?/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(Do Not
Cite or Quote)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
------- |