extent  of bioexchange  of the radioactive  label  into  natural



body constituents also could  not be determined in this  study.







9.1.3.   Metabolism.   Comprehensive studies designed  to fully characterize the



metabolic fate of ethylene oxide have not  been conducted.



     Significant concentrations of ethylene glycol were detected  in the plasma



of  four beagle dogs following the  intravenous administration of 25 or 75 mg/kg



ethylene oxide on  separate occasions (Martis et al. , 1982).   Urinary excretion



data  indicated that  7-2^% of the administered  dose was excreted  in  the urine



within 24  hours  as ethylene glycol; the  mean percentages  of the  low and high



doses  that  were excreted in  the  urine  were  13-5 +_  3-5%  and  14.2  + 8.1?,



respectively.


                                                       14
     Two urinary  metabolites were detected  when  [1 ,2- C]  ethylene  oxide was



administered to Sprague-Dawley  rats via a single  intraperitoneal  injection at



a dosage of 2  mg/kg (Jones and Wells,   1981).  The urinary  metabolites were S-



(2-hydroxyethyl)-cysteine  (9%  of  the   dose)  and   N-acetyl-S-(2-hydroxyethyl) -



cysteine  (33%  of  the  dose),  which  suggests that the metabolism  of ethylene



oxide  involved conjugation with glutathione.  A small percentage  of the dose


               14
was exhaled as   C0» and  as unchanged  ethylene oxide  (Section 9.1.4).
                                      9-3

-------
     In  the  inhalation  study  with  mice summarized in Section 9.1.2  (Ehrenberg




et  al. ,  1974),  the only urinary metabolite  characterized  was 7-hydroxyethyl-



guanine,  >*iich accounted  for a  minor amount  (0.007$)  of  the  total urinary




radioactivity.    Significant  alkylation  of  tissue  proteins  was   found,  and




alkylation  of  DNA was  confirmed  by  the identification  of a  high specific




activity  radiolabeled  7-hydroxye thy Iguan ine.    Gumming et  al.  (1981) reported




large differences  in  the patterns  of initial alkylation  as well as removal of




total  alkylation  products  from the  DNA  of various  tissues  (i.e.,  testis,




liver,  lung, kidney, spleen)  of mice following inhalation exposure to tritium-




labeled  ethylene   oxide.     Thus,   ethylene   oxide   distributes  and  reacts



extensively throughout the body.








9.1.4.   Elimination.    In  the inhalation  study  with mice  (Ehrenberg  et al. ,




1974) using  tritium-labeled  [1 ,2-%]-ethylene  oxide (see   Section  9.1.3),  it




was  found  that  78% (mean value)  of the absorbed radioactivity was excreted in




the  urine within  48 hours.   The  biological half-life in mice  was  reported to




be -9 minutes,  indicating rapid urinary  elimination.




     Approximately  43$   of  the  administered  radioactive   dose  of   [1 ,2-  C]




ethylene oxide  (2  mg/kg,  single  injection) was excreted in  the  urine of mice




over 50 hours,  most of which  (=40/5)  appeared within  18 hours of dosing (Jones




and  Wells, 1981).   Two urinary metabolites, S-(2-hy drox yeth y 1)-cy ste ine  and N-




acetyl-^-(2-hydroxyethyl)-cysteine  accounted   for  9  and  33%  of  the  dose,




respectively.   Within 6  hours, 1.5$ of the dose was  exhaled as   C0_ and ~\% as




unchanged  ethylene  oxide,   but   these  are  not   maximum  values  (exhaled




radioactivity was not sampled  at  later post-exposure  times).
                                     9-4

-------
     Martis  et al.  (1982)  investigated  the  elimination  kinetics  of   intra-




venously  administered  ethylene  oxide  in  beagle dogs.    Four dogs  received



single 25  and  75 mg/kg  injections of  the  compound  on  separate occasions,  and




venous blood  was sampled for ethylene oxide  and  ethylene glycol at 0,  0.08,




0.25, 0.5,  1.0,  2.0, 3.0, 4.0, 7.0, and 24 hours after administration.   It  was




found that  the ethylene oxide cleared  rapidly from the  plasma,  and  that  in  all




cases concentrations decreased  to  <2% of the zero-time  value  within 5  hours.




The plasma  concentration of  ethylene  oxide declined  exponentially,  and  first-




order rate  constants of 0.025 + 0.006 min~1  and  0.023 +  0.010  rain"   for  the




low  and   high  dosages,   respectively,   were  calculated  from   the   plasma




concentration  corresponded to  plasma  half-lives  of 29.3 + 5.7 min  and  36.5  +




18.5 min.   It was noted  that  the  lack of significant  differences  in  kinetic




parameters  (i.e.,  elimination  rate  constant,   plasma  half-life,  apparent




distribution volume, total  body clearance) at  the  two dose  levels indicates




that  the  elimination  kinetics  are  not dose-dependent.   Ethylene   glycol  was




formed quite   rapidly  following the  administration of  ethylene   oxide,   and




plasma concentrations reportedly exhibited  the characteristics of a metabolite




in a  one-compartment model;  maximum  plasma concentrations of  ethylene  glycol




were reached by 90 + 24.5 minutes (25 mg/kg)   and 120 + 42.4 minutes (75 mg/kg)




post-in ject ion.  Plasma concentration-time data for  ethylene  glycol following



the  intravenous  injection  of  35  and   106  mg/kg  of ethylene  glycol indicated




half-lives of 177.1 + 29.3 and  264.9 + 90 minutes, respectively.
                                     9-5

-------
9.2.  ACUTE, SUBCHRONIC,  AND CHRONIC  TOXICITY








9.2.1.  Effects in Humans.








     9.2.1.1.  ACUTE  EXPOSURE —  Case  reports indicate  that headache,  nausea,




vomiting,  dyspnea, and/or  respiratory  irritation  are common effects of  acute




inhalation  exposure   to  ethylene  oxide  (Greaves-Walker and  Greeson,  1932;




Blackwood and  Erskine, 1938; von  Oettingen,  1939;  Anonymous,  1947;  Sexton and




Henson, 19^9; Hollingsworth et al. ,  1956;  Curme and Johnston,  1952;  Salinas et




al. , 1981).  Symptoms of poisoning have been reported to be  delayed  by  several




hours  following  exposure.   Similar  effects (e.g.,  marked nausea and  profuse




vomiting), as well as mild  leukocytes is and blisters (discussed  subsequently),




developed  in  three chemical plant workers who were  dermally  drenched  with  1?




aqueous ethylene  oxide solution  (Sexton and Henson,  19**9).    Inhalation  expo-




sure  to  high  concentrations of  ethylene  oxide  for brief  periods has  been




associated with bronchitis, pulmonary edema, and emphysema  (Thiess,  1963),  as




well as convulsive movements (Salinas et al. ,  1981).  In a controlled study of




the effects of ethylene  oxide on  human  volunteers,  Greaves-lfelker and  Greeson



(1932)  observed  that  ethylene  oxide at =2200 ppm was  slightly  irritating  to




four subjects.  At a  5-fold higher concentration,  the compound had  a definite




effect on nasal mucosa within =10  seconds.




     Three  chemical  plant  workers  drenched  with   1/E  aqueous ethylene  oxide




solution developed marked nausea and profuse vomiting several hours  following




exposure  (Sexton and  Henson,  19^9).   Large vesiculated  blisters developed  in




the areas of exposed  skin,  and  two workers who had  complete  blood counts  taken




showed a mild leukocyte sis.




                                     9-6

-------
     Cobis (1977) reported a very  low incidence of health-re la ted effects due




to exposure to ethylene  oxide  in  Veteran's  Administration medical facilities.




Ethylene  oxide  was used  for  sterilization  purposes  in  162  hospitals  and  7




outpatient clinics  over an  average  of 8.2 years.    Only  12 employees were




reported  to have been  involved  in  exposure incidents,  and  symptoms included




watering  eyes, nausea, and  skin  irritation.   These  cases are currently being




followed  to determine  possible  exposure sequelae.    The  average  exposure con-




centration was not  given,  and  it  is  presumed  (although  not  stated)  that the




employees were exposed to ethylene  oxide vapor.




     The  derma to logical effects of ethylene  oxide contact have  been reviewed




by Taylor (1977).    Concentrated  ethylene  oxide  evaporates rapidly  from the




skin and  produces a freezing effect,  resulting  in burns  ranging  from   first-




through   third-degree  severity.     Ethylene  oxide  gas  retained  in  porous




materials that have  not been properly aired  can produce skin irritation.  Foot




burns  (Phillips  and fey,  19^9)  and hand burns  (Royce and  Vbore,  1955),  for




example,  have been  observed  in  workers  that  wore  ethylene  oxide-sterilized




rubber boots and rubber gloves, respectively.  Biro et al.   (1974) described a




hospital  incident in which  19  women were burned by  surgical gowns and  drapes




that had  been sterilized with ethylene  oxide.  Joyner  (1964)  found in a 2-year




retrospective  study of  medical records that ethylene  oxide  plant workers had



experienced exposure-re la ted  burns.




     Sexton and  Henson  (19^9)  described   the  derma to logical reactions that




occurred  in 6  men  whose skin was  directly  exposed to a  1/5  water  solution  of




ethylene  oxide  for  periods  ranging  from 15  minutes to 3  hours.   The men with




the maximum exposures  (2-3 hours)  exhibited  the  most marked cutaneous effects
                                      9-7

-------
(vesicular eruptions), but nausea  and  vomiting  were  the only systemic effects



noted.




     In a  subsequent  study,  Sexton and Benson (1950) applied 1-100?  solutions




of ethylene  oxide  to  the  skin of 8 volunteer  subjects  for  tine  intervals that




ranged  from  20  seconds to 95 minutes.   The magnitude  of skin injury appeared




to be  related to  the duration  of contact and  the  concentration.    The  most




hazardous  concentrations  of ethylene  oxide  were  in  the 50%  range,  since the




manifestation arbitrarily  examined  in  this study  (minimal  second-degree burn




demonstrated  as  an area  of  erythema with  one  or  more  superimposed vesicles)




was produced  in 45 seconds with this solution.   The  degree  of skin  injury was




proportionately  decreased  at concentrations  both  greater  and less  than 50%.




The  lowest ethylene  oxide concentration  investigated  (1?)  produced a  mild




reaction (erythema) after 50 minutes of exposure.   The milder  skin reactions




at concentrations >50% were attributed to  the  fact that the more concentrated




solutions  boil  vigorously,  thus  preventing  efficient  skin  penetration;  the




more dilute  solutions  lacked  sufficient  chemical  to  cause  injury  except after




prolonged  contact.     Delayed  skin  sensitization  developed  in   3   of  the  8



subjects.




     Shupack  et  al.   (1981)  demonstrated  that  human  skin  reactions  were




directly related to total dose when  exposures were to ethylene oxide that was




retained in permeable  materials.   In tests with 12 unsensitized  volunteers,  it




was found that patch  materials that  rapidly  lose  ethylene  oxide  (i.e.,  fabric




or rubber)  elicited  few  reactions,  even  at ethylene oxide levels as high  as




3000-5000 ppm after 4-8 hours of contact.   Patch  materials  that  lost ethylene




oxide slowly  produced  mild  skin reactions  (erythema plus  edema)  at material




levels as  low as  1700  ppm (PVC  film) and 1000 ppm (PVC blocks) after similar




                                     9-8

-------
durations  of contact.   Patches were  removed  from the subjects after 1, 2, 4,




and  8  hours; it was found  that  most of  the  ethylene  oxide diffused from the



fabric  and  rubber patches within  1  hour and  from  the  PVC  film  patches within 4




hours, but  that the PVC block retained a  substantial portion of ethylene oxide




residue at  4  hours.  In a  subsequent experiment the  same  subjects  (i.e., those




previously  exposed  in  the  first  experiment)  were exposed  to  patch materials




that retained  ethylene  oxide the  longest  (thick PVC blocks  and  petrolatum




applied to  Webril pads).   It was  found that the reactions were most  widespread




when the  ethylene  oxide  levels   in  these materials  were =1000  ppm; erythema




appeared  in  10 of  the 12  PVC block subjects and  10 of  the  12  petrolatum




subjects after  4-8  hours  of contact,  and cleared  within  3-4  days.  Reactions




were not elicited at nominal  levels of 10 or  100  ppm  ethylene  oxide  in  PVC or




petrolatum,  although  one  subject  who had  developed  sensitivity   to  1000  ppm




ethylene  oxide  in  PVC block in  the first  experiment  showed  a mild  delayed




reaction  to  100   ppm.    Little  or  no   reaction  developed  to  patches  that




contained  ethylene oxide by-products  that  were present in  the original patches




(i.e.,  ethylene  glycol and  ethylene  chlorohydrin),   indicating  that ethylene




oxide was the toxic  agent.




     Although incidental findings  in  the  Sexton and  Henson  (1950)  and  Shupack




et al.  (1981) experimental studies described above suggest that ethylene oxide




can  cause  skin  sensitization,  Thiess (1963)  did  not  observe  sensitization  in




ethylene oxide plant workers who  were challenged with a single dermal applica-




tion of 1%  after  an average of 10.4  years  of occupational  exposure.  Anaphy-




lactic reactions have been  observed  in patients using ethylene  oxide  steril-




ized plastic  tubing for  hemodialysis (Poothullil et al. ,  1975)  or  cardiac




catheter izat ion (Pessayre and  Trevoux,  1978).  These  symptoms included  uti-




                                      9-9

-------
caria,   breathlessness,  and  hypotension.    In a  follow-up  study on  a  patient




apparently sensitized by  contact  with  hemodialysis tubing,  Dolovich and  Bell



(1978)  illustrated  that  this  patient showed a positive  skin  test  response to




ethylene  oxide-serum  albumin  conjugate,  and  produced  in  vitro  histamine




release to this antigen.  This response indicates that a specific IgE antibody




to ethylene oxide had been induced in this patient.




     Clinical reports of  hemolysis  following  use  of ethylene  oxide  sterilized




plastic tubings have also been published (Hirose et al. ,  1963;  Clarke  et  al.,




1966).   Ethylene oxide,  rather  than  a  chemical reaction product,  is  impli-




cated,  since  this type of  effect can  be  prevented  by extensive aeration  of




ethylene oxide sterilized  plastic  devices.




          Ethylene oxide  vapors in high concentrations are  irritating  to the




eyes, but ocular contact with liquid ethylene oxide can cause  severe burns.  A




workman exposed  to  ethylene oxide  in an unstated  manner  was  reported  to  have




suffered a corneal burn, but healing was observed  within  48  hours  following a




cornea 1 denudement procedure  (McLaughlin,  1946).   Thiess  (1963) described two




cases of accidental eye injury with ethylene oxide.  A  nurse  was exposed  to a




direct  blast of ethylene  oxide from a  sterilizer  cartridge,  and developed an




epithelial keratitis of the cornea  within  3 hours.  Within  24  hours,  the eye




was entirely normal.  The second  case  involved a patient who  received a squirt




of liquid ethylene oxide (concentration not stated) in  the eye and was treated




immediately by  extensive  washing  with  water;  this  resulted  in only  irritation




of the conjunctivae that persisted for  =1 day.








     9.2.1.2.    SUBCHRONIC  AND  CHRONIC  EXPOSJRE  —  Limited  information   is




available on toxic effects of subchronic or chronic ethylene  oxide exposure  in



                                     9-10

-------
humans.   The  information  is  largely derived  from  clinical case reports  from




retrospective  mortality studies.



     Gross  et al.  (1979)  reported  on  four  cases of apparent ethylene oxide-




induced  neurotoxicity.   This  occurred  in a plant  in which a  sterilizer was




found  to  have leaked  for 2 months of operation.  The exact  levels of ethylene




oxide  were unknown, but  the four  individuals involved reported that they could




intermittently  smell  the  ethylene  oxide gas,   indicating  roughly a  level of




>700 ppm.    The  length  of  exposure  to  ethylene   oxide   from  the  leaking




sterilizer was 3  weeks  for cases  1 and  2, 2  weeks for case  3, and 2 months for




case 4.   Three of  the  four  cases had  worked  as sterilizer  operators  for >2




years  and were exposed  to  ethylene oxide  from the leaking sterilizer for 2, 3,




or 8 weeks;  the  fourth  had been an  operator for only 3  weeks and was exposed



for the duration.




     The  individual who had  been exposed to ethylene  oxide for 3  weeks had




noted  conjunctival and mucosal  irritation and transient blunting  of the senses




of smell  and  taste,  and  developed  headache,  nausea, vomiting,   and  lethargy




that was  followed by  acute encephalopathy (recurrent major  motor  seizures at




20-30  minute  intervals).   Two  of the  other three  operators were symptomatic




(i.e.,  headaches, numbness and weakness  in the  extremities,  fatiguability,  one




case of  memory/think ing  disturbances)  and had  abnormal  neurological examina-



tion results that were consistent with sensorimotor  neuorpathy.   Nerve conduc-




tion studies  were abnormal in  these three  operators, including  the  asympto-




matic  patient, and  were compatible  with  the diagnosis of  sensorimotor  neuro-




pathy.   Removal  from  exposure resulted  in relief of symptoms within  2  weeks.




Two of the individuals returned to work  under normal conditions of lower  ethy-




lene  oxide  exposure,  but  improvement  in  nerve  conduction  was not  observed;




                                     9-11

-------
significant  improvement was  noted,  however,   in   the   third  individual  who




returned to work in a position without ethylene oxide exposure.




     Jensen  (1977)  reported  that  three  workers  using  ethylene  oxide steril-




izers  were  hospitalized   for   neuropathy  of  the  lower  linbs.    Follow-up




indicated that these effects were  reversible.




     Jay  et al.  (1982)  found  that the  four  sterilizer  operators  described




above (Gross  et al. ,  1979),  who  were exposed  to excessive levels  of ethylene




oxide from a  leaking sterilizer and developed neurologic abnormalities,  subse-




quently  developed  cataracts.    The  operator  exposed for  2  months  developed




bilateral cataracts during the following 2 1/2 years; cataracts were diagnosed




in  the  other three operators upon  examination =3  1/2 years after  exposure to




the  leaking  sterilizer.   Eight  other  men  whose  work  involved  exposure  to




ethylene  oxide  sterilizers (6  of the 8  were sterilizer  operators),  but who




were  asymptomatic,  were  subjected  to  complete   ocular  examinations,   but




cataracts were  not  found.  Four of the 12 men, two  of  whom had  not worked on




the  leaking sterilizer,  had  increased central  corneal   thickness  with  normal




endothelial cell counts when  compared  with a control group  of 12  subjects of




higher  average  age (41 vs.  33  years).   None of  the patients  were examined




before  exposure  to  ethylene  oxide, but the  authors  believed  it  unlikely that




cataracts  would  occur by  chance  in persons  in  this age  range,  particularly




because none  of the patients  had  any systemic or ocular disease  that might be




associated with cataract formation.




     Hemoglobin values and lymphocyte counts were reported to be significantly




lower and  higher,  respectively,  in a group of Swedish  ethylene  oxide produc-




tion  workers when  compared  with  control subjects  (Ehrenberg  and  Hallstrom,




1967).  The design and  results of this  study  are  more completely described in



                                     9-12

-------
Section  9.5.2,  but  it   should  be  noted   that  the  production   workers  were




reported  to have been  exposed  for 2-20  years  (average 15  years)  to an unknown



level of the compound.




     Joyner (1964)  conducted  a retrospective  morbidity  study  of  37 male ethy-




lene  oxide  production  plant  workers.   These workers varied in age from 29-56




years and  were  exposed to typical concentrations of 5-10 ppm (range 0-55 ppm)




for 5-16 years (mean 10.7 years).  Age-matched controls consisted of 41  opera-




tors  (mean  length  of service,  11.7  years)  assigned to other production  units,




who had past exposure  to  many  different petrochemical industry agents, but had




never exhibited  clinical effects attributable to  systemic  chemical toxicity.




As  detailed  in  Sections 9.3  and  9.5,  no  significant  increase  in  health




problems  relative  to  controls  was  found.    This evaluation  should have  been




sufficient  to  identify  major  toxic  effects  of extended  low-level  ethylene




oxide exposure, although  limitations in the design of the study,  as well as an




insufficient period  of observation,  preclude  evaluation  of more  subtle  toxic




or carcinogenic responses.




     An  excess of  deaths  from specific  causes  (including  all  circulatory




causes)  other  than  certain  malignancies (Section 9.5) was  not  observed  in  a




group of 767  male  ethylene  oxide  workers from the  Texaco  Chemical Company




Plant in Port  Neches,  Texas  (Morgan et al. ,  1981).  These  cohort  members  had



been employed  for at least 5 years between January 1955 and  December 31,  1977,




and an industrial hygiene survey of the plant performed in  July,  1977,  showed




that the 8-hour TWA  exposure  to ethylene oxide was well below 50  ppm.




     Hogstedt  et al.  (1979a)  conducted  a  cohort   study of mortality among  89




full-time ethylene  oxide  production  workers,  86  intermittently  exposed  main-




tenance  workers, and a group  of 66 unexposed control workers  during the years




                                     9-13

-------
1961-1977.  As described in Section 9.5, exposure patterns were quite complex;




in  addition  to  ethylene  oxide  (concentrations  were  generally <50  mg/nr),




workers  were  exposed  at  different  times to  ethylene  dichloride,  ethylene




chlorohydrin, ethylene, low concentrations of bis(2-chloroethyl)ether,  as well




as  traces  of other chemicals.   It was found that the full-time exposed cohort




showed considerable  excess  mortality when  compared  with the  number expected




based  on  national  statistics.    The  excess  mortality arises  mainly  from




increased mortality due to stomach cancer and leukemia (Section 9.5), but also




from diseases of the circulatory  system.   When at least  1 year of exposure and




_>10 years of  induct ion-latency time  were  required  for  inclusion  in  the study,




there were  12 observed deaths attributed  to  the  circulatory  system (9  due to




coronary heart disease and 3  due  to cerebrovascular  disease),  compared  to the




expected  incidence  of 6.3;  this  difference  was  statistically   significant




(P<0.05).   The excess  mortality  was of the same  magnitude  in  a  restricted




cohort of  those with MO  years of employment in ethylene oxide production and




20 years of induct ion-latency  time (7 observed, 2.2 expected).








9.2.2.  Effects in Animals.








     9.2.2.1.  ACUTE EXPOSURE  —  The acute toxicity  of ethylene oxide is sum-




marized in Table 9-1 .  Exposure of  mice,  rats, guinea  pigs,  rabbits,  and dogs




to  lethal  levels  of ethylene  oxide has produced  symptoms  of mucous membrane




irritation and CNS depression, including lacrimation, nasal discharge,  saliva-




tion,  nausea,  vomiting,  diarrhea,  respiratory irritation, in coordination, and




convulsions (Sexton and Henson, 19^9; Hollingsworth  et al.,  1956;  Mine et al.,




1981).   Animals  that  survived  the initial exposures  showed  subsequent bron-




                                     9-14

-------
chit is,  pneumonia,  and  loss of  appetite,  with  delayed  symptoms  of apathy,




dyspnea, vomiting, paralysis (particularly of the hindquarters),  periodic con-



vulsions,  and  death  (Waite  et  al. ,  1930;  Hollingsworth et al. ,  1956).  Prompt




deaths  are usually  due to  lung  edema;  delayed deaths  frequently  result from




secondary  infections  in  the lungs,  although general systemic intoxication may




also be a  factor  (Hine et al. ,  1981 ).




     Pathological  findings  following  lethal  exposure to  ethylene  oxide  in




mice,   rats, and  guinea pigs showed  congestion of the  lungs,  hyperemia  of the




liver  and  kidneys,  and gray discoloration  of the  liver (Waite  et al. ,  1930).




Pathological  findings  after delayed death  caused  by  ethylene  oxide  included




emphysema  of  the lungs,  fatty  degeneration of the  liver,  cloudy  swelling  of




the kidney  tubules, and  congestion  of the  spleen and  brain (Hollingsworth  et




al. ,  1956).    Intravenously-administered  ethylene  oxide caused  congestion  in




all organs  of the rabbit  (Greaves-Walker and Greeson, 1932).   Zamlauski and




Cohen  (1976) have reported  that infusion of ethylene oxide in the rat at blood




levels of  0.45-4.5 mg/mS,  produced a significant decrease  (-30$)  in glomerular




filtration rate,  which indicates effects of  ethylene  oxide on kidney function.




     Ethylene   oxide  in  10  and 50%   aqueous solutions   produced  hyperemia and




edema   in shaved  rabbit  skin when  applied  through cotton pads for 1-60 minutes




(Hollingsworth et al. ,  1956).  Bruch (1973)  studied  the dermal irritation pro-




perties of 2-10$  aqueous  ethylene oxide  solutions  in guinea  pigs and  rabbits.




Subcutaneous  injection  in  the guinea  pig   resulted   in  ecchymoses and  skin




thickening, while intradermal injection and  topical  application  in  the  rabbit




resulted in  mild irritation.   Topical  or   intraderraal administration  of  1/J




ethylene oxide (0.5  m£),  thrice weekly for 3  weeks,  did not result in  sensiti-




zation in guinea  pigs  (Woodward and  Woodward,  1971).




                                     9-15

-------
     McDonald et al.  (1977)  studied  the  ocular effects of  varied  concentra-




tions of ethylene oxide  in  saline  applied  repeatedly  over a  6-hour  period  to




the eyes of rabbits.   They  observed  a  dose-dependent  increase  in congestion,




swelling,  discharge,  iritis,  and corneal cloudiness,  indicating  the  irritating




effect  of  ethylene  oxide  on  mucous  membranes and  corneal  epithelium.   The




maximum nondamaging  concentration  for  this  time  period  was  0.1/6  ethylene




oxide.   In another  study  of ocular  irritation  in  rabbit eyes,  Vfoodward  and




Woodward (1971)  found slight irritation following a single  application  of 10$




aqueous ethylene oxide (duration of exposure unknown),  and a no-effect concen-




tration of 2.1$ ethylene  oxide  was determined.  The higher  values  determined




in this study are probably the results of a different mode of application and,




therefore,  different  duration of exposure.








     9.2.2.2.   SUBCHRONIC  AND  CHRONIC EXPOSURE -- The  subchronic toxicity  of




inhaled ethylene oxide has been  investigated  in  a variety  of different animal




species by  different routes  of  exposure  (Hollingsworth  et al.,  1956;  Jacobson




et al. , 1956).   As  summarized  in Table 9-2,  symptoms of  poisoning  and patho-




logic  changes are similar  to those observed in acute studies, including lung,




kidney, and  liver damage,  and  neuropathy  of the hindquarter  and  testicular




tubule degeneration in some species.



     Hollingsworth  et  al.   (1956)  observed   neurotoxic  effects in  animals




following  inhalation  exposure   to  357  ppm ethylene oxide  vapor  for   several




weeks  (the  exposure  for  each   species  is presented  in  Table  9-2).    Rats,




rabbits, and monkeys  showed  paralysis and  atrophy  of the  muscles of the hind




limbs.   These   effects were  reversible  100-132  days after discontinuation  of




exposure.   Special studies  on  monkeys were carried  out  with repeated  (38-94)




                                      9-16

-------
                                                                             TABLE 9-2

                                                               Subchronic  Toxioity of Ethylene  Oxide
       Route
                           Species
                                                        Concentration
                                                      Number of
                                                      Exposures
                                                                                                               Effects
                                                                                                                                            Reference
     inhalation
     inhalation
UD

--3
     inhalation
     inhalation
20 rats (10/sex)
16 guinea pigs (8/sex)
 5 mice (female)
 2 rabbits (1/sex)
 1 monkey (female)

30 mice (female,  white)
20 rats (male,  white)
                                                             ppm
                                                         100  ppra
                        20 rats (10/sex)
                        10 mice (female)
                                 357  ppm
                        16 guinea pigs  (8/sex)
                                 357  ppm
up to 8 in 10 days
(7 h/d; 5 d/wk)
30 (6 h/d; 5 d/wk)
33-38 (7 h/d; 5 d/wk)
123 in 176 days (7 h/d;
5 d/wk)
                                                                                                        Death  in  all  animals.   Patholo-      Hollingsworth
                                                                                                        logic  changes in  lungs,  liver  and    et  al.,  1956
                                                                                                        kidneys similar to  those in  acute
                                                                                                        poisoning.
Weight loss, reddish nasal dis-     Jacobson et al.
charge, diarrhea, labored breath-   1956
ing, weakness of the hind legs,
and some deaths (13/20 exposed
and 0/20 control rats, and 2M/30
exposed and 3/30 control mice).
Fifteen additional rats or mice
were examined pathologically;
changes were limited to a few
cases of hemoslderosis in the
spleen that occurred late in the
exposure period.

Death in 10/10 mice (33 exposures)  Hollingsworth
and 18/20 rats (38 exposures)       et al., 1956
caused by secondary respiratory
infections.  Impairment of sensory
and motor function in rats prior to
death, resulting in reversible
hind leg muscle paralysis and
atrophy.

Growth depression, degeneration     Hollingsworth
of the testicular tubules with      et al., 1956
replacement fibrosis (males),
slight fatty degeneration of the
adrenal cortex (females).  No
nervous system effects or
mortality.           •

-------
                                                                       TABLE 9-2  (oont.)
Route Species Concentration
Inhalation 2 monkeys (I/sex) 357 ppm
2 monkeys (males) 357 ppm
Number of
Exposures
38-11 in 60 days
91 in 110 days
(both schedules 7 h/d;
5 d/wk)
Effects
Growth depression and characteris-
tic neurological impairment (e.g.,
hind limb paralysis and muscular
atrophy, poor or nonexistent knee
reflex, extensor reflex and
hindquarter/genitalla pain percep-
tion). No histopathologlc effects
of exposure.
Reference
Holllngsworth
et al.,1956
    inhalation
                       3 dogs (male, Beagle)
                            290  ppm
MD
 I
oo
    inhalation
                       20 rats
                                                    20M ppm
    inhalation
                       60 mice (30/sex)
                            250 ppm
    inhalation
8 guinea pig
1 rabbits (2/sex)
2 monkeys (female)
                                                    20t ppm
30 (6 h/d;  5 d/wk)
                                                  127-133 in  185-193  days
                                                  (7 h/d;  5 d/wk)
                                                                          50-55 (6 h/d;  5 d/wk)
127-157 in 176-226 days
(7 h/d; 5 d/wk)
Two of 3 exposed dogs showed        Jacobson et al.,
toxic signs that included vomiting, 1956
slight tremors, transient weakness
of the hind legs and decreases In
red blood cells, hemoglobin, and
hematocrlt.  Hematologlc parameters
normal in control dogs.  Lungs
showed congestion and alveolar
collapse and fatty changes in the
hindquarters were consistent with
muscular atrophy.

Weight loss, some deaths with       Holllngsworth
effects on lungs (congestion,       et al., 1956
hemorrhage, emphysema, atelecta-
sis) kidneys and testes (slight
degeneration of some tubules)
(slight cloudy swelling of tubules)

Signs of neuromuscular toxicity,    Snellings et al.
decreased red blood cell count,     1981a
packed cell volume and hemo-
globin concentration were ob-
in both sexes.  No histopathologic
effects were observed.

No effect on growth or mortality.   Hollingsworth
Evidence of paralysis/muscular      et al., 1956
atrophy in the rabbits and monkeys.
Slight edema and congestion noted
In rabbits' lungs.

-------
                                                 TAELE 9-2 (cent. )
Route Species Concentration
inhalation 20 rats 113 ppm
8 guinea pigs
U rabbits (2/sex)
2 monkeys (females)
inhalation 60 mice (30/sex) 100 ppm
Inhalation 30 mice (females, l*iite) 100 ppm
20 rats (male, Wiite)
inhalation 2tO rats (120/sex) 100 ppm
inhalation 3 dogs (male, Beagle) 100 ppm
inhalation 60 mice (30/sex) 50 ppm
inhalation 20 rats H9 ppm
8 guinea pigs
Number of
Exposures Effects
122-157 in 176-226 days Growth depression and a moderate
(7 h/d; 5 d/wk) increase in lung weights in rats
were the only adverse treatment-
related effects noted.
50-55 (6 h/d, 5 d/wk) Hunched posture and reduced loco-
motion observed in both sexes.
No histopatho logic effects were
observed.
130 (6 h/d, 5 d/wk) No clinical signs of toxiclty or
treatment related mortality
(3/20 exposed and 3/20 control
rats, and 8/30 exposed and U/30
control mice died). No significant
pathologic changes in additional
groups of 60 rats or mice.
T45 (6 h/d, 5 d/wk) Early deaths and decreased body
weight gain were observed
starting at week 1.
130 (6 h/d, 5 d/wk) Normochronic anemia (decreased
red blood cell, hemoglobin, and
hematocrit) indicated in 1 and
suggested in 1 of 3 dogs. No
changes in the 3rd exposed dog,
or in control dogs.
50-55 (6 hr/d, 5 d/wk) Hunched posture in males and re-
duced locomotion in females. No
histopatho logic effects were ob-
served.
127-131 in 180-18H days No adverse effects as Judged by
(7 h/d, 5 d/wk) general appearance, behavior,
Reference
Hollingsworth
et al. , 1956
aiellings et al. ,
198ta
Jacobson et al. ,
1956
Snellings et al. ,
198l|b
Jacobson et al. ,
1956
Sne llings et al. ,
198Ha
Hollingsworth
et al. , 1956
 M rabbits (2/sex)
10 mice ( female)
mortality, growth,  final body
and organ weights,  and gross
or microscopic pathologic
examination.

-------
                                                                    TAELE 9-2  (cent. )
Number of
Route Species Concentration Exposures
inhalation 210 rats (120/sex) 33 ppm 115 (6 h/d, 5 d/wk)

inhalation 60 mice (30/sex) 10 ppm 50-55 (6 h/d, 5 dA
-------
exposures  to this  level  of ethylene  oxide.   Knee  jerk  reflexes became  very




weak,  pain perception  in the hind  quarters decreased,  the creraasteric  reflex



was  elicited,   and  the  extensor  reflex  of the  palms  of  the hind  feet was




abolished.   Impairment of both sensory  and motor function at  the lumbar and




sacral level of the  spinal cord was indicated.   Exposure  of monkeys to a lower




level of ethylene  oxide  (204  ppm  for  176-226 days) produced partial paralysis




and  some  muscular atrophy  of the hind  legs with  moderate  suppression   of the




leg reflexes.   The Babinski reflex  was present after  this lower level exposure




to ethylene oxide.




     In  a  more recent  subchronic   study,  Snellings et  al.   (1981*3)  exposed




groups of  30  male and  30 female B6C3F1  mice to  the  vapors  of ethylene  oxide.




Exposures were  for  6  hours/day, 5 days/week for 10-11 weeks to nominal  levels




of 1,  10,  50,  100,  and 250 ppm.   No  effects were  observed on survival, body




weight or  histologic  sections of a variety of  organs.   At  the  three  higher




exposure levels,  however,  signs of neuromuscular  toxicity  were  observed.  In




both sexes  of the high exposure  group,  there  was a  statistically significant




increase  in hunched posture,  reduced  locomotion, and  righting  reflex.   The




former  two  were  observed in  the  100  ppm  group  and  in  males  and  females,




respectively, of the 50 ppm group.   The  abnormal righting reflex  was observed



only during  intermediate  testing  of females in  the  100  ppm group.   In   addi-




tion,  reduced toe pinch reflex was reported for  females tested at  intermediate




periods and reduced tail pinch reflex  was reported for males at termination in




the 250 ppm group.   Also at  termination in the high dose  group, hematologic




parameters, red blood cell count,  packed cell volume and  hemoglobin concentra-




tions were  decreased,  and  some changes  in  either absolute or  relative  organ
                                     9-21

-------
weights  were  observed.   In  this  study,  the neu ro muscular effects appeared  to




be  the most sensitive indicator of exposure to  ethylene  oxide.




     Preliminary results of a  chronic  inhalation study  conducted by  NIOSH have




been reported  (Lynch et al.,  1982a).   Male F344  rats (80 per treatment  group)




and male  Cynomolgus monkeys  (12 per  treatment  group)  were exposed  to either  50




or  100 ppm ethylene  oxide  for 7 hours/day, 5  days/week  for  24  months.  Addi-




tional details of the experimental design  are  presented  in Section 9.5,  but  it




should be noted  that the  rats were  included primarily  for  carcinogenicity




evaluation, and the monkeys used to  determine  target  organ toxicity.   A  number




of  indices  were  evaluated   including  body  weights,   hematology,   clinical




chemistry,  urinalysis,   opthalmology,  pulmonary   function,   neurophysiology,




neuropathology, gross and histopathology,  sister chromatid exchange  rates,  and




chromosomal  aberrations  in  peripheral  lymphocytes.    The  results  that  are




currently available are  summarized  below.




     As  detailed  in  Section 9.5,  weight gain  throughout most of the  exposure




and survival periods were significantly depressed  in  the rats at both  exposure




levels (Lynch  et al. ,  1982a).   Weight gain was significantly depressed  in the




treated monkeys beginning at  week  25.   The liver and spleen  of the  rats were




the only  organs in  which  histopathological evaluations  have been  completed,



but the  preliminary terminal  sacrifice  spleen data  indicate  a dose-related




induction of leukemia (Section 9.5).   Hematologic analyses showed no  statisti-




cally significant  change  in  red  blood cell  count in  the  treated  rats,  but




white  blood cell counts were highly  variable and reflected the  presence  of the




leukemia.  There were no differences in the red or  white blood  cell counts  in




either of the  monkey  groups,  although  increased   frequencies  of chromosomal
                                     9-22

-------
aberration  and  SCE were  observed  in  the  peripheral  lymphocytes  of  these




animals.



     In  another  2-year   inhalation   toxicity   study  in  rats  conducted  by




Snellings  et  al.   (1984b),  the  only  non-neoplastic  effect  reported  was  a




decrease in  body  weight gain.   As  described  more  fully in Section 9.5, groups




of  120 male  and  120   female  Fischer  344  rats were exposed  6 hours/day,  5




days/week  to  ethylene  oxide  at  target  levels of  0,  10,   33,  and  100  ppm.




Decreased  body  weight  gain  was  observed in  both   sexes  after  4 weeks  in the




high exposure  group and in  females of  the  33 ppm  group after 10 weeks.  Other




groups  were  similar to control animals.   Early deaths were reported  for the




high exposure  group and  were  likely  to  be tumor-related.   The incidence and




type of neoplastic  lesions are discussed in Section 9.5.




     Significant hematological  effects (i.e.,  anemia)  have  also been observed




in ethylene  oxide-exposed dogs.   Jacobson  et  al.  (1956) found  decreased red




blood  cell counts,  hemoglobin,  and hematocrit  in  2 of 3 beagle dogs that were




exposed to 292 ppm  ethylene oxide vapor  for  6 hours/day,  5  days/week  for 6




weeks.    Definite  (1  dog) and  suggestive (1 dog)  hematologic  effects  of the




same type  were also observed in 2 of  3  dogs  that were similarly exposed to 100




ppm ethylene oxide  for  6 months.    Woodward  and Woodward (1971) demonstrated a




dose-related  increase   in  anemia   in  dogs  that were administered  6-36  mg/kg




ethylene  oxide in  30  daily subcutaneous  injections.   Pathologic examination




showed  hyperplastic bone  marrow  and  ectopic  hematopoiesis.    Balazs  (1976),




however, was  unable to repeat  these  findings  in  beagle  dogs with an ethylene




oxide-glucose  solution  administered intravenously   over  the  same concentration




range  in a 21-day study.
                                      9-23

-------
     An oral feeding study using 1Q% ethylene oxide  in  olive  oil was  performed




on rats (Hollingsworth  et al.,  1956).  Rats fed 100 rag/kg ethylene oxide in  15




doses over  21  days  showed narked weight  loss,  gastric  irritation,  and  slight




liver damage.  Feeding of 30 mg/kg  in  22  doses produced no observable adverse




effects.








9.2.3-  Summary of Toxicity.  The primary  effects of acute  inhalation exposure




to high concentrations of ethylene  oxide  gas are  respiratory tract irritation




and  CNS depression.    Headache,  vomiting,  dyspnea,  and  diarrhea  are   common




systemic effects  of  vapor exposures  in  humans, and  excessive  exposures have




produced  bronchitis,   pulmonary  edema,  and  convulsive  movements.    Similar




effects have been observed in a variety of animal species, but paralysis (par-




ticularly of the hindquarters)  and periodic  convulsions frequently  preceded




death.  Death  in  ethylene oxide-exposed  laboratory animals  is  usually  due  to




lung edema or  secondary  lung infections,  and postmortem pathologic findings  in




other  organs  include  widespread hypereraia  and  congestion  (liver,  kidneys,




spleen) and  fatty degeneration  (liver).




     Derma to logical effects  following  skin  contact with  ethylene  oxide   in




humans  from  accidental or experimental exposure include  edema,  erythema,  and




vesiculation  with  possible  bleb   formation,  in   that  sequence.    Vesicle




formation is usually delayed (e.g.,  6-12  hours),  the magnitude  of skin  injury




appears  to   be related  to  concentration  and  duration  of  contact,   and  the




effects are  reversible.   Concentrated  ethylene  oxide evaporates from the skin




resulting in a freezing   effect,  but more  dilute  solutions penetrate  the skin




more  effectively,  resulting  in chemical  burning;  weak  solutions  lack   suffi-




cient chemical strength  to  cause  injury  except after prolonged contact.  Skin




                                     9-24

-------
burns have also been caused by residual ethylene oxide in clothing or footwear




treated or  accidentally contaminated with  the  compound.    Sens it izat ion  has




also been  associated  with repeated  dermal  exposure to ethylene  oxide  at the




sites of  contact.   Similar  dermal  irritative  effects  have been  observed  in




experimentally  exposed  rabbits  and  guinea  pigs,   but  sensitization was  not




demonstrated by  topical or  intradermal administration  in guinea pigs.   High




concentrations of ethylene oxide vapors are  irritating to  the eyes of humans




and animals, and direct ocular contact  with  liquid  ethylene oxide can produce




cornea 1 injury.




     Case reports indicate that  neurological  effects (e.g., headache/vomiting,




sensorimotor neuropathy, seizures) and  ocular effects (e.g., cataracts)  may be




primary effects of limited repeated exposure  to high levels of ethylene oxide,




and hematological effects  (reduced hemoglobin and  increased number  of lympho-




cytes)  have been noted  in  chronically  exposed ethylene  oxide production plant




workers.   Retrospective morbidity and mortality studies of ethylene oxide pro-




duction  workers do  not,  however,  suggest  chemical-related,  non-neoplastic




toxicity.   Subchronic   exposure  of  different  species  of  animals  to ethylene




oxide by  different routes of exposure  produced  effects similar  to  those seen




in  acute  studies;  symptoms of poisoning  primarily reflect neurotoxic  action




(e.g,  hindquarter  neuropathy)  and pathologic changes  generally  occur  in  the




lungs,  kidney,  and  liver (e.g., congestion and degenerative changes), although




testicular effects (e.g.,  tubule degeneration)  and hematologic effects (e.g.,




anemia) have also been  observed.
                                     9-25

-------
 9.3.  TERATOGENICITY  AND  REPRODUCTIVE TOXICITY








 9.3.1.    Teratogenic   Effects.      Batelle   Pacific   Northwest   Laboratories




 (Hackett  et al. ,   1982)  conducted  teratology  and reproductive studies  for  the




 National  Institute   for  Occupational  Safety  and  Health  investigating  the




 effects  of  ethylene  oxide  produced  by  inhalation exposure.   Pregnant  rabbits




 and  rats were  exposed  to  a  single dose of ethylene oxide  (150 ppm,  Union




 Carbide,  Linda  Lot  No.  01901,  99.7?  pure)  both prior to and during  the period




 of organogenesis.   Thirty New Zealand White  rabbits per group were  exposed  in




 three different regimes (filtered air alone,  ethylene  oxide exposure  on days




 7-19  of gestation,  and ethylene  oxide  exposure on days  1-19 of gestation).




 Forty-one Sprague-Dawley CD  rats per  group   were  exposed  according  to  four




 different schedules  (filtered  air  alone,  ethylene oxide exposure  on days 7-16




 of gestation, ethylene  oxide  exposure on  days 1-16 of  gestation,  and ethylene




 oxide exposure 3 weeks prior to mating and through days 1-16 of gestation).




     In  the rabbits, no  toxic effects  were   observed  in   the  mothers  (i.e.,




 changes  in  body  weight,  organ  weight,   histopathological  changes   in   the




 organs).  In  addition,  there  were no decreases  in  the percentage of pregnant



 animals  nor was there  any  indication of  adverse effect  on  the  fetus  (i.e.,




 decreases in fetal body  weight, crown  rump length,  sex ratios, or  morphologic



 alterations).




     In  the rats,  maternal toxicity was  observed with sporadic  decreases  in




 food consumption,  decreases  in body  weight, and increases  in kidney and  spleen




weights,   with   the  increases   in  spleen  weights  roughly   proportional  to  the




 duration of exposure.   Adverse effects were  also observed  in  the developing




 conceptus.  There  was an increase in resorptions in animals exposed both pre-




                                     9-26

-------
and postgestationally  with  a trend for  early  midgestational resorptions.  In




addition,  lowered   fetal body  weight,  decreased  crown-rump  length  and  an



increased  incidence  of incomplete  skeletal  ossification  were observed in  all




ethylene  oxide  exposed  offspring,  and  this  was  especially  pronounced  in




animals  exposed  both  pre- and postgestationally.   It was concluded  from this




study that exposures of  150  ppm  in rats caused significant adverse effects in




both the nether  and developing fetus; however,  since  only  one dose  was used in




this study, it is not known  whether these developmental effects  would  occur in




the absence of maternal toxicity.




     Because  of  concerns over adverse  reproductive effects which  could  occur




as a result of exposure  to  ethylene oxide or ethylene oxide  reaction products




left on  improperly  degassed  surgical  supplies,  LaBorde and  Kimmel  (1980) con-




ducted  studies  on  the effects  of  ethylene  oxide  administered  intravenously.




CD-1 mice  in  four  replicates of three  treatment  groups (10  animals per group)




were treated  with  0,  75, 150 mg/kg ethylene oxide (Eastman  Organic  Chemicals




Co., purity not  stated,  ethylene  oxide was  injected  in 5% dextrose solution).




The animals were exposed in  the following treatment periods  of gestation: days




4-6  (period  I),  days 6-8  (period  II),  days 8-10  (period  III)  and days  10-12




(period IV).



     Clinical  signs   of  maternal  toxicity   (weakness,    labored  breathing,




tremors, and  death) were  observed  in  animals injected with  150  mg/kg ethylene




oxide on gestational  days  4-6 (period I), 8-10  (period III)  and 10-12 (period




IV) but not in the group  injected  on  days 6-8  (period II).   Decreases  in mean




maternal body weight gain  were  observed in animals  in  period  I,  period III,




and period  IV and were  accompanied by  decreases  in  the  mean number of live




fetuses  in  periods III  and  IV.    Embryotoxicity as manifested  by  significant




                                     9-27

-------
reductions  in mean fetal weight  was observed in all  four periods at the  150




mg/kg  dose.   There was  no  significant change in  the  mean number of  implants




per  litter,  but there was a  reduction  in  the mean number of live  fetuses  per




litter (and  an  increase in  the  number  of  dead  and  resorbed  offspring)   in




periods  III  and IV at  the  150  mg/kg level.   An  increase  in  the percent  of




malformed  fetuses/litter was noted  in  periods II, III,  and IV  at 150  mg/kg




level,  but  in period  III  the incidence  did not achieve  statistical  signifi-




cance.   It  was concluded that the ethylene  oxide  exposure under these condi-




tions  was  selectively affecting  the development of the conceptus (as seen  by




skeletal  malformations and   embryotoxicity)  since  ethylene  oxide exposure   in




period II  (days 6-8  of  gestation)  produced  malformations and embryonic  death




while  not affecting the  mother (no  clinical  signs  of toxicity) ; however,  this




conclusion was tempered  somewhat by  the observation  of maternal  deaths   in




treatment groups  before  and  after  this time  period.   Although  there was  no




dose-response relationship  in the severity  of adverse effects  in  either  the




mother  or  fetus,  the  types of  malformations  seen   in  periods  II  and   III




appeared  to   follow a developmental  pattern.   The  authors reported  that   in




animals treated on days  6-8 cervical and upper thoracic vertebrae malformation




were  observed.   Animals  treated  on  days 8-10  had  defects primarily in   the



lower  thoracic region.




     Another study by  the same investigators (Jones-Price et al. , 1983) evalu-




ated the reproductive  effects of intravenous  injections  of  ethylene  oxide   in




rabbits.   New Zealand  White  rabbits  were  intravenously injected in two treat-




ment regimes;  0,  9,   18, or  36 mg/kg  ethylene  oxide  (source and  purity  not




reported)  on  days  6-14  of  gestation,  or 0,  18,  or 36  mg/kg on days 6-9   of
                                     9-28

-------
gestation.  Seventeen to 21 animals were examined  in  the  group  exposed  on days




6-9, 18-24 animals examined in the group  exposed  on days  6-14.



     Maternal toxioity was observed  in both  exposure  groups,  with more  severe




effects observed in the groups treated on days 6-14 than on days 6-9 of gesta-




tion.  Significant decreases  in maternal weight gains were observed during the




entire treatment at the  18 and  36  mg/kg  level.   These decreases included both




decreases  in  pregnancy  weight  gains and  decreases in  absolute  weight gains




(weight gained during pregnancy minus uterine weight).  No embryotoxic  effects




were observed in the day 6-9  treatment groups; however,  in the  6-14 day treat-




ment  group significant  dose-re la ted  trends  for  decreased   numbers  of  live




fetuses/litter and  resorptions/litter  were  observed.   At  the  36  mg/kg  level,




the  incidence of  resorptions/litter  was  statistically significantly different




from control levels.  The authors concluded that intravenous administration of




ethylene  oxide  in  rabbits  produced embryotoxicity,  though at  doses which also




produced significant maternal toxicity.




     LaBorde et al.  (1982) presented  data  at the  1982  Society of Toxicology




meeting regarding  the  teratogenic effects  of ethylene  chlorhydrin  (ECH),  a




reaction product of  ethylene  oxide,  in mice and rabbits.  ECH is produced by




the  interaction of ethylene  oxide and chloride  ions, so  it  is a  residue  of




ethylene oxide  that could  be  left  on medical devices after improper degassing




of  ethylene  oxide  during  sterilization.    Forty-one to  65   CD-1  mice  were




injected intravenously  with 60 mg/kg or 120 mg/kg ECH  (source not reported; in




5% sterile dextrose)  on days  4-6,  6-8, 8-10,  or 10-12 of gestation.  Seventeen




to 22 New  Zealand  White  rabbits were intravenously injected  with  9,  18, or 36




mg/kg ECH on days  6-14  of gestation.
                                     9-29

-------
     In  this  study,  no adverse effects were  observed  in  either the mother or




the  fetus  of the New  Zealand  White  rabbits;  however,  in  CD-1  mice, clinical




signs  of  toxicity  (weight loss  of >1 g  in  24  hours)  were observed  in the




mothers  in all  treatment periods at  the 120 rag/kg dose.   Maternal weight gain




during  the entire  treatment  period  and  during  pregnancy  were  significantly




reduced  at  the  120  mg/kg  level in day 4-6, 6-8,  and  10-12  groups.  There was




also a  trend  for increased resorptions/litter  in  animals  exposed on days 4-6




and  10-12  at  the 120  mg/kg  level.   At the 120 mg/kg dose  for all treatment




periods, there  was a  significant decrease in mean  fetal weight/litter.  At the




60  mg/kg  level  in  animals  exposed   on  days  8-10,  there  was  a significant




reduction  in  fetal  weight  in  the absence of maternal toxicity.   The authors




reported a trend for an increase in the number of malformed  fetuses  treated on




days 8-10;  however,  the incidence of this  effect did  not achieve statistical




significance.




     The conclusion  reached by LaBorde et al.  (1982) was that ECH administered




intravenously in  mice  produced embryo/fetal  toxicity and  possibly a  slight




increase in  malformations  at  maternally  toxic  doses.  At  the 60 mg/kg level,




in  animals treated   on days  8-10,   fetal  weight reductions  occurred   in  the




absence  of maternal toxicity;  therefore,  it was concluded  that ECH may pose a



hazard specific  to the developing  conceptus.




     Verret (1974)  investigated the  toxic  and  teratogenic  effects of ethylene




chlorohydrin (ECH) in the  developing  chick embryo.  ECH (source and purity not




reported) was administered via  the air cell during a p re-incubation period (0




hour) and  after 96  hours  of  incubation  at levels  equivalent to  10,  25,  50,




100, and 200  mg/kg.   The  control  groups were  treated with the vehicle (water)




or  left  untreated.   One hundred  eggs  were  used per group.   Ethylene chloro-




                                     9-30

-------
hydrin was  found  to  be toxic in this system with significantly increased mor-




tality (no hatch) at  levels >25 mg/kg at  the 0 hour  exposure, and  at levels



>12.5 mg/kg  at the  96 hour  exposure.   Statistically significant increases in




the  number  of structural anomalies were observed at two dose  levels  (50 and




100  mg/kg)  at the 0  hour exposure,  and  at four dose levels (12.5, 25, 50, and




100 mg/kg) at the 96 hour exposure.  The significance of these observations in




terms of mammalian effects  is not known, however, since  teratogenic effects in




chick embryos may not be predictive  of mammalian effects.








9.3.2.    Reproductive  Effects.     The  Carnegie-Me lion  Research  Institute




(Snellings  et al. ,   1982a)  conducted a one-gene rat ion  study  evaluating  the




reproductive  effects  of inhalation exposure  to ethylene oxide.   Thirty  male




and  female  Fischer-344  rats were continually exposed  to 10,  33,  and  100 ppra




ethylene  oxide with  the  control animals  exposed to  filtered air.   Prior to




cohabitation,  all groups  were  initially   exposed   to  ethylene  oxide for  6




hours/day, 5  days/week for  12  weeks.   After 1  week of cohabitation,  females




with  vaginal plugs  were  removed,  and  the other  females were  rotated   to  a




different male  to allow for mating  for another week.    At the  end  of 2  weeks




all  male  and female animals were  separated.   The males were  then  exposed  to




ethylene  oxide  for  6  hours/day, 7  days/week  for an additional 3  weeks.   The



females were exposed for 6 hours/day, 7  days/week from day 1  through  day  19 of




gestation.   On  day  20 of exposure,  females not pregnant were sacrificed.   The




pregnant  females  were allowed   to  deliver  and  5  days after  parturition  were




again exposed  to ethylene  oxide for 6  hours/day,   7  days/week  until day  21




postpar turn.
                                     9-31

-------
     The  following  criteria were  used  to  establish  fertility.    If  a  female




produced a  litter,  or if gross  examination  revealed  implantation sites after




staining,  she was  considered  fertile.    Any  female  not pregnant  after  two




different matings was considered  infertile.   If  the  male impregnated a  female




after  the  first  mating,  then  he was considered  fertile; any  male failing to




impregnate a  female  in  two  different mating periods was considered infertile.




By  this  criteria,  females exposed to 100 ppm had a higher incidence of  infer-




tility after  mating  with  a  male of  proven  fertility;  however,  this incidence




did not  achieve  statistical significance.   In  the males there was no decrease




in  fertility.  In the 100 ppm group, significantly more  females had lengthened




gestation periods  (time from vaginal plug  to  birth of  litter)  than the con-




trol,  10, or 33  ppm groups.   The control,  10, and  33  ppm groups had  gesta-




tional periods of 22 days,  while  the 100  ppra group had gestations  ranging from




22-31 days  (7/14  rats had  22 day gestation, V1U rats  had 23  day gestation,




3/1^ rats >25 day gestation).   Since most of  the animals  did  not have  exten-




sively long  gestational delays,  it  is not  clear  whether  this  lengthening  of




gestation represented a true adverse  biological effect.




     In  this  study  (Snellings et al. , 1982a),  the number of pups was signifi-




cantly reduced with  a  decrease  in the  number of implantation  sites at the 100




ppm level.   Of the  surviving pups,  however,  there was  no  effect on survival




after parturition.   In  the  parental  generation,  there  were no adverse effects




on body  weight or organ histology  (testes, epididymides,  accessory sex glands,




cervix,  uterus,  ovaries,  oviducts,  mammary  tissues).   In  the  F..a generation,




-25% of  the  animals  suffered  from sialoacryoadenitis virus  infection but this




infection appeared to be unrelated to the  ethylene oxide  exposure.
                                     9-32

-------
     It was  concluded from this study  (Snellings  et al., 1982a) that ethylene




oxide administered  to rats by  inhalation  has the  potential  to disrupt repro-



duction  by causing  an increased  incidence of  embryolethal  effects; however,




this embryotoxic effect was only observed  when the animals were exposed to the




highest dose  (100  ppm) and not at  the  lower doses  (10  or 33 ppm)  of ethylene




oxide.



     Snellings  et  al. (1982b)  have  also  examined teratological  aspects by




exposing  rats  by  inhalation  to  10,   33,  or  100  ppm   ethylene  oxide  for  6




hours/day  on  days  6-15 of gestation.   Exposure to  100 ppm caused depression of




fetal weight  gain,  but did not result in  fetal death  or abnormalities other




than  variations in  ossification  of  sternebrae  and  distal  thoracic vertebral




centra.








9.3-3.     Testicular  Effects.   Hollingsworth et  al. (1956)  investigated  the




acute and  chronic  toxicity of ethylene oxide  in a  variety  of animal species.




Positive  responses  related specifically  to the male  reproductive  system were




observed  in  hamsters  and  rats.    Eight guinea  pigs were  exposed  to  357  ppm




ethylene oxide  (commercial  grade  ethylene oxide,  97-98.6?  pure by  weight)  and




received  123  seven  hour  exposures over  a  176  day  period.    There was  only




moderate  growth reduction in  the  males;  however,  appreciable degeneration of



testicular histology  was  noted.   In another  phase  of  this  experiment,  both




rats and  hamsters were  exposed to  204 ppm ethylene oxide,  7 hours/day,  in




122-157 exposures given over an experimental  period of  176-226 days.   A slight




but not statistically  significant  decrease  in  testis weight  of rats and guinea




pigs was observed.   In rats,   there was histological evidence of  degeneration



of the testicular tubules.




                                      9-33

-------
     A  recent  study   sponsored  by  NIOSH  described  the  effects  of  inhaled




ethylene oxide on  semen production  in  Cynomologus  (Macaca  fasicularis)  monkeys




(Lynch  et  al. ,  1983).   The monkeys  were exposed to 50  and 100 ppm  ethylene




oxide,  (toion Carbide, 99.7?  pure)  7 hours/day, 5 days/week  for  2  years.   In




the  preliminary  range-finding  study,  only  two animals  per group  was  used.




Testicular  weights were  diminished in  animals exposed  to  100  ppm  ethylene




oxide but were only marginally decreased  in  those  exposed  at the  50  ppm level.




Similar  decreases  in  epididymal  weights were  reported.    Sperm  motility  was




significantly  reduced at  the 50  and  100  ppm  level,  both  in  terms  of  the




percentage of  motile sperm and  the  ability of  the  sperm  to  travel  a  given




distance in  a given  time  (drive range).  In the preliminary study,  the  sperm




concentration was  decreased  at the 50  and   100  ppm  levels.   In  a  subsequent




study with  larger numbers of  monkeys per  group  (8  or 9),  the same  types  of




adverse  testicular effects were  observed.    In this study,  there  was a  30%




decrease in  sperm  concentration,  30% reduction  in  motile sperm, and  a  3-  to  4-




fold decrease in distance  traveled  when the  animals were exposed  to  50 and  100




ppm  ethylene oxide;  however,  there  was no  effect  on  sperm head  morphology




(Lynch et al., 1983).




     In another study relevant to the effects of ethylene  oxide on the testis,




radiolabeled  ethylene  oxide  was detected  in  auto radiograms of  mouse gonads




(epididymis  and  testis)  20 minutes after intravenous  injection  (Appelgren  et




al. ,  1977).   Radioactivity was found  in  the epididymis up  to  24 hours  after




injection.     The  results  of the   dominant  lethal  mutagenicity   test  were




negative, although  inadequacies  in  this  study  prevent  a  firm conclusion from




being made  (see  Section 9.4).   This study  is  relevant to  testicular  effects




because it establishes that ethylene oxide has access to  the gonads.




                                     9-34

-------
9.3.1*.    Adverse Reproductive Outcome in Humans.  There is little  information




on  the  effects  of ethylene  oxide  on the  human  reproductive  system.   In  one



study a  comparison  was made between  the health  of 37 male employees involved




in ethylene oxide production with 41  men who  worked in other production units




(Joyner, 1964).   This  study evaluated many health endpoints including genito-




urinary  problems.   The  mean exposure period  was  10.7  years with  a general




level of exposure on  the  order of 5-10  ppm.   The  range  of exposure  levels




varied  from  0-55  ppm.   The  health  survey  of the  workers considered  the




following  information:   1)  the number of sick days  taken  in  a  10 year period




with  information on the  etiology and duration  of the  illness, 2)  any medical




diagnosis  entered   into   the medical  records  and  confirmed  by   an  outside




physician,   3)  any  visits  to  the  Medical  Division  related   to  respiratory,




gastrointestinal, or genitourinary  problems.   In this study there was a higher




incidence  of  chest  abnormalities   and  a  higher   incidence  of   absenteeism




attributable to  gastrointestinal and genitourinary causes;  however,  the higher




incidence  of  absenteeism  was attributable  to  a  single  individual in  each




category.   Therefore,  it  was concluded that  long  term exposure  to ethylene




oxide had  no adverse  health effect on the  men  involved  in ethylene  oxide




production.   Since this  study  did not deal specifically with  reproductive




health  problems,  it  is  of limited  value  in  determining  the  potential  of




ethylene oxide to cause adverse  reproductive effects.




     A study by  a  Russian investigator (Yakubova et  al. ,  1976)  reported  that




female workers  involved  in  ethylene oxide production  experienced a number of




gynecological and obstetrical  problems.   Tnese  problems included  diseases of




the  cervix,  Inflammation of the  uterus,   obstetric  anamnesis  (this  word,  as




well  as others  may  have  been  incorrectly  translated),   hypertonic  disease,




                                     9-35

-------
anemia,   toxicosis,  and   shortened  pregnancies.    These  observations  were

reported  in  an anecdotal  manner  with  no  presentation  of  actual  data  or

description of  methodologies.  This  study  is  therefore  of little value in the

scientific review of adverse reproductive effects.

     Holmberg (1979) and Hotaberg and Nurminen (1980) reported case studies of

a mother  exposed to a variety of  organic solvents.   These  studies describe an

adverse  reproductive  outcome in  a  woman  exposed  to alkylphenol and  dyes as

well as ethylene oxide.   It is not clear whether the two articles describe the

same woman  or two  different women.   Both  reports describe an  infant born with

hydrocephalus and  Holmberg  (1979) described a  child  with additional malforma-

tions  (cleft  palate,  double uterus,  polydactyly).    These  reports are  not

useful in establishing causal relationship  between ethylene oxide exposure and

congenital  malformations  because  ethylene  oxide  was not the  only  chemical

involved  and  because a  larger  population  size  would  have  to be  evaluated

before such an association  can be established.

     An  epidemiology  study  has  been  conducted concerning  the  effects  of

ethylene  oxide  exposure on  pregnancy  outcomes  in   nursing personnel.    This

report is the only one which adequately evaluates the possible causal associa-

tion between  ethylene  oxide exposure and  adverse human  reproductive effects,

and has been  reviewed in depth by an  Environnental  Protection Agency epidemi-

ologist  (Margosches,  1983).   In  cooperation  with  the  Finnish  investigators,

the data  have been critically analyzed and  reviewed.   The following  is  the

text of this evaluation:
          "In November,  1982,  K. Hemminki  et al.  published  a  study  of
     "Spontaneous  abortions  in  hospital  staff  engaged  in  sterilizing
     instruments with  chemical  agents"  in  the British  Medical  Journal.
     This study, encompassing all Finnish sterilizing  staff at that time,

                                     9-36

-------
claimed  adjusted  spontaneous  abortion  (s. a.)  rates  of  16.7%  for
"exposed"  and 5.6/5  for nonexposed  pregnancies among  these   staff.
The  report singled  out ethylene  oxide,  glutaraldehyde, and  formalde-
hyde use and  suggested  concentrations as low as 0.1-0.5 ppm ethylene
oxide   might  have  been  associated  with  adverse  outcomes.    In
particular,  among hospital-discharge-corroborated  pregnancies,   the
ethylene-oxide-exposed  s.a.  rate (22.6)  significantly  exceeded  the
control  s.a.  rate (9.2).   Also  rates  among  all pregnancies  exposed
to  ethylene  oxide or  to  glutaraldehyde differed significantly  from
rates among pregnancies not exposed.

     This  study  encompassed  staff  employed   in 1979  at  hospitals
throughout  Finland  (including   tuberculosis   sanitoria  and   mental
hospitals).    It  was  a  cohort  study  looking   at  past   events;
determination of exposure  status was based on  the  responses  to  two
questionnaires.     The  unit  for most  statistical  tabulations  and
analyses was  the pregnancy;  while not  uncommon  in  the   literature,
such  a  basis cannot   take  into consideration  the  relatedness of
sibling births or repeated  miscarriages  of a single  woman.

     The  cohorts,   sterilizing  staff and  controls  [hereafter also
called group  (1) and (2)],  were   identified by the head nurses at  the
study  hospitals.   The  former  were named  in  response  to a  first
questionnaire  that  also  queried  chemical  sterilizing  agent   use
history  at each hospital.   The  latter were obtained  as  members of
cluster  samples  from  auxiliary nurses  in  departments (not  including
chemical sterilization, x-ray, or  surgery) at  the time of distribu-
tion  of a  second  questionnaire  that   focussed  on pregnancy  and
employment history.   The investigators obtained  a  very  high  return
rate  (92%  among  sterilizing  staff,   91&  among  auxiliary   nurse
controls)  and studied the 645 (63$) ever pregnant women among  steri-
lizing  staff  and the 574  (55%}   ever pregnant  women among controls.
The 17 male sterilizing staff were not studied.

          While   the  study  population  was selected  on  the  basis of
hospital employment as sterilizing  staff  (1)  or  non-sterilizing-
staff auxiliary  nurses  (2),  Dr.  Hemminki classified pregnancies of
each group  (1)   member  according to  likelihood  of  exposure and  the
agent(s) present in order  to make  finer comparisons.  He considered
all pregnancies  occurring after  the first  use of ethylene  oxide at a
hospital to  be   exposed to ethylene  oxide  unless an  individual did
not work  at  the hospital  during a  particular pregnancy; similarly
for glutaraldehyde and  formaldehyde.   This was a fairly conservative
classification.   This  study  design  precluded  the examination  of the
questions  whether   spontaneous   abortions  were   related   to  an
individual's ever having been exposed to an agent.

     Another limiting factor  of  the  study design was the characteri-
zation  of  individual  exposures  in  purely qualitative terms.   Dr.
Hemminki  believes  that  typical exposures  have   averaged  <1  ppm
(measured  by  gas-tight syringes).   He  bases this  belief  on  papers
                             9-37

-------
     published  by  colleagues  at  the  Institute  of  Occupational  Health
     covering a 3-year period overlapping the close of  the  study period,
     on the  unchanged  instrumentation  of ethylene  oxide use  over  its 20
     or  so years  in  Finland,  and  on  the  measurement method's  1  ppm
     detection  limit.     He  did  not,  however,   make  any  unwarranted
     inferences   regarding    possible   dose-response    relationships.
     Nevertheless,    he   did  find   sizeable   differences   in   adjusted
     spontaneous  abortion   rates   in  both   nurses  and   sterilizing
     professionals  (these  are  2  education  levels  of sterilizing  staff)
     between ethylene-oxide-ex posed and non exposed,  age-adjusting (£30)
     the   rates  among   discharge-registry-identified   pregnancies,   the
     ethylene oxide-exposed s. a.  rate (16.1) also  exceeded  the rate (9.^)
     in control pregnancies but no longer significantly.  (Certain  of the
     pregnancies occurring during  1973 to 1979  could  be cross-identified
     through a national hospital discharge register and  parallel analyses
     were  carried out  on  this  set and  the  total  questionnaire-obtained
     set. )

          On the whole,  this study and its report paid close attention to
     the  possibilities  and  consequences of  such   typical  epidemic logic
     afflictions  as  reporting  and  recall  bias.    Additionally,   the
     methodology for  statistical analysis,  based  on  rates  adjusted  for
     such  concomitant  variables  as age,  parity, and  decade  of pregnancy
     by logistic regression,  is sound (although  there  may be good reasons
     to investigate  a  finer  categorization of age).   While  a "per woman"
     analysis,   the  analytic  methods   for  incorporating an  individual's
     pregnancy history  have not yet been  perfected  or standarized.

          Unfortunately,  the  investigators  introduced   a possible  source
     of bias through telling the supervisory nurses (who identified group
     (1) and selected group  (2)) the purpose of the study,  including  the
     names  of  the   agents of  interest.   Another  shortcoming  is  the
     impreciseness  with  which hospital exposure history was determined.
     Finally,  although  the authors   planned  a   priori  to  investigate
     relationships between ethylene oxide and  spontaneous  abortions  and,
     possibly,    other    adverse   pregnancy    outcomes,   the   underlying
     relatedness of  multiple pregnancies and  of certain of  the analyses
     (e.g.,  regroupings  of  the  same  pregnancies   to  look  at  different
     exposures) dilute  the strength of any associations  perceived in  this
     study.  Notwithstanding these limitations,  this work is sufficiently
     suggestive to  support  further  study of  the   possible  associations
     between ethylene  oxide exposure  and  adverse  pregnancy  outcomes or
     other reproduction effects."  (Margosches,  1983).


9.3.5.    Summary  of Teratogenicity and  Reproductive  Toxicity.   The potential

of ethylene  oxide  to  cause  teratogenic  or  adverse reproductive  effects  has

been  examined  in  four  animal  species  (mouse,  rat,  rabbit,  monkey) by  two

routes of administration (inhalation and  intravenous)  (Table 9-3).

                                  9-38

-------
                            TAK.E 9-3




Sumnary  of  Studies on Tera togenic ity  and  Reproductive Toxicity
Type of Route of
Study Administration
Te ra to logy iv



Teratology iv



Teratology/ Inhalation
Reproduction








Teratology/ Inhalation
Reproduction











Dose Level and
Species Time of Exposure
CD-1 mouse 0, 75, 150 rag/kg
day t-f>, 6-8, 8-10
or 10-12 of gestation

New Zealand 0, 18, 36 mg/kg
White rabbit day 6-9 of gestation;
0, 9, 18, 36 mg/kg
day 6-1t of gestation
Sprague- 150 ppm, 7 hr/day:
Dawley CD day 7-16 gestation,
rat day 1-16 gestation,
3 weeks pregestatlon
plus day 1 -1 6
gestations




New Zealand 150 ppm, 7 nr/day:
Kiite rabbits day 7-19 gestation,
day 1-19 gestation











Findings
1. Developmental toxic ity
at or near dose level
which produced maternal
toxlcity (150 mg/kg)
1. Developmental toxicity
only at levels which
were maternally toxic
(36 UK/kg, day 6-1 "4)
Te ra to logy :
1 . Retarded fetal
development
Reproduction :
1 . ffetemal toxicity
2. Increase In trau terine
mortality



1. No tera togenic or
reproductive effects












References
LaBorde and
Klmmel, 1980


Jones-Price
et al. , 1983


Battelle Pacific
Northwest Labora-
tories (NI031
210-80-0013)
Hackett et al. ,
1982




Battelle Pacific
Northwest Labora-
tories (NIOSH
contract No.
210-80-0013)
Hackett et al. ,
1982







Comments








Te ra to logy :
1. Inadequacies
a) no maternally
toxic dose
b) dose-response
not determined
Reproduction :
1 . Inadequacies
a) dose -response
not determined
Te ra to logy :
1 . Inadequacies
a) no maternally
toxic doses
b) no develop-
mental toxic
doses
c) dose-response
not determined
Reproduction
1. Inadequacies
a) do se -response
not determined

-------
                                                                        TABLE 9-3  (oont. )
jr
O
Type of Route of
Study Administration Species
•One genera- Inhalation Fischer 31t
tion reproduc- rats
tion














Teratology Inhalation Fischer 3m
rats

Chronic Inhalation Guinea pigs
tox ic ity
(nale
reproduction)
Dose Level and
Time of Exposure
0, 10, 33, 100 ppm
12 wks prior to Dating,
6 hr/day, 5 day/wk.
During gestation -
days 0 through day 19.
During lactation -
days 5 through 21.











0, 10, 33, 100 ppm
6 hr/day, days 6-15
of gestation
357 ppm, 123 7-hr
exposures In 176
days


Findings
1. No difference In F
fertility. No FQ
toxic ity.
2. No adverse effects on F
survival, growth rate,
or lactation.
3. Adverse reproductive
effects at highest dose,
100 ppm.
a) increased gestatlonal
length
b) decreased litter size
c) decreased implantation
sites (i.e., decreased
fecundity)
d) decreased fetuses/
implantation sites
(embryo lethal)
1 . Depression of fetal body
weight (slight)

1. Tubular degeneration of
test with replacement
fibrosls


References Comments
Carnegie -Me lion
Research
Institute, 1979
(Snelllngs et al. ,
1982a)*












Snelllngs et al. , 1. htatemal body weight
198lb not monitored during
treatment
Hollingsworth
et al. , 1956'-


                                                         20M ppm,  122-157
                                                         7-hr exposures in
                                                         176-226  days
1.  Slight  decrease  in  testes
   weight,  not  statistically
   significant.





Testicular Inhalation
tox ic ity





Rats 201 ppm, 122-157
7-hr exposures In
176-226 days


Cynomologus 50, 100 ppm,
monkeys 7 hrs/day for 2
years




1.


2.

1 .

2.

3.
1.

Slight decrease in testes
weight, not statistically
significant.
Testes: small, slight
degeneration of tubules.
Decreased testicular Lynch et al. , 1983
weight.
Decreased sperm
concentration.
Decreased sperm raotllity.
No change In sperm
morphology.
        •A variety  of experimental protocols were utilized;  only  those  which  provided positive  information on reproduction effects are noted  here

-------
                                                                         TAELE 9-3 (cent. )
I
-Cr
Type of
Study
Medical
survey of
workers




Medical
survey of
workers





Case study



Epidemiology
study






Route of Dose Level and
Administration Species Time of Exposure
Occupational 37 male Mean exposure time:
exposure workers 10.7 years.
General levels:
5-10 ppm



Occupational 282 female <0.2-0.3 mg/m3
exposure production
workers
259 female
management
coworkers
100 females
controls
Occupational Pregnant
exposure women


Occupational Pregnant <1 ppm
exposure women







Findings
1. No observed Increase
In male reproductive
disorders.




1. Gynecological dis-
orders, spontaneous
abortions, toxicosis,
decrease birth weights.




1. Infant with hydro-
cephalus


1. Ethylene oxide exposure
associated with an
Increase in spontan-
eous abortion





References Comments
Joyner, 1961 1. Small sample
size
2. Study did not
evaluate
ffertl lity or
test leu la r
function.
Yakubova et al. , 1. Difficulties in
1976 translated material
2. Little information
provided en experi-
mental design.
3 . Mu It ip le ex posu re s
to noise and high
tempera turea
Holmberg, 1979; 1. Mother exposed to
Halm berg and multiple chemicals
Nurmlnen, 1980 2. Only one Infant
studied
Hemmlnkl et al. , 1. Possible bias
1982 introduced by
supervisors
who categorized
participants in
this study
2. Limited exposure
data
        Teratology
                        In travenous
                                          CD-I  mouse
0, 60, 120 mg/kg on
days ")-6, 6-8, 8-10,
or 10-12 of gestation
                                                                                     Maternal toxic Ity at
                                                                                     120 mg/kg  fbr all
                                                                                     treatment periods
                                                                                     Embryo toxic Ity at 120
                                                                                     mg/kg for all treatment
                                                                                     periods and at 60 mg/kg
                                                                                     on days 8-10  (fetal
                                                                                     weight reduction).
LaBorde and  Kimroel,
1980

-------
                                                                        TAELE 9-3  (cent. )
I
-t
Type of
Study
Teratology





Teratology
and Toxic Ity




Route of Dose Level and
Administration Species Time of Exposure
Intravenous New Zealand 0, 9, 18, 36 rag/kg on
white rabbits days 6-14 of gestation




Air cell thick enbryo 0, 10, 25, 50, 100 or
Injection 200 ng/kg at 0 hour
incubation ; 0
5, 12.5, 25, 50 or
100 mg/kg at 96
hours incubation

Findings
1 . Ho effect on anther or
fetus




1. Ovo-toxic at levels
^25 rag/kg at 0 hour,
and >12.5 ing /kg at
96 hours.
2. Teratogenic to chick
embryo

References Comments
LaBorde et al. , 1. Inadequacies
1982 a) no maternally
toxic doses
b) no develop-
mentally toxic
doses
Verrett, 197« Uncertainties in
extrapolating avian
developmental effects
to those of manuals


        •Original study  performed by Carnegie-Me lion Research  Institute (Bayes, 1979); later published aa Snelllngs et al. ,  1982a

-------
     In a  teratology  study,  Hackett et  al.  (1982) reported  that  rats exposed




to a  single  150 ppm dose  of ethylene oxide  displayed  both  maternal  toxicity



(decreases in  food  consumption,  decreases in body weight,  increases  in  kidney




and  spleen  weights) and  toxicity  to  the developing  conceptus (increases  in




resorptions,   decreases  in fetal  weight, decreases  in  crown-rump  length,  and




increases  in   incomplete  skeletal  ossification).    Similar  effects  were  not




produced in rabbits exposed to 150 ppm ethylene oxide in this study.




     LaBorde and Kimmel (1980) administered 75 and  150  mg/kg ethylene oxide to




pregnant CD-1  mice  for several gestational  intervals.   The  animals  displayed




signs of maternal  and fetal  toxicity  at the highest dose  level.   There  were




maternal deaths with decreases  in the  number  of implants  per litter  and  an




increase in  the percentage  of malformed fetuses/litter.    The  malformations




appeared to  follow a  developmental pattern  and  in  at least  one  gestational




interval  (days  8-10  of  gestation) occurred  in  the  absence  of  significant




observed maternal toxicity.



     Similar  studies were  conducted  by  Jones-Price  et  al.   (1983)  on  the




effects of  18 and  36  mg/kg  ethylene  oxide  administered  intravenously  to  New




Zealand  rabbits.    Significant  maternal  toxicity  (decreased weight  gain)  was




observed in addition to embryotoxicity observed  in  the  offspring (decreases in




the   number    of    live   fetus/litter,   increases   in    the   number   of




resorptions/litter).    No  embryotoxicity  was  observed  in  the  absence  of




maternal toxicity.



     LaBorde   et   al.    (1982)    investigated   the   teratogenic   effect   of




intravenously  administered ethylene chlorohydrin (ECH), a  reaction product of




ethylene oxide, in  CD-1 mice and New Zealand  rabbits.   No  adverse  maternal or




embryotoxic effects were  produced  in  the rabbits.   In  the  mice  at  the highest

-------
dose  (120  mg/kg),  however,   severe  maternal weight  loss  with   increases  in




resorptions/litter and decreases  in  fetal weight  were  observed.   At  the  60




mg/kg  level,  with exposure  on  gestational days  8-10,  there  was significant




fetal weight loss in  the  absence  of  maternal  toxicity;  therefore,  the authors




concluded  that  ECH may be a specific  hazard  to  the developing  conceptus  at




this dose  level.   ECH has also been  reported  to produce  adverse  effects  in




developing  chick  embryos  (Verrett,  1974).     Structural  abnormalities  were




produced by 12.5-100  mg/kg of ECH when the egg was incubated with the chemical




for up to 96 hours.




     In a  one-gene ration  study  (Snellings  et  al. ,  1982a),  female rats exposed




by inhalation to  100 ppm  ethylene  oxide had  a higher  incidence of infertility




with indications  of a  longer gestational  period.   There was a decrease in the




number of pups produced by mothers exposed to  100  ppm ethylene oxide,  as well




as a  decrease  in  the  number of  implantation  sites.   However,  there  were  no




significant signs of  toxicity  in  the  mothers (no decreases  in body  weight  or




changes  in  organ histology).    The  sane group  (Snellings  et al. ,  1982b)




observed lowered  fetal  weights,  but not a substantial  level of malformations




in response  to  100 ppm ethylene  oxide administered   to rats by  inhalation  on




gestation days 6-15.



     Adverse effects on the  testis resulting  from ethylene  oxide exposure have




been  reported   for  the  hamster  and   rat  (Hollingsworth   et  al. ,   1956)  and




Cynonclogus  monkey  (Lynch et  al. ,  1983).   Hollingsworth  reported  testicular




degeneration  occurring  in  hamsters and   rats  exposed  to  ethylene  oxide  by




inhalation  (204-357  ppm).   Lynch et  al.  (1983)  reported  adverse  effects  on




sperm concentration and notility, but  not  morphology,  in  Cynomologus monkeys.




The monkeys  in this study were  exposed  over 2 years to  50 and  100 ppm ethylene




                                     9-44

-------
oxide by  inhalation.   In  mice,  radiolabeled ethylene oxide has  been found to




persist in  the  epididymis up to 24 hours  after a single  injection  (Appelgren




et al., 1977).



     Very  little information  exists  on  the  adverse  reproductive  effects of




ethylene  oxide  in  the human.   Medical surveys have described  effects ranging




from no  adverse reproductive  outcome  (Joyner,  1964)  to a variety of adverse




outcomes  (Yakubova  et al.,  1976).   The study  by  Joyner  (1964)  is  inadequate




because it  did  not  deal specifically  with  adverse  reproductive outcomes.   The




report by Yakubova  et  al.   (1976)  was presented  in  an anecdotal  manner and




therefore  is  of little scientific value.    A  case  report described by Holmberg




(1979)  and  Holmberg and  Nurminen  (1980)  indicated that one women  exposed  to a




variety  of  substances,   including  ethylene  oxide,  produced   an  infant  with




multiple  defects and  hydrocephalus.  However,  because  of the  multiple chemical




exposures   involved,  this  study  is  of  little  value  in  establishing  the




potential of ethylene oxide to cause adverse effects.




     A  recent   epidemiological  study  has   been   conducted   evaluating  the




pregnancy  outcome of  nursing personnel exposed  to ethylene oxide (Hemminki et




al., 1982).   Although  there were  problems in  the  study  design and  collection




of data,  the data are  sufficient  to  suggest  an association  between ethylene



oxide  exposure   and   spontaneous  abortion,  warranting  further examination of




adverse pregnancy outcomes.   Additional epidemiology studies would be helpful




to  more   firmly  establish  the  potential  of ethylene  oxide  to  cause adverse




reproductive effects  in humans.




     In   conclusion,    ethylene   oxide  produces   adverse  reproductive   and




teratogenic  effects  in  both  females   (maternal  toxicity,  depression  of  fetal




weight gain,  fetal death,  fetal malformation)  and males (reduced sperm numbers




                                     9-45

-------
and sperm  motility)  if the  concentration  of the  chemical  reaching the target




organ  is  sufficiently  high  or  if  exposure  at  lower  levels  is  sufficiently




long.     Thus,   the   experiments   in  which   ethylene  oxide  was  injected




intravenously  have  produced  more  detrimental  effects  than  the  short-term




inhalation experiments.   Even  short-term  inhalation experiments,  however,  have




resulted in suggestive  evidence  of  detrimental effects.  The  levels needed to




produce  the   developmental  effects  approach  or  equal  the  levels needed  to




produce  toxicity  in  the dams.    The effects  of ethylene  oxide  on  human




reproduction have not  been studied  in  depth,  although  one study indicates that




ethylene oxide  may be associated with spontaneous abortion  (Hemminki  et  al.,




1982).  Future studies are needed to establish this effect in humans.
9.4.  MUTAGENICITY




     Ethylene  oxide  has been  evaluated  for mutagenicity in  several different




systems  including  tests  in  bacteria,   fungi,   higher   plants,   Drosophila,




mammalian cells  in  vitro,  and rodents.   Effects  in humans  are  also reported.




The available  data concerning  the mutagenicity  of  ethylene  oxide are discussed




below and summarized in Tables 9-4  to  9-16.   The reader may also wish to refer




to other reviews of the mutagenic potential  of  ethylene oxide (e.g., Fishbein,




1976, Wolman 1979, Ehrenberg and Hussain, 1981,  and NIOSH, 1981).
                                      9-46

-------
                                  TAH.E 9-4

Summary of Mutagenicity  Testing of Ethylene Oxide:  Gene Mutations in Bacteria
Re f e re nc e
Rannug et al. ,
1976
Test
System Strains
Salmonella/ TA1535
micro some assay
(suspension
assay)
Activation
System
None
Chemical
In ft) r nation
Concentration tested:
0-95.5 mM
Source: Fluka
Results
Strong
positive
response
Comments
1. Ethylene oxide used as a
control.
2. Dose -dependent response.

positive
15-fold
Pfeiffer and   Salmonella/      TA98
Dunkelberg,     microsome  assay  TA100
1980           (plate test)     TA1535
                               TA1537
              None
                           Purity:  Not given

                           Solvent:   Cold  ethanol
                           Concentration  tested:
                           0-200 (moI/plate
                           (0-8.8 rag/plate)

                           Source:  J.T.  Baker
                                    Chemicals  BV
                                    Deventer,  The
                                    Netherlands

                           Purity:  99.7 %

                           Solvent:   Cold acetone
                 increase  in  revertants noted at
                 highest dose compared to negative
                 controls.

             3.   Five plates  used per dose.
Positive     1.   Dose-dependent response for TA1535
                 and  TA100.

             2.   Concurrent negative control values
                 not given.

             3.   Compared  to  lowest dose (20 (jnol/
                 plate),  revertant count at highest
                 dose  (200 uraol) was elevated
                 18-fold  for  TA1535 and 2.25-fold
                 for TA100.

             4.   Between  6 and 10 independent runs
                 were  done in duplicate for each
                 experiment.

-------
                                                                TAELB 9-4  (cent. )
Re fe re nc e
Test
System
Activation
Strains System
Chemical
Information Results
Comments
  Tanooka, 1979 Bacillus subtills
                spores (reversion
                to his*
                pro to trophy)
CD
HA 101
(his met
 leu)

TKJ 5211
(his net
 uv rA10)

TKJ 8201
(his net
 polA151)
                                                 None
Concentration  tested:
27.3% atmosphere  of
ethylene oxide gas
for times ranging
from 5-50 minutes.

Source: Daicide LS gas
        Daldo Oxygen Co.
        Tokyo,  Japan

Purity: 27.3 J ethylene
oxide 72 .1% Freon
Positivel.  Tests conducted in a polyethylene  bag;
response    *) x 10  spores placed on sterile
            filter inside bag.

        2.  Negative control values not provided.

        3.  Revertant values expressed as muta-
            tion frequency (6 x 10"  after
            5 minutes exposure and 8 x 10
            after 50 minutes exposure of HA 101
            and TKJ 5211).

        H.  Lethal and ma tag en ic effects were
            enhanced In the polA strain;  TKJ
            8201 was 10x no re sensitive than
            HA  101 and TKJ 5211.

-------
9.1.1.    Gene Mutation Studies.








     9.1.1.1.   PROKARYOTIC  TEST SYSTEMS  (Bacteria)  —  Several  investigators




have shown that ethylene  oxide  causes point mutations in bacteria (Table 9-1).




Ethylene  oxide  is  a  very  effective  sterilant  for  products  that would  be




damaged  by  other  sterilization  methods.   Bacillus  subtilis  var. niger  is




commonly  used  to  monitor the  effectiveness  of ethylene  oxide sterilization.




Jones and  Adams  (1981)  found that treatment of spores of  these  bacteria with




Pennges (12:88 ethylene oxide-Freon  mixture by  weight)  for 5 minutes increased




the number of  colony variants by five fold over the  spontaneous  level.   Forty




aberrant  isolates  (out  of 125 found) were  plated  five times in succession;  of




these,  11  reverted  to  typical appearance,   12  changed   to  other  atypical




appearances,  and  17 remained stable.   Although the  changed colonies were not




well-defined  genotypically  these  data  suggest  that ethylene  oxide  induced




mutations in the surviving spores.




     In  a study  by  Rannug  et  al.   (1976),  ethylene oxide was  chosen  as  a




positive  control  chemical  in   tests  of  other  chemical   substances  in  the




Salmonella/microsome  assay.    In  this  study,  strain  TA1535  was   exposed  to




concentrations of  ethylene oxide (purity not reported) ranging  from 0-95.5 mM



in  a  suspension  test without  addition  of an  exogenous  mammalian metabolic




activation  system  (Table 9-1).     A   statistically  significant  dose-related




response was observed (Figure 9-1) where the maximum killing was -20%.



     In  another   Salmonella  assay,   Pfeiffer  and  Dunkelberg  (1980)  exposed




strains TA98,  TA100,  TA1535, and  TA1537  to  concentrations of  ethylene  oxide




(99.7% pure  diluted in cold acetone)  ranging  from 0-200 p.M  (0-8.8 mg/plate)




(Table 9-4).   Between 6  and 10 trials  were  performed in duplicate.   A clear




                                      9-19

-------
  108
Q.
C/5
             y = 0.996x +6.02
             R = 0.9897
             p <0.01
                         24         36        48        60        72

                             ETHYLENE OXIDE CONCENTRATION, mM
84
96
        Figure 9-1.  Mutagenic response of Salmonella typhimurium strain TA 1535 exposed
        to ethylene oxide.

        Source: Rannug et al. (1976)
                                        9-50

-------
dose-dependent response  was observed for  the  base-pair substitution detecting



strains TA100 and TA1535 but not  for  the  frameshift detecting strains TA98 and



TA1537.   This result  is consistent  with  responses observed  with  other alky-



lating agents.



     Tanooka  (1979) exposed  spores  from three  different his~ Bacillus subtilis



strains  to an  ethylene oxide  gas  mixture (Daicide  LS,  comprised of 27.3%



ethylene oxide and  72.7$ freon gas)  in a  plastic  bag  (Table 9-4).   Histidine-



independent  revertants  were  selected  after  treatment;  a  repair-competent



strain and  a  uvrA repair-deficient strain  were  treated  for times ranging from



5-50 minutes.   Exposure-related  revertant  frequencies  were observed  for both


                             -6                                     -4
strains  (ranging  from  3 x  10   after 5 minutes  exposure to 2 x  10   after 50



minutes  exposure).   In  a  similar  experiment  conducted with  a  polA strain  a



significantly higher dose-related revertant rate was reported compared to that



obtained  with  the  repair  competent   and  uvrA   strains.     The  revertant



frequencies corresponding to  5  and 40 minutes of  exposure  were  about 8 x 10~



and  3  x  10  ,  respectively.   A  similarly elevated sensitivity of the polA



strain was  observed  for ethylene oxide-induced  toxicity.   No data were given



for negative controls for any of  the  strains.  The his+ revertants  produced in



the repair-competent strain  exposed to  ethylene oxide gas  for 30 minutes were



characterized, and 85/t of them were found  to contain suppressor  mutations; 15/f



were  true  revertants   as   measured  by  cotransformation   of  hisB+  with  the



neighboring trpC+ marker using  DNA extracted from each  his+ colony.   Although



this study was not conducted  using a "standard"  assay system, it does indicate



that ethylene oxide is  mutagenic  in B.  subtilis.



     The  positive responses  in   these  tests show  that  ethylene oxide  causes



genetic  damage  as  evidenced  by  induction of  mutations  in bacteria.    The



                                      9-51

-------
studies  described  below  show  that ethylene  oxide causes  genetic damage  in




higher organisms also.








     9.4.1.2.   EUKABYOTIC TEST  SYSTEMS








     9.4.1.2.1. Plants —  Kolmark  and  Kilbey (1968) studied  the  induction  of




ad+  revertants  in  Neurospora  crassa   strain   K3/17   (macroconidia)   after




treatment  with ethylene  oxide  (source  and  purity  not given).   Five  doses




ranging  from  0.0015-0.15M  were   employed,  but   the  corresponding  mutation




frequencies were not  reported  (Table  9-5).  The  purpose  of the  study  was  to




investigate the kinetics of nutation induction.   In this study,  ethylene oxide




was found to be 15-21  times more effective as a mutagen than diepoxybutane.




     Migliore et al.  (1982)  tested a series of  aliphatic  epoxides  for  their




ability  to  induce  forward  mutations  in  Schizosaccharomyces  pombe.   Ethylene




oxide treatment in liquid  suspension at concentrations  from 0.5-15 raM resulted




in dose-related  increases  in mutation frequency;  survival was reduced  -60%  at




the high dose.  One hundred-fold increases in nutation  frequency were  noted  at




the high dose  levels compared to the corresponding negative controls both with




and without  metabolic activation by phenobarbitone-induced  mouse  liver  S9  mix




(50.28 +  1.76  vs.  0.59  + 0.22  and  66.21 + 29.44  vs.  0.66 + 0.59 mutations/104




survivors, respectively).   The  ranking  of the  chemical  substances tested with




respect  to  their  relative specific activity was:   epichlorohydrin >  ethylene




oxide  >   glycidol  >   1 ,2-epoxybutane   >  1 ,1 ,1-trichloropropylene   oxide   >




propylene oxide > 2,3-epoxybutane.



     Ethylene  oxide is known to be a very effective mutagen  in higher plants.




l"kny  tests have been performed  in  which  ethylene oxide has been  shown  to  be




                                     9-52

-------
                                                                    TABLE  9-5
                          Sunmary of Mitagenicity Testing  of Ethylene  Oxide:  Gene Natation Teats in Lower Plants (Yeast)
   Reference
                      Test  System
                        Chemical Information
                                                             Results
                                                                                              Co amenta
  Kolmark  and  Kilbey, ad-3A revertants
  1968                In  Neuroapora era333
 I
Ul
   Migllore et al. ,
   1982
Forward ma tat ion 3 at
the ade locus In
Sehizocaacharomycea
Concentration tested:
ranged from 0-0.1  M
(0-6.2 g/i) ethylene  oxide.

Source:  Imperial  Chemical
         Industries Ltd.

Purity:  Not given

Solvent:  Distilled water


Source:  Montedison (Italy)

Purity:  99.70*
                                                        Do ae-re la ted
                                                        positive response
                      1.  Objective  of work was to study
                         kinetics of  mutation.

                      2.  Revertant  values given in Figure of
                         paper as mutation frequencies (i.e.
                         ad+/1o6 survivors).
Do se -re la ted
positive response
ponfce
Solvent:


Water and

DM SO






Without S9
Doae
(mM)
0
0.5
1.5
5
15
Survival

71
99
80
35
100
.78
.19
.3
.14
Mutation
.X10-4
0.
1.
4.
18
66
66 ± 0
89 ± 1
17 ± 0
.77 ±
.21 ±
Freq.
.59
.00
.75
0.72
29.44
Survival
100
100
76.64
100
42.87


With


S9
Mutation
.xlO-1*
0.
3.
7.
14
59 ± 0
32+0
15+0
.33 ±
50.28 +



Freq.
.22
.96
.24
7.62
1.76

-------
mutagenic.  The  results of these studies will not be analyzed in depth.  Most




were directed rautagenesis  tests conducted to generate desirable traits in food



crops.   The results of  two  tests,  in  which plants were treated with ethylene




oxide,  will be discussed  for  illustrative  purposes  (Ehrenberg et  al.,  1956;




Jana  and Roy,  1975).   Ehrenberg et al.  (1956)  administered  several chemical




substances  including  ethylene  oxide (purity not  given) to dry  and presoaked




barley  seeds which  were  subsequently  screened  for sterility  (dependent  on




chromosomal aberrations) and chlorophyll  mutations  (caused  by gene  mutations,




either  chromosomal or  ex trachomro so ma 1)  in  the developing  plants (Table 9-6).




The seeds were exposed to  ethylene  oxide  either as  a gas (dry seeds receiving




80? ethylene  oxide  for 6  days)  or  in  solution [seeds  were presoaked  in 0.12




and 0.03$  (0.27  and 0.07  M) solutions  for  2 hours].   Ethylene  oxide  induced




mutations in a  dose-dependent  manner as  can be  seen in Table 9-6.   A  5-fold




increase  in  lethal  nutations and a  33-fold  increase  in chlorophyll mutations




were observed.




     Jana and Roy  (1975) treated dry seeds of two  varieties  of  rice, IR8  and




Dular,   with 0.1-0.6?   (0.02-0.14  M;  pH 7.0))   ethylene   oxide  (purity  not




reported) solutions at  10°C  for  8 hours.   The seeds  were  sown  and  the  plants



grown and harvested.   Seeds  from single  plants were  collected  and  thoroughly




mixed  to  obtain  a random  sample  of seeds.   These  were then  grown  to  get  at




least 100 plants  from  treated  original seed for  the  next generation.   These




plants were scored for gene  mutations  affecting chlorophyll expression,  and  a




dose-related mutation frequency  was  observed (Table 9-6).   Although  negative




controls were  not reported,  and  the  spontaneous  mutation  frequency was  not




provided, about  three times  as many mutants  were  reported in offspring from




plants receiving the highest  dose compared to those receiving  the lowest dose.




                                     9-54

-------
                                                                       TABLE 9-6
                                  Summary  of Mutagenicity Testing of Ethylene Oxide:  Mutation Teats In Higher Plants
Reference
                        Test  System
                                                Chemical Information
                                                                                     Results
                                                                                                                      Comments
Ehrenberg et al.
1956
Lethal ( chromosomal)
and chlorophyll
(gene) nutations In
barley.
Vhen tested as a gas,  resting
seeds exposed to 80J ethylene
oxide for 6 days.  Wien tested
In solution,  partly preaoaked
seeds exposed to 0.03) and
0.12J (0.27 and 0.07 M)
ethylene oxide fbr 2 h at 20°C.
Postive response      1.   Third  generation progeny not
                          available  for analysis when report
                          written; positive  response nay be
                          due to extra chromosomal mutations.

                      2.   Mutagenic  response  observed after both
                          types  of treatment.

                      3.   Half-life  of ethylene oxide  in water
                          solution la around  100h at 20°C.
%
EtO
0
0.03
0.12
80
%
Sterility
t
5.7
9.5
22.1
* 2nd
generation
ch lorophyll
gene
mu ta t Ion s
0.05U
0.20
0.75
1.8
No. spikes
analyzed
15,861
2,510
1,872
989
treatment
condition
None
Solution
Solution
Gas

-------
                                                                           TABLE 9-6  (oont. )
        Reference
                                Test Systea
                         Chemical Information
                                                                                              Results
                                                                                                                               Comments
        Jana and Roy,
        1975
Chlorophyll gene
nutations In rice
(IR8 and  Dular)
Ul
CTv
Concentration tested:
ranged from 0 to 0.6<
ethylene oxide.   Seeds
treated fbr 8 hours at
10«C and pH 7.0

Source:  Bast man Organic
         Chemicals

Purity:  Not given

Solvent:  Not given
   Dose-re la ted
positive response
                                                                           % 2nd Generation Chlorophyll
                                                                                 Gene filiations
                                                                            jtEtO      Dular	IR8
1.  Objective of study UBS to study
    kinetics of mutation.

2.  Revertant values given in Figure
    In text as nutation frequencies.
                                                                            0.1    5.0 + 0.36    5.9 + 0.13

                                                                            0.3    7.0 + 0.37    7.0 + 0.30

                                                                            0.5   12.3 + 0.32   12.0 ± 0.19

                                                                            0.6   11.6 + 0.13   13.1 + 0.16

-------
The  positive responses observed  in  plants are  consistent with  the  bacterial




results and  show that ethylene oxide is rautagenic in plants.








     9.4.1.2.2. Animals —  Ethylene oxide has  also been  shown  to cause  both




gene and  chromosomal mutations in animals.   Bird  (1952)  injected adult  male




Drosophila  melanogaster  (Oregon   K)   with  0.5  and  0.8$  (0.11  and  0.18 M)




ethylene  oxide  to   test  its  ability  to  induce sex-linked  recessive  lethal




mutations  (Table  9-7).   The highest  dose  level approximated  the  LD     The




exact  amount administered and  the purity  of  the  sample  were  not  reported.




There  were  no sex-linked  recessive   lethals  in 494  offspring  of  untreated




flies.   Ten  lethals out of 713  offspring  (1.4$)  and  9  lethals  out of  198




offspring  (4.5$)  were  detected  after treatment  with 0.5$  and  0.8$  ethylene




oxide,  respectively.   The  dose-related  positive  response reported  indicates




ethylene oxide is mutagenic in Drosophila.




     Vfatson  (1966) fed  ethylene oxide  to male Oregon  K Drosophila melanogaster




to compare  the  induction  of sex-linked recessive mutations with  the  induction




of heritable  translocations.   A second objective of this  study was to compare




the  effect  on mutation yield  of  storing  sperm in  seminal  receptacles after




treatment with  alkylating agents.  A  positive  dose-related  increase in  both




endpoints resulted  from ethylene  oxide treatment  (Table 9-7).   For  the  sex-




linked recessive lethal test,  -3% lethals  were  detected at the low dose (0.4$




ethylene oxide) compared  to 1% at  the high  dose (0.7$  ethylene  oxide).   For




translocations  these values  were  =0.28   and   0.7$,   respectively.   Negative




control values were  not given.  Storage of ethylene oxide-treated sperm in the




seminal receptacles  for 6 days had  no effect  on   the  frequencies  of the two




types of genetic damage.




                                     9-57

-------
                                                                          TABLE 9-7

                                     Summary of Mutagenicity Testing of Ethylene Oxide:  Gene Mutation Tests  in  Insects
MD
 I
CO
Test
Reference System Strain
Bird, 1952 Drosophila Orgeon K:
melanogaster adult males
sex- linked
recessive
lethal test













Watson, 1966 Drosophila Oregon K:
melanogaster adult males
sex-linked
recessive
lethal test
and heritable
translocation
test

Chemical
Information Results
Ethylene oxide administered Dose-related 1.
by feeding, inhalation or positive
injection. (Data not pre- response
sented for first two routes
of of administration.) For 2.
injection experiments 0.5-5%
solutions administered to
20 males. Dosages >0.8J
lethal. 0.8J ethylene oxide
killed 50J of treated flies
while 0.5f ethylene oxide
did not affect viability
Source: Not given t No.
EtO Chromosomes
Purity: Not given
0 191
Solvent: O.W saline 0.5 713
0.8 198
Concentration tested: Positive dose- 1.
0, O.Ot, or 0.7* (0, related
0.09, or 0.16 M) ethylene response
oxide

Source: Not given
2.
Purity: Not given
Solvent: Not given 3.

Comments
Objective of experiment was to find most
effective method of administration for
routine testing.

Cannot determine germ cell stage
specificity.






No. %
Lethals Lethals

0 0
10 1.1|
9 4.5
Objective of experiment was to determine
effect of sperm storage in female seminal
receptacle on mutation frequency after
treatment with monofunctional and
bi functional alkylating agents.

Did not observe storage effect for ethylene
oxide with respect to either endpoint.
Cannot determine germ cell stage
                                                                                                            specificity.
                                                        Pre-stored
                                                        Post-stored
                                                                        0.7

                                                                        O.I)
itSIRL

 3.3
 3.6

 7.1

 3.3
 3.1
0.29
0.39

0.69

0.79
0.37
                                                                                                      % Trans.
0.08
0.1

0.1

0.2M
0.12
                                                                        0.7
                                                                                 6.8
                                                                                            0.60
                                                                                                        0.09

-------
                                                                     TABLE 9-7 (cont. )
U1
Reference
Lee, unpublished
Test
System Strain
Drosophlla
me la nog aster
sex -linked
recessive
lethal test
and gonadal
Chemical
In fo rma t ion He su It s
Source: Not given for
unlabeled ethylene
oxide ^H-ethylene
oxide from New
England Nuclear
sp. act. = 2.8
cl /mmo le
Comments
1. Objective of experiment was
the relation of exposure to
alkylatlon of germ cell DNA
to nutational response.

to determine
level of
                                                       Purity:  Not given
                                                                                  Exposure
                                                                             (umole/25  ml vial)

                                                                                    0

                                                                                    0.086

                                                                                    O.U3
    (Dose)
  Alkylation/
Nucleotlde x 1Q-3

     5.58

    22.3
% SLRL

0.12

0.35 + 0.07

0.92 + 0.2

-------
     Lee  (unpublished  data)  conducted parallel experiments with unlabeled  and


 ^H-labeled ethylene  oxide  to  determine:   1) the relationship between  exposure


 and  the  level of alkylation of germ  cell  DNA,   and  2)  the  relationship  of germ


 cell DNA alkylation to mutational responses  in  Drosophila melanogaster  males.


     For  both  the dosimetry  and  genetic  test  treatments,  ethylene  oxide  was


 given  to  the flies by adding 0.7  m& of  cold  water solutions  to  glass fiber


 paper  in  25  mJl scintillation vials   (0.086  or  0.43 (jmole/vial).   Immediately


 afterward 50 males were added to the vials  which  were  then sealed.  Treatment

                                      ill
 was  continued for  24  hours at 25°C.     C-Thymidine was also given to males in


 the dosimetry experiment.   The number of alkylations per nucleotide of DNA  was


 calculated based on  the 3n/ 1^C ratios in  purified  sperm DNA (to determine  the


 number of alkyl groups present)  and  the 1l*C/sperm cell ratio (to determine  the


 amount of sperm cell  DNA  in  the  extraction product).   The genetic data  showed


 ethylene oxide to be an effective mutagen since dose-re la ted increases  in sex-


 linked recessive lethals were  observed  (see Table 9-7).   Using the exposure-


 dose relationship  determined  from the dosimetry  experiments and  the genetic


 data, a doubling dose of 2.3  x 10   alkylations/nucleotide  was calculated.


     These studies show that ethylene oxide  is  distributed  to  the  gonads of a


higher eukaryote (Drosophila)  and causes heritable genetic  damage.




     9.4.1.2.3.  Nfemmalian  Cells in Culture  —  Three tests  have been conducted


 to  ascertain  the   ability  of  ethylene  oxide  to cause  gene  mutations  in


 mammalian cells in culture.   Brown et al.  (1979) reported  in an abstract that


polymethacrylate plastic  sheets and polypropylene  plastic  sheets  and  mesh


 sterilized by ethylene oxide gas adsorbed ethylene oxide molecules which could


be released  later  to exert a  mutagenic effect.   They placed the ethylene oxide


                                     9-60

-------
treated  plastic,   of  unspecified  size,   in  culture  flasks  containing L5178Y




TKV-  mouse  lymphoma  cells  for  3  days.   This  was  followed  by  dilution  in



ethylene  oxide-free  media for  3 days  prior to  selection  using BUdR.   Poly-




methacrylate  sheets  treated  for  18 hours with  pure  ethylene  oxide  were




estimated  to  release  8-40  |ig  ethylene  oxide  (as  measured  by  GC  into  the




flasks,  while  similarly  treated  polypropylene  sheets  and  meshes  released




5-100 |ig  ethylene  oxide.  Although the  spontaneous negative  control mutation




frequencies were not  given,  the  released  ethylene oxide was reported to result




in  a 2-  to  20-fold  increase  in  induced mutation frequency relative  to  the




controls  (Table 9-8).   It  was  not possible  to  evaluate this report critically




because it was presented in abstract form.




     Tan  et  al.  (1981) administered ethylene oxide (Matheson  Co.,  99.7? pure,




Dr.  R.  Gumming,  personal communication) to  Chinese   hamster  ovary  cells  at




concentrations ranging  upwards to 10  mM  in the medium.   Mutations at the HGPRT




locus  were  selected  after  5  hour  ethylene oxide  treatments  both  with  and




without an exogenous  metabolic activation system (39 mix  derived from Aroclor




1254-induced rat livers)  followed by a 16-18 hour recovery period and subcul-




turing for 1 week.  A dose-dependent positive response was obtained at concen-




trations  causing between 10? and 90? cell killing  (Figure 9-2)  both  with  and




without metabolic  activation.   The mutation frequency  at  the highest dose not



resulting in excess toxicity (<80? cell  killing) was roughly  10 times greater




than the  reported spontaneous frequency (see Table 9-8).




     Hatch  et al.  (1982) and  Dr.  S.  Nesnow  (personal  communication,  1983)




exposed Chinese hamster V-79 cells to ethylene  oxide  gas  at concentrations up




to  7500   ppm  and  selected for  ouabain-  and 6-thioguanine  resistant  mutants.




Significant  numbers of mutants were produced for both  genetic markers.  There




                                      9-61

-------
                                                                  TABLE 9-8

                               Summary of Mutagenicity Testing of Ethylene Oxide:  Mammalian  Cells  in  Culture
Reference
Test
System
Activation
System
Chemical
Information
Results Comments
Brown et al., L5178Y TK+/-
   1979        mouse lymphoma
              gene mutation
              assay
                 None
                                  Polymethacrylate (PUMA)
                                  plastic sheets and
                                  polypropylene (PP)
                                  plastic sheets and meshes
                                  sterilized for 18 h in
                                  pure gaseous ethylene oxide.
                                  PMMA retained ethylene oxide
                                  and established concentra-
                                  tions of 8-40 (ig/20 mi
                                  cultured medium (1-5 x
                                  10-5M ethylene oxide).
                                  PP retained ethylene oxide
                                  and established concentrations
                                  of 5-100 ng/ 20 ml in cultured
                                  medium.

                                  Source:  Not given

                                  Purity:  Not given

                                  Solvent: None
                                                2 to 20-fold
                                                induced
                                                mutation
                                                frequency
                                                observed
                                             Presented in abstract.

                                             Chemical concentrations measured by gas
                                             chromatography.

                                             Two ethylene oxide  metabolites also
                                             tested.  At the low, but unspecified,
                                             level tested, ethylene glycol residues
                                             did not produce an effect.  Chlorohydrin
                                             produced residues of 15-30 |ig/piece of PP.
                                             Direct addition of this compound to the
                                             medium resulted in a 2-3 x Induced
                                             mutation frequency.
Tan et al.,
  1981
HGPRT Chinese
Hamster Ovary
cell gene
mutation
assay
Liver S9 mix
from Aroclor
12511-induced
Sprague-
Dawley rats
Concentrations tested
0 to 10 mM
Dose-related
positive
response with
and without
activation
Concentrations and induced mutants
extrapolated from Figure 9-1 of text.

250-300 mutants/106 cells at high dose
both with and without activation compared
to 0-10 mutants/10" cells in negative
controls.

Direct acting rautagen.

Ethylene oxide both cytotoxic and mutagenic

-------
                                                         MUTATION FREQUENCY, HGPRT mutants x 1(r6/clonable cell
i
cri
CO
                   I
                   -1
                   o
                   CD
Q.

T3

CD
Q.
—h

O

3
                   O)
                   3
                   CO
                   00
     in
     CD

     c
                        03
                        to
                        OJ
                        3
     -a
      o
      3
      in
      CD
      O
      —h
      O
      I
      o
      o
                        o
                        CD
                        fD

                        (0

                        O
                        X
                                 O
                                 X
o
o
z
o
m
30

H

O
z

3
                                                                            RELATIVE SURVIVAL, percent

-------
was  a  dose-related  increase  in mutation  frequency.    The  response  at  the




highest dose was  20  times  greater  than negative control rates.   The  work  was




reported in an abstract.




     The studies  by  Brown  et al. (1979),  Tan et al.  (1981),  and  Hatch et al.




(1982)  indicate  that  ethylene  oxide  causes  gene  mutations  in  cultured




mammalian cells.








9.4.2.    Chromosome   Aberration  Studies.    Nkny  studies  have  shown  that




heritable chromosome  aberrations are  induced  in  plants after  ethylene oxide




exposure [e.g., Moutschen et al. (1968) in barley and Mackey  (1968) in wheat].




These  studies  will  not  be  discussed  in  this  report.    Nfost were  directed




mutagenesis  studies  designed to  obtain  desirable variants.   The  ability  of




ethylene oxide  to cause  such  mutations shows  it to be  an  effective clastogen




in plants.








     9.4.2.1.   DOMINANT  LETHAL  TESTS —  Ethylene  oxide causes  chromosome




damage  in both  mammalian germ cells and  somatic  cells  (Tables 9-9  to 9-13).




Ethylene oxide  has been  tested  in  dominant  lethal tests in both rats and mice




and has yielded a positive  response  in each (Table 9-9).   The precise nature




of the  damage causing dominant  lethal effects is not  known, but  there  is  a




good correlation between  chromosome breakage in germ cells and dominant lethal




effects (Nfetter and  Jaeger,  1975).   When  dominant lethal effects are observed




in the  offspring  of  treated  males,  it can be  concluded  that  the  test agent




reached  the  gonads  and  likely  caused  genetic  damage.   Embree et  al.  (1977)




conducted a dominant  lethal  test with  Long  Evans  rats.   Twe Ive-week-old males




inhaled  1000  ppm  ethylene  oxide  for 4 hours  (Matheson  Gas Products, Newark,




                                     9-64

-------
                                   TABLE  9-9

   Summary  of Mutageniclty Testing of Ethylene Oxide:   Dominant Lethal Testa

Reference
Embree et al. ,
1977
















Generoso
et al.,
I960














Test
System
Dominant
lethal assay
in Long Evans
rats














Dominant lethal
assay: male
mice T stock
(Experiment
I) and (101 x
C3H)F,
(Experiment
ID









Mating and
Sacrifice
Each male
placed with
2 virgin
females per
week for 10
weeks. Females
sacrificed on
the 17th day
after first
exposure to
male.







Experiment I:
Mated to 2
virgin (SEC
x C57BDF,
females about
12 weeks old.
Females replaced
when vaginal
plug observed.
Sacrificed 12-
15 days later.

Experiment II:
Mated to 2
virgins from
one of the
following stocks
Chemical
Information
12 week old male
animals exposed to
1000 ppm ethylene
oxide via inhalation
for t hours

Source: Not given

Purity: Not given









Single i.p. injection
of 150 mg/kg.
Maximum volume of
1 ml

Source: Eastman
Kodak Co.

Purity: Not given

Solvent: Double-
distilled
water





Results Comments
Positive response. 1. Animals exhibited toxicity but no deaths
Significant Increase resulted.
in postimplantational
fetal deaths during 2. Pattern of positive response indicates
first 5 weeks of the postmeiotic effect.
experiment


% Dead Implants
Week EtO Control

1 12» 2
2 30* 10
3 30» 1
19 8
5 10" 4
10 9 11
•P<0.05
Positive response 1. i.p. route of administration chosen to
observed for days mimic implanation of medical device.
2.5-11.5. Corresponds
to treated spermatozoa
and late spermatids.
During this period 12
to 31> dead implants
in treated group
compared to 3 to 5J
dead implants in
negative control group.
Little or no difference
in the yield of dominant
lethal mutations in male
postmeiotic germ cells
when mated to females
from different stocks.
T, (SEC x C57BDF,,
(101 x C3H)F,, or
(C3H x C57BDF!.
Sacrifice 12-15
days after
observation of
vaginal plug.

-------
              TABLE 9-9 (cent.)
Re fe re no e
Appelgren
et al. ,
1977

M3
1
cr>
ON



Test (feting and
System Sacrifice
Dominant lethal teles rated to
assay: mice 3 virgin
females per
week. Females
sacrificed
on 17th day
after first
exposure to a
male.

Chemical
Information Results
Single injection of Negative
either 0, 0.025, response
0.05 , or 0.1 g/kg
of ethylene oxide
given i.v.

Source: Not given

Purity: Not given

Comments
1. Reported data of dominant lethal test
from work by Bateman.

2. Positive controls showed a significant
dose-related positive response.

3. Highest dose is 1/3 that used by Generoso
et al. , 1980; route of administration
different from those used by Generoso
et al. and Embree et al. (1977).
Solvent:  Saline
                                                'I.  Conducted whole body autoradiography  study.
                                                    Determined ethylene oxide distribution  to
                                                    various tissues in the body,  Including
                                                    gonads after either Inhalation or  injection.

-------
                                   TABLE 9-10

Summary of Mutageniclty Testing of Ethylene Oxide:  Heritable Tranalocation Test





1
en
— 0

Reference
Generoso
et al.,
1980



Test
System Strains
Heritable T stock males
translocation treated and
mated to
(SEC x
C57BDF,
females
Chemical
Information
Single daily intra-
peritoneal Injection
of 0, 30, or 60
mg/kg of ethylene
oxide weekdays for
5 weeks

Results
Dose-related
positive
response




Comments


1. Shape of response curve consistent with
dose-squared kinetics.

2. Demonstrates capability of
to cause heritable genetic
in vivo.


ethylene oxide
damage in mice

                     Dose
                     (mg/kg)


                        0
                       30
                       60
                       60
Translocation Heterozygotes
Frequency           %
 0/822
 6/1456
38/U06
 6/72
0
1.32
9.36
8.33

-------
                                                                           TABLE 9-1 1


                                         Summary  of  Mutageniclty Testing  of Ethylene Oxide:  Chromosome Aberration Teats
I
CTi
OO
Reference
Fomenko and
Strekalova, 1973
Strekalova, 1971
Poirier and
Papadopulo, 1982
Teat
System
Chromosomal
aberrations
in bone marrow
from rata
Chromosome
aberrations in
bone marrow from
random bred
white rata
Chromosomal
aberrations in
the human
amniotic cell
line FL.
Chemical
Information
Concentration tested:
0.001-0.003 and
0.030-0.060 mg/liter
for 2, 4, 8, and 30 days
by inhalation
Source: Not given
Purity: Not given
Solvent: Not given
Concentration tested:
9 rag/kg per os
Source: Matheson Gas
products
Purity: Commercial
Grade
Reaulta Comments
Time-dependent poaitive 1. Method of preparing cells for analysis
response at highest dose not given.
2. Criteria for scoring aberrations not given.
3. Definition of terms not given.
1. Insufficient information for adequate
evaluation of results.
Positive response reported 1. Animals killed 24 and 48 hours after
treatment.
2. Chromosome preparations made from bone marrow
squashes.
3. Criteria for classification of aberrations
not defined.
4. Insufficient information for adequate evaluation
of results.
Dose-related poaitive 1. 1 hour vapor exposure.
response
2. Selected data presented only for cella harvested
72 hours after exposure.
EtO J Chromatid aberrations/ 100 cells
Dose Abnormal J
(mM) Metaphaaes Breaks Exchanges Survival
0 10.8 3.0 5.4 100
5 21.7 15.0 5.0 58
7.5 59.7 37.6 45.5 25
10 77.8 79.2 115.1 9.2

-------
                                                                     TABLE 9-12

                                       Summary of Mutagenioity Testing of Ethylene Oxide:  Micronucleus Tests
Reference
Test
System
Chemical
Information
Results
Comments
Appelgren et al.
  1978
Micronucleus
test:  NMRI
mice and
Sprague-
Dawley rats
Concentration tested:
0 to 0.3 g/kg (mice)
or 0 to 0.2 g/kg (rats)
via intravenous
injection 30 and 6
hours before the
animals are killed.

Source:  Not given

Purity:  Not given

Solvent: Cold water
Dose-dependent response in    1.
mice.  Increased Incidence
in rats,  but severe bone
marrow depression prevented   2.
further characterization.
    The animals given the highest doses died after
    the first or second injection.

    1000 polychromatic erythrocytes screened for
    micronuclei per animal.
Conan et al.,
  1979
Micronucleus
test:  Swiss
mice
Concentration tested:
Two injections.  Doses
ranged from 0-200 mg/kg
for i.p. injection, or
0-5 rag adsorbed to
Implanted plastic
devices.

Source:  Not given

Purity:  Not given

Solvent: Water
Dose-dependent positive
response after i.p.
injection.
Jenssen and
Ramel, 1980
Micronucleus
test:  CBA
mice (males)
Concentration tested
0-175 mg/kg

Source:  Fluka AC,
         Switzerland

Purity:  Not given

Solvent: Not given
Positive response
1.   Two-fold increase noted in micronucleus
    formation (0.33 + 0.10 in controls compared
    to 0.93 + 0.31  at 150 mg/kg.

-------
                                                                     TABLE 9-13

                            Summary  of  Mutagenicity Testing  of Ethylene  Oxide:   Chromosome  Mutations in Human Populations
Reference
Thiess et al.,
1981



Test
System
Chromsome
aberrations:
peripheral blood
of occupationally
exposed workers



Chemical
Information
Exposure:
1 . Long-term
(>20 years)
2. <20 years
3. Long-term plus
Results
Mutagenic effect Indicated
Aberrations excluding
gaps:
1. a. 3-5
b. 2.7
2. 2.3
3. 2.2
Comments
1. Workers were exposed to other alkylene
oxides besides ethylene oxide. Cannot
assign damage to one agent.



                                          accident

                                      4.   Accident

                                      5.   Control
                                             1.   1.1
Pero et al.,
  1981
Chromosome
aberrations:
peripheral
blood lymphocytes
from ethylene
oxide-exposed
workers
Exposure levels:
0.5-1.0 ppra in air
Suggestive positive
response for aberrations
excluding gaps.  Noted
only in comparison
1.   Both exposed groups had significantly higher
    levels of total aberrations (breaks and gaps)
    compared to the control group

-------
California,  purity  not given).   The  LC    is reported  to be  1462  ppm for  4




hours.   Erabree  et al.   (1977) reported signs  of toxicity after treatment  but




no deaths.   Immediately following treatment, each  male  was mated  to  two virgin




females per week for 10 weeks.  The  females were sacrificed  17  days  after  they




were  caged  with  a treated male.   Statistically  significant  (P<0.05)  increases




in the number of post imp Ian tat ion deaths were observed on  weeks 1, 2,  3, and  5




after treatment,  but not  on  other weeks,  indicating  that ethylene  oxide exerts




its effects on  post-meiotic cells.   It should be noted that  the  statistical




significance of  the  increases observed for weeks 1 and  5  may have been due to




low negative control values for the corresponding weeks.




     Generoso et al.   (1980)  also  observed an  increased incidence  in  postim-




plantation deaths in mice during the first 2 weeks after administration of 150




mg/kg ethylene  oxide  (Eastman Kodak,  purity  not reported) by  a single intra-




peritoneal  injection.  One dose of  200  mg/kg  ethylene oxide was shown  to  kill




10 of 12  mice.   The  testing  for  dominant  lethal  effects in this study was  done




in two  ways.   In  the  first  experiment,  T stock  males  treated with  ethylene




oxide were  mated to two  virgin (SEC  x C57BL)F..  females.   When a  female  was




impregnated, as  evidenced  by  the   observation  of  a vaginal  plug,  she   was




replaced  with  another  female.    These females  were also  replaced  after  the




observation  of  a vaginal  plug  and  so  forth   for  three  weeks  post-treatment.




The  females were sacrificed  12-15  days after  the observation  of the  vaginal




plug and were dissected to determine the  frequency of dominant  lethal effects.




A significant increase in post implantation deaths was observed  in  females  that




were  bred  with  treated males between  days 2.5  and   11.5  post-treatment  (from




12-31?  dead  implants  in  treated  group   compared  to  3-5?  dead  implants  in




negative control  group).   This  indicates  that late  spermatids  and  spermatozoa




                                     9-71

-------
are  sensitive to the  test  compound.   In  the  second  experiment (101 x  C3H)F


males were  injected with  ethylene  oxide and divided equally  into  four groups.


Four  days  post-treatment  they were mated  either to T stock,  (SEC x  C57BL)F


(101  x  C3H)F1}  or  (C3H  x  C57BL)F1  females.    The  females  were  checked  for


vaginal  plugs each  morning  until  the—S-th  day post-treatment and were  killed


for uterine analysis  12-15  days after  the  observation  of a vaginal plug.   The


purpose  of this experiment  was  to determine  whether  the different  stocks of


mice differed with respect  to  the  ability  of oocytes to  repair genetic  damage


induced  in the  treated  male  genome.    The  results  of  this experiment  were


consistent with  those of the  first experiment  in  showing  an increased  inci-


dence  of  post implantation  deaths.   However,  no  significant difference  was


observed when  (101 x  C3H)F.. -treated males were  mated  to females of  different


stocks.


     Appelgren  et  al.  (1977)  studied  the whole-body  distribution  of  radio-


labeled ethylene oxide  in mice and reported the  results of a dominant  lethal

                                    in
test.  Nfeile mice were  treated  with  [  C]  ethylene  oxide  (specific  activity not


given) by  inhalation or intravenous injection.   The animals  were  later  sacri-


ficed and  auto radiograms  of midsagittal sections  were  prepared.    The  auto-


radiograms  from  mice  that inhaled ethylene  oxide  differed qualitatively  from


those  that received   the material intravenously  in  only  one  respect:  the


mucosal membranes of  the  respiratory  tract  of animals  that  inhaled  the  com-


pound accumulated ethylene  oxide.   In  experiments  conducted using the  intra-


venous  route  of  administration,   ethylene  oxide  was present in  the  gonads


(epididymis and  testicle) 20 minutes  after administration.   Radioactivity was


still present  in the  epididymis 24  hours after  injection.
                                     9-72

-------
     These observations  that ethylene oxide reaches  the  gonads are consistent




with the  positive  dominant lethal responses reported by Erabree  et  al.  (1977)



and Generoso et al.   (1980).   However,  the results of the dominant lethal test




cited by  Appelgren et  al.  (1977)  were negative, in that  there  was no increase




in the  incidence  of  dominant lethal  mutations.   The  highest dose used in this




study was 100 mg/kg,  as compared to the  150  mg/kg  used  by Generoso  et  al.




(1980).   Since the  chemical  was  administered  by  intravenous injection  in the




study by  Appelgren  et  al.   (1977)  and  intraperitoneally  by Generoso  et  al.




(1980), it  is  not clear whether  the  apparently negative response in the study




of Appelgren (1977)  is  attributable  to  the difference in the dose or to other




factors.



     The  positive  dominant lethal tests  reported by Embree  et  al.  (1977)  and




Generoso  et al.   (1980)  indicate that  ethylene  oxide  reaches  the  germinal




tissue  in intact  mammals  and  causes  genetic damage.  Although  these tests do




not unambiguously  demonstrate  heritable effects caused by  ethylene  oxide,  the




positive  heritable translocation  test reported  by Generoso  et al. (1980) does.




Mouse-specific locus tests,  which measures  heritable gene mutations, are  now




underway  at Oak Ridge  National  Laboratory and  Research Triangle Institute,  and




the  results should  provide  additional  insight into  the ability  of ethylene




oxide to  cause heritable mutations in intact mammals.








     9.4.2.2.   HERITABLE  TRANSLOCATION  TEST  —  In conjunction with  their




study of  dominant  lethal effects,  Generoso et  al.  (1980) tested ethylene oxide




for  its  ability   to  cause  heritable  translocations  in  mice   (Table  9-10).




T stock male mice were given 0, 30, or  60 mg ethylene oxide per kg once daily,




weekdays,  for  5  weeks.    Immediately after the  last injection  each  male  was




                                      9-73

-------
caged with three  (SEC  x  C57BL)F1  females.   After 1 week the treated males were




removed,  and  the  females were  separated  from  each  other.    In  the  control




group,  each  male was  left  with  one of  the three females  for  -5 months after




the  first litters  were   born  in  order  to  produce  additional  progeny.   The




incidence of  heritable translocations was  as  follows:  negative control,  0%;




30 mg/kg, 1.32$; and 60 mg/kg, 9.36$.  These positive  results demonstrate that




ethylene oxide causes heritable chromosomal mutations in whole mammals.








     9.4.2.3.   CHROMOSOME  ABERRATION  TESTS —  The  ability  of ethylene oxide




to  cause well-defined  chromosomal  aberrations   (breaks,  rings,  inversions,




translocations,  etc.)  has  been   studied  by  several  investigators.    Some  of




these  studies have  been  discussed  above,  including  the positive  heritable




translocation  tests  (Watson,   1966   and  Generoso  et al.,  1980),  and  work




conducted with plants  (e.g., Jana  and  Roy,  1975).   Two additional experimental




studies  were  evaluated   (Table  9-11).    One  was  by  Fomenko  and  Strekalova




(1973),  who  administered   from   0.001-0.003 mg/Jl   or  from  0.030-0.060 mg/£




ethylene  oxide  (purity not reported)  by inhalation  for 2,  4, 8,  or 30 days to




white rats (strain unspecified).   A time-related increase  in total aberrations




in bone  marrow  cells  was noted in the high dose group (7.1-11.6$) compared to




the  negative  controls (3.0$).   The  significance  of  these results  cannot  be




determined,  however,  because of  deficiencies in  reporting how  the chromosomes




were prepared and in defining criteria for  scoring aberrations.




     Similarly, Strekalova  (1971)  reported that administration of one 9 mg/kg




dose of ethylene oxide  per os^ in aqueous solution resulted in  an  increased




incidence of  total aberrations  in bone  marrow  cells  scored  21 , and  in some-




cases,  48 hours  later;   the vague  manner  in  which  the  study   is  reported,




                                      9-74

-------
however, precludes an independent evaluation of  the  results.   The most notable




problem  is  that the  terms  and  the  criteria  for  scoring  aberrations are  not



defined.   Furthermore,   bone  marrow  squashes  were  used  to prepare  metaphase




chromosomes for analysis.  This  technique is not  suitable,  because  it does  not




yield high quality chromosome  spreads compared  to chromosome  preparations made




by the air-drying technique.




     Poirier  and Papadopoulo   (1982)  exposed  F1  cells   (derived  from  human




amnios)  to ethylene  oxide  (commercially  available from Matheson  Gas  Products)




at 5,  7.5, and 10  mM for 1 hour.  The corresponding number of cells  surviving




was  58,  25,   and 9.2?.   Three  separate experiments  were performed.    After




harvesting (at 48,   72  and 196  hours)  and  slide preparation,  150  metaphases




were scored for  each dose and  fixation time (50  from each  experiment).   Dose-




related  increases  in  chromatid aberration  were  found.    For  example,   at  48




hours after treatment the  frequency  of exchanges  (triradials,  'dicentric'  and




'centric' rings) per  100 cells was 5.9,  10.6, 56.7,  and  127.3  for  the corres-




ponding  treatments  of 0,  5,   7.5,  and  10  mM ethylene  oxide/1 hour  exposure




(Table 9-11).




     Ethylene   oxide  at  50  and  100 ppm 7 hours/day,  5 days/week  for  104  weeks




also significantly  increased   the  frequency of  chromatid/chromosomal  aberra-




tions  in peripheral  lymphocytes of  male  Cynomologus monkeys  (Lynch et  al.,




1982a;   Dr. D.  Lynch, personal  communication,  1983).   The response  was  dose-



related; roughly 4-fold  increases in  cells with one or  more  chromatid  and/or




chromosome aberrations  were  noted in the  high  dose animals  compared to  the



negative controls.
                                     9-75

-------
      9.4.2.4.   MICRONUCLEUS FORMATION  —  Three studies addressed  the ability




 of ethylene  oxide  to  induce micronuclei  (Table  9-12).  Appelgren et al.  (1978)



 treated  NMRI mice  by intravenous injection with  two  doses of ethylene  oxide




 ranging  from 50-300 rag/kg,  30 and 6  hours  before  sacrifice, and Sprague-Dawley




 rats according  to  the same regimen with doses  up  to 200 mg ethylene  oxide/kg.




 Mice given 300 mg/kg  died  after  the  first  injection.   Rats  given 200  rag/kg




 died after  the  second injection.   In  mice,  ethylene  oxide  caused  a highly




 significant  dose-related  increase in micronuclei.   At  the highest dose  there




 were 2.48? polychromatic  erythrocytes with  micronuclei compared to 0.52?  poly-




 chromatic  erythrocytes  with  micronuclei   in   the  negative   control  animals




 (P<0.001).   Rats also  exhibited  a statistically significant increase in micro-




 nuclei,  but  it was  not  shown to  be dose-related.   Toxicity to  the bone marrow




 confounded  the results.   The mid-dose level caused 1.08?  polychromatic eryth-




 rocytes  with micronuclei  compared to  0.49?  polychromatic erythrocytes  with




 micronuclei  in the negative controls  (P<0.05).




      Using  male  Swiss  mice,  Conan  et  al.  (1979) conducted  three different




 types  of experiments  to  assess  the  ability of ethylene oxide or  its metabo-




 lites, ethylene  glycol and 2-chloroethanol, to cause  micronuclei.   Ethylene



 glycol and 2-chloroethanol were  given  to  the  experimental animals  via  oral




 administration or  intraperitoneal  injection.  Ethylene  oxide  was  administered




 by   intraperitoneal     injection,  intravenous  injection  or  intraperitoneal




 implantation of gas sterilized medical  devices.   Implantation  of  the ethylene




 oxide gas  sterilized  medical devices  did  not induce elevated  numbers of poly-




 chromatic  erythrocytes with micronuclei.   Similarly,  when ethylene  oxide was




 injected  intravenously  (two  injections of  100  mg/kg 24 hours apart)  and the




animals  killed 6 hours after the  second injection,  no  statistically signifi-




                                     9-76

-------
cant   increase   in   micronucleus  formation  was  observed  after   treatment.




However, when ethylene oxide was given  in traper itonea lly a suggestive positive



response  was observed.    As  the dose of ethylene  oxide increased from 0-4000




rag/kg  (in traper itoneal),   the  percentage  of polychromatic  erythrocytes  with




micronuclei  increased from 0.23-0.47.




     Jenssen and  Ramel (1980)  used  CBA male mice  in  their  assessment  of the




ability  of  ethylene  oxide   to  cause  micronuclei.     Ethylene  oxide  was




administered  intraperitoneally  at dosages  up  to  175  mg/kg,  and micronuclei




were  scored  in  polychromatic  erythrocytes  24 hours  later.  The  response was




not clearly  dose-related,  but  a two-fold  increase  in micronuclei was observed




in the animals at the two highest doses (150 and 175 mg/kg) compared to values




for  the negative control animals  (0.93 +  0.31? and 0.66  +  0.1955 compared to




0.38+0.10$, respectively).




     The positive responses obtained  in the micronucleus tests of Appelgren et




al.  (1978)   and  of  Jenssen  and  Ramel  (1980)  indicate  that  ethylene  oxide




reaches bone marrow  and  exerts  a chromosome damaging (breakage and/or nondis-




junction) effect on heraatopoietic cells  of mammals.








9.4.3.    Chromosome  Mutations  in  Human Populations.   Three  studies  have been




conducted in which workers exposed to ethylene  oxide have been  monitored  fbr



the presence of chromosome damage in  peripheral blood lymphocytes.




     Ehrenberg  and Hallstrom  (1967)  monitored  eight workers  for  the presence




of chromosome aberrations in  peripheral lymphocytes eighteen  months  after an




acute  exposure  to high,   but  unspecified,   concentrations  of  ethylene  oxide.




Ten unexposed  persons  were  selected  as controls.   The  two groups were  not




characterized in the  report and  it  is not  known  how well the  control group was




                                     9-77

-------
matched  to  the  exposed group.   No  analyzable cells  were obtained  from  one




person  in the  exposed group.   All samples  were  coded and  an average  of 20




metaphase  plates  was analyzed  per remaining  person  (range  = 6-26).    Gross




chromosome  aberrations (i.e., chromosome  and chromatid breaks and exchanges,




supernumerary  chromosomes,  and one  case of endoreduplication) were elevated in




the  exposed  subjects  (17.5?)  compared  to  the  unexposed  control  subjects




(4.3?).    Chromosomal  effects  such  as  this  are  potentially heritable  and




represent  clear evidence of  genetic  damage.   The addition of chromosome  gap




data  to  these  values increased the  respective incidences to  30.2 and  16.5/5.




Because  of  the  small  size  of  the study  population  and  the  low number of




metaphase  spreads  analyzed,  the discriminating  power  of the study  was  not




great, and,  thus,  the elevated levels  of chromosome  damage   observed  in  the




exposed population was judged not to be a significant positive  effect.




     Thiess  et  al.  (1981)  monitored 43  humans  exposed to  ethylene oxide  and,




to  a lesser extent,  other alkylene  oxides for  the presence  of chromosomal




aberrations  (Table  9-13).    The workers  ranged in age  from  27-63 years  (x =




47.1 years).  Individuals were  divided  into four groups based  on  the  type and




extent of ethylene oxide exposure they had received:






     1.    Long-term exposure  (>20 years), 11 men



     2.    Less than 20 years of exposure, 6 men




     3.    Long-term exposure plus accident, 21 men




     4.    Accident (i.e., short-term high exposure to ethylene oxide),  5 men




Subjects in  the  first three groups worked  in plants where ethylene  oxide was




manufactured or  processed.    Personnel  in  the  fire  department  or maintenance




workers  comprised the  fourth  group.   The negative control  group included  male






                                     9-78

-------
office and  staff workers, none  of whom had  been  exposed to  radiation  at the




time  of  testing.   The  age  of  individuals  in  the  control  group  ranged  from



24-58 years  (x  =  38.6).   The  workplace was monitored for ethylene  oxide  by




means of  spot samples  for up  to  2-hour  periods  and  for  propylene  oxide  by




personal dosimeters  for up  to  10 hours over 12-hour  shifts.    Ethylene oxide




exposures were normally <5 ppm  but were  found to rise  to 1900  ppm for several




minutes during a plant  breakdown.   Levels of propylene oxide  were usually far




below the  maximum allowable  concentration of  100  ppra, but higher concentra-




tions were measured for brief periods.   The  percentage of aberrant metaphases,




excluding gaps,  in cells  cultured  from 70-72  hours  at 37°C  in  two control




groups was  1.4 and 1.   Based on  Fisher  exact test analysis of the data,  with




Yates correction,  significantly  increased incidences  of chromosomal aberra-




tions were  observed in Group I individuals  (>20  years exposure)  compared  to




the  control  group upon  examination  in  October  1978  (3.5%,  P<0.005).    An




increased incidence of  aberrant metaphases was  also noted when  these indivi-




duals were  examined again in August  1979  (2.7$,  P<0.05).    No  statistically




significant  increase  was  observed for  the  other groups.   The  significantly




increased rate of chromosome aberrations  (excluding  gaps)  in  workers exposed




to ethylene  oxide for >20 years  suggests  a  mutagenic  effect.   The results  do




not  conclusively  indict   ethylene  oxide  as  the  causative  agent,  however,




because the workers were  exposed  to  other  substances (such  as  ethylene chloro-




hydrin,  ethyleneimine, propylene oxide, etc.) which  may have caused or contri-




buted to the  effect.  Furthermore,  it  should  be noted  that the authors may not




have used an appropriate statistical test  in  their evaluation  of  the data.  In




performing  the  Fisher  exact test  one  must  assume  that  one aberration  is




independent of another  aberration.   Within   individuals  this   may  not  be  the




                                     9-79

-------
case.   If  a person has one aberration  he  may be more likely  to  have a second




aberration  particularly  if the damage  was induced  in  a stem  cell.    If  this




were the  case in the  study  by Thiess  et  al.  (1981) one of  the  basic assump-




tions  of  the  Fisher-Yates test,  that  of independence  of the  observations,




would  not  be  met.   A more  appropriate statistical  test,  therefore,  and  one




which the authors claimed to have used  (but did  not  report)  in their analysis,




is the  Mann-Whitney test.  Use  of the Mann-Whitney test to compare Group 1  and




the  control group shows  an  increased  (and  perhaps  biologically  significant)




but  not statistical difference between the  two groups  in regard  to aberra-




tions.




     Pero  et  al.  (1981)  also  found increased incidences  of chromosome aberra-




tions in factory workers  exposed  to  ethylene  oxide (Table  9-13).   The workers




were divided  into three  groups.   One  was an  unexposed control group and  two




were exposure  groups  (i.e.,  sterilizers and  packers) exposed  to  50% ethylene




oxide and  50%  methyl  formate  gas (0.5-1.0 ppm  ethylene  oxide) via inhalation.




Chromosome  breaks  and gaps  were  scored in  the peripheral  blood  lymphocytes




from these individuals.    Cells were  cultured for 72 hours  and 200 metaphases




were scored per  individual.    A statistically significant  increase  in chromo-




some gaps  plus  chromosome breaks was  observed  in  cells from  the sterilizer



ethylene  oxide-exposed group  (5 workers)  compared  to  the  control  group  (9




workers),  11-1^  in exposed  groups  compared  to  8.5%  in  controls,   (P<0.05).




With respect  to  breaks  alone,  however, a nonsignificant  (or  at  best  only  a




marginally  significant)   increase  was  noted  in  the  comparison  between steri-




lizers  and control groups (8.2  + 1.0%  compared  to  5.8 +  1.0?,  respectively,




P<0.15).   The comparison between the packer  (12 individuals),  6.2 +  0.9%,  and




control groups was not significant.




                                      9-80

-------
     The increased  incidences  of chromosome aberrations  in  peripheral lympo-




cytes noted in three studies of workers exposed  to  ethylene  oxide  are consis-



tent with  one  another  and with the  experimental animal  data showing ethylene




oxide to be clastogenic.   They  indicate  that similar  effects are  caused  in




humans as well.








9.4.4.    Other Studies Indicative of Genetic Damage.   Additional studies have




been conducted bearing on  the  genotoxicity  of ethylene  oxide  (Tables 9-14  to




9-16).  These  studies  do not measure mutagenic  events per se  in  that they  do




not demonstrate  the  induction  of heritable genetic alterations, but positive




results in  these  test  systems  do  show  that DNA has been damaged.   Such test




systems provide supporting evidence useful for qualitatively  assessing genetic




risk.








     9.4.4.1.   SCE FORMATION IN HUMAN  POPULATIONS  —  Three  studies have been




reviewed  here  concerning  the  induction  of  SCEs   in   humans  (Table  9-14).




Lambert and Lindblad  (1980)  studied peripheral  lymphocytes  from  five female




workers  in  a  German  sterilization  plant  to  determine  if ethylene  oxide




exposure causes  genotoxic effects  in  vivo as measured  by  SCE  formation.   A




description  of the  exposure  these workers  received  was not reported.   The




frequency  of  SCE  formation   in  exposed  individuals   (19.1/E)  was  increased




compared to  that  of the  unexposed  control group  (14.6$).   Although the small




sample  size and  uncharacterized  exposure  these workers received  preclude  a




definitive  assessment  of the  ability  of  ethylene oxide  to  cause  SCEs  in




humans,  the results are  considered to  indicate genetic toxicity  in  somatic




cells of the exposed workers.




                                     9-81

-------
CO
ro
                                                                           TABLE  9-14

                                     Summary of Mutagenicity Testing of Ethylene  Oxide:   SCE  Formation in Human Populations
Reference
Johnson and
Johnson,
1982
Garry et al. ,
1979
Test
System
Sister chromatic!
exchange
Induction and
chromosome
aberrations:
Industrial
workers
Sister chromatid
exchange induction:
Chemical
Information
Inhalation exposures
estimated to be: Low
relative exposure
(1 ppm), moderate
relative exposure
(1-10 ppm), high
relative exposure
(5-200 ppm).
Months
0
High Potential 33
Exposure
Low Potential 14
Exposure
Inside Controls 12
Outside Controls
Maximum exposures
estimated to be 36 ppm
Results Comments
Dose-response association 1. Levels of SCE remained elevated after
suggested termination of exposure.
2. Environmental exposure to ethylene oxide causes
increased SCE formation.
3. Report based on preliminary data from
relatively small sample population.
SCE Chromosome Aberrations
after Exposure Months after Exposure
606
35 1.5 2
15 1.1 0.9
12 0.6 0.78
8 — 0.5
Statistically significant 1. Air dried fluorescence plus Giemsa
increases in level of SCE chromosome preps.
                      peripheral blood
                      lymphocytes
                      collected from
                      hospital workers
(from average measure-
ments over one 8 hour
period).   Workers
divided into groups
based on  known exposures
to ethylene oxide and
symptoms  indicative of
exposure.
observed in exposed
Individuals compared to
controls (unexposed
laboratory personnel).
2.   20 metphases scored/individual.

-------
                                                                       TABLE  9-1M  (cont.)
Reference
Test
System
Chemical
Information
Results
Comments
     Yager,  1982
     and Yager
     et al.,
     1983
 l
CD
OJ
Sister chromatid
exchange induction:
peripheral blood
lymphocytes
collected from
hospital workers
Exposures determined by
individually monitoring
workers.  High exposure
group received a cumula-
tive dose >100 mg
while cumulative dose for
low exposure group was
<100 mg.
1.   Control group carefully matched to the exposed
    group for age, sex and personal habits.

2.   Exposure estimates based on breathing zone
    measurements and task frequency estimates.



Laurent
et al.,
1982



Sister chromatid No
exchange induction:
peripheral blood
lymphocytes collected
from hospital
workers.
Mean
Group Exposure (mg)
Control 7
Low exposure 13 7
High exposure 501 10
exposure estimates Exposed group had 1.
statistically significant
increase in SCEs compared
to control group. Range
of SCEs for the exposed group
was 9.61 - 17.57 compared to
a range of 7.0U - 8.52 for the
control group.
SCEs/
cell
.56 + 1.01
.76 + 1.05
.69 + 1.92
Control group
age, sex, and
group.



may not have been matched for
personal habits to the exposed

-------
                                                                            TABLE  9-15

                                    Summary of Mutagenioity Testing of Ethylene Oxide:  SCE Formation in Experimental Studies
Reference
Test
System
Chemical
Information Results
Comments
      Star,  1980
Sister chromatid
exchanges: Cultured
human fibroblasts
I
CO
-pr
                                            Concentrations tested:
                                            0 to 3600 ppm and
                                            residues from plastic
                                            children's endotracheal
                                            tubes treated with 1100
                                            mg/cm3 of pure ethylene
                                            oxide for 90 minutes
                                            followed by aeration
                                            from 21 to 96 hours
                                            after sterilization.
                      Source:  STERI-Gas cartriges
                               3M Germany GmbH,
                               Neuss

                      Purity:  Hot given

                      Solvent:  Dulbecco's Modified
                               Eagle's Medium
Toxic as well as mutagenic.
Significant increases in SCE
induction at 36 ppm.  Cyto-
toxiclty at 180 ppm and
higher
Cultures from skin biopsies used between
fifth and tenth subculture.

Insufficient data presented to evaluate
conclusions.
Yager and
Benz, 1982



Kligerman
et al.,
1983

Sister chromatid
exchange induction:
New Zealand White
rabbits


Sister chromatid
exchange induction:
CDF rats

Concentrations tested:
0, 10, 50, and 250 ppm
by inhalation
Source: Matheson
Dayton, OH
Concentrations tested:
0, 50, 150 and 150 ppm
for 1 or 3 days by
Inhalation
Positive response at 50 and
250 ppm exposures



Dose- and time-dependent
positive response


1 . Increased SCE levels decreased after exposure
ended but still remained above baseline
levels 15 weeks after exposure.


1. Significant increases at 50 ppm show effects
induced at levels to which workers have been
exposed. Until recently TWA was 50 ppm.

                                            Source:
                                                    Matheson Gas
                                                    Product
                                                                                                     2.   Data for 3 days exposure groups shown.
                                            Purity:  99.7*
     •Significantly  different  from  controls  by one-tailed Dunnett's
Concentration
0
50 + 7
110 + 17
111 + 33
SCEs/
Metaphase
7.5 + 0.5
9.1 + 1.3»
10.3 +1.3*
13.6 + 1.3»

-------
                                                                          TABLE 9-16

                                         Summary of Mutagenioity Testing of Ethylene Oxide:   Unscheduled DNA Synthesis

Reference
Gumming
et al.
(in press)








Test
System
Unscheduled DNA
synthesis:
testlcular DNA of
(101 x C3H)F!
mice






Chemical
Information
Concentration tested:

a. 600 and 800 ppm for
2, 14, 6, or 8 hours.
[3H] dThd administered
intratesticularly
immediately after
ethylene oxide
administration



Results Comments


a. Dose-dependent increase
in UDS over lower range
of doses tested (e.g.,
70 dpm/106 cells,
MB dpm/106 cells, and
8 dpm/106 cells for
800 ppm, 600 ppm, and
negative controls at
1 hours)
 I
OO
U1
                      b.  Same  as  above  except
                         [3H)  dThd  administered
                         at  different times
                         after termination of
                         exposure

                      c.  300 and  500 ppm 8 hr/day
                         for 5 days.  Aliquots
                         of  animals sacrificed
                         daily

                      d.  500 ppm  for 2,  4, 6, and
                         8 h.   6  animals given 80
                         mg/kg 3-methylchloran-
                         threne,  6  animals drank
                         water with 1 mg/m{,  sodium
                         phenobarbital  for  1 week
                         prior to exposure,  6
                         animals  uninduced controls
                                                                         b. UDS peaks 2 hours after end
                                                                            of exposure period at day
                                                                            5 for 300 ppm;  at day 1 for
                                                                            500 ppm
                                 Response peaked at day 5 for
                                 300 ppm; at day 1 for 500 ppm
                                                                            UDS response dramatically
                                                                            reduced In animals receiving
                                                                            mixed-function oxidase
                                                                            inducers
                                           Source:
                                                    Matheson Co.,  East
                                                    Rutherford, NJ
                                           Purity:  99.7*
     Pero et al.
     1981
Unscheduled DNA
synthesis:  Human
lymphocyte  cultures
Exposure levels: 0.5 to
1.0 ppm in air
Positive response
1.  UDS induced by exposure to N-acetoxy
   acetyl aminofluorene (NA-AAF).

2.  Decreases in NA-AAF-induced UDS measured
   biochemically and by autoradiography in
   lymphocytes from ethylene oxide-exposed
   workers.   UDS peaked at 2 mM exposure
   NAAAF.

-------
     In  a  preliminary,  unpublished  report Johnson and Johnson (1982) described




how  they monitored  workers at  three  sterilant facilities for the presence of



SCEs  and  chromosome  aberrations in  peripheral blood  lymphocytes.    Based on




environmental sampling  the workers  were  assigned  to one of the following cate-




gories  depending upon  the plant  site at  which  they  worked:    high relative




exposure (5-200  ppm),  moderate relative exposure (1-10  ppm),  and low relative




exposure (1  ppm).   The  numerical exposure values  represent the estimated range




of  an 8-hour  time-weighted-average inhalation exposure.    Employees  at  each




plant were further  categorized as  to  high  or  low potential  for ethylene oxide




exposure based on their job  description  and other  factors.   During  the course




of  the  study it  was noted that  the SCE  levels in  the control group of presum-




ably  unexposed workers  at Plant  III  were higher than  those of  other control




groups available  for  comparison at the time  (12/metaphase compared  to 7/meta-




phase).   The study was therefore  expanded  to include  an  additional  control




group, which  was  taken  from the  local community and  matched by  sex  and age to




potentially exposed Plant III employees.




     Preliminary  analysis of  the  data  indicates  a consistent  dose-response




trend at Plant III for  SCE induction both  at  the  original  monitoring and later




after 6 months with no  further ethylene  oxide  exposure  (mean values of 12, 14,




and  33  SCEs/metaphase  for internal controls,  low potential  exposure  and  high




potential  exposure  groups,  respectively,  compared  to 8  SCEs/raetaphase  for the




external control  groups).  A much less pronounced trend  was  noted at Plant II,




and  the  SCE  data for Plant  I  showed no significant  difference  between poten-




tially exposed and control groups.  Analysis of the chromosome aberration data




suggests a dose-related  increase  in  damage,   but  the  magnitude  of  the  dif-




ferences between groups is  not great.   Thus,  it  appears that a  dose-response




                                     9-86

-------
association  exists  between exposure  to  ethylene oxide  and  SCEs in humans and



that  the increased  levels  of SCEs  appears to  be stable,  perhaps suggesting



long-lived  adverse  effects caused  by human exposure  to ethylene oxide.  How-



ever,  it is  important  to bear  in mind  that  these  conclusions are  based  on



preliminary  data from a relatively  small  study population.



     In  a study of  12 ethylene  oxide  exposed workers from the instruments- and



materials-sterilization  areas  of  a  hospital,  Garry  et  al.  (1979)  reported



increased  SCE  levels  in  the  peripheral  blood  lymphocytes.     The   maximum



exposure concentration  sampled  15 feet from the sterilizer was estimated to be



36  ppm based on an infrared spectroscopy  measurement over  one  8-hour period



during the  course of the  study.   Individuals reporting upper respiratory irri-



tation had  statistically  significant  increases  in the  incidence  of SCEs com-



pared  to  the  control  population of 12  unexposed  persons  working  in  the



adjacent operating  room (10.3 +  1.8 vs. 6.4 + 0.47, P<0.01).



     Yager  (1982)  and Yager  et  al.  (1983) also  monitored 14 hospital  workers



exposed  to  ethylene oxide.   Thirteen  persons not  exposed   to  ethylene oxide



served as  matched  controls.    Cumulative exposure  doses  during the  6 months



prior to blood sampling were  estimated by monitoring air concentrations during


                                     TM
defined  tasks (using a Wilkes-Miran   1A  Gas  Analyzer)  and multiplying this



value by the  number of  sterilizer loads processed.   Based on these estimates,



the workers  were assigned  to  the low dose group (13 +  18 mg  ethylene oxide)  or



the high dose group (501  + 245  mg ethylene oxide).   An increased incidence  of



SCEs/cell was observed  in  the  high dose group  (10.7 + 1.92)  compared  to  the



low dose (7.8 + 1.05) and  unexposed control (7.56 + 1.01) groups.



     Laurent  et  al. (1982)  also  collected  peripheral  blood  from  hospital



workers exposed to   ethylene oxide.  Ten persons  in good health and not  exposed



                                      9-87

-------
 to  any known  toxicants  were selected as  the  negative control  group.   It  was




 not  reported whether the controls  were  matched for  sex,  smoking habits, etc.




 They  do  not appear to have been matched for age because the age  of  the  control




 group  ranged between 20  and 35 years while that of  the ethylene oxide  exposed




 workers  ranged between 23 and 51  years.   No  estimate was made of the exposure




 received  by the  sterilizer operators,  but they  had  a significantly elevated




 level  of SCE compared  to  the controls (13.02 + 2.294  vs. 7.86 + 0.479).




     The  increased  incidence  of  SCEs  observed  in  five  groups   of   workers




 exposed  to  ethylene  oxide does  not demonstrate  that mutations  occurred  but




 does  indicate  that  ethylene  oxide  can   cause  genotoxic  effects   in   somatic




 tissue of humans  in vivo.









     9.4.4.2.   SCE  FORMATION  IN  EXPERIMENTAL  STUDIES — Human  cells  in cul-




 ture  also  exhibited  increased  SCE levels after  exposure  to  ethylene oxide




 (Table 9-15).   Star  (1980)  exposed skin  fibroblast  cells from normal  healthy




human  tissue biopsies to 0-3600  ppm ethylene  oxide  or to  plastic  children's




endotracheal tubes  sterilized  with  1400 mg/cmj  ethylene  oxide  at  55°C  for 90




minutes followed  by  aeration in room air  for  varying  times from 24-96  hours.




The cell  lines were  kept frozen in  liquid nitrogen  and used between their  5th




and 10th subculture.   The placement  of the plastic  tubes  in  the culture medium




resulted in  ethylene oxide  concentrations  ranging  from 12-800  ppm as estimated




by GC  of  head  space  material.    Excessive  cell killing precluded  scoring SCEs




above 600 ppm  ethylene oxide for the experiment.   No statistically significant




increase in SCEs was noted  in the  experiment using  the endotracheal  tubes, but




a consistent,  apparently  dose-related,  rise  in SCSs  was noted  in this part of




the study at doses >217 ppm.   In the other set of experiments  a statistically




                                      9-88

-------
significant increase  in  SCE induction was reported  at  36 ppm; however, insuf-




ficient data were presented to permit an adequate evaluation of the results.



     A  membrane  dosimetry  system  was  developed  by  Garry et  al.  (1982)  to




enable the measurement and  determination of  dose-response relationships for in




vitro  exposure to  toxic gases.    Elevated  SCEs  were  observed  in peripheral




lymphocytes cultured  from healthy humans at  exposures as  low  10  (ig/mJ, (in the




media) during  a 20-minute  exposure period.  A  dose-related increase was noted




up to ethylene oxide  concentrations of 35  [ig/mJl  (the highest dose tested).  At




this  dose there  were about  20  SCEs/cell  compared  to  control  levels  of =5




SCEs/cell.



     Yager  (1982)  and  Yager  and  Benz   (1982)  administered  from  10-250  ppm




ethylene  oxide  gas  to 4-month-old  male  New  Zealand White  rabbits  via inhala-




tion.   Eight  animals were  placed  in  each exposure  chamber  for  6 hours/day, 5




days/week, for  12 weeks.  Blood  samples  were taken  from  the marginal ear vein




at  1,  7,  and   12 weeks  of  exposure  and  2, 7, and 15  weeks after exposure.




Three  animals  per  chamber  were  used for serial  blood  sampling  for  SCE and




hematological  assays  [i.e.,  red  cell   count  (total and  differential),  white




cell count, hematocrit, and hemoglobin concentration].   One animal was held in




reserve and four  animals were  sacrificed immediately at the end  of the 12-week



exposure  period for  analysis  of  reduced glutathione  (GSH)  in  liver and blood.




Positive  and  negative  controls  were  performed using  intraperitioneal injec-




tions  of  mitomycin  C and Hanks  balanced salt solution,  respectively,  at each




time  point.    Exposure  to   10  ppm did not cause a  detectable increase in the




incidence  of SCEs;  however, exposure to 50  and  250 ppm  did  cause  an  increase




in  SCEs  (9.^7  +  0.26  and  13.17  +  0.32, respectively)   that decreased after




exposure  ended, but  still remained  above baseline  levels (7.8 +  0.23)  15 weeks




                                      9-89

-------
after  exposure  (8.45  + 0.30).   Heraatological  and GSH  measurements from  the

animals  did not  differ  from controls.


     After  exposures  to  ethylene  oxide  of  0,  50,  150,  or 450  ppra  for  6


hours/day  for 1 or  3  days,  blood  was  removed  from  male CDF  rats  by cardiac


puncture,  cultured  in  the presence of  5-bromodeoxyuridine and scored for SCEs


and  chromosome  breakage  (Kligerman  et  al.,  1983).    No  significant  dose-

dependent  increase  in  chromosome  breakage  was  observed,  but  there   was  a


concentration  dependent increase  in SCEs.   Animals in the  highest  dose group


exposed  for  3 days had  13.6  + 1.3  SCEs/cell compared to  the control value of


7.8  +  0.5  SCEs/cell.   SCE induction was  also  significantly  elevated  after  3


days of  exposure to 50 ppm ethylene  oxide  (9.1  + 1.3)  showing effects in rats


at  levels  to  which workers   have  been  exposed.    There  was no  significant

reduction in mitotic activity  or slowing of cell kinetics.




     9.4.4.3.   UNSCHEDULED DNA  SYNTHESIS — Gumming et  al.  (in press)   tested


ethylene oxide for  its ability  to cause  unscheduled  DNA  synthesis in germ

cells  of  male  mice  after   inhalation  exposures.    Four  experiments  were


performed  in  which  hybrid mice  (101  x  C3H)F..  were treated with  99.7% pure

ethylene oxide  (Matheson Co.).  In  the  first experiment,  the effect of differ-

ential time  exposures  on  unscheduled  DNA  synthesis  induction  was  assessed.


Animals  were  treated with  600 and  800  ppm  ethylene oxide for 2-8 hours,  after


which exposed  animals  were anesthetized  with  metofane  and  injected intrates-


ticularly  with  [ H]thymidylic acid  (dThd).    A  dose-dependent increase  in


unscheduled DNA  synthesis  was  found over the lower  end  of the dose  range   for


the  first  4  hours of  exposure in  that a  higher response was  seen  at  800   ppm
                                    5
than at  600  ppm  (e.g.,   70  dpm/10   cells for 4-hour  exposure at 800   ppm


                                      9-90

-------
compared to 48  dpm/10  for 4-hour  exposure at  600  ppm;  controls  incorporated 8


dpm/10  cells).  Due to  the  toxioity of  ethylene  oxide  at  800 ppm it was only


possible  to measure  up   to  6 hours  exposure  for  this concentration.    In a


second  experiment,  ethylene  oxide administration  was  the  same  as above,   but

 •3
 [ HjdThd was administered to the  animals at  different  times after the end of


their ethylene  oxide exposure to  characterize the  unscheduled  DMA  synthesis


response at different times  after treatment.   Unscheduled  DNA  synthesis  was


found to increase with time  to a  peak 2 hours after the  end of the exposure


period,  and to  decrease  subsequently.  Two additional sets of experiments were


performed.   The first was a  workweek exposure  regimen of 300  or  500  ppm for 5


hours/day for 5 days, and the second involved  pretreatment of the animals with


mixed-function   oxidase inducers (either  a single  intraperitoneal injection of


80 mg/kg 3-methylcholanthrene or administration of drinking water containing 1


mg/mfl, phenobarbital for 1 week prior to  ethylene oxide  treatment).  Concerning


the workweek exposures,  little  effect was noted  after  the first two exposure


periods at 300  ppm.  An  effect was subsequently noted which  rose to a maximum


after the 5th  exposure period.   At 500  ppm the  maximum effect was seen after


the  first   exposure period.    Apparently,   increased   levels of  DNA  damage


occurred throughout the week, but after  the third exposure period the capacity


to respond  to  this damage appeared  to be  limited.


     Pero  et   al.   (1981,  1982)   treated  peripheral lymphocytes  taken  from


ethylene oxide   exposed workers with 10 mM N-ace toxy-2-ace ty lam inof luorene (NA-


AAF) for 1  hour and  subsequently measured the  incorporation of [ H] thymidylic


acid  into  DNA  to  detect unscheduled  DNA  synthesis (Table  9-16).    NA-AAF-


induced  unscheduled  DNA   synthesis was   found   to  be inversely related  to   the


duration  of  worker  exposure  to  ethylene   oxide  and   to   the   number  of


                                     9-91

-------
chromosome breaks observed.   This  suggests an inhibition of the cellular DNA-




repair  capacity  by  ethylene  oxide.   Biochemical and  auto radiography  studies




were  consistent with  this  response.   When NA-AAF-treated lymphocytes were ex-




posed to ethylene oxide, it was  found  that concentrations above 2 mM resulted




in inhibition of unscheduled DNA synthesis.




     As was  the case  for  the studies  of  SCE induction,  these  results  do not




show  that ethylene oxide is mutagenic but  do  indicate  it causes damage  to DNA




and  are  consistent  with   the   results  showing  that  ethylene   oxide  causes




nutations.








9.^.5.    Summary and Conclusions of the Mutagenicity  of Ethylene Oxide




Ethylene  oxide  has  been shown  to  induce  gene  mutations  in  bacteria,  fungi,




higher  plants,  Drosophila, and  cultured  mammalian  cells in  tests conducted




without  the  use of  exogenous hepatic  metabolic  activation   systems.    It  is




therefore a  direct-acting  mutagen.   Ethylene  oxide has  also  been  shown  to




induce  dominant  lethal effects  in mice and  rats; chromosomal  aberrations  in




higher  plants,  Drosophila,  mice,  and rats; and  micronuclei  in  mice and rats.




Based on these positive  findings in  different test  systems,  ethylene oxide  is




judged  to  be capable  of causing chromosomal aberrations.   It has also  been




shown to induce SCE  in rabbits,  rats  and humans.




     Tissue distribution studies have  shown that ethylene oxide   reaches the




gonads.   This  result is consistent  with evidence that  ethylene  oxide  causes




unscheduled  DNA synthesis  in  germ cells of male  mice and heritable mutations




in  insects  and  rodents  (i.e.,  sex-linked  recessive  lethals  and heritable




translocations in Drosophila,  dominant  lethals  in  rats  and mice and heritable
                                     9-92

-------
translocations  in  mice).   Ethylene oxide  can  therefore be  regarded  as rnuta-




genic both in somatic cells and in germ cells.



     Based on the available  data,  there  is overwhelming evidence that ethylehe




oxide is a direct-acting mutagen  that  has the potential  to  cause mutations in




the  cells of  exposed  human  tissue.    The observations  that  ethylene oxide




reaches and  reacts  with mammalian gonadal DNA,  and  causes  heritable mutations




in  intact mammals,   indicate  that  it  may be  capable  of  causing  heritable




mutations  in man  provided  that  the  pharmacokinetics  of  ethylene  oxide  in




humans  also  results  in its distribution  to  the DNA of  germ cells.   Thus,




ethylene oxide should be considered a potential human mutagen.
                                      9-93

-------
9.5.  CARCINOGENICITY









     The purpose of this section is to evaluate the likelihood that ethylene




oxide is a human carcinogen and, on the assumption that it is a human carcino-




gen, to provide a basis for estimating its public health impact and evaluating




its potency in relation to other carcinogens.  The evaluation of carcinogeni-




city depends heavily on animal bioassays and epidemiologic evidence.  However,




other factors, including mutagenicity, metabolism (particularly in relation to




interaction with DNA), and pharmacokinetic behavior, have an important bearing




on both the qualitative and the quantitative assessment of carcinogenicity.




The available information on these subjects is reviewed in other sections of




this document.  The carcinogenicity of ethylene oxide has also been evaluated




by the International Agency for Research on Cancer (1976).  This section pre-




sents 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 the carcinogenicity of ethylene




oxide.









9.5.1.  Animal Studies.  Only a few studies have been conducted to assess the




carcinogenicity of ethylene oxide.  Most of the reported studies have dealt




with subcutaneous administration and skin painting of the compound in mice and




intragastric administration in rats.  These studies are discussed briefly




herein.  Two lifetime inhalation studies in rats have been performed (Snellings




et al., 1981 and Lynch et al., 1982), and they will be described in detail.









     9.5.1.1.  MICE — Reyniers et al. (1964) conducted a study of female germ-




free mice that developed tumors (63/83) after being accidentally exposed to






                                   9-94

-------
ethylene oxide-treated  ground-corncob  bedding  for  150 days,  and were  moved  to




untreated bedding  for the  rest  of  their  lifespans.  These animals  developed




ovarian, lymphoid, and  pulmonary tumors.  Colony mates maintained  on  untreated




bedding did not develop  tumors.  All males exposed  to ethylene oxide-treated




bedding died, with necropsy  showing massive hemorrhage.  The causative agent




was not identified, since  chemical analysis of the  bedding was not done.  The




high number of tumors could  have been  due to other  chemicals (such as ethylene




glycol or 2-chloroethanol, both derived  from ethylene oxide) or to a  viral




agent, although the authors  believed that a viral agent was unlikely.  High




toxicity is indicated by these findings  in male mice.  Because germ-free mice




are T-lymphocyte deficient,  they may be  more susceptible than normal  animals to




tumor development, or the  tumor development may be  due to immune suppression.




At present, however, there is no evidence to support these hypotheses.




     Dunkelberg (1979) studied the oncogenic activity of ethylene oxide dis-




solved in tricaprylin and  administered subcutaneously to the interscapular area




of groups of 100 female NMRI mice in weekly dosages of 0.1, 0.3,  and  1.0 mg.




The incidence of spontaneous subcutaneous tumors in these mice was between 0




and 2%.  Preliminary results up to the 91st week of treatment showed that 6,




8, and 12 local tumors (sarcomas) occurred in mice  receiving total ethylene




oxide doses of 9.1, 27.3, and 91.0 mg,  respectively.  No local tumors occurred




in mice receiving no treatment or tricaprylin alone.  The number of tumors at




sites distant from the injection area was not significantly greater in the




group treated with ethylene oxide than in the two control groups.   The final




report of this study (Dunkelberg,  1981) covers  the period from the start  of




the study to 106 weeks,  at which time all of  the animals were sacrificed.   No




increase in tumors at  remote sites  was  observed.




     Lifetime skin painting studies with 10%  ethylene oxide in acetone (three





                                     9-95

-------
times weekly) were performed on 30 female mice by Van Duuren et al. (1965).




Application of 0.1 mL of ethylene oxide solution to the clipped dorsal skin




produced no tumors.  Median survival time for the mice was 493 days.  The




investigators indicated that rapid evaporation of the compound from the skin




was responsible for the negative results observed.









     9.5.1.2.  RATS — Walpole (1958) injected 12 rats subcutaneously with a




maximum total ethylene oxide dose of 1 g/kg (dissolved in arachis oil) over 94




days (dosing schedule not specified).  Rats were observed for their lifetimes




following treatment, and no tumors were observed.  Since the total amount of




ethylene oxide administered and the frequency of injection were not specified,




it is difficult to evaluate this negative result.




     Dunkelberg (1982) administered ethylene oxide intragastrically by gavage




at two dosages, 30 and 7.5 mg/kg body weight, to two groups of 50 female




Sprague-Dawley rats with empty stomachs twice weekly for a period of nearly 3




years, using salad oil as the solvent.  One group was treated with the solvent




alone, and the other group was left untreated.  A positive control group was




treated with3 -propiolactone.  The test substances were dissolved in 1 mL of




oil immediately before treatment.  The design of the experiment is summarized




in Table 9-17 and the results are summarized in Table 9-18.  Ethylene oxide




induced local tumors, mainly squamous cell carcinomas of the forestomach.  The




first tumor occurred in the 79th week.  The tumor rates were 62% in the 30




mg/kg group and 16% in the 7.5 mg/kg group.  In addition, carcinomas in situ,




papillomas, and reactive changes of the squamous epithelium of the forestomach




were observed in other animals.  An unspecified number of tumors occurred in




the glandular stomach.  Ethylene oxide did not induce tumors at sites away from




the point of administration.  Survival decreased in the positive control group.





                                    9-96

-------
            TABLE 9-17.  DESIGN SUMMARY FOR CARCINOGENICITY TESTING
                OF ETHYLENE OXIDE BY INTRAGASTRIC ADMINISTRATION
                             TO SPRAGUE-DAWLEY RATS


Group
Ethylene oxide I
Ethylene oxide II
Oil (vehicle)
Untreated
g -Propiolactone
Single
dose (mg/kg body wt)
(2x weekly)
30.0
7.5
1.0 mL
-
30.0

Average total
dose (mg/kg body wt)
5112
1186
-
-
2868

Number of
animals
50
50
50
50
50
SOURCE:   Adapted from Dunkelberg, 1982.
          TABLE 9-18.  TUMOR INDUCTION BY INTRAGASTRIC ADMINISTRATION
                OF ETHYLENE OXIDE IN FEMALE SPRAGUE-DAWLEY RATS
                               Number of rats with stomach lesions

Dose
7.5
30. Ob
Reactive
changes3
9
11
Carcinoma
in situ
4
4

Fibrosarcoma
0
2
Squamous cell
carcinoma
8
29
No stomach tumors were seen in either vehicle-controls or untreated controls.
aReactive changes of the squamous epithelium of the stomach comprised hyper-
 keratosis, hyperplasia, and papillomas.
'-'Fifteen animals from the ethylene oxide I group developed stomach tumors, of
 which 10 exhibited metastasis and invasive growth into neighboring organs.

SOURCE:   Adapted from Dunkelberg, 1982.
                                   9-97

-------
     Two other studies designed to test for chronic toxicity of ethylene oxide




reported no tumors; however, the exposure and observation periods were too




short to adequately test the carcinogenicity of ethylene oxide in rats,  mice,




monkeys, guinea pigs, and rabbits (Hollingsworth et al., 1956; Jacobson et al.,




1956),




     9.5.1.2.1.  Snellings et al. (1981) Inhalation Study — A 2-year inhala-




tion study (unpublished) was performed by Bushy Run Research Center, Pitts-




burgh, Pennsylvania (Snellings et al., 1981).  Fischer 344 rats were exposed




to 100, 33, and 10 ppm of ethylene oxide vapor by the inhalation route,  6




hours/day, 5 days/ week, for approximately 2 years.  Two groups were exposed




to untreated air under similar conditions.  Whole-body exposures were conducted




in a dynamic exposure system in which the vapor concentration levels were




determined by gas chromatography.  Initially, 120 rats per sex per group were




exposed, with interim sacrifices of 10 animals each at 6 and 12 months and 20




animals at 18 months to determine possible treatment-related effects.  Interim




and terminal evaluation included hematology, serum clinical chemistry, urinaly-




sis, body weight, organ weight, bone marrow cytogenetic studies, and gross and




histologic examinations.




     In the cytogenetic studies, no statistically significant differences were




noted for the "percentage of abnormal cells," the "average number of chromo-




somal aberrations per cell," or the "total number of chromosomal aberrations




(per rat)" for either males_or females exposed to ethylene oxide at  100 ppm




when compared with values obtained for the air-control groups.  However, sta-




tistically significant chromosomal aberrations have been found in other ethy-




lene oxide studies (see section on mutagenicity).




     Histopathologic examination was performed on all tissues of each air-con-




trol group and the 100 ppm  group at 6 months and at 12- and 18-month necropsy





                                   9-98

-------
intervals.  At 6, 12, and 18 months, for the two lower groups (10 and 33 ppm),




this histopathologic examination was performed only when the tissue had gross




lesions.  At the 24-month necropsy interval, the histopathologic examination




was performed on all tissues of rats in the 100 ppm group and both control




groups, and on potential target tissues, selected tissues, and tissues with




gross lesions in the two lower-dose groups  (10 and 33 ppm).




    During the 15th exposure month, all rats became infected with sialodacryo-




adenitis (SDA) virus infection.  Clinical signs of infection were noted during




the 62nd and 63rd exposure weeks.  After the 64th exposure week, the exposures




were temporarily terminated to permit recovery from the viral infection.  Very




low mortality had been observed prior to the infection of the initial 120 rats




per sex per exposure group; no more than five in any group of one sex had died




or were sacrificed because of a moribund condition.  During the 64th and 65th




exposure weeks, a total of 24 rats died.  There was a higher rate of mortality




among female rats in the 100 ppm exposure group than in any other group.  Gross




and microscopic examination of tissues of the animals that died during this




infection period revealed no pathologic findings sufficient to explain the




cause of death.  Most of the clinical signs associated with the infection




subsided after 2 weeks of no exposure, as the mortality rate and body weights




returned to preinfection values.   As a result,  the exposure was restarted.  No




increase in mortality in association with this disease had been reported in the




literature.




     According to Snellings et al. (1981),  the total numbers of rats that died




or were sacrificed in a moribund condition were 49, 39, 28, 31, and 29 for the




males and 53,  31, 25, 19, and 20 for the females in the 100 ppm, 33 ppm, 10




ppm, Air Control 1,  and Air Control II groups,  respectively.  One additional




male in the 33 ppm group and one female in Air Control Group I were acciden-




                                  9-99

-------
tally killed.




     The cumulative mortality data and statistical significances for male and




female rats are shown in Tables 9-19 and 9-20, respectively.  The cumulative




percentage dying in the 100 ppm group for both sexes was significantly higher




than that of controls for at least the last four exposure months of the study.




Very few significant differences were observed in males of the 33 ppm group.




     During the 15th exposure month, the mortality rate of females in the 100




ppm group increased significantly.  This increase was also noted for males in




the 100 ppm group and females in the 33 ppm group, but to a lesser degree.




Since the SDA virus may have contributed significantly to this mortality, the




data were re-evaluated by Snellings et al. (1981), using the number of rats




alive at the beginning of month 17 as the starting point.  This re-evaluation




eliminated the immediate effects of the SDA virus infection.  The results of




these calculations, presented in Tables 9-21 and 9-22, indicate a significant




increase in mortality in the 100 ppm group versus the controls for both males




and females, but the increased mortality was not significant until month 23 for




the males and month 22 for the females.  In no time interval was the cumulative




percentage mortality value for either sex in the 33 ppm group significantly




different from that of combined controls.  However, from the 21st month on, the




values for both sexes in the 33 ppm group were higher than those for both




control groups.  At no time were significant increases in mortality observed in




the 10 ppm exposure group of either sex.




     Of the many tumor types occurring in the Snellings et al. (1981) study,




six types, which may be treatment related, are reviewed here:  subcutaneous




fibroma, peritoneal mesothelioma, pancreatic adenoma, pituitary adenoma, brain




neoplasm, and mononuclear cell leukemia.  The authors presented no evidence




that the SDA viral infection increased the tumor incidence in the experimental




                                    9-100

-------
      TABLE 9-19.   CUMULATIVE PERCENTAGES OF MALE FISCHER 344 RATS THAT DIED OR
   WERE SACRIFICED IN A MORIBUND CONDITION AFTER EXPOSURE TO ETHYLENE OXIDE VAPOR3
Exposure
month
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
24.5
25.0
100 ppmb
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
1.0
3.0
7.0
7.0
7.0
10.4
11.7
18.2
24.7(~,~ja)
27.3(~>~>~)
44.2(a»c»c)
50.7(a,c,c)
55.9(3, b,c)
65'2(a'"'b)
Exposure concentration
Air
33 ppmb 10 ppm Control I
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.9
1.8
2.8
2.8
4.8
4.8
6.8
8.8
9.8
9.8
12.5
15.1
20.3
29.4(~>a>a)
36.o(~»b»b)
39>9(-,a,-)
42.5
54.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.9
0.9
0.9
0.9
0.9
0.9
0.9
1.9
2.9
2.9
2.9
8.0
10.6
14.4
18.3
25.9
31.0
38.3
0.0
0.0
0.0
0.0
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
1.8
1.8
1.8
2.9
5.1
5.1
9.0
11.5
17.9
21.8
12.9
34.6
41.9
Air
Control II
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
1.0
4.1
5.2
5.2
6.5
10.4
11.7
13.0
20.8
28.6
42.6
Combined
controls
0.0
0.0
0.0
0.0
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.9
1.9
1.9
3.5
5.2
5.2
7.8
11.0
14.8
17.4
25.2
31.6
42.3
aLife table analysis, adjusted for scheduled interim sacrifices,
Superscripts in parentheses denote values significantly higher than those of control
 groups.  First letter denotes degree of significance vs. Control I group; second
 letter denotes degree of significance vs. Control II group; third letter denotes
 degree of significance vs. combined controls (C-I plus C-II).

 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001   - = not significant

SOURCE:  Adapted from Snellings et al., 1981.
                                        9-101

-------
      TABLE  9-20.   CUMULATIVE  PERCENTAGES  OF  FEMALE  FISCHER  344 RATS  THAT  DIED  OR
     WERE  SACRIFICED  IN  A MORIBUND  CONDITION  AFTER EXPOSURE  TO ETHYLENE OXIDE VAPORa
Exposure
month 100 ppmb
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
24.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.8
1.8
1.8
2.8
3.9
16.o(b»b»b)
18.o(b»b»b)
21.l(c,c,c)
22.2(b»c>c)
25.o(a»c»c)
30.4(b>a»b)
34.4(b»a»b)
41.3(c,b,c)
49.5(c>b>c)
63.3(c»c»c)
70.0(c,c,c)
33 ppm
0.0
0.0
0.0
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
1.8
1.8
5.9
5.9
6.9
10.3
15.5
16.8
22.0
24.6
32.4
35.2
41.1
Exposure concentration
Air Air Combined
10 ppm Control I Control II controls
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
3.0
5.0
6.2
11.3
11.3
12.6
13.9
24.2
28.5
34.7
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
2.8
3.8
3.8
6.1
8.6
9.9
9.9
9.9
18.8
22.9
25.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.1
3.1
3.1
4.3
5.6
12.3
16.3
18.9
22.9
25.8
25.8
0.4
0.0
0.0
0.0
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
3.0
3.5
3.5
3.5
5.2
7.1
11.0
13.0
14.3
20.8
24.3
25.9
aLife table analysis, adjusted for scheduled interim sacrifices.
bSuperscripts in parentheses denote values significantly higher than those of control
 groups.  First letter denotes degree of significance vs. Control I group; second
 letter denotes degree of significance vs. Control II group; third letter denotes
 degree of significance vs. combined controls (C-I plus C-II).
 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001

SOURCE:  Adapted from Snellings et al., 1981.
not significant
                                       9-102

-------
 TABLE 9-21.   CUMULATIVE PERCENTAGES OF MALE FISCHER 344 RATS THAT WERE ALIVE AT THE
          BEGINNING OF MONTH 17,  BUT DIED OR WERE SACRIFICED IN A MORIBUND
            CONDITION AFTER SUBSEQUENT EXPOSURE TO ETHYLENE OXIDE VAPOR3
Exposure concentration
Exposure Air
month 100 ppmb 33 ppmb 10 ppm Control I
17
18
19
20
21
22
23
24
24.5
25.0
0.0 1.1
3.7 1.1
5.0 4.0
12.0 6.8
19.0 12.6
21.8 22.6
40.o(a,c,0 29.8(->a>-)
46.9<-,c,b) 34.1
52.5(~>b»b) 36.9
62.5(~>~>a) 49.8
1.0
1.0
1.0
6.2
8.8
12.8
16.7
24.5
29.7
37.1
1.0
3.3
3.3
7.2
9.9
16.4
20.3
28.2
33.4
40.8
Air
Control II
2.1
3.2
3.2
4.6
8.6
9.9
11.2
19.2
27.1
41.4
Combined
controls
1.6
3.3
3.3
5.9
9.2
13.2
15.8
23.7
30.3
41.2
aLife table analysis, adjusted for scheduled interim sacrifices.
Superscripts in parentheses denote values significantly higher than those of control
 groups.  First letter denotes degree of significance vs. Control I group; second
 letter denotes degree of significance vs. Control II group; third letter denotes
 degree of significance vs. combined controls (C-I plus C-II).

 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001   - = not significant

SOURCE:  Adapted from Snellings et al.,  1981.
                                       9-103

-------
  TABLE 9-22.  CUMULATIVE PERCENTAGES OF FEMALE FISCHER 344 RATS THAT WERE ALIVE AT
        THE BEGINNING OF MONTH  17, BUT DIED OR WERE SACRIFICED IN A MORIBUND
          CONDITION AFTER SUBSEQUENT EXPOSURE TO ETHYLENE OXIDE VAPOR3
Exposure
month 100 ppmb
17
18
19
20
21
22
23
24
24.5
3.7
5.1
8.4
15.1
20.1
28.4(b,-,a)
38.4 p > 0.01
0.01 > p > 0.001    c = p < 0.001   - = not significant
cControl I group differed significantly from Control II group at the p < 0.05
 level only for the 22-month mortality count.

SOURCE:   Adapted from Snellings et al., 1981.
                                     9-104

-------
groups.  The time to first  tumor for some neoplasms  (but not for mononuclear




cell leukemias) was decreased in the high-dose group as compared to controls,




as shown in Table 9-23.  Median time-to-tumor was not reduced.




     Histopathologic examinations were performed on tissues of all the rats in




the 100 ppm group and both  control groups.  In the 33 and 10 ppm groups, only




those tissues that had gross lesions were examined.  Therefore, some small




tumors in these two groups  may have been missed, yielding an erroneously low




estimate of tumors.




     In male rats sacrificed at 24 months, a statistically significant increase




in subcutaneous fibromas (10/28, 35.7%) was observed in the group exposed to




100 ppm ethylene oxide as compared with combined controls (3/91, 3.3%) (Table




9-24).  An increased prevalence of these tumors was also observed in the 10 ppm




group (8/48, 17%); however, this increase was not significant.  No increase in




subcutaneous fibromas was observed in the 33 ppm group.   The authors concluded




that the increased prevalence of subcutaneous fibromas in the 100 ppm group




represented an effect of treatment.  It should be noted, however, that histo-




logic examinations were performed only on skin sections  that showed gross




lesions; therefore, many tumors too small for gross detection were probably




missed.   When the incidences of this tumor type were added to those for animals




that died spontaneously or were euthanized when moribund,  the totals were even




higher in both the 100 and  10 ppm groups than in the controls (Table 9-25).




     An increase in the frequency of peritoneal mesothelioma was observed in




all of the male treatment groups sacrificed at 24 months (4/30 at 100 ppm,




4/39 at 33 ppm,  2/51 at 10 ppm vs.  1/48 for the Control  I group and 2/84 for




the Control II group) (Table 9-24).  Although the increase was riot significant




at any dose level, this enhanced prevalence in the 100 and 33 ppm groups is




considered a treatment-related effect.   This tumor was also found in a large





                                 9-105

-------
TABLE 9-23.  SUMMARY OF SELECTED TUMOR INCIDENCE COMPARISONS FOR MALE AND FEMALE
             FISCHER 344 RATS EXPOSED TO ETHYLENE OXIDE FOR 2 YEARS
Ethylene oxide
concentration
ppm
100
33C
10C
0-1
O-II
100
33C
10C
0-1
O-II
100
33c
10C
0-1
O-II
100
33c
10C
0-1
O-II

100
33C
10C
0-1
O-II
Time in months to:
Total number of rats

With tissues examined With tumora
Mononuclear cell leukemia - Males
119 26
81
79
116
118
25
21
20
18
Mononuclear cell leukemia - Females
113 28(c»b>c)
79
77
118
117
Peritoneal mesothelioma
119
91
89
114
116
Pituitary adenoma -
117
79
80
117
117
Pituitary adenoma -
117
90
90
119
116
24(c,c,c)
14
9
13
- Males
22(c,c,c)
y(a, a, a)
3
2
2
Males
27
16
27
28
22
Females
32
38
39
38
38
First
tumor
19
13
20
18
21
18
18
19
19
18
15
18
20
18
20
15
15
18
17
18

10
17
16
15
18
Median
tumorb
24
25
25
23
25
24
24
25
24
23
23
25
—
—
—
25
25
25
25
25

24
25
24
25
25
Superscripts in parentheses denote values significantly higher than those of
 control groups.  First letter denotes degree of significance vs.  Control I
 group; second letter denotes degree of significance vs. Control II group; third
 letter denotes degree of significance vs. combined controls (C-I  plus C-II).

 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001   -  = not significant

"Medians were not presented if the total number of a particular tumor was
 three or less.
C0nly organs with gross lesions were histologically examined from  this exposure
 level at the 6-, 12-, and 18-month sacrifice intervals.

SOURCE:  Adapted from Snellings et al., 1981.

                                      9-106

-------
         TABLE 9-24.  ETHYLENE OXIDE 2-YEAR VAPOR INHALATION STUDY:
    24-MONTH FINAL SACRIFICE FREQUENCY OF EXPOSURE-RELATED NEOPLASMS FOR
                     110- TO 116-WEEK-OLD FISCHER 344 RATS
                                           ppm of Ethylene Oxide

Organs/Findings/Sex   100a          33a            10a       Control I   Control II


Total number
examined grossly
  Male                 30            39             51            48        49
  Female               26            48             54            60        56

Pituitary
  Adenomas
  Male               12/29b        13/39          15/51          16/48     13/49

Pancreas0
  Adenomas
  Male                5/30          1/2            2/3            2/48      5/49

Subcutisd
  Fibromas
  Male               10/28(c>c>c)   1/34           8/48<-a»a»b)    1/44      2/47

Peritoneum
  Mesotheliomas
  Male                4/30          4/39           2/51           1/48      1/49

Spleen
  Mononuclear
  cell leukemias
  Male                8/30         10/39           9/51           5/48      8/49
  Female
14/48(b>b>b)    ll/54(-.~»a)    5/60      6/55
Superscripts in parentheses denote values significantly higher than those of control
 groups.  First letter denotes degree of significance vs. Control I group; second
 letter denotes degree of significance vs. Control II group; third letter denotes
 degree of significance vs. combined controls (C-I plus C-II).

 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001   - = not significant

bNumerator equals number of rats with specified finding.  Denominator equals
 number of rats for which specified tissues were examined.
"-Tissues from 33- and 10-ppm groups examined only if gross lesions were present.
 Since tissues were not examined from all rats, data from the 33- and 10—ppm
 groups were not statistically compared with data from other groups.
^Examined only if gross lesions were present (except flank region skin and
 subcutis, which was routinely examined microscopically).

SOURCE:  Adapted from Snellings et al., 1981.


                                         9-107

-------
    TABLE 9-25.  ETHYLENE OXIDE 2-YEAR VAPOR INHALATION STUDY:   FREQUENCY OF
         EXPOSURE-RELATED NEOPLASMS AT 24-MONTH FINAL SACRIFICE AND IN
       FISCHER 344 RATS DYING SPONTANEOUSLY OR EUTHANIZED WHEN MORIBUND3
Organs/Findings/Sex   100b
                                           ppm of Ethylene Oxide

                                     33b           10b      Control I   Control II
Pituitary
  Adenomas
  Male

Pancreas^
  Adenomas
  Male

Subcutis6
  Fibromas
  Male

Peritoneum
  Mesotheliomas
  Male

Spleen
  Mononuclear
   cell leukemias
  Male

  Female
24/79c
16/79
                                                  26/79
                2/32
                                                                24/79
                                                                 2/80
                     ll/80(b,-,a)   1/43



                     15/78(c,b,c)   3/75          10/77(b,a,b)   !/76



                     21/80(c.c»c)   6/80(~»-»a)    3/80          1/80
                     25/80
                                   23/80
               21/80

               14/80
                                           20/80

                                            9/80
19/78



 5/80



 3/78



 2/80




18/80

13/76
aConcerning the animals that died spontaneously or were euthanized when moribund,
 it was not specified whether tissues were examined microscopically only when
 gross lesions were present, or if all tissues were reviewed in this way.
 It is therefore assumed that all of the tissues from these animals were studied
 histologically, whether or not gross lesions were observed.  Not to have per-
 formed such studies would have yielded erroneously low frequencies of exposure-
 related neoplasms.
Superscripts in parentheses denote values significantly higher than those of control
 groups.  First letter denotes degree of significance vs. Control I group; second
 letter denotes degree of significance vs. Control II group; third letter denotes
 degree of significance vs. combined controls (C-I plus C-II).

 a = 0.05 > p > 0.01    b = 0.01 > p > 0.001    c = p < 0.001   - = not significant

cNumerator equals number of rats with specified finding.  Denominator equals
 number of rats for which specified tissues were examined.
"^Tissues from 33- and 10-ppm groups were examined only if gross lesions were
 present.  Since tissues were not examined from all rats, data from the 33- and
 10-ppm groups were not statistically compared with data from other groups.
eExamined only if gross lesions were present (except flank region skin and
 subcutis, which was routinely examined microscopically).

SOURCE:  Adapted from Snellings et al., 1981.

                                        9-108

-------
 number  of  treated  animals  that  died  spontaneously  or  were  euthanized when




 moribund.   When  the  tumor  incidence  in  this  latter group was  added  to  that  for




 animals  sacrificed at  24 months,  the numbers were  much  higher than  controls  and




 were  statistically significant  for the  high-dose group  versus controls  (21/80




 at  100 ppm,  6/80 at  33  ppm,  3/80  at  10  ppm vs.  1/80 for the Control I  group  and




 2/80  for the Control 11 group)  (Table 9-25).




      Pancreatic adenomas were statistically  significant for the male high-dose




 group sacrificed at  24  months and the animals that died spontaneously  or were




 euthanized  when moribund (11/80 at 100  ppm,  1/43 at 33  ppm, 2/32 at 10  ppm vs.




 2/80  in  the  Control  I group  and 5/80 in the Control II  group)  (Table 9-25).




 Tissues  from the 33  and 10 ppm  groups were examined only if gross lesions were




 present  in  the 24-month sacrifice group, which may explain the paucity  of




 tumors in these groups  (Table 9-24).  The denominator in Table 9-25, the number




 of  rats for  which  the specified tissue  was examined, may be erroneously high




 for the data combining  the 24-month  sacrifice with the  animals that died




 spontaneously or were euthanized  when moribund.




     While Tables  9-23 and 9-24 show no significant increase  in the frequency




 of pituitary adenomas in the groups  of  treated males,  Table 9-23 shows some




 indication of a decreased time-to-tumor.  In males, the first pituitary adenomas




 appeared at  15 months in the 100  and 33 ppm groups, and in the 17th or 18th




 month in all other groups;  in females,  the corresponding times were 10 months




 for the 100  ppm group versus at least 15 months for all other groups.   The




 time-to-tumor decreased significantly with increasing dose (p < 0.01 for males,




 p < 0.0001 for females), suggesting  that the normal incidence of pituitary




 adenomas was accelerated by exposure to ethylene oxide.




     An increased frequency of mononuclear cell leukemia was  observed  in the




ethylene oxide-treated animals at  the 24-month sacrifice interval (Table 9-24).





                                     9-109

-------
Statistical significance was observed in females in both the 100 and 33 ppm




groups versus combined controls (p < 0.01).  The responses for the 24-month




sacrifice were 15/26 (58%), 14/48 (29%), and 11/115 (10%) for the 100, 33, and




10 ppm groups and combined controls, respectively.  The frequencies for male




rats were not significantly increased in the treated versus the control groups.




     In females, the results for animals dying spontaneously or euthanized when




moribund and for those sacrificed at 24 months remained statistically signifi-




cant for the two higher-dose groups versus combined controls.  The frequencies




for females (Table 9-25) were 27/80 (34%), 24/80 (30%), 14/80 (18%), and 22/156




(14%) for the 100, 33, and 10 ppm groups and combined controls, respectively,




with statistically significant differences in the two higher-dose groups versus




combined controls (p < 0.01) and a significantly positive linear dose-response




trend (p < 0.01).  The trend became even stronger (p < 0.00001) when the pro-




portions were adjusted for early mortality.  These data suggest that exposure




to ethylene oxide not only increased the total incidence of leukemia but also




accelerated its rate of development (Figure 9-3).  The authors also reported




that the number of female rats with three or more tumors was significantly




(p < 0.001) increased in the 100 ppm group as compared to the controls.




     A letter to the U.S. Environmental Protection Agency (Browning, 1982)




stated that a recent histologic examination of all brain tissue from the




Snellings et al. (1981) study revealed the presence of primary brain neoplasms




(see Tables 9-26 and 9-27, and Table 9-33 in Section 9.5.3.3.2).  These tumors




were shown to be statistically significant by the Fisher Exact Test in both




males and females.




     In summary, ethylene oxide has produced significant increases of several




tumor types in rats.  A dose-related increase in mononuclear cell leukemia




occurred in female rats.  The occurrence of pituitary adenoma appeared to be




                                    9-110

-------
   60
   50


til  40
o
z
UJ
o
(E
MALE
                           I
                                      I
                                                           I
               20         40         60        80        100


                CONCENTRATION OF ETHVLENE OXIDE, ppm
        Figure 9-3.  Percentages of male and female Fischer 344 rats with
        histologically confirmed mononuclear cell leukemia at 24-month
        sacrifice.


        Source:  Snellings et al. (1981).
                                 9-111

-------
       TABLE 9-26.  ETHYLENE OXIDE 2-YEAR VAPOR INHALATION STUDY: FREQUENCY OF
                     PRIMARY BRAIN NEOPLASMS IN FISCHER 344 RATS
                                            Exposure level (ppm)


     Sex         100           33            10           0 (CI)            0 (CII)


                              18-month sacrifice8

   Male          0/20          1/20          0/20         0/20              0/20
   Female        1/20          0/20          0/20         1/20              0/20

                              24-month sacrifice5

   Male          3/30          1/39          0/51         1/48              0/49
   Female        2/26          2/48          0/51         0/60              0/56

                               Dead/euthanized moribund3

   Male          4/49          3/39          1/28         0/30              0/29
   Female        1/53          1/31          1/24         0/18              0/20

              18- and 24-month sacrifices and dead/euthanized moribund5
                                (Combined from above)

   Male          7/99          5/98          1/99         1/98              0/98
   Female        4/99          3/99          1/95         1/98              0/96

                                    2-year study*3
(Combined 6-,  12-,  18-,  and 24-month sacrifices and dead/euthanized moribund animals)

   Male          7/119         5/118         1/119         1/118             0/118
                 P=0.002C      p=0.017c

   Female        4/119         3/119         1/115         1/118             0/116
                 p=0.045c

   ^Numerator  equals the number of  brains with primary  neoplasms.   Denominator
    equals total number  of brains examined microscopically.
   ^Numerator  equals the number of  brains with neoplasms.   Denominator equals total
    number of  brains examined microscopically.  Although animals sacrificed at
    6 and 12 months  are  included, no brain neoplasms were  discovered in these
    groups.  The 6-  and  12-month animals  can be eliminated by subtracting 20 from
    each denominator.
   GFisher Exact Test.

   SOURCE:  Adapted  from Snellings  et al.,  1981.


                                     9-112

-------
          TABLE 9-27.  ETO 2-YEAR VAPOR  INHALATION STUDY:  FREQUENCY OF
                PRIMARY BRAIN NEOPLASM TYPES IN FISCHER  344 RATS
          (Combined data for 6-,  12-, 18-, and 24-month sacrifices, and
                       dead/euthanized moribund animals)
  Neoplasm type
  100
                                         Exposure level  (ppm)
33
10
0 (CI)
0 (CII)
Granular cell tumor

Astrocytoma/oligodendro-
  glioma/mixed glioma

Malignant reticulosis-
  microglioma
                                                   Males3
1/119      1/118     1/119     0/118       0/118
5/119      2/118     0/119     1/118       0/118
1/119      2/118     0/119     0/118       0/118
Granular cell tumor
Females3
1/119 1/119 0/118 1/118

0/116
Astrocytoma/oligodendro-
  glioma/mixed glioma          2/119      2/119     1/118     0/118       0/116

Malignant reticulosis-
  microglioma                  1/119      0/119     0/118     0/118       0/116

aNumerator equals the number of brains with primary neoplasms.  Denominator equals
 total number of brains examined microscopically.  Although animals sacrificed at
 6 and 12 months are included, no brain neoplasms were discovered in these groups.
 The 6- and 12-month animals can be eliminated by subtracting 20 from each
 denominator.

SOURCE:  Adapted from Snellings et al., 1981.
                                    9-113

-------
 accelerated  in  female  rats  exposed  to  100 ppm, although there was no  statisti-




 cally  increased incidence of  these  tumors.  The frequency of peritoneal meso-




 thelioma was  treatment-related  in the  male rats exposed to 100 and  33 ppm.




 Further, a significant  increase occurred in subcutaneous fibromas in male rats.




 Increases in  brain  neoplasms  were also observed in both sexes.









     9.5.1.2.2.  Lynch  et al. (1982) Inhalation Study  (NIOSH) — Another chronic




 inhalation study (unpublished draft) on ethylene oxide and propylene oxide was




 performed by  the National Institute for Occupational Safety and Health (NIOSH)




 (Lynch et al.,  1982).   In the present  report, only the preliminary  findings of




 the ethylene  oxide  section  of the study will be discussed.  Male Fischer 344




 rats (80 in each group) and 12 male cynomolgus monkeys were exposed to ethylene




 oxide at either 50  or 100 ppm for 7 hours/day, 5 days/week,  for 24 months.




 Each treatment  group consisted of 80 rats and 12 monkeys at the start of the




 study.   Rats  and monkeys were housed together in the same chambers during the




 7-hour exposure period.  Food and water were available ad libitum except during




 the exposure  periods.   In analyzing for carcinogenicity, only limited data were




 available for monkeys because of their longer lifespans; however,  the authors




 reported that there was no evidence of leukemia in any of the exposed monkeys.




     An overall statistically significant depression in weight gain was noted




 for ethylene oxide-exposed rats.  This development,  which appeared to begin at




about week 7 for the 100 ppm group and at week 15  for the 50 ppm group,  con-




 tinued  throughout the study.  Survival was  also adversely affected by exposure




 to ethylene oxide,  with estimated mean survival times of greater than 720 days




for the controls, 690 days for the 50 ppm group,  and 653 days for  the 100 ppm




group.   An outbreak of mycoplasma infection also caused an abrupt  decline in




survival at about 480 days into  the study.





                                     9-114

-------
     With respect  to pathology,  the  authors  reported that the livers  and

 spleens  of the  ethylene oxide-exposed rats were  the  only organs  for which

 histopathologic evaluations  were completed.   While the results are prelimi-

 nary  (Table 9-28),  the  data  obtained at  terminal sacrifice indicate that the

 incidence of leukemia followed a dose-response pattern ranging from 33.3% in

 controls  to 64.3%  in the 100 ppm group (p =  0.07,  Table 9-28).   The one-tailed

 test for  linear trend at terminal sacrifice  was  significant  at the p  <  0.05

 level.  Using a two-tailed test,  the significance  level was  p =  0.08.   These

 preliminary data,  therefore,  do  provide  some evidence  of ethylene  oxide-induced

 leukemia.   The  data  from moribund sacrifice  and  deaths (Table 9-28) merely

 accentuate both the  early toxicity and the mortality  in the  100  ppm group as

 compared  with the  other groups,  and  the  relatively high leukemia rates  in these

 rats.  Neither  these  rats nor  the total  was  significantly  higher than controls.
             TABLE 9-28.  LEUKEMIA INCIDENCE IN MALE FISCHER 344
                  RATS EXPOSED TO ETHYLENE OXIDE FOR 2 YEARS3
Treatment group
 Terminal
 sacrifice
  only (%)
                                            Leukemia incidence
Moribund
sacrifice
and death (%)
Terminal
sacrifice
plus moribund
sacrifice and
death (total)
Control

Ethylene oxide, 50 ppm

Ethylene oxide, 100 ppm
 7/21  (33.3%)

12/27  (44.4%)

 9/14  (64.3%)b
 5/18 (27.9%)

26/52 (50.0%)

21/62 (33.9%)
12/39 (30.8%)

38/79 (48.1%)

30/76 (39.5%)
aBased on histopathologic evaluation of spleens.
bp = 0.07 based on the one-tailed Fisher Exact Test.

SOURCE:   Lynch et al., 1982.
                                  9-115

-------
     Lynch et al. (1982) also reported that exposure to ethylene oxide signifi-




cantly increased the incidence of peritoneal mesotheliomas.   These tumors were




present on the tunica vaginalis surrounding the testes and epididymis, and




occasionally spread to the peritoneal cavity.  A nonsignificant increase in




pheochromocytoraas was observed in exposed groups (Table 9-29).




     Lynch et al. (1982) reported the following incidences of mixed-cell gliomas




in male rats:  0/76 in controls,  2/77 in the 50 ppm group, and 5/79 in the 100




ppm group.  The term "glioma" was used because the tumors contained both astro-




cyte and oligodendroglia cells within the tumor.  These findings are significant




because the above-described tumors are unusual in Fischer 344 rats.  Additional




data collected from this study are currently being evaluated, and a final com-




prehensive report is scheduled to be published within a year.









     9.5.1.2.3.  Summary of Animal Studies — The Snellings  et al. (1981)




study, which showed an increase in leukemia in Fischer 344 rats, is also




supported by a preliminary NIOSH study (Lynch et al., 1982,  Table 9-28) in




which an increase in leukemia appeared in rats of the same strain but of a




different sex and with mycoplasma instead of SDA viral infections.  Increases




in peritoneal mesotheliomas were observed in both studies (Snellings et al.,




1981 and Lynch et al., 1982), and significant increases in subcutaneous fibro-




mas in the males were observed in the Snellings study.  Snellings et al. (1981)




also concluded that the frequencies among female rats with more than two neo-




plasms were significantly greater for all three groups when compared to com-




bined controls.




     Further, both studies found significant increases in brain neoplasms, a




development that requires further review in terms of its possible value for




risk evaluation.  Like the finding of gliomas in male rats reported previously,




                                    9-116

-------
these studies are significant because brain neoplasms are unusual in the Fischer

344 strain of rats.

     In 1980, the National Toxicology Program (NTP) began a cancer bioassay in

B6C3F1 mice (inhalation exposure).  Exposure to ethylene oxide at 0, 5, and 100

ppm for 6 hours/day, 5 days/week began in August 1981.  The final report is

expected in mid-1985.
      TABLE 9-29.  INCIDENCE OF NEOPLASTIC LESIONS IN MALE FISCHER 344
                  RATS EXPOSED TO ETHYLENE OXIDE FOR 2 YEARS3
  Organs/Findings
                                              Exposure level (ppm)
Control
    50
 100
Adrenal
Pheochromocytomas
Brain
Gliomas (mixed-cell)
8/78
0/76
14/77
2/77
13/78
5/79
Body cavity
  Peritoneal mesotheliomas
Spleen
  Mononuclear cell leukemia
  3/78
 24/77
                                                                (p = 0.032)b
    9/79         21/79
           (p = 4.95 x 10~5)b
   38/79
(p = 0.22)b
30/76
aEach group consisted of 80 male rats.  Denominators of less than 80 reflect
 tissues accidentally lost or tissues that could not be examined histologically
 due to autolysis.
bFisher Exact Test.

SOURCE:  Lynch et al. , 1982.
                                    9-117

-------
9.5.2.  Epidemiologic Studies









     9.5.2.1.  JOYNER (1964) -- Joyner (1964) conducted a health evaluation of




employees at an ethylene oxide plant in Texas.  The evaluation included a phy-




sical examination of 37 male ethylene oxide operators, aged 29 to 56, and a




similar number of age-matched controls.  The operators were reported to have




been exposed to ethylene oxide at approximately 5 to 10 ppm for the durations




of their service.  The controls, who were chosen from operators assigned to




other production units, had been exposed to many different agents encountered




in the petrochemical industry.  The author stated that the mean length of




service for the control group was 11 2/3 years, as compared with 10 2/3 years




for the exposed group.  The author used company medical records for the period




1952-1963 to compare the exposed group and controls with respect to days lost




for illness, specific diagnoses, and initial visits for respiratory, gastro-




intestinal, or genitourinary complaints.   The author found that the ethylene




oxide operators who were currently employed exhibited less absenteeism, fewer




symptoms,  and fewer diagnosed illnesses (including malignant neoplasms) than




the controls.




     The author also reviewed the medical records of nine operators who had




experienced accidental exposures in the previous 10 years, and seven workers




other than operators who had experienced accidental exposures in the previous




8 years.  Twelve of the accidental exposures were reported to be dermal expo-




sures,  while three were reported to be inhalation exposures; one exposure was




reportedly to  "vapor."  Most of the dermal exposures produced burns.  The vapor




exposure produced conjunctivitis.  Two of the persons with inhalation exposure




suffered no symptoms; the third developed nausea and vomiting, which lasted




several hours.   The authors reported that the persons identified as having had




                                  9-118

-------
accidental exposures did not exhibit any recurring medical problems.  The one




person who had suffered symptoms from the inhalation exposure was no longer




with the company and could not be traced.




     Additionally, the author reviewed the medical records of eight persons who




had previously worked as ethylene oxide operators for 100 months or more but




who had since been transferred to another division.  Among persons formerly




employed as ethylene oxide operators for 100 months or more, no significant




differences were found in the incidence of illness, symptoms, complaints, or




absenteeism when compared to the study cohort or to controls; very few data




were presented in this regard, however.




     This study is inadequate for use in evaluating the carcinogenicity of




ethylene oxide for several reasons.  First, it is primarily a cross-sectional




study of ethylene oxide operators who were employed as such at the time of the




study.  Workers who had developed cancer would probably no longer have been




employed at the plant.  Secondly, the period of observation, which in this




study is the same as the duration of exposure for the current operators, may




have been too short to allow adequate assessment of a carcinogenic effect.




Cancer latency may be as long as 20 to 30 years; the longest observation period




among current operators in this study was 16 1/3 years.  The mean exposure for




current operators was 10 2/3 years.  For those with accidental exposures, the




longest follow-up was 10 years.  For the eight workers with over 100 months (8




1/3 years) of exposure, the length of follow-up was not indicated.   Third, the




sample sizes studied were so small that only an extremely large carcinogenic




effect could be detected.









     9.5.2.2.  EHRENBERG AND HALLSTROM (1967) — Ehrenberg and Hallstrom (1967)




conducted a hematologic investigation of workers at a factory that  manufactured





                                    9-119

-------
and used ethylene oxide.  A preliminary investigation in 1960 revealed certain




hematologic differences between 28 exposed persons who worked in an area of




the factory "where leakage of ethylene oxide from tube joints, pumps, etc. was




possible (and at least occasionally occurred)," and 26 controls in other




departments not working in contact with ethylene oxide.  The sex of the study




subjects was not reported.  The ages of persons in the exposed group were




reported to be about the same as those in the control group.  The exposed




persons were reported to have been active in the ethylene oxide department for




2 to 20 years, with an average of 15 years.  One case of leukemia (chronic




lymphatic type) was observed in the exposed group; the expected number of




leukemia cases in the exposed group was not reported.  No cases of leukemia




were found in the controls.  Three cases of anisocytosis were found in the




exposed group and none in the controls, a finding which the authors suggested




may indicate a disturbed bone marrow function.   Hemoglobin values were reported




to be significantly (p < 0.05) lower in the exposed group than in the controls,




and lymphocytes per mm^ were reported to be significantly (p < 0.01) higher in




the 27 exposed healthy persons than in the 20 healthy controls (the presence of




disease may affect the white blood cell count;  thus,  only "healthy" persons




were considered in the latter comparison).  It  should be noted that three




persons who were reported to have been accidentally exposed to high levels of




ethylene oxide were added to the exposed group  for the lymphocyte/mm^ compari-




son (for a total of 31 persons in the exposed group).  The authors did not




state where these three persons worked or even whether they worked in the




factory.




     Because of these differences relating to hemoglobin and lymphocytes, and




because ventilation was improved in the plant,  the authors did a second study




of the factory workers in 1961.  The second study was expanded to include all




                                    9-120

-------
 of  the  workers  in the  plant.   Workers  were  divided into four categories:   "66




 persons not  working with ethylene  oxide  (including the 1960 control group);




 86  persons  intermittently working  in ethylene  oxide premises;  54  persons  who




 had once been working  in contact with  ethylene oxide for some  period of time;




 and 37  persons  permanently working in  the ethylene oxide area  (including  the




 1960 exposed group)."   The only hematologic  analysis in the second  study  was




 for lymphocytes.   The  authors  found an elevated lymphocyte  count  in the ex-




 posed group  as  compared with controls, but this difference  was not  significant




 (p  > 0.05) for  either  healthy  individuals or the  total group.  The  authors




 suggested that  this  lack of a  significant difference could  possibly  be attri-




 buted to improved  ventilation  and  safety control  in the  factory,  the  small




 number  (17)  of  healthy  persons in  the  group  permanently  exposed (versus 27




 healthy  exposed individuals in the  1960 investigation),  and/or the  average age




 difference between  the  exposed and  control groups.   The  average age of the




 enlarged control group  was  reported  to be "significantly" lower than  that of




 the  exposed  group, and,  in general,  a  decrease  in  lymphocyte count with age was




 found.   A significant age  difference between the exposed group and  the controls




was  not  present in the  1960 examination.   It should  be noted that for those




persons  examined in the  1960 investigation,  a  significant difference  in average




lymphocyte count between the exposed group and  the controls occurred again when




the  two  groups were examined in 1961.




     The authors also compared the number of chromosome aberrations in eight




persons accidentally exposed to ethylene  oxide with  that in a control group of




10 persons,  and found that chromosome aberrations were significantly elevated




in the exposed group.  Details  of  the statistical analysis were not given.




     In conclusion, Ehrenberg and  Hallstrom (1967) found one leukemia case




among 28 workers exposed to ethylene oxide.   The authors indicated that the




                                   9-121

-------
probability of such an occurrence was small, but its statistical significance




was not calculated.  The results of the study also suggest that ethylene oxide




may elevate lymphocyte counts and reduce hemoglobin values.









     9.5.2.3.  HOGSTEDT ET AL.  (1979a, 1984) — A follow-up study for the years




1961 through 1977 of these same workers with regard to mortality and cancer




incidence was done by Hogstedt et al. (1979a).  A subsequent study by Hogstedt




et al. (1984) followed the mortality of the cohort for the years 1978 through




1982 and the incidence for the years 1978 through 1981.  A period of at least




10 years of follow-up from date of first employment was required in order for a




member of the cohort to be considered at risk.  The authors reported that the




workers in this factory were exposed to various chemicals.  During the period




from 1941 to 1947, it was estimated that the air concentrations were 5 mg/m-^




ethylene chlorohydrin, 100 mg/m^ ethylene dichloride, 0.05 mg/m^ bis(2-chloro-




ethyl) ether, and 600 mg/rn^ ethylene.  The authors also cited the possibility




that concentrations up to 1,199 times greater than those reported may have




occurred for short periods of time.   For ethylene oxide, the exposure was




reported to be probably < 25 mg/m^,  although there were occasional exposures to




the chemical at 1300 mg/m^ (odor threshold).  During the 1950s and until 1963,




the authors reported that the average air concentration of ethylene oxide in




the factory was probably 10 to 50 mg/m ,  although peaks above the odor thresh-




old still occurred.   Random samples  in the 1970s showed a range of 1 to 10




mg/m^ for ethylene oxide and 10 to 25 mg/m^ for propylene oxide, with the latter




concentrations occasionally being as high as 120 to 150 mg/m^.




     Hogstedt et al. (1979a) reported that the study included three subcohorts:




66 men who had never taken part in work involving exposure to ethylene oxide,




86 "intermittently" exposed men (maintenance workers),  and 89 men whose work




                                     9-122

-------
involved full-time exposure.  The same number of men were reported by Hog-




stedt et al. (1984) for the non-exposed and full-time exposed groups, but they




reported only 79 men in the "intermittently" exposed group.  No explanation




for this difference was provided by the authors.




     By the end of the second follow-up of these workers (Hogstedt et al.,




1984), a total of 12 cancer deaths had been observed in the full-time exposed




group, while only 4.8 were expected.  The Carcinogen Assessment Group (GAG)




calculated that the probability of this occurring was less than 0.01.  There




were no statistically significant differences between the observed and expec-




ted number of cancer deaths in the other two exposure groups.  Seven of the 12




cancer deaths seen in the fulltime exposed cohort were either from cancer of




the stomach (5 deaths) or from leukemia (2 deaths).  Deaths from both causes




were significantly elevated in comparison with the numbers expected (5 observed




versus 0.6 expected for stomach cancer, p < 0.01, as calculated by the GAG; and




2 observed versus 0.15 expected for leukemia deaths, p < 0.05, as calculated by




the GAG).  One of the leukemia deaths was from chronic lymphatic leukemia, and




the other was from acute myeloid leukemia.  The death from chronic lymphatic




leukemia may well have been the same case that was reported in the Ehrenberg




and Hallstrom (1967) study.  Excess mortality for cerebrovascular disease




during the period 1961 through 1982 was also statistically significant (5




observed, 1.5 expected; p < 0.05, as calculated by the GAG) among the full-




time operators.  Although the maintenance group showed no overall excess can-




cer mortality, the cancer deaths that occurred in this group were restricted




to cancers of the esophagus and stomach and to leukemia.  The leukemia death




was from chronic lymphatic leukemia.




     Hogstedt et al. (1984) examined the observed and expected numbers of




deaths from various causes, including different cancer sites, by 1-4, 5-9,




                                    9-123

-------
and 10+ years of exposure.  No response by length-of-exposure was found for




any of the causes of death, but the data were rather limited for this type of




analysis.




     Cases of cancer in the study group were determined using the Swedish




Cancer Registry.  No indication of the completeness of ascertainment of the




Registry was given.  Seventeen cases versus 7.9 expected (p < 0.01) were




identified among full-time exposed workers for the period 1961 through 1981.




These included three cases of leukemia versus 0.24 expected (p < 0.01, as




calculated by the GAG).  Included in the three leukemia cases were the indivi-




dual with acute myeloid leukemia and the individual with chronic lymphatic




leukemia, both of whom had died, as well as an individual with chronic myeloid




leukemia.  One case of stomach cancer was reported in addition to the five




cases in which individuals had died.  The expected number of stomach cancers




was not indicated.




     In summary, deaths from cancer of all sites, deaths from stomach cancer,




and deaths from leukemia were each significantly (p < 0.05) elevated among the




full-time exposed cohort.  The total number of malignancies and the number




of leukemia cases were also significantly (p < 0.01) elevated in this group.




Workers in the full-time exposed cohort were exposed to several chemical




agents, however, and the excess cancer incidence and mortality in this cohort




cannot necessarily be ascribed to the ethylene oxide exposure.









     9.5.2.4.  HOGSTEDT ET AL. (1979b, 1984) — Hogstedt et al. (1979b, 1984)




reported on the morbidity and mortality of a group of sterilizing operators




exposed to 50% ethylene oxide and 50% methyl formate.  The 1979b report indi-




cated that only seven persons worked with the sterilization process.  However,




treated boxes (supposedly containing the sterilized equipment) were stored in a




                                      9-124

-------
hall where  30 women worked, and  because  of  leakage  from  the  boxes,  the average




exposure in the  storage hall was  reportedly higher  than  in the sterilization




room.  Exposure  measurements made in  1977 showed storage hall concentrations of




2 to 70 ppm with 8-hour time-weighted average concentrations being  calculated




at 20± 10  ppm.  The concentration was 1,500 ppm inside  newly sterilized boxes




and 150 ppm on the floor outside  the  boxes.




     During the  period from 1968  to 1977, 70 persons had been employed at some




time in the storage hall, and another 160 had been  employed  in the  neighboring




rooms  or as  sterilizing operators.  Of these, 69 (63 women,  6 men)  and 134




(90 women,  44 men) had worked for a year or more, respectively (Hogstedt et




al., 1984).  In  the Hogstedt et al. (1979b) report, the authors indicated that




among  this  group of workers three cases of leukemia had occurred.   One of these




reported cases was actually a Waldenstrom's macroglobulinemia, however, and




it was subsequently reported (Hogstedt et al., 1984) that this case should have




been classified as a non-Hodgkin's lymphoma.  However, according to the Eighth




Revision of  the International Classification of Diseases (ICD),  which is the




revision used in the Hogstedt et  al. (1984) report, Waldenstrom's macro-




globulinemia is classified as a plasma protein abnormality and not as a neoplasm.




     The two leukemia cases were  among women who worked in the storage hall.




One of  the  cases was a woman who  began working in the storage hall in 1966,




was diagnosed with chronic myeloid leukemia in early 1972 at the age of 51,




and died in  1977.  The other case was a woman who began working in the storage




hall in 1968, was diagnosed with acute myelogenetic leukemia in early 1977




at the  age of 37, and as of July  1978 was reported to be in complete remis-




sion.   Hogstedt et al.  (1984),  however,  reported that she subsequently died




during  the extended follow-up period (1978-1982).  A third leukemia death was




reported by Hogstedt et al.  (1984).  This case was  a woman who had had inter-





                                      9-125

-------
mittent exposure, passing through the storage hall 2 to 4 times per day.  She




was employed during 1969-1972, and in 1979 was diagnosed with blast leukemia




and died the same year.




     Two males were reported to have died from cancer in the period 1968-1982.




One was the case of Waldenstrom's macroglobulinemia mentioned earlier.




As previously indicated, this death would not be considered a cancer death by




the Eighth Revision of the ICD.  The other death from cancer was not specified




by site.  Observed and expected cancer mortality, total mortality, and mortal-




ity from cancer of the lymphatic and hematopoietic systems for both the earlier




(Hogstedt et al., 1979b) and later (Hogstedt et al., 1984) observation periods




as well as the total observation period is reported in Table 9-30.




     With regard to morbidity, ten cases of cancer among the female sterilizer




workers had been reported to the Cancer Registry versus 5.2 expected  (p < 0.05,




as calculated by the GAG) during the period 1961-1981; there were three cases




of cancer among male sterilizer workers (excluding the case of Waldenstrom's




macroglobulinemia) versus 1.8 expected.  The excess morbidity among females was




mainly due to the three cases of leukemia (0.1 expected, p < 0.01, as calcula-




ted by the GAG) and two cases of malignant cancer of the cervix (0.4  expected).




The cases among men were due  to tumors of the stomach, colon, and rectum.




     Hogstedt et al. (1979b)  suggested that the  combination of ethylene oxide




and methyl formate may  produce a special carcinogenic  risk, since methyl for-




mate,  the authors indicated,  exhibits its antibacterial effect by affecting DNA




structure.  No  literature reference was cited by  the authors as to  this point,




however.  A literature  search conducted for the  Carcinogen Assessment Group by




the Environmental Mutagen Information Center  at  the Oak Ridge National  Labora-




tory  (Francis,  1985) failed to  find  any literature  citations for  mutagenicity




studies  of methyl formate.



                                     9-126

-------
             TABLE 9-30.  OBSERVED AND EXPECTED NUMBER OF DECEASED
                   AMONG 153 WOMEN AND 50 MEN WITH CONTINUOUS
                   OR INTERMITTENT EXPOSURE TO ETHYLENE OXIDE3
ICDb
Causes
of death
Women
1968-77 1978-82
Obs Exp Obs Exp
Men
1968-77 1978-82
Obs Exp Obs Exp
Women + Men
1968-82
Obs Exp
1-     Total          2   2.9     3   2.7    4   2.6    2   2.2      11  10.4
999

140-   All tumors     2   1.2     3   1.1    0   0.6    1   0.5       6   3.4
209

200-   Lymphatic      1   0.1     2   0.08C  0   0.06   0   0.05      3   0.3C
207    and hemato-
       poietic tissue


aThe observed case of Waldenstrom's macroglobulinemia has been deleted from
 this table.  It had appeared in the table by Hogstedt et al. (1984).
^International Classification of Diseases, Eighth Revision.
cp < 0.01, calculated by the GAG.

SOURCE:  Adapted from Hogstedt et al. , 1984.
                                       9-127

-------
     9.5.2.5.  MORGAN ET AL. (1981) — Morgan et al.  (1981) conducted a retro-




spective study of 767 workers potentially exposed to ethylene oxide who had




worked for at least 5 years at a Texaco Chemical Company plant in Port Neches,




Texas, between January 1955 and December 31, 1977.   The authors provided no




analysis of the cohort with respect to length of follow-up.  An industrial sur-




vey of the plant (performed in July 1977) showed that the 8-hour time-weighted




average exposure to ethylene oxide was "well below" 50 ppm, except in the area




around the tank car loading operations, where readings were as high as 6,000




ppm.  Among the 767 male workers potentially exposed to ethylene oxide in the




study cohort, there were 11 deaths from malignant neoplasms, where 15.24 would




have been expected on the basis of U.S. vital statistics.




     There were more deaths than expected from pancreatic cancer (SMR* = 377,




3 observed versus 0.8 expected), bladder cancer (SMR = 322, 1 observed versus




0.31 expected), brain and central nervous system cancer (SMR = 285, 2 observed




versus 0.7 expected), and Hodgkin's disease (SMR = 570, 2 observed versus 0.35




expected).  Although the 95% lower confidence limits for these SMRs were all




less than 100, the number of deaths from pancreatic cancer and the number of




deaths from Hodgkin's disease are each significantly (p < 0.05) more than




expected by hypothesis testing using the Poisson test.  Excess mortality from




leukemia was not found.  Because their study cohort was small and because




excess cases of leukemia following exposure to ethylene oxide were found in  the




studies by Hogstedt et al.  (1979a, b), the authors calculated the magnitude  of




the relative risk of mortality from leukemia, given the sample size of  the




cohort, that could be detected at  the  95% confidence level with a power of 80%.




This  relative risk was calculated  to be  10.5 (SMR  of 1050).   In conclusion,  it
 *Standardized  mortality  ratio.



                                    9-128

-------
should be stated that the observed mortalities from pancreatic cancer and from




Hodgkin's disease were each significantly elevated among the study cohort, and




that the study cohort may have been too small for an adequate evaluation of the




risk of mortality from leukemia or other cancer types.  Furthermore, there was




no indication by the authors that sufficient allowance had been made for a




cancer latency period.









     9.5.2.6.  THIESS ET AL. (1982) — Thiess et al. (1982) conducted a cohort




mortality study of 602 persons who had been employed for 6 months or longer in




the alkylene oxide (ethylene oxide/propylene oxide) production or processing




areas of nine BASF Aktiengesellschaft, Ludwigshafen plants in West Germany




during the period from 1928 to 1980.  Vital status was ascertained for 523 of




the 536 German employees in the cohort, while that of only 30 of the 66 non-




German employees could be determined.  Thus, the percentage of overall follow-




up in this study was 92% (553 of 602).  In addition to alkylene oxides, the




workers were reported to have been exposed to a variety of other compounds.




     The expected mortality for the total cohort and for those within the co-




hort who were observed for a minimum of 10-years was calculated using mortality




data for Ludwigshafen, Rhinehessia-Palatinate, and the Federal Republic of




Germany.  The observed and expected numbers of cancer deaths for those persons




observed for at least 10 years are reported in Table 9-31.  The observed number




of deaths from cancer of any site was not significantly (p < 0.05) higher than




that expected based on mortality data for Ludwigshafen, Rhinehessia-Palatinate,




or the Federal Republic of Germany.  Deaths from cancer of the brain among




alkylene oxide workers who were followed for at least 10 years did approach




statistical significance (p < 0.07), however, in comparison with those expected




based on Ludwigshafen or Rhinehessia-Palatinate mortality data.




                                    9-129

-------
              TABLE 9-31.  COMPARISON OF OBSERVED NUMBERS OF CANCER DEATHS IN BASF-AKTIENGESELLSCHAFT, LUDWIGSHAFEN  PLANTS
                    1928-80 FOR PERSONS HAVING 10 YEARS OF OBSERVATION FOLLOWING EXPOSURE TO ALKYLENE OXIDE WITH THAT
                    EXPECTED  BASED ON MORTALITY STATISTICS FOR RHINEHESSIA-PALATINATE  1970-75, LUDWIGSHAFEN 1970-75,
                               AND THE FEDERAL REPUBLIC OF GERMANY 1971-74, BY ICD CODE AND CAUSE OF DEATH
UJ
o



ICD No.a
151
156
162
188
191
193-199

Total of
140-199c
205
230-239



Cause of death
Malignant tumor
of the stomach
Malignant tumor of
the gall bladder
Malignant tumor
of the bronchi!
Malignant tumor of
the urinary bladder
Malignant tumor
of the brain
Squamous cell
carcinoma of unknown
primary site
malignant tumors in ICD
Myeloid leukemia
Tumor of unknown
character

Rhinehessia-
Palatinate
1970-75
Observed
deaths No. P-value
2 1.852 0.552
1 0.201 0.182
4 3.769 0.520
1 0.469 0.374
1 0.071 0.068

1 0.743 0.525
!0 — d — d
1 0.148 0.138
1 0.454 0.365


Ludwigshaf en
1970-75

No.
1.765
0.243
3.956
0.532
0.066

1.047
__d
0.145
0.426

P-value
0.527
0.216
0.568
0.413
0.064

— c
__d
0.135
0.347
Federal
Republic of
Germany
1971-74

No. P-value
2.033 — b
	 C 	 C
	 C 	 C
	 C 	 C
	 C 	 C

	 C 	 C
11.816 — d
0.756 0.531
	 c 	 c
            International Classification  of Diseases, Eighth Revision.
            bThe probability of observed deaths  occurring  by chance was not  provided by the authors because the observed
             deaths were fewer than expected.
            cThe authors did not report  the number  of  deaths that would be expected in the cohort based on Federal Republic
             of  Germany mortality rates  for individual tumor sites other  than  stomach and myeloid leukemia.
            dThe authors did not report  the number  of  deaths from tumor sites,  ICD 140-199, that would be expected based on
             Rhinehessia-Palatinate or Ludwigshafen mortality data.

            SOURCE:  Adapted from Thiess  et al. ,  1982.

-------
      The  authors  also compared the  observed number  of  cancer deaths  with that




 expected,  using an internal cohort  of  1,662 styrene workers.   The  minimum obser-




 vation  period  of  10 years  required  for the  comparison  in Table 9-31  was  not




 used  for  this  analysis.  Thus,  in Table 9-32,  there were 14  total  observed




 cancer  deaths,  as opposed  to 12 observed deaths  in  Table 9-31.   These  results




 are reported in Table 9-32.   The relative risk of death  from cancer  of all




 sites in  the alkylene oxide cohort  in  comparison to what would be  expected




 based on  cancer mortality  in the styrene cohort  was 1.48.  Assuming  that  the




 numbers of  observed and  expected deaths (14  and  9.44,  respectively)  are  both




 Poisson variables,  the difference between the two is not statistically signi-




 ficant  (p < 0.05).   In the  65-  to 74-year-old age group,  the  relative  risk was




 2.78.   If it is assumed  that  both the  observed and  expected  deaths are Poisson




 variables,  the  difference  between the  two is statistically significant at  p <




 0.05.   It should  be  noted  that  although the  authors  reported  in  tabular form




 that 10 cancer  deaths had  occurred  in  the 65- to 74-year-old  age group, the




 text indicated  that  11 had  occurred—a  difference that obviously would function




 to lower the probability of  the  number  of cancer deaths  that  was observed.  A




 major problem in  evaluating  this result, however, is that the workers in the




 alkylene oxide  cohort were exposed  to a  variety  of  chemicals  in addition to




 ethylene oxide, some  of which are known  or suspected carcinogens.  The authors




 did not compare the  alkylene oxide and  styrene cohorts with regard to the




 number of deaths  by  individual  tumor site.




     The authors  also analyzed  the cancer deaths by length of exposure, and did




not find a dose-response.  However,  they gave no indication that the mortality




analysis by length of employment had been adjusted for length of follow-up.




     In summary,  this study is inconclusive  as to whether persons exposed to




ethylene oxide  are at an excess  risk of death from cancer.  There was a signi-





                                    9-131

-------
  TABLE 9-32.  RELATIVE RISKS OF DEATH FROM CANCER IN THE ALKYLENE OXIDE COHORT
                  AS COMPARED WITH THE STYRENE COHORT, BY AGEa
Observed
Age group deaths
15-24
25-34
35-44
45-54
55-64 4
65-74 10
75-84
Total 14
Expected
deaths Relative risk
—
0.35
0.47
1.61
3.41 1.17
3.60 2.78
—
9.44 1.48
aln this analysis, a minimum observation period of 10 years was not made a
 requirement.

SOURCE:  Adapted from Thiess et al.,  1982.
                                    9-132

-------
 ficant  excess  number  of  cancer  deaths  in the  65-  to  74-year-old  age  group  in




 the  alkylene oxide  cohort,  as compared to that  expected  based  on the mortality




 data for  a  group  of styrene workers.   A fact  that  may  have  confounded this  re-




 sult is that the  alkylene oxide workers were  exposed to  a variety  of chemicals




 in addition to  ethylene  oxide,  some of which  are known or suspected  carcino-




 gens.   Deaths  from  cancer of any  particular site were  not found  to be signifi-




 cantly  (p < 0.05) in  excess when  the expected numbers  of deaths  for  those  sites




 were derived using mortality data for  Ludwigshafen or  Rhinehessia-Palatinate.




 Two  problems with this study are  the small sample  size and  the fact  that only




 a little more  than half  of  the cohort  was observed for 10 years  or more.   In




 regard  to leukemia mortality, for which Hogstedt et  al.  (1979a,  1984)  had  found




 an association  with ethylene oxide exposure,  the authors found that  for those




 persons who had had more than 10  years  of exposure,  one  case of  myeloid leuke-




 mia  occurred where only about 0.15 would have been expected based  on local mor-




 tality data, but this difference  was not statistically significant (p  < 0.05).









      9.5.2.7.    SCHNORR (1982) —  A proportionate mortality study by  Schnorr




 (1982) of decedents who had  been  members of District 1199 of the National




Hospital and Health Care Workers  Union  found  that  the  proportionate  mortality




ratio (PMR) for neoplasms of lymphatic  and hematopoietic tissue  (1CD code




200-209, Eighth Revision),  as well as  for other types  of tumors,  was signifi-




cantly elevated for certain  job categories (e.g.,  "service" and  "nursing")




that included  job titles of personnel exposed to ethylene oxide  (e.g., hospi-




tal central service employees,  registered nurses,  licensed practical nurses,




and nurse's aides).   Such job categories were relatively broad in their inclu-




sion of job titles,  however, and  the results of the study with regard to a




possible association of cancer  risk with ethylene oxide exposure  must therefore




                                   9-133

-------
 be  judged  inconclusive.









     9.5.2.8.  STUDIES  IN PROGRESS — Several cohort or case-control studies




 testing  the association of ethylene oxide exposure and the risk of cancer are




 currently  in progress or about to begin.  A cohort mortality study of approxi-




 mately 1,200 workers who were engaged in ethylene oxide production during




 their work history in the chemical industry in the Kanawha Valley, West




 Virginia,  is currently  being conducted by the National Institute for Occupa-




 tional Safety and Health (NIOSH) and the Union Carbide Corporation.  Nested




 case-control studies within the cohort will be done for certain kinds of




 deaths (e.g., leukemia).  The results of the cohort study will not be avail-




 able until early 1986.  The results of the case-control study will not be




 available until some time later (Rinsky, personal communication).




     NIOSH and the Health Industry Manufacturing Association are currently




 conducting a cohort mortality study of approximately 10,000 persons, consist-




 ing primarily of medical equipment manufacturing personnel who use ethylene




 oxide as a sterilant.   Some exposure information on this cohort is available,




 but only for 1978 onward.   The results will not be available until at least




 1987, and a published report is expected about a year later.




     The U.S. Environmental Protection Agency funded a case-control study of




 70 cases of cancer of the lymphatic and hematopoietic tissue and 140 controls




 in District 1199 of the National Hospital and Health Care Workers Union to




 determine if an association existed between such cancers and occupational




 exposure to ethylene oxide and/or other substances.   The study, conducted by




Dr. Jeanne Stellman of Columbia University,  has been completed but has not




yet been published as of the date of this writing.  No association between




 the cases and exposure  to ethylene oxide was reported to be found; however,



                                   9-134

-------
some problems with  regard  to  ascertainment  of exposure  did  exist  (Schnorr,




personal communication).










     9.5.2.9.  SUMMARY OF  EPIDEMIOLOGIC  STUDIES — Three epidemiologic studies




of persons occupationally  exposed  to ethylene oxide demonstrated  a significant




association between ethylene  oxide exposure and cancer  incidence  or mortality.




A study by Hogstedt et al.  (1979a, 1984) found significantly  (p < 0.05) in-




creased mortality for stomach cancer and leukemia and significantly (p < 0.01)




increased incidences of cancer of all sites and of leukemia among ethylene




oxide production workers.   Hogstedt et al.  (1979b, 1984) found significantly




(p < 0.05) increased incidences of leukemia and cancer  of all sites and sig-




nificantly (p < 0.01) increased mortality from leukemia among workers exposed




to ethylene oxide used as  a sterilant.  The study by Morgan et al. (1981)




found increased mortality  from pancreatic cancer and Hodgkin's disease that




is statistically significant  (p < 0.05) by hypothesis testing.




     Excess mortality from leukemia in the Hogstedt et  al.  (1979a, 1984)




study and excess incidences of leukemia in the Hogstedt et al. (1979b) study




were not limited to any particular types of leukemia.  Excess deaths from




leukemia in the Hogstedt et al. (1979a) study included one case of acute




myeloid leukemia and two cases of chronic lymphatic leukemia.  Excess cases




of leukemia in the Hogstedt et al. (1979b) study included one case of acute




myeloid leukemia, one case of chronic myeloid leukemia,  and one case of




"blast" leukemia.  The expected numbers of deaths or cases by type of leukemia




were not calculated in either study.





     It should be noted that  in all three of the above-referenced epidemio-




logic studies, exposure of the cohort to other chemicals besides ethylene




oxide was reported to have occurred or probably occurred.   In the Hogstedt et





                                    9-135

-------
 al.  (1979a)  study,  reports  were  made  of  exposure  to  several  chemicals,  of




 which  two, ethylene dichloride and  bis(2-chloroethyl)ether,  are  recognized




 carcinogens.   In the Hogstedt et al.  (1979b)  study,  ethylene oxide-exposed




 workers  experienced concurrent exposure  to methyl formate.   In the Morgan et




 al.  (1981) study,  there was no mention of exposure to chemicals  other than




 ethylene oxide,  but the fact that the study was conducted at a chemical plant




 would  suggest  that  exposure to other  chemicals did occur.









 9.5.3.   Quantitative  Estimation.  This quantitative  section  deals with the




 incremental unit  risk for ethylene  oxide in air and  the potency  of ethylene




 oxide  relative to other carcinogens that the GAG has evaluated.  The incre-




 mental unit risk estimate for an air pollutant is defined as the increased




 life-time cancer risk occurring  in a hypothetical population in which all in-




 dividuals are exposed continuously from birth throughout their lifetimes to a




 concentration of 1y g/m^ of  the agent in the air they breathe.   These cal-




 culations are done  to estimate in quantitative terms the impact of the agent




 as a carcinogen.  Incremental unit risk estimates are used for two purposes:




 1) to compare the carcinogenic potencies of several agents with each other,  and




 2) to give a crude  indication of the population risk that would be associated




with air or water exposure  to these agents,  if the actual exposures were known.




Hereinafter,  the term "unit  risk" will always refer to incremental unit risk.




     In the sections  that follow, the general assessment procedures used by




 the GAG are discussed.  These include animal-to-human extrapolation modeling,




 data selection, calculation  of human equivalent doses, extrapolation modeling




 from human epidemiologic studies, and interpretation of  the resulting esti-




mates.   Following this discussion, the CAG's unit risk calculations and rela-




 tive potency estimates are  presented.



                                     9-136

-------
     9.5.3.1.  PROCEDURES  FOR THE DETERMINATION OF UNIT  RISK FROM ANIMAL DATA




— In developing quantitative estimates of carcinogenic  risks, one  or  both of




two types of data are utilized:   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  con-




trary, that if a carcinogenic response occurs at the dose levels used  in the




study, then responses will also occur at all lower doses, at incidences




determined by an extrapolation model.




     There is, however, no solid scientific basis for any mathematical  extra-




polation model that relates carcinogen exposure to cancer risks at  the  ex-




tremely low concentrations that must be dealt with in evaluating environ-




mental 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 which risk model to use.  At the present




time, the dominant view of the carcinogenic process involves the concept that




most cancer-causing agents 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 that the quantal




type of biological response, which is characteristic of mutagenesis, is




associated with a linear nonthreshold dose-reponse 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 nonthreshold dose-response relationship is




                                     9-137

-------
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 aflatoxins in the diet).  There is also some evidence from animal experi-




ments that is consistent with the linear nonthreshold model (e.g., liver




tumors induced in mice by 2-acetylaminofluorene in the large-scale EDgi study




at the National Center for lexicological Research, and the initiation stage




of the two-stage carcinogenesis model in rat liver and mouse skin).




     Based on the above evidence of low-dose linearity, and because very few




compounds exhibit low-dose responses that are superlinear, the linear non-




threshold model has been adopted as the primary basis for risk extrapolation




in the low-dose region 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 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 nonthreshold




dose-response relationship at low doses is the linearized multistage model.




The multistage model employs enough arbitrary constants to be able to fit al-




most any monotonically increasing dose—response data, and it incorporates a




procedure for estimating the largest possible linear slope (in the 95% confi-




dence limit sense) at low extrapolated doses that is consistent with the data




at all dose levels of the experiment.









     9.5.3.1.1.  Description of the Low-Dose Animal Extrapolation Model — Let




P(d) represent the lifetime risk (probability) of cancer at dose d.  The multi-




stage model has the form



                                    9-138

-------
                P(d) =  1 - exp  [-(q0 + qLd + q2d2 +  ... +  qkdk)]
where
                          qt  >_ 0, i = 0,  1, 2,  ..., k





Equivalently,
                  Pt(d) =  1 - exp  [(qLd + q2d2
where
                              Pt(d) = P(d) - P(0)

                                        1 - P(0)
is the extra risk over background rate at dose d.



     The point estimate of the coefficients q^, i = 0,  1,2,  ..., k, and



consequently, the extra risk function, P(-(d), at any given dose d,  is calcu-



lated by maximizing the likelihood function of the data.



     The point estimate and the 95% upper confidence limit of the extra risk,



Pt(d), are calculated by using the computer program GLOBALB3, developed by



Howe (1983).  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, q*, on parameter q^.  Whenever q^ > 0, at



low doses the extra risk Pj-(d) has approximately the form Pj-(d) = qj x d.



Therefore, q^ x d is a 95% upper confidence limit on the extra risk, and



R/q^ 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, q|, is calculated by increasing q^ to a value q^ such that when the



log-Likelihood is remaximized subject to this fixed value qi for the linear



coefficient, the resulting maximum value of the log-likelihood LI satisfies



                                    9-139

-------
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,




Pt(d), is an improvement on the Crump et al. (1977) model.  The upper confi-




dence limit for the extra risk calculated at low doses is always linear.  This




is conceptually consistent with the linear nonthreshold concept discussed




earlier.  The slope, qn, is taken as an upper bound of the potency of the




chemical in inducing cancer at low doses.  (In the section calculating the




risk estimates, Pfc(d) will be abbreviated as P.)




     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 experi-




ment, including the control group.




     Whenever the multistage model does not fit the data sufficiently well,




data at the highest dose are deleted and the model is refit 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
                     x   -

is calculated where N. is the number of animals in the i   dose group, X. is




the number of animals in the i   dose group with a tumor response, P^ is the




probability of a response in the i^ dose group estimated by fitting the




multistage model to the data, and h is the number of remaining groups.  The



                                   9-1^0

-------
 fit  is  determined to be unacceptable whenever ^ ^  is larger than the cumu-
 lative  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 multi-
 stage coefficients.


     9.5.3.1.2.   Selection of  Data  — For some  chemicals,  a number of studies
 in different  animal  species,  strains,  and sexes, each run at  varying doses
 and  routes  of exposure,  are available.  In such cases,  choices  must be made
 as to which of  several  data sets are appropriate for use  with the chosen
 model.   The following are the  procedures  used by the GAG  in evaluating these
 data for the  purpose of  risk estimation:
     1.   The  data  on tumor incidence are  separated according  to organ sites
 or tumor types.  The dose and  tumor  incidence data set  used in  the model is
 the  set  in  which  tumor  incidence is  statistically  significantly higher than
 in controls for at least  one test dose level, and/or where  the  tumor  inci-
 dence rate  shows a statistically significant  trend with respect to dose level.
 The data  set  that  gives  the highest  estimate  of the  lifetime  carcinogenic
 risk, q^, is  selected in  most  cases.  However,  efforts  are  made to exclude
 data sets that 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  data  set
having the  larger sample  size  is selected  for calculating carcinogenic
potency.
     2.    If there are two  or more data sets of  comparable size  that are
identical with respect to  species,  strain, sex,  and  tumor sites,  the geo-
metric mean of q^, estimated from each of  these data sets,  is used  for risk
assessment.
                                   9-1U1

-------
     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.






     9.5.3.1.3.  Calculation of Human Equivalent Dosages from Animal Data — In



calculating human equivalent dosages, it is necessary to correct for differen-



ces in metabolism among species and for the variations in absorption factors



involved in different routes of administration.



     Following the suggestion of Mantel and Schneiderman (1975), it is



assumed that mg/surface area/day is an equivalent dose between species.



Since, to a close approximation, the surface area is proportional to the



two-thirds power of the weight, as would be the case for a perfect sphere,



the exposure in mg/day per two-thirds 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



The lifetime average exposure is then
                                 d -
                                     L  x W2/3
                                      e
When exposure is given in terms of mg/kg/day = m/Wr = s, the conversion is



                                     9-142

-------
simply






                                 m   = s x W1/3
                                2/3
                              rw
where r is the absorption rate for ethylene oxide (assumed to be 1).




     When exposure is via inhalation, as with ethylene oxide, dose calcula-




tions at experimental exposures of up to 100 ppm are performed under the




assumption that the compound is a completely water-soluble gas absorbed




proportionally to the amount of air breathed in.  While the GAG has previously




used an existing methodology to determine dose equivalency in such cases, for




ethylene oxide the total body dose resulting from exposure of male Fischer




344 rats to air concentrations of 100 ppm for 6 hours has been measured as




20.24 mg/kg (Tyler and McKelvey, 1980).  At 10 ppm exposures under similar




conditions, the measured dose was 2.7 mg/kg.  Since daily exposures in the




Snellings et al. (1981) study included 10 ppm and 100 ppm, the human equiva-




lent dosage for the above exposure is estimated as
           dh = 20.24 x 5/7*  (70/0.42)1/3 = 2.63 mg/kg/day for 100
   ppm
and
           d,  = 2.7 x 5/7*  (70/0.42)1/3 = 0.35 mg/kg/day for 10
ppm
where 0.42 kg is the average weight of the male rat in the Snellings et al.




(1981) study, 70 kg is the average weight of the adult human, and 5/7 is the




fraction of days exposed.  By interpolation, the 33 ppm exposure is estimated




as 0.94 mg/kg/day in human equivalent doses.







                                   9-TI3

-------
     9.5.3.1.4.  Calculation of the Unit Risk from Animal Studies — The risk




associated with d mg/kg2/3/day is obtained from GLOBAL83, and for most cases




of interest to risk assessment, the 95% upper-limit risk 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, it




is simply necessary to find the number of mg/kg^'-^/day that corresponds to




one unit of X, and substitute this number into the above relationship.  For




ethylene oxide, human equivalent doses will first be calculated and then




fitted, together with the observed responses, to the linearized multistage




model.  An equivalent method of calculating unit risk would be to use mg/kg/




day for the animal exposures and then to increase the j*-*1 polynomial coeffi-




cient by an amount
                     (Wh/Wa)J/3         j = 1, 2, ..., k
and use the mg/kg/day equivalents for the unit risk values.  In the section




of this document that presents unit risk calculations from animal data, the




final qi will always represent the upper-limit potency estimate for humans.
     9.5.3.1.5.  Interpretation of Quantitative Estimates — Unit risk esti-




mates based on animal bioassays are only approximate indications of absolute




risk in populations exposed to known carcinogen concentrations.  This is true




for several reasons.  First, there are important species differences in up-




take, metabolism, and organ distribution of carcinogens, as well as in target




site susceptibility, immunological responses, hormone function, dietary




factors, and disease.  Second, the concept of equivalent doses for humans as




compared to animals based on the relationship of weight to surface area is




virtually without experimental verification as regards carcinogenic response.





                                   9-1

-------
Finally,  human  populations  are  variable  with  respect  to  genetic  constitution




and diet,  living  environment, activity patterns,  and  other  cultural  factors.




     Unit  risk  estimates  can  give  rough  indications of the  relative  potencies




of given  agents as  compared with other carcinogens.   Such comparisons  are,  of




course, most  reliable when  based on  studies in which  the test  species,




strain, sex,  and  route of exposure are the same.




     The  quantitative aspects of assessing carcinogenic  risks  are  discussed




here because  of the possible usefulness  of this information  in the regulatory




decision-making process, e.g.,  in  setting regulatory  priorities, evaluating




the adequacy  of technology-based controls, etc.   However, the  uncertainty of




present estimations of cancer risks  to humans at  low  levels  of exposure




should be  recognized.  The  GAG  feels that, given  the  limited data available




from animal bioassays, especially at the high dosage  levels  required for




testing, almost nothing can be  known about the true shape of the dose-response




curve at low  environmental  levels.  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 is appreciably higher than




the estimated risk, but it  could very well be considerably lower.  The risk




estimates  presented in this document should not,  therefore, be regarded as




accurate representations of the true cancer risks even when the exposures are




accurately defined.  These estimates may, however, be factored into regula-




tory decisions to the extent that the concept of upper risk limits is found




to be  useful.









     9.5.3.1.6.   Alternative Methodological Approaches — The methods used  by




the GAG for quantitative assessment are consistently conservative in that they




tend to result in high estimates of risk.  This conservatism is primarily due





                                     9-745

-------
to the CAG's use of the linear nonthreshold extrapolation model in prefer-




ence to any one of a variety of other extrapolation models that would give




lower risk estimates.  For purposes of comparison, descriptions of these




alternative models (the one-hit, the probit, and the Weibull models) are




presented in Appendix 9A.




     Another method of risk estimation employed by the GAG involves the use




of animal bioassay data as the basis for extrapolation.  At present, the




CAG's approach is to utilize data corresponding to the most sensitive animal




responses in these studies.  An alternative approach would be to use the




average responses of all adequately tested bioassay animals.




     Extrapolations from animals to humans can also be made on the basis of




either relative weight or surface area.  The latter approach, which is used




by the GAG, has more of a basis in human pharmacological responses; however,




at the present time there is some question as to which of the two approaches




is more appropriate for use with carcinogens.  Given this uncertainty, the




GAG has chosen the most generally employed method, which is also the more




conservative of the two.  In the case of ethylene oxide inhalation studies,




the use of extrapolation based on surface area rather than weight increases




the unit risk estimates by a factor of 5.5 for the males and 6.8 for the




females.









     9.5.3.2.  HUMANS—MODEL FOR ESTIMATION OF UNIT RISK BASED ON HUMAN DATA




— Whenever possible, the GAG utilizes data from human epidemiologic studies




in preference to animal bioassay data.  If sufficiently valid exposure infor-




mation is available for a given compound, this information  is always used by




the GAG in its assessment.  If  the  results of such  studies  show carcinogenic




effects, the data are analyzed  to give estimates  of the linear dependence of




                                   9-146

-------
 cancer  rates  on  lifetime  average  doses  (equivalent  to the  factor  By  in the




 equation  below).   If  human epidemiologic  studies  show no carcinogenic  effects




 when  positive animal  evidence  is  available,  then  it  is assumed  that  a  risk




 does  exist, but  that  the  risk  is  smaller  than  could  have been observed in an




 epidemiologic study.   In  such  cases  it  is  assumed that the true incidence is




 just  below  the level  of detection in the  cohort studied, and calculations are




 then  made to  estimate an  upper limit of cancer incidence,  as determined




 largely by  the size of the cohort.




      Very little  information exists  that  can support extrapolation from high-




 exposure  occupational studies  to  situations  in which contamination is  at low




 environmental levels.  However, if a number  of simplifying assumptions are




 made, it  is possible  to construct a  crude  dose-response model whose  para-




 meters  can  be estimated using  vital  statistics, epidemiologic studies,  and




 estimates of  worker exposures.




      In human studies, responses  are measured  in  terms of  the relative  risk




 of an exposed cohort  as compared  to  a control  group.   The  mathematical  model




 employed by the GAG assumes that  for low exposures the lifetime probability




 of death from lung cancer  (or  any cancer), PQ, may be  represented by the




 linear equation







                                  P0  = A + bHX







where A is the lifetime probability  of death from cancer in the absence of




 the agent, and X  is the average lifetime exposure to environmental levels in




units such as  ppm.  The factor, by,  is the increased probability of cancer




associated with each unit  increase of the agent in air.




     If it is  assumed that R,   the relative risk of lung cancer  for exposed




workers as compared to the general population,  is independent of the length




                                   9-147

-------
or age  of  exposure and depends  only  on average  lifetime  exposure,  it  follows



that
                            R = P  = A + bH
                                P0   A + bH
or
                             RPO = A + bH  (K! + x2)






where Xj = lifetime average daily exposure to the agent for the general




population, X2 = lifetime average daily exposure to the agent in the occupa-




tional setting, and P0 = lifetime probability of dying of cancer with no or




negligible ethylene oxide exposure.  Substituting P0 = A + b^ X^ and rearrang-




ing gives






                               t^ = P0 (R - 1)/X2






     To use the above model, estimates of R and X2 must be obtained from ap-




propriate epidemiologic studies.  The value of PQ is derived by means of life-




table methodology from 1976 U.S. vital statistics records of age- and cause-




specific death rates for males.  For leukemia, the estimate of PQ is 0.0091.




This methodology is utilized by the GAG in the present document, in the sec-




tion on unit risk based on human data.









     9.5.3.3.   UNIT RISK ESTIMATES FOR ETHYLENE OXIDE









     9.5.3.3.1.  Unit Risk Estimate Based On Animal Studies — The two long-




term animal inhalation studies presented in the qualitative carcinogenicity




section of this document showed similar results,  both qualitatively and




                                   9-148

-------
 quantitatively,  for  the  males.   Both  studies  had  significantly  increased




 dose-related  incidences  of  peritoneal mesothelioraas  and  gliomas,  and  some




 increase  in mononuclear  cell  leukemias.   These  studies will  be  analyzed




 separately and  then  compared.









      9.5.3.3.1.1.  Snellings  et  al. (1981)  (Bushy Run).   This study exposed




 120 Fischer 344  rats  of  each  sex to three different  doses  (100  ppm, 33 ppm,




 and 10 ppm) of  ethylene  oxide vapor via  inhalation for 6  hours/day, 5 days/




 week, for approximately  2 years.  Comparable  untreated (air) control  groups




 were  also used.   Interim sacrifices were  conducted to evaluate  the time




 development of  treatment-related effects.




      The results  of  the  study show statistically significant increases in




 brain gliomas (highest dose group) and in mononuclear cell leukemias  in




 females in the  two highest  dose  groups, and peritoneal mesotheliomas and




 brain gliomas in  males in the two highest dose groups.  The above tumors all




 exhibited dose-response  trends.   Table 9-33 summarizes the pertinent data




 from  this study which the GAG has used in calculating potency estimates for




 ethylene oxide.    In connection with these data, it should be noted that the




 brain gliomas were not examined  histopathologically until after the results




 of the NIOSH study (Lynch et al., 1982) had alerted the Bushy Run researchers




 to the possibility of the occurrence of brain neoplasias.  For this reason,




 only  18-month, 24-month,  and dead/euthanized moribund denominator figures




were available for gliomas.   For the male peritoneal mesotheliomas and the




 female mononuclear cell  leukemias, the denominators in Table 9-33 correspond




 to the number of animals  alive when the first tumor of that type was  found.




In the males,  the first peritoneal mesothelioma was found at 15  months;  in




the females,  the first mononuclear cell leukemia was  found at 18 months  (see




                                   9-149

-------
                                   TABLE  9-33.  BUSHY RUN ETHYLENE  OXIDE  INHALATION  STUDY  IN FISCHER 344 RATS.
                              INCIDENCE OF  PERITONEAL MESOTHELIOMA  AND  BRAIN  GLIOMA  IN  MALES,  AND MONONUCLEAR CELL
                                  LEUKEMIA  AND  BRAIN GLIOMA3  IN FEMALES BY  DOSE  AMONG SURVIVORS  TO FIRST TUMOR.
                                           MAXIMUM LIKELIHOOD ESTIMATES  OF LINEAR TERM AND  95%  UPPER-LIMIT q*
Ul
O
Group
Males
Peritoneal meso./No. examined (X)c
p-values^
Brain glioraas/No. examined (%)e
p-values
Total
p-values
Human equivalent dose (mg/kg/day)'
Females
Mon. leukem/No. examined (%)S
p-values
Brain gliomas/No. examined (%)e
p-values
Total
p-values
Human equivalent dose (mg/kg/day )f

0 (combined)
4/187(2)
<0. 00001
1/196(0.5)
=0.0003
5/187(3)
<0. 00001
0
22/186(12)
<0. 00001
1/194(0.5)
=0.014
23/186(12)
<0. 00001
0
Exposure in air (ppm)
10 33
3/88(3) 7/82(8)
=0.02
1/99(1) 5/98(5)
=0.02
4/88(5) 12/82(15)
=0.0005
0.35 0.94
14/71(20) 24/72(33)
=0.08 0.0001
1/95(1) 3/99(3)
15/71(21) 27/72(38)
<0.0001
0.28 0.75
Linear term
estimates
q* b
MLE 1
100 qj (mg/kg/day)"1
22/96(22) 5.1xlO-2
<0.0001
7/99(7) 3.1x10-2
=0.002
29/96(30) l.lxlQ-1
<0.0001
2.63
28/73(38) 2.0x10-!
<0.0001
4/99(4) 2.0x10-2
=0.05
32/73(44) 2.5x10-!
<0.0001
2.11
1.1x10-1
5.0x10-2
1.7x10-1

2.9x10-!
4.0x10-2
3.5x10-1

                    Table 9-27.
               b95% upper-limit unit risk estimate.
               cNumber alive at 15 months.
               dFisher Exact Test vs.  combined controls  (one  tailed).   P-value under controls  is a one-sided Cochran-Armitage
                test for a dose-response trend.
               eTotal number examined less  6- and  12-month sacrifices.
               fBased on measured doses in  males of  20.24 and 2.7  mg/kg b.w.  following 6 hours'  exposure to ethylene oxide at 100 ppm
                and 10 pnm, respectively.  The animal-to-human dose equivalences  are based on  a  dose per surface area factor of
                (70/W ) '   , which increases unit risk estimates by factors of 5.5 for the males  and 6.8 for the females over
                dose per body weight equivalences.
               SNumber alive at 18 months.
               SOURCE:  Adapted from Snellings et  al.,  1981.

-------
also Table 9-23).




     As reported earlier, a dose of 20.24 mg/kg of body weight has  been




measured for male Fischer 344 rats exposed to ethylene oxide at  100 ppm under




conditions similar to  those of  the Snellings et al.  (1981) study.   For this




document, dose is assumed to be equivalent between species on the basis of




mg/surface area, or mg/body weights/3.  This means that a dose of 2.63 mg/kg




body weight given to a 70 kg human is assumed to produce an equivalent res-




ponse to that produced by 20.24 mg/kg in the male rat.  As discussed above




and as shown in Table 9-33, this method of determining dose equivalence




increases the unit risk estimates by factors of 5.5  for females and 6.8 for




males over estimates obtained on the basis of mg/kg  of body weight.




     Table 9-33, in presenting the total number of significant tumors by sex,




sums the total number of significant tumors over the smallest denominator.




This is done because time-to-tumor data on the gliomas are unavailable.  Com-




pared with the usual GAG procedure of counting the total number of animals




with significant tumors, the addition of total significant tumors, as is done




here,  increases the risk estimate very slightly.




     Calculations of the 95% upper-limit unit risk estimate,  based on the




linearized multistage model fitted to the data in Table 9-33, yield a high




value of q^ = 3.5 x 10  (mg/kg/day)  , based on total mononuclear cell leu-




kemias and brain gliomas in the female rats.   The responses of the males,




based on total peritoneal mesotheliomas and brain gliomas,  yield a value of




50% less,  qh = 1.7 x 10" (mg/kg/day)"1.  The higher estimate  is chosen for




safety purposes.




     To convert the above estimate to units of y g/nH for humans, the follow-




ing formula is used:




                                   9-151

-------
 1 mg/kg/day = 1 mg/kg/day x 70 kg x 1000p g/mg x day/20 m3 = 3.5 x 103y g/m3




or




                         ly g/m3 = 2.86 x 10~4 mg/kg/day.




The 95% upper-limit slope estimate in terms of y g/m3 is thus calculated as




q* = 3.5 x lO-iCmg/kg/day)'1 x 2.86 x 10~4(mg/kg/day) = 1.0 x 10~4(u g/m3)-1
 h                                           —  o—

                                            y g/m



     To convert from y g/m3 to ppm, the formula is




            1 ppm =  l'2 g   x 44>1 m.w. ethylene oxide x 10 y g x 1Q-6

                     10~3 m3        28'2 m
-------
study did not show a significant increase in these leukemias, only peritoneal

mesotheliomas and brain gliomas were used for risk assessment.  The results

of the potency calculations, shown in Table 9-34, are quantitatively nearly

identical to those in Table 9-33.  Based on the above analyses, the maximum

animal 95% upper-limit slope potency value is still q^ = 3.5 x 10~ (rag/kg/
       based on the total mononuclear cell leukemias and brain gliomas in

female rats in the Snellings et al. study.



     9.5.3.3.1.3.  Effects of Results on Different Dose Equivalence Assumptions

— OSHA Versus EPA Assessments.  The results of the above assessments depend,

to some extent, on the dose equivalence assumptions.  Dose equivalence in the

following discussion means the dose that will cause an equivalent response,

quantitatively, in both species.  The GAG has assumed that doses are equiva-

lent on the basis of mg per surface area, an assumption for which there is

some experimental evidence when first-order kinetics apply; for ethylene

oxide, first-order kinetics appear to apply at exposures up to 100 ppm (Tyler

and McKelvey, 1980).  As explained in an earlier section,  use of the surface

area correction increases the 95% upper-limit unit risk estimate by factors

of 5.5 for the males and 6.8 for the females over estimates obtained on the

basis of mg/kg/body weight.*  OSHA, which assumes equivalence on a mg/kg/

body weight basis, calculated exposures of 19.30 mg/kg/day for males and

23.94 mg/kg/day for females exposed to ethylene oxide at 100 ppm in the

Snellings et al. (1981) study, using EPA methodology (Federal Register 48[78]:
*Equivalence could also have been calculated directly on a ppm basis; this
 would have yielded a 95% upper-limit estimate approximately 1.8 times as
 high as that obtained on the basis of mg/kg/body weight.  EPA uses direct ppm
 equivalence for partially soluble gases and particulates.  Ethylene oxide
 can be considered a completely soluble gas.

                                      9-153

-------
  TABLE 9-34.  NIOSH ETHYLENE OXIDE INHALATION STUDY IN MALE FISCHER 344 RATS.
         INCIDENCE OF PERITONEAL MESOTHELIOMA AND BRAIN GLIOMAa BY DOSE,
                             AMONG TOTAL EXAMINED.
  ESTIMATES OF 95% UPPER-LIMIT RISK BASED ON HUMAN EQUIVALENT DOSE (mg/kg/day)
                                      Exposure in air (ppm)
                                                                        1
                                             50          100       (mg/kg/day)-1
Peritoneal mesothelioma/No.
examined
Brain glioma/No. examined
Total
Human equivalent dose
(mg/kg/day)f
3/78b
0/76C
3/78b
0
9/79
2/77
ll/79b
1.59
21/79b
5/79d
26/79b
3.06
l.OxlO-1
3.4xlO-2
1.3X10-1
—
aSee Table 9-29.
bp < 0.001.
cp < 0.01.
dp < 0.05.
ep-values noted beside control incidences represent values associated with a
 one-sided Cochran-Armitage test for a dose-response trend.
fHuman equivalent dose based on transforming ppm to mg/kg/day as in Table 9-33,
 except for an adjustment for 7 hours' exposure.

SOURCE:  Lynch et al., 1982.
                                  9-154

-------
 172-193)  for a completely soluble gas.   While the results for the males,




 19.30 mg/kg/day,  are within 5% of the dose measured by Tyler and McKelvey




 (1980) (20.24 mg/kg/day), EPA used the  more accurate measured dose in this




 case.   EPA then used the surface area correction factors  for animal-to-man




 equivalence.   Thus,  on the basis of  the difference in assumptions of  equiva-




 lent  dose alone,  the EPA risk numbers are  larger than OSHA's by  a factor  of




 about  6.




      One  other difference between the OSHA and EPA assessments,  both  based on




 the Bushy Run data,  is that the  EPA  added  total significant  tumors (mononuc-




 lear  cell leukemias  and brain gliomas for  the females), while OSHA used the




 total  number  of malignant tumor-bearing animals.   For EPA, this  led to factors




 higher by 50% for  the males and  20%  for the females.   The  result,  based on




 animal data,  is that  the EPA 95% upper-limit  unit  risk factor is  larger than




 that  of OSHA  by a  factor of about 8.









     9.5.3.3.2.  Comparison of Animal and  Human Inhalation Studies  — The pur-




 pose of this  section  is  to  determine  whether  or not  the extrapolated  risks




 from the  animal data  can reasonably  predict  the observed human results.  As




 presented in  Sections  9.5.1  and  9.5.2,  there  is  strong evidence for the carci-




 nogenicity of  ethylene  oxide  in  rats, while  the evidence in  humans  is limited,




 but suggestive of a leukemia  effect.  Table 9-35 summarizes  the leukemia evi-




 dence  in  humans.  As  can be  seen,  two of the  four  mortality  studies reviewed




had statistically significant increases in  the  relative risks of  leukemia.




None of the four studies was  particularly  revealing, however, since both the




 observed  and expected numbers were quite small.  Because the expected numbers




were so small, the 95% confidence  limits around  the relative risk are quite




large.  Even with statistically significant increased cancers, estimates of



                                   9-155

-------
            TABLE 9-35.   LEUKEMIA (ICD  204-207)  INCIDENCE AND MORTALITY:   ETHYLENE OXIDE EPIDEMIOLOGY.
                  INCLUDED ARE RELATIVE RISKS,  95% CONFIDENCE LIMITS,  NOMINAL EXPOSURE ESTIMATES,
                                     AND 95% CONFIDENCE LIMITS ON UNIT RISK


Study
Hogstedt (sterllant plant)
(1968-1977)
Storage hall
Adjacent area


Total (original, 1968-77)
(New data, 1968-1982)


N


(70)
(160)


(230)
(203)t>


Obs.


2
0


2
4


Exp.


0.03
0.07


0.10
0.3
Rela-
tive
risk.


66.7
0


20
13.3
(95%
confidence
limits)


8-240
0-52


2.4-72
3.6-34.2

Exposure
(ppm)


20 + 10 (8 hr TWA)
"exposed occasionally
on passing through"



95% confidence
limits on unit
risk3 (ppm~* )


0.07 - 2.5
No other
estimates
possible


    ICD 200-207

Hogstedt (prod, plant)
   a.  (1961-1982)
       >_ 10 yr latency
       >^  1 yr exposure
Direct exposure (89) 2 0.18 10.0
Intermittent exposure (79) 1 0.16 6.2
b. > 10 yr exposure
7 20 yr Induction-
latency (1961-1977)
Direct exposure 1 0.04 25
Intermittent exposure 1 0.1 10
Morgan et al.
(1955-1977) (767) 0 0.7 0
Thiess et al. (351) 1 0.15 6.8
(1928-1980)
> 10 yr exposure
1.3-40.1
0.1-42.8

0.3-139
0.1-55.6
0-5.2
0.1-37

< 14
"less exposure"



< 10
mostly < 5 ppm
(a few large variations)

aSee text.  Estimates  for most groups not possible, due to lack of information about exposure duration.
blncludes those employed at least one year.  Hogstedt et al. (1984) subcohorts constructed differently than
 earlier Hogstedt et al. (1979b) subcohorts.

-------
 hazard due  to  ethylene  oxide  are  difficult  because  of  the  small  sample  size.




      An even greater  problem  associated  with  determining potency estimation




 from human  studies  is the  general lack of exposure  information.   Table  9-35




 presents  some  ethylene  oxide  measurements based  on  either  8-hour time-weighted




 averages  or spot  measurements,  but  other important  data are  generally missing.




 Such necessary  information as average length  of  exposure,  average age at




 exposure, and  average length  of follow-up cannot even  be estimated  from three




 of  the four studies in  the literature.   Only  the subcohort of storage hall




 workers  in  the  sterilant plant  (Hogstedt et al.,  1979b, 1984) is consistent




 enough in terms of exposure conditions,  duration, and  follow-up  to  estimate




 95%  confidence  limits of ethylene oxide  potency.  This is  presented in  the




 following paragraphs.   The simplifying assumptions  add to  the uncertainty of




 the  estimate.




      The risk assessment done on  the basis of the Hogstedt et al. (1979b)




 study  probably underestimates the carcinogenic potency of  ethylene  oxide




 because of  two factors:   1) In this study, exposure started  in 1968 and ended




 in  1977—giving a maximum  latency period of only 9 years, whereas cancer




 usually involves a relatively long  latency period.  (However, the author




 states that leukemia  incidence in Hiroshima and Nagasaki due to  the atomic




 bomb  irradiation showed a  rapid increase that began shortly  after exposure




 and  reached a peak after 6  years.)  2) Since the study did not report the




 number of person-years  of  exposure, it is assumed for present purposes that




 all  of the  230 workers were exposed for  the full 9 years,  an assumption which




 tends  to underestimate  the  risk.  Another problem with this  study is that the




 gas used for sterilization was 50% ethylene oxide and 50% methyl formate.




Little is known about the  biological effects of methyl formate or of the




combination of  methyl formate with ethylene oxide.  However,  methyl formate





                                9-157

-------
is known to metabolize to formic acid, which is a normal body metabolite.  It




is assumed for present purposes that ethylene oxide was the only leukeraogen




in this study, although one of the cases (the man) had reported some contact




with benzene in laboratory work.




     Hogstedt et al. (1979b) states, in connection with exposures in the




factory studied, that infrared spectrophotometry and gas chromatography




measurements in 1977 showed values ranging from 2 to 70 ppm in the factory's




storage hall area.  The study also reports that the calculated 8-hour time-




weighted average ethylene oxide concentration in the breathing zone was




20 + 10 ppm, and that the concentration in the storage hall was higher than




in the sterilization room.  The accompanying table described the 70 storage




hall employees as having had 8-hour exposures, while all but seven of the




remaining employees were described as "occasionally exposed."




     Of the two leukemia cases (acute myeloid and chronic myeloid), both




people worked in the storage hall area, and neither had reported exposure to




benzene.  Because the two cases worked in the storage hall, the GAG has




chosen to estimate the expected number of leukemia cases for the persons who




worked only in that area rather than in the entire factory.  Based on the




reported expected leukemia incidence of 0.1 cases for the 230 exposed em-




ployees, we can estimate approximately (70/230) x 0.1 = 0.03 cases for the




group exposed in the storage hall.  Compared with the two observed cases,




this yields a ratio of observed to expected cases of (2/0.03) = 66.7.




     The estimated average exposure to ethylene oxide over the lifetime of




the workers is calculated as follows:






                 20 ppm x 8/24 hr x 240/365 days x 9/45.6 yr







                             exposure = 0.865 ppm






                                   9-158

-------
where 45.6 years is the mean age of the 70 storage hall employees at  the end

of the study period.

     The slope by of the lifetime probability of dying from leukemia  due to a

lifetime of breathing ethylene oxide at 1 ppm (Section 9.5.3.2.) is given by


                       b  = P  (R - 1) Xl
                                 X2


where PQ is the lifetime probability of dying* from leukemia in the United

States in the absence of ethylene oxide exposure, R is the relative risk, Xj

is the exposure of 1 ppm, and X2 is the exposure experienced by the factory

workers.  The relative risk R estimated above is 66.7; the exposure X2 is

given as 0.865 ppm.  The lifetime probability of death from leukemia in the

U.S. population is 0.0091.  Substituting these values in the above equation

gives


                    b  = 0.0091 (66.7 - 1) = o^Cppm)-1
                     H       0.865 ppm


The probability associated with breathing ethylene oxide at 1 ppm for a

lifetime is


                           P = 1 - e~bH (1 ppm) = 0.50



To convert ppm to  y g/m , the formula is
*PQ employs both leukemia incidence cases and leukemia mortality rates.
 While leukemia mortality in the younger ages « 55) can be closely equated
 with incidence, in the older age groups chronic forms predominate in inci-
 dence, with death often occurring from other causes.  Nevertheless, for this
 assessment it is assumed that although ethylene oxide would cause all types
 of leukemias, death will result from each case.  In this study, the leukemias
 in the two women were of the acute form.

                                 9-159

-------
               1 ppm = 1'2 8   x 44.1 tn.w. chemical x 10% g x 1Q-6
                       IQ-3 m3      28.2 m.w. air        g



                     = 1.9 x 103U g/m3



Thus the unit risk estimate in terms of y g/m3 is


           bR = O.egCppm)'1 x     1 ppm	 = 3.6 x 10~4(u g/m3)"1
                              1.9 x 103y g/m3


Similarly, if the 95% limits on the relative risk from Table 9-35 are sub-

stituted for the point estimate, the 95% confidence limits on the unit risk

become 7.0 x 10~^(ppm)~^ to 2.5(ppm)"~ .  This represents a range of 36 and

encompasses the 95% upper-limit incremental unit risk, q^ = 1.9 x 10~

ppm~l, extrapolated from the Snellings et al. (1981) study.

     Based on the above analysis, we conclude that the carcinogenic

potency estimates for ethylene oxide derived from human data do not contradict

the estimate based on the rat inhalation studies.  Because of the uncertain-

ties in the epidemiologic study, however,  the animal inhalation study is

chosen for the 95% upper-limit incremental unit risk estimate for ethylene

oxide.

     9.5.3.4.  RELATIVE POTENCY — One of the uses of the concept of unit risk

is to compare the relative potencies of carcinogens.  For the purposes of the

present analysis, potency is defined as the linear portion of the dose-response

curve, and is used to calculate the required unit risk factors.  To estimate

relative potency on a per-mole basis, the unit risk slope factor is multiplied

by the molecular weight of the compound, and the resulting number, expressed

in terms of (mmol/kg/day)~l, is called the "relative potency index."

     Figure 9-4 is a histogram representing the frequency distribution of rel-

tive potency indices for 54 chemicals that have been evaluated by the GAG as

                                   9-160

-------

20
18
16
14
>- 12
CJ
:REQUEI^
o
8
6
4
2
n

—
«.
—
_
—
- I
—
"

-1







II
l«
0


4th 3rd 2nd 1st

QUARTILE QUARTILE QUARTILE QUARTILE
1 x 10+ 4 x 10+ 2 x 10

—

PH




IS
vXvX;
"***•**"•***
«***•***•*•*
B
M

^ivvV
•
^^^



—
H
>x'&x*
XvX'X'
:•••:>$•$
:£:;!::|:x'
•x'x'x'i
—
—
—

B
|:::;:;X;S; —
X'X'X'X *X* • "t't* P"^
mm ETTEI s?s R EHEI
12345678
                          LOG OF POTENCY INDEX
Figure 9-4. Histogram representing the frequency distribution of the potency indices
of 54 suspect carcinogens evaluated by the Carcinogen Assessment Group.
                                 9-161

-------
suspect carcinogens.  The data summarized by the histogram are presented in




Table 9-36.  Where human data have been available for a compound, such data




have been used to calculate these indices.  Where no human data have been




available, data from animal oral studies have been used rather than data from




animal inhalation studies, since animal oral studies have been conducted for




most of these compounds, and their use allows potency comparisons by route.




     On the basis of mononuclear cell leukemias and gliomas in female rats in




the Snellings et al. (1981) inhalation study, the relative potency index for




ethylene oxide has been calculated as 1.54 x 10+^.  This number was derived




by multiplying the slope in units of (ing/kg/day)"* by the molecular weight of




ethylene oxide, which is 44.1.  For the rat study, this slope is 3.5 x 10~1




(mg/kg/day)+1.




     The potency index for ethylene oxide is thus 3.5 x 10"~1 x 44.1 =




1.54 x 10~1, putting ethylene oxide at the bottom of the third quartile of the




54 chemicals which the GAG has evaluated as suspect carcinogens.  It should




be noted that the ranking of these relative potency indices is subject to the




uncertainties involved in comparing a number of potency estimates for differ-




ent chemicals on the basis of varying routes of exposure in different species,




using studies whose quality varies widely.  Furthermore, all of these indices




are based on estimates of low-dose risk that have been calculated by means of




linear extrapolation from the observational range.  The indices are, therefore,




not valid for the comparison of potencies in the experimental or observational




range if linearity does not exist there.









9.5.4.  Summary.  Ethylene oxide has been shown to be carcinogenic in animals




in long-term studies by three different routes of administration (inhalation,




subcutaneous injection, and gavage).  The most relevant route for human expo-




                                     9-162

-------
      TABLE  9-36.  RELATIVE CARCINOGENIC POTENCIES AMONG 54 CHEMICALS EVALUATED BY THE CARCINOGEN ASSESSMENT GROUP
                                              AS SUSPECT HUMAN CARCINOGENS
Level
of evidence3
Compounds
Acrylonitrile
Aflatoxin B^
Aldrin
Allyl chloride
Arsenic
B[a]P
Benzene
Benzidene
Beryllium
1 ,3-Butadiene
Cadmium
Carbon tetrachloride
Chlordane
CAS Number
107-13-1
1162-65-8
309-00-2
107-05-1
7440-38-2
50-32-8
71-43-2
92-87-5
7440-41-7
106-99-0
7440-43-9
56-23-5
57-74-9
Humans
L
L
I

S
I
S
S
L
I
L
I
I
Animals
S
S
L

I
S
S
S
S
S
S
S
L
Grouping
based on
IARC
criteria
2A
2A
2B

1
2B
1
1
2A
2B
2A
2B
3
Slope
(mg/kg/day)"1
0.24(W)
2900
11.4
1.19x10-2
15(H)
11.5
2.9xlO-2(W)
234(W)
2.6
l.OxlO-id)
7.8(W)
l.SOxlQ-1
1.61
Molecular
weight
53.1
312.3
369.4
76.5
149.8
252.3
78
184.2
9
54.1
112.4
153.8
409.8
Potency
index
1x10+1
9xlO+5
4x10+3
9x10-!
2xlO+3
3x10+3
2x10°
4xlO+4
2xlO+1
5x10°
9xlO+2
2xlO+1
7x10+2
Order of
magnitude
(Iog10
index)
+1
+6
+4
0
+3
+3
0
+5
+1
+1
+3
+1
+3
aS = Sufficient evidence; L = Limited evidence; I = Inadequate evidence.
                                                                                     (continued on the following page)

-------
                                                TABLE 9-36.   (continued)
Level
of evidence3
Compounds
Chlorinated ethanes
1 , 2-Dichloroethane
hexachloroe thane
CAS Number

107-06-2
67-72-1
1,1,2, 2-Tetrachloroethane 79-34-5
1,1, 2-Trichloroethane
Chloroform
Chromium VI
DDT
Dichlorobenzidine
1 , 1-Dlchloroethylene
(Vinylidene chloride)
Dichlorome thane
(Methylene chloride)
Dieldrin
2,4-Dinitrotoluene
Diphenylhydrazine
Epichlorohydrin
Bis(2-chloroethyl)ether
79-00-5
67-66-3
7440-47-3
50-29-3
91-94-1
75-35-4

75-09-2

60-57-1
121-14-2
122-66-7
106-89-8
111-44-4
Humans

I
I
I
I
I
S
I
I
I

I

I
I
I
I
I
Animals

S
L
L
L
S
S
S
S
L

L

S
S
S
S
S
Grouping
based on
IARC
criteria

2B
3
3
3
2B
1
2B
2B
3

3

2B
2B
2B
2B
2B
Slope
(mg/kg/day)"1

6.9xlO-2
1.42xlO-2
0.20
5.73x10-2
7xlO-2
41(W)
0.34
1.69
1.17(1)

6.3xlO-4(I)

30.4
0.31
0.77
9.9x10-3
1.14
Molecular
weight

98.9
236.7
167.9
133.4
119.4
100
354.5
253.1
97

84.9

380.9
182
180
92.5
143
Potency
index

7x10°
3x10°
3xlO+1
8x10°
8x10°
4xlO+3
1x10+2
4x10+2
1x10+2

5xlO-2

1x10+4
6xlO+1
lxlO+2
9x10-!
2x10+2
Order of
magnitude
(Iog10
index)

+1
0
+1
+1
+1
+4
+2
+3
+2

-1

+4
+2
+2
0
+2
aS = Sufficient evidence; L = Limited evidence;  I  =  Inadequate  evidence.
                                                                                    (continued  on  the  following  page)

-------
                                                TABLE 9-36.  (continued)
Level
of evidence3
Compounds
Bi s ( ch lor ome thy 1) ether
Ethylene dibromide (EDB)
Ethylene oxide
Heptachlor
^ Hexachlorobenzene
0! Hexachlorobutadiene
Hexachlorocyclohexane
technical grade
alpha isomer
beta isomer
gamma isomer
Hexachlorodibenzodioxin
Nickel
Nitrosamines
Dimethvlnitrosamine
Die thy Initrosamine
Dibutylnitrosamine
CAS Number
542-88-1
106-93-4
75-21-8
76-44-8
118-74-1
87-68-3


319-84-6
319-85-7
58-89-9
34465-46-8
7440-02-0

62-75-9
55-18-5
924-16-3
Humans
S
I
L
I
I
I


I
I
I
I
L

I
I
I
Animals
S
S
S
S
S
L


S
L
L
S
S

S
S
S
Grouping
based on
IARC
criteria
1
2B
2A
2B
2B
3


2B
3
2B
2B
2A

2B
2B
2B
Slope Molecular
(mg/kg/day)"1 weight
9300(1)
41
3.5x10-1(1)
3.37
1.67
7.75x10-2

4.75
11.12
1.84
1.33
6.2xlO+3
1 .15(W)

25.9(not by qf)
43.5(not by q|)
5.43
115
187.9
44.1
373.3
284.4
261

290.9
290.9
290.9
290.9
391
58.7

74.1
102.1
158.2
Potency
index
lxlO+6
8xlO+3
2xlO+1
lxlO+3
5x10+2
2xlO+1

lxlO+3
3xlO+3
5x10+2
4x10+2
2xlO+6
7xlO+1

2xlO+3
4xlO+3
9x10+2
Order of
magnitude
(Iog10
index)
+6
+4
+1
+3
+3
+1

+3
+3
-1-3
+3
+6
+2

+3
+4
+3
aS = Sufficient evidence;  L = Limited evidence;  I  =  Inadequate  evidence.
                                                                                    (continued  on  the  following page)

-------
                                                TABLE 9-36.   (continued)
Level
of evidence3
Compounds CAS Number
N-nit rosopyrrolidine
N-nitroso-N-ethylurea
N-nit roso-N-methylurea
N-nitroso-diphenylamine
PCBs
Phenols
2,4, 6-Trichlorophenol
Tetrachlorodibenzo-
p-dioxin (TCDD)
Tetrachloroethylene
Toxaphene
Trichloroethylene
Vinyl chloride
930-55-2
759-73-9
684-93-5
86-30-6
1336-36-3

88-06-2

1746-01-6
127-18-4
8001-35-2
79-01-6
75-01-4
Humans
1
I
I
I
I

I

I
I
I
I
S
Animals
S
S
S
S
S

S

S
L
S
L/S
S
Grouping
based on
IARC
criteria
2B
2B
2B
2B
2B

2B

2B
3
2B
3/2B
1
Slope
(mg/kg/day)"1
2.13
32.9
302.6
4.92xlO-3
4.34

1.99x10-2

1.56x10+5
6.0x10-2
1.13
1.2x10-2
1.75x10-2(1)
Molecular
weight
100.2
117.1
103.1
198
324

197.4

322
165.8
414
131.4
62.5
Potency
index
2x10+2
4x10+3
3xlO+4
1x10°
lxlO+3

4x10°

5x1 0+7
IxlO1
5xlO+2
2x10°
1x10°
Order of
magnitude
(Iog10
index)
+2
+4
+4
0
+3

+1

+8
+1
+3
0
0
aS = Sufficient evidence; L = Limited evidence; I = Inadequate evidence.

Remarks:
1.  Animal slopes are 95% upper-limit slopes based on the linearized multistage model.  They 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).  Human slopes are point estimates based on the linear nonthreshold model.

2.  The potency index is a rounded-off slope in (mraol/kg/day)"^- and is calculated by multiplying the slopes in
    (mg/kg/day)-'- by the molecular weight of the compound.

3.  Not all of the carcinogenic potencies presented in this table represent the same degree of certainty.  All are
    subject to change as new evidence becomes available.

-------
 sure  is  inhalation.  Two  long-term  inhalation  studies  in  rats were  performed




 that  adequately  tested  the  carcinogenic potential of ethylene oxide by  inhala-




 tion:  the Bushy Run study  (Snellings  et  al.,  1981) and the NIOSH study  (Lynch




 et al.,  1982).   Snellings et al.  (1981) found  that ethylene oxide exposure




 resulted in  an increased  incidence  of  mononuclear cell leukemia in  females  in




 the two  highest  dose groups; this increase was dose related.  The test  for




 linear trend was highly significant  (p <  0.0001).  There was also a signifi-




 cant  (p  = 0.045) increase in gliomas at the highest dose, and the test  for




 linear trend was highly significant  (p =  0.014).  In males, incidences  of




 primary  brain neoplasm, peritoneal mesothelioma, and subcutaneous fibroma




 were  significantly elevated in at least two exposed groups.  The trend analy-




 sis was  significant for both mesotheliomas (p < 0.00001) and gliomas (p =




 0.003) in males.  In the NIOSH (Lynch et  al., 1982) study, which involved




 only male rats,  leukemia incidence was significantly increased at low doses




 only, while  gliomas (mixed-cell) and peritoneal mesotheliomas were  increased




 significantly in the high-dose groups.  For these latter two sites,   the dose-




 response trend tests were also statistically significant (p < 0.01).  Other




positive results for the carcinogenicity  of ethylene oxide were demonstrated




by subcutaneous injection in mice and intragastric administration in rats.




     Three epidemiologic studies of workers exposed to ethylene oxide demon-




strated significant (p < 0.05)  association between ethylene oxide exposure




and the occurrence of cancer.   Two of the studies (Hogstedt et al.,   1979a, b)




found an association between ethylene oxide exposure and the incidence of




leukemia.  Ethylene oxide was  not found to be associated with any particular




type of leukemia, however.  Other sites or types of cancer found to  be signi-




ficantly (p < 0.05) associated  with ethylene oxide exposure in an individual




study  include pancreatic cancer and Hodgkin's disease in the Morgan  et al.




                                    9-167

-------
(1981) study and stomach cancer in the Hogstedt et al. (1979a) study.  The




possibility of confounding due to other chemical agents cannot be excluded in




any of the studies, however.




     An upper-limit incremental unit risk estimate of 1.0 x 10~^( p g/m^)""!




for ethylene oxide has been calculated, using a linearized multistage model,




on total mononuclear cell leukemias and brain gliomas in female Fischer 344




rats from the Bushy Run (Snellings et al., 1981) study.  Extrapolation




from the human leukemia data results in a highly uncertain risk estimate due




to the small numbers of leukemia cases that were observed and expected.




Quantitative comparisons of human and animal inhalation studies do,  to the




extent possible, support each other.









9.5.5.  Conclusions.  Ethylene oxide has been shown to be carcinogenic in




animals by intragastric, subcutaneous injection, and inhalation routes of




exposure.  Three human studies show an association between ethylene  oxide




exposure and an excess risk of cancer, but each of these studies has some




limitations.  Other evidence, which is in the mutagenicity section of this




document, supports the conclusions for carcinogenicity in that ethylene




oxide is a direct-acting alkylating agent, it reacts with mammalian DNA, it




induces base-pair substitutions in the Ames test and gene mutations  in




plants and animals, and it breaks chromosomes of plants, animals, and humans




and causes DNA damage in the spermatids of mice.




     Using the weight-of-evidence criteria of EPA's Proposed Guidelines for




Carcinogen Risk Assessment (U.S. EPA, 1984), the Carcinogen Assessment Group




considers the animal evidence for carcinogenicity to be "sufficient" and




the human evidence to be "limited" bordering on inadequate.  Based on both




the animal and human findings, the overall EPA classification for ethylene




                                    9-168

-------
oxide is Group Bl, meaning that ethylene oxide should be considered as probably




carcinogenic to humans.  This Bl classification is qualified as bordering on




Group B2 because of the limitations in the human evidence.




     According to the IARC guidelines for evaluating carcinogen evidence (See




Appendix 9B), ethylene oxide would be classified in Group 2A.  This classifi-




cation is similarly qualified as bordering on Group 2B because of limitations




in the human evidence.  A Group 2 classification, whether 2A or 2B, means




that ethylene oxide should be considered as probably carcinogenic in humans.




     An upper-limit carcinogenic potency value of 3.5 x 10" 1 (mg/kg/day)"*




has been calculated based on total mononuclear cell leukemias and brain




gliomas in female Fischer 344 rats in the Snellings et al. (1981) study.  An




upper-limit incremental unit risk of 1.0 x 10"^ (y g/ra^)~^ has also been




estimated using a linearized multistage model.
                                    9-169

-------
                                   APPENDIX 9A




             COMPARISON OF RESULTS BY VARIOUS EXTRAPOLATION MODELS









     The estimate of unit risk from animals presented in the body of this




document was calculated by use of the linearized multistage model.  This non-




threshold model is part of a methodology for estimating a conservative linear




slope at low extrapolation doses that is usually consistent with the data at




all dose levels in an experiment.  The model holds that the most plausible




upper limits of risk are those predicted by linear extrapolations to low




levels of the dose-response relationship.




     Other nonthreshold models that have been used for risk extrapolation




are the one-hit, the log-Probit, and the Weibull models.  The one-hit




model is characterized by a continuous downward curvature, but is linear




at low doses.  Because of its functional form, the one-hit model can be con-




sidered the linear form or first stage of the multistage model.  This fact,




together with the downward curvature of the one-hit model, means that the




model will always yield low-level risk estimates that are at least as large




as those obtained with the multistage model.  In addition, whenever the data




can be fitted adequately to the one-hit model, estimates based on the one-hit




model and the multistage model will be comparable.




     The log-Probit and the Weibull models, because of their general "S" curv-




ature, are often used for the interpretation of toxicological data in the




observable range.  The low-dose upward curvatures of these two models usually




yield lower low-dose risk estimates than those of the one-hit or multistage




models.




     The log-Probit model was originally used in biological assay problems




such as  potency assessments of toxicants and drugs, and is most often used to






                                    A-l

-------
estimate such values as percentile lethal dose or percentile effective dose.




The log-Probit model was developed along strictly empirical lines, in studies




where it was observed that several log dose-response relationships followed




the cumulative normal probability distribution function, $ .  In fitting the




log-Probit model to cancer bioassay data, assuming an independent background,




this relationship becomes







         P(D;a,b,c) = c + (1-c)  $  (a+blog10 D)   a,b > 0<  c < I







where P is the proportion responding at dose D, c is an estimate of the back-




ground rate, a is an estimate of the standardized mean of individual toler-




ances, and b is an estimate of the log-Probit dose-response slope.




     The one-hit model arises from the theory that a single molecule of a




carcinogen has a quantifiable probability of transforming a single normal cell




into a cancer cell.  In this model, the probability distribution function is
                     P(D;a,b) = l-exp-(a+bd)   a,b > 0
where a and b are the parameter estimates (a = the background or zero dose




rate, and b = the linear component or slope of the dose-response model).  In




considering the added risk over background, incorporation of Abbott's correc-




tion leads to







                     P(D;b) = l-exp-(bd)    b > 0







Finally, a model from the theory of carcinogenesis arises from the multihit




model applied to multiple target cells.  This model,  known as the Weibull




model,  is of the form




                     P(D;b,k) = l-exp-(bdk)   b,k > 0






                                     A-2

-------
For the power of dose only, the restriction k > 0 has been placed on this




model.  When k > 1, the model yields low-dose estimates of risks that are




usually significantly lower than either the multistage or one-hit models,




both of which are linear at low doses.  All three of these models—the multi-




stage, the one-hit, and the Weibull—usually project risk estimates that are




significantly higher at low exposure levels than those projected by the




log-Probit model.




     The results of both the male and female rat data sets from the Bushy




Run (Snellings et al., 1981) study are presented in Table 9A-1.  Surprisingly,




for the female rats, both the Weibull and log-Probit models yielded larger




estimates of risk than the multistage model, which, in this case, produced




results identical to those produced by the one-hit model.  For the males, the




one-hit model produced the highest estimates and the log-Probit model produced




the lowest; in this case, the multistage, one-hit, and Weibull all produced




similar results.
                                     A-3

-------
                  TABLE A-l.  ESTIMATES OF HUMAN LOW-DOSE RISK BASED ON DATA FROM MALE AND FEMALE FISCHER  344  RATS
                            IN THE BUSHY RUN ETO INHALATION STUDY, AS DERIVED FROM FOUR DIFFERENT MODELS.
                            ALL ESTIMATES INCORPORATE ABBOTT'S CORRECTION FOR INDEPENDENT BACKGROUND RATE
Maximum likelihood estimates of
additional risks
Continuous
human
exposure Multistage One-hit Weibull
ppm model model model
Males
.001 3.1x10-4 1.3x10-4 2.5x10-5
0.01 3.1x10-3 1.3x10-3 3.3x10-4
0.1 3.1x10-2 1.3x10-2 4.3x10-3
1 8.4x10-2 1.2x10-1 5.5x10-2
Females
.001 1.4x10-5 3.6x10-3
0.01 1.4x10-3 1.4x10-2
0.1 1.4x10-2 5.0x10-2
1 1.3x10-1 1.7x10-1
Animal exposure 0, 33 ppm, 100 ppm 6 hours/day, 5 days/week.
DATA
Human eq. dose - mg/kg/day
Males 0 0.35 0.94 2.63 Females

Log-Probit Multistage
model model
3.1x10-1 9.2x10-5
1.6x10-6 9.2x10-4
8.6x10-4 9.2x10-3
5.4x10-2 8.8xlO-2
1.2x10-4 1.9x10-4
3.0x10-3 1.9x10-3
3.3x10-2 1.9x10-2
1.7x10-! 1.7x10-1

Human eq. dose - mg/kg/day
0 0.28 0.75 2.11
957. upper confidence limit
of additional risks
One-hit Weibull Log-Probit
model model model
1.6x10-4 1.3x10-4 5.0x10-'-*
1.6x10-3 1.4x10-3 1.5x10-5
1.6x10-2 1.3x10-2 3.9xUr-l
1.6x10-1 9.7x10-2 9.7x1(1-2
1.5xl()-2 1.0x10-'
4.3x10-2 1.5x10-2
l.lxlO"! 9.Jxl()-2
2.5xH)-l 2.5x11)-'


No. tumors/No, examined     5/187    4/88    12/82    29/96
23/186    15/71    27/72    32/73
Conversions for low doses:  Humans       1 mg/kg/day = 1.84 ppm in air

                            or           1 ppm air = .543 mg/kg/day

                            Multistage and one-hit models gave identical results in females.

-------
                                    APPENDIX  9B


        INTERNATIONAL  AGENCY  FOR RESEARCH  ON  CANCER  CLASSIFICATION  SYSTEM


                   FOR THE  EVALUATION  OF THE  CARCINOGENIC  RISK


                             OF  CHEMICALS  TO  HUMANS*




ASSESSMENT  OF  EVIDENCE FOR CARCINOGENICITY FROM  STUDIES IN  HUMANS


     Evidence  of  carcinogenicity  from human  studies  comes from  three main


sources:


     1.  Case  reports  of individual cancer patients  who were  exposed to  the


         chemical  or  process.


     2.  Descriptive  epidemiological  studies in  which the incidence of cancer


         in human  populations was found to vary  in  space or time with exposure


         to the agents.


     3.  Analytical epidemiological (case-control and cohort) studies in which


         individual exposure to the chemical or  group of chemicals was found


         to be associated with an increased  risk of  cancer.


     Three criteria must be met before a causal  association can be inferred


between exposure and cancer in humans:


     1.  There is  no identified bias which could explain the association.


     2.  The possibility of confounding has been considered and ruled out as


         explaining the association.


     3.  The association is unlikely to be due to chance.


     In general,  although a single study may be  indicative of a cause-effect


relationship,  confidence in inferring a causal association is increased when


several independent studies are concordant in showing the  association,  when
*Adapted from International Agency for Research on Cancer.  Monographs Sup-
 plement 4, Evaluation of the Carcinogenic Risk of Chemicals to Humans, 1982.
 pp. 11-14.                                                                 '
                                    3-1

-------
the association is strong, when there is a dose-response relationship, or




when a reduction in exposure is followed by a reduction in the incidence of




cancer.




     The degrees of evidence for carcinogenicity from studies in humans were




categorized as:




     1.  Sufficient evidence of carcinogenicity, which indicates that there




is a causal relationship between the agent and human cancer.




     2.  Limited evidence of carcinogenicity, which indicates that a causal




interpretation is credible, but that alternative explanations, such as chance,




bias, or confounding, could not adequately be excluded.




     3.  Inadequate evidence, which indicates that one of three conditions




prevailed:  (a) there were few pertinent data; (b) the available studies,




while showing evidence of association, did not exclude chance, bias, or




confounding; (c) studies were available which do not show evidence of carcin-




ogenicity.









ASSESSMENT OF EVIDENCE FOR CARCINOGENICITY FROM STUDIES IN EXPERIMENTAL




ANIMALS




     These assessments were classified into four groups:




     1.  Sufficient evidence of carcinogenicity, which indicates that there




is an increased incidence of malignant tumors:  (a) in multiple species or




strains;  or (b) in multiple experiments (preferably with different routes of




administration or using different dose levels);  or (c) to an unusual degree




with regard to incidence, site or type of tumor, or age at onset.  Additional




evidence may be provided by data on dose-response effects, as well as infor-




mation from short-term tests or on chemical structure.






                                      B-2

-------
     2.  Limited evidence of carcinogenicity, which means that the data sug-




gest a carcinogenic effect but are limited because:  (a) the studies involve




a single species, strain, or experiment; (b) the experiments are restricted




by inadequate dosage levels, inadequate duration of exposure to the agent,




inadequate period of follow-up, poor survival, too few animals, or inadequate




reporting; or (c) the neoplasms produced often occur spontaneously and, in the




past, have been difficult to classify as malignant by histological criteria




alone (e.g., lung and liver tumors in mice).




     3.  Inadequate evidence, which indicates that because of major qualita-




tive or quantitative limitations, the studies cannot be interpreted as showing




either the presence or absence of a carcinogenic effect; or that within the




limits of the tests used, the chemical is not carcinogenic.  The number of




negative studies is small, since, in general, studies that show no effect are




less likely to be published than those suggesting carcinogenicity.




     4.  No data indicates that data were not available to the Working Group.




     The categories sufficient evidence and limited evidence refer only to the




strength of the experimental evidence that these chemicals are carcinogenic




and not to the extent of their carcinogenic activity nor to the mechanism




involved.  The classification of any chemical may change as new information




becomes available.









EVALUATION OF CARCINOGENIC RISK TO HUMANS




     At present, no objective criteria exist to interpret data from studies




in experimental animals or from short-term tests directly in terms of human




risk.  Thus, in the absence of sufficient evidence from human studies, evalua-




tion of the carcinogenic risk to humans was based on consideration of both




the epidemiological and experimental evidence.  The breadth of the categories




                                     B-3

-------
of evidence defined above allows substantial variation within each.  The de-




cisions reached by the Working Group regarding overall risk incorporated these




differences, even though they could not always be reflected adequately in the




placement of an exposure into a particular category.




     The chemicals, groups of chemicals, industrial processes, or occupational




exposures were thus put into one of three groups:




Group 1




     The chemical, group of chemicals, industrial process, or occupational




exposure is carcinogenic to humans.  This category was used only when there




was sufficient evidence from epidemiological studies to support a causal




association between the exposure and cancer.




Group 2




     The chemical, group of chemicals, industrial process, or occupational




exposure is probably carcinogenic to humans.  This category includes exposures




for which, at one extreme,  the evidence of human carcinogenicity is almost




"sufficient," as well as exposures for which, at the other extreme, it is




inadequate.  To reflect this range, the category was divided into higher




(Group A) and lower (Group B) degrees of evidence.  Usually, category 2A was




reserved for exposures for which there was at least limited evidence of




carcinogenicity to humans.   The data from studies in experimental animals




played an important role in assigning studies to category 2, and particularly




those in Group B; thus, the combination of sufficient evidence in animals and




inadequate data in humans usually resulted in a classification of 2B.




     In some cases, the Working Group considered that the known chemical prop-




erties of a compound and the results from short-term tests allowed its trans-




fer from Group 3 to 2B or from Group 2B to 2A.




                                    B-4

-------
Group 3




     The chemical, group of chemicals, industrial process, or occupational




exposure cannot be classified as to its carcinogenicity to humans.
                                    B-5

-------
                                10.   REFERENCES

Alguire, D.E.  1973.  EthyLsne oxide gas.   Food Technology,   pp. 64-6?.


Altshuller,  A.P.  and  J.J.  Bufalini.    1965.    Photochemical aspects  of air
pollution:  A review.  Photochern. Photobiol.   4:  97-146.


Anderson,  S.R.   1971.   Ethylene  oxide  toxicity:   A study of tissue  reactions
to retained ethylene oxide.  J. Lab.  Clin.  Med.  77:  346-356.


Ando,  A.    1968.    Mutation  induction  in  rice  by radiation  combined  with
chemical  protectants  and  mutagens.    International   Atomic  Energy   Agency,
Vienna, Tech. Rep. Ser. 86, pp. 7-15.  (Cited  in NIOSH,  1977).


Anonymous.  1947.  Ethylene oxide poisoning.   Ind.  Hyg.  Newsletter.  7:  3-6.


Anonymous.  1977.  New process for preserving  wood.   Adhesives Age.  5:  56.


Anonymous.   1981 a.   Chemical  Profiles.   Ethylene  Oxide.   Chemical  Marketing
Reporter.  June  8, 1981.  p.  9.


Anonymous.   198lb.   Chemical  Profiles.  Ethylene  GLycol.  Chemical  Marketing
Reporter.  June 15,  1981.   p. 9.


Anonymous.  198lc.  Key Chemicals.   Ethylene Oxide.  Chem. Eng. News.   July 6,
1981.  p.  10.


Anonymous.   1982.   Aliphatic Organics Forecast.  Chemical Marketing  Reporter.
January 4, 1982.   p. 33.


Appelgren, L.,  G.  Eneroth and  C. Grant.    1977.   Studies  on ethylene oxide:
Miole body auto radiography and dominant lethal  test  in  mice.   Proc. Eur. Soc.
Toxicol.  18: 315-317.


Appelgren, L.E., G. Eneroth,  C.  Grant,  L .E.  Lanstrom and K. Tenshasen.  1978.
Testing of ethylene oxide  for mutagenicity using the micronucleus test  in mice
and rats.  Acta.  Pharmacol. Toxicol.   43(1):  69-71.
                                     10-1

-------
Balazs, T.   1976.   Toxicity of ethylene  oxide and chloroethanol.   Food and
Drug Administration Bylines.   7:  150-155.


Barnard, J.A. and R.K.Y. Lee.  1972.   Combustion of n-pentane  in a  shock  tube.
Combust. Sci. Techno 1.  6(3): 143-150.


Bartnicki,  E.W.   and  C.E.  Castro.    1969.    Biodehalogenation.    Pathway for
transhalogenation and  the  stereochemistry  of  epoxide  formation   from halo-
hydrins.  Biochemistry.   8(12):  1677-4680.


BASF  Wyandotte  Corporation.   1972.   Ethylene oxide handling  manual.   Report
No. 21.  BASF Wyandotte  Corporation,  Wyandotte,  Michigan.


Bayes,  A.L.     1979.     Ethylene  oxide  one-gene rat ion  reproduction  study.
(Carnegie -Me lion  Institute  of  Research).    Submission  to  the Environmental
Protection  Agency under  the Toxic Substances  Control  Act.   Citation number
8EHQ-0379-0248.   (Later published as Snelling  et al. ,  1982).


Ben-Yehoshua, S. and  P.  Krinsky.   1968.   Gas chromatography of ethylene  oxide
and its toxic residues.   J. Gas Chroma tog.   6:  350-351.


Ben-Yehoshua, S. , S.  Staler, A.  Goldner and P. Krinsky.  1971.  Toxic  residues
in  Hayani dates  after fumigation with  ethylene  oxide.  Pest.  Chem.  Proc.  Intl.
Congr. Pest. Chem. 2nd.   4: 593-605.


Bertsch,  W. ,  R.C. Chang  and  A.  Zlatkis.   1974.  The determination  of organic
volatiles  in  air  pollution  studies.    Characterization  of  profiles.    J.
Chromatog. Sci.   12:  175-182.


Binder, H.   1974.   Ethylene oxide and chlorohydrin in tobacco and  its smoke.
Fachliche Mitt.  Oesterr. Tabakregie 1974.  15: 294-301.


Binder,  H.   and  W.  Lindner.   1972.   Determination of ethylene  oxide in  the
smoke   of  treated   and  untreated  cigarettes.    Fachliche   Mitt.   Oesterr.
Tabakregie.  13:  215-220.


Bird,  M.   1952.   Chemical production of mutations in Drosophila:   Comparison
of techniques.   J.  Genet.  50: 480-485.
                                      10-2

-------
Biro,  L.,  A. Fishes  and  E. Price.   1974.  Ethylene  oxide burns,  a hospital
outbreak involving 19 womsn.  Arch. Dermatol.   110:  924.


Blackford, J.L.   1974.  Aery Ion itrile.   Chemical  Economics  Handbook.  Stanford
Research Institute, Nfenlo Park,  California.


Blackford,  J.L.   1976a.    Propylene  oxide.    Chemical  Economics Handbook.
Stanford Research Institute, Msnlo Park,  California.


Blackford,  J.L.    1976b.    Ethylene  oxide.    Chemical  Economics Handbook.
Stanford Research Institute, Msnlo Park,  California.


Blackwood, J.  and E.  Erskine.    1938.   Carboxide  poisoning.   U.S. Navy Med.
Bull.  36: 44-45.


Bogan,  D.J.  and  C.W.  Hand.  1978.   Absolute  rate constant,  kinetic   isotope
effect,  and  mechanism  of  the  reaction  of  ethylene  oxide  with  oxygen ( P)
atoms.  J. Phys. Chem.  82:  2067-2073.


Bridie, A.L.,  C.J.M.  Wblff  and  M. Winter.   1979.   The acute toxicity   of some
petrochemicals to goldfish.   Water Res.   13:  623-626.


Bronsted, J.N.,  M.  Kilpatrick  and  M.  Kilpatrick.    1929.   Kinetic  studies on
ethylene oxides.  J.  Amer.  Chem. Soc.   51:  428-461.


Brown, R.H.  and  C.J. Pumell.   1979.   Collection and analysis of trace  organic
vapour  pollutants in  ambient  atmospheres:   The  performance  of  a Tenax  GC
adsorbent tube.   J. Chroma tog.   178:  79-90.


Brown, A.M.,  C.  Bruch, E.  Jackson, K.  Krell,  B. Page and  M.  Thomas.   1979.
Increased mutation frequency  due to ethylene  oxide adsorbed  to plastics.   Iri
vitro.  15:  220-221.
Brown, D.  1970.  Determination of ethylene  oxide and  ethylene chlorohydrin in
plastic  and  rubber  surgical  equipment  sterilized  with  ethylene  oxide.   J.
Assoc. Off. Anal.  Chem.   53:  263-267.


Browning, J.B.   1982.   Union Carbide  Corporation.   (Aug.  5)  Letter to Office
of Toxic Substances,  U.S.  Environmental Protection Agency.
                                     10-3

-------
Bruch, C.   1973.   Industrial Sterilization.   Phillips,  G.  and W.  Miller,  eds.
Duke University Press,  Durham,  North  Carolina,  pp.  119-123.


Bruch,  C.  and  M.  Koesterer.   1961.   The  microbicidal activity  of gaseous
propylene oxide and its application  to powdered or  flaked foods.   J. Food  Sci.
26: 428.
Carlson, R.M. and  R.  Caple.   1977.   Chemical/biological implications of using
chlorine and ozone for disinfection.  EPA-600/3-77-066,  Environmental Research
Laboratory, Duluth, Minnesota.


Carpenter,  C.,  H. Smyth  and  U. Pozzani.    1949.   The  assay  of  acute  vapor
toxicity, and the grading and   interpretation of  results on 96  chemical com-
pounds.  J. Ind. Hyg.  Toxicol.   31:  3^3-349.


Casteignau,  G.   and   J .L.  Halary.     1972.    Identification  et  dosage  par
chromatographic   enphase gazause d'oxetanues  disubstitues en 3  et  de quelques
autres ethers-oxydes.   Bull.  Soc. Chem.  Fr.   420-428.


Castro,  C.E.  and E.W. Bartnicki.   1968.  Biodehalogenation.   Epoxidation  of
halohydrins.  Epoxide opening  and  transhalogenation  by a  flavobacterium  sp.
Biochemistry. 7: 3213-3218.


Cawse,  J.N.,  J.P. Henry,  M.W.   Swartzlander  and'P.H.  Wadia.   1980.   Ethylene
Oxide.    In;    Kirk-Othmer  Encyclopedia  of  Chemical  Technology,  3rd  ed.
Grayson,   M.  and  D.   Eckroth,   eds.    John  Wiley  and  Sons, Inc., New  York.
vol. 9, pp. 432-471.


Chemical Week:   1982 Buyers' Guide Issue.  1981.  McGraw-Hill,  Inc., New York.
p. 417.


Clarke,  C. ,  W.  Davidson  and  J. Johnson.  1966.   Hemolysis  of  blood following
exposure to an  Australian manufactered  plastic  tubing sterilized  by  means  of
ethylene oxide gas. Australia  (South) New Zealand J. Surg.  36: 53-55.


Cobis,  J.   1977.   Correspondence  from  the  Department of  Medicine  and  Surgery
(Code  134D),  Veterans   Administration,  Washington,   D. C. ,  March 17,   1977.
(Cited in Glaser, 1977).
                                     10-4

-------
Cogswell, S.A.   1980.   Chemical Economics Handbook.   Marketing Research Report
on  Ethylene  Oxide.  SRI  International,  Menlo Park,  CA.   January,  1980.   pp.
65*1.50323, 654.5032Q, 654.5032U.


Conan, L., B.  Foucault,  G. Siou, M. Chalgneau and G.  LeMoan.   1979.  Study of
the  mutagenic  action of ethylene  oxide,  ethylene glycol,  and 2-chloroethanol
residues  in  plastic material  sterilized  by ethylene  oxide.   Ann.  Falsif.
Expert. Chim.  72: 141-151.


Conway,   R.A.,   G.T.  Waggy,   M.H.  Spiegel   and  R.L.  Berglund.       1983.
Environmental fate and effects  of ethylene  oxide.   Environ.  Sci.  Technol.   17:
107-112.


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.


Crutzen,  P.J.  and J.  Fishman.    1977.    Average  concentrations of  OH  in  the
troposphere and  the  budgets of CH., CO,  H0,  and  CH-CC1     Geophys.  Res. Lett.
4: 321-324.                       H       d         J   5
Gumming, R.B., G.A. Sega,  C.Y. Horton  and  W.H.  Olson.   1981.   Degree of alkyl-
ation of DMA  in various  tissues  of the mouse following  inhalation  exposure  to
ethylene oxide.  Environ. Mutagen.  3(3):  343.


Gumming, R.B.,  T.A.  Michand, L.R.  Lewis  and W.H.  Olson.   1984.    Inhalation
mutagenesis in mammals.   I.   Patterns of unscheduled  DNA synthesis  induced  in
germ cells  of male mice  by  exposure  to ethylene  oxide   in air.    (In  press).
(Mutagen.  Res.).


Cupitt,   L.T.    1980.    Fate  of  toxic  and  hazardous  materials   in  the air
environment.  U.S. EPA.  EPA-600/3-80-084.   NTIS PB  80-221948,  35 pp.


Cupitt,  L.T.   1983.   Personal communication between Dr.  L. Cupitt  of ESRL and
Dr. D.A. Gray of Syracuse Research Corporation.   March  1983.


Curme,  G. and  F.  Johnston (eds.).  1952.    Glycols.  American  Chemical  Society
Monograph,  Series 114.   Reinhold, New  York.


Cvetanovic,  R.J.    1955.    Mercury photosensitized  decomposition  of ethylene
oxide.   Can. J. Chem.   33: 1684-1695.
                                     10-5

-------
Darnall,  K.R., A.C.  Lloyd,  A.M.  Winer and J.N. Pitts, Jr.   1976.   Reactivity
scale  fbr  atmospheric  hydrocarbons based on  reaction with  hydroxyl  radical.
Environ.  Sci. Technol.   10:  692-696.
DeBont, J.A.M. and  R.A.J.M. Albers.   1976.
An ton ie van Leeuwenhoek.  42(1):  78-80.
  Microbial metabolism of ethylene.
Billing, W.L.  1977.  Interphase transfer processes.
chloro me thanes,  ethanes,   ethylenes,   propanes,   and
aqueous solutions.   Comparisons with  theoretical predictions.
Technol.  11: 405-409.
           II. Evaporation rates of
            propylene   from  dilute
                      Environ.  Sci.
Dimitriades, B. and  S.B.  Joshi.   1977.   Application of reactivity criteria in
ox id ant-re la ted  emission  control  in  the   U.S.A.     In:   Inter.   Conf.   on
Photochemical  Oxidant  Pollution   and   Its   Control.    Dimitriades,  B.,  ed.
Research Triangle  Park,  North Carolina,  pp.  705-711.  EPA-600/3-77-011B.
Dobbs, A.J., B.C.  Gilbert and R.O.C. Norman.   1971.   Electron spin resonance
studies.  Part XXVII.   The  geometry  of  oxygen-substituted  alkyl  radicals.   J.
Chem. Soc.  A:  124-135.
Dobinaon, B., W. Hoffman  and  B.D.  Stack.
Groups.  Pergamon Press,  Oxford.
1969.   The Determination of Epoxide
Dolovich, J.  and  B.  Bell.   1978.   Allergy to a  produet(s)
gas:   Demonstration  of  IgE and  IgG antibodies and  hapten
Allerg. Clin. Immunol.   62:  30-32.
                  of ethylene  oxide
                  specificity.   J.
Dow Chemical  Co.   1977.   Alkylene  oxides from Dow.
110-551-77R, Midland,  Michigan.
          Dow Chemical Co.  Form No.
Dunkelberg,  H.    1979.    On   the  oncogenic  activity  of  ethylene   oxide  and
propylene oxide  in mice.   Br.  J.  Cancer.   39(4): 588-589.


Dunkelberg,  H.    1981.    Kanzerogene  Aktivitat  von  Sthylenoxid  und  seinen
Reaktions-produkten    2-Chloroethanal,    2-Bromoethanol,    Ethylenglykol   und
Diethylenglykol.    T.   Kanzerogenitat von  Ethylenoxid  in  Vergleich  zu  1,2-
Propylenoxid bei  Subkutaner Applikation  an  Mausen.  Zbl.  Bakt. Hyg.  I.  Abt.
Orig.  B 174: 383-404.
                                     10-6

-------
 Dunkelberg,  H.   1982.   Carcinogenicity of  ethylene  oxide and  1 ,2-propylene
 oxide upon  intragastric administration to rats.  Br. J. Cancer.  46:  924-933.


 Eastham,  A.M.  and  G.A.  Latremouille.    1952.   Ethylene  oxide V  reaction  with
 halide ions  in neutral and acid solution.  Can. J. Chem.   30:  169.


 Ehrenberg,  L.  and  A.  Gustaffson.   1957.  On  the  mutagenic action  of ethylene
 oxide and diepoxybutane in barley.  Hereditas.  43: 595-602.


 Ehrenberg,  L.  and T. Hallstrom.   196?.   Haema to logic  studies  on  persons occu-
 pational ly  exposed  to  ethylene  oxide.   In;   International  Atomic  Energy  Agency
 Report SM 92/96,  pp. 327-334.


 Ehrenberg,  L.  and  S. Hussain.    1981.    Genetic  toxicity  of some  important
 epoxides.   Mutat. Res.  86: 1-113.


 Ehrenberg, L., A. Gustaffson  and  U. Lundquist.   1959.  The mutagenic  effect of
 ionizing  radiation  and reactive  ethylene derivatives  in  barley.   Hereditas.
 45:  351-368.  (Cited in NIOSH, 1977).


 Ehrenberg, L., K. Hiesche, S. Osterman-Golker  and  I. Wennberg.   1974.  Evalua-
 tion  of genetic  risks of  alkylating  agents:   Tissue  doses  in  the  mouse  from
 air  contaminated with  ethylene oxide.   Mutat.  Res.   24: 83-103.


 Ehrenberg,  L. ,  A.  Gustaffson and U.  Lundqvist.    1956.    Chemically induced
mutation and sterility in  barley.  Acta.  Chem. Scand.  10:  492-494.   (Cited in
NIOSH,  1977).


 Embree,  J.  and  C.  Hine.    1975.    Mutagenicity  of ethylene  oxide.   Toxicol.
Appl. Pharmacol.   33:  172-173.


Embree, J.,  J. Lyon and C.H.  Hine.  1977.  The mutagenic potential of ethylene
 oxide using  the dominant  lethal assay  in rats.  Toxicol.  Appl. Pharmacol.  40:
261-267.


EOIC   (Ethylene   Oxide  Industry   Council  of  the   Chemical  Manufacturer's
Association).   1984.   Research directed at   the  development  of a  sensitive
analytical method for  the quantification of ethylene oxide and its metabolites
in biological  samples.  Final  report.    Prepared  by  Arthur D.  Little,  Inc.,
Cambridge, MA.
                                     10-7

-------
Faberge,  A.C.    1955.    Types of  chromosome  aberrations induced  by ethylene
oxide in maize.  Genetics.   40:  571.   (Cited  in NIOSH,  1977).


Fahmy, 0.  and  M.  Fahmy.   1970.  Gene elimination in carcinogenesis:   Reinter-
pretation of the somatic  nutation  theory.   Cancer  Res.  30:  195-205.


Fishbein, L.   1976.   Potential hazards of fumigant residues.   Environ.  Health
Perspect.  14:  39-45.
Fomenko, V.N. and E.E. Strekalova.  1973.  Mutagenic action of some industrial
poisons  as a  function  of concentration and  exposure time.    Toksikol.  Nov.
Prom. Khim. Veshchestv.   13:  51-57.


Francis, M.   1985.   Letter  to  Herman Gibb,  Carcinogen Assessment Group,  U.S.
Environmental Protection Agency.  Subject:   Search  request on methyl formate,
107-31-3.    Environmental   Mutagen,  Carcinogen,  and  Teratogen  Information
Program, Oak Ridge National  Laboratory,  Oak Ridge, TN.


Freed,  V.H.,  C.T.  Chiou and R. Haque.   1977.   ChenDdynamics:   Transport and
behavior of chemicals in  the environment - A  problem in environmental health.
Environ. Health Per spec.   20: 55-70.


Fritz,  B.,  K.  Lorenz,  W. Steinert and  R. Zellner.   1982.   Laboratory kinetic
investigations of the tropospheric oxidation  of selected industrial emissions.
Comm. Eur.  Communities  (Rep.) Eur. Phys. -Chem.  Behav.  Atmos. Pollut.   7624:
192-202.


Gaines,  Personal Communication.   Union Carbide Corporation.  203-794-2000.


Garry,  V.V. , J.  Hozier,  D.  Jacobs,  R.L. Wade and D.G. Gray.   1979.   Ethylene
oxide:  Evidence of human chromosomal  effects.  Environ. Muta.  1:  375-382.


Garry,  V.F.,  C.W.  Opp,  J.K.  Wiencke  and D.  Lakatua.   1982.   Ethylene oxide
induced  sister  chromatid  exchange   in  human  lymphocytes  using a  membrane
dosimetry system.  Pharmacol.  25:  214-221.


Garry,  V .F., C.W. Opp  and J.K.  Wiencke.  1981.  Ethylene-oxide induced sister
chromatid  exchange  in  human-lymphocytes using  a  membrane  dosimetry  system.
Environ. Mutagen.  3(3): 340.
                                      10-8

-------
Generoso,  W.M. ,  K.T.  Cain,  M.  Krishna,  C.W.  Sheu  and  R.M.  Gryder.    1980.
Heritable  trans location  and  dominant-lethal nutation  induction  with ethylene
oxide in mice.  Mil tat.  Res.  73(1): 133-142.


Gilmour, R.H.   1978.   Ethylene oxide  treatment  of cosmetics.   Soap Perfum.
Cosnet.  51(11):  498-499.


Corner,  R.  and  W.A.  Noyes.    1950.    Photochemical studies.   XLJI. Ethylene
oxide.  J. Am. Chem.  Soc.  72:  101-108.


Goncarlovs,  V.    1983.    Personal  communication  between  Dr.  D.A. Gray   of
Syracuse Research  Corporation  and  V.  Goncarlovs  of the  Office of Pesticides.
March, 1983.


Graedel, T.E.   1978.  Chemical  Conpounds  in  the Atmosphere.  Academic Press,
New York.  pp. 267, 272,  276.


Greaves-Walker, W. and C.  Greeson.   1932.  The  toxicity  of ethylene  oxide.   J.
Hyg.  32: 409-416.


Gritter, R.J.  and E.G.  Sabatino.   1964.   Free-radical chemistry  of cyclic
ethers.  VII.  Ultraviolet  photolysis of epoxides  and propylene sulfide  in the
liquid phase.  J.  Org.  Chem.   29:  1965-1967.


Gross, J.,  M. Haas and  T. Swift.   1979.  Ethylene oxide neurotoxicity:  Report
of four cases and  review of the literature.  Neurology.  29:  978-982.


Hackett, P.L., M. G.  Brown, R .L.  Buschbom et al.  1982.   Teratogenic study  of
ethylene and propylene oxide and n-butyl acetate.   Battelle,  Pacific  Northwest
Laboratories.  Prepared for NIOSH,  contract  number 210-80-0013.


Hanson,  C.  and   A.C.  Leo.    1979.    Substituent  Constants  for  Correlation
Analysis  in  Chemistry  and Biology.    John Wiley and  Sons,   Inc.,  New   York.
p. 176.


Hatch, G. G. , P.D.  Manay,  C. C.  Christensen, R. Lagenbach, C.R.  Goodheart and  S.
Nesnow.   1982.   Enhanced viral transformation  of primary  Syrian hamster lung
(V-79)  cells  exposed to  ethylene  oxide in sealed  treatment chambers.    Proc.
Am. Assoc.  Cancer  Res.  23: 74.
                                     10-9

-------
Hawley,  G.G.  (ed. ).    1981.   The Condensed Chemical Dictionary, 10th  ed. ,  Van
No strand Reinhold Company, New York.   p. 435.


Hemminki,  K. ,  P.  Mutinen,  I.  Saloniemi,   M.L.  Niemi and  H.  Vainis.    1982.
Spontaneous  abortions   in  hospital staff  engaged   in  sterilizing   instruments
with chemical agents.  Br. Med. J.  285: 1461-1463.


Hendry, D.G. , T. Mill,  L.  Piszkiewicz,  J .A. Howard  and H.K. Eigenmann.   1974.
A  critical review of  H-atom  transfer  in   the  liquid  phase.   J.  Phys.  Chem.
Ref. Data.  3:  937-978.
Hine,  C.  and  V.  Rowe.   1973.   Industrial  Hygiene  and Toxicology.   2nd  ed.
vol. 2, Interscience, New York.   pp.  1600-1601.


Hine,  C. ,  V.K.  Rowe,  E.R.  White,   K.I.  Darmer and  G.T.   Youngblood.    1981.
Expoxy  Compounds.    In:   Toxicology.   Clayton,  G.D.  and   F.E.  Clayton,  Eds.
Volume  2A  of  Patty's  Industrial   Hygiene  and  Toxicology.     Third   revised
edition.  John Wiley and Sons, Inc.,  NY.   pp.  2141-2258.


Hirose, T.,  R.  Goldstein  and C. Bailey.   1963.   Hemolysis  of blood  due  to
exposure to  different  types of  plastic  tubing  and the  influence of ethylene
oxide sterilization.  J. Thorac.  Cardiov.  Surg.   45: 245-251.


Hirose, C.    1974.   The microwave spectra  and  r0, rs,  and rm  structures  of
ethylene oxide.  Bull.  Chem. Soc. Japan.   47:  1311-1318.


Hogstedt,  C., 0. Rohlen, B.S. Berndtson, 0. Axelson and  L. Ehrenberg.    1979a.
A cohort study  of  mortality and  cancer incidence  in ethylene oxide production
workers.  Br. J. Ind. Med.   36:  276-280.


Hogstedt,  C. , N. MaMgrist and B. Wadman.   1979b.  Leukemia in workers  exposed
to ethylene oxide.   J.  Am.  Med.  Assoc.   241(11):  1132-1133.


Hogstedt,   C. ,  L.  Aringer  and   A.  Gustavasson.    1984.   Ethylene  oxide  and
cancer:  A   literature  review and   follow-up  of  two   epidemiologic  studies.
Accepted for publication in  Arbete  och  Halse.
Hollingsworth, R. ,  V.  Rowe, F.  Oyen,  D. McCollister and  H. Spencer.    1956.
Toxicity  of ethylene  oxide determined  on  experimental animals.    Am.  Med.
Assoc.  Arch. Ind. Health.   13:  217-227.
                                     10-10

-------
Holmberg,  P.C.    1979.    Central-nervous-system defects  in  children  born  to
mothers exposed to organic solvents.   Lancet.   2: 8135:  177-179.


HoMberg,  P.C.  and M.  Nurminen.    1980.    Congenital  defects of  the nervous
system and occupational  factors  during pregnancy.   A  case-re fere nee study.
Am.  J. Ind. fed.   1:  167-176.


Holmgren, A., N. Diding and G. Samuelsson.  1969.  Ethylene  oxide  treatment of
crude drugs.   Part V.  Formation  of  ethylene  chlorohydrin.    Acta Pharmacol.
Sciencia.  6: 33-36.


Howe, R.B.   1983.   GLOBAL 83.  An experimental program developed  for  the U.S.
Environmental Protection Agency as an unpdate  to GLOBAL 82:  A computer program
to  extrapolate  quantal animal toxicity data  to low  doses  (May 1982).   K.S.
Crump and  Co. ,  Inc. , Ruston, LA.   Unpublished.


Hughes,  K.J., R.W.  Hum and F.G. Edwards.   1959.   Separation and identifica-
tion  of   oxygenated  hydrocarbons  in  combustion  products   from  automotive
engines.     In:  Proc.  2nd  Int.  Symp.  on  Gas  Cnromatography,   East   Lansing,
Michigan,  p. 171-182.


Hughes, T.W. , D.R. Tierney and Z.S. Khan.   1979.   Measuring  fugitive emissions
from petrochemical plants.  Chem.  Eng. Prog.  August 1979, pp. 35-39.
Hussain, S.  and  L.  Ehrenberg.   1975.  Prophage inductive efficiency of alkyl-
ating agents and radiations.   Intl. J. Radiat. Biol. Relat. Stud. Phys.   Chem.
Med.  27:  355-362.


I ARC  (International  Agency for  Research  on  Cancer)   1976.   Ethylene  oxide.
In;   IARC  Monographs on the Evaluation of  Carcinogenic  Risk  of  Chemicals to
Man.  Lyon,  France.


IARC  (International  Agency for  Research  on  Cancer)  1982.    IARC  Monographs:
Evaluation of  the Carcinogenic Risk  of  Chemicals  to  Humans.   Supplement 4.
Lyon, France.


Ishishi, E.  and  Y.  Kobayashi.   1973.  Treatment  of a  waste solution obtained
in  the  manufacture  of epoxide by  chlorohydrination.   Japan Kokai 73,  10,012,
5 pp.
                                     10-11

-------
Jacobson,  K.,  E.  Hackley and  L.  Feinsilver.   1956.   The toxicity  of inhaled
ethylene oxide and propylene oxide vapors.  Arch. Ind. Health.  13: 237-2M.
Jana, M.K. and K.  Roy.   1975.   Effectiveness and efficiency of methyl methane-
sulfonate and ethylene oxide for the induction  of  mutations  in mice.   Mutagen.
Res.  28(2): 211-215.  (Cited in NIOSH, 1977).
Jay, W.M., T.R. Swift and D.S. Hull.
oxide exposure to cataract formation.
           1982.  Possible  relationship  of ethylene
            Am.  J.  Ophthalmol.   93(6):  727-732.
Jefferson  Chemical  Co.
Houston, Texas.
Undated  a.    Ethylene  oxide  technical  brochure.
Jefferson  Chemical  Co.
Houston, Texas.
Undated  b.    Propylene  oxide  technical  brochure.
Jensen,  E.    1977.    Union  Carbide  Corp.  (unpublished).
Register, 43: 3812,  1978).
                                  (Cited in  Federal
Jenssen, D.  and C. Ramel.   1980.   The  micronucleus  test as part  of  a short-
term mutagenicity  test program for the  prediction  of  carcinogenicity evaluated
by 143 agents tested.  Mutat. Res.   75(2): 191-202.


Johnson and  Johnson.   1982.  Preliminary report of pilot  research chromosome
study of workers at  sites where  ethylene oxide  gas is utilized as a sterilant.
Unpublished  report  available  from  Dr.  J.   Paul   Jones,  Director  of  Health
Sciences, Johnson and Johnson,  New Brunswick, NJ, p. 18.


Jones, A.R. and G. Wells.   1981.  The  comparative  metabolism of 2-bromoethanol
and ethylene oxide in the rat.   Xenobiotica.   11(11):  763-770.


Jones, L.A. and D.M. Adams.  1981.   Mutations in Bacillus  subtilis var.  niger
(Bacillus   globigii)   spores   induced   by   ethylene   oxide.      Sporulation
Germination, Proc.  Int.  Spore Conf.  VIII.   Levinson,  H.S., A.L.  Sonenshein,
and D.J. Tipper, Eds.  American Society of Microbiology, Washington, DC.
Jones-Price, C., C.A. Kimmel, T.A. Marks, T.A. LeDoux,  J.R.  Reel,  P.W.  Fisher,
L.  Langhoff-Paschke  and M.  Marr.   Teratologic evaluation  of ethylene  oxide
(CAS  No.  75-71-8) in  New Zealand  White rabbits.    Final report  (RB80--EO).
NIEHS Contract N01-ES-2127.
                                     10-12

-------
Joshi,  S.B.,  M.C.  Dodge  and  J.J. Bufalini.   1982.   Reactivities of  selected
organic compounds and contamination effects.  Atmos.  Environ.   16:  1301-1310.


Joyner,  R.    196*1.    Chronic  toxicity of ethylene  oxide.    Arch.  Environ.
Health.  8: 700-710.


Jukes, T.H.   1964.   Choline.   Kirk-Othmer  Encyclopedia  of Chemical Technology,
2nd ed.  Interscience, New York.  5:  405.


Kadoi,  M.  and I.  Kataoka.  1968.  Purification of a  waste  liquor  from produc-
tion of an oxirane compound.  Japan Patent  74,   25,119,  5  pp.


Kauhanen,  K.    1977.   Stanford  Research  Institute  (unpublished).   (Cited  in
Federal Register,  43: 3912, 1978).


Klapproth,   E.M.    1976.    Ethylene  glycol.    Chemical  Economics  Handbook.
Stanford Research  Institute, Menlo Park,  California.


Kligerman,  A.D.,  G.L. Erenson,  M.E.  Phelps and J.L. Wilmer.   1983.   Sister-
chroraatid  exchange  induction  in peripheral blood  lymphocytes  of rats  exposed
to ethylene oxide by inhalation.   Mutat. Res.   120:  37-44.


Kolmark, H.G. and B.J.  Kilbey.    1968.  Kinetic  studies of mutation induction
by epoxides in Neurospora crassa.  Mol.  Gen. Genet.   101:  89-98.


Koskikallio, J. and E. Whalley.   1959.  Effect of pressure on the spontaneous
and the base-catalyzed hydrolysis of epoxides.   Can.  J.  Chem.   37:  783-787.


Krell,  K. ,  E.D. Jacobson  and  K. Selby.    1979.   Mutagenic effect  on L5178Y
mouse  lymphoma cells  by growth  in ethylene  oxide-sterilized  polycarbonate
flasks.  In Vitro.   15(5): 326-328.


Kurginski,   E.R.     1979.    Personal  Communication   to   C.   Glasgow,  U.S.
Environmental  Protection   Agency,  October  22.    Dow  Chemical  Co.,   Midland,
Michigan.


LaBorde, J.B.  and  C.A. Kimmel.   1980.   The  teratogenicity of ethylene oxide
administered intravenously to  mice.   Toxicol.  Appl. Pharmacol.  56(1):  16-22.
                                     10-13

-------
LaBorde,  J.B.,   C.A.  Kimmel,  C.  Jones-Price,  T.A.  Marks,  and  T.A.  Ledoux.
1982.    Teratogenic  evaluation of ethylene  chlorohydrin  (ECH)  in  mice and
rabbits.  Society of Toxicology Meeting,  Boston,  MA.


Lamty,  G., R.  Maleq, C.  Selve, A.  Sivade  and  J.  Wylde.   1975.  Mechanisms of
ring  opening  of oxirans by acids in  aqueous  and nonaqueous  solvents.   J.
Chem. Soc. Perkin II: 1119-1124.


Lambert, B. and A.  Lindblad.  1980.  Sister chromatid exchange  and  chromosome
aberrations  in   lymphocytes  of laboratory  personnel.   J.  Toxicol.  Environ.
Health.  6(5-6): 1237-1243.


Landels, S.P.   1976.  Fumigants and nematicides.   Chemical  Economics Handbook.
Stanford Research Institute,  Menlo Park,  California.


Lapkin,  M.  1965.  Epoxides.   Kirk-Othmer  Encyclopedia of Chemical Technology,
2nd ed.  Interscience, New York.  8:  288-289.


Laurent,  C.H.,  J.  Frederic and F. Marechal.   1982.   Etude des effects   cyto-
genetiques  d'intoxication a  1'oxyde  d'ethylene.    C.R.  Soc.   Biol.    176:
733-735.
Lee,  W.R.    Unpublished.    Final  report on  project  "Relation  of mutation
frequency  to dose  of ethylene  oxide  in germ  cells."    Submitted  to  South
Carolina Pesticides  Epidemic logic  Study Center,  Preventive Medicine  Division,
Department of Family Practice,  Medical University of  South  Carolina.   12 pp.


Lichtenstein, H.J.  and  G.H.  Twigg.    1948.   Trans. Faraday  Soc.   44:   905.
(Cited in Mabey and Mill, 1978).


Lichtenwalter,  G.D.  and  G.H.  Riesser.   1964.    Chlorohydrins.    Kirk-Othmer
Encyclopedia of Chemical Technology,  2nd ed.   5:  304-324.


Liepins,  R.,  F.  Mixon,  C. Hudak and  T.B. Parsons.   1977.   Industrial process
profiles for  environmental use:   Chapter 6.   The Industrial Organic  Chemicals
Industry.   Washington,  D. C.:   U.S.  Environmental Protection  Agency.   Avail.
NTIS PB281-478.   pp. 6-370 to 6-373.


Lindgren, D.  and  L.  Vincent.   1966.   In:   Advances  in  Pest Control  Research.
vol. 5, Metcalf, R.,  ed.  Interscience, New  York.  85 pp.
                                     10-14

-------
Lindgren,  D.  and K. Sulovska.   1969.   The mutagenic effect of  low concentra-
tions of ethylene  oxide  in  air (abstract  only).   Hereditas.  63: 460.   (Cited
in NIOSH,  1977).


Lindgren,  D. ,  L. Vincent and H. Krohne.   1954.   Relative effectiveness  of ten
fumigants  to adults of eight  species  of  stored-product  insects.    J.  Econ.
Entomol.   47: 923-926.


Lindgren,  D.L.,  W.B.  Sinclair and  L.E.  Vincent.    1968.   Residues  in raw and
processed  foods  resulting from post-harvest  insecticidal treatment.   Residue
Reviews.   21: 93-95.


Logan,  J.A.,  M.J.  Prather,  S.C. Wofsy  and M.B.  McElroy.   1981.  Tropospheric
chemistry:  A global perspective.  J. Geophys. Res.   86:  7210-7254.


Long,  F.A. and  J.G.  Pritchard.    1956.   Hydrolysis  of substituted  ethylene
oxides  in  H?0^°  solutions.  J. Araer. Chem. Soc.  78: 2663-2667.


Long, F.A., J.G. Pritchard  and F.E.  Stafford.   1957.   Entropies  of  activation,
and  mechanism for  the   acid-catalyzed  hydrolysis  of  ethylene   oxide  and  its
derivatives.  J. Amer.  Chem. Soc.   79:  2362-2364.


Loveless,  A.  and P.  Wheat ley.   1966.    In;    Genetic  and  Allied  Effects of
Alkylating  Agents.    Pennsylvania  State  University  Press,  University  Park,
Pennsylvania,  vol. 131.


Lynch,  D.W.,  T.R.  Lewis, W.J. Moorman, P.S.  Sabharwal  and J.R.  Burg.   1982a.
Chronic  inhalation  toxicity  of  ethylene  oxide   in  rats  and   monkeys.    A
Preliminary  Report.    Presented  at  the  21st  Annual  Society   of   Toxicology
Meeting,   Boston,   MA,   February   22-26,   1982.     National   Institute   for
Occupational Safety and  Health, Division  of Biomedical and Behavioral  Science,
4676 Columbia Parkway,  Cincinnati, OH.


Lynch,  D. , T. Lewis, W.  Moorman,  et al.   1982b.  Carcinogenic and  toxicologic
effects of inhaled ethylene oxide and propylene  oxide  in F344  rats.   National
Institute  for  Occupational  Safety   and  Health,  Washington,  D. C.   (Unpublished
draft.)


Lynch,  D.W. , T.R. Lewis, W.J. Moorman,  P.S. Sabharwal,  J.B. Lai and  J.R. Burg.
1983.   Toxic  and mutagenic  effects  of  ethylene  oxide  and  propylene  oxide on
spermatogenic function  in Cynomologus monkeys.  Society  of Toxicology  Meeting.
Las Vegas,  NV.
                                     10-15

-------
Mabey, W.  and T. Mill.   1978.   Critical review of hydrolysis  of  organic  com-
pounds in  water under  environmental  conditions.    J. Phys.  Chem. Ref.  Data.
7(2): 383-415.


Mackay,  D.  and P.J.  Leinonen.    1975.   Evaporation  of  low-solubility  contam-
inants from water bodies to atmosphere.  Environ.  Sci. Technol.   9: 1178-1180.


Mackey,  J.    1968.   Mutagenesis  in  volgare wheat.   Hereditas.   59:  505-517.
(Cited in NIOSH, 1977).


Mantel,  M.  and  M.A.  Schneiderman.     1975.    Estimating  "safe"  levels,   a
hazardous undertaking.  Cancer Res.  35:  1379-1386.
March,  J.    1977.   Advanced  Organic  Chemistry:   Reactions,  Mechanisms,  and
Structure.  2nd ed.  McGraw-Hill Book Co.,  New York.


Margosches, E.H.   May 13,  1983.   Memorandum to the  Record  U.S.  Environmental
Protection Agency.   EtO and  the  study of  spontaneous abortions published  in
the British Medical J. 20 November 1982.


Martis, L., R.  Kroes,  T.D.  Darby and E.F.  Woods.   1982.   Disposition kinetics
of ethylene oxide, ethylene glycol and  2-chlorethanol  in  the dog.   J. Toxicol.
Environ. Health.  10: 847-856.


Matter, B.E.  and  I. Jaeger.   1975.   The cytogenetic  basis of  dominant lethal
mutations  in mice.   Studies with  TEM,  EMS,  and  6-mercaptopurine.   Mutat.  Res.
33: 251-260.


MCA   (Manufacturing  Chemists  Association).   1971.   Chemical Safety Data Sheet
SD-38, Ethylene Oxide.  Manufacturing Chemists Association, Washington, D.C.


McDonald,  T.,  K.  Kasten, R.  Hervey, S. Gregg,  A.  Borgmann and  T.  Hurchison.
1977.  Acute ocular  toxicity  for normal and irritated  rabbit eyes and subacute
ocular  toxicity  for  ethylene  oxide,   ethylene  chlorohydrin,   and  ethylene
glycol.  Bull. Parenter. Drug Assoc.  31: 25-32.


McLaughlin, R.   1946.   Chemical burns  of the  human cornea.   Am. J.   Ophthal.
29: 1355-1362.
                                     10-16

-------
Mehra, P.N. and  S.K.  Mann.   1975.   Cytological effects of chemical mutagens on
Pterotheca  falconeri.   Part 2.    Monofunctional  alkylating  agents.   Nucleus
(Calcutta).  L7(3): 167-182.  (Cited in NIOSH, 1977).


Migliore,  L.,  A.M.  Rossi   and  N.  Loprieno.    1982.    Mutagenic  action  of
structurally   related  alkene  oxides  on   Schizosaccharomyces   pombe;     The
influence,  'in vitro',  of  mouse-liver  metabolizing system.  Mutat.  Res.   102:
425-437.
Mishmash, H.E. and  C.E. Meloan.   1972.   Indirect spectrophotometric determina-
tion of nanomole quantities of oxiranes.  Anal. Chem.  44: 835-836.


Morgan, R.W.,  K.W.  Claxton, B.J.  Divine,  S.D. Kaplan and V.B.  Harris.   1981.
Mortality among ethylene oxide workers.  J. Occup. Med.  23(11): 767-770.
Morris,  J.C.    1975.   Formation  of halogenated  organics  by  chlorination  of
water  supplies.    EPA   600/1-75-002,  U.S.  Environmental  Protection  Agency,
Washington, D.C.


Moutschen,  J.,  M.  Moutschen-Dahmen and  L. Ehrenberg.    1968.   Note on  the
chromosome breaking  activity  of ethylene oxide and ethyleneimine.   Hereditas.
60: 267-269.  (Cited in NIOSH, 1977).
NAS  (National  Academy  of Science).   1976.
National Academy of Sciences, Washington, DC.
Vapor-phase  organic  pollutants.
NFPA (National Fire Prevention Association).   1975.   Hazardous chemicals data.
Fire Protection Guide on Hazardous Materials.  Boston, Massachusetts.


NIOSH (National Institute for Occupational Safety  and Health).   1977.   Special
occupational  hazard  review  with  control  recommendations  for  the  use  of
ethylene  oxide as  a sterilant  in  medical  facilities.   Glaser,  Z.,  67 pp.,
Department of Health, Education, and Welfare Publication  No.   77-200,  National
Institute for Occupational Safety and Health, Cincinnati,  Ohio.
NIOSH  (National  Institute  for Occupational  Safety  and Heath).
Current  Intelligence  Bulletin  No.  35.   Ethylene  Oxide  (EtO):
Carcinogenicity,  May 22, 1981.
                   1981.   NIOSH
                    Evidence  of
NTP  (National  Toxicology Program).  1981.   National Toxicology Program,  1980.
Department of Health and Human Services.   U.S. Public Health Service.
                                     10-17

-------
Ong, E.   1948.   Chemistry and Uses  of  Insecticides.   Reinhold Publishing Co.,
New York.  p. 150.
Oxirane  Corporation.   Undated.   Propylene Oxide Technical  Bulletin,  Houston,
Texas.
Patty,  F.A.  (ed.).   1973-   Industrial  Hygiene and  Toxicology.    2nd  revised
edition, vol. 2.  Interscience, New York.  p. 1626.


Pellizzari,  E.D.,  J.E.  Bunch,  R.E.  Berkley and J.  McRae.    1976.   Collection
and analysis of trace organic  vapor pollutants in  ambient atmospheres.   The
performance  of  a  Tenax  GC  cartridge  sampler  for  hazardous  vapors.    Anal.
Letters.  9: 45-63.
Pero, R.W.,  T.  Bryngelsson, B. Widegren,  B.  Hogstedt and H.  Welinder.   1982.
A  reduced   capacity   for  unscheduled  DNA  synthesis  in   lymphocytes   from
individuals  exposed to  propylene  oxide and ethylene  oxide.   Mutat.  Res.   104:
193-200.


Pero, R.W.,  B.  Widegren,  B. Hogstedt and F.  Mitelraan.   1981.  In vivo and  in
vitro  ethylene  oxide  exposure  of  human  lymphocytes  assessed  by  chemical
stimulation of unscheduled DNA synthesis.  Mutat. Res.  83:  271-289.


Pessayre, D. and A. Trevoux.   1978.    Collapse  after  endovascular exploration:
The  role  of  ethylene  oxide  or  endotoxin?    LaNouvelle  Presse   Medicale.
39: 3558.


Pfeiffer,  E.  and H.  Dunkelberg.   1980.   Mutagenicity  of  ethylene  oxide  and
propylene oxide and of the glycols and halohydrins  formed from them  during  the
fumigation of foodstuffs.  Toxicol.   18:  115-118.


Pfeilsticker, K., G.  Fabricius and G.  Timme.    1975.   Simultaneous gas-chroma-
tographic determination  of  ethylene  oxide,  ethylene chlorohydrin,  and ethylene
glycol in grain.  Z. Lebensm-Unters.  Forsch.   158:  21-25.


Phillips, C.  and S. Kaye.   1949.  The sterilizing action  of gaseous ethylene
oxide.  Part I. Review.  Am. J. Hyg.   50: 270-279.


Poirier,  V.  and D.  Papadopoulo.    1982.   Chromosomal abberations  induced  by
ethylene  oxide  in  a  human  amniotic  cell  line  in  vitro.   Mutat.  Res.    104:
255-260.


                                     10-18

-------
Poothullil, J.,  A.  Shimizu, R. Day  and J.  Dolovich.   1975.   Anaphylaxis from
the product(s) of ethylene oxide gas.  Ann.  Intern. Med.  82: 58-60.


Pritchard,  J.G.  and F.A.  Long.    1956.   Hydrolysis of  ethylene  oxide deriva-
tives in deuterium oxide-water mixtures.  J. Amer. Chera. Soc.  78: 6008-6013.


Pritchard,  J.G.  and I.A. Siddiqui.   1973.   Activation  parameters and mechan-
isms  of the  acid-catalyzed hydrolysis of  epoxides.    J.  Chem.  Soc.  Perkin
II: 452-457.


Qazi,  A.H.  and  N.H.  Ketcham.   1977.   A  new method for  monitoring personal
exposure  to ethylene oxide  in the  occupational  environment.   Am.  Ind.  Hyg.
Assoc. J.   38: 635-647.


Rannug,  U.,  R.  Goethe  and C.A.  Wachtmeister.    1976.   The  mutagenicity  of
chloroacetaldehyde,  chloroethylene  oxide,   2-chloroethanol,  and  chloroacetic
acid,  conceivable  metabolites  of  vinyl  chloride.     Chem.  Biol.  Interact.
12: 251-263.


Rathburn,  R.E.  and D.Y. Tai.   1981.   Technique  for  determining the volatili-
zation  coefficients  of priority pollutants in streams.   Water Res.   15:  243-
250.   (Cited  in Conway et al., 1983).


Reyniers,  V.,  M.  Sacksteder and L.  Ashburn.   1964.   Multiple tumors in female
germ-free  inbred  albino mice  exposed  to bedding treated  with  ethylene oxide.
J. Natl. Cancer Inst.  32:  1045-1057.


Roberts,  J.,   L.  Allison,  P.  Prichett  and  K.  Riddle.   1943.   Preliminary
studies on  soil sterilization with ethylene oxide.  J. Bacteriol.  45: 40.


Romano,  S.J.  and  J.A.  Renner.   1975.   Comparison  of  analytical  methods for
residual ethylene oxide analysis.  J. Pharmacol. Sci.  64:  1412-1417.


Romano, S.J., J.A.  Renner  and  P.M. Leitner.  1973.  Gas chromatographic deter-
mination   of  residual   oxide   by   head   space  analysis.      Anal.   Chem.
45: 2327-2330.


Rowell,  R.M.  and D.I.  Gutzrner.   1975.  Chemical modification  of wood:   Reac-
tions of alkylene oxides with southern  yellow pine.  Wood Sci.  7: 240-246.
                                      10-19

-------
Rowell,  R.M.,  D.I. Gutzmer,  I.E.  Sadio  and  R.E.  Kinney.   1976.   Effects of
alkylene  oxide  treatments  on  dimensional   stability  of  wood.    Wood  Sci.
9: 51-54.
Rowlands,  D.G.    1971.    The  metabolism of  contact  insecticides in  stored
grains.  II. 1966-1969.  Residue Reviews.  34: 91-161.
Roy,  K.  and  M.K.  Jana.   1973.   Chemically  induced  chlorophyll  mutation  in
rice.  Indian Agric.  17(4): 301-308.  (Cited in NIOSH, 1977).


Royce,  A.  and  W.  Moore.   1955.   Occupational  dermatitis caused  by ethylene
oxide.  Br. J. Ind. Med.  12: 169.


Salinas, E.,  L.  Sasich,  D.H. Hall, R.M. Kennedy and H. Morriss.   1981.   Acute
ethylene oxide intoxication.  Drug Intel. Clinical Pharmacy.  15(5): 384-386.


Schnorr, T.M.    1982.    The  health  of  health care  workers:    A  proportionate
mortality study.  A thesis in epidemiology  presented  to  the graduate faculties
of  the  University  of Pennsylvania in  partial fulfillment  of  the requirements
for the degree of Master of Science.


Schultze, H.C.   1965.   Ethylene  Oxide.   Kirk-Othmer  Encyclopedia  of Chemical
Technology, 2nd ed.  Interscience, New York.  8:  523-558.


Scudamore,   K.A.  and  S.G.  Heuser.   1971.   Ethylene  oxide and its persistent
reaction products in wheat flour  and other  commodities:   Residues from fumiga-
tion or sterilization, and effects of processing.  Pest.  Sci.   2(2): 80-91.


Sega, G.A.,  R.B.  Gumming,  J.G.  Owens,  C.Y.  Horton  and L.R. Lewis.    1981.
Alkylation  pattern  in developing mouse  sperm,  sperm DNA,  and  protamine  after
inhalation of ethylene-oxide.  Environ.  Mutagen.   3(3): 371.


Seizinger,   D.E.  and B.  Dimitriades.   1972.   Oxygenates  in exhaust  from simple
hydrocarbon fuels.   J. Air Pollut. Contr. Assoc.   22(1):  47-51.


Sexton,  R.  and  E.  Henson.   1949.  Dermatological injuries by ethylene  oxide.
J. Ind.  Hyg. Toxicol.   31: 297-300.


Sexton,   R.  and  E.  Henson.    1950.    Experimental ethylene  oxide  human  skin
injuries.  Arch. Ind.  Hyg. Occup. Health.  2:  549-564.


                                     10-20

-------
Shenderova, R.I., G.K. Alsopova,  I.E.  Shaklinazarov and V.A.  Linetskii.  1972.
Purification of  waste waters from  simultaneous  production of  propylene oxide
and styrene oxide.  Khirn. Prom. (Moscow).  48: 746-7^7.


Shupack, J.L., S.R. Andersen and  S.J.  Romano.   1981.   Human  skin  reactions to
ethylene oxide.  J. Lab. Clin.  Med.  98(5): 723-729.


Sickles, J.E., II, R.S.  Wright,  C.R. Sutcliff, A.L. Blackard  and  D.P. Dayton.
1980.   Smog chamber  studies of  the reactivity of  volatile organic compounds.
Proc. Ann. Meet. Air Pollut. Contr. Assoc. Paper 80-501, 16 pp.


Sittig, M.   1962.   Organic Chemical Processes.   The Noyes Press,  Pearl River,
New Jersey.


Sittig, M.   1965.   Acrylic Acid  and  Esters.  Noyes  Development  Corporation,
Park Ridge, New Jersey.


Skeehan,  A.    1959.    A  portable  sterilizer  for the  application  of  ethylene
oxide gas in ophthalmology.  Am.  J. Ophthalraol.  47: 86.


Smith,  H.H.  and  T.A.  Lotfy.   1954.  Comparative effects  of  certain chemicals
on Tradescantia chromosomes as observed  at pollen tube mitosis.  Am.  J.   Bot.
41: 589-593.  (Cited in NIOSH,  1977).


Smyth,  H.,  J. Seaton and  L.  Fischer.   1941.    Single dose  toxicity  of  some
glycols and their derivatives.   J. Ind. Hyg. Toxicol.  23: 259-268.


Snellings, W., J.  Pringle,  et  al.   1979.   Teratology and reproduction studies
with rats exposed to  10,  33, or  100 ppm of ethylene oxide.   Abstract, Society
of Toxicology, 18th meeting, New Orleans, March 1979.


Snellings, W.M., C.S.  Weil  and R.R. Maronpot.   1981.   Final  report,  ethylene
oxide,  two  years  inhalation  study.    Submitted  to   the  U.S.  Environmental
Protection Agency by Bushy Run Research Center, Pittsburg, PA.


Snellings, W.M., J.P.  Zelenak  and C.S. Weil.   1982a.   Effects on  reproduction
in  Fischer  344  rats   exposed   to  ethylene   oxide  by  inhalation  for   one
generation.  Toxicol.  Appl. Pharmacol.   63: 382-388.
                                     10-21

-------
Snellings,  W.M.,  R.R.  Maronpot,   J.P.   Zelenak   and  C.P.  Laffoon.    1982b.
Teratology study in  Fischer  344 rats exposed to ethylene  oxide by inhalation.
Toxicol. Appl. Pharmacol.  64: 476-481.


Snellings, W.M., C.S.  Weil and  R.R.  Maronpot.   1984a.   A subchronic inhalation
study on the toxicologic potential of  ethylene  oxide in  B6C3F1  mice.   Toxicol.
Appl. Pharmacol.  76: 510-518.


Snellings, W.M.,  C.S.  Weil and  R.R.  Marenpot.   1984b.  A two-year inhalation
study of  the carcinogenic potential of  ethylene  oxide in  Fischer 344  rats.
Toxicol. Appl. Pharmacol.  75: 105-117.


Spasovski, M.,  V.  Khristeva, K. Pernov,  V.  Kirkov, T.  Dryanovska,  Z.  Panova,
G. Bobev,  N.  Gincheva  and S. Ivanova.   1980.   Health state of the  workers in
the   production  of   ethylene  and   ethylene  oxide.     Khig.   Zdraveopaz.
23(1): 41-47.
Spencer,  E.F.,  Jr.    1971.    Pollution  control  in  the  chemical  industry,
Chapter 14.    In;    Industrial  Pollution  Control  Handbook,  H.F.  Lund,  ed.
McGraw-Hill Book Co., New York.


SRI International  (Stanford Research Institute  International).   1977.   A Study
of  Industrial  Data  on  Candidate  Chemicals   for  Testing.     Contract  No.
68-01-4109, U.S.  Environmental Protection Agency,  Office of  Toxic Substances,
Washington, D.C.


SRI International  (Stanford Research  Institute International).    198la.   1981
Directory  of   Chemical  Producers:    United  States of  America.    Menlo  Park,
California:  SRI International, pp. 594-595.


SRI International  (Stanford Research Institute  International).    198lb.   1981
Directory  of  Chemical  Producers:   United States  of America.   Supplement II.
Menlo Park, California:   SRI International, p. 51.


Star,  E.G.   1980.  Mutagenic  and  cytotoxic effect of  ethylene  oxide  on human
cell cultures.  Zentralbl. Bakteriol.  170(5-6): 548-556.
Star, E.G.,  J.  Caselitz and T. Loening.   1980.   The effect  of  ethylene oxide
and  chemical disinfectant  residues upon  the larynx  and  tracheal mucosa  of
rabbits.  Zentralbl. Bakteriol.  170(5-6):  539-547.
                                     10-22

-------
Steinberg,  S.    1977.    Analytical  methods  for  determining ethylene  oxide  in
plastics.    Saf.  Health  Plastics,  Natl.   Tech.  Conf.  Soc.  Plastics  Eng.,
pp. 141-116.


Stijve, T.,  R.  Kalsbach and G. Eyring.   1976.   Determination of occurrence of
ethylene  chlorohydrin residues  in foods fumigated with  ethylene  oxide.   Mitt.
Geb. Lebensmittelunters. Hyg.  67(4): 403-428.


Strekalova,  E.    1971.    Mutagenic action of  ethylene  oxide  on  mammals.
Toksikol. Nov. Prom. Khim. Veshchestv.  12: 72-78.


Sulovska, K., D. Lindgren, G. Eriksson  and L. Ehrenberg.   1969.   The mutagenic
effect   of   low  concentrations  of  ethylene   oxide  in  air.     Hereditas.
62(1/2): 264-266.   (Cited in NIOSH,  1977).


Sykes,  G.    1964.    Disinfection  and  Sterilization.    J.P.  Lippincott  Co.,
Philadelphia, Pennsylvania, pp. 202-227.


Systems Application, Inc.   1982.   Human exposure to atmospheric concentrations
of  selected chemicals.    Appendix   A-14  Ethylene Oxide.   U.S.  Environmental
Protection Agency,  Research Triangle Park, NC.


Tan,  E.L.,  R.B.  Gumming and A.W. Hsie.   1981.    Mutagenicity  and cytotoxicity
of ethylene  oxide in the CHO/HGPRT system.  Environ. Mutagen.  3: 683-686.


Tanooka,  H.   1979.   Application  of  Bacillus subtilis spores  in  the detection
of gas mutagens:  A case of ethylene oxide.  Mutat. Res.  64(6): 433-435.


Taylor,  D.G.   1977a.   NIOSH  Manual of Analytical  Methods,  2nd  ed.,  vol. 2.
U.S.  Department  of Health,  Education,  and Welfare,  National  Institute  for
Occupational Safety and Health, Cincinnati, Ohio.


Taylor,  D.G.   1977b.    NIOSH  Manual of  Analytical  Methods,   2nd  ed.,  vol. 3.
U.S.  Department  of Health,  Education, and Welfare,  National  Institute  for
Occupational Safety and Health, Cincinnati,  Ohio.


Taylor,  J.    1977.    Dermatologic  hazards from   ethylene  oxide.    Curtis.
19: 189-192.
                                      10-23

-------
Taylor,  G.    1979.    Mutagenicity  testing  of  ethylene  oxide
Communication  to division directors, NIOSH, January 8, 1979.
                                         (unpublished).
Thiess,  A.    1963.   Considerations of  health damage through  the  influence of
ethylene oxide.  Arch. Toxikol.  20: 127-140.
Thiess,  A.M.,  H. Schwegler,  I.  Fleig and  W.G.  Stocker.   1981.   Mutagenicity
study of workers exposed  to  alkylene  oxides (ethylene oxide - propylene oxide)
and derivatives.  J. Occup. Med.  23(5): 343-347.


Thiess,  A.M.,  R. Frentzel-Beyme,  R.  Link and  W.G.  Stocker.   1982.   Mortality
study on employees  exposed to alkylene oxides (ethylene oxide/propylene oxide)
and their  derivatives.   In;   Proceedings  of the  International  Symposium  on
Prevention of Occupational Cancer, Helsinki, Finland,  p. 249-259.


TSCA (Toxic Substances  Control Act).   1980.   TSCA  Chemical Assessment Series.
Chemical  Screening:     Initial  Evaluations  of   Substantial  Risk  Notices,
Section 8(e).  U.S.  Environmental Protection  Agency,  Office of Pesticides and
Toxic Substances, Washington, D.C.


Tyler,   T.R.  and  J.A.   McKelvey.    1980.   Dose-dependent  disposition of  C
labeled  ethylene  oxide  in   rats.     Carnegie-Mellon  Institute  of  Research,
Pittsburgh, PA.
U.S. EPA.   1976.   Frequency of Organic  Compounds  Identified in Water.
600/4-76-, U.S. Environmental Protection Agency, Washington,  D.C.
                                                  EPA-
U.S.  EPA.      1980.
contaminants: Epoxides.
EPA-560/11-80-005.
Investigation  of   selected   potential  environmental
U.S. Environmental  Protection  Agency,  Washington,  DC.
U.S.  EPA.    1984.    Performance  evaluation of  the  Sciex TAGA   600  mobile
laboratory  for  ambient  air  monitoring.   Summary report.   Advanced  Analysis
Techniques  Branch,   Environmental Monitoring  Div.,  Environmental  Monitoring
Systems Laboratory,  RTP, NC.
U.S. EPA.   1984.   Porposed guidelines  for  carcinogen risk assessment.
Reg. 49(227):  46294-46301.
                                                  Fed.
                                     10-24

-------
USITC  (U.S.  International  Trade  Commission).    1976.    Synthetic  Organic
Chemicals:  United States  Production  and Sales, 197^.   USITC  Publication 776,
p. 203.


USITC  (U.S.  International  Trade  Commission).     1977a.    Synthetic  Organic
Chemicals:  United States  Production  and Sales, 1975.   USITC  Publication 804,
p. 198.


USITC  (U.S.  International  Trade  Commission).     1977b.    Synthetic  Organic
Chemicals:  United States  Production  and Sales, 1976.   USITC  Publication 833,
p. 303.


USITC  (U.S.  International  Trade  Commission).    1978.    Synthetic  Organic
Chemicals:  United States  Production  and Sales, 1977.   USITC  Publication 920,
p. 361.


USITC  (U.S.  International  Trade  Commission).    1979.    Synthetic  Organic
Chemicals:  United States  Production  and Sales,  1978.   USITC Publication 1001,
p. 313.


USITC  (U.S.  International  Trade  Commission).    1980.    Synthetic  Organic
Chemicals:  United States  Production  and Sales,  1979.   USITC Publication 1099,
p. 269.


USITC  (U.S.  International  Trade  Commission).    1981.    Synthetic  Organic
Chemicals:  United States  Production  and Sales,  1980.   USITC Publication 1183,
p. 265.


USTC  (U.S.  Tariff Commission).   1971*.   Synthetic Organic  Chemicals:   United
States Production and Sales, 1972.  TC Publication 681, p. 206.


USTC  (U.S.  Tariff Commission).   1975.   Synthetic Organic  Chemicals:   United
States Production and Sales, 1973.  TC Publication 728, p. 205.


Van  Duuren,  B.,  L.  Orris  and  N.  Nelson.   1965.   Carcinogenicity of epoxides,
lactones, and peroxy compounds, II.  J. Natl. Cancer Inst.  35: 707-717.


Veith,  G.D.,  K.J.  Macek,   S.R.  Petrocelli  and J.  Carroll.   1979.   Federal
Register  15926 (March 15).
                                     10-25

-------
 Verrett,  M.J.   1974.   Memorandum  from M.J.  Verrett,  Reproduction  Physiology
 Branch, FDA,  to  Tibor  Balazs,  Bureau  of Drug,  FDA.  Investigation of the  toxic
 and  teratogenic  effects of  2-chloroethanol  to  the  developing chicken  embryo.


 Verschueren,  K.   1977.   Handbook  of Environmental Data  on Organic  Chemicals.
 Van  Nostrand  Reinhold, NY.   p.  322.


 Virtanen,  P.O.I, and T.  Kuokkanen.   1973.   Effect of pressure on the rates of
 acid-catalyzed   methanolysis   of   3-substituted   1,2-epoxypropanes.      Suom.
 Kemistilehti  B.   46(10): 267-270.


 von  Oettingen,  W.   1939.   Ethylene oxide,   In:   Supplement  to Occupation and
 Health:    Encyclopedia of  Hygiene,  Pathology,  and Social Welfare.   Geneva,
 International Labor  Office.
Waite,  C.,  F.  Patty  and W.  Yant.   1930.   Acute responses of  guinea pigs to
vapors  of some  new commercial organic  compounds.   IV.   Ethylene oxide.   Publ.
Health  Rep.   45:  1832-1843.


Wallace,  T.J.  and R.J.  Gritter.    1963.   Free  radical  chemistry  of  cyclic
ethers.     IV.   Free   radical  rearrangement   of  epoxides.     Tetrahedron.
19: 657-665.
Walpole, A.   1958.   Carcinogenic action of alkylating agents.  Ann. N.Y. Acad.
Sci.  68: 750-761.


Watson, W.A.F.   1966.   Further  evidence of an essential difference between the
genetical  effects  of mono-  and bifunctional  alkylation  agents.    Mutat.  Res.
3: 455-457.
Weast, R.C.   1972.   Handbook  of Chemistry and Physics.  53rd ed.  The Chemical
Rubber Co., Cleveland, Ohio.  p. C-287.


Wesley, F.,  B.  Rourke and  0.  Darbishire.   1965.  The  formation of persistent
toxic  chlorohydrins  in foodstuffs by  fumigation with ethylene  oxide  and with
propylene oxide.  J. Food Sci.  30: 1037.


Whelton,   R.,  H.  Phaff,  E. Mruk  and  C. Fisher.    1946.    Control of  micro-
biological food spoilage by fumigation with epoxides.  Food Ind.  18: 23-25.
                                     10-26

-------
Wolman, S.R.   1979.   Mutational  consequences  of exposure  to  ethylene oxide.
J. Environ. Pathol. Toxicol.  2: 1289-1303.


Woodward,   G.  and  M.  Woodward.    1971.   Toxicity of residuals  from  ethylene
oxide  gas  sterilization.    Proc.  of  the  Health  Ind.  Assoc.  Tech.  Symp.,
Washington, DC.  pp. HJO-161.


Yager,  J.L.W.    1982.   Sister  chromatid  exchange as a biological  monitor for
workplace  exposure:    A  comparative  study  of  exposure-response  to  ethylene
oxide.   Berkeley,  CA:   The  University  of  California.   Northern California
Occupational  Health  Center  Report No. NCOHC 82-B-2.   March  1982.    p.  163.
Ph.D. Dissertation.
Yager,  J.L.W.  and  D. Benz.    1982.    Sister  chromatid  exchanges  induced  in
rabbit  lymphocytes  by  ethylene  oxide  after  inhalation  exposure.    Environ.
Mutagen.  4: 121-134.


Yager, J.W., C.J. Hines  and R.C. Spear.   1983.   Exposure to ethylene oxide at
work  increases sister  chromatid exchanges  in  human peripheral  lymphocytes.
Science.  219:  1221-1223.


Yakabova, Z.N., N.A.  Shamove,  F.A. Miftakhove  and L.F.   Shilova.   1976.   Gyn-
ecological  disorders  in  workers  engaged in ethylene  oxide  production.   Kazan.
Med. Zh.  57(6): 558-560.
Young, H.  and  R.  Busbey.   1935.   References  to the use  of  ethylene oxide for
pest  control.   U.S. Department  of Agriculture, Bureau  of Entomology.   Plant
Quarantine, April 1935.


Zamlauski,  M.J.  and G.J.  Cohen.   1976.   The  effects  of aortic  infusion  of
ethylene   oxide  renal   function   in   the   rat.     Toxicol.   Appl.  Pharmacol.
38(2): 283-295.
                                     10-27

-------