?/EPA
                United States
                Environmental Protection
                Agency
                Office of Health and
                Environmental Assessment
                Washington DC 20460
EPA-600/8-84-009A
April 1984
External Review Draft
                Research and Development
Health  Assessment
Document for
Ethylene Oxide
 Review
 Draft
 (Do Not
 Cite or Quote)
                               NOTICE

                This document is a preliminary draft. It has not been formally
                released by EPA and should not at this stage be construed to
                represent Agency policy. It is being circulated for comment on its
                technical accuracy and policy implications.

-------
                                            EPA-600/8-84-009A
(Do Not                                                April 1984
Cite or Quote)                                External Review Draft
     Health  Assessment Document
              for  Ethylene Oxide
                            NOTICE
 This document isa preliminary draft. It has not been formally released by the U.S. Environmental
 Protection Agency and should not at this stage be construed to represent Agency policy. It
 being circulated for comment on its technical accuracy and policy implications.
               U.S. ENVIRONMENTAL PROTECTION AGENCY
                  Office of Research and Development
               Office of Health and Environmental Assessment
               Environmental Criteria and Assessment Office
               Research Triangle Park, North Carolina 27711

-------
                                  DISCLAIMER



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




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




not constitute endorsement or recommendation for use.
                                       ii

-------
                                    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

-------
     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

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

-------
     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.
        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.
        Dharm  V. Singh, D.V.M., Ph.D.»
        Todd W. Thorslund, Sc.D.

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

the  preparation  of  sections  on  mutagenicity, 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 mutagencity of EDC.
        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)

     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

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
                                      vi

-------
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.
Pels 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

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
                              vii

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

-------
                               TABLE OF CONTENTS
LIST OF TABLES	      xii

LIST OF FIGURES	      xvi

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  TORTIONAL 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, Volume %	      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
          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
                                      ix

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

                                                                           Page

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   Ethanolamines	      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	       6-2
     6.3  ETHYLENE OXIDE FATE IN SOIL	       6-4
     6.4  ETHYLENE OXIDE FATE IN THE ATMOSPHERE	       6-4
     6.5  DEGRADATION IN COMMODITIES AND MANUFACTURED PRODUCTS	       6-7
     6.6  BIOACCUMULATION IN AQUATIC ORGANISMS	      6-10
     6.7  SUMMARY	      6-10

-------
                           TABLE OF CONTENTS (cont.)
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	      9-5

          9.2.1   Effects in Humans	      9-5
          9.2.2   Effects in Animals	     9-14
          9.2.3   Summary of Toxicity	     9-23

     9.3  TERATOGENICITY AND REPRODUCTIVE TOXICITY	     9-24
     9.4  MUTAGENICITY	     9-45
     9.5  CARCINOGENICITY	     9-93

          9.5.1   Animal Studies	     9-93
          9.5.2   Epidemiologic Studies	    9-116
          9.5.3   Quantitative Estimation	    9-131
          9.5.4   Summary	    9-158
          9.5.5   Conclusions	    9-160

     Appendix 9A 	       A1
     Appendix 9B 	       B1

10.   REFERENCES	     10-1
                                      xi

-------
                                LIST OF TABLES


Table                                                                      Page

3-1     Manufacturers'  Specifications for Ethylene Oxide	       3-6

3-2     Typical Reactions of Ethylene Oxide	       3-9

3-3     Hydrolysis Kinetics of Ethylene Oxide	      3-12

3-4     Specific Rates of Reaction of Anions and Lewis Bases with
          Ethylene Oxide	      3-13

4-1     Breakthrough and Safe Sampling Volumes for Propylene Oxide
          with Several Sorbents	       4-2

5-1     Ethylene Oxide Production	       5-2

5-2     Ethylene Oxide Producers, Plant Sites, Capacities,  Processes,
          and Technology	       5-3

5-3     Ranges of Reaction Systems Variables in the Direct
          Air-Oxidation of Ethylene Oxide	       5-9

5-4     Ranges of Reaction Systems Variables in the Direct
          Oxygen-Oxidation of Ethylene Oxide	      5-11

5-5     Users and Use Sites of Ethylene Oxide	      5-14

5-6     Typical Vent Gas Composition for Both Air- and Oxygen-Based
          Ethylene Oxide Plants	      5-21

8-1     Acute Aquatic Toxicity of Ethylene Oxide	       8-3

9-1     Acute Toxicity of Ethylene Oxide	       9-2

9-2     Subchronic Toxicity of Ethylene Oxide	      9-17

9-3     Summary of Studies	      9-39

9-4     Summary of Mutagenicity Testing of EtO: Gene Mutations in
          Bacteria	      9-47

9-5     Summary of Mutagenicity Testing of EtO: Gene Mutation Tests
          in Lower Plants	      9-52

9-6     Summary of Mutagenicity Testing of EtO: Mutation Tests in
          Higher Plants	      9-54
                                      xii

-------
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


9-20



Summary of Mutagenicity Testing of EtO: Gene Mutation Tests

Summary of Mutagenicity Testing of EtO: Mammalian Cells

Summary of Mutagenicity Testing of EtO: Dominant Lethal
Tests 	
Summary of Mutagenicity Testing of EtO: Heritable

Summary of Mutagenicity Testing of EtO: Chromosome

Summary of Mutagenicity Testing of EtO: Micronucleus
Tests 	
Summary of Mutagenicity Testing of EtO: Chromosome

Summary of Mutagenicity Testing of EtO: SCE Formation

Summary of Mutagenicity Testing of EtO: SCE Formation

Summary of Mutagenicity Testing of EtO: Unscheduled
DNA Synthesis 	
Design Summary for Carcinogenicity Testing of EtO by
Intragastric Administration to Sprague-Dawley Rats 	
Tumor Induction by Intragastric Administration of EtO in

Cumulative Percentages of Male Fischer 34U Rats that Died
or were Sacrificed in a Moribund Condition After Exposure
to EtO Vapor 	
Cumulative Percentages of Male Fischer 3*M Rats that Died
or were Sacrificed in a Moribund Condition After Exposure
to EtO Vapor 	
Page

9-57

9-61

9-64

9-66

9-67

9-68

9-69

9-80

9-82

9-83

9-96

9-96


9-99


9-100
          xiii

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

Table                                                                      Page

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 EtO Vapor	      9-101

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 EtO Vapor	      9-102

9-23    Summary of Selected Tumor Incidence Comparisons for Male
          and Female Fischer 344 Rats Exposed to EtO for
          Two Years	      9-104

9-24    EtO 2-Year Vapor Inhalation Study:   24-Month Final
          Sacrificed Frequency of Exposure-Related Neoplasms
          for 110- to 116-Week-Old Fischer 344 Rats	      9-105

9-25    EtO 2-Year Vapor Inhalation Study:   Frequency of Exposure-
          Related Neoplasms at 24-Month Final Sacrificed and in
          Fischer 344 Rats Dying Spontaneously or Euthanized
          When Moribund	      9-107

9-26    EtO 2-Year Vapor Inhalation Study:   Frequency of Primary
          Brain Neoplasms in Fischer 344 Rats	      9-111

9-27    EtO 2-Year Vapor Inhalation Study:   Frequency of Primary
          Brain Neoplasms Types in Fischer 344 Rats	      9-112

9-28    Leukemia Incidence in Male Fischer 344 Rats Exposed
          to EtO for 2 Years	      9-114

9-29    Incidence of Neoplastic Lesions in Male Fischer 344
          Rats Exposed to EtO for 2 Years	      9-114

9-30    Comparison of Observed Numbers of Cancer Deaths in
          Base-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	      9-126

9-31    Related Risks of Death From Cancer in the Alkylene Oxide
          Cohort as Compared With the Styrene Cohort, By Age	      9-128
                                       xiv

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


Table                                                                      Page

9-32    Bushy Run EtO Inhalation Study in Fischer 311 Rats.
          Incidence of Peritoneal Mesothelioma and Brain Glioma
          in Males, and Mononuclear Cell Leukemia and Brain
          Glioma in Females by Dose Among Survivors to First
          Tumor	    9-111

9-33    NIOSH EtO Inhalation Study in Male Fischer 311 Rats.
          Incidence of Peritoneal Mesothelioma and Brain Glioma
          by Dose, Among Total Examined.  Estimates of 95%
          Upper-Limit Risk Based on Human Equivalent Dose	    9-11?

9-31    Relative Carcinogenic Potencies Among 51 Chemicals
          Evaluated by the Carcinogen Assessment Group as
          Suspect Human Carcinogens	    9-155
                                      xv

-------
                                LIST OF FIGURES


Figure

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

9-1     Mutagenic Response of Salmonella Typhimurium Strain
          TA1535 Exposed to Ethylene Oxide ............................     9-^9
9-2     Mutagenic Response of CHO Cells to EtO ........................     9-62
9-3     Percentages of Male and Female Fischer 3W Rats with
          Histologically Confirmed Mononuclear Cell Leukemia
          at 21-Month Sacrificed ......................................    9-110

9-4     Histogram Representing the Frequency Distribution of the
          Potency Indices of 5*J Suspect Carcinogens Evaluated
          by the Carcinogen Assessment Group ..........................    9-154
                                      xvi

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








     The  largest single use  of ethylene  oxide is as  an intermediate  in the



synthesis of ethylene glycol.   However,  small amounts of this epoxide are used



as a  sterilant  or pesticide in  commodities,  Pharmaceuticals,  medical devices,



tobacco,  and  other items.   Although  this  use is only  a  small fraction of the



total  epoxide  consumption,  it  represents  a  considerable potential  for human



exposure.



     The  pharmacokinetics of  ethylene oxide have not been studied extensively.



There were no studies  found in  the literature  searched concerning the absorp-



tion  of this chemical.   However,  the toxicity data  suggests that absorption



occurs  via  the  respiratory  and  gastrointestinal   tracts.     Two  studies



(Ehrenberg  et al.f  197^;  Appelgren  et  al.f  1977) have  shown  that  ethylene



oxide  is  widely distributed in  various  tissues (liver,  kidney,  lung, testes,



brain,  spleen,  and  intestinal  mucosa)  following inhalation exposure and intra-



venous administration.



     Acute exposure to ethylene oxide in  humans  has  resulted  in symptoms of



headache,  vomiting,  dyspnea,   and  diarrhea.    Similar  symptoms  have  been



reported  by Blackwood  and  Erskine (1938),  Cobis  (1977),  and Anonymous (191?).



Thiess  (1963)  reported that  high concentrations of  ethylene oxide  for brief



periods produced bronchitis,  pulmonary edema,  and  emphysema.   Salinas  et al.



(1981)  reported neurological symptoms  (convulsive  movements)  in a woman who



was exposed for a brief period to 500 ppm ethylene oxide.



     Anaphylactic  reactions have  been  observed  in   patients using  ethylene



oxide sterilized  plastic tubing for hemodialysis (Poothullll  et  al.,  1975) or



                                      1-1

-------
cardiac catheterization  (Pessayre and Trevoux,  1978).   Hirose et  al.  (1953)



and Clarke et al.  (1966)  reported hemolysis in  patients following  the  use of



ethylene oxide sterilized plastic tubings.



     The acute toxic effects of ethylene  oxide  in  laboratory animals have been



extensively  reviewed.     Exposure  to  concentrated  ethylene  oxide  produces



systemic poisoning,  with symptoms  of salivation,  nausea,  vomiting, diarrhea,



convulsion, and death  (Hine and  Rowe, 1973).  Symptoms  of  respiratory irrita-



tion,   incoordination,  and cardiac  arrhythmia  have also been  reported (Sexton



and Benson, 19^9).



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



documented.  Both  Jensen (1977)  and  Gross et  al.  (1979) have  reported neuro-



toxicity in humans following inhalation exposure to ethylene oxide.



     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).   Hackett  et  al.  (1982) reported  that  rats,  but not  rabbits,



exposed to 150  ppm ethylene oxide  administered by inhalation  displayed signs



of maternal  toxicity and toxicity to the  developing conceptus.   Laborde and



Kimmel  (1980)   and  Kimmel   (1982)  reported that  150  mg/kg  ethylene  oxide



administered intravenously  to mice and  rabbits  caused maternal  toxicity and



developmental  toxicity.    Laborde   et   al.   (1982)   reported  that  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  developmental effects



without significant toxicity when administered ECH at  60  mg/kg intravenously.



In a  one-generational  study,  Snellings  et  al. (1982)  reported that  100 ppm



                                      1-2

-------
ethylene oxide  administered  by inhalation  caused  severe adverse effects   in-



cluding a  higher  incidence of  infertility,  longer gestational periods,  a de-



crease  in  the  number  of  pups  born,  and  a  decrease  in  the  number  of



implantation  sites.   Hollingsworth  et al.  (1956)  observed  testicular  degen-



eration in hamsters and  rats  inhaling  204 to 357  ppm ethylene  oxide.   In a



more recent study, Lynch  et  al.  (1983) reported adverse  effects  on sperm con-



centration motility but  not  morphology in  Cynomologuous  monkeys  exposed to 50



and 100 ppm ethylene oxide by  inhalation.   In humans,  Hemminki et  al.   (1982)



conducted  an   epidemiologic  study  of  nursing personnel  exposed  to  ethylene



oxide and  found an association  between ethylene oxide exposure and spontaneous



abortion.



     In conclusion,  the  available  information  indicates that ethylene oxide



produces developmental  toxicity  in  laboratory  animals when  conducted  at  or



near maternally toxic doses.   Ethylene oxide produces adverse reproductive ef-



fects and  testicular toxicity at  levels lower than those which produce general



toxicity.    Finally, ethylene  oxide  is reported  to be  associated  with spon-



taneous abortions  in  nursing personnel  exposed to ethylene  oxide  in hospital



occupational settings.



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



fungi, higher  plants,  Drosophila, and  cultured mammalian cells  in tests con-



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



is therefore a direct-acting mutagen.   Strong positive responses  were found in



bacteria (10-fold  to  18-fold  increase over  negative  controls),  higher plants



(33-fold   increase),  and  mammalian   cells  in  culture  (2-fold  to  20-fold



increases).    Less strong,  but  clearly   positive,   reponses  were  found  in



Drosophila (2-fold to 3-fold increases).  Based on  these positive  findings in




                                      1-3

-------
different test systems  in  a  wide range of organisms,  ethylene  oxide is judged




to be capable of causing gene mutations.




     Ethylene oxide  has also been shown  to  be clastogenic, in  that it causes




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 exhange (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 mu-




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




table  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 mu-




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




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




     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.




While  these  studies  have  shortcomings  and  are  not  definitive,  they  do,




                                      1-4

-------
nevertheless,  constitute  limited, bordering  on  inadequate,  evidence for human



carcinogenicity  under  the International  Agency  for Research on  Cancer  (IARC)




classification scheme for the evaluation  of carcinogenic risk to humans.




     Positive  results  for the  carcinogencity of ethylene oxide  have been ob-




tained  by  subcutaneous  injection in  mice and by intragastric administration in




rats.   In  addition,  two  long-term  chronic  inhalation studies  in  rats have




shown statistically significant responses for leukemia, brain tumors and peri-




toneal  mesothelioma.  The animal evidence is sufficient under the IARC classi-




fication  system for  experimental animals.   Considering  both  the  animal and




human evidence for carcinogenicity,  especially leukemias  in both  humans and




rats, the  Carcinogen Assessment Group  (CAG)  classifies ethylene oxide as being




probably   carcinogenic  to  humans   and,   therefore,   belonging  in  the  IARC




Group 2A.  Because of the  very  few human cancers,  however, this classification




borders on a Group 2B  classification.    Assuming  that ethylene oxide  is car-




cinogenic  in humans,  upper-limit potency estimates have  been calculated using




both the animal and human data base.




     Estimates  of  carcinogenic  relative potencies of ethylene oxide  in rats




and humans suggest that humans  may be  more sensitive  than animals  to the car-




cinogenic  effects  of  ethylene  oxide. Supporting this  suggestion are data in-




dicating that  humans  have greater sensitivity  than rats  to  chromosome abnor-




malities  induced  by  ethylene   oxide  exposure.    The  unit  risk estimate  of




lifetime cancer  risk  resulting  from continuous  exposure to  air  that contains




an ethylene  oxide concentration  of 1 ng/m3 for humans is 3.6 x 10~^, while the




95$ upper-limit estimate for animals based  on rat  studies is 1.0 x  1(H*.   The
                                      1-5

-------
potency index  for  ethylene oxide, which  is  based on both the  unit risk value



and molecular  weight,  is in  the  lower part  of the  the  third quartile  of 54



suspect carcinogens evaluated by the CAG.
                                      1-6

-------
                               2.   INTRODUCTION



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




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




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




Standards  to   support   decision-making  regarding   possible   regulations   of




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




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




information needs of many  government agencies and  private groups  that  may be




involved in decision-making activities related to ethylene oxide.




     In  tne  development  of  the  assessment  document,  existing  scientific




literature has  been surveyed in detail.   Key studies have been evaluated  and




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




related characteristics are qualitatively identified.




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




the likelihood  for  its  exposure  to  humans, and  the possible  effect on man  and




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




integrated  into  a  format  designed  as  the basis  for performing  risk assess-




ments.   When  appropriate,  the  authors  of  the  document have  attempted  to




identify gaps in current knowledge that limit risk evaluation capabilities.
                                      2-1

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

     Synonyms:  1,2-epoxyethane
                ethylene oxide
                oxirane

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

     Structural formula:

               H     H
               I     I
            H—C	C—H
                \/
                  0

     Molecular formula:
3.3.  TORTIONAL ANGLES AND BOND DISTANCES (Hirose, 1974)'

                                        Tortional
                                         Angles

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




3.4.1.   Description.   Ethylene  oxide  is  a  colorless,  flammable  gas  which




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




et al.f 1980;  Hawley, 1981).








3.4.2.  Molecular Weight.




     44.05 (Weast, 1980)








3.4.3.  Melting Point.




     -111°C (Weast, 1980)








3.'I.1*.  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.



     d]JJ:  0.8824 (Weast, 1980)








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




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

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

     1.3597 (Weast, 1980)



3.1-9.  Vapor Pressure (Cawse et al., 1980).

      Temperature               Vapor Pressure

          °C                  kPa           Torr
-40
-30
-20
-10
0
10
20
30
MO
50
60
70
80
90
100
8.35
15.05
25.73
12.00
65.82
99.51
115.8
207.7
288.1
391.7
521.2
681.0
875.1
1108.7
1385.1
62.6
112.9
193.0
315.0
193.7
716.6
1093
1558
2163
2938
3909
5108
6566
8315
10390
3-1.10.  Aqueous Solubility3 (Cawse et al., 1980).
Pressure
kPa
20
27
10
53
67
80
93
101
torr
150
202.5
300.0
397.5
502.5
600.0
697.5
757.5
5°C
15
60
105
162
210
NT
NT
NT
Temperature
10°C
33
16
76
120
178
291
NT
NT
20*C
20
29
19
71
101
131
170
195
      Solubility in al vapor/mi 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
HO
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
0.0
-0.9
-1.6 (eutectic)
5.6
8.9
10.4
11.1 (max)
10.4
9.3
7.8
6.0
3.7
0.0
-112.5
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

-------
3.4.14.  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.^.16.  Heat of Combustion at 25°C (Cawse et al., 1980).

     5.17 kJ/mol

     1.24 kCal/mol



3.4.17.  Log Octanol/Water Partition Coefficient.

     -0.30 (Hansch and Leo, 1979)



3.4.18.  Ultraviolet Spectroscopic Data (Weast, 1980).

     X = 169 nm (gas)

     log e = 3.58

     \2 - 171 nm (gas)

     log e = 3.57



3.5.  PHYSICAL PROPERTIES AND DESCRIPTION OF COMMERCIAL ETHYLENE OXIDE

     The physical properties  and description of  commercial  ethylene  oxide are

described in Table 3-1.


                                      3-5

-------
                                                  TABLE 3-1

                             Manufacturers' Specifications for Ethylene Oxide3'

Purity, wt % min
Hater, wt % max
Aldehydes, as acetaldehyde, wt % max
Acidity, as acetic acid, wt % max
C02, wt % max
Total Cl as Cl~, wt % max
Nonvolatile residue, g/100 mi, 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.

CPresently, 0.005 g/.100 ml in Dow ethylene oxide (Kurginski, Dow Chemical Co.)
NA = Not available; wt = weight; max = maximum; min = minimum

-------
     Commercial grade ethylene oxide has  a  purity of >99.9$.  Specific impuri-




ties 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 previously described.








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  to 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  to 14,000,




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




                                      3-7

-------
ranging from  90,000  to 4 x  10,  are formed by coordinate  anionic polymeriza-

tion.   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  primary  catalyst  for  this  process  is  rust,  and  no

inhibitor has been found.




3.6.4.  Other Reactions.  Table 3-2  lists  a number of other reactions ethylene

oxide undergoes that are representative of its chemistry.




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  paralleled hydrolysis;  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  spontaneous  (I),  acid-catalyzed  (II),  and  alkali-catalyzed hydrolyses

(III):
                       k
     (I)  C H^O + HO ——>HOCH CH OH
                                      3-8

-------
                                   TABLE 3-2



                      Typical Reactions of Ethylene Oxide
1.   Crown Ethers



     n H0C - CHD Cataly3t> cyclic
        2 \   /  s


           0
2.   Hydrolysis
     H,C - CH~ + H00 - > HOCH0CH_OH


      2\ /

         0
3.   Reaction with Alcohols



     H0C - CH0 + ROH 	> R-0-CH-CH.OH      -> R_04CH_CH,.,0} H
      2 y  .  2                  22    oxide          2  2  n


         0
 .   Reaction with Organic Acids and Acid Anhydrides
     RCOOCOR + H.C - CH0 - > RCOOCH_CH_OH         > RC004CH0CH.O^ H
                ti v   /  2            22    oxide           2  2  n


                   0
5.   Reaction with Ammonia and Primary and Secondary Amines
     R-NH_ + H_C - CH0 - > R-NH-CH0CH0OH         > R-NH4CH0CH00} H
         d    d \   I  e.             d  d.    oxide           2  2  n


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



     H2C - CH2 + RSH 	> RSCH2CH2OH



         0
                                      3-9

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

      /—^              H2°       /^>               -          /^>
      (ON + H,C - CH0 —	>    < O N-CH_-CH0OH + OH  	>    (O N + HOCH.CH-OH
      \ J     2 «.   /  2           \—i    2   2                  \—i       2  2

                 0
8.   With Phenols
              H2C - CH2

                  0
9.   With Hydrogen Cyanide

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

-------
                       k
     (II'  C2HijO + H20* -=-*  HOCH2CH2OH

                       kB
    (III   C2H40 + H20 —=-*  HOCH?CH2OH
                      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, 19*18)
           Log kg =  9-312 - 75.3/RT (Lichtenstein  and  Twigg,  1948^


     Epoxides  can   also  react  with nucleophiles (anions or  Lewis  bases)  by
pathways  which  parallel hydrolysis  (reaction with  water or  hydroxide).    The
chemistry,  although similar  to hydrolysis, is more  complex.   The epoxide  ring
can be cleaved by spontaneous  reaction or  by acid-catalyzed reaction:
                                             +OH
1
•?\
,c^0 + x"H
1
1
C
1
1
k -COH
h H20 — JU. _^
"l
1
kx -COH
1
Table  3-4  summarizes specific rate  constants for reactions  of ethylene  oxide
with various  anions.   The consensus  agrees  that the  spontaneous reaction  is
S..2, but  disagreement exists  whether acid catalyzed epoxide  ring opening  is
A1-like or  A2-like  (Long et  al.,  1957;  Lamaty et  al., 1975;  Pritchard and
                                      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-1S-1)
5.3«a
NR
9.3°
NR
NR
NR
9
NR
10. Od
16. 9e
Specific Rate Constant
kfl x 10? kB x 101*
(S-1) (M-1S-1)
3.613 NR
M.2b 0.65t>
6.75° 1.Qd
5.62f»8 NR
6.17f)h m
e.Glfii NR
5.56J 1.1
5.8k NR
NR NR
NR NR
aBronsted et al., 1929
 Lichtenstein and Twigg,
cEastham and Latremouille, 1952
 Pritchard and Long, 1956
eLong et al., 1957
 Conway et al., 1983
gRiver water pH 7.H
h
 Sterile river water pH 7.4
1Sterile distilled water
•'Long and Pritchard, 1956
kKoskikallio and Whalley, 1959
NR = Not reported                     3-12

-------
                                   TABLE 3-H



   Spec:fie  Rates of  Reaction  of  Anions  and Lewis  Bases  with  Ethylene Oxide
Lewis Base
or Anion Temperature K
Ci~ 293
298
298
300
Br" 293
298
Pyridine 291
10S a
(JL/mole - sec)
NR
NR
0.3056
NR
NR
NR
200 ( water) d
102kx
(8,2 /mole2 - sec)

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




discussion of the mechanism is beyond the scope of this review.




     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:
                         rate

                    determining
                                                    -C5H5N
     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 T "A°H00   " "B'OH'^epox                        (1)
                               3




          -dC

                   = 'k,r,-C,. + k . C. .Cu n+)C
           dt
where C. . , k . , and k  . refer  to  the  concentration and specific rate constants




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




contributions,  as given in Equation 3:
       dC
         epox
         dt

-------
     The  relative  importance of chemical hydrolysis and reaction with  chloride

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

butions   (chlorohydrin  to  glycol  ratios)  were  estimated  for  freshwater and

marine  water  (NaCl  concentration  of  1%  or  0.513M).   The  following  specific

rate constants  from  Tables  3-3  and 3- ^ were  utilized:
                         k     0.661 x  10~6  s"1
                                    _3   _i   _i
                         k     9 x  10 ° M   s
                                     U    1    1
                         kg    1 x  10   M"   s

                         k     0.305 x  10~6  M"1  s"1
                          y              POO
                               3.6-7 x 10~* VT" sT*
Estimates  were calculated for  pH 5, 7,  and  9, which  is  approximately the  pH

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

hydrin/glvcol  ratios  (for sea water  reactions)  are summarized  below:


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

                            PH       5      7       9

  Freshwater                       256     291     291
  Saline Solution
    0.85$  (physiological)                  273
    1*                             2*40     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 31 1, 265, and  224

hours,  respectively.   The half-lives  for river water  (pH 7.1),  sterile river

water  (pH  7.1), and  sterile  distilled  water  reported by  these authors were

311, 310,  and  293  hours, respectively.   The  chlorohydrin/glycol ratio experi-

mentally determined  by  Conway  et al.  (1983)  was 0.11  and 0.23 for  1  and  3%

saline solutions.

                                      3-15

-------
     From the data presented, some understanding of  the  fate  of ethylene oxide


in biological fluids  can  be determined.   The  hydrolysis half-life  in  physio-


logical saline (0.85$) is 273 hours  or  11.H  days.   This long a half-life would


clearly allow 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 (e.g., RS~,  PhNH?) present in biological  systems,  are known to be


more  nucleophilic  than  pyridine,   and  may  react  with  ethylene  oxide  in


biological systems much more rapidly than either water or chloride.


     Hydrolysis  or hydrolysis- type  reactions  are also  the most significant


industrial  reactions  of  ethylene  oxide.    Ethylene  glycol  is the  hydrolysis


product;  higher  glycols  (diethylene,   triethylene,  and  polyethylene  glycols)


and glycol  ethers  are the  result  of  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  as described  above.   For example,


reaction  of  e^hylene  oxide  and  nonylphenol  yields  nonylphenoxypolyethoxy-


ethanol, a non-^onic,  surface-active agent (Blackford, 1976a).


       0

    / \
n CH
    2
     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:
                                      3-16

-------
                     0


                nH2C - CH2 	> H2N 4 CH2CH20 ^n H
where n  is  typically  1  to 4.   Choline is prepared  by  reacting trimethylamine



with ethyl ene oxide (Jukes, 1964):
              ) N + H2C - CH2
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  LS  of  particular importance  in  determining its  fate  in the atmosphere.



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



radical.



     Only  one  reported  study  of  the  reaction  of  the hydroxyl  radical with



ethylene  oxide  was  found  in the  available  literature.   Fritz et  al.   (1982)



reported  the   results  of  a   study  utilizing   a  laser  photolysis/resonance



fluorescence  (LPRF)  unit  designed  to study  the reactions of OH radicals with



anthropogenic  pollutants.   Hydroxyl  radical was generated  by HNO_  photolysis



and radical  concentrations  were measured  by the system.   The authors  studied



the  reaction over  three  temperatures,  297,  377, and  435K, at  10  torr  (Ar).



The following relations were reported:
and
     k(297K) = (8'° - 1'6) x  10~111 cm3/molec-
     k,_,<, =  M.I + 0.4) x  10~   exp  (-1460/T) cnr/molec.  S.
       (. i)        ~
                                      3-17

-------
where the error limits are  the  >90?  confidence  limit (3o).   The mechanism they




reported  involves  hydrogen  abstraction,  followed  by  ring  opening,  reaction



with  oxygen,  NO,  and  finally  decomposition to  carbon monoxide  and  formalde-




hyde.   Ring  opening  may take  place either  before  O* addition  or after  NO



reaction.
                                      3-18

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








U.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




amh;ent air monitoring studies and  found  them to be inappropriate for ethylene




oxide.   Although  most  of  'he 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 (very similar to  ethylene  oxide).    Table




4-1 compares  the  breakthrough volumes  for  several  sorbents.   The  effect of




humidity on  the breakthrough  volume was  tested for  Tenax GC.   Breakthrough




volume  increased  from 4.0 to  4.5 fc/g when  humidity  was increased  from  41 to




92%.  Pellizzari et al. (1976) also  examined  the effect of storage time on the




recovery  of diepoxybutane  (300  ng) loaded  onto Tenax  GC cartridges.   They




desorbed it  thermally and analyzed  it by  GC.   When analysis  was  immediate,




recovery was  100$.   After  the  loaded cartridge was  stored  for 1  week, the




recovery dropped  to  76$.    Combined transport  (6  days)  and storage  yielded




recoveries of  75  and 64$ 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

-------
                                    TABLE U-1

            Breakthrough and  Safe  Sampling Volumes  for  Propylene  Oxide
                              with Several Sorbents
 Sorbent
                           Breakthrough Volume
                             fc/g (sorbent)a
aPellizzari et al., 1976

 Brown and Purnell, 1979
 Mesh size
Safe Sampling Volume
      U/g)b
PEL Carbon
PCB Carbon

SAL9190
MI808
Tenax GC (35/60)c
Porapak Q (100/120)
Chromosorb 101 (60/80)
Chromosorb 102 (60/80)
Chromosorb 101 (60/80)
36
in

10
21
1
il
ij
8
>36
9

10
10
6
1
1
1
2
9
                                     4-2

-------
 ethylene  oxide behave  similarly.   Brown  and Purnell  (1979)  have noted  that,

 under  the  conditions  of the  test  (5 to 600 mi/minute flow rate, <100 ppm  vapor

 concentration,  <20°C, and <95$ relative  humidity),  the breakthrough volume  is


 not  <50$  of the  retention volume, and  a  safe  sampling  volume is  50$ of the

 retention  volume.    Thus,  it  appears  that  if propylene  oxide  behavior  is


 analogous  *o  ethylene oxide, then  ethylene  oxide will  likely be detected  only

 rarely using  solid adsorbants.  even if it is present,  since  the great  majority


 of monitoring  studies use air samples  larger  than  the  breakthrough volume for

 ethylene oxide.


     The  National  Institute for  Occupational  Safety  and Health  'NIOStH has

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

 iQ77'.   The -  procedure calls for the sampling of 5 I  of  air through  glass

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

 tubes  mounted  in  series are used;  the front and back-up  tubes contain  400 and


 °00 mg,  respectively,  of  charcoal.     The  front   and  back-up   sections are

 individually  measured  for epoxide.   If  the back-up portion  contains  >25%  of


 the  epoxide,   the  analysis  is  not  considered  valid.    The  method   suggests

 desorbing  the  epoxide  with  carbon disulfide.    The  required  solvent amount  is


 2.0 m&  for ethylene  oxide.    Aliquots of  the  desorbed  solutions are   then

 analyzed by GC with flame ionization  detection.   NIOSH (1977) conducted  tests

 on the analytical parameters.   Ethylene oxide  was  examined  at concentrations

 from 11  to 176 mg/m3  (23  to 98 ppm);  precision (CV_,)  was 0.103  (or  standard

                      •3
deviation  of  9.3  mg/nr), and accuracy  was  0.9$  lower  than  the  "true" value.


NIOSH  (1977)  recommended  sample   concentrations  of  20  to 270 mg/m   for   this

method for industrial hygiene monitoring.

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




comparison of three methods  for  sampling ethylene  oxide in surgical equipment.




The  study was  administered through  the  Z79  Subcommittee  on  Ethyiene  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  freezing them  in a




cold trap.   The  sample is  then vaporized and  an aliquot  is  removed with a




vacuum  syringe  for GC  analysis.  Romano  and  Renner  (1975"> reported  that  the




method  requires  greater time and equipment  than  the  other techniques and is




subject  to  errors from  equipment  leaks.   Its advantages  are  that  ii-  is  the




most sensitive, and since the  sample injected into  the GC is a  vapor,  column




life is  long.  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,




problems from impurities  in the  solvent  and  extraction of other compounds  from




the plastics, the  reduced  lifetime of  columns  because of these impurities,  and




low sensitivity.   In  headspace analysis,  the  sample is  placed  into a   vial



which  is equipped with  a  septum for gas  withdrawal by syringe.   The epoxide




partitions  between the  sample  and  headspace  gases.   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.




     Romano  et  al. (1973)  reported  that  the  headspace technique  has  a  lower




limit  of 0.1 ppm  and  that the technique  can be  automated.  Romano and Renner




(1975) evaluated results  for the three  methods at  six laboratories by analysis

-------
of variance.   Among  overall  methods, there  were no  significant  differences;




howeve", slight differences between laboratories were detected.




     Ben-Yehoshua et al.  (197D extracted  fruit  pulp by blending it with 50 mi




of  anal v-*.oal   grade  acetone   for   30 seconds,   filtering  the  homogenate  to




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




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




accurate to +5$.



     Scudamore and  Heuser (197D  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 mJ,




solvent'p 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




mJ, of methanol using  continuous agitation for 24  hours.   Recovery  of  ethylene




oxide (25 ppm) was 73$ and standard deviation  (with  GC analysis) was 1.70 ppm.




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




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




Brown  (1970)   could  separate   ethylene   oxide  and   its  degradation   product,




ethylene chlorohydrin.   Samples were extracted  with p-xylene  (3 days  contact)




or  co-sweep  distillation.    The  three  column  system  consisted  of:    I.




Fluorisil, II.  acid-celite,  and  III. Fluorisil.   The p_-xylene solution  was




passed  through  Column I; ethylene  chlorohydrin  remained  fixed  in  the column




and ethylene  oxide  passed  through.   The  ethylene  oxide   solution  was passed




through  the  acid-celite column which converted  it  to  ethylene chlorohydrin.




Column  III  retained the  ethylene  chlorohydrin,  which  was  subsequently eluted




with  petroleum  ether.    The  sample was  concentrated  with   a  Kuderna-Danish




                                      1-5

-------
apparatus, and  then analyzed by  GC.   Brown (1970)  reported values as  low as




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



described.








4.2.  ANALYSIS




     Thus  far,  GC  analysis for ethylene  oxide has only  used  flame-ionization




detection  or  thermal  conductivity detection.    Neither detection system  is




selective, so the  epoxides must  be  separated  from all  interferences,  and the




choice of  analytical  column depends on  potential interferences.   Columns for




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




102 (Taylor,  i977a,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;




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.




     Other  analytical  methods   include  various  wet   chemical  techniques.




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




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




Mishmash  and  Meloan  ( 1972)  reported  perhaps the  most recent  use  of  this




approach.    Butylene oxide  was  hydrolyzed to  its glycol, then  the glycol was




oxidized with periodic acid.    Residual  oxidant  was  analyzed  by  adding Cdlp-




starch,   and   then  measuring  the   starch-I_  complex  concentration  at  590 nm.




They claimed a detection limit in the nmole range.

-------
                         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




nameplate  capacity  is  rated  at  520  million  pounds/year  (Anonymous,  198la).




Dow will add  MOO million pounds/year capacity onto  its Plaquemine,  Louisiana,




facility during  the  fourth quarter of 1983.   Union Carbide is building  a MOO




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 or not  to  move  the  former's idle  Guayanilla, Puerto  Rico,  facility




(rated at  300 million  pounds/year) to Beaumont, Texas,  to  be  operated jointly




by both.








5.1.3.  Production Methods and Processes.








     5.1.3.1.  INTRODUCTION — The majority  of information in this  section was




obtained from Cawse et al. (1980).
                                      5-1

-------
                                   TABLE 5-1

                                                  a b
                         Ethylene Oxide Production '
Year
WPa
1981
1980
1979
1978
1977
1976
1975
1974
1973
1972
Production
5200
4937
5220
5665
5012
4364
4184
4467
3893
4167
3962
(2359)
( 2240)
( 2368)
(2570)
(2273)
(1980)
(1898)
( 2026)
(1766)
(1890)
(1797)
Sales0
NA
NA
531
560
525
549
439
409
457
501
454


(241)
(254)
(238)
(249)
(199)
(186)
(207)
(227)
(206)
aSource:  USTC, 1974,  1975;  USITC,  1976,  1977a,  1977b,  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
Company
Annual
Location Capacity13
Process
Oxidant
Technology
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.

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

Olin Corp., Olin Chems. Group

PPG Industries, Inc.
  Chems. Group, Chem. Div.-U.S.
Geismar, LA         181 (216)

Lake Charles, LA    225 (101)
Clear Lake, TX

Freeport,  TX
Plaquemine, LA
Longview, TX

Bayport, TX
    (191)

260d (117)
450® (203)
oxygen

oxygen


oxygen

air
air
195 (88)      oxygen

520 (23I»)f     NA
Shell

Shell


Shell

Dow
Dow



Shell

NA
Joliet, IL          230  (10*0     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.
Cheras. and Plastics Div.
Union Carbide Carbie, Inc., Subsid.
Location
Geisraar, 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
 Sources:  Anonymous, 198la; SRI International,  198la,b;  Cawse, 1980

 Capacities are expressed in millions of pounds;  capacities in millions of kilograms are in parentheses.

°Plant 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.
f-f
 Expansion of 400 million pounds/year ( 180 million kg/year)  is due in the fourth quarter of 1983.
f
 Under construction

NA = Not available

-------
      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


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


 drawback of the  direct  oxidation process is  the  loss  of  =25 to  30% of the


 ethylene to carbon  dioxide and water.





      5.1.3.?.   DIRECT  OXIDATION  — The overall  reaction for direct oxidation


 can  be  represented  as  follows:
          H9C=CH,  + TiO- -* H0C - CH0
            22      2         2 \  /   2

                                   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  of the


process 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


minor components such as  formaldehyde and acetaldehyde.


     Ethylene conversion  to ethylene oxide  per pass  in the main  reactors is 20


to 50$.   Oxidation inhibitors (e.g., vinyl  chloride,  ethylene dichloride) are


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


                                      5-5

-------
       Main
      reactor
 Main
absorber
 Purge
reactor
 Purge
absorber
                                                                  Desorber
                                                             Stripper      Refiner
 StMm
Coolant
                                                                                                      Ethylene
                                                                                                       oxide
                                             Figure 5-1

                 Schematic  for air-based  ethylene oxidation (Schultze,  1965)

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




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



      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
                                      5-7

-------
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.




     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  mole % purity.




     The  specific  conditions   used  to  operate   ethylene   oxide  plants  are




proprietary.   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




substantially  pure  oxygen,  the  recycle  gas is  almost entirely unconverted




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




is still produced in the oxygen system, and  since  this has a negative effect




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




                                      5-8

-------
                                   TABLE 5-3

               Ranges of Reaction System Variables in the Direct
                        Air-Oxidation of Ethylene  Oxide^
          Variable                                           Range
ethylene, mole %                                        2-10

oxygen, mole %                                          M-8

carbon dioxide, mole %                                  5-10

ethane, mole %                                          0-1.0

temperature, °C                                         220-277

pressure, MPa (psi)                                     1-3

space velocity , h~                                     2000-^500

pressure drop, kPa (torr)                               H1-152 (308-1140)

conversion, %                                           20-65

selectivity or yield (mole basis, %)                    63-75
 Source:  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

-------
carbon dioxile removal unit, an argon vent  is  also  required.   Argon is a major




impurity in  oxygen  and can build  up  to the extent of 30  to 40  mole %.   In




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




plant is much smaller than from an air-based plant.




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




usually contains  high hydrocarbon  concentrations.   In such  cases,  the  purge




stream can  be  used  readily 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 medium capacity  units  (<50,000 t/year),




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




ation  facility.   For medium to  large plants (75,000  to  150,000  t/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  to the produc-




tion 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




                                      5-10

-------
                                   TABLE 5-4

                   Ranges  of Reaction  System Variables  in  the
                   Direct  Oxygen-Oxidation  of Ethylene  Oxidea
         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

-------
chlorohydrin  process  is attractive  commercially only  when a  good supply  of




captive low-cost chlorine and lime or caustic soda  is available.   Also,  satis-




factory markets or disposal facilities are needed for the  by-products  produced




(Schultze, 1965).



     The  chlorohydrin  process  starts  by  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  to 150  pounds/1000  pounds  ethylene  oxide)




and bis(2-chloroethyl)ether (=70 to  90 pounds/1000  pounds  ethylene oxide), are




formed  during the chlorohydrin  formation;  acetaldehyde  (5 to  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 HOCH2CH2C1 + Ca(OH)2 	»• 2 CH2C




                                          0
     Ethylene chlorohydrin is formed  in  the  lower section of a reaction tower.



Gases are  separated  from the dilute  chlorohydrin solution in  the  top section




and  the  vent gases from  the  condensing  apparatus pass in  series  to water and




caustic  scrubbers,  where  residual chlorine  and  HC1 gas  are  removed  before




recycling  the unreacted  ethylene.   The  aqueous  chlorohydrin  solution is mixed




with a 10$ solution of milk  of  lime at the inlet  to  the  hydrolyzer  (Schultze,




1965).



     The crude  ethylene  oxide product from the hydrolyzer contains about 77.5$




ethylene  oxide,  10$  water,  12$  chlorinated  organic  compounds   (principally




                                      5-12

-------
1,2-dichloroethane   and   bis(2-chloroethyl)ether),   and   0.5%   acetaldehyde

together with  small  amounts of hydrocarbon  gases.   This  crude  ethylene oxide

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:


                                         Pounds5    Percent of Total

     Ethylene glycol                    3-2 x  1066          62?
     Nonionic surface-active agents     0.62 x 10,-          12$
     Glycol ethers                      0.31 x 10?           6%
     Ethanolamines                      0.26 x 10fi           5%
     Miscellaneous applications         0.78 x 10           15$
     (higher glycols, urethane
       polyols, sterilant,  fumigant,
       export)

     Source:  Anonymous,  1981a

      Based on 1982 production  estimates of 5200 x 10  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
Ul
1
         Company
                           Location
               Ethylene  Glycol  Diethylene   Ethanol-   Triethylene  Polyethylene
                Glycol   Ethers    Glycol     amine       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  4-
Clear Lake, TX    4-
Freeport, TX      +
Plaquemine, LA    +
Midland, MI
Longview, TX      +
Morris, IL        +
Brandenburg, KY   4-
Beaumont, TX      4-
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  ovide, -  indicates non-users  of  ethylene oxide

-------
(Blackford,  19?6b).   Ethylene glycol  is mainly used  for  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  mercaptans, 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
                                     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,  which include  ethylene  glycol monomethyl,  monoethyl,




and  monobutyl  ethers;  diethylene  and  triethylene  monoethyl,  monomethyl,  and




monobutyl  ethers  (Cogswell,   1980).   Solvent  applications dominate the  many




uses of glycol ethers.   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  the  three ethanol-




amines is  dependent  upon the ratio of reactants used.  About 25 to 30% of all




ethanolamines are used for  soaps  and detergents, 5  to  20% for  gas condition-




ing,  10$  by  the metal industry,  8% for textiles,  5 to 15? for toilet goods,




and the remainder in varied applications (Blackford, 1976b).








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




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




group  goes  into  production  of polyether polyols  for flexible  polyurethane




foams.   In 1978,  about  100  million pounds (45  million kg) of  ethylene oxide




were consumed in these polyols (Cogswell, 1980).




     Approximately  17  million pounds of ethylene  oxide  are  used  annually to




make the medicinals, choline and choline chloride (Cogswell, 1980).




     Approximately  10 million pounds of ethylene  oxide are  used  annually in




the manufacture of  hydroxyethyl  starch,  which is a semi-synthetic gum used in




                                     5-16

-------
 textile  sizing  and adhesives  (Cogswell,  1980).    Hydroxyethyl  cellulose  is




 produced  by  reacting cellulose with  ethylene  oxide.   About  25  million pounds




 ( 11  million kg)  of ethylene  oxide  are used  annually  to make  these adhesive



 additives (Cogswell, 1980).




     Arylethanolaraines are  made  by reacting ethylene  oxide with either aniline




 or  aniline  derivatives.   It  is  estimated  that  3  million pounds  (1.4 million




 kg)  of  ethylene  oxide  are  used  annually  for arylethanolamines  (Cogswell,




 1980).  They are used as intermediates for monoazo dyestuffs.




     Acetal  copolymer  resins  are  produced  by catalytically  copolymerizing




 1,3,5-trioxane  with a  cyclic ether having  at  least two  adjacent  carbon atoms




 (e.g.,  ethylene  oxide).    Ethylene  oxide  consumption  for  these   resins  is




 believed  to  have amounted  to -2 to  3 million  pounds/year (0.9  to 1.4 million



 kg) from  1977 to 1978 (Cogswell,  1980).




     Like nonionic surface-active  agents,  ethylene oxide  is used  to produce




 ethoxylated  cationic  surface-active  agents.    Several million  pounds  of




 ethylene  oxide  are  used  annually  to produce these  cationic agents,  such  as




 ethoxylated  (coconut oil  alkyl)  amine,  ethoxylated  (tallow  alkyl)  amine,  and




 various ethoxylated fatty acid amino amides (Blackford,  1976).




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




 food and  cosmetic  sterilant,  and  in  hospital  sterilization  (Gilmour,  1978).




 In 1975, an estimated 0.1  million pounds of ethylene oxide were  used for fumi-




gant purposes  (Landels,  1976).   By  contrast, Dow  Chemical  (Kurginski,  1979)




has estimated  that 0.2$ of production (=10 million  pounds/year)  of ethylene




oxide is used as a fumigant; however,  the exact amount  is not available.
                                     5-17

-------
5.2.7.   Discontinued  Uses of Epoxides.   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:








     H2C - CH2 + HCN 	>• HOCH.CH2CN 	»• CH2 = CHCN + H20








In  1956,  American  Cyanamid Company  closed  down  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.   No other




significant discontinued uses of ethylene oxide are known.








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 to 30$ (by weight) of the epoxide to the wood.








5.2.9.   Alternatives  to Uses for Ethylene  Oxide.  More than 99% of the United




States'  production  of  ethylene  oxide  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  production  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  capacity  represents =25$ of the  total  industry  ethylene glycol




capacity.




     As  far  as the  other  compounds  synthesized from ethylene  oxide are con-




cerned,  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 equal  10 million




pounds;  Kurginski,  1979).   Since there  are  many commercial  furaigants  avail-




able,  it  seems possible  that many of its  fumigant uses might be replaced by an




alternative fumigant.








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  (N2,  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

-------
                                                                              p
     The  analogous  vent  stream  from  an  oxygen-based  system  is  about  10


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.


     Approximate amounts of vented reaction stream  have  been estimated.   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  =4  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  about  2 x 10   pounds (9.09 x 10  kg)  (SAI,  1982).


     Process waters for ethylene  oxide manufacture  and use appear to  be minor


problems with respect to waste treatment.   The major aqueous waste is draw-off


from separator  bottoms (Liepins et  al.,  1977).  The process  water  is  recycled


in  its  manufacture  and its primary use 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  specific details were  provided  on treat-


ment methods.   The  waste water will  contain  high  BOD, but  inorganic  composi-


tion and refractory organics  appear minimal problems with  ethylene oxide manu-


facture or ethylene glycol production  from ethylene  oxide  (Sittig,  1962, 1965;


Spencer,  1971).    Conventional  water  treatment   (including  filtration  and


                                      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
C0_ 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

-------
flocculation) with  a biological treatment  appears sufficient  (Spencer,  1971;




Shenderova et al., 1972).




     There is no solid waste associated with ethylene oxide manufacture.








5.3«2.  Handling, Transport, and Storage.   Ethylene oxide could  be  emitted to




the  atmosphere   as  the  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 following paragraphs discuss potential  releases of  epoxides without making




any attempt to establish relative importance.




     Bulk  shipments  of ethylene oxide  are  commonly  made by  railroad  freight




tanker;  the  sizes  of the  tankers are  commonly  10,000  and   20,000  gallons.




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




transfer  utilizes  nitrogen pressurization  to  =50 psi  or pumping.    Faulty




equipment  or over-pressurization  can cause  epoxide emissions.   Small  amounts




spilled during handling could also release some ethylene oxide.




                                     5-22

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




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




scope  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 would  rupture.    No  information was  available  to  predict




how often  the minor release accidents do,  in  fact,  occur or on 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  preferrably be stored in  an  area




detached from  the  plant site  and  storage  tanks should  be  diked.   Ethylene




oxide  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  in the  environment  of  ethylene oxide  is  probably the




combustion  of  hydrocarbon fuels.   Hughes  et  al.  (1959)   utilized  gas-liquid




partition  chromatography to separate and  identify  oxygenated derivatives  of




hydrocarbons that were  found in the  combustion products of hydrocarbon fuels.




                                      5-23

-------
Among  the  oxygenated  combustion  products identified  were ethylene  oxide  and




propylene oxide.   Barnard  and  Lee  (1972)  identified  these  compounds  in  the



oxygenated  combustion  products  from  n-pentane   combustion.    Seizinger  and



Dimitriades  (1972)  identified  ethylene  oxide  as  a  component of  automobile




exhaust.  The  fuels  used were simple hydrocarbons,  not gasoline,  but all were




components of  gasoline; no  lead  was present.    Stationary  sources  of hydro-



carbon  combustion may also emit  large  quantities of  these  compounds into the




environment.




     Ethylene  oxide  has been identified in tobacco  smoke  (Binder and Lindner,




1972;  Binder,  197*0.    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 concentra-




tion  of unfumigated  tobacco was 0.02  \ig/mSL,  while  fumigated tobacco  had  a




concentration  of 0.05  (ig/mS,  and extensively  fumigated tobacco had a concentra-




tion  of 0.30  ng/m£.    Binder  determined the ethylene oxide  content  of smoke




from unfumigated tobacco as 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  (HAS, 1976).




     Water  disinfection  has  the  potential  to  convert  olefins  to  epoxides.




Olefin  conversion during chlorination would  proceed by the same  route as for




chlorohydrination  production of   the  epoxide.     However,  this  process would




require conversion of ethylene  to the  chlorohydrin  (Morris,  1975; Carlson and




Caple,  1977).  Since  ethylene is  very volatile, this process seems unlikely.

-------
 5.14. SUMMARY




     This  section discusses production, uses, and emissions of  ethylene  oxide.




 Ethylene  oxide  is  produced  virtually  exclusively  by direct  oxidation  using




 either  air or  oxygen.    Its 1981  production volume was  H937 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 major  source.   Total air emissions  from




 production have   been estimated  to be  around 2 million  pounds based  on  1978




 production (50 x 10  pounds).   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.




     Greater  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; when  used  as  such, ethylene  oxide  emissions  might  be signif-




 icant.   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  on  transport was not sufficient to develop a




 definite  description of 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  follow parallel pathways  with  respect to products.   Its degrada-




 tion  in  water, soil,  commodities,  and manufactured  products proceeds through




 ionic  reactions.   Its  degradation  in  the atmosphere  has  not  been well charac-




 terized   with   respect   to   processes  or   products.     Available   information




 indicates that  it  is very reactive in photochemical  smog cycle  reactions.   No




 information was available on  whether  ionic reactions (e.g.,  with  water vapor




 or water  within aerosols) significantly contributed  to its degradation  in the



atmosphere.
                                      6-1

-------
6.2. ETHYLENE OXIDE FATE IN WATER




     Ethylene oxide  will degrade  in  water by  hydrolysis  and  related  nucleo-




philic reactions; aqueous radical  reactions will not  be  a  significant process.




The  hydrolysis   chemistry  of  ethylene oxide  has  been  discussed  in  Section




3.6.U,  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 pro 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/fc.




These  variations  in  hydrolysis  rates  are  well  within  the  error  limits




discussed by Mabey and Mill (1978)  in regard to hydrolysis  experiments.




     It  is interesting  to  note that the presence  of  a microbial  population in




the unfiltered river water did  not  materially decrease the half-life  of  the




ethylene  oxide   merely  from  related   hydrolysis  reactions  with  the  moieties




present  in biological  systems.  Although  the microbial  concentration  was  not




reported,  the  lack  of  a  significant  change  in  rate may indicate  that  such




reactions are not significant in river water.   In  addition,  it should be noted



that the half-life of 12 to  14 days  allows for sufficient  time for exposure of




ethylene oxide to biota and possibly humans,  although the  latter  would be much




less  likely  given  the  addition  of  hypochlorite   in water treatment  plants.




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 to 10.




     Evaporation from water also appears  to be a  significant  process.   Conway




et al. (1983) reported  the calculated  relative desorption  coefficient a, (ad =




Kd  (ethylene oxide)/Kd(02),  Kd  is  the  desorption coefficient)  to   be  0.31,



                                      6-2

-------
0.3^, and 0.36  for  10,  20,  and 30°C water.  Experimental values for 22°C water


are 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  water  body with a  rate dependent upon the  actual oxygen-transfer rate


in  a specific  system.    The  rate  of  desorption  will  be  less than  that for


volatile  low  solubility  organics  such  as  toluene,  benzene,  and chloroform,


which have an a  of around  0.65 (Rathburn  and Tai,  1981).


     Conway et  al.  (1983)  also measured  the  biochemical  oxygen  demand  (BOD)


using  2 m£ of  domestic sewage/BOD  bottle.   They found  the  biooxidation as a


percent  of  theoretical  to  be  5,  22,  40,  and  52% on  days  5,  10,  15,  and 20,


respectively.    Conway et  al. (1983)  suggested that in a  sewage treatment  plant


where  the microbial  population is  much  higher,   biodegradation  may  be very


fast.   However, from  their data  it is  not possible to determine whether or not


the chemical actually being degraded is  ethylene oxide or ethylene glycol from


hydrolysis, since  the hydrolysis  half-life,  which is =14  days,  is similar  to


the BOD half-life of a little  less than 20 days.


     Hendry et  al.  (197*0  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  105  M~1  s~1  for ethylene  oxide.   Given  an

                                                       I h
alkyloxy radical concentration in  ambient water of 10    M,  the  half-life for


this process  is =6  years.    Hence, for ethylene  oxide, hydrolysis and evapora-


tion appear to  be  the dominant fate processes,  while no definitive statement


regarding biodegradation can be made.
                                      6-3

-------
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,  however,




that given  the major  components  of  soil, the  half-life  of ethylene  oxide  in




soil would be shorter than in water.








6.4.  ETHYLENE OXIDE FATE IN THE ATMOSPHERE




     Epoxide  degradation  in  the  atmosphere can  be  inferred  from  information




derived from  their oxidation  by free-radical pathways.   Little direct informa-




tion on epoxide behavior in the environment was available.




     Atmospheric  reactivity   of  volatile   organics  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.   The most  important  is in choosing the appropriate




hydroxyl radical concentration.  The second  difficulty is more fundamental and




questions  the appropriateness  of chosing  the  OH  reaction  as  the  dominant




removal  mechanism.    A number  of different modeling  and direct  measurement




efforts have  been expended  in  determining the hydroxyl  radical  concentration




in  the  atmosphere.   These  have  provided a wide range  of values  of varying




accuracy for  both  average and altitude specific concentrations.   A reasonable




compromise  for an average  OH concentration appears to  be 1   x  10  molecules




cm   based  on more recent modeling efforts  (Cupitt,  1983).   For  ground level




concentrations, the values  may be somewhat higher,  possibly around 1.3 to 1.4




x  10  molecules cm~3 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 lifetimes of ethylene

oxide vary between  215  days (using the lower limit  of  the Arrhenius equation)

and  100  days  (using the upper limit  of the Arrhenius equation) for a hydroxyl
                                   r               o
radical  concentration of  1 x  10"  molecules  cm" ,  and  159 days  (using the

lower  limit  of the Arrhenius  equation} and 74 days  (using  the upper limit of

the  Arrhenius  equation).   Thus,  given the limits, the  lifetimes  vary between

215  and  74 days.   This  lifetime is in sharp contrast  with  other  ethers which

are  significantly  shorter.    For   example,  tetrahydrofuran,  a five  membered

cyclic ether,  has a  lifetime  of =1 day.   Fritz  et  al.  (1982) suggested that


this is  due  to the distorted  sp  bonds in ethylene  oxide that  give  rise to a

hydrogen  abstraction  activation  energy  of  5.8 kcal/mol,   rather  than  the

standard 2.8 kcal/mol.


     Bogan and  Hand (1978) determined  the  absolute  rate constant  of the reac-

tion oxygen  atoms  [0(3P)]  with ethylene  oxide  to   be  (6.3  +  0.18) x  10


cc/mcule-sec at 300 K.  This  rate is  several  orders of magnitude slower than


the hydroxyl radical  reaction, and  yields  a half-life  of  1400 years,  given an

atmospheric 0(3P) concentration of 2.5 x 10  molecules/cc (Graedel, 1978).


     Sickles et al.  (1980) used a  Teflon  smog chamber and  the rate  of  ozone

production to  rank 19 compounds relative  to propane.   The  chambers  were out-

doors and  irradiated  with  sunlight.   Purified air,  an organic compound,  and

NO,,  were added  before sunrise  to multiple  chambers;   ethylene oxide to  NCL


ratio at  the  onset of  each experiment was 4:0.067.   Sickles  et al.  (1980)

found  ethylene  oxide  much less  reactive  than propane,  with  ethylene  oxide


being the  fifth least  reactive  compound  tested.   The order  of  reactivities

found was:
                                      6-5

-------
               Acrylonitrile
               Perohloroethylene
               Ethanol
               Ethylacetate
               Acetone
               Methanol
               Acetic acid
               Propane
               Ethylene dichloride
               Acetylene
               Chloroform
               Dimethyl formamide
               Benzaldehyde
               Methylene chloride
               Pyridine
               Ethylene oxide
               Methyl chloroform
               Phenol
               Acetonitrile
               Nitrobenzene

When compared  to  an indoor  smog  chamber study  (Dimitriades  and Joshi, 1977),

the relative ordering of compounds was similar.

     All of  these results indicate  that  ethylene oxide  is  relatively unreac-

tive in the atmosphere  compared  to  other ethers.  Given  the  shortest lifetime

of  7^  days  for  hydroxyl  radical   reaction,  and  the  production  volume  and

volatility, it should be possible to detect  ethylene  oxide in ambient air, yet

no  reports  confirming  its  detection were found  in  the  available  literature.

This lends support to the  possibility  that a reaction  or  reactions other than

hydroxyl radical, possibly ionic, are  dominant.   This possibility is supported

further by  the facility with which  ethylene oxide undergoes  ionic reactions,

yet no  such  reactions  have been identified  by  atmospheric chemists.  Nonethe-

less,  the  possibility  that   such  a  reaction  may   be  dominant  remains.

Atmospheric constituents such as suspended particles  may catalyze the decompo-

sition  of  ethylene  oxide  as  well  as  water  vapor  and  other  nucleophilic

species.  Thus, with the  information currently  available,  no definitive state-
                                      6-6

-------
ments  can  be  made  regarding  the  atmospheric  fate or  lifetime  of  ethylene




oxide.








6.5.  DEGRADATION IN COMMODITIES AND MANUFACTURED PRODUCTS




     Ethylene oxide  is  registered  in the United  States  for use  as a  fumigant




or  sterilant on  several  stored  food  commodities  and  manufactured  products




(Goncarlovs,  1983).    These   include  as   a   fumigant  on  furs,  bulk  food



containers,  food  containers,  stored  grain, stored  fruits,  stored processed




foods,  garments,  stored  herbs,  stored  spices,   furniture,   aircraft,  buses,




railroad  cars,   laboratory  animal  bedding,  and  tobacco  products.    As  a




sterilant,  it is  used principally  on  hospital equipment  and  Pharmaceuticals.




The use  of  ethylene oxide as a  fumigant  is  chiefly to protect stored products




from either insect or microbial  destruction.   The fate of this epoxide and its




residue  are  especially  important in those materials, commodities, and products




coming  into close  contact  with humans,  such  as surgical  equipment,  Pharma-




ceuticals,  and  food  service and  packaging materials  (Wesley  et  al.,  1965;




Alguire, 1973; Holmgren et al., 1969; Gilraour,  1978).




     Delineation  of ethylene  oxide fate in  these materials  has  established




that it  will degrade to glycol and  halohydrin or evaporate.   The  degradation



could  result  from chemical or  enzymatic  activity or  from some combination of




the two.   The  halohydrin formation requires  epoxide reaction  with inorganic




halide.   The halide could be  naturally present,  be added,  or be derived from




organic  halides.  Bromide  ion  often comes from degraded  methyl bromide, which




is also  a fumigant (Rowlands, 1971; Lindgren et al., 1968).




     Scudamore and  Heuser (1971) evaluated  ethylene oxide fate  for a variety




of  treated  commodities.  They  examined  degradation and  apparent vaporization




                                      6-7

-------
of  ethylene  oxide  and  its  residues.   The  losses  of  the  ethylene  oxide,



ethylene chlorohydrin,  and ethylene bromohydrin  were measured  over a  1-year




period.  Apparent first  order specific rate constants,  k,  were  calculated for




epoxide dissipation.   The  rate constant, k  combined losses from  the  degrada-




tion (chemical and metabolic pathways), k   and vaporization, k •
                    k =
The glycols  (ethylene and  diethylene)  were only  determined once at  either 6




months or 1 year  after  treatment.   Effects considered  included  ethylene oxide




treatment (dose  and temperature during  application),  moisture  content  of the




commodity, storage  temperature,  and  storage  in closed  containers or  in open




trays.  Ethylene  oxide residues  rapidly  dissipated.   While its estimated half-




life  was  longest at  10°C  in  sealed containers,  it never  exceeded  2  weeks.




Increasing the ethylene  oxide  dose had a varied 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, caused non-linear  correlations.    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  normally dissipate  from  treated  commodities,  but  under  some  circum-




stances, 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
                                      6-8

-------
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  an empty




container  and ascribed  this  to apparent  adsorption to  container  walls.   The




larger losses  experienced  with  2.1  kg of dates  in  the  container resulted  from




ethylene oxide uptake  by the  fruit.   The ethylene oxide  loss  in  treated dates,




which  were  left  in  open  containers,  was  attributed  to  degradation  to  the




chlorohydrin and glycol  combined with volatilization.




     The available  information  on fate  of ethylene oxide  applied  to manufac-




tured  goods was  not as  extensive  as  that  on  its  fate  in  commodities.   All




available  information  suggested  that  its   behavior  in  manufactured  products




corresponds  to the  pathways of degradation and volatilization  described above.




     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?  vaporized in the first




day, and  no residual  ethylene  oxide remained  after 5 days.   Ethylene oxide




loss  from  cream  cheese  wrappers  primarily consisted of  its  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.






                                      6-9

-------
Holmgren  et  al.  (1969)  measured 0  to  1500  ppm chlorohydrin  on  21  ethylene



oxide treated drugs.







6.6.  BIOACCUMULATION IN AQUATIC ORGANISMS



     Specific   experimental   information  regarding  the   bioaccumulation  of



ethylene  oxide  in aquatic  organisms  is  not  available.    Veith  et  al.  (1979)



have  suggested,  however,   the  use  of  the  following  equation  to  calculate



bioconcentration factors (BCF):
               log BCF = 0.76 log K   -0.23
                                   o w
where K   is the  partition  coefficient  between octanol and water.   Using this



equation and the  log K   of -0.30, reported by  Hansch and  Leo (1979), the BCF
                       ow


for whole fish was calculated to be 0.31*.
6.7. SUMMARY



     This  section  discusses  the  results  of  studies  relating  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  on the order of



12 to  14  days at  298  K.   Lower  temperatures  will  lengthen  the  half-life; pH



changes will  have  a minimal  effect.   Volatilization will also be a significant



process although  less   so  than for  sparingly  soluble  solutes  (e.g., toluene,



chloroform,  benzene).   There is  no conclusive evidence  indicating microbial



degradation  will be  significant;  however, the components  of  sewage  sludge may



react  rapidly with ethylene  oxide.   The  fate  of  ethylene oxide  in soil will



likely be similar  to water; its half-life will probably be shorter.



                                      6-10

-------
     The  fate  of  ethylene  oxide  in the  atmosphere  is not  clear  from the




information presented  in  the  available literature.   Rate constants are avail-




able  for  hydroxyl  radical  and oyxgen atom  [0(  P)]  reactions as  well as smog




chamber studies.   All  predict a  lifetime  of sufficient  length to  allow for




measurement,  the shortest  calculated  lifetime being  71*  days.   Nonetheless,  no




confirmed  reports  detailing  the  measurement of  ambient  levels  of ethylene




oxide were  found.   The possibility of nucleophilic  reactions in the gas  phase




may explain this.




     In commodities,  food containers, and  manufactured  goods,  ethylene  oxide




appears to volatilize or hydrolyze  to glycol  or  halohydrin with a half-life on




the order of 2 weeks.
                                     6-11

-------
                     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 identified.
                                      7-1

-------
     No  monitoring  data  was  available  for  ethylene  oxide  in  biological


tissues,  except  for  some tissue  distribution  studies.    Since  epoxides  are


reactive alkylating agents, it  is  reasonable  to  expect such results (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.  (197U)


tentatively identified  ethylene oxide in  the ambient air  near  the University


of Houston.  However,  the authors  used  Tenax as an adsorbant  for  trapping air


contaminants and its  use casts doubt on their  tentative identification, since


Tenax does not adequately retain ethylene oxide.


     U.S. EPA  (1976)  listed one monitoring  observation for ethylene  oxide in


water.   It  was  observed  in the  effluent from a chemical  plant  in Bandenburg,


Kentucky.   No other  epoxide  observation was reported.   U.S. EPA (1976)  also


noted  observations  of  ethylene  halohydrin,  but  its  origin  might  be  from


industrial wastes rather  than residues from epoxide.


     No other reports  of ethylene  oxide  in  ambient  air  or water were  found,


yet  SAI  (1982)  reported  that  the maximum  exposure  concentration  level  of

                                                           •3
ethylene  oxide,  based  on dispersion  models, was  5  (ig/m  (2.77 ppb).   One


reason  for  this  could  be its reactivity,  but another reason could be the lack


of an adequate sampling  method  (see Section  ^).  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  samples are  rarely as  high


as workplace samples  (particularly in the case  of  a  reactive  molecule such as


ethylene oxide), the  well documented NIOSH  (1977)  method  becomes inadequate.



                                      7-2

-------
The problem is compounded by  the  fact  that  few,  if any, monitoring studies are




undertaken  to  identify only  one compound  in the  environment.   Thus,  these




studies must assume  some  compromise between completeness and  speed,  making it




impossible to optimize conditions for any one compound.



     Several studies  have  examined  the  residues  of ethylene  oxide  applied to




commodities  and   manufactured goods  as  a  fumigant  and  disinfectant.    The




information  on  residues  in  commercial  products  is  discussed here.   Another




portion of  this  report (Section  6.5)  describes  investigations on the  fate of




this epoxide.   The present  section differs from the  previous section in that




the information here concerns residues in actual commercial products.




     Scudamore and  Heuser (1971) evaluated  ethylene  oxide  and its metabolites




in commercially treated products, and  also  did  some fate studies (discussed in




Section  6.5).    While  they  never  detected  ethylene  oxide  in  commercial




products,  they  did  find ethylene  chlorohydrin  residues  ranging  from  10 to




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.    Their  review 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 is




helpful.  Over 5 billion  pounds  (>2  billion  kg)  of ethylene oxide are produced




yearly.   The vast  majority is  used  captively  as  a synthetic  intermediate.




Possibly,  10  million  pounds  (4.5  million  kg)  are  used  for  fumigation/




sterilization,  which  includes  food  commodities,  medical  devices,  Pharma-




ceuticals, and cosmetics.   This use constitutes  the  only documented potential




exposure to  ethylene  oxide;  however, the extent  of  this exposure  needs  to be




determined.    Ethylene  oxide  also  appears  to  be  a   product  of incomplete




combustion,  and  has  been  identified   in  automobile  and  diesel   exhaust  and




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 miss ethylene  oxide even if present.




Several  studies  have examined  the persistence and  fate of  ethylene oxide in




commodities and commercial goods including food, medical supplies,  and drugs.
                                      7-M

-------
                             8.   ECOLOGICAL EFFECTS

8.1. MICROORGANISMS AND INSECTS
     Ethylene   oxide   is  utilized   as   a  fumigant   for   foods   and  spices
(particularly  grains),  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,  rickettsia and viruses.    Sykes  (1964),  for example, reported
that  exposures to gaseous  ethylene  oxide at  concentrations  of 1  to  10$ will
kill    Bacillus   globigii,    Staphylococcus     aureus,    Escherichia   coli,
Chromobacterium  prodigiosum,  and  Mycobacterium  phlei  within a   few  hours.
Roberts et  al.  (1943)  found that 10% gaseous ethylene  oxide will kill Bacillus
anthracoides  in  8 hours.   Ethylene oxide also produced significant sporicidal
activity  against  dry bacterial  spores  (Bruch  and  Koesterer,  1961).  Exposure
of Bacillus subtilis spores  to  1 to  2% vapor  concentrations  of ethylene oxide
killed  >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.,  1943).   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 the stored product insect population at  a concentration range of

                                      8-1

-------
6  to  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 Lindgren  as  intermediate  between that  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 ambient  exposure  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 8*1-90 mg/i, for  fish,  a mean
48-hour LC50 of  212 mg/J,  for  Daphnia,  and a mean 46-hour LC    of  745 mg/Jl for
brine  shrimp  (Table  8-1).    LC50  values for  the hydrolysis product ethylene
glycol were >10,000 rag/A for  the above  species  except goldfish  (which were not
tested) (Conway  et  al.,  1983).   If reacted  to  form ethylene chlorohydrin, the
96-hour LCc  for fathead minnows was  about 90 rag/A (Conway et al.,   1983).
                                      8-2

-------
                                             TABLE 8-1
                               ute Aquatic Toxicity of Ethylene Oxidec
Test Procedure
range-finding , static, aerated

range-finding , static, sealed
under oxygen
definitive static acute (no
aeration)
i static acute
u>
static acute


static acute


LC,-0 (95% Confidence limits), rag/J,
Test Organism 24 fir 48 hr 96 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
NA

NA

89 (63-125)

NA

300
137 (83-179)
200 (150-243)
>500
1000
490
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.    No  pertinent 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   gastro-




intestinal 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 and intravenous administration.




     Ehrenberg et  al.  (197^)  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




radioactivity  (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




that were  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 as  that  seen




following   intravenous  injection,  except   for   a  high  initial  labeling  of




respiratory mucosa  (data  not  shown).    Two  minutes  after   the   injections,




concentrations  of radioactivity  2 to  3  times  those seen  in the  blood were






                                      9-1

-------
                                                     TABLE 9-1




                                         Acute Toxicity of Ethylene Oxide
Route
oral
oral
oral
oral
oral
ihl.
ihl.
ro 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
LCn
low
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

-------
observed 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  autoradiographs,  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 4  beagle dogs following  the  intravenous administration  of 25  mg/kg  or 75



mg/kg  ethylene oxide on  separate occasions  (Martis  et al.,  1982).   Urinary



excretion data indicated  that 7-24?  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.

                                                       •jlj
     Two urinary  metabolites  were detected when [1,2-  C]  ethylene oxide was



administered  to Sprague-Dawley rats  via single intraperitoneal injection  at a



dosage of 2 mg/kg  (Jones  and  Wells,  1981).   The urinary metabolites were ^-(2-



hydroxyethyl)  cysteine   (9$   of  the  dose)   and  N-acetyl-J3-(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   C02 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.,  197*0,  the only urinary  metabolite  characterized was  7-hydroxyethyl-



guanine,  which  accounted  for a  minor amount  (0.007$) of  the  total  urinary



radioactivity.   Significant  alkylation  of  tissue   proteins  was  found,  but



alkylation of DNA was confirmed by  the high  specific activity of tritium as 7-



hydroxyethylguanine.  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 conducted  by Ehrenberg



et  al.  (1974)  using  tritium labeled  [1,2- H]  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, thus indicating rapid urinary elimination.


                                                                            14
     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$) appeared within  18 hours  of dosing  (Jones



and Wells,  1981).   Two  urinary metabolites,  S!-(2-hydroxyethyl) cysteine  and N-



acetyl-S-(2-hydroxyethyl)  cysteine accounted  for  9  and   33$   of  the  dose,


                                                                 14
respectively.  Within 6 hours,  1.5$ of the dose was exhaled as   CCL and 1$ as



unchanged   ethylene   oxide,   but   these   are   not  maximum  values   (exhaled



radioactivity was not sampled at later  post-exposure times).



     Martis   et   al.   (1982)  investigated  the   elimination   kinetics  of



intravenously administered ethylene oxide  in beagle  dogs.   Four dogs  received





                                      9-4

-------
single  25  and  75  mg/kg injections  of  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  min~  for the




low  and high  dosages,  respectively, were  calculated from the plasma concentra-




tion-time data  using  a  curve-fitting computer  program.   These rate  constants




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  concentra-




tions  reportedly  exhibited the  characteristics  of  a  metabolite  in  a  one-




compartment  model;  maximum  plasma  concentrations  of  ethylene gylcol  were




reached by  90 +  24.5 minutes  (25  mg/kg) and  120 +  42.4 minutes  (75 mg/kg)




post-injection.   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.









9.2.  ACUTE, SUBCHRONIC, AND CHRONIC TOXICITY









9.2.1.  Effects in Humans.
                                      9-5

-------
     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  (Greaves-Walker and Greeson,  1932;  Blackwood and Erskine,



1938;  von  Oettingen,   1939;   Anonymous,   19*17;  Sexton  and   Benson,   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  leukocytosis  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).    Inhalataion




exposure  to high concentrations  of  ethylene  oxide for  brief periods has been




associated  with  bronchitis,  pulmonary edema,  and  emphysema (Theiss,  1963),  as




well as  convulsive movements  (Salinas et  al.,  1981).   In a controlled study of




the  effects of ethylene oxide on human volunteers,  Greaves-Walker 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,  who  were drenched with  "\% aqueous ethylene



oxide  solution,  developed  marked nausea and  profuse  vomiting  several  hours




following  exposure  (Sexton  and  Henson,  1949).   Large  vesiculated blisters




developed in the areas of exposed skin, and two workers  who had complete blood




counts taken showed a mild leukocytosis.




     Cobis  (1977)  reported  a  very low incidence of  health-related 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






                                      9-6

-------
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




concentration was not given, and it is  presumed  (although  not stated)  that the




employees were exposed to ethylene oxide vapor.



     The dermatological  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  Kay,  19^9) and  hand  burns  (Royce and  Moore,  1955), for




example,  have been  observed  in  workers  that  wore  ethylene  oxide-sterilized




rubber boots  and rubber gloves,  respectively.  Biro et al.   (197*0 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-related burns.



     Sexton   and  Henson  (1949)  described the  dermatological  reactions  that




occurred in  6 men whose  skin was directly exposed to a  1/t  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




(vesicular eruptions),  but  nausea and  vomiting  were  the  only  systemic  effects




noted.




      In  a  subsequent  study,  Sexton   and Henson  (1950)  applied  1   to  100?




solutions  of  ethylene  oxide   to the  skin of  8 volunteer   subjects  for  time




intervals  that ranged  from 20  seconds  to  95 minutes.  The  magnitude of skin






                                      9-7

-------
injury appeared  to  be related to  the  duration  of contact and  the   concentra-



tion.   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  to  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 durations of contact.  Patches  were  removed  from  the subjects after 1,




2, M  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






                                      9-8

-------
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 approximated




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  to U



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, Theiss  (1963)  did not  observe  sensitization in




ethylene  oxide  plant  workers  who  were  challenged  with  a  single  dermal




application  of 1$ after an  average  of  10.U  years of occupational exposure.



Anaphylactic reactions  have  been  observed  in  patients  using  ethylene oxide




sterilized  plastic  tubing  for   hemodialysis   (Poothullil  et   al.,   1975)  or




cardiac catheterization  (Pessayre and  Trevoux,  1978).   These symptoms included




uticaria, breathlessness, and  hypotension.   In a  follow-up study  on a patient




apparently sensitized  to 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
                                      9-9

-------
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   usage  of   ethylene  oxide



sterilized  plastic  tubings  have also  been  published   (Hirose  et  al.,  1953;




Clarke  et  al.,   1966).    Ethylene  oxide,  rather  than a  chemical  reaction



product,  is  implicated,  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




corneal 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  only in irritation




of the  conjunctivae that persisted for about  1 day.








     9.2.1.2.    SUBCHRONIC  AND  CHRONIC  EXPOSURE  ~ Limited   information  is




available on  toxic effects of  subchronic or  chronic ethylene oxide exposure in




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






                                     9-10

-------
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,  roughly  indicating 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 M.  Three of  the four cases had worked  as sterilizer  operators  for more



than  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/thinking disturbances) and  had  abnormal  neurological examination




results  that were  consistent with  sensorimotor neuorpathy.   Nerve conduction




studies  were  abnormal  in the   three  operators,   including  the asymptomatic




patient, and were   compatible  with  the  diagnosis  of  sensorimotor neuropathy.




Removal  from exposure resulted in  relief  of symptoms within  2 weeks.   Two of




the  individuals returned  to  work  under normal  conditions of  lower ethylene




oxide  exposure,   but   improvement   in   nerve  conduction  was  not  observed;




significant  improvement  was   noted,  however,   in  the  third  individual  who




returned to work in a position without ethylene oxide exposure.
                                      9-11

-------
     Jensen  (1977)  reported  that  three workers  using ethylene  oxide  steril-



izers  were  hospitalized  for  neuropathy  of  the  lower  limbs.    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,



subsequently  developed  cataracts.   One  of  the  operators  was exposed  to the




leaking sterilizer  for  2 months  and  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  (11 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




production  workers   when  compared  with   control   subjects   (Ehrenberg  and




Hallstrom, 1967).   The  design and  results  of this  study are  more  completely




described in Section  13-5,  but  it should be noted that the production workers
                                      9-12

-------
were reported to  have  been exposed for 2-20 years (average of  15  years)  to  an




unknown level of compound.



     Joyner  (1964)   conducted  a  retrospective  morbidity  study  of  37  male




ethylene 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




operators  (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 9.4  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  time-weighted average exposure  to ethylene  oxide was well




below 50 ppm.




     Hogstedt et  al.  (1979)  conducted a  cohort  study of mortality  among  89




full-time   ethylene  oxide   production  workers,   86   intermittently   exposed




maintenance workers,  and a  group  of  66  unexposed control workers  during the




                                     9-13

-------
years  1961-1977.   As  described  in Section  9.5,  exposure patterns  were quite



complex;  in  addition  to  ethylene  oxide  (concentrations  were generallay  <50



mg/m ),  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  a considerable  excess  mortality  when



compared  with the expected number 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  or more  of  induction-latency



time were  required  for inclusion,  there were  12 observed deaths  attributed to



the  circulatory  system  (9  due  to  coronary  heart  disease  and  3  due  to



cerebrovascular disease),  with an  expected  incidence of 6.3;  this  difference



was statistically significant  (P<0.05).   The excess  mortality is of  the same



magnitude  in  a  restricted cohort of  those with 10  or more years  of employment



in  ethylene  oxide   production  and  20  years  of  induction-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




summarized 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  central  nervous  system  depression,   including



lacrimation,  nasal    discharge,   salivation,  nausea,   vomiting,   diarrhea,



respiratory  irritation,  incoordination, and  convulsions  (Sexton and  Henson,






                                      9-1*1

-------
19^9;  Hollingsworth  et  al.,  1956; Hine  et al.,  1981).  Animals  that survived
the initial  exposures  showed  subsequent  bronchitis,  pneumonia,  and loss  of
appetite,  with  delayed  symptoms  of  apathy,  dyspnea,  vomiting,  paralysis
(particularly of  the hindquarters),  periodic  convulsions  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 dealyed  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.^5  to  1.5   mg/mA  produced  a   significant   decrease  (-30$)  in
glomerular  filtration  rate,  which   indicates effects  of  ethylene oxide  on
kidney  function.
     Ethylene oxide  in 10/6  and 50%  aqueous  solutions  produced  hyperemia and
edema  in  shaved  rabbit  skin when  applied  through  cotton  pads  for   1  to
60 minutes  (Hollingsworth et  al.,  1956).   Bruch  (1973)  studied  the  dermal
irritation properties of  2 to  10$ 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

                                      9-15

-------
application in the rabbit resulted in mild  irritation.   Topical  or intradermal
administration of  1%  ethylene  oxide  (0.5 mfc), thrice  weekly for  3  weeks,  did
not result in sensitization in guinea pigs (Woodward and Woodward, 1971).
     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$  ethylene
oxide.   In another  study  of  ocular  irritation in  rabbit eyes,  Woodward and
Woodward  (1971)  found slight irritation  following a  single  application of 10$
aqueous  ethylene  oxide  (duration  of   exposure  unknown),  and  a  no-effect
concentration  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
pathologic changes are  similar to those observed in  acute  studies  with  lung,
kidney, and liver  damage occurring, and with neuropathy of  the hindquarter and
testicular tubule  degeneration occurring  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,

                                      9-16

-------
                                                              TABLE 9-2

                                                Subohronio Toxioity of Ethylene Oxide
      Route
   Species
                                         Concentration
                     Number of
                     Exposures
                                 Effects
                                    Reference
    inhalation
    inhalation
vo
I
    Inhalation
20 rats (10/aex)
16 guinea pigs (8/sex)
 5 mice (female)
 2 rabbits (1/sex)
 1 monkey (female)

30 mice (female, white)
20 rats (male, white)
841 ppm
100 ppm
up to 8 in 10 days
(7 h/d; 5 d/wk)
30 (6 h/d; 5 d/wk)
20 rats (10/sex)
10 mice (female)
                                          357 ppm
                     33-38 (7 h/d; 5 d/wk)
    inhalation    16 guinea pigs (8/sex)    357 ppm
                                               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-
charge, diarrhea, labored breath-
Ing, weakness of the hind legs,
and some deaths (13/20 exposed
and 0/20 control rats, and 2U/30
exposed and 3/30 control mice).
Fifteen additional rats or mice
were examined pathologically;
changes were limited to a few
cases of hemosiderosis in the
spleen that occurred late in the
exposure period.

Death in 10/10 mice (33 exposures)
and 18/20 rats (38 exposures)
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
of the testicular tubules with
replacement fibrosis (males),
slight fatty degeneration of the
adrenal cortex (females).  No
nervous system effects or
mortality.
Jacobson et al.,
1956
                                                              Hollingsworth
                                                              et al.,
                                                              1956
                                                                                   Hollingsworth
                                                                                   et al., 1956

-------
                                                          TABLE 9-2 (cont.)
    Route
   Species
                                          Concentration
                                            Number of
                                            Exposures
                                                                                                 Effects
                                                                                Reference
    inhalation
2 monkeys (1/sex)
2 monkeys (males)
357 ppm
357 ppm
    inhalation   3  dogs  (male,  Beagle)
                             290 ppm
\£>
 I
00
    inhalation   20 rats
                             201 ppm
    Inhalation
 8 guinea pig
 H rabbits  (2/sex)
 2 monkeys  (female)
201 ppm
38-iM in 60 days
91 in 110 days
(both schedules 7 h/d;
5 d/wk)
               30 (6 h/d;  5 d/wk)
               127-133 in 185-193 days
               (7 h/d; 5 d/wk)
127-157 in 176-226 days
(7 h/d; 5 d/wk)
Growth depression and charaoteris-  Holllngsworth
tic neurological impairment (e.g.,  et al.,
hind limb paralysis and muscular    1956
atrophy, poor or nonexistent knee
reflex, extensor reflex and
hindquarter/genltalia pain percep-
tion).  No histopathologic effects
of exposure.

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
hematocrit.  Hematologic 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)

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.

-------
                                                                ABLE 9-2 (oont.)
     Route
   Species
                                                noentration
               Number of
               Exposures
                                                                                                 Effects
                                                                                                     Reference
     inhalation
     inhalation
     Inhalation
vo
 I
20 rats
 8 guinea pigs
 H rabbits (2/sex)
 2 monkeys (females)
113 ppm
30 mice (females,  White)  100 ppm
20 rats (male,  White)
 3 dogs (male,  Beagle)    100 ppm
122-157 in 176-226 days Growth depression and a moderate    Hollingsworth
(7 h/d; 5 d/wk)         increase in lung weights in rats    et al., 1956
                        were the only adverse treatment-
                        related effects noted.
                130 (6 h/d,  5 d/wk)
                130 (6 h/d,  5 d/wk)
                        No clinical signs of toxicity or
                        treatment related mortality
                        (3/20 exposed and 3/20 control
                        rats, and 8/30 exposed and 1/30
                        control mice died).  No significant
                        pathologic changes in additional
                        groups of 60 rats or mice.

                        Nonnochronic anemia (decreased
                        RGC, Hb and hematocrit) indicated
                        in 1 and suggested in 1 of 3 dogs.
                        No changes in the 3rd exposed dog,
                        or in control dogs.
Jacobson et al.,
1956
                                                                                                                             Jacobson et al.,
     inhalation         20 rats
                         8 guinea pigs
                         4 rabbits (2/sex)
                        10 mice (female)
     oral (Intubation)  5 rats (female)
     oral (intubation)  rats (female)
                          19 ppm
                         h/d, 5 d/wk)
                         100 mg/kg
                127-131 in 180-184 days No adverse effects as Judged by
                (7 h/d, 5 d/wk)         general appearance, behavior,
                                        mortality, growth, final body
                                        and organ weights, and gross
                                        or microscopic pathologic
                                        examination.
                 15 doses in 21 days
                (5 d/wk)
                          10 or 3 mg/kg   22 doses in 30 days
                                         (5 d/wk)
                        Weight loss, gastric irritation
                        and slight liver damage, but no
                        mortality.

                        No evidence of adverse effect
                        as Indicated by growth, hema-
                        tology, blood urea nitrogen
                        determinations, organ weights
                        or gross microscopic pathology.
                                                            Hollingsworth
                                                            et al., 1956
Hollingsworth
et al., 1956
                                                                            Hollingsworth
                                                                            et al., 1956

-------
                                                        TABLE 9-2 (cont.)
Route
s.c.
vO
to s.c.
O
s.c.
i.v.
Species
rats
rats
dogs
dogs
Concentration
54 mg/kg
18 mg/kg
36 mg/kg
36 mg/kg
Number of
Exposures
30
30
30
21
Effects
Weight loss, injection site
hemorrhage and inflammation.
No observed effect.
Anemia, hyper plastic bone
marrow, and ectoplc
hematopoiesis.
No observed anemia, other
observations not mentioned.
Reference
Hoi lings worth
et al., 1956
Hoi lings worth
et al., 1956
Woodward and
Woodward, 1971
Balazs, 1976
d = day; h = hour;  wk = week

-------
rabbits, and  monkeys  showed paralysis and  atrophy  of the muscles of  the hind




limbs.   These effects were reversible after  discontinuation of exposure  for




100 to  132 days.   Special  studies on monkeys  were carried out with  repeated




(38-94)  exposures to this level  of ethylene oxide.   Knee  jerk  reflexes  became




very  weak,  pain perception in  the  hind quarters  decreased,  the  cremasteric




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.




     Preliminary results of  a  chronic  inhalation  study  conducted by  NIOSH have




been reported (Lynch  et  al., 1982).   Male  F344  rats (80  per  treatment  group)




and male cynomolgus monkeys  (12  per  treatment  group)  were  exposed  to either 50




ppm or  100 ppm  ethylene  oxide  for  7 hours/day,  5 days/week  for  2*1  months.




Additional details  of  the epxerimental design are  presented  in Section  9.5 of




this study, but  it should be noted that  the  rats were included primarily for



carcinogenicity evaluation,  and  that the  monkeys  were  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 cymphocytes.   The




results that are currently available are summarized  below.




     As detailed in Section 9»3» weight gain  throughout most of the study and




survival  were significantly  depressed in  the rats  at both  exposure  levels




                                     9-21

-------
(Lynch et  al.,  1982).   Weight gain was  significantly  depressed  in the treated



monkeys beginning at week 25.  The livers 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 9.3).    Hematologic  anlayses  showed no  statistically



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




aberration  and  sister  chromated  exchanges  were  observed  in   the  peripheral



lymphocytes of these animals.



     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  (Jacobson  et  al.,  1956).    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) was  unable  to  repeat these findings in



beagle  dogs,  however,  with  an  ethylene  oxide-glucose  solution  administered



intravenously over the same concentration range in a 21-day study.



     An oral feeding study using 10$  ethylene oxide  in olive oil was performed



on rats (Hoilingsworth et al.,  1956).  Rats  fed 100 mg/kg ethylene oxide in 15






                                      9-22

-------
doses  over  21  days showed marked  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  central  nervous  system  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  (particularly  of   the  hindquarters)  and periodic  convulsions




frequently preceded death.   Death  in  ethylene  oxide-exposed laboratory animals




is  usually  due   to  lung  edema  or  secondary  infections  in  the  lungs,  and




postmortem  pathologic  findings  in other  organs  include widespread  hyperemia




and congestion (e.g., liver, kidneys,  spleen) and fatty degeneration (liver).




     Dermatological effects of ethylene oxide following  skin contact in humans




following  accidental or experimental  exposure  include  edema,  erythema,  and




vesiculation with  possible bleb formation.   These changes  typically  progress




in  the above  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 sufficient  chemical  strength  to  cause




injury except  after  prolonged  contact.   Skin  burns  have also  been  caused by




residual  ethylene  oxide   in  clothing  or  footwear   that   was   treated  or






                                      9-23

-------
accidentally  contaminated  with  the  compound.   Sensitization  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



corneal 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 elevated  lymphocytes) have




been  noted  in  chronically  exposed ethylene  oxide  production plant workers.




Retrospective  morbidity  and  mortality studies  of  ethylene   oxide production




workers  do  not  suggest,  however,  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 been observed.








9.3.  TERATOGENICITY  AND REPRODUCTIVE TOXICITY




     Batelle Pacific Northeast  Laboratories (Hackett  et al.,  1982)  conducted




teratology  and   reproductive   studies   for   the   National  Institute  for






                                      9-24

-------
Occupational  Safety  and  Health investigating  the effects  of ethylene  oxide

(EtO)  produced  by inhalation  exposure.   Rabbits  and rats  were  exposed to  a

single dose  of  EtO 150 ppm,  (Union  Carbide,  Linda Lot No.  01901,  99-756 pure)

both prior to, and during the  time of  organogenesis.   Thirty New  Zealand white

rabbits per group were exposed in  three  different  regimes (filtered air alone,

EtO exposure  on  days  7-19  of gestation,  and  EtO 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,  EtO exposure on  days  7-16 of

gestation,  EtO  exposure  on  days  1-16  of  gestation, and  EtO exposure  three

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,  increases  in kidney  and spleen

weights with  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- and post-

gestationally with a trend for early midgestational resorptions.   In addition,

fetal  body  weight,  decreases  in crown-rump length and increases  in  incomplete

skeletal ossification were observed  in all  EtO  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  mother and developing fetus.   However,
                                      9-25

-------
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 EtO or  EtO reaction  products  left on  improperly



degassed surgical supplies, LaBorde and Kimrael  (1980) conducted  studies on the



effects of EtO administered intravenously.   CD-1  mice  in  four replicates  of



three  treatment  groups (10  animals  per group)  were treated with  0,  75,  150



mg/kg EtO (Eastman Organic  Chemicals Co.  purity not stated, EtO was  injected



in 5% dextrose solution) .   The animals  were  exposed  in  the following treatment



periods of  gestation;  days  1-6  (period I),  days  6-8 (period  II),  days  8-10



(period III)  and 10-12 (period IV).



     Clinical  signs of maternal  toxicity (weakness,  labored  breathing,  tremors



and death)  were observed in animals injected  with  150 mg/kg  EtO  on  gestational



days  4-6 (Period  I), days  8-10  (Period III),  and  days  10-12 (Period IV)  but



not 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  was  accompanied



by  decreases   in  the  mean  number  of  live   fetuses  in periods III  and  IV.



Embryotoxicity, as  manifested by significant 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



reduction in the mean number of  live fetuses per litter (and also  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  were



noted  in  periods II,  III  and IV  at 150  mg/kg level,  but  in  period III the



incidence did  not achieve  statistical significance.   It  was  concluded that the



EtO   exposure,  under   these  conditions,   was   selectively   affecting   the



                                      9-26

-------
development  of  the conceptus (skeletal  malformations  and embrotoxicity)  since




EtO exposure in  period  III (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  because  maternal




deaths  were  observed in group  III  before and  after  days  6-8  of gestation.




Although  there  was no  dose-response relationship in  the severity  of adverse




effects  in  either the mother  or fetus,  the types of  malformations 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  (Kimmel et  al.,  1982)  evaluated




the reproductive  effects of intravenous injections  of  EtO  in rabbits.   This




study  was  reported briefly in a poster session presented at  the 1982 Society




of Toxicology meeting.   New Zealand white  rabbits were intravenously injected




in  two treatment  regimes;  0,  9,  18 or 36 mg/kg EtO (source and  purity not




reported) on days 6-14  of gestation, or 0,   18 or  36 mg/kg  on days  6-9  of




gestation.   Seventeen to twenty-one  animals were examined in the group exposed




on days  6-9, eighteen to  twenty-four animals  examined  in  the  group exposed on



days 6-14.




     Maternal toxicity 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




gestation.   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




                                      9-27

-------
embryotoxic effects were  observed  in  the day 6-9 treatment  groups,  however in




the 6-14  day treatment group  a significant  dose-related  trend for  decreased




numbers of  live  fetuses/litter and reaorptions/litter  were  observed.   At  the




36  mg/kg   level,   the   incidence   of  resorptions/litter   was   statistically




significantly different from control  levels.  Therefore, the authors concluded




that  intravenous administration  of  EtO in  rabbits  produced  embrotoxicity,




however only at doses which also produce 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 EtO in mice and  rabbits.   Since ECH  is produced  by  the




interaction  of EtO  and  chloride  ions, it  is a residue of  EtO that  could be




left on medical  devices after  improper degassing of  EtO during sterilization.




Forty-one  to sixty-five CD-1 mice  were  intravenously  injected with 60 mg/kg or




120 mg/kg ECH  (source  not reported, ECH was  injected in 5%  sterile dextrose)




on  days 4-6,  6-8,  8-10,  or  10-12  of  gestation.   Seventeen  to  twenty-two  New




Zealand white rabbits were intravenously injected  with  9,  18, or  36 mg/kg  ECH




on days 6-14 of gestation.




     In this study, no  adverse  effect  was 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  gram or  more  in  24 hours) were  observed in  the




mothers in all treatment  periods at the 120 mg/kg dose.  Maternal weight gain




during  the  entire  treatment   period  and during  pregnancy  were  significantly




reduced at the 120  mg/kg  level on   days 4-6,  6-8  and  10-12.   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






                                      9-28

-------
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.   However,  at the 60 mg/kg




level,   in  animals  treated  on days  8-10,  fetal  weight  reductions  occurred




without 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 pre-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  a water vehicle or




left untreated.  One hundred eggs were used per group.




     Ethylene  chlorohydrin  was   found   to   be  toxic  in  this  system  with




significantly  increased mortality (no  hatch) at  levels X25  mg/kg at the 0 hour




exposure, and at levels  VI2.5 mg/kg  at  the 96 hour  exposure.    Statistically




significant increases in  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  however, is  not   known since




teratogenic  effect  in  chick  embryos may  not  be predictive   of  mammalian




effects.




                                      9-29

-------
REPRODUCTIVE EFFECTS




     The Carnegie-Mellon Research Institute  (Snellings  et  al.,  1982) conducted




a  one-generation  study  evaluating  the  effects   of   EtO  exposure  due  to




inhalation.  Thirty  male  and female Fischer-S1^ rats were continually exposed




to  10,  33  and  100  ppm  ethylene oxide  with  the  control  animals  exposed  to




filtered air.  Prior  to cohabitation, all  groups  were initially exposed to EtO




for  6  hours/day,  5  days/week  for 12 weeks.   After  one week  of  cohabitation




females with  vaginal plugs  were removed,  and  the  other  animals  were rotated




with a different male to allow for mating for another week.   At the end of two




weeks all male and  female animals were  separated.   The  males were then exposed




to  EtO  for  6 hours/day,  7 days/week  for  an additional  three  weeks.    The




females were  exposed for  6 hours/day,  7  days/week  from  day  one  through day




nineteen of gestation.  On  the twentieth day  of exposure,  females  not pregnant




were sacrificed.   The pregnant  females  were allowed to deliver and five days




after parturition were again exposed to EtO  for 6 hours/day,  7 days/week until




day 21 postpartum.




     The following  criteria were  used  to  establish fertility.   If  a female




produced a  litter,  or if gross  examination  revealed implantation  sites  after




staining,  then she  was  considered fertile.   Any  female,  not becoming 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  infertility  after mating with  a  male  of  proven  fertility.




However, this  incidence  did not  achieve statistical  significance.   In the




males  there  was  no  increase  in  infertility.     In  the  100    ppm  group,






                                     9-30

-------
significantly  more  females had lengthened gestational  period  (time of vaginal




plug to litter) than the control, 10 or  33 ppm  groups.   The control,  10 and 33




ppm  groups had gestational  periods of  22  days, while  the 100  ppm  group had




gestations  ranging  from 22  to  31  days  (7/14 rats  had 22  day  gestation,  4/11*




rats  had  23  day gestation,  3/14  rats had  greater  than  25  day  gestation) .




However,  since  most of the  animals  did not have extensively  long  gestational




delays,  it is  not  clear whether this  lengthening of  gestation  represented a




true adverse biological effect.




     In   this   study   (Snellings  et   al.,   1982),   the  number  of  pups  was




significantly  reduced  with a decrease  in  the  number of  implantation sites at




the  100  ppm level.   However,  of  the  surviving  pups,  there was no  effect on




survival  after  parturition.   In the parental generation, there was no adverse




effect on bodyweight  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 EtO exposure.




     It  was  concluded  from this  study  (Snelling  et  al.,   1982)  that  EtO




administered to rats has the potential  to  disrupt reproduction  by causing an




increased  incidence  of embryolethality.  However,  this embryotoxic effect was




only observed when  the  animals  were exposed  to  the highest  dose  ( 100 ppm)  and




not at the lower doses (10,33 ppm)  of EtO.








TESTICULAR EFFECTS




     Hollingsworth  et  al.  (1956) investigated  the acute and  chronic toxicity




of  EtO   in  a  variety  of  animal  species.     Positive   responses  related




specifically to the male  reproductive  system  were  observed  in hamsters  and




                                     9-31

-------
rats.   Eight guinea pigs  were exposed to  357  ppm ethylene oxide  (commercial




grade EtO, 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  EtO,  7  hours/day,  in  122 to  157 exposures  given  over  an  experimental




period  of 176 to  226  days.   Only  slight but  not  statistically  significant




decreases in testis weight of  rats and  guinea pigs were observed.  However in




rats,  there  was   histological  evidence   for  a  degeneration   of  testicular




tubules.




     A  recent study sponsored by the NIOSH  described the effects of  inhaled




EtO on  semen  production  in Cynomologus (Macaca fasicularis) monkeys  (Lynch et




al., 1983).   The  monkeys  were exposed by inhalation to  50  and 100 ppm EtO,




(Union  Carbide,   99.7$)  7 hours/day,  5  days/week   for  2  years.    In  the




preliminary  range-finding   study,  only  two   animals  per  group  were  used.




Testicular weight  was  diminished  in animals  exposed  to  100 ppm EtO but were




only marginally  decreased  in those  exposed  at  the  50  ppm  level.    Similar




decreases  in epididymal  weights  were  also   reported.    Sperm motility  was




significantly reduced  at  the 50 and  100 ppm  level, both  in  terms   of the




percent 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  level.   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 3-4 fold
                                      9-32

-------
increase  in  drive range when  the  animals  were exposed to 50  and  100 ppm EtO.




However,  there was no effect on sperm head morphology (Lynch et al.,  1983).




     In   another  study   relevant   to   the  effects  of  EtO  on  the  testis,




radiolabeled EtO  was  detected  in  autoradiograms of mice gonads (epididyrais and




testis)  20  minutes  after intravenous  injection  (Appelgren  et  al.,  1977).




Radioactivity was found in the epididyrais  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 a




discussion of  this  study  in Mutagenicity  Section).   This study is relevant to




testicular effects because it establishes  that  EtO has access to the  gonads.








ADVERSE REPRODUCTIVE OUTCOME IN HUMANS




     There  is  little  information  relating   to  the  effects  of  EtO  on  the




reproductive system in humans.  In  one  study  a comparison was made between the




health of 37 male employees involved in EtO  production with 41 men who worked




in  other  production units (Joyner,  1964).   This  study  evaluated many health




endpoints including genitourinary  problems.   The mean  exposure period was 10.7




years with a general level of exposure on  the  order  of  5  to  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   cause.




                                     9-33

-------
However,  the  higher  incidence of  absenteeism was  attributable  to a  single




individual  in  each  category.    Therefore,   it  was  concluded  that  long  term




exposure  to EtO  had  no  adverse  health effect  on  the  men  involved  in  EtO




production.    However,  since  this  study  did  not  deal  specifically  with




reproductive  health  problems,  it  is  of  limited  value  in  determining  the




potential of EtO to cause adverse reproductive effects.




     A  study  by  a Russian  investigator  (Yakubova, 1976)  reported  that  female




workers involved  in EtO production experienced a number of  gynecological  and




obstetrical  problems.    These  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,  anemia,  toxicosis,




and shortened pregnancies.   In this study,  the observations were  reported in




an  anecdotal  manner  with  no  presentation  of  actual  data  or  description of




methodologies.  Therefore,  it  is  of little  value in the scientific review of




adverse reproductive effects.




     Holmberg (1979) and Holmberg and Nurminen  (1980)  reported case studies of




a mother exposed  to a  variety  of  organic solvents.  These studies  describe an




adverse  reproductive  outcome  of  a woman  exposed to  alkylphenol and dyes as




well as EtO.   The  same woman  may have  been described in both these reports,



however,  this  was not made  clear  in the articles.   Both reports  describe an




infant  born with  hydrocephalus and  Holmberg  (1979)   described  a  child  with




additional  malformations  (cleft palate,  double  uterus,  polydactyly).    These




reports  are  not  useful  in  establishing  causal  relationship  between  EtO




exposure  and  congenital malformations  because  EtO  was not  the  only chemical




involved.   A larger population  size would  have to be  evaluated before  such an




association can be established.






                                     9-34

-------
     An  epidemiology  study has  been conducted  concerning  the effects  of EtO




exposure on pregnancy outcomes  in  nursing personnel.  This  report  is the only




one  which adequately  evaluates the possible  causal association  between EtO




exposure  and  adverse  human  reproductive effects,  and  has  been  reviewed  in




depth by  an Environmental Protection Agency epidemiologist (Margosches, 1983).




In  cooperation  with  the  Finnish  investigators the  data  has  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,  claimed adjusted  spontaneous abortion  (s.a.)  rates  of




16.7$  for "exposed"  and  5.6/1  for  nonexposed  pregnancies  among  these staff.




The  report  singled out  ethylene oxide,  glutaraldehyde,  and  formaldehyde use




and  suggested  concentrations  as   low  as  0.1-0.5  ppm  EtO 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  EtO 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.




                                      9-35

-------
     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




distribution   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$)  ever pregnant  women among sterilizing  staff  and  the  57^  (55$)  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




EtO at a  hospital  to  be  exposed to  EtO 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  characterization of




individual  exposures in purely  qualitative  terms.  Dr. Hemminki believes that




typical exposures have averaged <1 ppm (measured by gas-tight  syringes).   He




                                     9-36

-------
bases  this  belief  on  papers  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 EtO  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-exposed  and non  exposed  age-adjusting (£30)  the rates among




discharge-registry-identified  pregnancies,  the EtO-exposed  s.a.  rate  (16.1)




also   exceeded  the  rate  (9.4)   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  epidemiologic  afflictions  as




reporting  and  recall  bias.   Additionally,  the  methodology for  statistical



analysis,  based on  rates  adjusted  for  such  concommitant  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




                                      9-37

-------
interest.    Another  shortcoming   is  the  impreciseness  with  which  hospital




exposure  history  was  determined.    Finally,  although  the  authors planned  a



priori to  investigate  relationships between EtO  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  EtO  exposure  and adverse  pregnancy  outcomes  or  other




reproduction effects."  (Margosches, 1983).








SUMMARY OF TERATOGENICITY AND REPRODUCTIVE TOXICITY




     The  potential  of ethylene oxide  (EtO)  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).




     In a  teratology  study, Hackett  et  al.,  1982 reported that  rats exposed to




a single  150 ppm  dose of EtO 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).  However, similar  effects were not produced




in rabbits exposed to  150 ppm EtO in this study.




     LaBorde and Kimmel,  1980,  administered 75  and 150 mg/kg EtO to 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






                                      9-38

-------
                                                                     TABLE  9-3

                                                                 Summary of Studies
Type of
Study
Houte of
Administration Species
Dose of Level and
Time of Exposure
Findings
References Comments
       Teratology
iv
CD-I mouse
                        0, 75,  150 mg/kg
                        day 1-6, 6-8, 8-10
                        or 10-12 of gestation
                                       1.  Developmental toxicity LaBorde and
                                          at or near dose level  Klnmel, 1980
                                          which produced maternal
                                          toxicity (150 mg/kg)
U)
vo
Teratology iv



Teratology/ Inhalation
Reproduction








Teratology/ Inhalation
Reproduction











New Zealand 0, 18, 36 mg/kg
white rabbit day 6-9 of gestation;
0, 9, 18, 36 mg/kg
day 6-1*1 of gestation
Sprague- 150 ppm, 7 hr/day:
Dawley CD day 7-16 gestation,
rat day 1-16 gestation,
3 weeks pregestation
plus day 1-16
gestations




New Zealand 150 ppm, 7 hr/day:
white rabbits day 7-19 gestation,
day 1-19 gestation










1 . Developmental toxicity
only at levels which
were maternally toxic
(36 mg/kg, day 6-11)
Teratology:
1. Retarded fetal
development
Reproduction:
1 . Maternal toxicity
2. Increase intrauterine
mortality



1 . No teratogenic or
reproductive effects











Kimmel et al. ,
1982


Battelle Pacific
Northwest Labora-
tories (NIOSH
210-80-0013)
Hackett et al. ,
1982




Battelle Pacific
Northwest Labora-
tories (NIOSH
contract No.
210-80-0013)
Hackett et al. ,
1982










Teratology:
1 . Inadequacies
a) no maternal
toxic doses
b) dose response
not determined
Reproduction:
1 . Inadequacies
a) dose-response
not determined
Teratology:
1. Inadequacies
a) no maternally
toxic doses
b) no develop-
mental toxic
doses
c) dose-response
not determined
Reproduction
1 . Inadequacies
a) dose-response
not determined

-------
                                                                        TABLE 9-3 (cont.)
<£>

Jr
O
Type of Route of
Study Administration Species
•One genera- Inhalation Fisher
tion reproduc- 311 rats
tion
Chronic Inhalation Guinea pigs
toxicity
(male
reproduction)
Dose of Level and
Time of Exposure
0, 10, 33, 100 ppm
12 wks prior to mating,
6 hr/day, 5 day/wk.
During gestation -
days 0 through day 19.
During lactation -
days 5 through 2 1 .
357 ppm, 123 7-hr
exposure in 176
days
Findings
1 . No difference in FQ
fertility. No F
toxicity.
2. No adverse effects on F^
survival, growth rate,
or lactation.
3. Adverse reproductive
effects at highest dose,
100 ppm.
a) increased gestational
length
b) decreased litter size
c) decreased implantation
sites (i.e., decreased
fecundity)
d) decreased fetuses/
implantation sites
( embryo lethal)
1 . Tubular degeneration of
tests with replacement
fibrosis
References Comments
Carnegie-Mellon
Research
Institute, 1979
(Spelling et al. ,
1982)
Hollingsworth
et al., 1956
                                         Rats
204 ppm, 122 to 157
7-hr exposures in
176 to 226 days

201 ppm, 122 to 157
7-hr exposures in
176 to 226 days
1. Slight decrease in testes
   weight, not statistically
   significant.

1. Slight decrease in testes
   weight, not statistically
   significant.
2. Testes: small, slight
    degeneration of tubules.
Testicular Inhalation
toxioity
Cynomologous 50, 100 ppm,
monkeys 7 hrs/day for 2
years
1.
2.
3.
1.
Decreased testicular
weight.
Decreased sperm
concentration.
Decreased sperm motility.
No change in sperm
morphology.
Lynch et al., 1983
       A variety of experimental  protocols were  utilized, only those which provided positive  information on reproduction effects are noted here.

-------
                                                               TABLE 9-3 (oont.)
VO
Type of
Study
Medical
survey of
workers




Medical
survey of
workers





Case study



Route of Dose of 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 female



Findings References
1. No observed increase Joyner, 1964
in male reproductive
disorders.




1. Gynecological dis- Yakubova, 1976
orders, spontaneous
abortions, toxicosis,
decrease birth weights.




1. Infant with hydro- Holmberg, 1979,
cephalus Holmberg and
Nurminen, 1980


Comments
1 . Small sample
size
2. Study did not
evaluate
fertility or
testicular
function.
1. Difficulties in
translated material
2. Little information
provided on experi-
mental design.
3. Multiple exposures
to noise and high
temperatures
1 . Mother exposed to
multiple chemicals.
2. Only one infant
studied.
Epidemiology Occupational Pregnant <1 ppm
study exposure female






1. EtO exposure Hemminki et al.,
associated with an 1982
increase in spontan-
eous abortion




1 . Possible bias
introduced by
supervisors
who categorized
participants in
this study.
2. Limited exposure
data
       Teratology   Intravenous     CD-1  mouse
0, 60, 120 rag/kg
day 4-6,  6-8, 8-10,
10-12 of gestation
Maternal toxicity at
120 mg/kg for all
treatment periods
Embryotoxicity at 120
mg/kg for all treatment
periods and at 60 mg/kg
on days 8-10 (fetal
weight reduction).
LaBorde and Kimmel,
1980

-------
                                                                   TABLE 9-3 (oont.)
VO

-Cr
ru
Type of Route of
Study Administration Species
Teratology Intravenous New Zealand
white rabbits




Teratology Air cell Chick embryo
and Toxicity injection




Dose of Level and
Time of Exposure
0, 9, 18, 36 mg/kg
day 6-T4 of gestation




0, 10, 25, 50, 100
200, mg/kg at 0 hour
incubation; 0
5, 12.5, 25, 50,
100 mg/kg at 96
hours incubation

Findings References
1. No effect on mother or LaBorde et al.,
fetus 1982




1. Ovo-toxic at levels Verrett, 1974
>25 mg/kg at 0 hour,
and >_12.5 mg/kg at
96 hours.
2. Teratogenlc to chick
embryo

Comments
1 . Inadequacies
a) no maternally
toxic doses
b) no develop-
mentally toxic
doses
Uncertainties in
extrapolating avian
developmental effects
to those of mammals


      "Original study performed by Carnegie-Mellon Research Institute (Bayes,  1979);  later  published  as  Snellings  et  al.,  1982

-------
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 maternal toxicity.



     Similar studies were  conducted  by Kimmel et al.,  1982,  on  the effects of



18  and  36  mg/kg  EtO  administered   intravenously  to  New  Zealand  rabbits.



Significant  maternal  toxicity  (decreases  in weight  gain)  were  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 investigated  the  teratogenic  effect of intravenously administered



ethylene  chlorohydrin  (ECH), a reaction product of EtO,  in CD-1 mice and New



Zealand rabbits.   No adverse  maternal or  embryotoxic  effects were  produced in



the rabbits.   However  in the mice,  at the  highest  dose  (120  mg/kg)  severe



maternal  weight  loss  with increases  in  resorptions/litter  and decreases  in



fetal weight were observed.  At  the  60 mg/kg level, 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 was  also reported  to  produce



adverse  effects   in  developing  chick  embryos   (Verrett,   1974).    Structural



abnormalities were   produced  by   12.5  to  100  mg/kg of ECH  when the egg was



incubated with the chemical for up to 96 hours.



     In a one generational  study,  (Snellings  et  al.,  1982)  female rats exposed



by  inhalation  to  100  ppm EtO   had  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  EtO,  as  well  as  a




                                      9-43

-------
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).




     Adverse  effects  on  the  testis  resulting  from  EtO exposure  have   been




reported for the hamster and rat (Hollingsworth et al.,  1956)  and Cynoraologous




monkey (Lynch  et al.,  1983).   Hollingsworth reported testicular degeneration




occuring in hamsters  and  rats  exposed to EtO by  inhalation  (204 to 357  ppm) •




Lynch et al. (1983) reported adverse  effects on sperm concentration, motility,




but not  morphology  in Cynoraologous monkeys.   The monkeys in this study  were




exposed over  two years to 50 and  100 ppm  EtO  by inhalation.   In  mice radio-




labeled EtO 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 EtO




in the human.   Medical surveys have  described  either no adverse reproductive




outcome (Joyner, 1964)  or  a variety of adverse outcome  (Yakabova,  1976).   The




study by Joyner, 1964 is inadequate because it does not  deal specifically with




adverse reproductive  outcomes.   The report by Yakabova,  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 EtO  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 EtO to cause adverse effects.




     A recent  epidemiology  study has been  conducted   evaluating  the  pregnancy




outcome of nursing personnel exposed  to EtO (Hemminki et al., 1982).  Although




there were  problems  in the study  design  and collection of  data,  the  data  is




                                      9-44

-------
sufficient  to suggest  an  association between  EtO  exposure and  spontaneous



abortion  and  warrants   further  examination  of  adverse  pregnancy  outcomes.




Additional epidemiology  studies  would  be  helpful to more  firmly  establish the




potential of EtO to cause adverse reproductive effects in humans.




     In  conclusion,   EtO appears  to  be  capable  of producing  developmental




toxicity,  i.e.,  structural defects, in  utero death,  growth  retardation,  and




infertility in laboratory animals.   The levels needed to produce  these effects




approach or equal the levels  needed  to produce toxicity in the dams.   EtO has




been  shown  to produce  adverse  testioular  effects  (testicular  degeneration,




poor semen quality)  and was found to accumulate  in  the epidiymus.   The effects




of  EtO  on  human  reproduction have  not  been studied  in  depth,  although one




study indicates that EtO may  be  associated with  spontaneous abortion (Hemminki




et al.,  1982).  Future studies are needed to establish this effect in humans.








9.1.  MUTAGENICITY




     Ethylene  oxide  (EtO)  has  been  evaluated   for  mutagenicity  in  several




different  systems   including  tests  in  bacteria,  fungi,  higher  plants,




Drosophila, mammalian cells iji vitro,  and rodents.   Effects in humans are also




reported.  The available data concerning  the  mutagenicity  of EtO  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 EtO  (e.g., Fishbein,  1976,




Wolman 1979, Ehrenberg and Hussain,  1981,  and NIOSH, 1981).








GENE MUTATION STUDIES




Prokaryotic Test Systems (Bacteria)

-------
     Several investigators  have shown that  ethylene  oxide (EtO)  causes  point



mutations  in  bacteria (Table  9-*0 •    EtO 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  EtO



sterilization.   Jones and Adams  (1981) found  that  treatment  of spores of these



bacteria  with   Pennges  (12:88   EtO-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  were  not



well-defined genotypically  these data  suggest that  EtO  induced  mutations in



the surviving spores.



     In a  study by  Rannug et al. (1976), EtO 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  EtO



(purity not reported) ranging  from  0 to  95.5 mM  in a  suspension  test without



addition  of an exogenous mammalian  metabolic activiation system  (Table  9-*0 •



A  statistically  significant dose-related response was observed  (Figure  9-1)



where the maximum killing was *2Q%.



     In  another  Salmonella assay,   Pfeiffer  and Dunkelberg  (1980)  exposed




strains TA98,  TA100,  TA1535,  and TA1537  to  concentrations of  EtO (99.7?  pure



diluted in cold  acetone)  ranging from 0 to  200 uM (0  to  8.8 mg/plate)  (Table



9-M).   Between  6  and  10  trials  were  performed  and  each  was  conducted  in



duplicate.   A clear dose-dependent  response  was  observed  for the base-pair



substitution detecting  strains  TA100  and TA1535  but  not for  the frameshift
                                      9-16

-------
                                                                       TABLE 9-t
                                          Summary of Mutagenicity Testing of EtO:  Gene Mutations in Bacteria

Reference
Rannug et al.
1976




Test Activation
System Strains System
Salmonella/ TA1535 None
microsome assay
(suspension/
assay)


Chemical
Information
Concentration tested:
0 to 95.5 mM

Source : Fluka

Purity: Not given

Results
Strong
positive
response




Comments
1. Eto used as a positive control.

2. Dose-dependent response. 15-fold
increase in revertants noted at
highest dose compared to negative
controls.
VO
 I
      Pfeiffer  and
      Dunkelberg,
      1980
Salmonella/     TA98
microsome assay TA100
(plate test)    TA1535
                TA1537
                                                   None
Solvent:  Cold ethanol


Concentration tested:
0 to 200 [imol/plate
(0 to 8.8 mg/plate)

Source:  J.T. Baker
         Chemicals BV
         Deventer, The
         Netherlands

Purity:  99.1%

Solvent:  Cold acetone
                                                                                                        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 \imol/
                 plate), revertant count at highest
                 dose (200 (imole) 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.

-------
TABLE 9-t (oont.)
Reference
Tanooka, 1979





j

3


Test
System
Bacillus subtilis
spores ( reversion
to his*
pro to trophy)







Strains
HA 101
(his met
leu)

TKJ 5211
(his met
uvrATO)

TKJ 8201
(his met
polA151)
Activation Chemical
System Information Results
None Concentration tested: Positive
27. 3% atmosphere of EtO response
gas for times ranging
from 5 to 50 minutes.

Source: Daicide LS gas
Daido Oxygen Co.
Tokyo , Japan

Purity: 27.3* EtO
72. 7> Freon
Comments
1 . Tests conducted in a polyethylene bag;
t 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 10~3
after 50 minutes exposure of HA 101
and TKJ 5211).
                                    1.  Lethal and mutagenic effects were
                                        enhanced in the polA strain; TKJ
                                        8201 was 10x more sensitive than
                                        HA  101 and TKJ 5211.

-------
                                                                         FIGURE  9-1
                                    MUTAGENIC RESPONSE OF Salmonella typhimurtum STRAIN TA 1636 EXPOSED TO ETHYLENE OXIDE
                     108 r
 I
J=
vo
Muttntt/
plat*
y - 0.996 x + 6.02
R « 0.9897
p<0.01
                                                                 36            48

                                                            EtO Concentration (mM)
                          Rannugatal. (1976)

-------
detecting  strains  TA98 and TA1537.   This result is  consistent  with responses



observed for other alkylating agents.



     Tanooka (1979)  exposed spores  from three different his- Bacillus subtilis



strains  to an  EtO gas mixture (Daicide  LS  comprised  of  27.3$  EtO  and 72.7$



freon  gas)  in  a plastic  bag  (Table  9-H).    Histidine-independent  revertants



were  selected  after treatment;  a repair-competent strain  and a  uvrA repair-



deficient  strain  were  treated  for  times   ranging   from   5  to  50  minutes.



Exposure-related revertant frequencies  were observed  for both  strains (ranging


            -6                                     U
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 the  results  with  the



repair competent and uvrA  strains.   The revertant frequencies  corresponding to



5 and  MO  minutes of exposure were about 8 x  10   ,  and 3 x 10   ,  respectively.



A similarly elevated  sensitivity of  the  polA  strain was  observed  for  EtO-



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 EtO  gas for   30 minutes were characterized,  and 85$ of  them were found to



contain suppressor mutations;   15$ were  true revertants as  measured by cotrans-



formation 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 EtO is mutagenic in B.   subtilis.



     The positive responses in these tests show  that  EtO causes  genetic damage



as evidenced  by induction of mutations  in  bacteria.   The studies described



below show that  EtO causes genetic damage in  higher organisms too.
                                     9-50

-------
Eukaryotic Test Systems



Plants—



     Yeast—Kolmark  and  Kilbey (1968)  studied the  induction of ad+ revertants



in  Neurospora crassa  strain  K3/17 (macroconidia)  after  treatment with  EtO



(source  and  purity not  given).   Five  doses ranging from  0.0015  to 0.15M were



employed, but  the corresponding mutation frequencies were  not  reported (Table



9-5) .   The purpose  of the work was to  study kinetics  of mutation 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.    EtO



treatment in liquid  suspension  at  concentrations  from  0.5 to 15 mM resulted in



dose-related increases in mutation frequency; survival  was  reduced  about  60$



at  the high dose.   One hundred fold increases in mutation frequency were noted



at  the high  dose  levels  compared  to  the corresponding  negative  controls both



with and without  metabolic activation  by phenoballitone-induced mouse liver S9



mix (50.28  +  1.76 vs.  0.59 +  0.22 and  66.21 +  29.U4  vs.  0.66 +  0.59 muta-



tions/101!  survivors,  respectively).   The  ranking  of the  chemical substances



tested with respect  to their relative specific activity was epichlorohydrin >



EtO >  glycidol >  1,2-epoxybutane > 1,1,1-trichloropropylene oxide  > propylene



oxide >  2,3-epoxybutane.



     Angiosperms—EtO  is  known to be  a  very   effective  mutagen of higher



plants.   Many tests  have been  performed  in which  EtO  has  been shown  to  be



mutagenic.  The results of these studies will not  be  analyzed  in depth.  Most



were directed mutagenesis  tests  conducted  to generate  desirable traits in food



crops.   The results  of two tests,  in  which plants  were  treated with EtO, will




                                     9-51

-------
                                                                        TABLE 9-5
                                  Summary of Mutagenicity Testing of EtO:  Gene Mutation Tests in Lower Plants  (Yeast)
   Reference
Test System
Chemical Information
                                                                                      Results
                                                                                                                        Comments
   Kolmark and Kilbey,  ad-3A nevertants
   1968                 in Neurospora crassa
                         Concentration tested:
                         ranged from 0 to 0.1 M
                         (0 to 6.2 g/i) EtO.

                         Source:  Imperial Chemical
                                  Industries Ltd.

                         Purity:  Not given

                         Solvent:  Distilled water
                                Dose-related
                                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 /10  survivors).
f> Migliore et al.
vj» 1982
ro
Forward mutations at
the ade locus in
Schlzocaooharomyces
Source:  Montedison (Italy)

Purity:  99.70J
Dose-related
positive response
ponbe
Solvent:
Water and
DMSO

Without S9
Dose
(mM)
0
0.5
1.5
5
15
Survival
100
71.78
99.19
80.3
35. 11
Mutation Freq.
.x10
0.66 + 0.59
1.89 + 1.00
1.17 ± 0.75
18.77 * 0.72
66.21 + 29. 11
Survival
100
100
76.61
100
12.87

With S9
Mutation Freq.
.x10"^
0.59 ± 0.22
3.32 + 0.96
7.15 ± 0.21
11.33 ± 7.62
50.28 + 1.76

-------
be discussed  for  illustrative purposes (Ehrenberg  et  al., 1956, and  Jana  and



Roy,  1975).  Ehrenberg et al.  (1956)  administered  several chemical  substances,



including  EtO  (purity  not  given),   to dry  and  presoaked barley  seeds  and



screened for sterility (dependent  on chromosomal aberrations)  and  chlorophyll



mutations  (caused  by gene mutations,  either chromosomal  or  extrachomrosomal)



in the developing plants (Table 9-6).  The seeds were  exposed  to EtO either as



a  gas  (dry  seeds receiving  80%  EtO  for 6  days)  or in  solution.   For  the



solution  exposure experiments,  the  seeds  were presoaked in  0.12  and  0.03?



(0.27 and  0.07 M) solutions  for  2  hours.    EtO  induced mutations  in a  dose-



dependent manner  as  can  be seen in  Table 9-6.   A fivefold increase in  lethal



mutations 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 EtO (purity not given) solutions  from  0.1  to 0.6$ (0.02 to 0.14 M)



at 10°C for 8  hours  at pH  7.0.  The seeds were sown and  the  plants  were  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.



     The  positive   responses   observed   in   plants  is   consistent  with  the



bacterial results and shows EtO is mutagenic in plants.
                                      9-53

-------
                                TABLE 9-6
Summary of Mutagenicity Testing of EtO:   Mutation Tests in Higher Plants
Reference
Ehrenberg et al.
1956
vo
V71
XT




Teat System Chemical Information
Results
Lethal (chromosomal) When tested as a gas, resting ' Postive response
and chlorophyll seeds exposed to 80$ EtO for
(gene) mutations in 6 days. When tested in
barley. solution, partly presoaked
seeds exposed to 0.03$ and
0.12J (0.27 and 0.07 M) EtO
for 2 h at 20°C.

% %
EtO Sterility
0 4
0.03 5.7
0.12 9.5
80 ~22.1

% 2 nd
generation
chlorophyll
gene
mutations
0.051
0.20
0.75
1.8
Comments
1 . Third generation progeny not
available for analysis when report
written; positive response may be
due to extra chromosomal mutations.
2. Mutagenic response observed after both
types of treatment.
3. Half-life of EtO in water solution
Is around 100h at 20 *C.
No. spikes treatment
analyzed condition
15,861 None
2,510 Solution
1,872 Solution
989 Gas

-------
                                                                    TABLE 9-6 (oont.)
Reference
Jana and Roy,
1975
Teat System
Chlorophyll gene
mutations in rice
Chemical Information
Concentration tested:
ranged from 0 to 0.6J
Results
Dose-related
positive response
Comments
1. Objective of study was
kinetics of mutation.

to study
                         (IRS  and  Dular)
Ul
Ul
EtO.  Seeds treated
for 8 hours at 10°C
and pH 7.0

Source:  Eastman Organic
         Chemicals

Purity:  Not given

Solvent:  Not given
                                                                    % 2nd  Generation  Chlorophyll
                                                                           Gene Mutations
                                                                     
-------
Animals—



     Insects—EtO  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)  EtO 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$  EtO,  respectively.    The  dose-related   positive response  reported



indicates EtO is mutagenic in Drosophila.



     Watson (1966)  fed EtO 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  EtO   treatment (Table   9-7).    For  the  sex-linked



recessive lethal  test,  about 3%  lethals were detected  at the low dose (0.4$



EtO)  compared to  1%   at the high dose  (0.7% EtO).   For  translocations these



values were  ==0.28$ and 0.7$, respectively.   Negative control values  were  not



given.  Storage of EtO-treated sperm in  the  seminal  receptacles for 6 days  had



no effect on the frequencies of the two types of genetic damage.



     Lee (unpublished) conducted  parallel experiments  with unlabeled  and  3H-



labeled EtO to determine:






     1.   The relation of epxosure to level of alkylation of germ cell DNA.






                                     9-56

-------
                                                                   TABLE 9-7

                                    Summary of Mutagenicity Testing of EtO:   Gene Mutation Tests in Insects
 I
Ln
--4
Test
Reference System Strain
Bird, 1952 Drosophila Orgeon K:
melanogaster adult males
aex-linked
recessive
lethal test












Watson, 1966 Drosophila Oregon K:
melanogaster adult males
sex-linked
recessive
lethal test
and heritable
translocation
test

Chemical
Information Results
EtO administered by feeding, Dose-related 1.
inhalation or injection. positive
(Data not presented for response
first two routes of
administration.) For 2.
injection experiments 0.59
to 5% solutions administered
to 20 males. Dosages >0.8J
lethal. 0.8% EtO killed 50$
of treated flies while 0.5%
EtO did not affect viability
Source: Not given % No.
EtO Chromosomes
Purity: Not given
0 W
Solvent: 0.<4» saline 0.5 713
0.8 198
Concentration tested: Positive dose- 1.
0, O.OH, or 0.7* (0, related
0.09, or 0.16 M) EtO response

Source: Not given

Purity: Not given 2.

Solvent: Not given

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.U
9 1.5
Objective of experiment was to determine
effect of sperm storage in female seminal
receptacle on mutation frequency after
treatment with monofunctlonal and
bifunctional alkylating agents.

Did not observe storage effect for EtO
with respect to either endpolnt.

                                                 Pre-stored


                                                                 0.7

                                                 Post-stored      0.4
7.1

3.3
3.1
              3.  Cannot determine germ cell stage
                  specifloity.

            % Trans.
% Trans.    % Lethal

  0.29        0.08
  0.39        0.1

  0.69        0.1

  0.79        0.2U
  0.37        0.12
                                                                 0.7
6.8
  0.60
                       0.09

-------
                                                               TABLE 9-7 (oont.)
u»
00
Reference
Lee, unpublished
Teat
System
Drosophila
melanogaater
sex-linked
recessive
lethal test
and gonadal
Chemical
Strain Information Results
Source: Not given for
unlabeled EtO
3H-EtO from New
England Nuclear
sp. act. =2.8
ci/mmote
Comments
1. Objective of experiment was
the relation of exposure to
alkylation of germ cell DNA
to mutational response.

to determine
level of
alkylation
Purity:  Not given
                                                                            Exposure
                                                                       (qmole/25 ml vial)

                                                                              0

                                                                              0.086

                                                                              O.H3
                                                (Dose)
                                              Alkylation/
                                            Nucleotide x 1Q-3

                                                 5.58

                                                22.3

-------
     2.   The relation of  germ cell DNA alkylation  to  mutational  responses in
          Drosophila melonegaster 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

mi- scintillation vials  (0.086 or 0.^3  (imole/vial).   Immediately afterwards 50

males were  added  to the vials which were  sealed  and treatment  was continued

for 2U hours  at 25°C.    l4C-Thymidine was  also  given to males in the dosimetry

experiment and  alkylations  per nucleotide  of DNA were  calculated  based on the

SH/I^C ratios  in  purified  sperm  DNA (to determine  the  number  of alkyl groups

present) and the  1^C/sperm cell ratio  (to  determine the amount  of sperm cell

DNA  in  the extraction  product).    The  genetic   data  showed  EtO  to  be  an

effective mutagen  as dose-related  increases  in sex-linked  recessive  lethals

were   observed (see Table  9-7).   Using the  exposure-dose relation 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 EtO  is  distributed  to the  gonads  of a higher

eukaryote (Drosophila) and  causes heritable genetic damage.

     Mammalian  Cells  in   Culture  —  Three  tests  have  been  conducted  to

ascertain the  ability of  EtO to cause gene mutations  in mammalian cells in

culture.   Brown et  al.  (1979)  reported  in an  abstract that polymethacrylate

(PMMA)  plastic  sheets  and   polypropylene  (PP)  plastic  sheets   and  meshes

sterilized  by  EtO  gas  adsorbed EtO  molecules  which could be released later to

exert a mutagenic effect.   They  placed  the EtO treated plastic, of  unspecified

size, in culture flasks  containing  L5178Y  TK+/~ mouse lymphoma cells for  three

days.   This was followed  by dilution  in  EtO-free media  for 3  days  prior to

selection using BUdR.    PMMA  sheets  treated  for  18  hours with  pure EtO were
                                      9-59

-------
estimated to  release  8 to UO  [ig  EtO  (as measured by gas  chromatography)  into




the flasks, while similarly treated PP sheets  and meshes released 5  to  100 ug




EtO.  Although  the  spontaneous negative  control mutation  frequencies were not




given, the released EtO was reported  to  result in a 2- to  20-fold  increase in



induced mutation  frequency  relative to  the  controls (see Table 9-8).   It was




not possible  to evaluate  this report critically,  because it was presented in




abstract form.




     Tan  et   al.  (1981)  administered  EtO (Matheson Co.,  99.7$  pure,  Dr.  R.




Gumming,  personal communication)  to  Chinese  hamster  ovary  cells  at  concen-




trations ranging  upwards to 10 mM in  the medium.   Mutations at the HGPRT locus




were selected after  5 hour  EtO treatments both  with and  without an  exogenous




metabolic  activation  system   (S9 mix derived from  Aroclor  1254-induced  rat




livers)  followed  by  a 16-18  hour  recovery  period  and  subculturing for  one




week.    A dose-dependent  positive  response  was  obtained at  concentrations




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.  Stephen Nesnow  personal communication (1983)



exposed Chinese hamster V-79  cells  to EtO 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  was a dose-




related increase  in mutation  frequency.   The  response for the highest dose was




20  times  greater  than negative out rats  at  reported to be repeatable but  this




could not be verified because the work was reported in an abstract.
                                      9-60

-------
                                                             TABLE 9-8

                                Summary of Mutagenioity Testing of EtO:  Mammalian Cells in Culture
Reference
Brown et al.,
1979
Teat Activation
System System
L5178Y TKV- None
mouse lymphoma
gene mutation
assay
Chemical
Information
Polymethacrylate (PMMA)
plastic sheets and
polypropylene (PP)
plastic sheets and meshes
sterilized for 18 h in
Results
2 to 20-fold
induced
mutation
frequency
observed
Comments
1. Presented in abstract.
2. Chemical concentrations measured by gas
chromatography.
                                           pure gaseous EtO.  PMMA
                                           retained EtO and established
                                           concentrations of 8-HO (ig/20
                                           mi cultured medium (1-5 x
                                           10-5M EtO).  PP retained
                                           EtO and established
                                           concentrations of 5-100 ug/
                                           20 ml in cultured medium.

                                           Source:  Not given

                                           Purity:  Not given

                                           Solvent: None
                                                                                 Two EtO metabolites also tested.   At the
                                                                                 low, but unspecified,  level tested,
                                                                                 ethylene glycol residues did not  produce
                                                                                 an effect.   Chlorohydrin produced residues
                                                                                 of 15-30 ug/piece of PP.  Direct  addition
                                                                                 of this compound to the medium resulted
                                                                                 in a 2-3 x  induced mutation frequency.
Tan et al.,
  1981
CHO-KT-BHI,
HGPRT Chinese
Hamster Ovary
cell gene
nutation
assay
Liver 39 mix
from Aroclor
1254-lnduced
Sprague-
Dawley rats
Concentrations tested
0 to 10 mM
Dose-related
positive
response with
and without
activation
1.  Concentrations and induced mutants
    extrapolated from Figure 9-1 of text.

2.  250-300 mutants/10^ oells at high dose
    both with and without activation compared
    to 0-10 mutants/10^ cells in negative
    controls.

3.  Direct acting mutagen.

4.  EtO both cytotoxic and mutagenlc.

-------
                                            Mutation Frequency  (HGPRT mutant* x  10"6/clonable cell)
                       §
                       s
                       .
                       8
                  5
                  o
v£>
cr>
ro
                  SI
                  5'
z



o
m
Z
O
33
                                                                                                                       Vt
                                                                                                                       m
                                                                                                                       O

                                                                                                                       O
                                                                                                                       o
                                                                                                                       V)
                                                                                                                       m
                                                                                                                       5
      o
      G
      50
      PI
                                                                Relative Survival (%)

-------
     The studies  by Brown et  al.  (1979), Tan  et  al.  (1981) and  Hatch et al.



(1982) indicate that EtO causes gene mutations in cultured mammalian cells.








CHROMOSOME ABERRATION STUDIES




     Many studies  have  shown  that  heritable chromosome aberrations are induced




in  plants  after  EtO exposure  (e.g.,  Moutschen  et  al.  [1968]  in  barley and




Mackey [1968]  in  wheat).   These studies  will  not  be discussed  in this report.




Most were directed mutagenesis studies designed to  obtain  desirable variants.




The  ability  of EtO  to  cause such  mutations  shows   it  to  be  an effective




clastogen in plants.








Dominant Lethal Tests




     EtO causes chromosome damage both  in mammalian  germ cells  and  somatic




cells (Tables  9-9  to 9-13).    EtO  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  (Matter  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.




Twelve-week-old males inhaled  1000 ppm  EtO  for 4 hours (Matheson Gas Products,




Newark,  California,  purity  not given) .   The LC^ is  reported  to  be  1462 ppra




per 4 hours.   Embree 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




                                      9-63

-------
                                                                        TABLE  9-9

                                             Summary of Mutagenicity Testing of EtO:  Dominant Lethal Tests
v£>

 4=

Reference
Embree et al.,
1977
















Generoso
et al.,
1980














Test
System
Dominant
lethal assay
in Long Evans
rats














Dominant lethal
assay: male
mice T stock
(Experiment
I) and (101 x
C3H)Fi
( 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 EtO
via inhalation
for !J 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» U
1) 9 8
5 10* 4
10 9 11
•PO.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 J dead implants
in treated group
compared to 3 to 5%
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 C57BL)F1t
                                    (101 x C3H)Ft, or
                                    (C3H x C57BDF!.
                                    Sacrifice 12-15
                                    days after
                                    observation of
                                    vaginal plug.

-------
                                                                         TABLE  9-9  (oont.)
vo
 I
ui
Reference
Appelgren
et al. ,
1977







Test Mating and
System Sacrifice
Dominant lethal Males mated to
assay: mice 3 virgin
females per
week. Females
sacrificed
on 17th day
after first
exposure to a
male.

Chemical
Information
Single Injection of
either 0, 0.025,
0.05, or 0.1 g/kg
of EtO given i.v.

Source: Not given

Purity: Not given

Solvent: Saline
Results Comments
Negative 1. Reported data of dominant lethal test
response 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).
                                                                                              4.   Conducted whole body autoradlography study.
                                                                                                  Determined EtO distributed to various
                                                                                                  tissues in the body, including gonads,
                                                                                                  after either inhalation or injection.

-------
                                                                  TABLE 9-10




                                     Summary of Mutagenlcity Testing of EtO:   Heritable  Translocation  Test
vo
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 EtO weekdays
for 5 weeks
Dose Tranalocat
(mg/kg) Frequency
0 0/822
30 6/456
60 38/406
60 6/72
Results Comments
Dose-related 1. Shape of response curve consistent with
positive dose-squared kinetics.
response
2. Demonstrates capability of EtO to cause
heritable genetic damage in mice in vivo.
;ion Heterozygotes
0
1.32
9.36
8.33

-------
                                                                TABLE 9-11

                                   Summary of Mutagenicity Testing of EtO:  Chromosome Aberration Tests
Reference
Test
System
Chemical
Information Results
Comments
Fomenko and      Chromosomal
Strekalova, 1973 aberrations
                 in bone marrow
                 from rats
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
Time-dependent positive
response at highest dose
                                                                                          1.   Method of preparing cells for analysis
                                                                                              not given.

                                                                                          2.   Criteria for scoring aberrations not given.

                                                                                          3.   Definition of terms not given.

                                                                                          U.   Insufficient information for adequate
                                                                                              evaluation of results.
Strekalova, 1971 Chromosome
                 aberrations in
                 bone marrow from
 I               random bred
 9J              white rats
                                   Concentration tested:
                                   9 mg/kg per os
                         Positive response reported
                               1.   Animals  killed  21  and U8 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.
Polrier and      Chromosomal
Papadopulo, 1982 aberrations in
                 the human
                 amniotic cell
                 line FL.
Source:  Matheson Gas
         products

Purity:  Commercial
         Grade
 Dose-related positive
 response
                                                                                           1.   1 hour vapor exposure.

                                                                                           2.   Selected data  presented only for cells harvested
                                                                                               72 hours after exposure.
EtO
Dose
(DM)
0
5
7.5
10
%
Abnormal
Metaphases
10.8
21.7
59.7
77.8
Chromatid aberrations/ 100

cells
%
Breaks Exchanges Survival
3.0 5.1
15.0 5.0
37.6 15.5
79.2 115.1
100
58
25
9.2

-------
                                                                  TABLE 9-12

                                          Summary of Mutagenlclty Testing  of EtO:  Micronucleus Tests
Reference
Teat
System
Chemical
Information Results
Comments
   Appelgren et al.
     1978
Micronucleus  Concentration tested:
test:  NHRI
mice and
Sprague-
Dawley rats
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.
nice.  Increased incidence
in rats, but severe bone
marrow depression prevented    2.
further characterization.
The animals given the highest doaes died after
the first or second injection.

1000 polychromatic erythrocytes screened for
micronuclei per animal.
   Conan et al.,
     1979
ON
oo
Micronucleus
test:  Swiss
mice
Concentration tested:
Two injections.   Doses
ranged from 0-200 mg/kg
for i.p. injection,  or
0-5 mg adsorbed  to
implanted plastic
devices.

Source:  Not given

Purity:  Not given

Solvent: Water
Dose-dependent positive
response after i.p.
injection.
   Jenssen and       Micronucleus  Concentration tested
   Ramel, 1980       test:  CBA    0-175 rag/kg
                     mice (males)
                                   Source:   Pluka AG,
                                            Switzerland

                                   Purity:   Not given

                                   Solvent: Not given
                                         Positive response
                                                         1.  Two-fold increase noted in mioronucleus
                                                             formation (0.33 + 0.10 In controls compared
                                                             to 0.93 + 0.31 a£ 150 mg/kg.

-------
                                                                TABLE  9-13

                            Summary of Mutagenicity Testing of EtO:  Chromosome  Mutations  in  Human Populations
Reference
Test
System
Chemical
Information Results
Comment a
Theis3 et al., Chromaome
  1981         aberrations:
               peripheral blood
               of occupationally
               exposed workers   Exposure:
VO

VO
                                 1.  Long-term
                                     (>20 years)

                                 2.  <20 years

                                 3.  Long-term plus
                                     accident

                                 It.  Accident

                                 5.  Control
                                             Mutagenic effect indicated


                                             Aberrations excluding
                                             gaps:

                                             1.  a.  3.5
                                                 b.  2.7

                                             2.  2.3

                                             3.  2.2


                                             1.  1.4

                                             5.  a.  1.JJ
                                                 b.  1
                                                                                              1.   Workers  were  exposed  to  other alkylene
                                                                                                  oxides besides  EtO.   Cannot assign  damage
                                                                                                  to  one agent.
Pero et al.,
  1981
Chromosome
aberrations:
peripheral
blood lymphocytes
from EtO exposed
workers
                                    Exposure levels:
                                    0.5 to 1.0 ppm
                                    in air
Suggestive positive
response for aberrations
excluding gaps.  Noted
only in comparison
Both exposed groups has significantly higher
levels of total aberrations (breaks and gaps)
compared to the control group

-------
after  caging  with  a   treated  male.     Statistically   significant   (P<0.05)



increases in  postimplantation deaths were  observed on  weeks  1, 2,  3>  and  5



after  treatment,  but not  other weeks,  indicating EtO  exerts  its effects  on



postmeiotic cells.   It  should be  noted that  the  statistical significance  of



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  two weeks after administration  of



150 mg/kg  EtO (Eastman  Kodak,  purity not  given)  by a  single  intraperitoneal



injection.   One  dose of 200  mg/kg EtO was  shown  to kill  10  out of  12  mice.



The testing  for  dominant lethal effects in  this study was done two ways.   In



the first experiment, T  stock males treated  with EtO were mated  to two  virgin



(SEC x  C57BL)F.  females.   When  females were impregnated, as evidenced  by  the



observation of a  vaginal plug,  they were replaced  with  other females.   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  to  15



days after the observation  of the  vaginal  plug and were  dissected to  determine



the frequency of  dominant  lethal effects.    A  significant increase in postim-



plantation deaths  was  observed  in females  that were bred with  treated  males



between  days  2.5 and  11.5 post-treatment  (from  12  to  31$  dead implants  in



treated  group  compared  to  3  to 5% dead implants  in negative control group).



This indicates that  late  spermatids  and spermatozoa are  sensitive to  the test



compound.  In the second  experiment (101 x C3H)F.  males  were injected with EtO



and divided  equally into  four  groups.    Four days  post-treatment  they  were



mated  either  to  T stock,  (SEC  x CSTBDF^  (101 x C3H)F1t or (C3H x  C57BL)F1



females.   The females  were checked  for vaginal plugs  each morning until  the



                                      9-70

-------
8th day  post-treatment and  were killed  for uterine  analysis  12  to  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  incidence of postimplantation deaths.




However,   no significant  difference was  observed  when (101   x  aSH^-treated




males were mated to females of different stocks.



     Appelgren  et  al.  (1977)   studied  the  whole-body distribution  of radio-




labeled EtO in  mice and reported the  results of a  dominant lethal  test.   Male




mice were treated with [   C] ethylene  oxide (sp. act. not given) by inhalation




or  intravenous  (i.v.) injection.    The  animals  were  later sacrificed  and




autoradiograms  of midsagittal  sections  were prepared.   The autoradiograms from




mice  that  inhaled  EtO 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  compound  accumulated  EtO.   In




experiments conducted  using  the i.v.  route  of  administration,  EtO  was present




in  the  gonads  (epididymis  and  testicle)  20   minutes  after administration.




Radioactivity  was  still  present  in the  epididymis 24 hours  after injection.




These  observations  that  EtO  reaches  the  gonads  are   consistent   with  the




positive  dominant  lethal  responses  reported   by  Embree  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 i.v. injection in the  study by




                                      9-71

-------
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  attributed 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 EtO reaches the germinal  tissue in intact




mammals and causes genetic damage.   Although these  tests  do not unambiguously




demonstrate heritable effects  caused  by  EtO,  the  positive heritable  trans-




location 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  EtO  to  cause  heritable




mutations in intact mammals.








Heritable Translocation Test




     In conjunction  with their study of dominant  lethal  effects,  Generoso  et




al. (1980)  tested  EtO  for its ability  to cause  heritable translocations  in




mice (Table 9-10).   T  stock  male  mice were given 0,  30,  or 60 mg  EtO  per  kg




once daily, weekdays,  for  5 weeks.   Immediately after  the last  injection each




male was caged with  three  (SEC  x C57BL)F1  females.   After one 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




EtO causes heritable chromosomal mutations  in whole mammals.




                                     9-72

-------
Chromosome Aberration Tests



     The ability of EtO  to  cause  well-defined chromosomal aberrations (breaks,




rings,  inversion,  translocations,  etc.)  has  been  studied by  several investi-




gators.  Some  of these studies have been  discussed  previously.   These include




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 to 0.003 rag/liter or




from 0.030 to 0.060 mg/liter EtO (purity not  given)  by  inhalation for 2,  U,  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




to  11.6J5) 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  EtO  per os_ in aqueous  solution  resulted  in an increased  incidence  in




total aberrations  in  bone  marrow  cells  scored  21  and,  to a  lesser extent,  48



hours  later;   the  vague  manner  in  which the study  is  reported,  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.
                                      9-73

-------
     Poirier  and  Papadopoulo  (1982)  exposed  F1   cells  (derived  from  human




amnios)  to EtO (commercially available  from Matheson Gas Products) at  5,  7.5,




and 10 mM  for 1  hour.  The corresponding cell survivors was  58,  25,  and 9.2$,




respectively.   Three separate experiments  were performed.   After  harvesting




(at 48, 72,  0196h)  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  'centric1  rings)  per 100




cells was 5.9, 10.6, 56.7, and 127.3  for  the corresponding  treatments of 0,  5,




7.5, and 10 mM EtO/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




aberrations  in  peripheral  lymphocytes  of  male Cynomolgus  monkeys  (Lynch  et




al., 1982;  Dr. D. Lynch, personal communication  1983).  The response  was dose-




related;   roughly  four-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.








Micronucleus Formation




     Three studies  addressed  the ability of EtO to induce  micronuclei (Table




9-12).  Appelgren  et al.  (1978)  treated  NMRI mice  by  i.v.   injection with two




doses of EtO  ranging  from  50 to  300  mg/kg,  30 and 6 hours before sacrifice and




Sprague-Dawley rats  according  to the same  regimen  with  doses up to  200 mg/kg




EtO.   Mice given  300 mg/kg died  after  the  first  injection.  Rats  given 200




mg/kg  died  after  the  second   injection.    In  mice,  EtO  caused  a  highly




significant dose-related  increase  in micronuclei.   At the  highest  dose there




                                      9-74

-------
were 2.48$  polychromatic  erythrocytes  (PCE)  with micronuclei compared to 0.52%




PCE  with  micronuclei in  the negative  control  animals  (P<0.001).    Rats  also




exhibited a  statistically significant  increase in micronuclei,  but  it was not




shown to be  dose-related.   Toxicity  of the bone marrow confounded the results.




The  mid-dose level  caused  1.08$  PCE  with  micronuclei  compared to  0.49$ PCE




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  EtO,  or its metabolites ethylene




glycol  and   2-chloroethanol,  to cause micronuclei.   Ethylene  glycol  and  2-




chloroethanol were  given  to the experimental  animals via  oral administration




or i.p. injection.   EtO was administered  by i.p.  injection, i.v. injection or




i.p. implantation of gas  sterilized  medical devices.  Implantation  of the EtO




gas   sterilized  medical   devices  did   not   induce  elevated  numbers  of




polychromatic erythrocytes  with micronuclei.  Similarly,  when EtO was injected




i.v. (two injections of 100 mg/kg 24  hours  apart)  and the animals were killed




6 hours after the second  injection,  no statistically significant  increase in




micronucleus  formation  was  observed after treatment.   However, when  EtO was




given  i.p.   a  suggestive  positive  response   was  observed.    In  order  of



increasing  doses  of EtO  (from 0 to  4000 mg/kg  i.p.),  the  percentage  of PCE




with micronuclei ranged from 0.23 to 0.47.




     Jenssen  and  Ramel (1980)  used  CBA male mice  in their  assessment  of the




ability of EtO  to cause micronuclei.   EtO was  administered  i.p.  at  dosages up




to  175  mg/kg,  and  micronuclei  was 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
                                      9-75

-------
(150  and  175 mg/kg)  compared  to negative  control animals  (0.93  + 0.31$  and




0.66 + 0.19% 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  EtO reaches  bone




marrow  and  exerts   a  chromosome  damaging  (breakage  and/or  nondisjunction)




effect on hematopoietic cells of mammals.








CHROMOSOME MUTATIONS IN HUMAN POPULATIONS




     Three  studies  have been  conducted  in which  workers exposed to  EtO  have




been  monitored   for  the induction  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 EtO.  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 matched 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 persons (range =  6  to  26) .   Gross  chromosome aberra-




tions  (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  gaps  to  these values




increased the respective incidences to  30.2$  and  16.5$.   Because  of the small




size of the study population and the low  number of metaphase spreads analyzed,




                                     9-76

-------
the discriminating  power of  the study  is  not great  and,  thus,  the  elevated



levels of chromosome  damage  observed in the  exposed population is  judged  not




to be a significant positive effect.



     Theiss et  al.  U981)  monitored 43  humans exposed to EtO and  to  a lesser




extent  other  alkylene  oxides  for  the presence  of  chromosomal  aberrations




(Table  9-13).    The  workers  ranged from  27  to  63  years  (x  =  47.1  years).




Exposed individuals  were  categorized into  four groups based  on the  type  and




extent of EtO exposure they had received:






     1.   Long-term exposure (more than 20 years),  11 men.




     2.   Less than 20 years exposure,  6 men.




     3.   Long-term exposure plus accident, 21 men.




     4.   Accident (i.e., short-term high exposure to EtO), 5 men.




Subjects in the first three groups  worked  in  plants  where EtO was manufactured




or  processed.    Personnel   in  the fire  department  or  maintenance  workers




comprised the fourth  group.   The  negative  control group  included  male 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  to 58




years  (x  =  38.6).    The  work  place was  monitored  for  EtO  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  ppm, but  higher concentrations  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
                                      9-77

-------
and  1.    Based  on  Fisher  exact  test  analysis  of  the  data,  with  Yates




correction, significantly increased incidences of chromosomal  aberrations 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  individuals were




subsequently  examined  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  EtO for  more  than  20  years  suggests a  mutagenic  effect.   However,  the




results do not conclusively  indict  EtO  as  the causative  agent,  because  the




workers  were   exposed   to   other   substances   (e.g.,   ethylene  chlorohydrin,




ethyleneimine,  propylene  oxide,  etc.)  which  may have caused or  contributed 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




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  Theiss  et  al.   (1981)   one   of the  basic




assumptions   of  the   Fisher-Yates   test,  that   of   independence   of  the




observations, would not  be  met.   A more  appropriate statistical test,  and one




which  the   authors  claimed   to  have  used (but  have  not  reported)  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 aberrations.




                                     9-78

-------
      Pero   et   al.   (1981)   also  found  increased  incidences  of  chromosome




 aberrations  in  factory  workers exposed to EtO (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? EtO  and 5Q%




 methyl  formate  gas  (0.5 to 1.0 ppm EtO) 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  chromosome gaps  plus




 chromosome  Dreaks  was observed in cells from  the sterilizer EtO-exposed group




 (5  workers)  compared to  the  control  group  (9  workers),  11-14/6  in  exposed




 groups  compared  to  8.5%  in  controls, (P<0.05).   However,  with  respect  to




 breaks  alone,   a  nonsignificant  (or at  best  only  a  marginally significant)




 increase was noted in  the comparison  between sterilizers 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.




     The   increased  incidences   of  chromosome  aberrations   in  peripheral




 lyrapocytes  noted  in  three  studies of  workers exposed  to  EtO  are consistent




 with  one another  and  with  the  experimental  animal  data showing  EtO  to  be




 clastogenic.  They indicate that similar effects are caused in humans as well.








 OTHER STUDIES INDICATIVE OF MUTAGENIC DAMAGE




     Additional studies  have been  conducted  bearing  on the  genotoxicity of EtO




(Tables 9-14 to 9-16).   These studies do  not measure  mutagenic  events per  s_e_




in  that   they   do   not   demonstrate   the  induction   of   heritable   genetic




alterations, but  positive  results in  these  test systems do show that  DNA has




                                      9-79

-------
                                                                    TABLE 9-11

                                   Summary of Mutagenioity Testing of EtO:  SCE Formation  In Human Populations
   Reference
Test
System
Chemical
Information
                                                               Results
                                                                                                                  Comments
   Johnson and   Sister chromatid
   Johnson, 1982 exchange
                 induction and
                 chromosome
                 aberrations:
                 Industrial
                 workers
                    Inhalation  exposures
                    estimated to  be:  Low
                    relative exposure
                    (1  ppm), moderate
                    relative exposure
                    (1-10 ppm), high
                    relative exposure
                    (5-200 ppm).
                           Dose-response association
                           suggested
                      1.  Levels of SCE remained elevated after
                          termination of exposure.

                      2.  Environmental exposure to EtO causes increased
                          SCE formation.

                      3.  Report based on preliminary data from
                          relatively small sample population.
                                                               SCE
                                                        Months after Exposure
                                                            0           6
                                                                 Chromosome  Aberrations
                                                                 Months  after  Exposure
                                                                     0            6
VO
oo
o
                    High  Potential
                    Exposure

                    Low Potential
                    Exposure
                       33
35
                                                                       15
1.5
                                                                                     1.1
                                                                                                 0.9
Inside Controls 12
Outside Controls
12
8
0.6
0.78
0.5
   Garry et al., Sister  chromatid
     1979       exchange  induction:
                peripheral blood
                lymphocytes
                collected from
                hospital workers
                    Maximum  exposures
                    estimated  to be 36 ppm
                    (from  average measure-
                    ments  over one 8 hour
                    period).   Workers
                    divided  into groups
                    based  on known exposures
                    to EtO and symptoms
                    indicative of exposure.
                           Statistically significant
                           increases in level of SCE
                           observed in exposed
                           individuals compared to
                           controls (unexposed
                           laboratory personnel).
                      1.   Air dried fluorescence plus Glemsa
                          chromosome preps.

                      2.   20 metphases scored/individual.

-------
                                                              TABLE 9-11 (cont.)
Reference
Test
System
Chemical
Information Results
Comments
Yager, 1982  Sister chromatid
and Yager    exchange  induction:
et al., 1983 peripheral blood
             lymphocytes
             collected from
             hospital  workers
vo
oo
Exposures determined by
individually monitoring
workers.   High exposure
group received a cumula-
tive dose MOO mg
while cumulative dose for
low exposure group was
<100 mg.
                                                         Group
                                       Mean
                                       Exposure (mg)
                      Control

                      Low exposure

                      High exposure
                                                                               13

                                                                              501
                              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.
                                                                                                SCEs/
                                                                                                cell
                               7.56 + 1.01

                               7.76 ± 1.05

                              10.69 + 1.92
Laurent      Sister chromatid
et al.,      exchange induction:
1982         peripheral blood
             lymphocytes collected
             from hospital
             workers.
No exposure estimates
Exposed group had
statistically significant
increase in SCEs compared
to control group was range
of SCEs for the exposed group
was 9.61 - 17.57 compared to
a range of 7.01 - 8.52 for the
control group.
1.   Control group may not have been matched for
    age, sex, and personal habits to the exposed
    group.

-------
                                                                     TABLE 9-15

                                   Summary of Mutagenioity Testing of EtO:  SCE Formation in Experimental Studies
Reference
Test
Sys tern
Chemical
Information Results
Comments
   Star,  1980
Sister chromatic!
exchanges: Cultured
human fibroblasts
Concentrations tested:
0 to 3600 ppm and
residues from plastic
children's endotracheal
tubes treated with 1100
mg/cm3 of pure EtO for
90 minutes followed by
aeration from 24 to 96
hours after sterilization.
                                                              Toxic as well as mutagenlc.   1.
                                                              Significant increases in SCE
                                                              induction at 36 ppm.  Cyto-
                                                              toxicity at 180 ppm and       ?.
                                                              higher
Cultures from skin biopsies used between
fifth and tenth subculture.

Insufficient data presented to evaluate
conclusions.
00-
                                      Source:   STERI-Gas  cartriges
                                               3M Germany GmbH,
                                               Neuss

                                      Purity:   Not given

                                      Solvent:  Dulbecco's Modified
                                               Eagle's  Medium
Yager and    Sister chromatid
Benz, 1982   exchange induction:
             New Zealand white
             rabbits
                                      Concentrations  tested:
                                      0,  10,  50, and  250 ppm
                                      by  Inhalation

                                      Source:   Matheson
                                               Dayton, OH
                                                 Positive response at 50 and
                                                 250 ppm exposures
                                                             Increased SCE levels decreased after exposure
                                                             ended but still remained above baseline
                                                             levels 15 weeks after exposure.
   Kligernan    Sister  chromatid
   et  al.,  1983 exchange  induction:
               CDF  rats
                      Concentrations tested:
                      0, 50, 150 and 450 ppm
                      for 1 or 3 days by
                      inhalation
                           Dose and time dependent
                           positive response
                                      Source:
                                              Matheson Gas
                                              Product
                                      Purity:   99.7J
                                                                Concentration
                                                                   50 ± 7
                                                                   140 ± 17
                                                                   144 + 33
                                                                                            1.  Significant increases at 50 ppm show effects
                                                                                                induced at levels to which workers have been
                                                                                                exposed.  Until recently TWA was 50 ppm.

                                                                                            2.  Data for 3 days exposure groups shown.
                                                                       SCEs/
                                                                     Metaphase

                                                                      7.5 ± 0.5
                                                                      9.1 ± 1.3»
                                                                     10.3 ± 1.3s
                                                                     13.6 + 1.3»
   •Significantly different  from controls by one-tailed Dunnett's  test

-------
                                                                 TABLE 9-16

                                     Summary of Mutagenlcity Testing of EtO:  Unscheduled DNA Synthesis

Reference
Cunning
et al.
(in press)




Test
System
Unscheduled DNA
synthesis:
testicular DNA of
(101 x C3H)Fi
mice


Chemical
Information
Concentration tested:

a. 600 and 800 ppm for
2, U, 6, or 8 hours.
[3H] dThd administered
intratesticularly
immediately after
administration

Results Comments


a. Dose-dependent increase
in UDS over lower range
of doses tested (e.g.,
70 dpm/106 cells,
18 dpm/100 cells, and
8 dpm/106 cells for
VO
oo
CO
b. Same as above except
   [3H] dThd administered
   at different times
   after termination of
   exposure

o. 300 and 500 ppm 8 h/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-methyl
   chloranthrene, 6 animals
   drank water with 1 mg/mH
   sodium phenobarbital for
   1 week prior to exposure,
   6 animals uninduced
   controls

Source:  Matheson Co., East
         Rutherford, NJ

Purity:  99.1%
                                                                   800 ppm, 600 ppm, and
                                                                   negative controls at
                                                                   4 hours)

                                                                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
                                                                d. UDS response dramatically
                                                                   reduced in animals receiving
                                                                   mixed-function oxidase
                                                                   inducers
  Pero et  al., Unscheduled DNA     Exposure levels: 0.5 to
  1981         synthesis: Human    1.0 ppm in air
              lymphocyte cultures
                              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 EtO-exposed workers.  UDS
   peaked at 2 mM exposure NA-AAF.

-------
been  damaged.    Such  test  systems  provide  supporting  evidence  useful  for



qualitatively assessing genetic risk.








SCE Formation in Human Populations




     Three  studies  have  been  reviewed  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 EtO 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 was increased




(19-1$) compared  to the unexposed control group (1U.6J).   Although  the small




sample  size  and  uncharacterized  exposure  these workers  received  preclude  a




definitive  assessment of  the  ability  of  EtO  to  cause  SCEs  in  humans,  the




results are  considered to indicate  genetic toxicity in somatic cells  of the



exposed workers.




     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




categories 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 EtO 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 presumably




                                     9-8H

-------
unexposed workers  at  Plant III were higher  than those of other control groups




available for  comparison at  the  time  (12/metaphase  compared  to 7/metaphase).



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.




     The  preliminary  analysis of  data  indicates  a  consistent dose-response




trend at Plant III  for SCE induction both at  an original monitoring and later




after  6 months  of no  further EtO  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/metaphase 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




potentially exposed and  control groups.   Analysis of the  chromosome aberration




data  suggests  a  dose-related  increase  in  damage,  but  the  magnitude  of




differences  between groups  is not great.    Thus,   it appears  that  a  dose-




response association  exists  between  exposure  to EtO  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  EtO.    However,  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  EtO  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  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 irritation had statistically




                                     9-85

-------
significant  increases  in  the  incidence   of  SCEs  compared  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  hospital  workers  (14)


exposed  to EtO.    Thirteen persons  not  exposed  to   EtO  served  as  matched


controls.    Cumulative  exposure  doses  during the  6  months prior  to  blood


sampling were  estimated by monitoring air concentrations during defined  tasks

                      TM
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  low  exposure  dose  group  (13 ± 18 mg EtO) or the high exposure


dose  group (501  +_ 245  mg  EtO).    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 EtO.    Ten  persons in good health and  not exposed 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 EtO exposed  workers


ranged between  23  and 51 years.  No estimate was  made of the exposure received


by the sterilizers but they had a significantly elevated  level of  SCE compared


to the controls (13-02 + 2.294  vs.  7.86 + 0.479).


     The increased incidences  of  SCEs  observed  in  five   groups of  workers


exposed to EtO  do  not demonstrate mutations but do  indicate that EtO can cause


genotoxic effects in somatic tissue of humans in vivo.
                                      9-86

-------
SCE Formation in Experimental Studies



     Human cells in  culture  also exhibited increased SCE levels after exposure




to EtO  (Table 9-15).   Star  (1980) exposed  skin  fibroblast cells  from normal




healthy human  tissue biopsies from 0 to  3600 ppm EtO or to plastic children's



endotracheal  tubes  sterilized  with 1400  mg/cm3  EtO  at 55°C  for  90 minutes




followed  by  aeration in room air  for varying times from 24 to  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  EtO concentrations  ranging  from 12 to 800 ppm  as  estimated by gas




chromatography  of  head  space   material.    Excessive  cell killing  precluded




scoring SCEs above 600 ppm  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 SCEs  was  noted  in this part of




the study at doses >217 ppm.   In  the other set of  experiments a statistically




significant  increase  in   SCE  induction  was  reported  at  36   ppra.    However,




insufficient  data  are 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 as  little  as  10 ng/mS,  (in  the




media) during  a 20-minute exposure period.   A dose-related increase was noted




up to EtO concentrations  of  35  (ig/rni  (the highest dose tested).   At this dose




there  were  about  20  SCEs/cell  compared  to control   levels  of  roughly  5




SCEs/cell.
                                      9-87

-------
     Yager (1982)  and Yager and  Benz (1982)  administered  from  10 to 250  ppm




EtO  gas   to  four-month-old  male  New  Zealand  white rabbits  via  inhalation.



Eight animals were  placed  in each exposure chamber  and  exposed  6  hours/day,  5




days/week, for  12  weeks.  Blood  samples  were obtained  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  (i.p.)




injections 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  (9.47 + 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 after  exposure (8.45 + 0.30).   Hematological and  GSH measurements from




the animals did not differ from controls.




     After exposures  to  EtO  of 0,  50, 150 or 450 ppm 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-bromodoxyuridine and scored for  SCEs  and  chromosome breakage




(Kilgerman  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 to  50 ppm  (9.1  +




                                      9-88

-------
 1.3)  showing effects at levels  to  which workers have been exposed.  There was



 no significant reduction in mitotic activity or  slowing  of  cell kinetics.








 Unscheduled  DNA Synthesis




      Gumming et al.  (in press) tested EtO for its ability  to  cause UDS  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 EtO




 (Matheson  Co.).    In the  first experiment,  the  effect of  differential  time




 exposures  on UDS  induction  was assessed.   Animals were  treated  with 600 and




 800  ppm  EtO  from  2 to 8 hours,  after which  exposed animals were  anaesthetized




 with  metofane and injected  intratesticularly  with [ H]thymidylic acid  (dThd).




 A  dose-dependent  increase in  UDS 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  than at 600  ppm (e.g.,  70 dpm/10  cells  for 4-hour exposure at  800 ppm




 compared to  48 dpm/10   for 4-hour exposure at 600 ppm;  controls incorporated 8




 dpm/10   cells).  Due to the toxicity of  EtO  at 800 ppm it was only possible to




 measure  up   to  6  hours  exposure  for this   concentration.     In  a  second




 experiment,  EtO  administration  was  the  same   as  above,  but   [ H]dThd  was




 administered  to the  animals  at different times after removal  from EtO exposure




 to characterize the  UDS response at different times  after  treatment.   UDS was




 found  to increase  with time  to a peak  2 hours  after the  end of the exposure




 period  and  to  fall  afterwards.   Two  additional  sets  of   experiments  were




 performed.   The first was  a  work week exposure regimen  of  (300 and 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  i.p.  injection of  80




mg/kg  3-methylcholanthrene or  administration  of  drinking water  containing  1






                                      9-89

-------
mg/mi phenobarbital  for  1  week prior  to  EtO treatment).  Concerning  the  work


week 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  EtO


exposed workers with  10  mM  N-acetoxy-2-acetylaminofluorene  (NA-AFF)  for 1  hour

                                                  3
and subsequently  measured  the incorporation of  [ H]  thymidylic  acid  into  DNA


to  detect  unscheduled DNA  synthesis  (UDS)   (Table  9-16).  NA-AAF-induced  UDS


was found  to  be  inversely  related  to the duration  of worker exposure  to  EtO


and to  the  number of chromosome breaks observed.   This suggests  an inhibition


of  the  cellular DNA-repair capacity  by EtO.  Biochemical and  autoradiography


studies  were  consistent  with  this  response.  When  NA-AFF-treated lymphocytes


were exposed  to EtO,  it  was found that  concentrations  above  2 mM  resulted in


inhibition of UDS.


     As  was  the  case  for  the studies  of sister chromatid exchange induction


these results do  not  show that EtO  is mutagenic  but  do  indicate it  causes


damage  to  DNA  and  are  consistent  with  the  results  showing  the  EtO  causes


mutations.




SUMMARY AND CONCLUSION ON THE MUTAGENICITY OF ETHYLENE OXIDE


     Ethylene oxide  (EtO) 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.



                                      9-90

-------
It is  therefore a  direct-acting  mutagen.   Strong positive responses were found




in  bacteria (10  to 18-fold  increase over  negative  controls),  higher plants




(33-fold  increase),  and mammalian cells  in culture (2  to  20-fold increases).




Less  strong,  but  clearly  positive,  responses  were  found in  Drosophila (two-




fold  to  three-fold increases).   Based on  these positive findings in different




test  systems  in a wide range of  organisms, EtO  is  judged  to be  capable  of




causing gene mutations.




     EtO  has  also  been shown to  be clastogenic,  in  that it  causes 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,  EtO 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 EtO  reaches  the gonads.  This




result  is  consistent with evidence  that  EtO 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).   EtO can therefore be regarded as  mutagenic both in  somatic  cells  and



in germ cells.




     Based  on the  available data,  there  is overwhelming evidence that EtO is a




direct-acting mutagen  that has  the potential to cause mutations  in  the cells




of exposed  human  tissue.   The observations that EtO 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  EtO in  humans  also  results   in  its




                                      9-91

-------
distribution to the DNA of germ  cells.   Thus,  EtO should be considered to be a




potential human mutagen.
                                      9-92

-------
9.5  CARCINOGENICITY




     The purpose of this section is to evaluate the likelihood that ethylene




oxide (ETO) is a human carcinogen and, on the assumption that it is a human




carcinogen, to provide a basis for estimating its public health impact and




evaluating its potency in relation to other carcinogens.  The evaluation of




carcinogenicity depends heavily on animal bioassays and epidemiclogic evidence.




However, other factors, including mutagenicity, metabolism (particularly in




relation to interaction with DNA), and pharmacokinetic behavior, have an impor-




tant bearing on both the qualitative and the quantitative assessment of carcino-




genicity.  The available information on these subjects is reviewed in other




sections of this document.  The carcinogenicity of ETO has also been evaluated




by the International Agency for Research on Cancer (1976).  This section presents




an evaluation of the animal bioassays, the human epidemiologic evidence, the




quantitative aspects of assessment, and, finally, a summary and conclusions




dealing with all of the relevant aspects of the carcinogenicity of ETO.




9.5.1  Animal Studies




     Only a few studies have been conducted to assess the carcinogenicity of ETO.




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 ETO-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
                                        9-93

-------
 tumors.  All males exposed to ETO-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  author  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 develop-




 ment may be  due  to immune suppression.  At  present, however,  there  is  no  evidence




 to support these hypotheses.




      Dunkelberg  (1979)  studied the  oncogenic activity of ETO  dissolved 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 ETO 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 ETO 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 sacrified.   No increase in tumors at remote sites was  observed.




     Lifetime skin painting studies with 10% ETO in acetone (three  times




weekly) were performed on 30  female mice by Van Duuren et al. (1965).




Application of 0.1 mL of ETO  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.







                                         9-94

-------
 9.5.1.2  Rats—Walpole  (1958) injected 12 rats subcutaneously with a maximum




 total ETO 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 ETO administered and




 the frequency of injection were not specified, it is difficult to evaluate this




 negative result.




     Dunkelberg (1982)  administered ETO 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 with




 B-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 •  ETO 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.




 ETO did not induce tumors at sites away from the point of administration.




 Survival decreased in the positive control group.




     Two other studies designed to test for chronic toxicity of ETO reported




no tumors;  however,  the exposure and observation periods were too short to




 adequately  test the  carcinogenic!ty of ETO 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 inhalation




study  (unpublished)  was performed by Bushy Run Research Center, Pittsburgh,
                                       9-95

-------
     TABLE  9-17.  DESIGN SUMMARY FOR CARCINOGENICITY TESTING OF ETO BY
            INTRAGASTRIC ADMINISTRATION TO SPRAGUE-DAWLEY RATS
                         (adapted from Dunkelberg 1982)
Group
Ethylene oxide I
Ethylene oxide II
Oil (vehicle)
Untreated
£ -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
TABLE4'18 .  TUMOR INDUCTION BY INTRAGASTRIC ADMINISTRATION OF ETO IN FEMALE
                              SPRAGUE-DAWLEY RATS
                         (adapted from Dunkelberg 1982)
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.
Pennsylvania (Snellings et al. 1981).  Fischer 344 rats were exposed to 100, 33,

and 10 ppm of ETO 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.
                                       9-96

-------
 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, urinalysis, 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 chromosomal




 aberrations per cell," or the "total number of chromosomal aberrations (per rat)"




 for either males or females exposed to ETO at 100 ppm when compared with values




 obtained for the air-control groups.  However, statistically significant




 chromosomal aberrations have been found in other ETO studies (see section on




 mutagenicity).




     Histopathologic examination was performed on all tissues of each air-




 control group and the 100 ppm group at 6 months and at 12- and 18-month necropsy




 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
                                       9-97

-------
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 I, and Air Control II groups, respectively.  One additional male in




the 33 ppm group and one female in Air Control Group I were accidentally killed.




     The cumulative mortality data and statistical significances for male and




female rats are shown in Tables 9-19 and 9-20t  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 9-21an(j 9-22 f indicate a significant
                                     9-98

-------
       TABLE 9-19.  CUMULATIVE PERCENTAGES OF MALE FISCHER 344 RATS THAT DIED
       OR WERE SACRIFICED IN A MORIBUND CONDITION AFTER EXPOSURE TO ETO VAPOR3
                       (adapted from Snellings et al. 1981)
Exposure
month 100 ppm^
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
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
27.3(-f-»->
44.2(a,c,c)
50.7(a,c,c)
55.9(a,b,c)
65.2(a,-,b)
Exposure concentration
Air
33 ppin^ 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)
36.0(-.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).
       0.05 > P > 0.01
0.01 > P > 0.001
c = P < 0.001
not significant
                                          9-99

-------
        TABLE 9-20 .   CUMULATIVE PERCENTAGES OF  FEMALE  FISCHER 344  RATS THAT  DIED
        OR WERE SACRIFICED IN A MORIBUND CONDITION AFTER  EXPOSURE  TO  ETO VAPOR*
                         (adapted from Snellings  et  al.  1981)
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»
18.o(b.
21.1
25.o(a>
30.4(b,
34.4(b>
41.3 P > 0.01
b = 0.01 > P > 0.001
P < 0.001
                                        9-100

-------
 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 ETO VAPOR3
                        (adapted from Snellings et al.  1981)
Exposure concentration
Exposure Air
month 100 ppm^ 33 ppm 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,c) 29.8<-.a,->
46.9(-»c»b) 34.1
52.5O,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
                                        9-101

-------
  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 ETO  VAPOR3
                        (adapted from Snellings  et  al.  1981)
Exposure
month      100 ppm^       33 ppm
      Exposure concentration

                     Air         Air         Combined
         10 ppm    Control I   Control II    controls
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(a,-,b)
55.2(c,c,c)
63.4(c,c,c)
1.1
4.7
10.2
11.6
17.1
19.9
28.2
31.2
37.4
2.1
3.2
8.6
8.6
9.9
11.2
21.8
26.2
32.6
0.0
2.3
5.0
6.3
6.3
6.3C
15.5
19.8
22.9
0.0
1.2
2.6
9.4
13.6
16. 3C
20.4
23.4
23.4
0.0
1.8
3.8
7.8
9.8
11.2
17.9
21.6
23.2
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
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.
                                         9-102

-------
 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,




 five 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




 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 ETO 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 sub-




cutaneous fibromas was observed in the 33 ppm group.  The authors concluded
                                     9-103

-------
 TABLE 9_23-  SUMMARY OF SELECTED TUMOR INCIDENCE COMPARISONS FOR MALE AND FEMALE
                 FISCHER 344 RATS EXPOSED TO ETO FOR TWO YEARS
                      (adapted from Snellings et al. 1981)
Ethylene oxide
concentration
    ppm
        Total number of rats
With tissues examined      With tumor3
Time in months to:
First       Median
tumor       turner^
Mononuclear cell leukemia - Males
100
33C
IOC
0-1
O-II
100
33C
10C
0-1
O-II
100
33C
IOC
0-1
O-II

100
33^
IOC
0-1
O-II

100
33C
IOC
0-1
O-II
119
81
79
116
118
26
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
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
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).
^Medians were not presented if the total number of a particular tumor was
 three or less.
  a « 0.05 > P > 0.01    b = 0.01 > P > 0.001    c = P < 0.001   - = not significant

C0nly organs with gross lesions were histologically examined from this exposure
 level at the 6-, 12-, and 18-month sacrifice intervals.
                                        9-104

-------
         TABLE 9-24  •  ETO 2-YEAR VAPOR INHALATION STUDY:  24-MONTH FINAL
             SACRIFICE FREQUENCY OF EXPOSURE-RELATED NEOPLASMS FOR
                     110- TO 116-WEEK-OLD FISCHER 344 RATS
                      (adapted from Snellings et al. 1981)
Organs/Findings/Sex   100a
              ppm of Ethylene Oxide

       33a            10a       Control I   Control II
Total number
examined grossly
  Male
  Female

Pituitary
  Adenomas
  Male

Pancreas0
  Adenomas
  Male

Subcutis^
  Fibromas
  Male

Peritoneum
  Mesotheliomas
  Male

Spleen
  Mononuclear
  cell leukemias
  Male

  Female
                       30
                       26
12/29b
                      5/30
        39
        48
                                   13/39
       1/2
                               51
                               54
                     15/51
                              2/3
 48
 60
49
56
16/48     13/49
10/28(c,c,c)   ^34
                      4/30
       4/39
 8/30         10/39

15/26b>
                              2/51
                      9/51
 2/48
                                                                  1/44
 1/48
5/49
                                               2/47
1/49
                                                                  5/48

                                                                  5/60
           8/49

           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).
      0.05 > P > 0.01
0.01 > P > 0.001    c = P < 0.001   - = not significant
^Numerator equals number of rats with specified finding.   Denominator equals
 number of rats for which specified tissues were examined.
cTissues 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).
                                         9-105

-------
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 not  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 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




II 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
                                        9-106

-------
 TABLE  9-25.  ETO 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
                      (adapted from Snellings et al. 1981)

                                           ppm of Ethylene Oxide

Organs/Findings/Sex   100b           33b           10b      Control I   Control II
Pituitary
Adenomas
Male
Pancreas^
Adenomas
Male
Subcutis6
Fibromas
Male
24/79^ 16/79
ll/80(b»~»a) 1/43
15/78(c,b,c) 3/75
26/79 24/79
2/32 2/80
10/77(b>a,b) 1/76
19/78
5/80
3/78
Peritoneum
  Mesotheliomas
  Male               21/80(c»c»c)   6/80(-»~»a)    3/30          1/80       2/80

Spleen
  Mononuclear
   cell leukemias
  Male               25/80         23/80          21/80         20/80      18/80

  Female             27/80(c,a,c)  24/80(b»a,b)   14/30          9/80      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).
                                       9-107

-------
 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 9-2$nd 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 ETO.




     An  increased frequency of mononuclear cell leukemia was  observed in




 the ETO-treated animals at the 24-month sacrifice interval (Table 9-24).




 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
                                       9-108

-------
 trend  (P < 0.01).  The  trend  became even stronger (P < 0.00001) when the




 proportions were adjusted  for early mortality.  These data suggest that




 exposure to ETO 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 (Tables 9-26 >




  9_27  » and  9-32 )•  These tumors were shown to be statistically significant by




 the Fisher Exact Test in both males and females.




     In summary, ETO 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 accelerated in




 female rats exposed to 100 ppm, although there was no statistically increased




 incidence of these tumors.  The frequency of peritoneal mesothelioma 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  National Institute for Occupational Safety and Health Inhalation




 Study (Lynch et al. 1982).  Another chronic inhalation study (unpublished




 draft) on ETO and propylene oxide (PO) 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 ETO section of the study will be




discussed.   Male Fischer 344 rats (80 in each group) and 12 male cynomolgous




monkeys were exposed to ETO 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
                                     9-109

-------
           60
                                            Male
                                                           •a
                      20       40        60       80
                          Concentration of Ethylene Oxide, PPM
100
Figure 9.3 .  Percentages  of male  and  female  Fischer 344 rats with histologically
              confirmed mononuclear  cell leukemia at 24-month sacrifice.
              (Snellings et al.  1981)
                                     9-110

-------
            TABLE g_26 •   ETO 2-YEAR VAPOR INHALATION STUDY:  FREQUENCY OF
                    PRIMARY BRAIN NEOPLASMS IN FISCHER 344 RATS
                         (adapted from Snellings et al. 1981)
                                            Exposure level (ppm)


     Sex         100           33            10           0 (CI)             0 (CII)
                              18-month sacrifice3
   Male          0/20          1/20          0/20         0/20              0/20
   Female        1/20          0/20          0/20         1/20              0/20

                              24-month sacrifice3

   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  moribund3
                                (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

                                   Two-year  studyb
(Combined 6-,  12-,  18-,  and  24-month sacrifices and  dead/euthanized  moribund animals)

   Male          7/119b         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

   aNumerator  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.
   cFisher Exact  Test.
                                           9-111

-------
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 carcinogen!city,

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.
          TABLE 9-27  .  gTO 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)
                      (adapted from Snellings et al. 1981)
  Neoplasm type
                                         Exposure level (ppm)
100
33
10
0 (CI)
0 (CII)
Granular cell tumor
Astrocytoma/oligodendro-
glioma/mixed glioma
Malignant reticulosis-
microglioma
Granular cell tumor
Astrocytoma/oligodendro-
glioma/mixed glioma
Malignant reticulosis-
microglioma

1/119
5/119
1/119

1/119
2/119
1/119

1/118
2/118
2/118

1/119
2/119
0/119
Males3
1/119
0/119
0/119
Females8
0/118
1/118
0/118

0/118
1/118
0/118

1/118
0/118
0/118

0/118
0/118
0/118

0/116
0/116
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.

                                       9-112

-------
     An  overall  statistically  significant depression in weight gain was noted




 for ETO-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, continued throughout




 the study.   Survival was also  adversely affected by exposure to ETO, 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.




     With respect  to pathology, the authors report that the livers and spleens




 of  the ETO-exposed rats were the only organs for which histopathologic evalua-




 tions were completed.  While the results are preliminary (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 ETO-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.




     Lynch et al. (1982) also reported that exposure to ETO significantly




 increased the incidence of peri'toneal mesotheliomas.  These tumors were present




 on the tunica vaginalis surrounding the testes and epididymis,  and occasionally




 spread to the peritoneal cavity.  A non-significant increase in pheochromocytomas




was observed in exposed groups (Table 9-29  )•
                                    9-113

-------
                 TABLE 9-28 .  LEUKEMIA INCIDENCE IN MALE FISCHER 344
                          RATS3 EXPOSED TO ETO FOR 2 YEARS
                                (Lynch et al.  1982)
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.
          TABLE  9-29 .  INCIDENCE OF NEOPLASTIC LESIONS IN MALE FISCHER 344
                          RATS EXPOSED TO ETO FOR 2 YEARS3
                                (Lynch et al. 1982)
  Organs/Findings
                                                      Exposure level (ppm)
                Control
              50
      100
Adrenal
  Pheochromocytomas
                  8/78
             14/77
     13/78
Brain
  Gliomas (mixed cell)
Body cavity
  Peritoneal mesotheliomas
Spleen
  Mononuclear cell leukemia
                  0/76
                  3/78
                 24/77
              2/77          5/79
                        (P = 0.032)b
              9/79         21/79
                   (P = 4.94 x 10~5)b
             38/79
          (P = 0.22)
     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.
                                           9-114

-------
     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




 comprehensive  report  is scheduled  to be published within a year.




 9.5.1.3  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  fibromas 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 neoplasms were significantly greater for




 all  three groups when compared to combined 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, 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 ETO at 0,  5,  and 100 ppm for 6




hours per day,  5 days per week began in August 1981.  The final report  is




expected in mid-1984.
                                      9-115

-------
9.5.2  Epidemiologic Studies




9.5.2.1  Joyner (1964)—Joyner (1964) conducted a health evaluation of employees




at an ETO plant in Texas.  The evaluation included a physical examination of 37




male ETO operators, aged 29 to 56, and a similar number of age-matched controls.




The operators were reported to have been exposed to ETO at approximately 5-10




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,




gastrointestinal,  or genitourinary complaints.  The author found that the ETO




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 exposures,




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 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 was reportedly not available for follow-up.
                                       9-116

-------
     Additionally, the author reviewed the medical records of eight persons who




had previously worked as ETO operators for 100 months or more but who had since




been transferred to another division.  Among persons formerly employed as ETO




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 little data was presented in this regard,




however.




     This study is inadequate for use in evaluating the carcinogenicity of ETO




for several reasons.  First, it is primarily a cross-sectional study of ETO




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 amoung 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 and used




ETO.  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 ETO.  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
                                        9-117

-------
control group.  The exposed persons were reported to have been active in the




ETO 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 ETO were




added to the exposed group for the lymphocyte/mm^ comparison (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




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 exposed




group as compared with controls, but this  difference was not significant (P >
                                         9-118

-------
 0.05)  for  either  healthy  individuals  or  the  total group.  The authors  suggested




 that this  lack  of a significant  difference could possibly be attributed  to




 improved ventilation  and  safety  control  in the factory, the small number (17)




 of  healthy  persons in the group  permanently  exposed  (vs. 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 ETO 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 ETO.  The authors indicated that the probability of such an




 occurrence was  small, but its statistical significance was not calculated.  The




 result of the study also suggested that ETO may elevate lymphocyte counts and




 reduce hemoglobin values.




 9.5.2.3  Hogstedt et  al» (1979a)—A follow-up study of these same workers with




regard to mortality and cancer incidence was done by Hogstedt et al. (1979a).




The follow-up period  included the years from 1961 to 1977.   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
                                       9-119

-------
 5 mg/m3 ethylene chlorohydrin,  100 mg/m3 ethylene dichloride, 0.05 mg/m




 3 bis(2-chloroethyl)ether,  and  600 mg/m3 ethylene.  The authors also cited




 the  possibility  that  concentrations up to 1000 times greater than those reported




 may  have  occurred  for short periods of time.  For ETO, the exposure was reported




 to be  probably < 25 mg/m3,  although there were occasional exposures to the




 chemical  at  1300 mg/m3  (odor threshold).  During the 1950s and until 1963, the




 authors  reported that the average air concentration of ETO in the factory was




 probably  10  to 50  mg/m3, although peaks above the odor threshold still occurred.




 Random samples in  the 1970s  showed a range of 1 to 10 mg/m3 for ETO and 10 to




 25 mg/m3  for propylene  oxide, with the latter concentrations occasionally being




 as high as 120 to  150 mg/m3.




     The  study included three subcohorts composed of 66 men who had never taken




 part in work involving  exposure to ETO, 86 intermittently exposed men (maintenance




workers), and 89 men whose work involved full-time exposure.  In the full-time




 exposed group, a total  of 9  cancer deaths were observed while only 3.4 were




expected  (P  < 0.01).  There were no statistically significant differences between




the observed and expected number of cancer deaths in the other two exposure




groups.   Five of the nine cancer deaths seen in the full-time exposed cohort




were either  from cancer of the stomach (three deaths)  or from leukemia (two




deaths).  Deaths from both causes were significantly (P < 0.01) elevated in




comparison with the numbers expected (3 observed versus 0.4 expected for stomach




cancer and 2 observed versus 0.14 expected for leukemia deaths).  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.




Although  the maintenance group showed  no overall excess cancer mortality,  the




cancer deaths that occurred in this group were restricted to cancers of the
                                       9-120

-------
esophagus, stomach, and lymphatic system.  The lymphatic system cancer death




was  from chronic lymphatic leukemia.




     Cases of cancer in surviving subjects were identified by the Swedish




Cancer Registry.  By this method, two cases were identified among full-time




exposed workers (testis cancer and urinary bladder cancer), two cases among




maintenance workers (glottis cancer and prostate cancer), and one case (thyroid




cancer) among unexposed workers.  This raised the total number of cancer cases




(both living and dead) identified during the follow-up period among full-time




exposed workers to 11, with an expected number of 5.9 (P < 0.05).  The expected




number of cases by tumor site was not indicated.   Among maintenance workers




and  unexposed workers, the total numbers of observed cases were raised to three




and  two, respectively.  The expected number of total cancer cases for these two




latter groups was not reported.




     In summary, deaths from cancer of all sites, deaths from stomach cancer, and




deaths from leukemia were each significantly (P < 0.01) elevated among the full-




time exposed cohort.  The total number of malignancies was also significantly




(P < 0.05) 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




ETO exposure.




9.5.2.4  Hogstedt et al. (1979b)—Hogstedt et al. (1979b) reported three cases




of leukemia among workers in a small factory in Sweden between 1972 and 1977, in




a study of a different population than the one studied by Ehrenberg and Halstrom




(1967) and Hogstedt et al. (1979a).  The factory had used 50% ETO and 50%




methyl formate since 1968 for sterilizing hospital equipment.  The number of




persons who worked with the actual sterilization procedure was few, but the
                                        9-121

-------
treated boxes were stored in a hall where 30 women worked.  Because of leakage




from the treated boxes, "the average exposure in the storage hall was actually




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




1500 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 neighbor-




ing rooms or as sterilizing operators.  The expected number of leukemia cases




in this group for the above period would have been 0.2.  This was calculated by




multiplying the person-years of observation by the sex- and age-specific national




leukemia incidence for 1972.




     The first of the three reported leukemia cases was that of a woman who




had begun working in the storage hall in 1966.  In 1972, at the age of 51, she




was diagnosed as having chronic myeloid leukemia, and died in 1977.  The




second case was that of a woman who had begun working in the storage hall in




1968, and in early 1977, at the age of 37, was diagnosed as having acute myelo-




genetic leukemia.  As of July 1978, her leukemia was in complete remission.




The third reported case involved a man who had been the local manager of the




plant since 1965.  It was estimated that his exposure to ETO was 3 hours per




week.  In 1974, at the age of 56, he was diagnosed as having primary macro-




globulinemia, and died in 1976.  The authors stated that the two women had




not been exposed to radiation, benzene, or other leukemia-inducing agents, but




that the man had had occasional contact with benzene.




     It should be noted that primary or Waldenstrom's macroglobulinemia, as was




diagnosed in the plant manager, is not considered a leukemia under the Interna-




tional Classification of Diseases (ICD), Eighth Revision.  As a result, the
                                        9-122

-------
 Carcinogen Assessment Group  (GAG)  requested  clarification from the primary




 author  of the  study on  the classification of  the case of macroglobulinemia as




 a  leukemia case.   In correspondence  to  the GAG  (Hogstedt 1983), Hogstedt stated




 that Waldenstrom1s macroglobulinemia was recognized by most experts in Sweden




 in the  late  1970s  to be a type  of  leukemia,  but was considered by early 1983 to




 be a type of non-Hodgkin's lymphoma.  In the  same letter to the GAG, Hogstedt




 indicated, as  he  had in an earlier letter to  Dr. Peter Infante of the U.S. Occupa-




 tional  Safety  and  Health Administration (Hogstedt 1981), that he and his fellow




 authors, since publication of their  1976 study, had calculated an expected number




 of leukemia  cases  based on Swedish incidence  data for 1968-77.  They found that




 the the expected  number was  0.1, as  opposed  to  the 0.2 that had been calculated




 from 1972 Swedish incidence  data and reported in the Hogstedt et al. (1979b)




 article.  The  probability of the occurrence of  two cases of leukemia (excluding




 the Waldenstrom's macroglobulinemia  case), given the expectation of 0.1, is




 less than 0.01.   (Had the expected number of  leukemuia cases been 0.2, the




 probability would have  been  less than 0.02.)  The probability that two cases




would occur in the group working in  the storage hall (where the two leukemia




 cases worked)  is even lower, however, because the expected number of cases,




0.1, was calculated for  the  entire population of the factory.




     Hogstedt et al. also suggested that the combination of ETO and methyl




formate may produce a special carcinogenic risk, since methyl formate, 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 Environ-




mental  Mutagen Information Center at the Oak Ridge National Laboratory (Stafford




1983) failed to find any literature citations for mutagenicity studies of methyl




formate.
                                      9-123

-------
9.5.2.5  Morgan et al.  (1981)—Morgan et al.  (1981)  conducted a retrospective




study of 767 workers potentially exposed to ETO who  had worked for at least 5




years at the 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  survey  of the plant




(performed in July 1977) showed that the 8-hour time-weighted average exposure




to ETO was "well below" 50 ppm, except in the area around the tank car loading




operations, where readings were as high as 6000 ppm.  Among the 767 male workers




potentially exposed to  ETO 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 were 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 ETO 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 (an SMR of  1050).  In conclusion, it




should be stated that the observed mortalities from  pancreatic cancer and from
*Standardized mortality rate.
                                       9-124

-------
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  Theiss et al. (1982)—Theiss et al. (1982) conducted a cohort mortality




study of 602 persons who had been employed for six 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 cohort




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 9-30 •  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




followed for at least 10 years did, however, approach statistical significance




(P < 0.07) in comparison with those expected based on Ludwigshafen or Rhinehessia-




Palatinate mortality data.




     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-
                                       9-125

-------
    TABLE 9-30-  COMPARISON OF OBSERVED NUMBERS OF CANCER DEATHS IN BASF-AKTIENGESKLLSCHAFT, 1,1'UWIGSIIAFEN  PUNTS
         1928-80 FOR PERSONS HAVING 10 YEARS OF OBSERVATION FOLLOWING EXPOSURE TO ALKYLENE OXIDE WITH  THAT
         EXPECTF.D BASED ON MORTALITY STATISTICS FOR RHINEHESSIA-PALATINATE  1970-75, LUOUIGSHAFEN 19/0-75,
                    AND THE FEDERAL REPUBLIC OF GERMANY 1971-74, BY ICD CODE AND CAUSE OF DEATH
                                         (adapted from Theiss et al. 1982)





ICD No.a Cause of death
140-199C Malignant tumors
151 Malignant tumor
of the stomach
156 Malignant tumor of
the gall bladder
1
I—1 162 Malignant tumor
O\ of the bronchll
\88 Malignant tumor of
the urinary bladder
191 Malignant tumor
of the brain




Observed
deaths
10

2

1


4

1

1

Rhinehes6* la-
Palatlnate
1970-75

No. P-value
—

1.852 0.552

0.201 0.182


3.769 0.520

0.469 0.374

0.071 0.068




Ludwlgs ha fen
1970-75

No.
—

1.765

0.243


3.956

0.532

0.066

P-value
—

0.527

0.216


0.568

0.413

0.064
Federal
Republic of
(ie rmany
1971-74
P-
No. valu
11.816 — b

2.033 — b

	 C 	 C


_-C 	 C

	 C 	 C

__c — c
193-199         Squamous cell
                carcinoma of unknown
                primary site

205             Myelold leukemia

230-239         Tumor of unknown
                character
                                            1
0.743    0.525

0.148    0.138


0.454    0.365
1.047    --<•

0.145    0.135


0.426    0.347
                                                                                                     0.756   0.531
                                                                                                     — c    — c
International Classification of Diseasi's Code, Eighth Revision.
'The 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 1n the cohort based on Federal Republic
 of Germany mortality rates for Individual tumor sites other than stomach and myelold leukemia.

-------
 vation  period  of  10 years  required  for  the  comparison  in Table 9-14 was not




 used  for  this  analysis.  Thus,  in Table 9-31  ,  there were  14  total observed




 cancer  deaths,  as  opposed  to  12 observed  deaths in Table  9-30 •  These results




 are reported in Table  9-31 .  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 significant




 (P <  0.05).  In the 65-74-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 65-74-year-old age group, the text indicated




 that  11 had occurred—a difference  that obviously would function to lower the




 probability of  cancer  deaths.   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




ETO are at an excess risk of death  from cancer.  There was a significant




excess number of cancer deaths in the age group 65-74 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 result is that the alky-
                                      9-127

-------
  TABLE 9-31 .  RELATIVE RISKS OF DEATH FROM CANCER IN THE ALKYLENE OXIDE COHORT
                  AS COMPARED WITH THE STYRENE COHORT,  BY AGE
                       (adapted from Theiss et al.  1982)a
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.


lene oxide workers were exposed to a variety of chemicals in addition to ETO,

some of which are known or suspected carcinogens.   Deaths from cancer of any

particular site were not found to be significantly  (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 of the 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. It should be noted that in

regard to leukemia, for which Hogstedt (1979a, b) had found an association with

ETO exposure, the authors found that for those persons who had had more than 10

years of exposure, one case of myeloid leukemia occurred where only about 0.15

would have been expected based on local mortality data, but this difference was

not statistically significant at P < 0.05.
                                     9-128

-------
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 (ICD code 200-209, 8th




Revision), as well as for other types of tumors, was significantly elevated for




certain job categories (e.g., "service" and "nursing") that included job titles




of personnel exposed to ETO (e.g., hospital central service employees, registered




nurses, licensed practical nurses, and nurse's aides).  Such job categories




were relatively broad in their inclusion of job titles, however, and the results




of the study with regard to a possible association of cancer risk with ETO




exposure must therefore be judged inconclusive.




9.5.2.8  Studies in Progress—Several cohort or case-control studies testing the




association of ETO exposure and the risk of cancer are currently in progress or




about to begin.  A cohort mortality study of approximately 1000 ETO production




workers in the Kanawha Valley, West Virginia,  is currently being conducted by




NIOSH and the Union Carbide Corporation.  The  results of this study will not be




available until at least mid-1984.  NIOSH and  the Health Industry Manufacturing




Association are currently discussing plans for a cohort mortality study of




medical equipment manufacturing personnel who  use ETO as a sterilant.  If the




study is initiated, the results will not be available until at least 1985.




     The U.S. Environmental Protection Agency  is currently funding a case-




control study of cases of cancer of the lymphatic and hematopoietic tissue




among District 1199 of the National Hospital and Health Care Workers Union




to determine if an association exists between  such cancers and occupational




exposure to ETO and/or other substances.  The  study,  which includes 63 cases




and 126 controls, is being conducted by Dr. Jeanne Stellman of Columbia Univer-




sity.  The results are expected to be available in late 1984.
                                       9-129

-------
 9.5.2.9  Summary of Epidemiologic Studies—In summary, three epidemiologic




 studies  of  persons occupationally exposed to ETO found a significant association




 between  ETO exposure and either cancer incidence or mortality.  The study by




 Hogstedt et al. (1979a) found significantly (P < 0.01) increased mortality for




 stomach  cancer and leukemia among ETO production workers.  Hogstedt et al.




 (1979b)  found a significantly (P < 0.05) increased leukemia incidence among




 workers  exposed to ETO used as a sterilant.  The study by Morgan et al. (1981)




 found significantly (P < 0.05) increased mortality from pancreatic cancer and




 Hodgkin's disease.




     Excess mortality from leukemia in the Hogstedt et al. (1979a) 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 and one case of




 chronic myeloid 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 epidemiologic




studies, exposure of the cohort to other chemicals besides ETO was reported to




have occurred or probably occurred.  In the Hogstedt et 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, ETO-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 ETO, but the fact that the study




was conducted at a chemical plant would suggest that exposure to other chemicals




did occur.
                                      9-130

-------
 9.5.3   Quantitative  Estimation




     This  quantitative  section  deals with  the unit  risk  for ETO in air, and




 the  potency  of ETO relative  to  other carcinogens that the CAG has evaluated.




 The  unit risk estimate  for an air  pollutant  is defined as the lifetime  cancer




 risk occurring in a  hypothetical population  in which all individuals are




 exposed continuously from birth throughout their lifetimes to a concentration




 of 1 ug/m^ of the agent  in the  air they breathe.  These  calculations are done




 to estimate  in quantitative  terms  the impact of the agent as a carcinogen.  Unit




 risk estimates are used  for  two purposes:  1) to compare the carcinogenic




 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.




     In the  sections that follow,  the general assessment procedures used by the




 CAG  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 estimates.




 Following this discussion, the  CAG's unit risk calculations and relative potency




 estimates are presented.




 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 contrary, 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 extrapo-
                                        9-131

-------
lation model that relates carcinogen exposure to cancer risks at the extremely




low concentrations that must be dealt with in evaluating environmental hazards.




For practical reasons,  such low levels of risk cannot  be measured directly




either by animal experiments or by epidemiologic studies.  We must,  therefore,




depend on our current understanding of the mechanisms  of carcinogenesis for




guidance as to 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 non-threshold




dose-reponse relationship.  Indeed, there is substantial evidence from mutageni-




city 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 parti-




cularly true at the lower end of the dose-response curve; at higher doses,




there can be an upward curvature, probably reflecting the effects of multistage




processes on the mutagenic response.  The linear non-threshold dose-response




relationship is also consistent with the relatively few epidemiologic studies




of cancer responses to specific agents that contain enough information to make




the evaluation possible (e.g., radiation-induced leukemia, breast and thyroid




cancer, skin cancer induced by arsenic in drinking water, liver cancer induced




by aflatoxins in the diet).  There is also some evidence from animal experiments




that is consistent with the linear non-threshold model (e.g., liver tumors




induced in mice by 2-acetylaminofluorene in the large-scale EDgi study at the




National Center for Toxicological Research, and the initiation stage of the




two-stage carcinogenesis model in rat liver and mouse skin).




     Because its  scientific basis, although limited, is the best of any of the
                                    9-132

-------
 current mathematical extrapolation models, 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 non-threshold

 dose-response relationship at low doses is the linearized multistage model.

 The multistage model employs enough arbitary constants to be able to fit almost

 any monotonically increasing dose-response data, and it incorporates a procedure

 for estimating the largest possible linear slope (in the 95% confidence limit

 sense) at low extrapolated doses that is consistent with the data at all dose

 levels of the experiment.


 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 multistage

 model has the form

                P(d) = 1 - exp [-(q0 + qxd + q2d2 + ... + qRdk)]

 where

                          q± >. 0, i = 0, 1, 2, .... k

 Equivalently,


                  Pt(d) = 1 - exp [(qid + q2d2 + ... + qkdk)]
where
                              Pt(d) = P(d) - P(0)
                                       1 - P(0)
is the extra risk over background rate at dose d.
                                         9-133

-------
     The point estimate of the coefficients q-^, i = 0, 1,2, ..., k, and conse-




quently, the extra risk function, Pt(d), at any given dose d, is calculated 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 GLOBAL79, developed by




Crump and Watson (1979).  At low doses, upper 95% confidence limits on the




extra risk and lower 95% confidence limits on the dose producing a given risk




are determined from a 95% upper confidence limit, q*, on parameter qj.  Whenever




qi > 0, at low doses the extra risk Pt(d) has approximately the form Pt(d) =




q* 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 qj to a value q* such that when the log-likeli-




hood is remaximized subject to this fixed value q* for the linear coefficient,




the resulting maximum value of the log-likelihood Lj satisfies the equation




                             2 (L0 - LL) = 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 confidence limit for




the extra risk calculated at low doses is always linear.  This is conceptually




consistent with the linear non-threshold concept discussed earlier.  The slope,




q*, is taken as an upper bound of the potency of the chemical in inducing cancer




at low doses.  (In the section calculating the risk estimates, Pt(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 experiment,
                                     9-134

-------
 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


                          h


                     X2  - I
                             Nipi  d-Pi)
                         1-1


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 ic" dose group estimated by fitting the


multistage model to the data, and h is the number of remaining groups.  The


fit is determined to be unacceptable whenever X^ is larger than the cumulative


99% point of the chi-square distribution with f degrees of freedom, where f


equals the number of dose groups minus the number of non-zero multistage


coefficients.


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 incidence


rate shows a  statistically significant trend with respect to dose level.  The
                                       9-135

-------
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 geometric




mean of q*, estimates from each of these data sets, is used for risk assess-




ment.  The geometric mean of numbers Aj, A2, ..., Am is defined as






                            (A, x A9 x ... x A ) 'm
     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.  In calculating human




equivalent dosages, it is necessary to correct for differences in metabolism




among species and for the variations in absorption factors involved in different




routes of administration.




     Following the suggestion of Mantel and Schneiderman (1977), 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 2/3 power of the




weight, as would be the case for a perfect sphere, the exposure in mg/day per




2/3 power of the weight is also considered to be equivalent exposure.  In an




animal experiment, this equivalent dose is computed in the following manner:
                                       9-136

-------
      Let



      Lg = 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



simply



                                 m   = s x W1/3
where r is the absorption rate for ETO (assumed to be 1).



     When exposure is via inhalation, as with ETO, dose calculations at experi-



mental 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 ETO 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 equivalent 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
           dh = 2.7 x 5/7 * (70/0.42)173 = 0.35 mg/kg/day for 10 ppm
                                        9-137

-------
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.




9.5.3.1.4  Calculation of the unit risk from animal studies.  The risk associated




with d mg/kg2/3/day is obtained from GLOBAL79, and for most cases of interest




to risk assessment, can be adequately approximated by P(d) = 1 - exp (-q*d).  A




"unit risk"  in units X is simply the risk corresponding to an exposure of X = 1.




To estimate  this value, 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 ETO, 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 jt*1 polynomial coefficient 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




q* will always represent the upper-limit potency estimate for humans.






9.5.3.1.5  Interpretation of quantitative  estimates.  Unit risk estimates based




on animal bioassays are only approximate indications of absolute risk in popula-




tions exposed to known carcinogen concentrations.   This is true for several




reasons.   First, there are important species differences in uptake,  metabolism,




and organ distribution of carcinogens,  a well as in target site susceptibility,




immunological responses,  hormone function,  dietary factors, and disease.  Second,
                                     9-138

-------
 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.  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 estima-




 tions of cancer risks to humans at low levels of exposure should be recognized.




 The CAG 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 regulatory 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 CAG




 for quantitative assessment are consistently conservative in that they tend to







                                       9-139

-------
 result  in  high  estimates  of  risk.  This  conservatism  is  primarily  due  to  the




 CAG's use  of  the  linear non-threshold extrapolation model  in  preference 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 B.




     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 CAG, 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 CAG has  chosen




 the most generally employed  method,  which is also the more conservative of  the




 two.  In the case  of ETO  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 CAG utilizes data from human epidemiologic studies in preference




 to animal  bioassay data.  If sufficiently valid exposure information is available




for a given compound, this information is always used by the CAG in its assess-




ment.   If  the results of such studies show carcinogenic effects, the data are




analyzed to give estimates of the linear dependence of cancer rates on lifetime




average doses (equivalent to the factor BJJ 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  parameters 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, BH,  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 or




 age of exposure and depends only on average lifetime exposure, it  follows that

-------
                                         BH (X,+ X9)
                                PO   A + BH x K!

or

                             RP0 = A + BH (Xi + 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 PQ = lifetime probability of dying of cancer with no or

negligible ETO exposure.  Substituting PQ = A + By Xj and rearranging gives


                               BH = P0 (R - 1)/X2


     To use the above model, estimates of R and X2 must be obtained from appro-

priate 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 section

on unit risk based on human studies.


9.5.3.3  ETO Unit Risk Estimates

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 quantitatively,

for the males.  Both studies had significantly increased dose-related incidences

of peritoneal mesotheliomas 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 ETO vapor via inhalation for 6 hours/day, 5 days/week,  for approximately
                                        9-142

-------
 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  mesotheliotnas  and brain gliomas in males




 in the two highest dose  groups, all of which  exhibited dose-response  trends.




 Table  9-32 summarizes the  pertinent data from this study  that the  CAG has  used




 in calculating potency estimates  for ETO.   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  9-32 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 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 ETO 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




weight^'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 response to that produced by  20.24




mg/kg in the male rat.  As discussed above  and as shown in Table 9-32  , this

-------
VO
                                        TABLE  9-32  BUSHY RUN ETO INHALATION STUDY IN FISCHER 344 RATS.
                              INCIDENCE OF PERITONEAL MESOTHELIOMA AND BRAIN GLIOMA IN MALES, AND MONONUCLEAR CELL
                                  LEUKEMIA AND BRAIN GLIOMA" IN FEMALES BY DOSE AMONG SURVIVORS TO FIRST TUMOR
                                                            (Snelllngs et al. 1981 )

Group
Males
Peritoneal meso./No. examined (Z)c "
P-valuesd •
Brain glioraas/No. examined (Z)8
P-values
Total
P-values
Hunan equivalent dose (nig/kg/day)6 •
Females
Mon. leukem/No. examined (Z)* '
P-values
Brain gllomas/No. examined (Z)R'
P-values
Total
P-values
Human equivalent dose (mg/kg/day)e •

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
10
3/88(3)
1/99(1)
4/88(5)
0.35

14/71(20)
=0.08
1/95(1)
15/71(21)
0.28
air (ppm)
33
7/82(8)
=0.02
5/98(5)
=0.02
12/82(15)
=0.0005
0.94

24/72(33)
0.0001
3/99(3)
27/72(38)
<.0001
0.75


<;"
100 (mg/kg/day)-'
22/96(22)
<0.0001
7/99(7)
-0.002
29/96(30)
<0.0001
2.63

28/73(38)
<.0001
4/99(4)
=0.05
32/73(44)
<.0001
2.11
1.1x10
5. OxlO-2
1.7x10"'


2.9x10-'
4.0xlO"2
3.5x10-'

               3See Table
               *>95Z upper-limit unit  risk estimate.
               cNumber alive at 15 months.
               dFlsher Exact Test vs.  combined  controls  (one tailed).  P-value under controls  is  a  one-sided Cochran-Armttage
                test for a dose-response trend.
               eBased on measured doses in  males  of  20.24 and 2.7 mg/kg b.w. following 6 hours' exposure  to ETO  at  100 ppm
                and 10 ppm respectively.  The animal-to-human dose equivalences are based  on a dose  per surface  area  factor  of
                (70/W )''3, 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.
               'Number alive at 18 months.
               (51'otal number examined less  6- and 12-month sacrifices.

-------
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 9-32  , 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.  Compared


with the usual CAG 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  9-32 , yield a high


value of q* = 3.5 x 10~1 (mg/kg/day)~^, based on total mononuclear cell leukemias
          h

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,


q* = 1.7 x ICTkmg/kg/day)-1.
 h

     To convert the above estimate to units of ug/m3 for humans, the following


formula is used:



  1 mg/kg/day = 1 mg/kg/day x 70 kg x 1000 ug/mg x day/20 m3 = 3.5 x 103 ug/m3


or


                        1 ug/m3 = 2.86 x 10-4 mg/kg/day.



The 95% upper-limit slope estimate in terms of ug/m3 is thus calculated as



  q* - 3.5 x KT1 (mg/kg/day)"1 x 2.86 x 10~4 (mg/kg/day) = 1.0 x
   h                                             ug/mj




     To convert  from ug/m3 to ppm,  the formula is
                                     9-145

-------
                     1.2 g  x 44.1  m.w.  ETO x 106  ug  x 10~6
                     10~3m3    28.2 m.w. air  g



                  = 1.9 x 103 ug/m3



     The lifetime probability of cancer  from continuously  breathing 1  ppm ETO


in air is thus calculated as follows:



           P = 1.0 x IQ-^ug/m3)"1  x 1.9 x 103 ug/m3   = 1.9  x IQ-^ppm)-1
                                         ppm



     9.5.3.3.1.2  Lynch et al. (1982)  (NIOSH).  The NIOSH  study (Lynch et al.


1982) in which male Fischer 344 rats were exposed  to  ETO at  50 ppm and 100 ppm


7 hours/day, 5 days/week for 2 years,  produced results very  similar to those


of the Bushy Run study (Snellings et al. 1981). The  results, shown in Table


qoo  , show statistically significant increases and  dose-response trends in


brain gliomas and peritoneal mesotheliomas; there  is  a significant increase in


mononuclear cell leukemias only at the lower dose, and no  significant  dose-response


trend.  Since incidence of this leukemia in controls  was over 30% in this study,


and since the Snellings et al. (1981)  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 9-33  ,


are quantitatively nearly identical to those in Table 9-32 .  Based on the above


analyses, the maximum animal slope potency value is still  q* = 3.5 x 10-i


(mg/kg/day)~l 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 vs. EPA assessments.  The results of the above assessments depend to some


extent on the dose equivalence assumptions.  Dose  equivalence in the following
                                     9-146

-------
 TABLE  9-33 .  NIOSH ETO INHALATION STUDY IN MALE 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 (mg/kg/day)
                              (Lynch et al. 1982)
                                      Exposure in air (ppm)
                                             50          100       (mg/kg/day)-1
Peritoneal mesothelioma/No.
examined
Brain glioma/No. examined
Total
3/78^
0/76d
3/78<*
9/79
2/77
ll/79b
21/79d
5/79^
26/79d
1.0x10-1
3.4xlO-2
1.3X10-1
Human equivalent dose
  (mg/kg/day)f                    0          1.59         3.06

aSee Table 9-13.
bp < 0.05.
CP < 0.01.
dp < 0.001.
eP-values beside control incidences represent values associated with a one-sided
Cochran-Armitage test for a dose-response trend.
^Human equivalent dose based on transforming ppm to mg/kg/day as in Table
 except for an adjustment for 7 hours' exposure.
                                        9-

-------
discussion means the dose which will cause an equivalent response, quantitatively,

in both species.  The GAG has assumed that doses are equivalent on the basis of

mg per surface area, an assumption for which there is some experimental evidence

when first-order kinetics apply; for ETO, 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 equi-

valence on an 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 ETO at 100 ppm in the

Snellings et al.  (1981) study, using EPA methodology (Federal Register 48[78]:

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 equivalent 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 (Snellings et al.

1981), is that the EPA added total significant tumors (mononuclear 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.
*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.  ETO can be considered a completely
 soluble gas.
                                     9-148

-------
     OSHA did not use human studies in its risk analysis.  As seen in the follow-




ing section, EPA's use of human studies increases this risk factor by an additional




factor of 3.6, so that EPA's final value is larger than OSHA's by a factor of




about 30.  Finally, the estimate based on human data predicts only leukemia




mortality due to ETO exposure.  According to Calleman et al. (1978), alkylating




compounds such as ETO could induce a spectrum of cancers, of which leukemia,




because of its shorter latent periods, would be the first to appear.  Thus,




since OSHA's analysis based on animals predicts risks for all cancers, and




EPA's upper-limit based on humans predicts risks only for leukemias, the possi-




bility exists that even EPA's values are not protective enough.




9.5.3.3.2  Unit risk estimate based on human studies.  In estimating the carcino-




genic potency of ETO on the basis of human data, researchers have focused their




attention on dose-response data for leukemia.  In two studies (Hogstedt 1979a




and Hogstedt 1979b), increased leukemias were evident.  However, one of these




studies (Hogstedt 1979a), described increased leukemias among production and




maintenance workers who had been exposed to multiple carcinogens, including




ethylene dichloride and ethylene chlorohydrin.  For this reason, only the second




study (Hogstedt 1979b) is used in the present analysis.




     The risk assessment done on the basis of the Hogstedt et al. (1979b) study




probably underestimates the carcinogenic potency of ETO 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
                                    9-119

-------
years, an assumption which tends to underestimate the risk.  Another problem




with this study is that the gas used for sterilization was 50% ETO and 50%




methyl formate.  Little is known about the biological effects of methyl formate




or of the combination of methyl formate with ETO.  However, methyl formate is




known to metabolize to formic acid, which is a normal body metabolite.  It is




assumed for present purposes that ETO was the only leukemogen in this study,




although one of the cases (the man) had reported some contact with benzene in




laboratory work.




     Hogstedt (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 ETO 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 employees, we can estimate approx-




imately (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) = 65.7.




     The estimated average exposure to ETO over the lifetime of the workers is




calculated as follows:
                                      9-150

-------
                 20 ppm x 8/24 hrs x 240/365 days x 9/45.6 yrs

                   exposure = 0.865 ppm

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 ETO at 1 ppm is given by


                       ,  . P0(R - 1) Xl
                                 *2


where PQ is the lifetime probability in the U.S. of dying* from leukemia in the

absence of ETO 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 65.7; 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


                    0.0091 (65.7 - 1) x 1 ppm = 0.68 (ppm)"1
               bH =          0.865 ppm


The probability associated with breathing ETO at 1 ppm for a lifetime is


                           P = l-e~bH (1 ppm) = 0.49


To convert ppm to ug/m3, the formula is


                     . 1.2 gm    44.1 m.w. chemical x _10_ ug x jQ~6
                       10-3 m3 x    28.2 m.w. air     gm

                     • 1.9 x 103 ug/m3
*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 incidence,
 with death often occurring from other causes.  Nevertheless, for this assessment
 it is assumed that although ETO 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.

                                       9-151

-------
Thus the unit risk estimate in terms of ug/m3 is

           bfl = 0.68 (ppm)""1 x     1 ppm	 = 3.6 x 10~4 (ug/m3)"1
                               1.9 x 103 ug/m3

This compares with an upper-limit estimate of 1.0 x 10~^ (ug/m3)"^ based

on the Bushy Run animal study (Snellings et al. 1981).  The estimate based on

the human study (Hogstedt 1979b) is 3.6 times as high.

     Because this estimate is based on only two human leukemias,  it raises

questions about the suitability of the human response for risk assessment

purposes.  Furthermore, OSHA in its analysis relied only on the Snellings et

al. (1981) study for its risk assessment.  The reasons the CAG has chosen to

use the Hogstedt (1979b) data are as follows:

     1.  Human data extrapolations are nearly always preferable to animal data

         extrapolations because of species and specific target organ.

     2.  Exposure in this study was actually measured.

     3.  The storage hall employees represented a fairly homogeneous group.

     4.  Both leukemia cases were in young women whose exposures and latent

         periods were less than 9 years.

     5.  Human cancer data are fully supported by animal cancer data showing

         strong dose-response relationships between ETO and leukemia and between

         ETO and other cancers.  Since the human study related only 9 years of

         exposure and follow-up histories, it is quite probable that further

         follow-up will show excesses in other cancers.

     6.  The higher potency estimates in the human cancer study are further sup-

         ported by human data showing a strong dose-response relationship between

         exposure to ETO and the frequency of chromosome abnormalities and sister

         chromatid exchange.  Humans appear to be 100 times as sensitive as rats

         with respect to ETO-caused chromosome abnormalities.


                                       9-152

-------
    Although the qualitative evidence supports ETO as a human leukemogen, the




estimation of potency from only two cases presents enough uncertainty from a




quantitative standpoint that such an estimate can be considered only roughly




approximate at best.  As such, the GAG considers it to be at the upper end of




the range, with the lower end being the highest estimates based on animal data.




This range is:




                  1.0 x 1CT4 (ug/m3)'1 - 3.6 x 10~4 (ug/m3)'1






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 relative




potency indices for 54 chemicals that have been evaluated by the CAG as suspect




carcinogens.  The data summarized by the histogram are presented in Table 9-34 .




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 leukemias in two women exposed to ETO for up to 9 years




(Hogstedt 1979b), the relative potency index for ETO has been calculated as




5.6 x 10+*.  This number was derived by multiplying the slope in units of




(mg/kg/day)~l by the molecular weight of ETO, which is 44.1.  Based on the
                                     9-153

-------
f

-------
TABLE 9-34 .   RELATIVE CARCINOGENIC POTENCIES AMONG  54 CHEMICALS  EVALUATED BY
     THE CARCINOGEN ASSESSMENT GROUP AS SUSPECT HUMAN CARCINOGENS1»2»3
Slope
Compound (mg/kg/day)""1
Acrylonitrile
Aflatoxin Bj
Aldrin
Allyl chloride
Arsenic
B[a]P
Benzene
Benzidene
Beryllium
Cadmium
Carbon tetrachloride
Chlordane
Chlorinated ethanes
1 , 2-dichloroethane
hexachloroe thane
1, 1,2,2-tetrachloroethane
1,1, 1-trichloroethane
1,1, 2-tr ichloroethane
Chloroform
Chromium
DDT
Dichlorobenzidine
1 , 1-dichloroethylene
Dieldrin
0.24(W)
2924
11.4
1.19xlO~2
15 (H)
11.5
5.2xlO~2(W)
234(W)
1.40
6.65 (W)
1. 30x10-!
1.61
6.9xlO~2
1.42xlO-2
0.20
1.6xlO-3
5.73xlO-2
7xlO~2
41 (W)
8.42
1.69
1.47x10-1(1)
30.4
Molecular
weight
53.1
312.3
369.4
76.5
149.8
252.3
78
184.2
9
112.4
153.8
409.8
98.9
236.7
167.9
133.4
133.4
119.4
100
354.5
253.1
97
380.9
Potency
index
1*10+1
9xlO+5
4x1 0+3
9x1 O-1
2x1 0+3
3x1 0+3
4x10°
4x1 0+4
1x10+1
7x1 0+2
2xlO+1
7xlO+2
7x10°
3x1 0°
3x10+1
2x10-1
8x1 00
8x10°
4x1 0+3
3x1 0+3
4xlO+2
1x10+1
lx!0«
Order of
magnitude
(logio
index)
+1
+6
+4
0
+3
+3
+1
+5
+1
+3
+1
+3
+1
0
+1
-1
+1
-l-l
+4
+3
+3
+1
+4
                                              (continued on the following page)
                                      9-155

-------
TABLE 9-34 .  (continued)
Slope Molecular
Compound (mg/kg/day)"1 weight
Dinitrotoluene
Diphenylhydrazine
Epichlorohydrin
Bis(2-chloroethyl)ether
Bis(chloromethyl)ether
Ethylene dibromide (EDB)
Ethylene oxide
Heptachlor
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclohexane
technical grade
alpha isomer
beta isomer
gamma isomer
Hexachlorodibenzodioxin
Methylene chloride
Nickel
Nitrosamines
Dimethylnitrosamine
Diethylnitrosamine
Diethylnitrosamine
Dibutylnitrosamine
N-nitrosopyrrolidine
N-nitroso-N-ethylurea
N-nitroso-N-methylurea
N-nitroso-diphenylamine
PCBs
0.31
0.77
9.9xlO-3
1.14
9300(1)
8.51
1.26(1)
3.37
1.67
7.75xlO~2
4.75
11.12
1.84
1.33
1. 1x10+4
6.3x10-4
1.15(U)
25. 9 (not by q*)
43.5(not by q*)
43.5(not by q*)
5.43
2.13
32.9
302.6
4.92xlO-3
4.34
182
180
92.5
143
115
187.9
44.1
373.3
284.4
261
290.9
290.9
290.9
290.9
391
84.9
58.7
74.1
102.1
102.1
158.2
100.2
117.1
103.1
198
324
Potency
index
6xlO+1
1x1 0+2
9x10-!
2x1 0+2
1x1 0+6
2x10+3
6x1 0+1
1x10+3
5xlO+2
2xlO+1
1x10+3
3x10+3
5xlO+2
4x10+2
4x1 0+6
5x10-2
7xlO+1
2x10+3
4xlO+3
4x10+3
9xlO+2
2xlO+2
4x10+3
3x10+4
1x10°
1x10+3
Order of
magnitude
(logio
index)
+2
+2
0
+2
+6
+3
+2
+3
+3
+1
+3
+3
+3
+3
+7
-1
+2
+3
+4
+4
+3
+2
44
44
0
+3
         9-156
(continued on the following page)

-------
TABLE  _
                                      (continued)


Slope
Compound (mg/kg/day)""1
Phenols
2,4,6-trichlorophenol
Tetrachlorodibenzo-p-dioxin
Tetrachloroethylene
Toxaphene
Trichloroethylene
Vinyl chloride
Remarks:
1 . Animal slopes are 95%
1.99xlO-2
1.5 6x1 0+5
3.5xlO-2
1.13
1.9xlO-2
1.75x10-2(1)

upper-limit sl<
Order of
magnitude
Molecular Potency (logio
weight index index)
197.4 4x10° +1
322 5xlO+7 +8
165.8 6x10° +1
414 5x1 0+2 +3
131.4 2.5x10° 0
62.5 1x10° 0

apes 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 non-threshold model.

2.  The potency index is a rounded-off slope in (mMol/kg/day)"^ and is calcu-
    lated by multiplying the slopes in (mg/kg/day)~1 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.
                                    9-157

-------
Hogstedt (1979b) study, this slope is 3.6 x 10~^ (ug/m3)"1.  The transformation

from ug/m3 to mg/kg/day is performed as follows:


              1 ug x 20 m3 x 1 mg x   1   = 2.86 x ICT4 mg/kg/day
              m3      day    1000   70 kg


The unit risk slope can then be converted as given below:


               b = 3.6 x 10-4 (Ug/m3)-l x l Ug/m3
                                          2.86 x 10~4 (mg/kg/day)

                              - 1.26 (mg/kg/day)"1


     The potency index for ETO is thus 1.26 x 44.1 = 5.6 x 10+1, putting it in

the third quartile of the 54 chemicals which the GAG has evaluated as suspect

carcinogens.  If the lower part of the range is used, the potency index for ETO

would be 3.5 x 10"1 x 44.1 = 1.5 x 10+1, which would also rank it in the third

quartile.  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 different 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 calcula-

ted 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 — Positive results for the carcinogenicity of ETO have been

obtained by subcutaneous injection in mice and intragastric administration in

rats.  Two long-term chronic animal studies were performed that adequately tested

the carcinogenic potential of ETO by inhalation:  the Bushy Run study (Snellings

et al. 1981) and that of NIOSH (Lynch et al. 1982).  Snellings et al. (1981)

indicated that ETO exposure resulted in an increased incidence of mononuclear
                                      9-158

-------
cell leukemia in females in the two highest dose groups; this increase was dose-




related.  The test for a linear trend was highly significant (P < 0.0001).  There




was a significant (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 analysis




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




high-dose exposed groups.  For these latter two sites, the dose-response trend




tests were also statistically significant (P < .01).




     Three epidemiclogic studies of workers exposed to ETO demonstrated signifi-




cant (P < 0.05) association between ETO exposure and the occurrence of cancer.




Two of the studies (Hogstedt et al. 1979a and Hogstedt et al. 1979b) found an




association between ETO exposure and incidence of or death from leukemia.  ETO




was not found to be specific for any particular type of leukemia, however.




Other sites or types of cancer found to be significantly (P < 0.05) associated




with ETO exposure include pancreatic cancer and Hodgkin's disease in the Morgan




et al. (1981) study and stomach cancer in the Hogstedt et al. (1979a) study.




     A range of unit risk estimates for ETO has been calculated from both




animal and human data.   The lower end of the range was a 95% upper-limit estimate




based on total mononuclear cell leukemias and brain gliomas in female Fischer 344




rats in the Bushy Run study.  The higher end was based on human leukemias in the




Hogstedt (1979b) study.  The unit risk estimates from the animal data were




calculated from a linearized multistage model,  while the human data estimate was




calculated using a relative risk model.  The unit risk estimate based on human
                                      9-159

-------
data is 3.6 times as high as that based on animal data.  Extrapolation from the




human leukemia data results in a highly uncertain estimate due to the small




number of leukemia cases recorded.  Of interest is the fact that humans have




quantitatively greater sensitivity to ETO than do rats, as evidenced by the data




for chromosome abnormalities.




     Using the above-referenced extrapolation procedures, the range of estimates




of lifetime cancer risk resulting from continuous exposure to air that contains




an ETO concentration of 1 ug/m^ is calculated to be 1.0 x 10"^ - 3.6 x 10"^.




The plausibility of these estimates is enhanced when clear evidence of mutagenicity




exists, as is the case with ETO.




9.5.5  Conclusions




     ETO is a direct-acting alkylating agent.  It reacts with mammalian DNA




primarily at the N-7 position of guanine.  It induces base-pair substitutions




in the Ames test, and gene mutations in plants and animals.  It also breaks




chromosomes of plants, animals, and humans, and causes DNA. damage in the




spermatids of mice.  The weight of the available evidence indicates that ETO is




a direct-acting mutagen.




     Using the criteria of the International Agency for Research on Cancer (IARC)




for assessing the evidence of carcinogen!city from studies in humans (Appendix A),




the GAG considers that the human data for ETO constitutes limited, bordering on




inadequate, evidence that ETO is a human carcinogen.  The GAG finds the animal




evidence of the carcinogenic!ty of ETO to be sufficient.  On the basis of its




analysis of the human, animal, and mutagenic data cited herein, the GAG classifies




ETO as being probably carcinogenic to humans and therefore as belonging in IARC




Group 2A.   The GAG would qualify this classification as bordering on Group 2B,




however, because of limitations in the human evidence.  (See Appendix A for a




description of the IARC categories.)
                                     9-160

-------
     Estimates of the relative potencies of ETO in animals and humans, made on




the basis of leukemias and brain gliomas in animals and leukemias in human




studies, suggest that humans may be more susceptible than animals to the carcino-




genic effects of ETO.  The unit risk estimate for ETO in humans is 3.6 x 10"^




(ug/m3)-!, while the estimate for animals, based on studies in Fischer 344 rats,




is 1.0 x 10~4 (ug/m3)"1.




     The potency index of a chemical, as calculated by the CAG, is based on both




its unit risk and its molecular weight.  For ETO, which has a molecular weight




of 99, the potency index based on human inhalation is 5.6 x lO"1"*.  The potency




index based on animal data is 1.5 x 10+1.  These indices rank ETO either in the




third or the fourth quartile, respectively, of the 54 suspect carcinogens




evaluated by the CAG.
                                        9-161

-------
                                   APPENDIX  9-A

          INTERNATIONAL AGENCY FOR RESEARCH ON CANCER (IARC) CRITERIA
               FOR EVALUATION OF THE CARCINOGENICITY OF CHEMICALS*
ASSESSMENT OF EVIDENCE FOR CARCINOGENICITY FROM STUDIES IN HUMANS

     Evidence of carcinogenic!ty 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 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.
*Intemational Agency for Research on Cancer.  1982.  IARC Monographs:
Evaluation of the Carcinogenic Risk of Chemicals to Humans, Supplement 4.
Lyon, France.
                                        A-l

-------
     The  degrees  of evidence  for carcinogen!city from studies in humans were




 categorized as:




     i.   Sufficient evidence  of carcinogenic!ty, which indicates that there is




 a  causal  relationship between the agent and human cancer.




    ii.   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.




   iii.   Inadequate evidence, which indicates that one of three condidtions pre-




 vailed:   (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 carcinogenicity.






ASSESSMENT OF EVIDENCE FOR CARCINOGENICITY FROM STUDIES IN EXPERIMENTAL ANIMALS




     These assessments were classified into four groups:




     i.  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 information from short-term tests or on




chemical structure.




    ii.  Limited evidence of carcinogenicity, which means that the data suggest a




carcinogenic effect but are limited because:  (a) the studies involve a single




species, strain, or experiment;  or (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
                                       A-2

-------
 to  classify as malignant by  histological criteria alone (e.g., lung and liver




 tumors in mice).




    iii.  Inadequate evidence, which  indicates that because of major qualitative




 or  quantitative limitations, the studies cannot be interpreted as showing




 either the presence or absence of a  carcinogenic effect; 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.




    iv.  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, evaluation of




the carcinogenic risk to humans was  based on consideration of both the




epidemiological and experimental evidence.  The breadth of the categories of




evidence defined above allows substantial variation within each.  The decisions




reached by the 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 chemical,  groups of  chemicals,  industrial processes or occupational




exposures were thus put into  one  of three groups:
                                       A-3

-------
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 epidemiclogical 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 classification




of 2B.




     In some cases, the Working Group considered that the known chemical properties




of a compound and the results from short-term tests allowed its transfer from




Group 3 to 2B or from Group 2B to 2A.




Group 3




     The chemical, group of chemicals, Industrial process or occupational exposure




cannot be classified as to its carcinogenicity to humans.
                                       A-4

-------
                                   APPENDIX  9-B




              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 non-threshold 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  obser-




vable 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 estimate




such values as percentile lethal dose or percentile effective dose.  The log-
                                        B-l

-------
 Probit model 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 < 1






 where P is the proportion responding at dose D, c is an estimate of the




 background rate, a is an estimate of the standardized mean of individual




 tolerances, 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 = background or zero dose rate, and




b = linear component or slope of the dose-response model).  In considering the




added risk over background, incorporation of Abbott's correction 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




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
                                         B-2

-------
usually significantly lower than either the multistage or one-hit models, which




are linear at low doses.  All three of these models—the multistage, 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 B-l.  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 pro-




duced the highest estimates and the log-Probit model produced the lowest; in




this case, the multistage, one-hit, and Weibull all produced similar results.
                                       B-3

-------
                  TABLE  B-l.   ESTIMATES OF HUMAN UM-UOSF. RISK BASF.U ON DATA FROM MAI> *NU FbMALF.  FISDItR J44 RATS
                             IN THE BUSHY RUN ETO INITIATION STUDY, AS DKRIVhU FROM k'llllR IHFFF.RKN1  MODELS.
                             ALL KSTIMATES INCORPORATE ABBOTT'S CORRECTION FOR INUKPI-NDENT BACKGROUND RATE
f-ontlnuous
human
exposure
ppm
Males
.001
0.01
O.I
1
Fenales
f .001
0.01
O.I
1
Maximum likelihood estimates of
additional risks
Multistage One-hit Welbull
model model model

3.1x10-* 1.3x10-* 2.5x10-5
3.1x10-3 1.3x10-3 3.3x10-*
3.1x10-2 1.3x10-2 4.3x10-3
8.4x10-2 1.2x10-1 5.5x10-2

1.4x10-5 3.fixlO-J
1.4x10-3 1.4x10-2
1.4x10-2 5.0x10-2
1.3x10-' 1.7x10-1

Log-Prohlt
model

3.1x10-1
1.6x10-6
8.6x10-*
5.4x10-2

1.2x10-*
3.0x10-3
3.3x10-2
1.7x10-1
95Z upper confidence limit
of additional risks
Multistage One-hit Welbull
model model model

9.2x10-5 1.6x10-* 1.3x10-*
9.2x10-* 1.6x10-3 1.4x10-3
9.2x10-3 1.6x10-2 1.3x10-2
8.8x10-2 1.6x10-' 9.7x10-2

1.9x10-* 1.5x10-2
1.9x10-3 4.3x10-2
1.9x10-2 1.1x10-'
1.7x10-1 2.5x10-1

Log-Ptobtt
model

5.0x10-9
1.5x10-5
3.9x10-3
9.7x10-2

1. OxKr3
1.5x10-2
9.2x10*2
2.5x10-'
Animal exposure 0, 33 ppm,  100 ppm 6 hours/day,  5 days/week.

DATA
Males

No. tumors/No, examined
   Human eq. dose - mg/kg/day

  0      0.35    0.94     2.63

5/187    4/88    12/82    29/96
               Human eq. dose -  mg/kg/day

Females      0       0.28      0.75     2.11

           23/186    15/71    27/72     12/7J
Conversions {or 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  pave  Identical results In females.

-------
                                10.   REFERENCES

Alguire, D.E.  1973.  Ethylene oxide gas.  Food Technology,   pp.  64-67.


Altshuller,  A.P.  and  J.J.  Bufalini.    1965.    Photochemical aspects  of  air
pollution:  A review.  Photochem.  Photobiol.  4: 97-146.


Anderson, S.R.   1971.   Ethylene  oxide  toxicity:   A study of  tissue  reactions
to retained ethylene oxide.  Jour. 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.   198la.   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.   19810.   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.  Encroth  and C. Grant.    1977.   Studies  on ethylene oxide:
Whole body autoradiography  and  dominant  lethal test in mice.  Proc.  Eur.  Soc.
Toxicol.  18: 315-317.


Appelgren, L.E.,  G.  Encroth,  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 ri-pentane in a shock tube.
Combust. Sci. Technol.  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):  4677-1680.


BASF Wyandotte Corporation.   1972.   Ethylene oxide  handling manual.   Report
No. 21.  BASF Wyandotte Corporation,  Wyandotte, Michigan.


Bayes,  A.L.     1979.    Ethylene  oxide  one-generation   reproduction   study.
(Carnegie-Mellon  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.  Jour. Gas Chromatog.  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.   Jour.  Genet.  50:  480-485.
                                      10-2

-------
Biro, L.f  A.  Fishes  and  E. Price.   1971.   Ethylene  oxide burns, a  hospital
outbreak involving 19 women.  Arch. Dermatol.   110:  921.


Blackford, J.L.   197*4.  Acrylonitrile.   Chemical  Economics  Handbook.   Stanford
Research Institute, Menlo Park, California.


Blackford,  J.L.     1976a.    Propylene  oxide.    Chemical  Economics  Handbook.
Stanford Research Institute, Menlo Park, California.


Blackford,  J.L.     1976b.    Ethylene   oxide.    Chemical  Economics  Handbook.
Stanford Research Institute, Menlo Park, California.


Blackwood,  J.  and E.  Erskine.   1938.   Carboxide  poisoning.   U.S.  Navy Med.
Bull.  36: 4U-M5.


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.  Jour. Phys. Chem.   82: 2067-2073.


Bridie, A.L.,  C.J.M.  Wolff 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.  Jour. Amer.  Chem. Soc.  51: M28-161.


Brown, R.H. and C.J.  Purnell.   1979.   Collection  and Analysis of trace organic
vapour  pollutants  in  ambient atmospheres:  The  performance  of  a   Tenax  GC
adsorbent  tube.  J. Chromatog.  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.   In
vitro.  15: 220-221.


Brown, D.   1970.   Determination of ethylene oxide and ethylene chlorohydrin in
plastic and rubber surgical equipment  sterilized with ethylene  oxide.   Jour.
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.   Jour.  Food
Sci.  26: 428.
Calleman, L.J., L.  Ehrenberg,  B.  Jansson, S. Osteran-Golkar, D.  Segerbach,  K.
Svensson  and  C.A.  Wachtmeister.    1978.   Monitoring  and  risk assessment  by
means of  alkyl  groups  in hemoglobin in persons  occupationally  exposed to ETO.
J.  Environ. Pathol. Toxicol.  2:  427.

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.  Jour. 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.  420-128.


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.  Hendry, 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 Jour. Surg.  36: 53-55.
                                      10-4

-------
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).


Cogswell, S.A.  1980.   Chemical  Economics  Handbook.   Marketing Research Report
on Ethylene  Oxide.   SRI International,  Menlo Park,  CA.   January,  1980.   pp.
654.5032S, 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: 137-151.


Crump,  K.S.  and  W.W.  Watson.    1979.    GLOBAL 79.    A  Fortran  program  to
extrapolate  dichotomous animal  carcinogenicity  data to  low dose.   National
Institute of Environmental Health Science.  Contract No. 1-ES-2123.


Crutzen,  P.J.  and  J.  Fishman.    1977.   Average  concentrations  of  OH  in the
troposphere  and the  budgets  of CHj., CO, H?,  and  CH-CC1-.   Geophys. Res. Lett.
4: 321-324.


Gumming,  R.B., G.A.  Sega,  C.Y.  Horton  and W.H. Olson.  1981.  Degree  of alkyl-
ation  of DNA in various tissues of the mouse following inhalation  exposure to
ethylene  oxide.   Environm. Mutag.   3(3): 343.


Gumming,  R.B.,   T.A.   Michand,   L.R.   Lewis  and   W.H.   Olson.      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).
(Mutat.   Res.).


Cupitt,  L.T.    1980.    Fate  of  toxic  and  hazardous  materials  in  the air
environment.  U.S. EPA Report No.   600/3-80-084.  NTIS PB 80-221948,  35 pp.
                                      10-5

-------
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 111.  Reinhold,  New York.


Cvetanovic,   R.J.    1955.    Mercury photosensitized  decomposition of  ethylene
oxide.  Can. Jour. Chem.  33: 1681-1695.


Darnall, K.R.,  A.C.  Lloyd,  A.M. Winer and J.N.  Pitts,  Jr.  1976.   Reactivity
scale  for  atmospheric hydrocarbons based  on reaction  with hydroxyl  radical.
Environm. Sci. Technol.  10: 692-696.


DeBont, J.A.M.  and R.A.J.M.  Albers.   1976.  Microbial  metabolism of ethylene.
Antonie van Leeuwenhoek.  12(1): 78-80.
Billing, W.L.   1977.  Interphase  transfer  processes.   II.  Evaporation rates of
chloromethanes,   ethanes,  ethylenes,  propanes,  and   propylene   from  dilute
aqueous solutions.   Comparisons with theoretical predictions.  Environm.  Sci.
Technol.  11: 105-409.
Dimitriades, B. and  S.B.  Joshi.   1977.  Application  of  reactivity criteria in
oxidant-related  emission  control   in  the  U.S.A.    In;    EPA-600/3-77-001B.
Inter. Conf. on Photochemical Oxidant Pollution  and Its  Control.   Dimitriades,
B., ed.  Research Triangle Park,  North Carolina,   pp.  705-711.


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.
Jour. Chem. Soc.  A:  121-135.


Dobinson, B., W. Hoffman and  B.D.  Stack.   1969.   The Determination of Epoxide
Groups.  Pergamon Press, Oxford.


Dolovich, J. and  B.  Bell.   1978.   Allergy to a  product(s)  of  ethylene oxide
gas:   Demonstration  of IgE and IgG antibodies  and hapten specificity.   Jour.
Allerg. Clin. Immunol.  62: 30-32.


Dow Chemical Co.   1977.  Alkylene oxides from Dow.   Dow Chemical Co. Form No.
110-551-77R, Midland, Michigan.
                                      10-6

-------
Dunkelberg,  H.    1979.    On  the  oncogenic  activity  of  ethylene  oxide  and
propylene oxide in mice.  Brit. Jour. Cancer.   39(4): 588-589.


Dunkelberg,  H.    1981.    Kanzerogene  Aktivitat  von Ethylenoxid  und  seinen
Reaktions-produkten   2-Chloroethanal,    2-Bromoethanol,    Ethylenglykol    und
Diethylenglykol.   I.    Kanzerogenitat  von Ethylenoxid  im  Vergleich zu  1,2-
Propylenoxid  bei  Subkutaner Applikation  an Mausen.   Zbl.  Bakt.  Hyg. I.  Abt.
Orig.  B 174: 383-404.


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.   1967.   Haematologic studies on persons occu-
pationally 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.
                                      10-7

-------
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.


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 mutation theory.  Cancer Res.  30: 195-205.


FASEB  (Federation of American  Societies   for Experimental  Biology).    1974.
Biological  data  books,   Vol.  Ill,  second edition,  Altman,  P.L.  and  D.S.
Dittmen, Eds.  Library of Congress No.  72-87738.  Bethesda, MD.


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.


Freed, V.H.,  C.T.  Chiou and  R.  Haque.   1977.   Chemodynamics:    Transport and
behavior of chemicals in  the  environment - A  problem in  environmental  health.
Environm. Health Perspec.  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.
                                      10-8

-------
Garry, V.F.,  C.W.  Opp and J.K. Wlencke.   1981.   Ethylene-oxide induced sister
chromatid  exchange  in  human-lymphocytes  using  a  membrane  dosimetry  system.
Environ. Mutat.  3(3): 340.


Generoso,  W.M., K.T.  Cain,  M.  Krishna,  C.W.  Sheu  and  R.M.  Gryder.   1980.
Heritable  translocation  and  dominant-lethal  mutation induction  with ethylene
oxide in mice.  Mutat. Res.   73(1): 133-142.


Gilmour,  R.H.    1978.   Ethylene  oxide treatment of  cosmetics.    Soap  Perfura.
Cosmet.  51(11): 498-199.
Corner,  R.  and  W.A.  Noyes.    1950.    Photochemical  studies.    XLII.  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  Compounds in the Atmosphere.   Academic Press,
New York.  pp. 267, 272,  276.


Greaves-Walker, W.  and  C.  Greeson.    1932.   The  toxicity of  ethylene oxide.
Jour. Hyg.  32: 409-416.


Critter,  R.J.  and  E.C.  Sabatino.    1964.    Free-radical chemistry  of cyclic
ethers.   VII.  Ultraviolet  photolysis of epoxides and  propylene  sulfide in the
liquid  phase.  Jour. 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.


Hansch,  C.  and  A.C.  Leo.    1979.     Substituent  Constants  for  Correlation
Analysis  in  Chemistry and  Biology.     John  Wiley  and Sons,  Inc.,  New  York.
p. 176.
                                      10-9

-------
Hatch, G.G., P.D. Mamay, 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.


Hawley, G.G.  (ed.).   1981.   The  Condensed Chemical Dictionary,  10th  ed.,  Van
Nostrand 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.  Brit. Med.  I.   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.  Jour.  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. 211+1-2258.


Hirose, T.,  R.  Goldstein  and C.  Bailey.    1953-   Hemolysis  of blood  due  to
exposure to  different  types of  plastic tubing and  the  influence  of ethylene
oxide sterilization.  Jour.  Thorac. Cardiov.  Surg.  45:  245-251.


Hirose, C.    1974.   The microwave spectra and r ,  r ,   and  r  structures  of
ethylene oxide.   Bull.  Chem. Soc. Japan.  47: 131T-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. Malmgrist  and B.  Wadman.   1979b.  Leukemia in workers exposed
to ethylene oxide.  Jour. Amer.  Med. Assoc.  241(11): 1132-1133.
                                     10-10

-------
 Hogstedt,  C.    1981.   Department of Occupational  Medicine,  Regional Hospital,
 Orebro,  Sweden.  Letter to Peter Infante,  Office  of Carcinogen Identification
 and  Classification,  Occupational Safety and Health Administration, Washington,
 D.C.  Subject:  Response to questions on study of  ethylene oxide workers.


 Hogstedt,  C.    1983.    Department  of Occupational Medicine,  National Board of
 Occupational  Safety  and  Health  and Karolinska  Hospital,   Stockholm,  Sweden.
 Letter   to  Herman  Gibb,   Carcinogen  Assessment  Group,  U.S.  Environmental
 Protection Agency.    Subject:    Response  to   questions regarding  study  of
 ethylene oxide-exposed  workers.


 Hollingsworth,  R., V.  Rowe,  F. Oyen,  D.  McCollister  and  H.  Spencer.   1956.
 Toxicity of  ethylene  oxide determined  on experimental  animals.    Amer.  Med.
 Assoc. Arch.  Ind.  Health.   13: 217-227.
Holmberg,  P.C.   1979.   Central-nervous-system  defects  in  children  born  to
mothers exposed to organic solvents.  Lancet.  2:8135:177-179.


Holmberg,  P.C. and  M. Nurminen.    1980.   Congenital  defects of  the nervous
system  and occupational  factors during  pregnancy.    A  case-reference study.
Am.  J. Ind. Med.  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.


Hughes,  K.J.,  R.W. Hurn  and  F.G.  Edwards.   1959.   Separation and identifica-
tion   of  oxygenated  hydrocarbons   in   combustion  products   from  automotive
engines.    Gas  Chromatography,  International  Symposium,  2nd, East  Lansing,
Michigan, pp.  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. Jour.  Radiat. Biol.  Relat.  Stud.  Phys.
Chem. Med.  27: 355-362.


IARC  (International  Agency  for  Research on  Cancer)   1976.    Ethylene  oxide.
In;   IARC  Monographs on  the  Evaluation of Carcinogenic  Risk of  Chemicals  to
Man.  Lyon,  France.
                                     10-11

-------
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.


Jacobson, K.,  E.  Hackley and  L.  Feinsilver.    1956.   The toxicity  of inhaled
ethylene oxide and propylene oxide vapors.   Arch. Ind.  Health.  13:  237-211.


Jana, M.K.  and  K.  Roy.   1975.   Effectiveness  and  efficiency  of ethyl methane-
sulfonate and ethylene  oxide for the induction  of  mutations  in rice.   Mutat.
Res.  28(2): 211-215.  (Cited in NIOSH,  1977).


Jay, W.M.,  T.R. Swift and D.S. Hull.  1982.  Possible  relationship  of ethylene
oxide exposure to cataract formation.  Am.  J.  Ophthalmol.  93(6): 727-732.


Jefferson  Chemical  Co.    Undated   a.    Ethylene  oxide  technical  brochure.
Houston, Texas.


Jefferson  Chemical  Co.    Undated  b.   Propylene   oxide  technical  brochure.
Houston, Texas.


Jensen,   E.    1977.   Union  Carbide  Corp.  (unpublished).    (Cited   in  Federal
Register, 43: 3812, 1978).


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.
                                     10-12

-------
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.


Joyner,  R.    1964.    Chronic  toxicity  of  ethylene  oxide.   Arch.   Environm.
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).


Kimmel, C.A., J.B. LaBorde, C.  Jones-Pride,  T.A.  Ledoux  and T.A.  Marks.   1982.
Fetal  development  in New Zealand white   (NZW)  rabbits  treated  i.v.  with
ethylene oxide (EtO)  during pregnancy.   Society of  Toxicology Meeting, Boston,
MA.
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-
chromatid 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.t  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.
                                     10-13

-------
Kurginski,  E.R.     1979.     Personnel  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  teratogenioity of ethylene  oxide
administered intravenously to mice.   Toxicol.  Appl.  Pharmaeol.   56(1):  16-22.


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.


Lamaty, G., R. Maleq, C. Selve, A. Sivade and J. Wylde.   1975.   Mechanisms of
ring  opening  of oxirans by  acids in  aqueous  and nonaqueous solvents.   Jour.
Chem. Soc. Perkin II: 1119-1121.


Lambert, B, and A.  Lindblad.  1980.    Sister  chromatid exchange and chromosome
aberrations  in  lymphocytes  of laboratory  personnel.    J.  Toxicol.  Environ.
Health.  6(5-6): 1237-12U3.


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  Epidemiologic  Study  Center,   Preventive Medicine Division,
Department of Family Practice, Medical University of South Carolina, p. 12.


Lichtenstein,  H.J.  and  G.H. Twigg.   19^8.   Trans.  Faraday  Soc.   14:  905.
(Cited in Mabey and Mill, 1978).


Liohtenwalter,  G.D.  and G.H. Riesser.    1964.    Chlorohydrins.    Kirk-Othmer
Encyclopedia of Chemical Technology,  2nd  ed.   5:  304-32M.
                                     10-1U

-------
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,  B.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.


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.   Jour.  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.  an£  J.G.  Pritchard.    1956.    Hydrolysis  of  substituted  ethylene
oxides in H^O   solutions.  Jour. Amer. 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.  Jour. Amer. Chem. Soc.  79: 2362-2364.


Loveless, A.  and  P.  Wheatley.   1966.   In;   Genetic and  Allied Effects  of
Alkylating  Agents.    Pennsylvania  State  University  Press,  University  Park,
Pennsylvania,   vol. 131.


Lynch, D.,  T.  Lewis, W. Moorman, et al.    1982.   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.)
                                     10-15

-------
Lynch, D.W.,  T.R.  Lewis, W.J.  Moorman,  P.S. Sabharwal  and  J.R.  Burg.   1982.
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,
H676 Columbia Parkway, Cincinnati,  OH.


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.


Mabey, W. and T. Mill.   1978.  Critical review of hydrolysis  of  organic  com-
pounds in water  under environmental  conditions.   Jour. Phys. Chera.  Ref.  Data.
7(2): 383-415.
Mackay, D.  and P.J.  Leinonen.    1975.   Evaporation of  low-solubility  contam-
inants  from  water   bodies  to   atmosphere.     Environm.   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.     1977.    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 Journal 20 November 1982.


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.
                                     10-16

-------
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.    Amer.  Jour.
Ophthal.  29:  1355-1362.


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-U37.
Mishmash, H.E. and  C.E.  Meloan.   1972.   Indirect spectrophotometric determina-
tion of nanomole quantities of oxiranes.  Anal. Chem.  UU: 835-836.


Morgan, R.W.,  K.W.  Claxton, B.J.  Divine,  S.D. Kaplan and V.B.  Harris.   1981.
Mortality among ethylene oxide workers.  Jour. 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.   Vapor-phase organic  pollutants.
National Academy of Sciences,  Washington, D.C.


NFPA (National Fire  Prevention Association).   1975.   Hazardous chemicals data.
Fire Protection Guide on Hazardous Materials.  Boston, Massachusetts.
                                     10-17

-------
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).   1981.   NIOSH
Current  Intelligence Bulletin  No.  35.    Ethylene  Oxide   (EtO):   Evidence  of
Carcinogenicity, May 22, 1981.


NTP  (National  Toxicology Program).  1981.   National Toxicology  Program,  1980.
Department of Health and Human Services.  U.S.  Public Health Service.


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. Mitelman.   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.
                                     10-18

-------
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.


Pfeilstieker, 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.  Amer. Jour. 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.


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.      Jour.   Amer.    Chem.   Soc.
78: 6008-6013.


Pritchard, J.G.  and I.A. Siddiqui.   1973-    Activation parameters and mechan-
isms  of  the  acid-catalyzed hydrolysis  of epoxides.   Jour.  Chem.  Soc. Perkin
II: 452-457.


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.
Jour. Natl.  Cancer Inst.  32: 1045-1057.


Roberts,  J.,  L. Allison,  P.  Prichett   and  K.  Riddle.   1943.    Preliminary
studies  on soil  sterilization with ethylene oxide.   Jour.  Bacter.   45:  40.


                                     10-19

-------
Romano, S.J.  and J.A.  Renner.    1975.   Comparison  of analytical methods  for
residual ethylene oxide analysis.   Jour.  Pharaacol.  Sci.   64:  1412-1417.

Romano, S.J., J.A. Renner and P.M. Leitner.   1973.   Gas  ohromatographic  deter-
mination   of  residual   oxide   by   head   space   analysis.      Anal.   Chem.
45: 2327-2330.


Rowell, R.M.  and D.I. Gutzmer.   1975.   Chemical modification of wood:   Reac-
tions of alkylene oxides with southern yellow pine.   Wood Sci.   7:  240-246.

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.  Brit. Jour. 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.  Cumming,  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.   Environm. Mutag.  3(3): 371.
                                     10-20

-------
Seizinger, D.E. and  B.  Dimitriades.   1972.   Oxygenates  in  exhaust from simple
hydrocarbon fuels.  Jour. Air Pollut. Contr. Assoc.  22(1):  47-51.


Sexton, R.  and E. Benson.   1949.   Dermatological injuries  by  ethylene oxide.
Jour. 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.


Shenderova, R.I.,  O.K.  Alsopova, I.B. Shaklinazarov  and  V.A.  Linetskii.  1972.
Purification  of  waste waters from simultaneous  production  of  propylene oxide
and styrene oxide.  Khim. Prom.  (Moscow).  48: 746-747.


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.  Amer. Jour. Ophthal.  47: 86.


Smith,  H.H.  and T.A.  Lotfy.   1954.   Comparative  effects  of certain  chemicals
on Tradescantia  chromosomes as observed at  pollen tube mitosis.   Amer. Jour.
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.   Jour. 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.
                                      10-21

-------
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.   1982.   Effects  on  reproduction
in  Fischer  344  rats  exposed   to   ethylene   oxide   by   inhalation  for  one
generation.  Toxicol. Appl. Pharmacol.  63:  382-388.


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.


Stafford,  R.S.    1983.    Letter  to Herman  Gibb of  the Carcinogen  Assessment
Group, U.S. Environmental Protection  Agency.  Subject:  Information on methyl
formate.    Environmental  Mutagen Information Center,   Oak   Ridge  National
Laboratory, Oak Ridge,  TN.


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. 141-146.


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(1): 403-128.


Strekalova,  E.    1971.     Mutagenic action  of  ethylene  oxide  on  mammals.
Toksikol. Nov. Prom. Khim. Veshchestv.  12: 72.


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.  Lipincott  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  (unpublished).
Communication to division directors,  NIOSH,  January 8,  1979.


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.  Jour. 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.


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.   EPA-
600/4-76-,  U.S.  Environmental Protection  Agency, Washington,  D.C.


U.S.   EPA.      1980.     Investigation  of   selected   potential  environmental
contaminants: Epoxides.  U.S. Environmental  Protection Agency,  Washington,  DC.
EPA-560/11-80-005.


USITC  (U.S.   International  Trade  Commission).     1976.    Synthetic  Organic
Chemicals:   United States Production  and Sales,  1974.   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.
                                     10-24

-------
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).   1974.   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.  Jour. Natl.  Cancer Inst.  35: 707-717.


Veith,  G.D.,  K.J.  Macek,   S.R.  Petrocelli  and  J.  Carroll.   1979.   Federal
Register 15926 (March 15).


Verrett, M.J.    197**.   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.
                                     10-25

-------
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: 155-457.


Weast, R.C.   1980.   Handbook  of Chemistry and Physics,  61st ed.  The Chemical
Rubber Co., Cleveland, Ohio.  p. C-318.


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.  Jour.  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.


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, D.C.  pp. 140-161.
                                     10-26

-------
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.
Mutag.  4: 121-13**.


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.


Yakubova, Z.N.,  N.A.  Shamova,  F.A.  Miftakhova 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.
                                     TO-27
                                                         • US GOVERNMENT PRINTING OFFICE 1(64 - 759-OI.J/8620

-------