EPA/600/R-95/099
                                                     April 1998
        Chloroethane Carcinogenicity
               (CAS No. 75-00-3)
National Center for Environmental Assessment—Washington Office
            Office of Research and Development
           U.S. Environmental Protection Agency
                  Washington, DC

-------
                                  DISCLAIMER

      This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
                                        11

-------
                            CONTENTS


LISTOFTABLES	iv

LIST OF FIGURES	.... iv

ABSTRACT	 v

PREFACE 		 vii

AUTHORS AND REVIEWERS	viii

1.  INTRODUCTION	1

2.  CARCINOGENICnT OF CHLOROETHANE	4

3.  CANCER BIOASSAY RESULTS	5
   3.1.  F344/NRATS	,	5
   3.2.  B6C3F, MICE	 9
   3.3.  MUTAGENICITY OF CHLQROETHANE	12

4.  DISCUSSION ..,	.13

5.  SUMMARY	19

6.  REFERENCES	21
                                ill

-------
                                LIST OF TABLES








1.  Vapor pressure of comparable compounds (mm Hg)	2





2.  Survival of F344/N rats at 2 years	6





3.  Rat leukemia incidence (2-year bioassay)	<	6





4.  Tumors of F344/N rats at 2 years	7





5.  Survival of B6C3F, miceat2years  ....	9





6.  Tumors of B6C3F, mice at 2 years	,	10





7.  Summary of tumors in F344/N rats and B6C3F, mice at 2 years	 15
                                LIST OF FIGURES








1.  Chloroethane chemical structure	,..	;	1





2.  SN2 mechanism of nucleophilic substitution of the chlorine in the Chloroethane molecule ... 3





3.  Potential mechanism of carcinogenesis of Chloroethane 	18
                                        IV

-------
                                     ABSTRACT

    Chloroethane (CE; CAS No. 75-00-3) can alkylatc cellular constituents but binding studies
do not exist. CE is mutagcnic in four tester strains of Salmonella typhimurium, plus and minus
S9 activating fraction.  A National Toxicology Program study tested CE carcinogenicity at one
dose 15,000 ppm in F344/N rats and B6C3F, mice: 6 hrs/day, 5 days/wk for 102 wks. Male rats
responded with basal cell carcinomas, keratoacanthomas, squamous cell carcinomas, and
trichoepitheliomas. The rat skin cancer incidence is 8/46 (17%) vs controls at 5/49 (10%)
(p=0.23) vs historical controls at 2/300 (0.7%) (p=2xlO"*). Female rats showed brain
astrocytomas at an incidence of 3/50 (6%) vs 0/50 (0%) controls vs 1/297 (0.3%) in historical
controls. Only comparisons with the historical control are significant. Female mice responded
with malignant and metastasizing endo- and myo-metrial cancers. They spread to the lung,
ovary, lymph nodes, kidney, adrenal gland, pancreas, mesentery, urinary bladder, spleen, and
heart. Supporting CE carcinogenicity is that bromoethane (BE), an analogue, produces a similar
spectrum of tumors—Jung, pheochromocytomas, and brain tumors in F344/N rats and uterine
tumors in B6C3F, female  mice. The structure-activity relationship lends weight to CE
carcinogenicity. The SAR, positive mutagenicity, and an exceptional degree and severity of
carcinogenicity all indicate that CE causes cancer in rodents and is probably carcinogenic to
similarly exposed humans.  CE is classified by the human inhalation route as Category B2
according to the U.S. Environmental Protection Agency's 1986 Guidelines for Carcinogen Risk
Assessment, and according to Proposed Guidelines for Carcinogen Risk Assessment, it is
classified as a likely human carcinogen.

-------
                                     PREFACE

    Chlorocthanc (CE) is a potentially hazardous air pollutant (HAP) that has been listed in the
1990 Clean Air Act Amendment, Section 112b. This report is an assessment of the
carcinogcnicity of inhaled chlorocthane in rodents, F344/N rats and B6C3F, mice. This report.
has been prepared by the National Center for Environmental Assessment-Washington Office
(NCEA-W).  It represents a weight-of-evidence approach, and it represents the summary
NCEA-W scientific position on chlorocthane carcinogenicity.
    This document was developed originally as a draft to assist the Office of Health and
Environmental Assessment Air Program Committee to construct a test rule in association with
the Office of Prevention, Pesticides, and Toxic Substances. On December 7,1994, chloroethane
draft was presented orally and lo the Carcinogen Risk Assessment Verification Endeavor
(CRAVE). A final version of the CE support document was sent to CRAVE along with an
Information System (IRIS) database IRIS summary on April 24,1995. The CE support document
was reviewed again administratively in February 1997 by NCEA-W, updated, and approved in
April 1998.
                                         VI

-------
                          AUTHORS AND REVIEWERS
PRIMARY AUTHOR (Hazard Evaluation)

James W. Holder, Ph.D.
Toxicologist/Cancer
• Effects Identification and Characterization Group
National Center for Environmental Assessment-Washington Office
 CONTRIBUTING AUTHOR (Cancer Potency Estimation)

 Jennifer Jinot
 Environmental Health Scientist
 Quantitative Risk Methods Group
 National Center for Environmental Assessment-Washington Office
 REVIEWERS (National Center for Environmental Assessment, Office of Research and
 Development)

 Charli Hiremath
 Debdas Mukerjee
 Sheila Rosenthal
 Cheryl Scott
 OUTSIDE REVIEWER

 John R. Bucher, Toxicology Branch Chief, Environmental Toxicology Program, National
 Institute of Environmental Health Sciences, Public Health Service, Research Triangle Park, NC
ACKNOWLEDGMENT

    The authors wish to express their appreciation to Dr. Aparna Koppikar of the Exposure
Analysis and Risk Characterization Group/NCEA for her valuable input to the discussions of
human astrocytoma incidences.
                                       vii

-------
1.  INTRODUCTION
                                   H     H
                                    I       I
                            H—C—C-C1
                                    I       I
                                   H     H
       This report is a characterization of chloroethane (CE, CAS No. 75-00-3) carcinogenicity.
CE is a potentially hazardous air pollutant (HAP) listed in the 1990 Clean Air Act Amendment,
Section 112b. Alternate chemical names or trade name products are ethyl chloride,
ir onochloroethanc, Kelene, Narcotile, muriatic ether, ether chloridum, and chloryl anesthetic.
The chemical formula of CE is CH3CH2C1, the molecular weight is 64.51, and the structural
formula is presented in figure 1.
       CE is produced commercially by the free radical
chlorination of ethane or by bubbling ethylene (CH2=CH2)g
through (HC1),. CE production in the United States in 1985.
was >460 million pounds (NTP, 1989a). CE can be reacted
with lead and a free radical initiator to make tetraethyl lead, an
antiknock compound for gasoline.  The manufacturing of
tetraethyl lead has been the largest source of human exposures
to CE, but these exposures have declined precipitously due to
reduced use of leaded gasoline. At present, chloroethane is
commonly used as an industrial solvent, a chemical
intermediate, and a blowing agent such as in styrene plastic
manufacture. At a production level of nearly one-half billion
pounds in the United States, there is potential for involuntary
inhalation exposure.
       The compound is flammable, especially in the gaseous state, and the flash point is -50eC
(closed cup), and the explosive limits are 3.8% to 14.8% (v/v).  Other physical properties of CE
are:  melting point = -136.4°C, boiling point = 12.3T, density = 0.9214, vapor density = 2.22
(air = 1.00), and vapor pressure =1,199 mm Hg (at 25 °C). CE's volatility at 20 °C is presented
in table 1 (number 12), compared with the volatility of other reference gaseous compounds. The
d'.ta in table 1 indicate that CE is quite volatile, and this further suggests that CE can be an
inhalation toxicant wherever it is stored, handled, or disposed.
       CE is chemically stable under neutral, metal-free conditions. However, CE can
chemically react with water, which is bipolar, at the C-l carbon atom (attached to the chlorine
atom) even under slightly basic conditions and prolonged reaction times. A base such as OH",
                          Figure 1. Chloroethane
                          chemical structure.
                          Chloroethane is a small,
                          hydrophobic molecule in
                          which C-1  is susceptible
                          to nucleophilic attack due
                          to the polarity of the C-CI
                          bond (see  also figure 2).

-------
       Table 1.  Vapor pressure of comparable compounds (mm Hg)
Number
1
2
3
4
5
6
7
8
9
10
11
12
Compound
Naphthalene
Ethylene dibromide
Water
Dichic.jethane
•Benzene
Carbon tetrachloride
Hexane
Chloropropane
Ethyl ether
Bromoe thane
Acetaldehyde
Chloroethane
Vapor pressure
0.53
10.1
17.3
60.6
74.6
76.4
120.0
278.1
290.8
/
475.0
764.3
1,002.3
which is symbolized as B" in figure 2, acts as an activator or catalyst in an SN2 reaction, thereby
generating an activated ethyl group and releasing a Cl" ion. For example, ethyl alcohol
CH3CH2-OH forms from CE hydrolysis where water is the nucleophile and a proton is released.
It may be anticipated that, at sufficiently high input of CE in cell water, an ethanolic acid solution
could be formed locally in situ that might be toxic.
       CE is a colorless gas with a pungent odor that is similar to that of ethyl ether, but at high
concentrations, CE gas has a burning nasal sensation and taste. The Occupational Safety and
Health Administration has recommended a threshold limit of 1,000 ppm  CE (2,600 mg/m3). CE
is a skin and eye irritant. The International Technical Information Institute (ITU) set the TC^
level equal to 1,300 ppm CE (TTII, 1979).  Excessive inhaled CE doses lead to central nervous
system suppression, headache, nausea, and lack of coordination (ataxia).  Prolonged or high
exposures produce feelings of inebriation, cardiac arrhythmias, unconsciousness, and cardiac
arrest. The mechanism of cardiac interference is likely to be by vagal nerve stimulation, which
can be reversed by atropine administration.

-------
                                                   can further react
v^n3v*
rijv-'i
chloroethane
T D — |Vrfn3v^ri2'^ I
attacking reaction intermediate
nucleophile
- v^n3un2o
activated
ethyl group
T ^1
leaving
group
Figure 2. SN2 mechanism of nucleophilic substitution of the chlorine in the chloroethane
       CE has been used as a local dermal anesthetic in humans; however, this use in humans
has declined ever since the chemical alkylation (ethylation) property of CE has been recognized
(see figure 2). However, CE is still used as a local percutaneous anesthetic in veterinary
procedures.  The mechanism of topical anesthesia is that heat is rapidly transferred from the skin
to the liquid CE raising it to the boiling point with rapid evaporation (expansion) from the skin,
leaving the skin frozen.
       Metabolism data for CE were not located in the literature, and this is a data gap.  Based
on its low molecular weight and high volatility, it is expected that inhaled CE would have free
access to the compartments of the corpus.  CE is  expected to deposit (to some degree) in the fat
depots but would not be a lingering corporal contaminant. The mixed function oxygenase system
should oxidize CE by C-l oxidation to acetaldehyde, which in turn is further oxidized to acetate.
The acetate would then be catabolized from the two-carbon pool by the trichloroacetic acid cycle.
Glutathione  (GSH) transferases should conjugate GSH and CE to GS-ethyl for elimination. Most
metabolic products should be passed through the urine.

-------
                   2.  CARCINOGENICITY OF CHLQROETHANE

       A Health Effects Assessment (HEA) document was developed previously on CE (U.S.
EPA, 1987). This 1987 document did not cite any cancer bioassays because such assays were not
available at that time. Mutagenicity was reviewed, and CE was found to be mutagenic in four
test strains of Salmonella typhimurium with and without S9 metabolic activation. Thus, using
only mulagenicity data, CE was classified in the HEA as an International Agency for Research on
Cancer Group 3 carcinogen, which translates to the U.S. Environmental Protection Agency
(EPA) cancer classification of Category D, insufficient data to characterize the carcinogenicity of
CE.      .
       Carcinogenicity data have become available since the 1987 HEA report.  A National
Toxicology Program (NTP) study was started on March 17,1982, and was reported as a final
report in 1989 (NTP, 1989a).  The NTP study is the only cancer study appearing in the cancer
literature since 1989. The results of this NTP study (final report no. 346) provide the bases of the
current EPA assessment of CE carcinogenicity by inhalation. The NTP study was designed to
determine the cancer effects of inhalation exposure to CE. The lower explosive limit for CE is
38,000 ppm in air, and to be safe, the carcinogenicity testing dose was set below this  limit at
15,000 ppm. Both F344/N rats and B6C3F, mice were exposed to CE at 15,000 ppm (39,000
mg/m3) only, and no other dose group was employed. CE was introduced into the inhalation
chamber as a 99.5% pure gas that was stable during the test and did not degrade. The animals
were dosed with CE for 6 hours/day, 5  days/week for 102 weeks. Groups of 50 F344/N rats or
B6C3F, mice (obtained from the Frederick Cancer Research  Facility in Maryland) for each sex
were used as control or treated animals. The treated group was compared with air-dosed control
animals for carcinogenicity. Good animal husbandry and good laboratory practices were
apparently observed at Battelle Pacific Northwest Laboratories where the rodent inhalation
exposure study was performed.
       The tissues that were necropsied were adrenal glands, brain, bronchial lymph nodes,
clitoral or preputial gland, cecum, urinary bladder, esophagus, gallbladder, trachea, tissue masses
with regional lymph nodes and any gross lesions, heart, thymus, thyroid, ileum, jejunum, rectum,
kidneys, spleen, sternebrae, salivary glands, larynx, liver, lungs, bronchi, mammary gland,
mandibular lymph nodes, snout, pancreas, parathyroid glands, pituitary, prostate, testes, and
epididymis or ovaries. Tumor discovery was either from (1)  early adventitial death, (2) early
death by other pathologic means, or (3) planned autopsy.

-------
                         3. CANCER BIOASSAY RESULTS

3.1. F344/N RATS
      The male rat lifetime growth curves were assessed by measuring body weights. The male
rat group exposed to 15,000 ppm CE had approximately 3% to 10% decreased body weights
compared with the concurrent control group from week 40 to termination. The body weights of
female rats exposed to 15,000 ppm showed a decrease of 5% to 13% compared with female
controls from week 15 to study termination. Although these are hot large differences, the female
body weight decrement shows that the maximum tolerated dose (MTD) was approximated
(^ 10%).  Male and female survival rates were computed by the Kaplan-Meier method and
showed an apparent precipitous decrement in both male control and treated groups (NTP, 1989a,
p. 38). Survival at terminus was low for male rats: 32% for controls and 16% for treated group.
Female F344/N rats showed better terminal survivals: 62% for controls and 44% for the treated
group. However, the differences between  the survival rates of controls and the treated groups in
either male or female rats were not statistically significant (table 2), so a treatment-related effect
on survival was not observed.     ,
       A high number of mononuclear cell leukemias were found in a number of tissues in both
the control and 15,000 ppm CE groups (table 3). This leukemic condition may account for the
lowered survivals'in both treated and untreated groups at the end of the 2-year NTP study, but it
was reasoned not to have compromised the study (NTP, 1989a).
       The male and female F344/N rat tumor occurrences are listed in table 4. CE may be
associated with low incidences of total skin tumors in male rats and with brain tumors in female
rats. The total tumor response in male rat skin seems to show that skin and certain skin
appendages are displaying a cancer response. Because skin is exposed to CE under the fur in the
inhalation chamber during the 102 weeks,  there is some dermal exposure. When compared with
the concurrent control incidence, that is, 5/49 (10%) versus 8/46 (17%), the male rat malignant
whole skin response is not statistically increased (p=0.23). The first skin tumor, a subcutaneous
fibroma,  occurred at 79 weeks in the treated group. Moreover,'the rates are not significantly
increased in the treatment group when rates are adjusted for animals dying before the first skin
tumor. The comparison in this case is 5/42 (12%) versus 8/42 (19%),/*=0.27.
      When the skin tumors of the treated group are compared with those of the historical
inhalation controls from the same testing laboratory, there is a statistically significant increase in

-------
       Table 2.  Survival of F344/N rats at 2 years'
Sex/treatment
groups
Males
Females
Controls
Survival
16/50(32%)
31/50(62%)
s,«
98
—
15,000 ppm chloroethane
Survival
8/50(16%)
22/50 (44%)
s,«
92
97
Probability
of survival
effect (p)
0.161
0.083
  Survival is defined as the number of animals alive at study termination divided by the starting number of animals in that
  group. The percentage survival is presented in parentheses. Sl/2 is the time in weeks that it takes to decrease to 50%
  survival compared with the start of the study. When survival is >50% at study termination, no Sl/2 exists by definition, and
  a dash is indicated.

       Table 3. Rat leukemia incidence (2-year bioassay)
Sex
Males
Females
Controls
33/50 (66%)
' (87.6% adj.)
20/50 (40%)
(48.1% adj.)
15,000 ppm
chloroethane
36/50 (72%)
(96.9% adj.)
26/50(52%)
(52% adj.)
Probability of
survival effect (p)
0.33
0.16
epithelial cancers: 2/300 (OJ%) versus 8/46 (17.4%),/7=2xlO'6.  Similarly, when NTP controls
from noninhalation historical experiments are compared with the treated group (28/1,936 [1.4%]
vs. 8/46 [17.4%],/?=8xlO'5),.there is also a statistically significant increase in epithelial skin
tumors.
       Historical incidence rates can be characterized. For example, tumor incidences may be
subjectively ranked:  (1) incidence rates <0.5% are rare, (2) incidences occurring >0.5% but <2%
may be considered uncommon, and (3) incidences >2% are generally common to aging test
rodents.  These definitions are operational, not absolute, and they represent expert judgment.  In
this bioassay, the historical malignant skin tumor incidence is 0.7%, and NTP incidence is 1.4%
where both are designated as uncommon tumor incidences. On the other hand, the above
observed control skin incidence is 10% (5/49) (table 4).  Comparing either the observed  or
historical control incidences to the treated group  incidences lead to different conclusions: there

-------
       Table 4.  Tumors of F344/N rats at 2 years
Sex

Males




Females

Controls
keratoacanthoma = 4/49 (8%)
fibroma =1/49 (2%)
total = 5/49 (10%)
adjusted to first appearance of tumor
(79 weeks)
(42 males)
tumor incidence = 5/42 (12%)
skin
historical controls = 2/300
(inhalation) (0.7%)
skin
historical controls = 30/1,936
(noninhalation) (2%)
astrocy_tomas = none in controls
adjusted to animals on test at 0 weeks
(46 females)
tumor incidence = 0/50 (0%)
historical astrocytoma controls =
1/297 (inhalation studies) (0.3%)
historical astrocytoma controls =
23/1,969 (all studies)ll.l%)
15,000 ppm chloroethane
basal cell carcinomas « 3/46 (7%)
keratoacanthoma = 2/46 (4%)
squamous cell carcinoma = 1/46 (2%)
trichoepithelioma a 1/46 (2%)
lip, squamous cell carcinoma = 1/46
(2%)
total = 8/46 (17%) __
adjusted to first appearance of tumor (79
weeks in treated group)
(42 males)
tumor incidence = 8/42 (19%)
see above, 8/46
see above, 8/46
astrocytomas » 3/50 (6%)
adjusted to first appearance of tumor at
52.weeks (49 females) tumor incidence
= 3/49(6.1%)
see above, 3/50
see above, 3/50
•
Estimate of
p value*
0.23
0.27
2.0 xlO'6"
1.3 x 10'6"
0.12
0.12
0.01"
0.02"
  The p value is the likelihood (probability) that the assumption of a positive cancer effect is in error. Usually piO.05 is
  taken as a reasonably significant level of certainty to continue to assume there is a positive cancer effect
" Designates statistical significance in a Rsher's exact test comparison.- Data taken from NTP report no. 346 (NTP, 1989a).

is a statistically significant increase when historical skin controls are considered but not when the
study concurrent control is considered as the reference control.
       In the female rats, brain astrocytomas occurred at a low incidence of 3/50 (6%) (table 4).
In analyzing the significance of this low incidence brain tiimor, it is known that astrocytomas are
not common in most strains of rat or in humans. So low incidences could be a sign of
carcinogenicity. There is extra concern when they do occur because such a tumor type in the

-------
brain has fatal implications in rodents and humans. When compared statistically with the
concurrent control (0/50 [0%] vs. 3/50 [6%]), the response yields statistical insignificance
(/?=0.12), which suggests that there may be no effect. The same may be stated when the adjusted
rates arc examined by subtracting the number of animals dying before the first astrocytoma
appears (52 weeks): 0/46 Versus 3/49, p=0.12.
       When rare tumors occur, the tumor rates require special consideration. Uncommon or
rare tumor incidences may not indicate a statistical increase when compared with their respecJve
concurrent control incidences.  This is because the number of trials (i.e., the number at risk in the
control and treated groups) is small, «50/sex/group, and a larger number of animals (in this case,
at the 95% level of confidence, * 150/sex/group) is needed to statistically score a rare
tumorigenic event. Accordingly, when the observed incidence (3/50) is compared with historical
pooled control incidence (1/297) from the same testing laboratory (Battelle Pacific Northwest
Laboratories), the statistically significant increase in astrocytomas is/?=0.01 (table 4). Note that
the larger denominator affects the statistical inference in the case of rare tumors. Similarly, when
the observed 3/50 astrocytomas in female F344/N rats are compared with the incidence of all
experimentally discovered astrocytomas in NTP studies (23/1,969), the statistical significance is
p=0.02(table4).
       The 3/50 (6%) astrocytomas response in female F344/N rats is statistically significant
when compared with historical controls but not with the concurrent controls. The observed and
historical control  incidences present different conclusions, that is,  a statistically significant
increase in astrocytomas is seen when-historical controls are considered but not when the study
concurrent control is considered.
       Further analysis shows, however, that Battelle Pacific Northwest Laboratories had a
singular prior incidence of 3/50 (6%) astrocytomas in a female concurrent control group of
F344/N rats. 'This singular control brain tumor incidence happens to be commensurate with the
brain response in  the 15,000 ppm CE group (table 4). Thus, if a past concurrent control
incidence can reach as high as 3/50 (6%), the apparent statistical significance of the dosed group
response—also an incidence  of 3/50 (6%)—becomes less important. Moreover, in past NTP
studies, the average astrocytoma incidence is 0.9% (18/1,969) and the range is 0% to 6% in
female F344/N rats. Here, too, it is observed  that an incidence level as high as 6% of
astrocytoma cancers may be observed in concurrent  controls.
       It is determined, then, that this female rat astrocytoma effect may be real but is marginal
if it is real. Sensitivity analysis indicates that only one more rat with an astrocytoma would have

-------
shifted the concern for a real response.  Therefore, the female rat brain response is designated as
equivocal evidence for carcinogenicity.
       It is concluded that both the male F344/N rats (skin tumors) and female rats (brain
tumors) present equivocal sets of evidence for carcinogenicity. This means that the rat data are
not negative, but they cannot be used in regulatory decisions as a positive bioassay cancer site.
However, both marginal sites of skin and brain may suggest clues to the mechanism of action of
CE carcinogenicity.
3.2. B6C3F,MICE
       The growth was comparable between the no-dose control and the 15,000 ppm CE-treated
groups, as measured by mean body weights for male and female B6C3F, mice. Mice survivals
are shown in table 5.  Survivals in the 15,000 ppm group were significantly lower than survivals
in the control mice (NTP, 1989a) for both the males and females. Male mice died earlier than the
females:  male mice survivals reached the 50% survival rate at 73 weeks, whereas female mice
reached the 50% survival rate at 89 weeks.
       There were few cancer incidence observations that could be interpreted as carcinogenic
responses in the male mice. Generally, the observed male B6C3F, mice cancer occurrences are
considered random and usual for aging mice of this strain (table 6). There was the suggestion of
a male lung response, with 10/48 (20.8%) responding versus 5/50 (10%) in control male B6C3F,
mice (p=0.11). The lung tumors were benign and composed of mostly adenomas (8/10 [80%] in

       Table 5, Survival of B6C3F, mice at 2 years1
Sex
Males
Females
Controls
Survival
28/50 (56%)
32/50 (64%)
s,«
^_
—
15,000 ppm
chloroethane
Survival
11/50(22%)
2/50 (4%)
S1rt
73
89
Probability
of survival
effect
(P)
3.91 x 10*}
<10'8
* Survival is the number or animals alive at study termination divided by the starting number of animals in that group.
  The percentage survival is presented in parentheses. S1/2 is the time in weeks that it takes to decrease to 50% survival
  compared with the start of the study. When survival is >50% at study termination, no Sm exists by definition and a dash
  is indicated; the survival only is presented for this group in the previous.column.

-------
      Table 6. Tumors of B6C3F, mice at 2 years
Sex
Males
Females
Controls
early deaths,
urinary tract infections,
no tumors of interest
uterine
carcinoma = 1/49 (2%)
(not endometrial)
uterine
carcinoma = 1/46
(corrected for time to 1st tumor,
which was at 67 weeks)
historical controls = 4/1,371
(inhalation studies)(0.29%)
historical controls = 3/951
(corn oil)(0.32%)
uterine
lymphomas = 1/49 (2%)
15,000 ppm chloroethane
early deaths,
urinary tract infections,
no tumors of interest
uterine
carcinomas = 43/50 (86%)
uterine
carcinoma = 43/48
(corrected for time to first
tumor, which was at 67
weeks) ,
cf. above, 43/50
cf. above, 43/50
uterine
lymphomas = 7/50 (14%) '
(P)
no male
cancer
effects

-------
Therefore, the male B6C3F, mouse lung response is considered inadequate to determine
inhalation carcinogenicity.
       The female B6C3F, mouse survival also was reduced significantly (p<10'8, table 5).  The
study diagnosis in female mice is that they died of complications caused by uterine carcinomas
(NTP,  1989a). Therefore, reduced survival in female mice was not incidental but rather was due
to the onset of cancer. The female B6C3F, mice responded to inhaled CE with 43 primary  .
endometrial tumors out of a total of 50 female mice (table 6). This is a primary uterine cancer
incidence of 86%. The endometrium of the uterus is the mucous and glandular lining that
contains columnar epithelial cells and is surrounded by a smooth muscular layer called the
myomctrium. The endometrial tumors caused by CE were highly malignant because they (1)
spread from the endometrium to the surrounding  myometrium, and (2) upon tumor progression,
then metastasized to many distal organs.  Second-site tumors, which had their origin in the
uterine primary site, were observed in 34/43 female B6C3F, mice (79% of responders). That is,
68% of the original 50 treated female B6C3F, mice responded with frank, malignant, and
metastasizing cancers. This uterine carcinogenic response and the subsequent metastases show
clear evidence of carcinogenicity caused by CE in female B6C3F, mice.
       There was an additional primary liver carcinogenic (6%) response in female B6C3Fi
mice.  Control liver rates were 0/49 (0%) adenomas and 3/49 (6%) hepatocellular carcinomas,
whereas the 15,000 ppm-treated group had 1/48 (2%) adenomas and 7/48 (15%) carcinomas.
The combined liver response is 3/49 (6%) and 8/48 (17%), which is a significant difference from
control liver rates (p=0.025). There were increases in hematopoietic cancer involvement with CE
treatment, including increases of a number of white cell types in bone marrow, lymph nodes.
(uterine, iliac, mediastinal, mandibular), spleen, and thymus. These effects are difficult to
differentiate from the secondary metastatic effect or second primary site effects by CE.
Nevertheless, these responses lend support to the powerful carcinogenic effects of CE in female
mice.
       The organ types and the number of female B6C3F, mice affected by metastasized uterine
cells were:
             Lung (23)                               •      Pancreas (7)
       •      Ovary (22)                              •      Mesentery (7)
             Lymph nodes (18)                        •      Urinary bladder (7)
             Kidney (8)                              •      Spleen (5)
             Adrenal gland (8)                         •      Heart (4)
                                          11

-------
Organs that had disseminated uterine cells, but to a lesser extent, were:
       •      Colon                           •      Liver
       •      Stomach                         •      Small intestine
       •      Gallbladder                      •      Ureter

3.3.  MUTAGENIGITY OF CHLOROETHANE
       Genotoxicity is useful in assessing carcinogenicity of a suspected environmental
carcinogen.  CE has tested positive for mutagenicity in Salmonella in two studies, one by Zeiger
et al. (1992) and the other by NTP (1989a). In both cases, the experiments were carried out in
desiccators because of the volatility of CE. Zeiger et al. (1992) tested CE and 310 other
chemicals under code in the presence or absence of liver S9 from Aroclor-induced male Sprague-
Dawley rats and Syrian hamsters. CE was tested at 0.002 to 0.017 moles per desiccator in strains
TA 100 and TA 1535 in Ames assays. In strain TA 100, CE produced less than a twofold
maximal increase in mutation over background in the absence of exogenous metabolic activation
or in the presence of rat liver S9. These were questionable responses. There was nearly a
twofold maximal increase with hamster liver +S9, which is a weak response. However, CE was
clearly positive in strain TA 535; maximal increases were 4.5-fold in the absence of S9,6-fold
with hamster S9, and 7-fold with rat S9, all with dose-response relationships.
       Genetic toxicology studies described in the NTP bioassay report on CE (NTP, 1989a)
were carried out in  strains TA 98, TA 100, and TA 1535 in the absence or presence of S9 as
described above in  Zeiger et al. (1992). The doses tested were 10 and 20 ug per plate. Testing in
strain TA 100 was negative without S9, equivocal with hamster S9, and positive with rat S9.
Testing in strain TA 1535 was  clearly positive with and without hamster or rat S9; increases in
mutation over background ranged from 7-fold to 34-fold.
       In addition to the above two studies, Ricco et al. (1983) published an abstract CE was
tested in strains TA 98, TA 100, TA 1535, and TA 1537 with and without Aroclor 1254-induced
S9 derived from male and female Osborne-Mendel rats and B6C3F, mice. As in the studies
described above, this study also was performed in desiccators.  The bacteria were exposed to CE
vapor over at least three dose levels. The abstract states that CE was mutagenic both with and
without metabolic activation. It did not state the dose levels or the Salmonella strains that tested
positive.  No data were presented.
                                          12

-------
                                   4.  DISCUSSION

       CE is a volatile industrial solvent that has narcotic and toxic properties at high inhaled
doses (NTP, 1989a). A 2-year cancer bioassay in rodents has been reported and is the sole source
of the current surrogate cancer toxicology on CE (NTP, 1989a). This report includes studies in
F344/N rats and B6C3F, mice. The nonstandard NTP protocol (only one-dose group), but with
apparently good laboratory practices, was used by Battelle Pacific Northwest Laboratories.
       The results suggest that male F344/N rats may not have significantly responded with
whole skin tumors (epidermis, dermis, and appendages) (table 4). Concurrent controls are
considered the most relevant comparison unless something is known to be experimentally wrong
with the concurrent control group.  Because nothing was reported to be wrong with the
concurrent control, the adjusted concurrent control comparison is being used (p=0.23), which
suggests a lack of cancer effect in the Udn of male F344/N rats. Whereas historical control
comparisons suggest statistically increased skin carcinogenic responses in male rats, the
concurrent control comparison indicates no positive carcinogenicity in male rat skin.  This
historical and concurrent control comparison indicates that the male F344/N rat skin cancer
response is a marginal cancer effect, at most.
       The female F344/N rat astrocytoma response is also equivocal because a low-level
response of 6% was found (0/46 [0%] vs. 3/49 [6%],/>=0.12). However, astrocytomas are
uncommon cancers in rodents and are rare in humans (personal communication, A. Koppikar,
NCEA-W/ORD/U.S. EPA).  Astrocytomas are often malignant tumors, sometimes invasive, and
contain varying amounts of fibrillar stroma.  They are tumors of concern when they occur in test
animals. The response was only 3/50, which presented no significance in a Fisher's exact test
when compared with concurrent control female rats (0/50), but it did show statistical significance
when similarly compared with historical control F344/N rats (table 4). Experience at Battelle
Pacific Northwest Laboratories, where the bioassay was conducted, with a single control group
with 3/50 astrocytomas (the same as the responding group in the current bioassay) suggests that
the historical control comparison may be less important than the concurrent control comparison.
Moreover, an upper-range limit of astrocytoma occurrence of 6% in NTP historical control
astrocytoma incidences also suggests the current putative brain response is equivocal.
       The male F344/N rats showed equivocal skin effects, and the female F344/N rats  showed
equivocal evidence in the brain. Both the skin and brain cancer responses are suggestive, mainly
by comparisons with historical controls, but they are only marginally positive at the most or are
false negatives at the least.

                                          13

-------
       Survivals were poor in male B6C3F, mice, but tumor occurrences are what one would
expect in aging 2-year-old male mice.  There was the suggestion of a lung response in male mice
(mostly adenomas and not carcinomas). However, the lung rates adjusted for mortality were 5/28
in treated versus 9/30 in control groups. These rates suggested no lung response (p=0.22).
Because so many male mice died before study termination, there can be no assurance that more
lung tumors might not have resulted if all males had lived to the end of the bioassay. That is, the
statistical power was sufficiently reduced so as to compromise the male mouse results being used
to infer cancer response. Thus, in this study the male B6C3F, mice results arc considered not
declarative and are inadequate to determine carcinogenicity in humans.
       There was a marginal liver response (first tumor at 81 weeks) in female B6C3F, mice
inhaling 15,000 ppm (3/45 vs. 7/37,p=0.09), but this was not considered a statistically relevant
cancer response in this group. However, a strong uterine carcinogenic response was observed in
female B6C3F, mice (0/49 vs. 43/50, /?=<10~8). These uterine tumors were highly malignant,
metastatic, and aggressive. Dissemination occurred in many distal organ sites, that is, 16 sites.
The complications from these tumors were reasoned to  be the cause of poor survival in these
'female B6C3F, mice (NTP, 1989a). These earlier-than-normal cancer-related deaths lend even
more credence to the carcinogenic effects of CE  in female B6C3Fi mice. The data indicate clear
evidence of CE carcinogenicity in female B6C3F, mice, which is determined to be useful in
predicting human cancer.
       A summary of the rodent surrogate cancer results is presented in table 7. The information
presented indicates equivocal carcinogenicity in the F344/N rat (male and female) and strong
evidence for carcinogenicity in the female B6C3F, mouse.
       Structure-activity relationships (SARs) are useful in assessing CE carcinogenicity.
Bromoethane (CH3CH2Br) is a structural analogue to CE. Bromoethane is a volatile compound
but less so than CE (table 1). While bromoethane has not been categorized as to carcinogenicity,
it was tested for carcinogenicity at inhaled concentrations of 100 ppm, 200 ppm, and 400 ppm
(NTP, 1989b). Female B6C3F, mice responded  to inhaled bromoethane with uterine
adenocarcinomas, carcinomas, and squamous cell carcinomas. The uterine responses at 100
ppm, 200 ppm, and 400 ppm doses were 4/50 (8%), 5/47 (11%), and 27/48 (56%) respectively
(NTP, 1989b). When the control incidence, 0/50 (0%),  is compared with the 400 ppm dose
                                          14

-------
       Table 7. Summary of tumors in F344/N rats and B6C3F, mice at 2 years*
      Sex
F344/N rat
B6C3F, mouse
     Males
                Marginal evidence
                • skin tumors
                      Inadequate for carcinogenicity
                      determination

                                       (0)
                Equivocal evidence
                • brain tumors
    Females
                              (±1
                      Clear uterine cancer evidence and
                      metastasis to 16 secondary organ sites.
                      Weak liver primary response.
                      Hematopoietic response in a number of
                      tissues and lymph nodes.

                      	      (strong positive)	
' Conclusions based on results and data taken from NTP report no. 346 (NTP, 1989a).

incidence, 27/48 (56%), a statistically significant increase in uterine cancer is observed (p=
<10'8). Bromoethane causes uterine cancer in mice, just as CE causes uterine cancer in mice.
                                                                  \
       Bromoethane also causes low-level brain tumors in male rats: 0/45 in controls and 3/50
(6%) at 100 ppm, 0/50 at 200 ppm, and 0/50 at 400 ppm.  This is not a statistically significant
cancer trend.  In the low-dose bromoethane group (100 ppm), the response level was the same as
the 15,000 ppm CE inhalation level. Male F344/N rats responded to inhaled bromoethane with 5
granular cell brain tumors in 150 rats summed over the 3 dose groups (0/48 in controls and 3/50,
1/50, and 1/50 in the treated groups). Again,  this is not a positive trend statistically but rather is a
low-dose response. No tumors of this type have been seen at Battelle Pacific Northwest
Laboratories, and only 0.2% have been seen in all the NTP studies. Gliomas, including
astrocytes, occurred in 3/150 bromoethane-treated rats. This brain tumor response in
bromoethane-treated (inhalation) rats also demonstrates organ site concordance to the brain
response in chloroethane-exposed (inhalation) rats.
       Taken together, the structural analogues CE and bromoethane have in common the
following:  a uterine 'and low-incidence-level  brain response.  The conjunction of these
results—both uncommon tumor types— in different bioassays indicates a response pattern that is
unlikely to occur by chance alone. Both the uterine and brain carcinogenic responses
demonstrate (1) organ site concordance in rodents, (2) the involvement of similar cancer
mechanisms between the two haloethanes, (3) the replication of these respective bioassays, and
                                          15

-------
(4) a pattern of carcinogenicity applicable to assessing cancer hazard for these monohalogenated
ethanes in humans.
       Further structural comparisons show that 1,2-dichloroethane (Category B2) produced a
marginal uterine carcinogenic response in female mice gavaged for 78 weeks (NCI, 1978). The
doses and uterine adenocarcinoma responses in these female mice were 148 mg/kg (3/49) and
299 mg/kg (4/47). The uterine tumors were not statistically increased. If this experiment had
been prolonged to 2 years as in the CE bioassay, more uterine tumors may have resulted from
1,2-dichloroethane.  Therefore, the SAR comparison to 1,2-dichloroethane is only suggestive.  It
is notable that higher halogenated ethane analogues (1,1-dichloroethane, 1,1,2-trichloroethane,
1,1,2,2-tetrachloroethane, pentachloroethane, and hexachloroethane) do not appear to cause
uterine tumors.
       The direct effect on the respiratory tract of halogenated ethanes, including CE, is not
clear.  There does not appear to be a CE-induced respiratory neoplastic response in the NTP
inhalation bioassay. Other portions of the respiratory tract, such as the nasal  cavity, do not seem
to be responding to CE exposure with a chronic oncogenic response.  This  is not true for
bromoethane and dibromoethane (ethylene dibromide); both of these compounds cause
respiratory lesions and neoplasms (NTP, 1989b). It  is not known why CE does not affect the
respiratory system at 15,000 ppm.
       CE is a direct, base-pair substitution mutagen in Salmonella. This direct mutagenicity
makes CE like other alkylators a candidate carcinogen by a direct DNA-based mechanism.  In
vitro and in vivo studies in mammalian mutagenicity systems are needed to further characterize
the mutagenic potential of this chemical.
       The monosubstituted ethanes, such as CE, can form primary alkyl carbonium ions
(CH3CH2+). This requires a formal charge separation in moving Cl" away from CH3CH2+ ions.
Primary alkyl carbonium ions are relatively unstable compared with tertiary carbonium ions, for
example, tert-bulyl carbonium ion.  As such, the CE carbonium ion, if it forms at all, would have
a relatively short half-life in solution. A short ion half-life means fewer chances to react in the
active carbonium ion state. Thus, the primary carbonium ion concentration would have to be
higher to be toxic by ethylation.  This pathway mechanism is likely to be minor or nonexistent.
       Another potential metabolic mechanism is that CH3CH2C1 can react by an S^2 reaction
with a cellular intermediate that is basic or is a nucleophile, making it electron-rich. For
example, chemical intermediates are normally and purposely activated this  way in biosynthesis.
Similarly, the CH3CH2C1 could react with a basic intermediate B: to form the intermediate, and
then react with nucleophiles in the cell such as DNA and proteins to form altered cellular
                                           16

-------
macromolecules (figure 3). These chemical changes may then lead to oncogenic sequelae.
Relevant macromolecular conjugation information was not found in the literature, and it exists as
a CE data gap.
       Under oxidative conditions, CE can react with cellular water: CH3CH2C1  + H2O -
CH3CH2OH + HG1. The ethyl alcohol from CE metabolism can proceed to acetaldehyde and
then finally to acetic acid. Under other conditions, B' can be a nuclcotide (in DNA or RNA) or a
cellular protein, and these macromolecules are ethylated. Glutathionc (GSH) would be expected
to transfer the ethyl group from GS-ethyl. If these ethylatibn events are in excess of normal
excision and/or repair metabolism, then toxicity can be expected to ensue. Low-level ethylations
are likely managed by the cell so that toxic effects, including cancer, do not ensue.
       It-might be expected that, since CE (15,000 ppm) and bromoethane (400 ppm) both cause
a similar pattern of oncogenicity in brain, skin, and mainly uterine cancer, there may be a
hormonal mechanism of action causing tumor promotion and progression.  Working on this
"hormonal" thesis, Bucher et al. (1995) looked for changes in estrous cyclicity in B6C3F,
females and did not find significant differences in the mean estrous cycle length (=5.1 days) at
the cancerous doses of either haloethane. Minor changes in time observed among proestrus,
estrus, metestrus, and diestrus, when carefully compared with controls, were judged to be not
significant.  Neither circulating estradiol nor progesterone levels measured in these studies varied
significantly with doses commensurate with oncogenicity (Bucher et al., 1995). Bucher and his
colleagues interpreted this to mean that the uterine cancer responses were not based on
predisposed changes in uterine hormones or in the estrous cycle. Later occurring hormone
effects (i.e., >21 days) still may be involved and are not ruled out by these studies. Bucher and
his colleagues speculated, that since there is no reason to suspect specific uterine organ
sequestration of CE or its metabolites, there must be other mechanisms such as oncogene

        chloroethane activator       B                       activated ethyl group
                                  I                               .    .
  CH3CH2CI  +   B:     -   CH3 - C - H   +   H     -       CH3CH2-B   +      Cl:

                                 Cl
              targets
                              time |  chloroethane-derived residues bound 1
  CH3CH2B   + [DNA, proteins!  -»  I  to cellular critical molecules         / -» toxic effects
                              toxicity

Figure 3. Potential mechanism of carcinogenesis of chloroethane.
                                           17

-------
 activation or mutation spectrum that should be further explored (Bucher et al., 1995).  Moreover,
we are most curious whether this strong uterine cancer response in mice is site concordant with
humans, that is, do they share the same mechanism and site of action?
      CE's metabolic mechanisms remain to be directly demonstrated by experiments. The
postulated mechanisms presented here should be considered when the cancer response of this
xenobiotic is considered. These mechanisms likely become more important at high CE exposure
levels, whereas at low exposure levels CE may be accommodated by normal steady-state
xenobiotic metabolism. This is not known at this time.  Metabolic information is needed on the
dosimetry of CE's metabolic intermediates and their cellular effects, especially concerning
oncogene activation, suppressor inactivation, DNA alkylation, and mutation spectra.
                                          18

-------