United States
Environmental Protection
Agency
Office of Health and
Environmental Assessment
Washington, DC 20460
Research and Development
EPA/600/S8-84/004F Dec. 1987
v°/ERA Project Summary
Health Assessment Document for
Chloroform
The Office of Health and Environ-
mental Assessment of the Office of
Research and Development, Environ-
mental Protection Agency (EPA), has
prepared this health assessment to serve
as a "source document" for EPA use.
While the assessment was originally
initiated for use in evaluating chloroform
as a toxic air pollutant under provisions
of Section 112 of the Clean Air Act, the
scope of the assessment covers other
exposure pathways such that it is useful
to other media-specific programs in the
EPA.
In the development of the assessment
document, the scientific literature has
been inventoried, key studies have been
evaluated, and conclusions have been
prepared in order to qualitatively and
quantitatively identify the toxicity of
chloroform. Toxic effect exposure levels
and other measures of dose-response
are discussed, where appropriate, to
place the nature of the health responses
in perspective with chloroform levels
in the environment.
Information regarding sources, of
chloroform release to the environment,
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 environment. While
the available information is presented
as accurately as possible, it is acknow-
ledged 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.
This Project Summary was developed
by EPA's Environmental Criteria and
Assessment Office, Research Triangle
Park, NC, to highlight the key findings
of the health assessment document (see
Protect Report ordering Information at
back).
Introduction
Chloroform (CHCI3) is a colorless,
volatile, nonflammable, liquid used pri-
marily in the production of chlorodi-
fluoromethane (90%) and for export (5%).
Nonconsumptive uses (5%) include use
as a solvent, as a cleaning agent, and as
a fumigant ingredient. Although chloro-
form production and capacity have de-
clined recently, 1981 data place direct
production of chloroform in the United
States at 184 million kg, with indirect
production estimated at 13.2 million kg.
Also, based on 1981 data, the amount of
chloroform in the U.S. emitted to air is
estimated to be 7.2 million kg, with dis-
charges to water of 2.6 million kg, and
discharges on the land of 0.6 million kg.
Chloroform is ubiquitous in the environ-
ment, having been found in urban and
non-urban locations. It has a characteris-
tic odor and is detectable at about 200
ppm. There have been reports of a
northern hemisphere background average
of 14 ppt (10"'2v/v), with an average in
the southern hemisphere of <5 ppt, and
a global average of 8 ppt. However, a
more recent report suggests the ratio of
hemispheric concentrations (north vs.
south) may be less dramatic, more on the
order of 1.6. For the most part, urban
ambient air concentrations remain n1000
ppt, and rural or remote locations can be
<10 ppt. There are some notable excep-
tions, however, but the reasons for them
are not readily apparent. The highest
values reported were in Rutherford, New
Jersey (31,000 ppt), and Niagara Falls,
New York (21,611 ppt).
-------
Physical and Chemical
Properties, and Analysis
Hydroxyl radical oxidation is the primary
atmospheric reaction of chloroform. Based
on the rate constant for reaction with
chloroform, a half-life of 11.5 weeks is
expected. The principal products from
this reaction are HCI and C02. It has
been estimated that roughly 1% of the
tropospheric chloroform will diffuse into
the stratosphere, based on a lifetime of
0.2 to 0.3 years and a troposphere-to-
stratosphere turnover time of 30 years.
An EXAMS model of chloroform in water
confirms other data suggesting that the
major removal process for chloroform in
water is evaporation.
The best analytical method for detection
of chloroform appears to be gas chro-
matography with electron capture or
electrolytic conductivity detection. This
gives a detection limit of <5 ppt.
Pharmacokinetics
The pharmacokinetics and metabolism
of chloroform have been studied in both
humans and experimental animals.
Chloroform is rapidly and extensively
absorbed through the respiratory and
gastrointestinal tracts. Absorption through
the skin could be significant only in in-
stances of contact with liquid chloroform.
The available data suggest that, in a
human at rest, at least 2 hours are re-
quired to reach an apparent equilibrium
of the body with the inhaled chloroform
concentration. The magnitude of chloro-
form uptake into the body (dose or body
burden) is directly proportional to the
concentration of chloroform in the in-
spired air, the duration of exposure, and
the respiratory minute volume.
The absorption of chloroform from the
gastrointestinal tract appears to be
virtually complete, judging from recovery
of unchanged chloroform and metabolities
in the exhaled air of humans and in the
exhaled air, urine, feces, and carcass of
experimental animals. Chloroform given
in a corn oil vehicle to experimental
animals is absorbed more slowly than
chloroform given in water. Peak blood
levels occurred at —1 hour after oral
administration of chloroform in olive oil
to humans or animals.
Following inhalation or ingestion ex-
posure, the highest concentrations of
chloroform are found in tissues with
higher lipid contents. Results from the
administration of 14C-labeled chloroform
to animals indicate that the distribution
of radioactivity (reflecting both chloroform
and its metabolities) may be affected by
the route of exposure. Oral administration
appeared to result in the accumulation of
a greater proportion of radioactivity in the
liver than did inhalation exposure. Dif-
ferences in the distribution of chloroform
and its metabolities between male and
female animals were found only in mice
and not in rats or squirrel monkeys. The
kidneys of male mice accumulated strik-
ingly more radioactivity than did those of
female mice.
Chloroform is oxidized via microsomal
cytochrome P-450 to trichloromethanol,
which spontaneously dehydrochlorinates
to the toxic and reactive intermediate
compound, phosgene. The end products
of the phosgene reaction with cellular
water are CO2 and hydrochloric acid, but
significant amounts of phosgene and
other reactive intermediates bind coval-
ently to tissue macromolecules or con-
jugate with cysteine and glutathione.
Covalent binding of the reactive inter-
mediates to macromolecules is considered
to be responsible for the hepato- and
nephrotoxicity of chloroform. While the
liver is the primary site for chloroform
metabolism, other tissues, including the
kidney, can also metabolize chloroform.
There is no evidence to suggest any
qualitative difference for chloroform
metabolic pathways in mice, rats, and
humans. Interspecies comparisons of the
magnitude of chloroform metabolism have
been made only for the oral route.
Metabolism of chloroform across species,
including mice, rats, squirrel monkeys,
and humans is proportional to the surface
area of the species. The end metabolite,
C02, is excreted in expired air. Dose-
dependent pulmonary exhalation is the
principal route of excretion for unmetabo-
lized chloroform. Small amounts of
chloroform metabolites are excreted in
the urine and feces. Results from obser-
vations in humans suggest that chloro-
form metabolism is rate limited.
Regardless of the route of entry into
the body, chloroform is excreted un-
changed through the lungs and eliminated
via metabolism, with the primary stable
metabolite, C02, also being excreted
through the lungs. High concentrations
of unchanged chloroform have been found
in the bile of squirrel monkeys after oral
administration, but not in the urine or
feces. The inorganic chloride generated
from chloroform metabolism is excreted
via the urine.
Decay curves for the pulmonary excre-
tion of unchanged chloroform in humans
appear to consist of three exponential
components. The terminal component,
thought to correspond to elimination from
adipose tissue, had a half-time of 36
hours. This long half-time of chloroform
residence in the human fat compartment
indicates that fatty tissue concentrations
of chloroform will not achieve steady-
state equilibrium conditions with exposure
concentrations until 6 to 7 days of con-
tinuous exposure to ambient concentra-
tions, or longer for repetitive daily
exposures in the workplace. Conversely,
the long residence time of chloroform in
the fat compartments of humans indicates
that complete desorption of chloroform
from these compartments requires 6 to 7
days in chloroform-free environs.
Health Effects Overview
Neurological, hepatic, renal, and cardiac
effects have been associated with ex-
posure to chloroform. These effects have
been documented in humans as well as
in experimental animals. In addition,
studies with animals indicate that chloro-
form is carcinogenic and may be
teratogenic.
Evidence of chloroform's effects on
humans has been obtained primarily
during the use of this chemical as an
inhalation anesthetic. In addition to
depression of the central nervous system,
chloroform anesthesia was associated
with cardiac arrhythmias (and some cases
of cardiac arrest), hepatic necrosis and
fatty degeneration, polyuria, albuminuria,
and in cases of severe poisoning, renal
tubular necrosis. When used for obstetri-
cal anesthesia, chloroform was likely to
produce respiratory depression in the
infant. Humans exposed experimentally
to chloroform for 20 to 30 minutes have
reported dizziness, headache, and tired-
ness at concentrations >1000 ppm, and
light intoxication at concentrations above
4000 ppm.
Similar symptoms occurred in workers
employed in the manufacture of lozenges
containing chloroform; exposure concen-
trations ranged from 20 to 237 ppm, with
occasional brief exposure to =1000 ppm.
Additional complaints were of gastro-
intestinal distress, and frequent scalding
urination. The only other report of adverse
effects stemming from occupational ex-
posure to chloroform was of enlargement
of the liver.
Acute inhalation experiments with
animals revealed that single exposures
to 100 ppm were sufficient to produce
mild hepatic effects in mice. The exposure
level that would produce mild renal effects
is not known, but toxic effects occurred
in the kidneys of male mice exposed to 5
mg/L (1025 ppm). In subchronic inhala-
tion experiments, histological evidence ol
-------
mild hepato- and nephrotoxicity occurred
in rats with exposures to as low as 25
ppm, 7 hours/day for 6 months. The
effects were reversible if exposure was
terminated, and did not occur when ex-
posure was limited to 4 hours/day.
Information on the effects of acute and
long-term oral exposure to chloroform is
available primarily from experiments with
animals. Human data are mainly in the
form of case reports and involve the
abuse of medications containing not only
chloroform, but other potentially toxic
ingredients as well; however, a fatal dose
of as little as 1 /3 ounce was reported. As
with inhalation exposure, the primary
effects of oral exposure were hepatic and
renal damage. Narcosis also occurred
with high doses. Subchronic and chronic
toxicity experiments with rats, mice, and
dogs did not clearly establish a no-effect
level of exposure for systemic toxicity.
Although a dose level of 17 mg/kg/day
of chloroform produced no adverse effect
in four strains of mice, the lowest dosage
tested, 15 mg/kg/day, elevated some
clinical chemistry indices of hepatic
damage in dogs and appeared to affect a
component of the reticuloendothelial
system (histiocytes) in their livers.
No controlled studies have been per-
formed to define dose-response thres-
holds for neurological or cardiac effects
of ingested or inhaled chloroform. It is
not known whether subtle impairment of
neurological or cardiac function might
occur at levels as low as or lower than
those which affect the liver.
Several substances that are of interest
because of accidental or intentional
human exposure have been shown to
modify the systemic toxicity of chloroform,
usually by modifying the metabolism of
chloroform to the reactive intermediate.
Examples of substances that potentiate
chloroform-induced toxicity are ethanol,
PBBs, ketones, and steroids, while those
that appear to protect against toxicity
include disulfiram and high carbohydrate
diets.
On the basis of presently available data,
no definitive conclusion can be reached
concerning the mutagenicity of chloro-
form. However, evidence from studies
measuring binding to macromolecules,
DNA damage, and mitotic arrest suggest
that chloroform may be mutagenic.
Chloroform has the potential for causing
adverse reproductive effects in pregnancy
maintenance, delays in fetal development,
and the production of terata in laboratory
animals. The studies which administered
chloroform by inhalation 7 hours/day
reported more severe outcomes than
other studies that administered chloro-
form by intubation, once or twice a day.
The adverse effects produced in the con-
ceptus were observed in association with
maternal toxicity; however, the type and
severity of effects appeared to be specific
to the conceptus, affecting development
to a much greater degree than the occur-
rence of maternal toxicity. It is concluded
that chloroform is a potential develop-
mental toxicant. The results of a pre-
liminary study indicate that chloroform
has no significant adverse behavioral
effect on the fetus and produces embryo-
toxic effects only at maternally toxic levels.
The carcinogenic potential of chloroform
has been experimentally evaluated in
several animal species and by epidemiolo-
gic surveys including chronic animal
studies. In all of these studies chloroform
was administered by the oral route and
not by inhalation. However, a carcino-
genic response from chloroform exposure
is not expected to be dependent upon the
route of assimilation into the body al-
though the magnitude of the response
may vary.
Evidence for the carcinogenicity of
chloroform in experimental animals in-
cludes: statistically significant increases
in renal epithelial tumors in male
Osborne-Mendel rats; hepatocellular
carcinomas in male and female B6C3F,
mice; kidney tumors in male ICI mice;
and hepatomas in female Strain A mice
and NIC mice. Chloroform has also been
shown to promote growth and metastasis
of murine tumors. In these cancer studies
the carcinogenicity of chloroform is
organ-specific, occurring primarily in liver
and kidney, which are also the target
organs of acute chloroform toxicity and
covalent binding.
The carcinogenicity of chloroform in
test animals was first investigated in
1945. Although the number of animals in
each test group was small and the
mortality was high at the higher doses,
an increased incidence of hepatomas was
observed in Strain A mice and confirmed
in 1967 in a study using NIC mice.
Chloroform was administered in oil by
gavage in both studies.
In 1976, male and female B6C3F, mice,
chloroform-treated by corn oil gavage,
showed highly significant dose-dependent
increases in hepatocellular carcinomas,
with metastases to the lungs in some
mice. In a similar study, statistically
significant dose-dependent increases of
kidney epithelial tumors were found in
male Osborne-Mendel rats. In another
study, kidney tumors were observed in
male ICI mice administered chloroform in
either toothpaste or arachis oil.
In the most recently published study
(1985), chloroform administered in the
drinking water of male Osborne-Mendel
rats induced a statistically significant in-
crease in the incidence of renal tumors,
thus supporting the findings from the
earlier study in which chloroform was
adminstered in corn oil by gavage. Female
BaC3F, mice, however, did not show an
increase in the incidence of liver tumors
when chloroform was administered in
the drinking water. This was inconsistent
with the positive findings reported, in
previous investigations of chloroform.oil
gavage treatment of mice. The lack of
response of the mice in the drinking
water study versus the highly significant
response of these mice when chloroform
was given in corn oil vehicle as a single
bolus, suggests that chloroform-induced
heptatocellular carcinomas in this strain
of mice may be related to chloroform
absorption patterns, the dosing regimen,
peak blood levels of chloroform, and
target tissue levels of its reactive inter-
mediate metabolites. The corn oil carrier
has not been shown to induce an increase
in the incidence of liver tumors in mice.
Other studies of chloroform carcino-
genicity have shown negative results.
Treatment with a gavage dose of chloro-
form in toothpaste did not produce a
carcinogenic response in female ICI mice
or in male mice of the CBA, C57BL, and
CF/1 strains, nor was a carcinogenic
response observed in male or female
Sprague-Dawley rats given chloroform in
toothpaste by gavage, but early mortality
was high in control and treatment groups.
Gavage doses of chloroform in toothpaste
did not cause a carcinogenic response in
male and female beagle dogs treated for
over 7 years, although there was an
increased incidence of hepatic nodular
hyperplasia. The daily chloroform doses
given to m'rce and rats in toothpaste or
arachis oil were lower than those given
in corn oil or drinking water in studies
showing a positive carcinogenic response.
In newborn (C57 x DBA2-F1) mice given
subcutaneous doses during the initial 8
days of life and observed for their life-
times, a carcinogenic effect of chloroform
was not evident. The doses levels used
appeared well below a maximum tolerated
dose and the period of treatment was
quite short. In Strain A mice, chloroform
was ineffective at maximally tolerated
and lower doses in a pulmonary adenoma
bioassay. However, other chemicals that
have shown carcinogenic activity in dif-
ferent tests were ineffective in this par-
ticular Strain A mouse pulmonary
-------
adenoma bioassay. Chloroform does not
induce transformation of Syrian baby
hamster kidney cells (BHK-21/C1 13) in
vitro.
While no cancer epidemiologic studies
have evaluated chloroform by itself,
several studies have been made of popu-
lations with chlorinated drinking water,
in which chloroform is the predominant
chlorinated hydrocarbon compound. Small
increases in rectal, bladder, and colon
cancer were consistently observed by
several case-control and ecological
studies, several of which are statistically
significant. Because other possible car-
cinogens were present along with chloro-
form, it is impossible to identify chloroform
as the sole carcinogenic agent. Therefore,
the epidemiologic evidence for the car-
cinogenicity of chloroform is considered
inadequate.
It is generally accepted that the car-
cinogenic activity of chloroform resides
in its highly reactive intermediate metabo-
lities. Irreversible binding of chloroform
metabolites to cellular macromolecules
supports several theoretical concepts of
the mechanism(s) for its carcinogenicity,
including the possibility that chloroform
may act as a promoter in animal tissues
in addition to having complete carcinogen
properties. Available data on chloroform
metabolism and pharmacokinetics per-
tinent to the conditions of the carcino-
genicity bioassays are used in the
extrapolation of the dose-carcinogenic
response relationships of laboratory
animals to humans. There is no difference
in absorption of chloroform across species.
Also, there is no evidence to suggest any
qualitative difference in the metabolic
pathways or profiles of mice, rats, and
humans for chloroform. An experimental
basis exists for determining relative
amounts of chloroform metabolized in
various species, including man, and this
information has been used in the unit
risk derivation for chloroform.
Based on EPA's proposed Carcinogen
Risk Assessment Guidelines, chloroform
is classified as having sufficient animal
evidence for carcinogenicity and inade-
quate epidemiologic evidence. The overall
wetght-of-evidence classification is group
B2, meaning that chloroform is probably
carcinogenic in humans.
The derivation of cancer risk values is
based on the assumption of a nonthres-
hold mechanism for cancer induction,
and consequently mathematical extrapo-
lation models consistent with this as-
sumption are utilized.
Five data sets are used to estimate the
carcinogenic risk of chloroform. The end
points include liver tumors in female mice,
liver tumors in male mice, kidney tumors
in male rats, and kidney tumors in male
mice. The unit risk values at 1 mg/kg/day,
calculated by the linearized multistage
model on the basis of these data sets, are
comparable. The risk value is useful for
estimating the possible magnitude of the
public health impact. The upper-bound
incremental cancer risk derived from the
geomatric mean of 4 data sets, chloroform
gavage studies which showed a statisti-
cally significant increase of hepatocellular
carcinomas in mice, is 8.1 x 10~2 per
mg/kg/day. The carcinogen cancer as-
sessment group (CAG) potency index for
chloroform (defined as the slope x mole-
cular weight) is 1 x 101, ranking it in the
lowest quartile of 55 chemicals that the
CAG has evaluated as suspect carcino-
gens. The upper-bound estimate of the
incremental cancer risk due to ingesting
1 ftg/L of chloroform in drinking water is
2.3 x 10'6. The upper-bound estimate of
the incremental cancer risk due to in-
haling 1 Aig/m3 of chloroform in air based
upon positive gavage carcinogenicity
studies is 2.3 x 1f>5. The upper-bound
nature of these estimates is such that the
true risk is not likely to exceed this value
and may be lower.
Although the nonthreshold mathe-
matical risk extrapolation model is con-
servative based upon a public health point
of view, the correction used in the cal-
culation of a human equivalent dose is
scientifically conservative and may lead
to an overestimate of the amount of
chloroform metabolized in the test
animals, and hence underestimate the
risk. In addition, experimental data that
include covalent binding in human tissues
suggest that humans may have a greater
than expected capacity to metabolize
chloroform when compared to rodents,
again indicating the possibility of under-
estimating the risk for humans.
This Project Summary was prepared by staff of Environmental Criteria and
Assessment Office, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711.
Si Duk Lee is the EPA Project Officer (see below/
The complete report, entitled "Health Assessment Document for Chloroform,"
(Order No. PB 86-105004; Cost: $36.95, subject to change/ will be available
only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Officer can be contacted at:
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S8-84/004F
00003^9 PS
------- |