-------
f
I
a."
S1
• 1
.00300 25°C .00400
1/T °K
Source: Arthur 0. Little, Inc., based on Hodgman (1962).
FIGURE 3 VAPOR PRESSURES OF TRIHALOMETHANES AS A FUNCTION OF TEMPERATURE
29
-------
Chloroform;
Bromoform;
. . -2090 + 7.83, r » 0.9999
log P a —-—
where P equals vapor pressure (mm Hg) and T equals temperature (°K).
The vapor pressures of chloroform and bromoform at 25°C were cal-
culated from these equations. The slopes of these lines correspond to
latent heats of vaporization of 7.9 fecal/mole for chloroform and 9.6
kcal/moie .for bromoform. In order to estimate the 25*C vapor pressures
of the mixed trihalomethanes, the latent heat of vaporization was
estimated to be 8.7 kcal/mole, the average of the corresponding values
for CHC13 and CHBr^. The estimated latent heat of vaporization (AHvap)
was then combined with the reported boiling point data (Hodgman 1962)
in order to estimate vapor pressures of 50 mm Hg for bromodichloromethane
and 20 mm Hg for dibromochloromethane at 25°C.
The probable range of solubilities for the mixed bromo/chloro
trihalomethanes may be defined by the 65 mM saturation concentration
for-chloroform in water (Lyman 1978) and the 4.9 mM value for bromoform
(Stecher et._§!. 1968). An average solubility of about 30 mM was esti-
mated for the mixed trihalomethanes. This would correspond to about
5000 mg/1 of bromodichloromethane and about 6000 mg/1 of dibromochloro-
methane at saturation.
The octanol-water partition coefficient provides a useful indica-
tion of tendencies for water-sediment and water-lipid partitioning.
Literature data on this property were available only for chloroform (90,
Hansch and Leo 1979). The value shown in Table 3 for bromoform was
estimated from that compound's water solubility, according.to the linear
free-energy relationship approach of Hansch_et_al. (1968). For alkyl
halides, the Hansch .et al. (1968) correlation between S (molar) and
partition coefficient (dimensionless) is:
log P = 0.681 - 0.819 log S.
The estimated value of P for bromoform is, therefore, 370. The parti-
tion coefficient, like the vapor pressure and solubility, appears to
vary by no more than an order of magnitude for the chemicals under
consideration. A value of 250 was selected as an estimate for CHBrCl2
and CHBr2Cl.
The Henry's Law constant listed in Table 3 is simply the ratio of
che vapor pressure at 25°C co Che water solubility (with no corrections
30
-------
for activity coefficients). This parameter is fundamentally related to
the tendency of a solute to escape from water into the atmosphere. Once
again, the range in values for chloroform and bromofora is not laree. An
average value of 2000 mm Hg - L/mole was assumed for CHBr?Cl and CHBrC^-
C. MONITORING DATA
1. Introduction
This section summarizes data regarding the presence of trihalometh-
ane in the environmental media air, water, and biota. In general, limited
data are available, especially for bromoform, bromodichloromethane, and
dlbromochloromethane, and no data were found regarding levels of trihalo-
methanes in soil. Levels of chloroform in POTW influent and effluent are
discussed in Section IV.D.7.
2. Water
Seawater
The only monitoring data available for trihalomethanes in seawater
were for chloroform. Pearson and McConnell (1975) found a maximum
chloroform concentration of 1.0 ug/1 in Liverpool Bay, U.K. Su- and
Goldberg (1976) found mean levels of 0.05 ug/1 in open waters of the
East Pacific, and average concentrations of 0.009-0.012 ug/1 closer to
shore. Murray and Riley (1973) reported mean chloroform concentrations
of 0.0008 ug/1 in the Northeast Atlantic.
b. Freshwater
STORET data provided the most complete survey of ambient concentra-
tions of trihalomethanes in surface waters of the U.S. However, the
data are far from complete: only 168 unremarked^ measurements have been
reported for chloroform, the most commonly monitored trihalomethane.
Summaries of the STORET data concerning trihalomethanes are given
in Tables 4-7. Most of the chloroform concentrations were between 1 ug/1
and 10 ug/1, and only rarely did they exceed 100 ug/1. Most of the data
for bromoform were remarked; almost all of these concentrations were
between 1 ug/1 and 10 ug/1 as well. For bromodichloromethane, the
majority of the measured concentrations fell between 0.1 ug/1 and 1.0
ug/1, with the exception of the Pacific Northwest, for which the reported
levels were approximately one order of magnitude higher. Two-thirds of
A large number of data entries for trihalomethanes in the STORET system
are "remarked." This means that some note regarding the data has been
entered. In general, the notation refers to a value less than the
detection liait. Thus, there is a large amount of uncertainty associ-
ated with "remarked" data.
31
-------
TABLE 4. CONCENTRATIONS OF TOTAL CHLOROFORM DETECTED
IN SURFACE WATERS OF THE U.S., 1970-1979
River Basin
•
New England
Mid Atlantic
Southeast
Great Lakes
Ohio
Tennessee
Upper Mississippi
Lower Mississippi
Souris and Red of North
Missouri
Arkansas and Red
Western Gulf
Rio Grande and Pecos
Upper.Colorado
Lower Colorado
Great Basin
Pacific Northwest
California
Alaska
Hawaii
United States 168 100 39 49 12 <1
Unremarked data.
2
Remarked data.
w<
Sami
U1
59
7
20
15
7
22
9
7
7
15
3
sles (
R2 :
4
8
3
16
12
4
3
1
5
2
2
2
2
1
24
11
3.1-1 us/1
100
18
30
47
41
14
23
33
100
100
100
33
100
100
100
19
92
1-10 us/1 10-100 uc/1
0
67
70
39
59
43
27
67
0
0
0
67
0
0
0
81
8
0
15
0
14
0
43
46
0
0
0
0
0
0
0
0
0
0
100-1.000 ut., i
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
Source: U.S. Environmental Protection Agency, STORET (1979)
32
-------
TABLE 5. CONCENTRATIONS OF TOTAL BROMOFORM DETECTED
IN SURFACE WATERS OF THE U.S., 1970-1979
River Basin
New England
Mid Atlantic
Southeast
Great Lakes
Ohio
Tennessee
Upper Mississippi
Lower Mississippi
Souris and Red of North
Missouri
Arkansas and Red
Western Gulf*
Rio Grande and Pecos
Upper Colorado
Lower Colorado
Great Basin
Pacific Northwest
California
Alaska
Hawaii
Slo Percentage of Observations at Concentration
Samples
Ul R*
4
4 12
4
8
18
1 1
6
4
1
6
' 2
4
2
2.
1
29
12 12
0.1-1 us/1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
1-10 ug/1
100
100
100
100
100
50
100
100
100
100
100
75
100
100
100
97
100
10-100 us/1
0
0
0
0
0
50
0
0
0
0
0
25
0
0
0
0
0
United States
17 114
97
"TJnremarked data.
remarked data.
Source: U.S. Environmental Procection Agency, 5TORET (1979)
33
-------
TABLE 6. CONCENTRATIONS OF TOTAL BROMODICHLOROMETHANE
DETECTED IN SURFACE WATERS OF THE U.S., 1970-1979
River Basin
New England
Mid Atlantic
Southeast
Great Lakes
Ohio
Tennessee
Upper Mississippi
Lower Mississippi
Souris and Red of North
Missouri
Arkansas and Red
Western Gulf
Rio Grande and Pecos
Upper Colorado
Lower Colorado
Great Basin
Pacific Northwest
California
Alaska
Hawaii
No Percentage of Observations at Concentration
Samoles
U-L
4
1
2
2
2
2
i
.2
R*
4
16
4
7
18
1
6
4
1
6
2
3
2
2
1
29
12
0.1-1 ug/1
100
60
100
88
60
33
75
67
100
100
100
100
100
100
100
14
100
1-10 us/1
0
35
0
0
40
0
25
17
0
0
0
0
0
0
0
86
0
.10-100 ug/1
0
5
0
12
0
67
0
16
0
0
0
0
0
0
0
0-
0
United States
Unremarked data.
'Remarked data.
25 118
66
31
Source: U.S. Environmental Protection Agency, STORET (1979)
-------
TABLE 7. CONCENTRATIONS OF TOTAL DIBROMOCHLORMETHANE
DETECTED IN SURFACE WATERS OF THE U.S., 1970-1979
River Basin
New England
Mid Atlantic
Southeast
Great Lakes
Ohio
Tennessee
Upper Mississippi
Lower Mississippi
Souris and Red of North
Missouri
Arkansas and Red
Western Gulf
Rio Grande and Pecos
Upper Colorado
Lower Colorado
Great Basin
Pacific Northwest
California
Alaska
Hawaii
No Percentage of Observations at Concentration
Samples
Ul R2
4
2 12
4
1 7
18
1
6
4
h 1
6
2
3
2
2
1
29
12 12
1
0.1-1.0 us/1
100
86
100
.88
67
100
100
100
100
100
100
100
100
100
100
17
25
100
1-10 us/1
0
14
0
12
33
0
0
0
0
0
0
0
0
0
0
83
75
0
United States
15 115
80
20
hremarked Data.
?
"Remarked Data.
Source: U.S. Environmental Protection Agency, STORET (1979)
35
-------
Che dibromochloromethane concentrations also fell between 0.1 ug/1 and
1.0 ug/1, although for California and the Pacific Northwest concentra-
tions were roughly one order of magnitude greater. The apparent higher
levels of these two chemicals in the Pacific Northwest and California are
probably due to a higher detection limit for data from these areas.
Chloroform is the only trihalomethane for which' data were found on
concentrations in precipitation. Pearson and McConnell (1975) reported
maximum concentrations of 0.2 ug/1 in rainfall near Runcorn, U.K. Su
and Goldberg (1976) found a mean level of 17 ug/1 in rainwater collected
in Southern California. The same authors reported chloroform levels in
snow collected in different parts of North America, ranging from 3 ug/1
to 90 ug/1.
3. Sediment
The only information available on trihalomethane concentrations in
sediment was from Pearson and McConnell (1975) who found maximum chloro-
form residues of 4 mg/kg in Liverpool Bay, U.K.
4. Aquatic Organisms
STORE! data on trihalomethane residues in fish were limited. Only
one unremarked measurement was reported for each of the four compounds.
All four measurements were taken in the Pacific Northwest basin, and
the concentrations reported all fell between 1 mg/kg and 10 mg/kg. The
remarked data, however, indicated concentrations from 100 mg/kg up to
100 mg/kg for each of the compounds.
Chloroform is the only trihalomethane compound for which.any
detailed residue data were found. The data of Pearson and McConnell
(1975) as shown in Table 8 include residue analyses for 21 species of
marine finfish' and shellfish, as well as plankton. The plankton had the
lowest overall residue levels (0.02-5 ug/kg, wet weight); shellfish
residues ranged from 2 ug/kg to 180 ug/kg. Residues in finfish were
analyzed in flesh, liver, and gastrointestinal tract. Residues of up to
50 ug/kg were found in mackerel flesh; concentrations in liver and
gastrointestinal tract were generally one-third to one-tenth of those
found in flesh.
5. Air
Chloroform is the only trihalomethane compound for which data on
concentrations in air were found. Harsch .et_al. (1977) have reported
that continental background levels of chloroform in the atmosphere
range between 0.04 ug/m3 and 0.13 ug/m^. Oceanic background levels, as
reported in NAS (1978), are somewhat higher, with mean concentrations
ranging from 0.13 ug/m^ to 0.20 ug/m^.
In the urban environment, ambient atmospheric levels of chlorofom
may be considerably higher as a result of anthropogenic sources. The
data of Harsch et_ al. (1977) suggesc that concentrations in city air
are highly dependent upon automobile activity. Urban air (Pullman,
36
-------
TABLE 8. CHLOROFORM RESIDUES DETECTED IN MARINE ORGANISMS
Organism
Chloroform Concentration
(ug/kg wee weight)
Plankton 0.02-5
Mussel (Mytilus edulis) 3-10
Cockle (Cerastodenna, edule) 4-150
Oyster (Ostrea edulis) 3
Whelk (Buceinum undatum) • 10
Slipper limpet (Crepidula fornicata) ' • 6
Crab (Cancer pagurus) 3-180
Shore crab (Carcinus maenas) 15
Hermit crab (Eupagurus bernhardus) 20-73
Sand shrimp (Crangon crangon) 45
Starfish (Asterias rubens) 13
Sunstar (Solaster sp.) 3
Sea urchin (Echinus esculentus) 2
Flounder (Platychthys flesus). flesh/liver 21/6
Mackerel (Scomber scomb'rus), flesh/liver 50/18
Dab (Limanda limanda), flesh 23
Plaice (Pleuronectes platessa). flesh 17
Sole (Solea solea). flesh/gastrointestinal tract 26/9
Red gurnard (Aspitrigla cuculus). flesh/gastrointestinal 21/2
Scad (Trachurus trachurus), flesh 48
Pout (Trisopterus luscus), flesh 15
Spurdog (Squalus acanthias), flesh 110
Source: Pearson and McConnell (1975)
-------
Washington) in the early morning contained 0.04 ug/m^ chloroform; in
light traffic the level was 0.10 ug/m3; and in heavy traffic with no
breeze, the concentration rose to 0.44 ug/m^. Chloroform levels in
automobile exhaust varied between 0.32 ug/m3 and 33 ug/m^, depending on
car model and year. (The car equipped with a catalytic converter emitted
fumes with chloroform levels approximately two orders of magnitude lower
than a car that lacked this device.) Lillian est, al. (1975) found a
wider range of urban ambient levels, ranging from <0.05 ug/m^ to 74 ug/m^;
the maximum value was reported for Bayonne, N.J.
Pellizzari et al. (1979) have sampled four highly industrial areas
for halogenated organic compounds. Chloroform was commonly found in
these samples, although none of the other trihalomethanes was reported.
Table 9 summarizes the results. It can be seen that chloroform levels
in air were highly variable, with a maximum of 139 ug/m3 reported in the
New Jersey area.
X
Harsch and Rasmussen (1977) have conducted extensive monitoring in
various indoor environments,.particularly in commercial establishments.
Concentrations ranged from 0.07 ug/m3 in the dining area of a restaurant
to 3.6 ug/m3 in the aisle of a food store containing household cleaners.
However, chloroform levels rarely exceeded 0.5 ug/m3, and most concen-
trations were similar to tropospheric background levels.
However, Pellizzari e,t al. (1979) sampled basements of houses in
the Old Love Canal area, and found levels of chloroform ranging from
0.46 ug/m3 to 13.48 ug/m3 in the seven samples taken. Upstairs in the
same homes concentrations from trace levels to 15.3 ug/nn chloroform were
found.
D. ENVIRONMENTAL PATHWAYS AND FATE
1. Overview
This section examines the sources of trihalomethanes released to
the environment, as discussed in Chapter III, the chemistry of their
inadvertent production, and the fate of these chemicals in various
environmental compartments. Most of this discussion is centered around
chloroform, but the general fate of the other trihalomethanes is expected
to be similar.
2. Sources to the Environment
The largest of the identified sources of chloroform released to
the environment are pulp and paper bleaching, the chlorination of
drinking water and wastewater, and the use of this chemical in pharma-
ceutical extractions. The greater portion of these releases goes to
air (-^17,000 kkg/yr), while about 900 kkg/yr are released to water.
figure 4 shows a general representation of the releases of chloroform,
as well as their ultimate distribution in the environment.
38
-------
TABLE 9. CHLOROFORM CONCENTRATIONS DETECTED IN
AMBIENT AIR OF INDUSTRIAL AREAS
Concentration
(ug/n»3)
City Mean Range No. Detected/No. Samples
Niagara Falls and Buffalo, N.Y. 89,000 1050-105,461 9/9
Rahway/Woodbridge, Boundbrook 47,000 T - 98.625'
and Fassaic, N.J.
75/93
Baton Rouge, Geismar, and
Plaquemine, LA
5,500 181-11,742
78/87
Houston, Deer Park and
Pasadena, TX
1,000 T - 53,846
29/30
irace
Source: Pellizzari et al. (1979)
39
-------
4--
O
Ti*nt|MMl 11 mil
l Souitci
Reaction wviih hydfoxy rtdicnl
(lu~2 3 months)
Aic< Sourc*
Aiim.t|4iciii:
Lniiuiuni
Point
Souic*
AllllUiptKliC
tniniM.cn
Wei and Dry
fdlloul
SulljCt Wilui 4111!
Phoiochnmical UvyitilHiou in kVaiot
litowl
Traiitpoii 10 Deep Waioci wd SoJiincnls sli
(lor.l n:s.iloi»:u limit)
I e jthliuj 10 0«
-------
3. Inadvertent Formation of Trlhalome thanes
Since Che formation of trihalomethanes upon chlorination has been
shown to be an important source of these chemicals in the environment,
it is important to consider the chemistry of these reactions.
Fulvic and humic acids constitute the major portion of the organic
matter in surface or ground waters and have been demonstrated to be
precursors for halo forms when treated with chlorine (Rook 1974) . The
reactions cited are direct analogs to the haloform reaction; these
reactions are initiated by the hypochlorite Ion present in chlorinated
waters .
One suggested mechanism for trihalome thane formation during
industrial bleaching and water purification processes is given in
reaction 1 (Hendrickson .et al. 1970) . This reaction is a direct P,
analog of the haloform reaction and is seen with methyl ketones (-C-CB^) or
compounds that can be readily oxidized to produce methyl ketones. The
reaction is base catalyzed; since each halogen introduced increases the
acidity of the remaining hydrogens, the intermediates are unstable and
trihalomethanes are produced.
0 0
R-C-CH3 + 3X2 + 40H -»• CHX3 + 3X + R-C-0 + 3H20 (1)
where X = Cl and/ or Br (Hendrickson £t al. 1970)
The presence of bromide ion in aqueous systems has been shown to
reduce substantially the yield of chloroform, while increasing levels of
dichlorobromomethane, bromodichlorome thane, and bromoform (Bunn et al.
1975). Available hypochlorite ion oxidizes bromide to give bromine,
which then undergoes the "haloform" type reaction to form tne appropri-
ate bromochloromethanes as shown in reaction 1.
The above mechanism is certainly applicable to industrial chlorina-
tion processes; however, the pH range of municipal water chlorination
is much closer to neutral. For this reason, and the observation that a
variety of degradation products other than trihalomethanes were produced
in nonbasic aqueous solutions, a mechanism other than the base-catalyzed
halogenation was suspected. For chlorination in aqueous solutions, Rook
(1977) proposes a slightly different pathway and has determined that the
most reactive site for the haloform reaction with fulvic acids at
neutral pH is at the carbon between two meta-OH groups. His mechanism
is outlined in reaction 2. The incorporation of bromine is according
to the pathway outlined above.
o
OH UOH
- 4 HOC:
'(Rook 1977) (2)
C-OH
-------
4. Fate in the Atmospheric Environment
\
Versar (1979a,b) has concluded on Che basis of a review of Che licer-
aCure that;"once chloroform is in Che troposphere, reaction with hydroxyl
radicals is;1, Che primary mechanism for removal. They reported, based upon
Che work o$ Spence at «1. (1976), Hansc (1978), and Che U.S. EPA (1975),
chac che produce of chis reaccion is a CC13 radical, which is oxidized
Co produce phosgene (COC12) and chlorine oxide (CIO). The estimated
lifetime is reported Co be 2-3 months, based on reported lifetimes of
0.19 years and 0.32 years reported by Cox et al. (1976) and Yung et al.
(1975). Bromo-alkanes generally have rate conscants for reaction with
hydroxyl radical that are the same or slightly higher than those for
the chlorinated analogs. Thus, the atmospheric residence times of
CHBrCl2» CHBr2Cl and CHBr3 are also expected Co be on Che order of 2 co
3 months.
Direct photochemical degradation of trihalomethanes in the tropo-
sphere is not likely to be important because these species do not absorb
UV radiation at wavelengths that penetrate the ozone layer. The esti-
mated atmospheric residence time is sufficiently short that extensive
transport to the stratosphere, where direct photochemical process could
occur, may not be a major pathway. Export from the atmosphere by rainout
is a ..possible, but not probable fate. The magnitude of the vapor pres-
sure in comparison with the solubility suggests that an insignificant
fraction of chloroform in the atmosphere would be associated with water
droplets or dust particles. Since the other trihalomethanes have similar
Henry's Law constants, they also are expected to be resistant to rainout.
Thus, trihalomethanes reaching the atmosphere may travel consider-
able distances before reaction with the hydroxyl radical results in
degradation.
5. Fate in the Aquatic Environment
Of the numerous modes by which the trihalomethanes might be removed
from the aquatic environment, volatilization is expected to be the most
significant. A laboratory study by Billing et. al. (1975) indicated that
the half-life for evaporation of chloroform from a stirred aqueous
solution was about 20 minutes, compared with a 1-minute prediction based
on vapor pressure and solubility. In fact, the results obtained strongly
suggest that the mass transport phenomena were limiting the evaporation
rate, since the same half-life (20 minutes) was observed for several
alkylhalides with vapor pressures from 19 mm to 426 mm.
The rate conscants for hydrolysis of the bromo/chloro trihalometh-
anes, as compiled by Habey and Mill (1977), are summarized in Table 10.
The reactions are base catalyzed, but the race constants for che basic
hydrolysis are aoc high enough to decrease significantly the predicted
hydrolytic half-life within the normal environmental ?H range (?H 5-3
for most natural waters). Chloroform, for example, has a predicted half-
life of_>500 days at pH 7 due to neutral hydrolysis, while the basic
42
-------
TABLE 10. RATE CONSTANTS FOR HYDROLYSIS OF TRIHALOMETHANES
Rate Constant
Temp. Neutral Basic
Compound (°C) lO^sec"1 104kOHM~1sec"L
Chloroform 25 1.621
100 7.29
25 0.60
Bromodichloromethane 25 16.0
Dibromochloromethane 25 8.01
Bromoform '25 3.20
Considered to be "suspect" value by Mabey and Mill (1977)
because of poor temperature control during measurement.
If the 100'C value is considered to be reliable, a 25° value
much lower than 1.62 (ca 0.07 perhaps) would be more likely
for a reaction with a typical activation energy.
Source: Mabey and Mill (1977)
-------
hydrolysis reaction at pH 8 vould give a 500-year half-Life. The base-
catalyzed reaction rate constants of Table 10 are included because they-;
provide one indication of the relative reactivity of the trihalomethanes.
With regard to basic hydrolysis, Che relative reactivities are: chloror-
form (1), bromoform (5), dibromochloromethane (13), bromodichloromethane
(27). It is not clear that the same order of magnitude vould be observed
for the neutral reactions, because the reaction mechanism may be signifiV
cantly different. However, it appears probable that the mixed trihalo-
methanes have considerably shorter hydrolytic half-lives than bromoform
or chloroform. It also appears likely that hydrolysis is not fast
enough under any conditions to compete with volatilization as a pathway
for loss from the aquatic compartment.
The octanol/water partition coefficient of 90 suggests that a
significant portion of chloroform in the hydrosphere is associated with
suspended solids and sediments (Lyman 1978). However, there is no
experimental evidence to support this as an important pathway, and no
monitoring data exist 'for chloroform in sediment. Thus, it is not clear
that adsorption competes with volatilization as an important fate path-
way.
Similarly, no information is available regarding the biodegradation
of trihalomethane in aqueous systems. However, again it is unlikely that
this mechanism could compete with volatilization.
Bioaccumulation does not appear to represent an important fate
pathway, although the octanol-water partition coefficient of 90 suggests
that accumulation may occur. In a summary of bioaccumulation studies on
various aquatic animals, U.S. EPA (1978) found a mean bioconcentration
factor of 6, following 14 days of exposure to a 110 ug/1 solution of
chloroform. The half-life of chloroform in the tissues was found to be
less than 24 hours. Preliminary accumulation tests on channel catfish
by Anderson _et_al. (1979) indicated a bioconcentration factor of approxi-
mately 10 over ambient water levels, with the chloroform levels in fat,
eggs, and gastrointestinal system at least twice those of levels in
other tissues.
Anderson et al. (1979) also conducted a 28-day uptake/28-day
depuration study of bromoform, using menhaden, oysters, shrimp, and two
species of clams (all marine species). After 24 hours or exposure to
bromoform, all species were found to have bromoform residues in the
tissues, and within a 48-hour depuration period bromoform residues
disappeared. In the highest bromoform concentrations (0.21 mg/1 to
19 mg/1, depending on the species), tissue levels were usually lower
than those in the water; in low concentrations (0.03 mg/1) the reverse
was true.
6. Fate in Soil and Sediment
Little is known regarding the fate of trihalomethanes in soils and
sediment. Volatilization is likely from surface soils, and is probably
-------
Che dominant pathway. However, some movement may occur, as evidenced by
some detectable levels of chloroform in ground waters (Coniglio e£ al.
1979).
Lyman (1978) suggests that some chemical degradation may occur.
Adsorption is also not likely to be important because of the high solu-
bility and vapor pressure of the trihalomethanes. The importance of
biodegradation is unknown.
One recent study by Wilson et al. (1980) investigated the fate of
chloroform and dichlorobromomethane in a sandy soil. In a soil column
receiving solutions of these chemicals, most was volatilized (66-122%)
and 31-48% was recovered in the column effluent. Actually, more was
recovered than applied. This excess was attributed to an overestimation
of the amounts volatilized. Degradation in the column was thought to
be insignificant. These authors calculated retardation factors, which
represent the interstitial water velocity/velocity of the pollutant.
The finding that these factors were less than 1.5 for both chemicals
suggests that little adsorption was occurring.
Thus, trihalomethanes are likely to be volatilized from surface
soils. Otherwise, they are subject to migration into aquifers.
7. Fate in Water Distribution Systems
Because of the importance of the chlorination process in trihalo-
methane formation, some consideration was given to the fate of trihalo-
methanes in water distribution systems. An EPA study of the sources of
priority pollutants found in publicly owned treatment works provides a
means of estimating the extent of loss and regeneration of trihalomethanes
in aqueous solutions (Levins it al. 1979a). Samples were collected at the
POTW, as well as at various source locations (residential, commercial, and
industrial). The summary data from four separate studies are presented in
Table 11. The concentrations of the trihalomethanes were consistently
higher in the samples of tap water than in the POTW influent samples, indi-
cating either dilution or volatilization is occurring. The relative
proportions of the concentrations of the four trihalomethanes are fairly
consistent from the tap to the sewers to the influent. Data for bromo-
form are inconclusive because of the very low concentrations reported.
The data indicate that the rates of dilution or volatilization are similar
for the four compounds. They also suggest that levels of these compounds
are not being augmented by mechanisms other than chlorination. The sources
sampled in this study did not include any paper and pulp industries.
Table 12 presents the mass flows for one city (Levins e_t al_. 1979b).
It can be seen that significant percentages of the trihalomethanes pro-
duced on chlorination are lost between the tap and the influent, either
through volatilization or consumptive use of tap water. Furthermore, a
comparison of the sum of the mass flows for each of the scaled-up source
categories with the mass flow at the influent shows that very little loss
takes place within the sewer system.
-------
TABLE 11. TRIIIA1.0METIIANE CONCENTRATIONS AT VARIOUS LOCATIONS
IN POTW SYSTEMS (averaged for four cities)
Compound
Location
Tap
Water
Sewers
Serving
Residential
areas
Sewers
Serving
Commercial
areas
Sewers
Serving
Industrial
areas
Influent to
POTW
Det. Mean Det. Mean Det. Mean Det. Mean Det. Mean
Freq. Cone. Freq. Cone. ' Freq. Cone. Freq. Cone. Preq. Cone.
(%) (UR/D (%) (iie/l) (%) (ug/l) (Z) dig/1) (%) (iiB/l)
Chloroform
100
27
91
100
100
12
100
Itr01110J iclilorometliane
100
50
57
11
0.2
llromoform
33
0.8
I)J liromoclilorometliane
58
43
0.7 57
22
0.2
Source: Levins et al. (1979a)
-------
TABLE 12. MASS FLOW ANALYSIS OF POTW
DATA FOR TRIHALOMETHANES
Tap .
Compound Water'
Mass Flow (kg/dav)
Known
Sources to Influent Co
POTW3 POTW4
Influent/Sum Tap/Influent
Chloroform
1.85
0.45
0.54
1.2
3.4
Bromodichloromethane
0.06
1.0
18.3
Oibromochloromethane 0. 71
0.04
0.09
2.2
7.9
Calculated by use of the assumption that tap flow equals influent flow.
flow of trihalomethanes at the tap.
Sum of all mass flows to POTW.
Measure of mass flow at influent to POTW.
Source: Levins £t al. (1979b)
-------
E. MODELING OF CHLOROFORM IN THE ENVIRONMENT
1. Estimated Environmental Distribution
A preliminary estimate of the environmental distribution of the
bromo/chloro trihalomethanes was attempted by application of the simple
model of Neely (1978). Neely's model consists of regression equations
that correlate the percentage distribution of a chemical among air,
water, and soil compartments with the Henry's Law constant and the
aqueous solubility. These equations predict that, at equilibrium, all
of the halomethanes under consideration would be almost entirely in the
air compartment, with 1% or less in either the water or soil compart-
ments. It is important to note, however, that the equilibrium distribu-
tion among phases may not be attained in an actual situation. For
volatile species such as trihalomethanes, the intercompartmental
distribution is affected by kinetic, as well as thermodynamic, factors.
The probable distribution of the residual amounts of trihalomethanes
between water and soil/sediment within the aqueous compartment can be
predicted from the soil adsorption models developed at the U.S. EPA
Athens laboratory (Means €£ al. 1978, Karickhoff ej: al. 1979). This
suggests that the ratio of sediment concentration to water concentration
is related to the water-octanol partition coefficient:
£
sediment „ _P^ x organic content of sediment
c i. o*
water
For a typical river sediment containing 2% organic carbon or less,
the sediment to water concentration ratio for the halomethanes ranges
from about 1 (chloroform) to about 5 (bromoform). Trace amounts of
trihalomethanes that have not (yet) volatilized may thus be expected to
be found in both soil and aqueous phases.
2. EXAMS Model Results
For the purpose of examining the probable fate of chloroform in
various aquatic environments under conditions of continuous discharge,
the EXAMS (Exposure Assessment Modelling System; U.S. EPA-SERL, Athens,
Ga.) model AETOX 1 was implemented.
Rate constants for the primary processes thought to influence the
fate of chloroform in aqueous environments were estimated (SRI 1980)
and are presented in Table 13.
A loading rate for trlchloromethane was estimated using the follow-
ing assumptions for a pulping plant: 0.1 Ib trichloromethane discharged
per ton of pulp production, 500 tons of production per day, therefore,
2 Ibs or 1 kg trichloromethane released per hour (JRB Associates 1930
and Andrew Caren, NCASI, personal connaunieation 1980). This astizrate
is not meant to be representative of the industry but to provide a
reasonable example of a discharge rata.)
48
-------
TABLE 13. RATE CONSTANTS USED IN EXAMS ANALYSIS
OF THE AQUATIC FATE OF CHLOROFORM
Process
Race Constant
Units
Volatilization 0.583
dimensionless
Hydrolysis (base-catalyzed) 0.23
mole [OH"]-1 hr"1
Hydrolysis (neutral) 5.40xlO"9
hr
-1
Oxidation (water column) 0.70
mole (oxidants) hr
Oxidation (sediment) 360.0
mole (oxidants)' hr
Source: SRI (1980)
49
-------
Six prototype systems were simulated in order to provide a range
of environmental conditions: pond, eutrophic and oligotrophic lakes,
and three river systems (average, turbid, and coastal plain).
As would be expected, in relatively static systems (pond and lakes)
in which physical transport processes did not dominate, volatilization
was the most important removal mechanism, responsible for a 93-96% loss
of the equilibrium chloroform mass. Table 14 presents information on
the distribution and transformation of chloroform in the different
systems. In all the river systems (1-km segments), transport downstream
alone accounted for at least 85% of the removal, and at a much faster
rate than volatilization (on the order of hours rather than days). The
overall time for self-purification (following cessation of discharge)
was thus over 2 months for the lakes, approximately 1 month for the
pond, and less than 2 days for the rivers.
Table 15 presents the simulated chloroform concentrations in differ-
ent environmental compartments (water column, sediment, plankton and
benthos) at steady-state conditions. Water concentrations were approxi-
mately 0.1-3 mg/1 in the pond and lake systems and considerably lower
(due to dilution and flow rates), 1-10 ug/1, in the river systems.
Sediment concentrations were variable, ranging from -U3.8 ug/kg up to
about 8* mg/kg. The highest levels were reached in the pond, where the
rates of both volatilization and physical transport were not fast enough
to prevent accumulation slightly above water column concentrations.
Concentrations in biota were generally one to two orders of magnitude
above levels in water.
Based on the results of the EXAMS run and dependent on the assump-
tions of the model and the rate constants used as input, some conclusions
can be drawn about the potential environmental behavior of chloroform.
In relatively static systems with slow flow rates (e.g. lakes), persis-
tence is a function of volatilization. In more dynamic river systems,
physical transport processes are much more competitive for chloroform,
removing it before volatilization can reduce the mass significantly.
Transformation processes and bioaccumulation will not significantly
reduce water concentrations. The sediment layer does not appear to
absorb chloroform at levels much above water concentrations; in fact,
these levels are sometimes lower than water levels. Since volatilization
is such an important parameter, then the environmental factors that
Increase volatilization rates — conditions of high temperature, high
wind speeds, turbulence of the water, and others — would increase the
rate of loss of chloroform.
Figure 5 further illustrates the difference between a static sys-
tem's (eutrophic lake) and a dynamic system's (river) response following
the termination of a chloroform discharge (after a system equilibrium has
been reached). Within the 12 hours following the last discharge, there
is little change in chloroform concentration that is primarily controlled
by volatilization. In Che river system, however, physical transport
processes reduce the concentration significantly by five orders of
50
-------
TABLE 14. Tilt PATE OF CIILOKOFOKM IN VARIOUS CliNKKAUZEU A()UATIC SYSTEMS
Percent Difitribiitlan
Percent Lost by Various Processes
Plllltl
till iii|>lili. l.ilku
III IKIII mplili: l.lliiLint'a Initial illstrlhutIon.
the results
cleans In'g
-------
TABLE IS. STEADY-STATE CONCENTRATIONS OK CIII.OROFOKH IN VARIOUS GENERALIZED
AQUATIC SYSTEMS RESULTING FROM CONTINUOUS DISCHARGE AT KATE OP 1.0 kg/ltr
Maximum Concentrations
Sjalciu l.ujdlilB
Uater
Dissolved
JEIL'IL
Mater
Total
(«e/l)
Bottom
Sediment
lit
Klvel
1 .O kg hr
-1
2.S
0. IJ
2.S
0.13
1.4
7.7
5.1 X 10~3 1.2 X 10-2
0. U 0.13 1.7 X ID"3 4.1 X 10~3
9.9 X IO-* 9.9 X 10'* 3.0 X 10~4 1.1 X 10'3
38
1.9
2.0
22
7.9 x 10-2
2.6 X l(l~2
55
300
340
9.9 X
9.9 X 10-* 6.3 X 10~* 8.6 X UT4
1.5 X 10~2 4.7 X Ifl"1 0.90
1.5 X 10~2 9.6 X 10~3 0.1)9
Total
Dally
l-oad (kg/day)
24
24
24
24
Co.iiHl I- 1.1 ill
Rivur
9.3 X 10-3 9.3 x 10-3 3.0 x 10~3 1.6 X 10"2
0.14
4.6 X 10~2 8.0.
24 '
Ml .l.i u ^ I inn 1.1 luil by EXAMS (U.S. EPA-KKHL. ALliunb, tta.) model (sec text for further information).
-------
10'1
ID'2
1C'3
f 10-4
I
a e
•5 10'5
,-6
S 10'
te
I
I 10'8
c
u 10'9
Iff10
Eutro'phic Lake Ecosystem
River Ecosystem
_L
I
_L
I
34567
(Mrs following last discharge)
8 9
10 11
FIGURE 5 CONCENTRATION OF TRICHLOROMETHANE IN WATER COLUMN
FOLLOWING CESSATION OF DISCHARGE AT 1.0 KG HR'1
12
53
-------
magnitude, in the first 5 hours. Although the rate of removal by physi-
cal transport is almost entirely system dependent (e.g. dependent on
flow rate, etc.), one can assume that through many systems most of the
chloroform would be removed from the vicinity of discharge before
significant volatilization occurs.
F. SUMMARY
Trihalomethanes, especially chloroform, are commonly found in the
natural waters of the United States. Levels in the open ocean appear
to be in the range of 1-10 ug/1. The other trihalomethanes are not
commonly detected, and when found are at lower concentrations than
chloroform.
Chloroform is also found in ambient air, both indoors and outdoors.
Continental background levels range from 0.04 ug/m3 to 0.13 ug/m3, while
marine background levels are somewhat higher. Levels in urban areas are
highly variable, ranging from <0.05 ug/m3 to 90 ug/m3, with the maximum
occurring in a highly industrialized area.
The fate of trihalomethanes is largely controlled by their volatil-
ity. While much of the chloroform is originally released to the air,
that reaching water is rapidly volatilized. The half-life of volatili-
zation from a stirred aqueous solution is 20 minutes. The importance of
hydrolysis, adsorption and biodegradation appear low compared with vola-
tilization.
Once reaching the atmosphere, chloroform and the other trihalometh-
anes as well, probably travel considerable distances before degradation
occurs. The lifetime in the troposphere has been estimated to be 2-3
months, attributable primarily to reaction with hydroxyl radicals.
Direct photochemical degradation of trihalomethane, as well as rainout,
are not expected to be important pathways.
Volatilization is also likely from soil surfaces. Biodegradation
and adsorption are not likely to be important fate processes. Thus,
trihalomethanes that are not volatilized are subject to rapid movement
downward.
A partitioning model suggests that at equilibrium, all of the
trihalomethanes would be primarily (>99%) in the air compartment.
EXAMS also showed that volatilization was the dominant fate process
in lakes and ponds. In rivers, physical transport over the short
stretch of river modelled dominated volatilization.
-------
REFERENCES
Anderson, D.R. ^et_al. Quarterly Progress Report Covering Period January 1
through March 31, 1979: Biocide by-products in aquatic environments.
PNL-2988. U.S. Nuclear Regulatory Commission; 1979.
Bunn, W.W.; Hass, B.B.; Deane, E.R.; Kloepfer, A.D. Formation of tri-
halomethanes by chlorination of surface water. Environ. Lett. 10(3):205-
213; 1975.
Coniglio, W.A.; Miller, K.; MacKeever, D. The occurrence of volatile
organics in drinking water. Briefing. Washington,D.C.: Criteria and
Standards-Division, U.S. Environmental Protection Agency; March 6, 1980.
Cox, R.A.; Derwent, R.G.; Eggleton, A.E.J.; Lovelock, J.E. Photochemical
oxidation of halocarbons in the troposphere. Atmospheric Environ. 10:305-
308; 1976. (As cited in Versar 1979b).
Dilling, W.L.; Tefertiller, N.B.; Kallos, 6.J. Evaporation rates of
methylene chloride, chloroform, 1,1,1-trichloroethane, trichloroethylene,
tetrachloroethylene, and other chlorinated compounds in dilute aqueous
solutions. Environ. Sci. Techno1. 9(9):833-838; 1975.
s~
Hansch, C.; Quinlan, J.E.; Lawrence, G.L. The linear free energy rela-
tionship between partition coefficients and the aqueous solubility of
organic liquids. J. Org. Chem. 33:347-350; 1968.
Hansch, C.; Leo, A. Substituent constants for correlation analysis in
chemistry and biology. New York, NY: John Wiley and Sons; 1979.
Hanst, P.L. Noxious trace gases in the air, Part II: halogenated
pollutants. Chemistry 51(2):6-12; 1978. (As cited in Versar 1979b).
Harsch, D.; Rasmussen, R.A. Identification of methyl bromide in urban
air. Unpubl. Contract No. WA 6-99-2922-J. Washington, DC: Office of
Research and Development, U.S. Environmental Protection Agency; 1977.
Harsch, D.E.; Rasmussen, R.A.; Pierotti, D. Identification of a poten-
tial source of chloroform in urban air. Chemosphere 11:769-775; 1977.
Hendrickson, J.B.; Cram, D.J.; Hammond, G.S. Organic chemistry. 3rd ed.
New York, SY: McGraw Hill; 1970.
Hodgman, C.O. ed in chief. Handbook of chemistry and physics. 43rd ed.
Cleveland, OH: Chemical Rubber Publishing Company; 1962.
JRB Associates. Level I materials balance. Chloroform. Draft report.
Contract Mo. 68-01-5793, Task Order 13. Washingcon, DC: Office of
Pesticides and Toxic Substances, U.S. Environmental Protection Agency;
1980.
55
-------
Karickhoff, S.W.; Brown, D.S.; Scott, T.A. Sorption of hydrophobia
pollutants on natural sediments. Water Res. 13:241-248; 1979.
Levins, P.; Adams, J.; Brenner, P.; Coons, S.; Thrun, K.; Harris, G.;
Wechsler, A. Sources of toxic pollutants found in influents to sewage
treatment plants. VI. Integrated interpretation Part I. Contract
No. 68-01-3857. Washington, DC: U.S. Environmental Protection Agency;
1979a.
Levins, P.; Adams, J.; Brenner, P.; Coons, S.; Thrun, K.; Varone, J.
Sources of toxic pollutants in influents to sewage treatment plants. III.
Coldwater Creek Drainage Basin, St. Louis, Missouri. Contract No. 68-01-
3857. Washington, DC: U.S. Environmental Protection Agency; 1979b.
•>
Lillian, D.; Singh, H.B.; Appelby, A.; Lobban, L.; Arnts, R.; Gumpert, R.;
Hagne, R.; Toomey, J.; Kazazis,'J.; Antell, M.; Hansen, D.; Scott, B.
Atmospheric fates of halogenated compounds. Environ. Sci. Technol. 9(12):
1042-1048; 1975.
Lyman, W.S. Literature review — problem definition studies on 13
selected chemicals. Chloroform. Contract DAMD17-77-C-7037. Washington,
DC: U.S. Army Medical Research and Development Command; 1978.
Mabey, W.; Mill, T. Critical review of hydrolysis of organic compounds
in water under environmental conditions. J. Phys. Chem. Ref. Data 7:383-
415; 1978.
Means, J.C.; Hassett, J.J.; Wood, S.B.; Banwart, W.L. Sorption properties
of energy-related pollutants and sediments. Jones, P.M.; Leber, P. eds.
Polynuclear aromatic hydrocarbons. Ann Arbor, MI: Ann Arbor Science;
1978: p. 327-340.
Murray, A.J.; Riley, J.P. Occurrence of some chlorinated aliphatic
hydrocarbons in the environment. Nature 242:37-38; 1973. (As cited in
NAS 1978).
National Academy of Sciences (NAS). Chloroform, carbon tetrachloride,
and other halomethanes. An environmental assessment. Washington, DC:
NAS; 1978. 294 p.
Neely, W.B. An integrated approach to assessing the potential impact of
organic chemicals in the environment. Proceedings of the workshop on
philosophy and implementation of hazard assessment procedures for chemical
sub. in the aqu. environment, Waterville Valley, NH, August 14-18, 1978.
Pearson, C.R.; McConnell, G. Chlorinated GI and C2 hydrocarbons in the
marine environment. Proc. Roy. Soc. London B 189:305-322; 1975.
-------
Pellizzari, E.D.; Erickson, M.C.; Zweidinger, R.A. Formulation of a
preliminary assessment of halogenated organic compounds in man and
environmental media. Washington, DC: U.S. Environmental Protection .
Agency; 1979. Available from: NTIS, Springfield, VA; PB 80 112170.
Rook, J.J. Formation of haloforms during chlorination of natural waters.
Water Treatment Exam. 23(2):234-243; 1974.
Rook, J.J. Chlorination reactions of fulvic acids in natural waters.
Environ. Sci. Technol. 11(5):478-482; 1977.
Spence, J.W.; Hanst, P.L; Gay, B.W. Jr. Atmospheric oxidation of methyl
chloride, methylene chloride, and chloroform. Air Poll. Control Assoc.
26(10):994-996; 1976. (As cited in Versar 1979b).
Stanford Research Institute (SRI). Estimates of physical-chemical
properties of organic priority pollutants. Preliminary draft.
Washington, DC: Monitoring and Data Support Division, U.S. Environ-
mental Protection Agency; 1980.
Stecher, P.G. jst al. eds. The Merck index. 8th ed. Rahway, NJ: Merck
and Co., Inc.; 1968.
Su, C.; Goldberg, E.D. Environmental concentrations and fluxes of some
halocarbons. Windom, H.L,.; Duce, R.A. eds. Marine pollutant transfer.
Lexington, MA: D.C. Heath and Co. (As cited in MAS 1978).
U.S. Environmental Protection Agency (U.S. EPA). Report on the problem
of halogenated air pollutants and stratospheric ozone. Report EPA 600/
9-75-008. Research Triangle Park, NC: Office of Research and Develop-
ment, U.S. EPA; 1975.
U.S. Environmental Protection Agency (U.S. EPA). In-depth studies on
health and environmental impacts of selected water pollutants. Contract
No. 68-01-4646. Washington, DC: Office of Water Planning and Standards,
U.S. EPA; 1978.
U.S. Environmental Protection Agency (U.S. EPA). STORET. Washington,
DC: Monitoring and Data Support Division, U.S. EPA; 1979.
Versar, Inc. Non-aquatic fate. Trihalomethane. Draft report for
Monitoring and Data Support Division, U.S. Environmental Protection
Agency, Washington, D.C.; 1979a.
Versar, Inc. Trihalomethane: statement of probable fate. Callahan,
M.A.; Slimak, M.W.; Gabel, N.W.; May, I.P.; Fowler, C.F. ; Freed, J.R.;
Jennings, P.; Durfee, R.L.; Whitmore, F.C.; Maestri, B.; Mabey, W.R.;
Holt, 3.R.; Gould, C. Water-related environmental fate of 129 priority
pollutants. I. Report No. EPA-440/4-79-029a. Washington, DC: Moni-
toring and Data Support Division, U.S. Environmental Protection Agency;
1979b.
-------
Wilson, J.T.; Enfield, C.G.; Dunlap, W.J.; Cosby, R.L.; Foster, D.A.;
Baskin, L.B. Transport and fate of selected organic pollutants in a
sandy soil. Ada, OK: Robert S. Kerr Environmental Research Laboratory,
U.S. Environmental Protection Agency; 1980.
Yung, Y.L.; McElroy, M.B.; Wofsy, S.C. Atmospheric halocarbons: a
discussion with emphasis on chloroform. Geophysical Research Letters
2(9):397-399; 1975. (As cited in Versar 1979b).
53
-------
• CHAPTER V
EFFECTS AND EXPOSURE — HUMANS
A. HUMAN TOXICITY
1. Chloroform .
a. Introduction
Prior to the mid-1940s, chloroform was widely used as an anesthetic
and, until recently, was utilized in many pharmaceutical preparations
(e.g., cough medicine, mouthwashes, dentifrices, linaments). The possi-
bility of chloroform carcinogenicity in rodents has raised concern about
long-term, low level exposure in man, particularly in view of recent
findings of chloroform in concentrations up to 300 ug/1 in the drinking
water of municipalities throughout the United States. This section of*
the report examines the extensive data base on the toxic effects of
chloroform in man and laboratory animals.
b. Metabolism and Bioaccumulation
Chloroform (CHC13) is rapidly absorbed through the lungs, through
the gastrointestinal tract if ingested, and, to a lesser extent, through
intact skin (Brown et .al. 1974, Van Dyke .et al. 1964). With some
species variation, chloroform is partially excreted unchanged and par-
tially metabolized to carbon dioxide (C02) and unidentified urinary
metabolites in mice, rats, monkeys, and humans (NAS 1977). Pohl et al.
(1977) recently reported that chloroform was metabolically activated to
phosgene (COC12) by liver microsomes of phenobarbital-pretreated rats.
Mice exhibit the highest conversion rate to CC^, metabolizing
approximately 80% of a 60-tng/kg oral dose of ^Q^ (Brown e_t al. 1974).
No differences in the conversion rate due to strain (Brown e_t al. 1974)
or sex (Taylor &t_ al. 1974) were noted. Tissue distribution of the
radio-label, however, showed striking sex differences: males had higher
levels of radioactivity (^ threefold) bound to the renal cortex and
medulla, while females accumulated the label in the liver to a greater
extent than males (Taylor _et al. 1974).
In rats administered 60 mg/kg 1^CHC13 by gavage, 66% of the dose
was metabolized to l^cc^, with an additional 20% of the dose eliminated
unchanged in expired air (Brown et al. 1974). Administration of a much
larger dose (1,484 mg/kg) incraduodenally resulted in a much smaller
conversion to C02 (4%); the major portion of the dose (70%) was elimi-
nated unchanged in expired air (Van Dyke ec al. 1964).
59
-------
Only 17% of an orally administered dose of 14CHCl3 (60 mg/kg) was
metabolized to ^ C02 in the squirrel monkey; 78% of the dose was excreted
unchanged in expired air (Brown _e£ al. 1974).
In eight healthy human volunteers, between 17.8% and 66.6% of
unchanged chloroform was recovered in expired air within an 8-hour-
period following oral administration of a gelatin capsule containing
500 mg of chloroform (^ 7 mg/kg) in olive oil. The wide range of values
is partially attributable to body weight variations (58-80 kg); obese
subjects excreted a smaller proportion of the dose through the lungs as
unchanged chloroform. The maximum pulmonary excretion occurred between
40 and 120 minutes post-dosing. Less than 1% of the dose was detected
in urine (Fry ej: al. 1972). In a separate experiment, two subjects, one
male and one female, swallowed a capsule containing 500 mg of l^c-labelle
chloroform. A total of 50.6% and 48.5% of the dose was accounted for as
exhaled 13C02 ^or the male an<* female subject, respectively (Pry et al.
1972).
Poobalsingham and Payne (1978) studied the rate of alveolar uptake
of chloroform in 16 patients undergoing elective surgery. Eight patients
spontaneously breathed known concentrations (2-2.5%) of chloroform; the
second group were artificially ventilated with a fixed inspired concen-
tration of 1% chloroform. Initial uptake of chloroform was rapid in
both groups, approaching a plateau after 40-45 minutes. Expressed as
a percentage .of the equilibrium value, the arterial concentration of
chloroform increased more steeply with controlled ventilation than with
spontaneous breathing. In spontaneously breathing patients, the arterial
concentration reached approximately 25% of the inspired concentration
after about 1 hour of exposure compared with an equilibration of approxi-
mately 41% in the same time period under controlled ventilation.
Increasing concentrations of chloroform in the blood of spontaneously
breathing patients causes respiratory depression sufficient to reduce
alveolar ventilation. Elimination following chloroform withdrawal was
rapid and exponential in nature.
An average chloroform concentration of 51 ug/kg (range: 5-68
ug/kg wet tissue) was detected in postmortem samples of body fat taken
from eight subjects between 48 and 82 years of age. Concentrations of
1.0 ug/kg to 10 ug/kg were present in kidney, liver, and brain tissues
(McConnell .et al. 1975).
In summation, chloroform is rapidly absorbed from the lungs and
gastrointestinal tract, and to a lesser extent, through intact skin.
With some species differences,'chloroform is excreted partially unchanged
and partially metabolized to CO2 and unidentified urinary metabolites.
Body organs contain 1-10 ug/kg, with 5-68 ug/kg chloroform detected in
bodv fat.
-------
c. Animal Studies
i. Carcinogenesis
Chloroform was implicated as a possible carcinogen more than 30
years ago by Eschenbrenner and Miller (1945). They administered 30 oral
doses of 0, 150, 300, 600, 1200 or 2400 ing/kg chloroform to groups of
five male and female strain A mice at 4-day intervals. Males in the
three highest treatment groups and females receiving 2400 mg/kg chloro-
form died within 48 hours. When the mice were 8 months of age, non-
metastasizing hepatomas and cirrhosis were found in all surviving
females given 600 mg/kg or 1200 mg/kg. No hepatomas were seen in
either male or female mice at the lower dosage levels.
An abstract of another study of limited'duration reports that
Rudall (1967) gave 24 mice (strain unspecified) 0.1 ml or a 40% solution
of chloroform orally twice weekly for 6 months. Three hepatomas were
found in the five survivors. No additional information was given.
USF-grade chloroform in corn oil was administered by gavage to
groups of 50 male and female mice (B6C3F1 strain) at a predetermined
maximally tolerated dose and one-half this amount 5 days per week for
78 weeks. The dose of chloroform was changed during the course of the
experiment; time-weighted-average doses were 138 mg/kg and 277 mg/kg for
males and 238 mg/kg and 477 mg/kg for females. The vehicle control
group consisted of 20 mice of each sex.
Frank hepatocellular carcinoma was found in 36% to 98% of treated
mice (p<0.001) compared with 0% to 6% incidence in controls (see
Table 16). Nodular hyperplasia of the liver was observed in several
low-dose males that had not developed hepatocellular carcinoma
(NCI 1976, Powers and Voelker 1976, Renne et al. 1976).
A concurrent study was conducted with groups of 50, 52-day-old
Osborne-Mendel rats given time-weighted oral doses of 90 mg/kg and
180 mg/kg chloroform for males and 100 mg/kg and 200 mg/kg for females
for 78 weeks. A significant (p=0.0016) increase in the incidence of
kidney epithelial tumors was seen in males: 8% at 90 mg/kg; 24% at
180 mg/kg compared with none in controls. An increase in thyroid tumors
was observed in female rats, but was not considered as resulting from
the chloroform exposure (NCI 1976, Renne e± a±. 1976).
Another recent series of studies (Roe jat jil. 1979, Palmer et al.
1979, Heywood jt_al. 1979) examined the carcinogenic effects of feeding
chloroform to Sprague-Dawley rats, beagles, and four strains of mice at
dosage levels ranging from 15 mg/kg to 60 mg/kg. Except for an excess of
renal tumors in male ICI/CFLP mice at the 60-mg/kg treatment level, ao
carcinogenic effects './era reported.
In the first study, four strains of mice (male and female ICI/CFL?,
aale C5736, CF/1 and C3A) vera administered chlorofonn in a coothpasca
OJ.
-------
TABLE 16. INCIDENCE OF HEPATOCELLULAR CARCINOMA
IN B6C3F1 MICE EXPOSED TO CHLOROFORM
Experimental
Group
No. Mice with Carcinoma
(% incidence)
Male
Female
High CHC1
Low CHC13
Control
- Dose
Dose
44/45
18/50
1/18
(98)
(36)
(6)
39/41
36/45
0/20
(95)
(80)
(0)
Source: NCI (1976), Powers and Voelker (1976), Renne et a_l. (1976)
TABLE 17. DISTRIBUTION OF NODULES AND FATTY CYSTS IN THE
LIVERS OF CHLOROFORM-TREATED' BEAGLE DOGS
Treatment
Croup
30 mg CHC13 /kg/day
15 mg CHOLj /kg/day
vehicle control
untreated control
No. Dogs with Nodules
(% incidence)
Male
^^^•^HM^
0/7 (0)
1/7 (14)
0/15 (0)
1/7 (14)
Female
4/8 (50)
1/8 (13)
3/12 (25)
1/5 (20)
No. Dogs with Fatty Cysts
(% incidence)
Male
7/7 (100)
6/7 (86)
8/15 (53)
2/7 (29)
Female
7/8 (88)
5/8 (63
3/12 (25)
1/5 (20)
Source: Heywood £t al. (1979)
62
-------
vehicle by gavage 6 days per week for 80 weeks, followed by 16-24 weeks
of observation. No overall increase in neoplastic changes was observed,
except for an increase in tumors of the renal cortex in male ICI/CFLF
mice (9/52) at the 60-mg/kg level compared with vehicle (6/260) and
untreated (1/52) controls. No overall increase in tumors was observed
in males of this strain given 17 mg/kg/day of chloroform nor in female
ICI/CFLP or male C57BL, CBA or CF/1 strain mice given up to 60 mg/kg/day
of chloroform for 1.5 years (Roe e£ al. 1979).
In a separate experiment, male ICI/CFLP mice were given 60 mg/kg
chloroform in arachis oil, as well as in the toothpaste vehicle. There
were 12/52 renal tumors in the CHCl3/arachis-oil-treated group compared
with 1/52 in arachis oil controls; mice given 60 mg/kg chloroform in the
'toothpaste vehicle had 5/52 renal tumors (Roe e£ al. 1979).
In the second study, caesarian-derived SPF Sprague-Dawley rats
(50 males, 50 females) were intubated 6 days per week with 60 mg/kg
chloroform in a toothpaste vehicle for 80 weeks, followed by 15 weeks
of observation. The only noteworthy changes were a depression of plasma
(but not erythrocyte) cholinesterase in females, which reversed itself
upon cessation of treatment, and a decrease in absolute and relative
liver weights of females killed at termination of the study. Females,
but not males, possess plasma butyrlcholinesterase activity, which was
inhibited by chloroform treatment. No clear histologic evidence of
toxic effects on liver or kidney were seen in either sex. A wide spec-
trum of-different types of neoplasms was seen but did not appear to be
related to treatment (Palmer £t al. 1979).
In the third investigation of this series, Heywood and co-workers
(1979) gave beagles 0, 15 or 30 mg/kg chloroform in a toothpaste vehicle
(in a gelatin capsule) orally 6 days per week for 7.2 years; the dogs
were maintained without treatment for an additional 20-24 weeks. There
were 8 dogs of each sex in untreated controls and each treatment level;
16 dogs of each sex were in the vehicle control group. The only signifi-
cant toxic response was a moderate rise in serum glutamic pyruvic trans-
aminase (SGPT) levels, which peaked during the sixth year of the study.
Most of the dogs reverted to normal SGPT values during the post-treatment
period. No palpable tumors were noted, but a few subcutaneous nodules
thought to be of mammary gland origin were detected. Possible chloro-
form-related aggregations of vacuolated histiocytes ("fatty cysts") were
seen in the livers of several dogs at postmortem; the distribution of
nodules of altered hepatocytes and fatty cysts in the liver is presented
in Table 17. The distribution of fatty cysts appears to be dose-related.
A small number of macroscopically and microscopically visible neoplasms
was seen but were predominantly age-related. No neoplasia of liver or
kidney was found.
Capel and Williams (1978) investigated the effect of chloroform on
the growth of some murine tumors (Lewis lung carcinoma, B16 melanoma,
and Ehrlich ascites). Groups of male TO strain mice were given 0, 0.15
or 15 mg/kg of chloroform in their drinking water for varying periods of
time either prior to and post-inoculation or only post-inoculation with
1 x 10^ cunor cells. Exposure co 0.15 tng/kg chloroform did noc enhance
the growth or aetastasis of Lewis Lung carcinoma or significantly
-------
increase the number of Ehrlich ascices cumor cells. Ac Che 15-mg/kg
chloroform treatment level, however, a significanc increase in foci of
Lewis lung and Ehrlich Cumor cells was noted. Similarly, both chloroform
exposure levels resulted in invasion of a greater percentage of organs
by B16 melanoma Chan Chac seen in controls.
Roe .ec al. (1968) gave an unspecified number of newborn (C57xDBA
mice either a single subcutaneous 200-ug dose of chloroform before they
were 24 hours old or eight daily 200-ug injections during the first week
of life. No evidence of carcinogenicity was observed after 77 to 80 weeks.
Thus, ingestion of 90-138 mg/kg of chloroform has been shown to
induce both hepatic and renal tumors in experimental animals, although
other experiments have yielded contradictory results. The negative
results may reflect different vehicles employed to administer chloroform,
lack of sensitivity-of various test species, populations too small to
measure statistically significant and/or the lower dosage levels admini-
stered in the Roe eit al. (1979) and Palmer _e_t al. (1979) studies compared
with the NCI bioassay (1976).
ii. Mutagenesis
No indications of mutagenic activity have been reported in the
scientific literature for chloroform. No significant mutations were
found at .the 8-azaguanine locus in Chinese hamster lung fibroblast cells
following 24-hour exposure to 3% chloroform in culture (Sturrock 1977).
•
Negative results have also been reported for chloroform vapors in
microbiol assays conducted with Salmonella typhimurium strains TA100,
TA1535 (base-pair mutants) and TA1538 (frameshift mutant), both in the
presence and absence of liver microsomal activation (Simmon and Tardiff
1978, Uehleke et al. 1977).
Chloroform also did not induce DNA damage (single-strand breaks)
in liver DNA of rats given 200-400 mg/kg of chloroform orally. DNA
damage was measured by alkaline elution assay (Petzold and Swenberg
1978).
iii. Teratogenesis
Schwetz_et_al. (1974) exposed pregnant Sprague-Dawley rats to 30,
100 or 300 mg/kg chloroform in the diet, 7 hours a day on days 6 through
15 of gestation. At 30 mg/kg (diet), an increase in wavy ribs and
delayed skull ossification were observed. Ac 100 mg/kg, a significant
incidence of fetal abnormalities such as acuadia, imperforate anus,
subcutaneous edema, missing ribs, and delayed ossification of both skull
and sternebrae were noted. At the 300 mg/kg exposure level, maternal
weight gain and food consumption were markedly reduced. The conception
race in Chis group was only 15% (3/20) compared with 88% (63/77) in the
control group. The aon-pregnant bred females in Chis treatment group
exhibited vascular evidence of implantation in the aesometrium, buc no
intrauterine evidence of implantation. The authors suggest that chloro-
form aay interfere vich Che process of iaplancacion. Litters from Che
chrea dams exposed co 300 tag/kg chloroform had a reduced number of live
-------
healing (Torkelson ec al. 1976). A single occluded 24-hour application of
fetuses and decreased fetal body measurements. The number of fetal
resorptions was also increased. These effects do not appear to be
related to the anorexia seen in the dams at this exposure level, in that
the same degree of starvation (3.7 g food per rat per day on days 6
through 15 of gestation) without exposure to chloroform was neither
embryo- nor feto-toxic in a special control group maintained on starva-
tion rations.
In another study, rats administered oral dosages up to 126 mg/kg
chloroform on days 6-15 of gestation and rabbits given up to 50 mg/kg
orally on days 6-18 of gestation showed no evidence of teratogenicity.
Offspring of both species had reduced birth weight at these maternally
toxic levels, but not at 50 mg/kg and 35 mg/kg treatment levels for rats and
rabbits, respectively, when compared with the short daily oral exposure.
iv. Other lexicological Effects
Available data indicate species-, strain-, and sex-dependent variations
in the acute toxicity of chloroform in laboratory animals (see Table 18).
Oral LD50 (50% lethal dose) values range from 119 mg/kg to 1400 mg/kg in
the mouse to 2000 mg/kg in the rat.
In mice, chloroform toxicity varies according to strain and sex.
The oral 1.050 for chloroform was four times higher in C57BL/6J males
than in DBA/2J males with first generation offspring of these two strains
(B6D2F1/J), exhibiting a value midway between the parental strains. In
addition, the histopathological response to acute exposure to chloroform
varied with sex. At doses > 252 mg/kg, males of all three genotypes
exhibited necrosis of the proximal convoluted renal tubules and hepatic
centrilobular necrosis; only renal tubular necrosis was seen at doses
less than 252 mg/kg of chloroform. Females had a similar threshold to
hepatic damage without developing renal lesions. Immature or castrated
adult males were also resistant to chloroform-induced renal toxicity, but
both castrated males and females treated with testosterone became sensi-
tive to chloroform-induced renal toxicity (Hill e£ al. 1975, Vessell
et al. 1976, Hill 1978).
Acute oral exposure to 250 mg/kg chloroform produced fatty infiltra-
tion and necrosis of the liver and kidney damage in rats (Torkelson et al.
1976). Jones and co-workers (1958) noted similar findings in mice given
350 mg/kg chloroform orally. The hepatotoxic effects of chloroform have
been attributed to the interaction of a reactive metabolite of chloroform
with tissue proteins and appears to be related to two factors: (1) the
rate of biotransformation and (2) the availability of endogenous hepatic
antioxidant, reduced glutathione (GSH). In rodents, hepatic necrosis can
be enhanced by pretreatment with hepatic microsomal enxyme inducers and
conversely reduced by inhibiting biotransformation (Cresteil_et_al. 1979,
Ilett _et _al. 1973). Supplying antioxidants such as GSH also reduces the
severity of liver necrosis associated with chloroform treatment; the
actual mechanism by which GSH reduces tissue damage is not known (Stevens
and Anders 1978).
One or cwo 2&-hour applications of chloroform to shaved rabbit
abdomen produced hypersraia, necrosis, scab formation, and delayed
-------
TABLE 18. ACUTE TOXICITY OF CHLOROFORM TN LABORATORY ANIMALS
Species
Sex
Route
LPsn (me/kg)
Reference
a\
Mmisi! TCI Swiss F
fCl Swiss M
C5UL/6J M
H6D2F1/J M
DUA/2J M
Swiss Webster M
Itat M
300 to 470 g M
80 to 160 g M
16 to 50 g M-i-F
Dog M+F
Kdhhir
oral
oral
oral
oral
oral
intraperltoneal
subcutaneous
inhalation
oral
oral
oral
oral
oral
intraperitoneal
Inhalation
inhalation
1400 (1120-1680)
1120 ( 789-1590)
489 ( 386- 594)
297 ( 237- 356)
119 ( 103- 148)
1,781 (1,484-1,929)
704
3
134 mg/m
2,000
1,182 (1,038-1,336)
1,336 (1,182-1,632)
445 ( 297- 742)
800
1,483
490 mg/m
289 mg/m3
Bowman et al. (1978)
Bowman et al. (1978)
Hill e£ al. (1975)
Hill £t a\_. (1975)
Hill £t al. (1975)
Klassen and Plaa (1966)
RTECS (1977)
Torkelson et^ al. (1976)
Torkelson et^ _al. (1976)
Torkelson e± al. (1976)
Kimura ^it al. (1971)
Klmura e± al. (1971)
RTliCS (1977)
Brownlee eit _al. (1953)
RTECS (1977)
RTECS (1977)
-------
1,000 mg/kg to 3,980 aig/kg of chloroform to rabbit abdomen produced
extensive necrosis and body weight loss. After 2 weeks, necropsy revealed
degeneration of the kidney tubules; the livers were not grossly affected
(Torkelson £t al. 1976).
Instillation into the eyes of rabbits resulted in slight conjuncti-
val irritation and corneal injury, which was not reduced by washing the
eye 30 seconds later (Torkelson ££ at% 1975).
Torkelson and co-workers (1976) exposed rats to 123, 245 or 417
mg/m3 chloroform by inhalation 7 hours per day for 138 and 144 days
during a 195- to 203-day period. All males displayed an increased
relative kidney weight, cloudy swelling of the renal tubular epithelium,
and centrilobular granular degeneration with focal areas of necrosis
throughout the liver. Decreased body weight was observed in males at the
50 mg/kg and 85 mg/kg levels, with pneumonia and an increase in the rela-
tive liver weight present in the 85 mg/kg male rats. Only an increase
in relative kidney weight was noted in females at the lowest treatment
level; at the two higher exposure levels, liver and kidney pathology was
similar to that seen in male rats. Male rats allowed to recover for 6
weeks after their last exposure to 25 mg/kg chloroform, exhibited no
pathological findings. Even at the 85 mg/kg treatment^level, terminal
blood counts and urine analyses were unaffected and serum levels of urea
nitrogen, glutamic-pyruvic transaminase, and alkaline phosphatase were
within normal limits, despite liver damage, observed histologically.
Dogs exposed to 123 mg/nr* chloroform, and rabbits and guinea pigs
exposed by inhalation to 123 mg/m^ or 417 mg/m^ under the same condi-
tions showed inconsistent liver, kidney, and lung changes (Torkelson
.et al. 1976).
In another study, Miklashevskii (1966) administered 0.4 mg/kg or
35 mg/kg chloroform orally to male guinea pigs for 5 months (daily
administration implied, but not stated). No effects were noted at the
0.4-mg/kg dose. At 35 mg/kg, only two of the six guinea pigs survived
beyond 3 months. The blood albumin-globulin ratio and blood catalase
activity decreased by the end of the first and second months, respec-
tively. Animals that died exhibited fatty infiltration, necrosis, and
cirrhosis of liver parenchyma, lipoid degeneration, proliferation of
interstitial cells in the myocardium, and acute edema of the submucosal
and muscular layers of the stomach.
No mortality was seen in dogs given 30 mg/kg to 120 mg/kg chloro-
form orally for 12-18 weeks. Vomiting occurred at times in dogs given
30, 60 or 120 mg/kg of chloroform and one of two dogs given 120 nig/kg/day
became jaundiced after 4-5 weeks. Marked body weight losses and
increased levels of serum glutamic pyruvic transaminase were observed
in dogs given 60, 90 or 120 mg/kg/day. At necropsy, liver discoloration
and increased liver weight were noted in dogs receiving 45 aig/kg/day and
above, but not at the 30-sig/kg treatment level. At 50, 90 and 120 mg/kg,
hepatocyta enlargement and vacuolation were seen, together with deposition
of fat vithin the heoatocytes (Heywood ac al. 1979).
-------
Thus, the acute toxicity of chloroform is species-, strain-, and '
sex-dependent. Oral LD$Q values range from 119 tag/kg to 2000 nig/kg, .'.
with indications of renal and hepatic necrosis. Dermal contact with /' ,
chloroform for 24 hours resulted in tissue necrosis in rabbits and '\i
instillation into eyes resulted in ocular injury. Subchronic inhalation ''
exposure to chloroform produced liver and kidney pathology in rats, but
findings were inconsistent in dogs, rabbits, and guinea pigs.
d. Human Studies
Although there are many documented fatalities from chloroform-
induced anesthesia, as well as cases of accidental or intentional
ingestion, limited information is available on controlled human exposure
to chloroform. Ingestion of 120 ml of chloroform has been survived
(Schroeder 1965), while serious illness occurred in another individual
after ingestion of only 5 ml (Winslow and Gerstner 1978).
Signs of chloroform poisoning include a characteristic sweetish
odor on the breath, dilated pupils, cold and clammy skin, initial
excitation alternating with apathy, loss of sensation, abolition of
motor functions, prostration, unconsciousness, and eventual death.
Central lobular necrosis of the liver and renal damage are the most
outstanding pathological findings (Winslow and Gerstner 1978).
litmus and Moser (1975) reported patchy pulmonary infiltrates and
acute respiratory distress in a 21-year old man who intentionally
injected himself intravenously with a 5-tnl bolus of reagent-grade
chloroform. Evidence of acute hemolysis was also noted. No hepatic
or renal toxicity was seen, probably because of the route of administra-
tion and the probable rapid pulmonary clearance of the injected chloro-
form. The subject returned to normal pulmonary functional status
within 3 weeks.
Desalva ejt al. (1975) found no indications of hepatotoxicity in
chronic users of a dentifrice and mouthwash containing 3.4% and 0.43%
chloroform, respectively, over a one- to five-year period. Estimated
daily ingestion was 0.3-0.96 mg/kg per day. Some reversible hepatotox-
icity was seen, however, in a ten-year clinical study with patients who
ingested a cough suppressant containing 1.6 g to 2.6 g of chloroform daily
(^ 23-37 mg/kg per day) (Wallace 1959).
Dermal exposure to chloroform may cause irritation, erythema,
hyper emia, and destruction of the epithelium (Ma It en _et_ al. 1968). Eye
contact produces burning, redness of conjunctival tissue, and possible
damage to the corneal epithelium. Recovery generally occurs in one to
three days (Grant 1974).
The carcinogenic effects of chloroform in rodents, and the detec-
tion of chloroform-in concentrations up to 300 ug/1 in the drinking
63
-------
water of municipalities throughout the United States have raised concern
on the human health impact of long-term, low-level exposure to chloro-
form in public water supplies. Several epidemiologic studies have
suggested an association between cancer mortality rates and levels of
chloroform in drinking water (Page £t, al. 1976, De Rouen and Diem 1977,
Cantor jet al. 1978, Alavanga j£ a!L. 1978). Many of these studies,
however, are deficient in a number of areas such as: lack of individual
exposure data, variations in concentrations in water data, and use of
current rather than historical exposure data. Furthermore, cancer rates
can be confounded by population mobility patterns; regional trends in
diet, tobacco, and alcohol consumption; as well as possible contributions
of other sources such as occupation, local extent of industrialization,
air pollution, etc. (Kraybill 1978).
In a more direct attempt to assess relative cancer risk associated •
with exposure to chlorinated drinking water, Alavanga and co-workers
(1978) conducted the only case-paired study cited in the scientific
literature. These investigators compared individual death certificate
data for all females in seven New York State counties for the years
1968 to 1970 that died of gastrointestinal and urinary tract cancer
with a corresponding set of matched controls. In three of seven coun-
ties studied, the combined gastrointestinal and urinary tract cancers
were significantly higher in areas supplied by chlorinated water, but
there was no significant difference in the other four counties. Further-
more, the results were complicated by the overriding finding that both
cancer and chlorination correlated with urban living and this common
relationship would result in the apparent correlation between cancer
and chlorination.
Thus, although positive correlations between chloroform in drinking
water and increased cartcfcr incidence may be apparent, causal relation-
ships have not been proven on the basis of results from these indirect
studies.
2. Bromoform
a. Metabolism
Absorption of bromoform may occur by inhalation, from the gastro-
intestinal tract, and, to a certain extent, through the skin. Lucas
(1928) reported that following administration of bromoform to rabbits
either rectally or by inhalation, inorganic bromides could be detected
in tissues and urine. Anders and associates (1978) found substantial,
dose-dependent elevations in blood carbon monoxide levels in male rats
given 1 tnmol/kg bromoform intraperitoneally. Enzyme induction by
pretreatment with phenobarbital (but not 3-methylcholanthrene) markedly
increased the blood carbon monoxide levels seen after bromoforn admini-
stration compared to saline-treated controls. The administration of
SKF 525-A, a known inhibitor of drug metabolism, significantly inhibited
in vivo aetaboLisin of bromofora to carbon monoxide.
69
-------
b. Animal Studies
1. Carcinogenesis
The is s _e_t al. (1977) reported Chat bromoform produced a positive
response in a pulmonary adenoma bioassay. The administration to strain A
male mice of 48 mg/kg bromoform intraperitoneally three times per week
for 23 weeks resulted in 1.13± 0.36 lung tumors per mouse p • 0.041)
compared with 0.27 t 0.15 tumors for vehicle controls. A higher bromo-
form dosage (100 mg/kg), however, failed to produce a significant
pulmonary response (0.67 t 0.21 lung tumors; p ** 0.136) under the same
treatment conditions.
Bromoform is currently being tested for carcinogenicity by the
standard bioassay protocol of the National Cancer Institute.
ii. Mutagenesis
Bromoform vapor tests positive in the Ames system (Simmon and
Tardiff 1978). Exposure to the vapor phase of 30 ul of bromoform was
mutagenic in Salmonella typhimurium TA 100 with or without metabolic
activation. A weak response was also reported for strain TA 1535. exposed
to the vapor phase of 100-ul of bromoform.
iii. Teratogenesis
No data are available regarding teratogenic effects due to bromo-
form.
iv. Other Toxicologic Effects
Little is known of the toxicity of bromoform. Bowman et al. (1978)
recently reported oral LD50 values of 1400 mg/kg (1205-1595) and 1550
mg/kg (1165-2065) in male and female ICR Swiss mice. Males appeared to
be more sensitive than females to the acute lethal effects of bromoform.
Ataxia, sedation, and anesthesia occurred within 60 minutes of oral
administration of 1000 mg/kg of bromoform, but not at lower dosages.
Sedation persisted for approximately 4 hours. Subcutaneous LDso values
of 410 mg/kg and 1820 mg/kg bromoform have also been reported for the
rabbit and mouse, respectively (RTECS 1977, Kutob and Plaa 1962).
In dogs, deep narcosis occurred after 8-minute exposure to 142,100
mg/m-3 bromoform; mortality occurred after 60 minutes. Exposure to the
same concentration for 30 minutes produced deep narcosis, but the dogs
survived (Sax 1979).
Inhalation by rats of 0.25 mg bromoform per liter of air for 4 hours
per day for 2 months produced disorders in prothrorabin synthesis and
glycogenesis in the liver and reduced renal filtration capacity; the
threshold concentration was 0.05 mg/liter (Dykan 1962).
70
-------
c. Human Studies
Little information is available on the adverse health effects of
bromoform in man. Sax (1979) reports that inhalation of small amounts
of bromoform causes irritation provoking lacrimation, salivation and
facial hyperemia. Poisonings with bromoform have been reported (Fatty
1963). Symptoms include listlessness, headache, vertigo, unconscious-
ness, loss of reflexes, and occasionally convulsions. The primary cause
of death is respiratory failure.
3. Dibromochloromethane
a. Metabolism
Anders et al. (1978) foun'd a slight elevation of blood carbon mono-
xide levels in male rats given 1 mmol/kg of dibromochloromethane by
intraperitoneal injection. At 2 hours, blood levels of carbon monoxide
were approximately 400 mmol CO/ml compared with a value of 50 mmol CO/ml
for controls.
b. Animal Studies
i. Carcinogenesis
No data are currently available concerning the carcinogenicity of
dibromochloromethane. The compound is currently being tested (June
1979) for carcinogenicity by the National Cancer Institute according
to its standard bioassay protocol.
ii. Hutagenesis
Simmon and Tardiff (1978) reported a positive mutagenic effect in
Salmonella typhimurium TA 100 (with or without metabolic activation) '
exposed to vapor from 10 ul of dibromochloromethane in a dessicator.
iii. Teratogenesis
No data are available concerning the teratogenesis of dibromochloro-
methane.
iv. Other Toxicoloqic Effects
Toxicological data on dibromochloromethane are sparse. Bowman et al.
(1978) recently reported oral IS^Q values of 800 mg/kg (667-960) and
1200 mg/kg (945-1524) dibromochloromethane tor male and female ICR Swiss
mice, respectively. Males appeared to be more sensitive than females
to the acute lethal effects of dibromochloromethane. Sedation and
anesthesia occurred within 30 ainutes following oral administration of
500 mg/kg of this compound and persisted for approximately 4 hours. Ac
necropsy, livers appeared co have fancy iafiltration, che kidneys vere
pale, and hemorrhaging was noted in the brain, lungs, and adrenals.
-------
c. Human Studies
No human data are available.
4. Dichlorobromomethane
a. Metabolism
Male rats injected intraperitoneally with 1 mmol/kg dichlorobromo-
methane revealed only a small elevation in blood levels of carbon
monoxide. At 2 hours following injection, blood carbon monoxide levels
were approximately 50 mmol and 120 mmol CO/ml for control and dichloro-
bromomethane-treated rats, respectively (Anders jat al. 1978).
b. Animal Studies
i. Carcinogenesis
Theiss et al. (1977) reported that dichlorobromomethane produced an
elevated response in a pulmonary adenoma bioassay that approached statis-
tical significance. Strain A male mice were injected intraperitoneally
three times per week for 24 weeks with 0, 20, 40 or 100 mg/kg dichloro-
bromome thane. The incidence of lung tumors per mouse (* SE) were 0.27 *
0.15, 0.20 ± 0.11 (p =0.724), 0.25 ± 0.11 (p =0.930) and 0.85 ± 0.27
(p » 0.067), respectively.
•
Dichlorobromomethane is currently being tested (June 1979) by the
National Cancer Institute according to Its standard bioassay protocol.
ii. Mutagenesis
Dichlorobromomethane (50 ul) was mutagenic in Salmonella typhimurium
strain TA 100, both in the presence and absence of metabolic activation
(Simmon and Tardiff 1978).
iii. Teratogenesis
No data are available concerning teratogenic effects of dichloro-
bromome thane.
iv. Other Toxicologic Effects
As is the case for dibromochloromethane, the toxicologic data avail-
able for dichlorobromomethane are quite limited. Male ICR Swiss mice
appear to be more sensitive than females to the acute lethal effects of
dichlorobromomethane. Oral LD$Q values of 450 mg/kg (326-621) and 900
mg/kg (811-999) of dichlorobromomethane were recorded for males and
females of this strain, respectively. Sedation and anesthesia occurred
within 30 minutes after oral administration of 500 mg/kg of this compound
and persisted for approximately & hours. At necropsy, livers appeared
to have fatty infiltration, kidneys were pale, and hetnorrhaging was
observed in the adranals, Lungs, and brain (Bowman ££ a.1. L973).
72
-------
c. Human Studies
No data are available.
5. Overview
Trihalomethanes are present in the chlorinated drinking water
supply of municipalities throughout the United States, Several epidemi-
ological studies have suggested a correlation between ingestion of
chlorinated water and a higher incidence of cancer. The four trihalo-
methanes of concern are chloroform, bromoform, dibromochloromethane, and
dichlorobromomethane.
Extensive toxieologic data on chloroform indicate that it is rapidly
absorbed through the lungs, from the gastrointestinal tract and, to a
lesser extent, through intact skin. Species variations exist with regard
to the metabolic handling of chloroform, but some conversion to carbon
dioxide does occur.
Ingestion of 90-132 mg/kg of chloroform has been shown to induce
both hepatic and renal tumors in experimental animals, although other
experiments at lower dosage levels have yielded contradictory results.
No indications of mutagenic activity have been reported.
Administration by inhalation to rats during days 6-15 of gestation
resulted in a high incidence of fetal resorption and a few cases of
acudate fetuses. No evidence of teratogenicity was found in another
study, in which pregnant rats and rabbits were given up to 126 and 50
mg/kg, respectively, by oral intubation.
Species-, strain-, and sex-related differences exist with respect
to the acute lethal effects of chloroform, and both acute and prolonged
exposure to chloroform results in liver necrosis and kidney damage.
In man, ingestion of 120 ml of chloroform has been survived, but
serious illness has been reported following ingestion of only 5 ml.
Dermal contact produces burning pain within a few minutes of exposure
and, depending on dose, erythema, hyperemia, and vesication may also
occur.
Limited toxieologic information is available for bromoform,
dibromochloromethane, and dichlorobromomethane. However, because of
structural similarities to chloroform, all three compounds are cause
for concern with regard to carcinogenic effects. Preliminary results
in lung adenoma bioassays with bromoform and dichlorobromomethane
support this concern. In addition, all three of the lesser crihalo-
methanes are mutagenic in the Ames test. To date, other adverse health
effects have not been demonstrated, and therefore cannot be quanticaced.
/ j
-------
3. . HUMAN EXPOSURE
1. Introduction
This section quantifies the exposure of persons in the IKS. to
trihalomethanes. Several authors have attempted such an analysis
previously (NAS 1978, U.S. EPA 1979a). However this discussion will
include some new data, as well as briefly review previous estimations
of exposure. The exposure routes considered include ingestion via
drinking water and food, inhalation, and dermal contact.
2. Ingestion
a. Drinking Water
the formation -of trihalomethanes during chlorination of water
supplies has been discussed in Chapters III and IV. The concentrations
of these chemicals in drinking water have been measured in several
nationwide surveys. The National Organics Reconnaissance Survey (NORS)
(Symons et al. 1975) included sampling of trihalomethanes in raw and
finished water from 80 water supplies. The National Organic Monitoring
Survey (NOMS)(U.S. EPA 1978) also included analysis for these chemicals*in
addition to others, in samples taken from 113 community water supplies
during a .12-month period. The results from these surveys have been
analyzed and presented in numerous ways, and have been used in various
risk assessments (NAS 1978, U.S. EPA 1979a, Reitz et al. 1978). As a
result, these data are not discussed extensively here. The reader is
referred to the original sources and the discussions cited above.
Table 19 shows the concentrations of trihalomethanes found in the
NORS and NOMS surveys (Symons .et.al. 1975, U.S. EPA 1978). Figure 6
shows the distribution of total trihalomethanes in drinking water from
NOMS. Concentrations of total trihalomethanes are greater than 200 ug/1
(see Table 20) and chloroform levels exceed 100 ug/1 in many of the
locations sampled.
The formation of trihalomethanes in chlorinated water supplies has
been discussed in Chapter IV. It is obvious that water source, treat-
ment method, and analytical method influence the observed levels of
trihalomethanes. The samples collected for NORS were of finished water
at the plant, while the terminal samples reported in NOMS had been
allowed Co react with residual chlorine. Sylvia ejt al. (1978) showed
that samples collected in the distribution system had much higher levels
than those collected at the plant. U.S. EPA (1978) showed that the
terminal samples contained higher levels than those that had been
quenched (sodium thiosulfate added to react with residual chlorine).
However, measurements taken in the distribution system generally showed
levels between those of the quenched and terminal samples. Therefore,
use of these values either underestimates or overestimates actual expo-
sure Levels.
-------
TABLE 19. CONCENTRATIONS OF CHLOROFORM, BROMOFORM, BROMODICHLOROMETHANE,
AND DIBROMOCHLOROMETHANE AND TOTAL TRIHALOMETHANE5 IN WATER
SUPPLIES FROM MORS AND MOMS
Concentration (mg/1)
SORS
T cihaloBBChanes
Chlorofora
Median
Mean
Range
Bromofonn
Median
Mean
Range
0.021
HF^O.Sll
0.005
NF-0.092
MOMS
Phase I
0.027
0.043
XF-0. 271
LD2
0.003
NF-0.039
Phase II
(summer)
0.059
0.083
XF-0. 47
LD
0.006
MF-0.280
Phase III
(wincer)
Qaenehed Terminal
0.022
0.035
SF-0.20
LD
0.002
NF-0.137
0.066
0.069
NF-0.540
LD
0.004
.NF-0.190
0
Dib romoehlorome chane
Median
Mean
Range
0.001
SF-0.100
LD
0.003
SF-0.19
0.006
0.012
NT-0.290
0.002
0.006
IJF-0. 116
0.003
0.011
NF-0.250
Bromodiehloromethane
Median
Mean
Range
0.006
NF-0.116
0.010
0.013
XF-0. 183
0.016
0.018
XF-0. 180
0.006
0.009
SF-0.072
0.011
0.017
SF-0.125
local Trihalomechanes
Median
Mean
Range
0.027
0.067
NF-0.482
0.043
0.068
XF-0. 457
0.087
0.117
XF-0. 784
0.037
0.053
HF-0. 295
0.074
0.100
XF-0. 69 5
hfF • noc found.
-LD • less Chan dececcion limic.
Source: U.S. EPA (1978)
-------
25
20
I
<3
15
£
I
i 10
f
I
0
"5
0-10
10-50 50-100 100-200
Concentration (yg/l)
>200
Source: U.S. EPA (1978)
FIGURE 6 DISTRIBUTION OF CONCENTRATIONS OF TOTAL TRIHALOMETHANES
IN MAJOR U.S. DRINKING WATER SUPPLIES
76
-------
TABLE 20. SAMPLING LOCATIONS FOR NOMS WITH
CONCENTRATIONS OF TOTAL TRIHALOMETHANES
GREATER THAN 200 ug/1
Location
Annondale, VA
Brownsville, TX
Camden, AR
Cape Cirardeau, MO
Charleston, SC
Columbus, OH
Hagerstown, MD
Houston, TX
Huntington, WV
Huron, SD
Illwaco, WA
Jackson, MS
Louisville, KY
Melbourne, FL
Montgomery, AL
Newport, RI
Norfolk, VA
Oklahoma City, OK
Omaha, NE
Tampa, FL
Terre Bonne Parish, LA
Wheeling, WV
Concentration
230
670
250
290
230
250
200
460
230
260
260
270
240
780
210
310
250
200
210
340
230
310
Phase II; summer terminal.
Source: U.S. EPA (1978)
77
-------
Table 21 shows the exposures of persons co chloroform estimated
with various assumptions regarding daily intake of water and the concen-
tration of chloroform. The range of exposures was estimated to be 0.02-
1.2 mg chloroform per day. These estimates are based on the assumption
Chat all water used in water-based beverages comes from the same source
of tap water. Though this assumption may be valid for beverages such as
coffee and tea," chloroform levels in soft drinks, beer, wine, etc. would
correspond to the levels in drinking water in che area of manufacture
and not necessarily to the levels in the area of consumption.
Table 22 shows median and maximum exposures for all of the four
trihalomethanes. Though some high concentrations of bromofora,
dibromochloromethane, and bromodichloromethane are found in drinking
water, exposure levels for these other trihalomethanes are generally
much lower than exposure levels for chloroform.
b- Food
Little information is available on- the levels of chloroform in
food. McConnell e_t al. (1975) apparently have done the only work to
date. They analyzed various foods in Great Britain and found chloroform
levels ranging from 0 ug/kg to 33 ug/kg. The higher levels were found
in cheeses, olive oil, tea packets, and potatoes. The source of this
contaminant is unknown, although chloroform has been used as a pesti-
cide. It is possible that chloroform may be formed in the food from
other organics. HAS (1978) used these data to estimate human exposure
to chloroform through food. For the minimum intake of food containing
the minimum concentration of chloroform, exposure is estimated to be
0.0006 ing/day; for the maximum intake and maximum concentration, the
exposure would be 0.04 mg/day. The average is estimated to be 0.006
ing/day. These estimates are based on concentrations of chloroform in
uncooked food; since cooking is likely to vaporize some of the chloro-
form, the estimated exposure levels are probably overstated. No
information is available on the levels of the other trihalomethanes in'
food.
3. Air
Section IV-c described levels of chloroform in both urban and rural
air. In order to simplify these data for use in exposure estimates, the
following assumptions were made:
Concentration of
Exposure Setting Chloroform (ug/m^)
rural atmosphere 0.07
urban atmosphere 1
industrial acaosphere 50
indoor acnospnere 0.5
78
-------
TABLE 21. ESTIMATED HUMAN EXPOSURES TO CHLOROFORM VTA DRINKING WATER IN THE U.S.
—I
vO
_ Exposure Category
AiltilL
medJan concentration
(0.059 nig/1)
maxImi mi concentration
(0.540 rog/1)
Cliilil
median concentration
(0.059 mg/1)
maximum concentration
(0.540 mg/1)
Estimated Exposure to Chloroform (me/day)
1,2
MlnJmum Intake Maximum Intake Reference Male Reference Female
0.02
0.20
0.03
0.29
0.13
1.18
0.05
0.43
0.10
0.90
0.06
0.51
0.07
0.65
I
Includes tan water and water-based beverages.
2lnuikus - Adult - 365-2180 ml/day. Taken from ICRP .(1975)
Children - 540-790 ml/day
reference male - 1650 ml/day
reference female - 1200 ml/day
reference child - 950 ml/day (It Is unclear why this Is not within the range shown above.)
Source: Arthur D. Little, Inc.
-------
TABLE 22. ESTIMATED HUMAN EXPOSURES TO
TRIHALOMETHANES VIA DRINKING WATER
Daily exposure (mg/day)
Assuming Maximum Adult Assuming Reference M <=
Intake^ and Maximum Intake^ and Median
Trihalomethane Concentration (iag/1) Concentration in Water Concentration in Water
Median Maximum
Chloroform 0-059 0.540 1.2 0.1
Bromoform 0.004 0.280 0.6 0.007
•>
Dibromochloromethane 0.004 0.290 0.6 0.007
Bromodichloromethane 0.014 0.180 0.4 0.02
2.18 liter per day
2
1*65 liter per day
Source: Arthur D. Little, Inc.
30
-------
These levels were selected co represent typical conditions, although
there is considerable controversy over what may be considered typical
since the range in values is so great. The value for the indoor occupa-
tional atmosphere is intended to represent concentrations commonly found
in such locations as beauty parlors, drug stores, photoduplicating rooms,
etc. Presumably persons working in these situations would be exposed for
8-hour periods. Persons using these services would be exposed for shorter
periods, perhaps up to 4 hours per exposure, as in the case of beauty
parlors.
Table 23 gives the estimated exposures associated with the above
concentrations. These estimates were developed for a respiratory volume
for light activity of 19200 1 for 16 hours and 360"0 1 for 8 hours of resting
(ICRP 1975). Higher respiratory volumes, such as those for athletes or
other more active people, would result in correspondingly higher expo-
sure levels.
Monitoring data are insufficient to permit an estimate of the size of
the population exposed at these levels. The urban exposure level shown in
Table 23 may be quite common, and exposures at this level for 24 hours would
not be unusual. The estimate of 0.5 mg/day would be expected for persons who
work in the highly industrialized cities in the U.S. Unfortunately, no
estimates of exposure can be made for the other trihalomethanes, but the
exposure levels would be expected to be considerably lower.
4. Dermal Contact
Absorption of trihalomethanes through the skin would be expected to
occur in areas where the water supply is chlorinated. Blank (personal
communication, as cited in Beech 1980) has measured the permeability con-
stant of 125 x 10" 3 cm/hr for chloroform. For the median concentration
of 60 ug/1 in water, an exposure of 0.15 mg/hr is estimated for a total
body exposure. If total body immersion occurs for 2 hours/day, the total
maximum exposure would be 0.3 mg/day. A more typical immersion would be
0.2 hours/day including bathing and dishwashing (U.S. EPA 1979c), resulting
in an average exposure of 0.03 mg/day. Similar calculations can be made
for the other trihalomethanes. Using the same permeability constant and
median concentrations in water, exposures of 0.002-0.006 mg/day were
estimated for this route. Maximum exposures ranged from 0.08-0.14 mg/day.
Beech est al. (1980) have measured the concentration of trihalo-
methanes in swimming pools in Florida and found mean concentrations of
156 ug/1 in freshwater pools (mainly chloroform) and 657 ug/1 in saline
pools (mainly bromoform). The maximum concentrations were 430 ug/1 and
1278 ug/1, respectively. These authors suggested that trihalomethanes
are continuously formed in swimming pool water. Using the permeability
constant as above for chloroform and bromoform, several estimates of
exposure vere made for a six-year old child, as shown in Table 24.
31
-------
TABLE 23. ESTIMATED CHLOROFORM EXPOSURES VIA INHALATION
Chloroform
Exposure Situation Concentration (ug/tn3) Exposure (ing/day)
Rural . 0.07 0.002
Urban (24 hours) 1 0.02
Urban (8 hours) indoor (16 hours) 1, 0.5 0.02
Industrial (8 hours) indoor (16 hours) 50,0.5 0.5
Source: Arthur D. Little, Inc. •
32
-------
TABLE 24. EXPOSURE OF A CHILD TO TRIHALOMETHANES
IN SWIMMING POOLS
Exposure (mg/day)
Typical^ Maximum
Type of Pool
freshwater 0.2 1.3
(chloroform)
saltwater 0.7 4.2
(bromoform)
1 42
Assumes a total exposure (0.88 x 10 cm ) of 1 hour per day at
the mean concentrations of 156 ug/1 and 657 ug/1 for chloroform
and bromoform, respectively.
2
Assumes an exposure of 3 hours per day at the maximum concentrations
of 430 ug/1 and 1278 ug/L respectively.
Note: The permeability constant of 125 x 10 cm/hour was used
for all calculations.
Source: AOL estimates, from Beech (1980)
33
-------
5. Conclusions
Both inhalation and ingestion of chloroform in drinking water
appear to be important exposure routes. The relative importance of
these two exposure routes depends upon the location. Inhalation may
be more important in highly industrialized areas, whereas chloroform
in drinking water may be more important where water is chlorinated and
levels in air are low. Ingestion of chloroform in food does not appear
to be an important exposure route. Dermal contact of persons spending
long periods in swimming pools may result in high exposures to chloroform.
Exposure to the other trihalomethanes occurs primarily through
drinking water, although dermal contact with these compounds may result
in significant exposure in some situations. The results in Table 24
suggest that dermal exposures can be high for persons spending a signifi-
cant amount of time in swimming pools. It should be noted that-inhalation
exposures would probably be occurring at the same time. Beech (1980)
calculated that inhalation exposures would be about one-half as large as
dermal exposures under the same conditions.
-------
REFERENCES
Alavanga, M.; Goldstein, I.; Susser, M. A case control study of gastro-
intestinal and urinary tract cancer mortality and drinking water chlor-
ination. Chapter 30. Jolley, R.L.; Gorchev, H.; Hamilton, D.H. Jr.
eds.' Water chlorination: environmental impact and health effects.
Vol. 2. Ann Arbor, Michigan: Ann Arbor Science Publishers, Inc.; 1978:
395-409.
Anders, M.W.; Stevens, J.L.; Sprague, R.W.; Shaath, Z.; Ahmed; A.E.
Metabolism of haloforms to carbon monoxide. II. _In vivo studies. Drug
Metab. 6(5):556-560; 1978.
Beech, J.A. Estimated worst case trihalomethane body burden of a child
using a swimming pool. Medical Hypotheses 6:303-307; 1980.
Beech, J.A.; Diaz, R.; Ordaz, C.; Falomeque, B. Nitrates, chlorates
and"trihalomethanes in swimming pool water. Accepted for publication.
Am. J. Public Health; 1980.
Bowman, F.; Borzelleca, J.F.; Munson, A.E. The toxicity of some halo-
methanes in mice. Toxicol. Appl. Pharmacol. 44(1) -.213-216; 1978.
Brown, D.M.; Langley, P.F.; Smith, D.; Taylor, D.C. Metabolism of
chloroform - I: the metabolism of (14C) chloroform by different species.
Xenobiotica 4(3):151-163; 1974.
Brownlee, K.A. e^ al. The up-and-down method with small samples. Amer.
Statist. Assoc. J. 48:262; 1933. (As cited in U.S. EPA 1977).
Capel, I.D.; Williams, D.C. The effect of chloroform ingestion on the
growth of some murine tumours. IRCS Medical Science 6:435; 1978.
Cantor, K.P.; Hoover, R.; Mason, T.J.; McCabe, L.J. Associations of
cancer mortality with halomethanes in drinking water. J. Nat. Cancer
Inst. 61(4):979-985; 1978.
Cresteil, T.; Beaune, P.; Leroux, J.P.; Lange, M.; Mansuy, D. Biotrans-
formation of chloroform by rat and human liver microsomes; in vitro
effect on some enzyme activities and mechanism of irreversible binding to
tnacromolecules. Chem. Biol. Interact. 24(2):153-165; 1979.
DeRouen, T.A.; Diem, J.E. Relationships between cancer mortality in
Louisiana drinking-water source and other possible causative agents.
Incidence of Cancer in Humans. Hiatt, H.H.; Watson, J.D.; Winston, J.A.
eds. Proceedings of the Cold Spring Harbor Conferences on origins of
human cancer. Vol. 4. Cold Spring Harbor: Cold Spring Harbor Laboratory;
1975: 1&0-1A3.
35
-------
DeSalva, S.; Volpe, A.; Leigh, G.; Regan, T. Long-term safety studies
of a chloroform-containing dentifrice and mouth-rinse in man. Food
Cosmet. Toxicol. 13(5):529-532; 1975.
Dykan, V.A. Changes in liver and kidney functions due to methylene
bromide and bromoform. Nauchn. Tr. Ukr. Nauch.-Issled. Inst. Gig. Tr.
Profzabol. 29:82-90; 1962. (As cited in HAS 1977).
Eschenbrenner, A.B.; Miller, E. Induction of hepatomas in mice by
repeated oral administration of chloroform with observation on sex
differences. J. Nat. Cancer Inst. 5:251-255; 1945.
Fry, B.J.; Taylor, I.; Hathaway, D.E. Pulmonary eliminations of chloro-
form and its metabolites in man. Arch. Int. Phannacodyn. Ther. 196:98;
1972.
Grant, W.M. Toxicology of the eye. Second Edition. Springfield,
IL: Charles C. Thomas; 1974: 267-268. (As cited in U.S. EPA 1979a).
Heywood, R.; Sortwell, R.J.; Noel, P.R.; Street, A.E.; Prentice, D.E.;
Roe, F.J.; Wadsworth, P.F.; Worden, A.N.; Van Abbe, N.J. Safety evalua-
tion of toothpaste containing chloroform. III. Long-term study in beagle
dogs. J.. Environ. Pathol. Toxicol. 2(3):835-851; 1979.
•
Hill, R.'N.; Clemens, T.L.; Liu, D.K. e£ al. Genetic control of chloro-
form toxicity in mice. Science 190:159-161; 1975.
Hill, R.N. Differential toxicity of chloroform in the mouse. Ann. N.Y.
Acad. Sci. 298:170-175; 1978.
Ilet, K.F.; Reid, W.P.; Sipes, I.G.; Krishna, G. Chloroform toxicity in
mice: correlation of renal and hepatic necrosis with covalent binding
of metabolites to tissue macromolecules. Exp. Mol. Path. 19:215-229;
1973. (As cited in U.S. EPA 1979b).
International Agency for Research on Cancer (IARC). Working group on the
evaluation of the carcinogenic risk of chemicals to man. IARC
monograph on the evaluation of the carcinogenic risk of chemicals to man.
1:61-65; 1971.
International Commission on Radiological Protection (ICRP). Report of
the Task Group on Reference Man. New York, NY: Pergamon Press; Adapted
October 1974.
Jones, W.M.; Marguis, G.; Stephen, C.R. Hepatotoxicity of inhalation
anesthetic drugs. Anesthesiology 19:715-723; 1958. (As cited in U.S.
EPA 1979a).
Kimura, S.T.; Ebert, O.M.; Dodge, P.W. Acute toxicity and liai:s of
solvent residue for 16 organic solvents. Toxicol. Appl. Pharmacoi.
19:699-704; 1971.
36
-------
Klassen, C.D.; Flaa, G.L. Relative effects of chlorinated hydrocarbons
on liver and kidney function in mice. Toxicol. Appl. Pharmacol. 9:139-
151; 1966.
Kraybill, H.F. Carcinogenesis induced by trace contaminants in potable
water. Bull. N.Y. Acad. Med.- 54(4) :413-427; 1978.
Kucob, S.D.; Plaa, G.L. A procedure for estimating the hepatoxic poten-
tial of certain industrial solvents. Toxicol. Appl. Pharmacol. 4:354-
361; 1962.
Lucas, G.H.S. A study of the fate and toxicity of bromine and chlorine
containing anesthetics. J. Pharm. Exp. Ther. 34:223-227; 1928.
Maiten, J.E.; .Spruit, D.; Boemaars, H.G.M.; deKeizer, M.J.M. Horny
layer injury by solvents-. Verufsdermatosen 16:135-147; 1968. (As cited
in U.S. EPA 1979b).
McConnell, G.; Ferguson, D.M.; Pearson, C.R. Chlorinated hydrocarbons
and the environment. Endeavor 34(121):13-18; 1975.
Miklashevskii, V.E.; Tugarinova, V.N.; Rakhmanina, M.L.; Yakovleva, G.P.
Toxicity of chloroform administered perorally. Hyg. Sanit. 31:320-322;
1966.
National Academy of Sciences (NAS). Drinking water and health.
Washington, DC: NAS; 1977.
National Academy of Sciences (NAS). Chloroform, carbon tetrachloride,
and other halomethanes: An environmental assessment. Washington, DC:
NAS; 1978. 294 p.
National Cancer Institute (NCI). Report on carcinogenesis bioassay of
chloroform. March 1, 1976.
Page, T.; Harris, R.H.; Estein, S.S. Drinking water and cancer mortality
in Louisiana. Science 193:55-57; 1976.
Palmer, A.K.; Street, A.E.; Roe, F.J.; Worden, A.N.; Van Abbe, N.J.
Safety evaluation of toothpaste containing chloroform. II. Long-term
studies in rats. J. Environ. Pathol. Toxicol. 2(3):321-833; 1979.
Patty, F.A. Industrial hygiene and toxicology, Vol. II. 2nd Ed. New
York, NY: Interscience Publishers; 1963.
Petzold, G.L.; Swenberg, J.A. Detection of DNA damage induced in vivo
following exposure of rats to carcinogens. Cancer Res. 38(6):1589-1594;
1978.
?ohl, L.R.; Booshan, 3.; Whittaker, N.F.; Krishna, G. Phosgene: A
mecabolice of chlorofora. Biochem. Biophys. Res. Commun. 79(3):53&-59I;
1977.
37
-------
Poobalasingham, N.; Payne,. J.P. The uptake and elimination of chloro-
form in man. Br. J. Anesth. 50(4):325-329; 1978.
Powers, M.B.; Voelker, R.W. Evaluation of the oncogenic potential of
chloroform by long-term oral administration in rodents. (Abstract)
Toxicol. Appl. Pharmacol. 37(1):179; 1976.
Registry of Toxic Effects of Chemical Substances (RTECS). Washington,
DC: Department of Health, Education, and Welfare, National Institute of
Occupational Safety and Health; 1977.
Reitz, R.H.; Gehring, P.J.; Park, C.N. Carcinogenic risk estimation for
chloroform - an alternative to EPA's procedures. J. Food & Cosmet.
Toxicol. 16:511-514; 1978.
Renne, R.A.; Ferrell, J.F.; Voelker, R.E.; Powers, M.B. Pathology of
long-term oral administration of chloroform in rodents. Toxicol. Appl.
Pharmacol. 37(1):179-180; 1976.
Roe, F.J.C.; Carter, R.L.; Mitchley, B.C.V. Tests of miscellaneous sub-
stances for carcinogenesis. Test of chloroform and 8-hydroxyquinoline for
carcinogenicity using newborn mice. A.R. Brit. Emp. Cane. Camp. 46:13;
1968. (As cited in IARC 1971).
Roe, F.J.C.; Palmar, A.K.; Worden, A.N.; Van Abbe, N.J. Safety evalua-
tion of toothpaste containing chloroform. Part I. Long-term studies in
mice. Abstract. Plaa, G.L.; Duncan, W.A. eds. International congress
on toxicology; 1978. New York, NY: Academic Press; 1979.
Rudali, G. A-propos de 1'activite oncogene de quelques hydrocarbures
halogenes utilises en therapeutique. UICC Monogr. Ser. 7:138; 1967.
(As cited in IARC 1971).
/
Sax, N.I. Dangerous properties of industrial materials. 5th ed. New
York, NY: Van Nostrand Reinhold Company; 1979.
Schroeder, H.G. Acute and delayed chloroform poisoning — a case report.
Brit. J. Anaesth. 37:972-975; 1965.
Schwetz, B.A.; Leong, B.K.J.; Gehring, P.J. Embryo- and feto-toxicity
of inhaled chloroform in rats. Toxicol. Appl. Pharmacol. 28:442-451;
1974.
Simmon, V.F.; Tardiff, R.G. The mutagenic activity of halogenated
compounds found in chlorinated drinking water. Water Chlorination
2:417-431; 1978.
Stevens, J.L.; Anders, M.W. Studies on the mechanism of Che metabolism
of haloforns co carbon monoxide. Toxicol. Appl. Pharmacol. 45(1):297;
1978.
33
-------
SCurrock, J. Lack of mutagenic efface of haloethane or chloroform on
cultured cells using Che azaguanine test system. Brit. J. Anaesth.
49(3):207-210; 1977.
Sylvia, A.E.; Miller, J.D.; Baxter, W.S. Preliminary investigations
into the presence of haloforms in Massachusetts drinking water supplies.
J. M.E. Water Works. March: 30-40; 1978.
Symons, J.M.; Bellar, T.A.; Carswell, J.K.; OeMarco, J.; Kropp, K.L.;
Robeck, G.G.; Seeger, D.R.; Slocum, C.J.; Smith, B.L.; Stevens, A.A.
National organics reconnaissance survey for halogenated organics. J.
Am. Water Works Assn. 67:634-646; 1975.
Taylor, D.C.; Brown, D.M.; Keeble, R.; Langley, P.F. Metabolism of
chloroform - II: A sex difference in the metabolism of (14C) chloroform
in mice. Xenobiotica 4(3):165-174; 1974.
Theiss, J.D.; Stoner, G.D.; Shimkin, M.B.; Weisburger, E.K. Test for
carcinogenicity of organic contaminants of United States drinking waters
by pulmonary tumor response in Strain A mice. Cancer Res. 37:2717-2720;
1977.
Thompson, D.J.; Warner, S.D.; Robinson, V.B. Teratology studies on
orally administered chloroform in the rat and rabbit. Toxicol. Appl.
Pharmacol. 29:348-357; 1974.
Timms, R.M.; Moser, K.M. Toxicity secondary to intravenously admini-
stered chloroform in humans. Arch. Intern. Med. 135:1601-1603; 1975.
Torkelson, T.R.; Oyen, F.; Rowe, V.K. The toxicity of chloroform as
determined by single and repeated exposure of laboratory animals. Am.
Md. Hyg. Assoc. J. 37(12):697-705; 1976.
Uehleke, H.; Werner, T.; Griem, H.; Kramer, M. Metabolic activation of
haloalkanes and tests in vitro for mutagenicity. Xenobiotica 7(7):393-
400; 1977.
U.S. Environmental Protection Agency (U.S. EPA). Water quality criterion
document. Interim draft. No. II. Washington, DC: Office of Water
Planning and Standards, U.S. EPA; November 3, 1977.
U.S. Environmental Protection Agency (U.S. EPA). National organics
monitoring survey. Unpubl. Washington, DC: Technical Support Division,
Office of Water Supply, U.S. EPA; 1978.
U.S. Environmental Protection Agency (U.S. EPA). Statement of basis and
purpose for an amendment to the National Interim Primary Drinking Water
Regulations on Trihalomethanes. Washington, DC: Office of Drinking
Water, Criteria and Standards Division, U.S. E?A; 1979a.
39
-------
U.S. Environmental Protection Agency (U.S. EFA). Chloroform — The
consent decree ambient water quality criteria document. Washington, DC:
Office of Water1 Planning and Standards, U.S. EPA; 1979b.
U.S. Environmental Protection Agency (U.S. EFA). Identification and
evaluation of waterborae routes of exposure from other than food and
drinking water. .Report No. EPA-440/4-79-016. Washington, DC: Office
of Water Planning and Standards, U.S. EPA; 1979c.
Van Dyke, R.A.; Chemoweth, M.B.; Van Poznak, A. Metabolism of volatile
anesthetics - I: conversion in vivo of several anesthetics to 1*C02
and chloride. Biochem. Pharmacol. 13:1239-1247; 1964.
Vessell, E.S.; Lang, C.M.; White, W.J.; Passanante, G.T.; Hill, R.N.
_et al. Environmental and genetic factors affecting the response of
laboratory animals to drugs. Fed. Proc. 35:1125-1132; 1976.
Wallace, C.J. Hepatitis and nephrosis due to cough syrup containing
chloroform. Calif. Med. 73:442; 1959. (As cited in U.S. EPA 1979a).
Winslow, S.G.; Gerstner, H.B. Health aspects of chloroform: a review.
Drug. Chem. Toxicol. 1(3):259-276; 1978.
90
-------
CHAPTER VI
EFFECTS AND EXPOSURE — AQUATIC BIOTA
A. EFFECTS OF TRIHALOMETHANES ON AQUATIC ORGANISMS
1. Introduction
This section provides information about the levels of trihalometh-
anes that disrupt the normal behavior and metabolic processes of aquatic
organisms. Despite a thorough literature search, no toxicity data for
dibromochloromethane or dichlorobromomethane were found. Moreover,
information on chloroform and bromoform was limited. In assessing
threshold effects levels, it is important to note that, with few excep-
tions, the bioassays were conducted under static conditions, and the
toxicant concentrations were calculated rather than measured. Conse-
quently, the LC5Q values (concentrations lethal to 50% of test organisms)
reported could underestimate actual toxicity, as a result of toxicant
evaporation and metabolism by test organisms.
2. Chloroform
Earlier studies on the aquatic toxicity of chloroform reported
"negative" (avoidance) reactions as a measure of effects levels. Clayberg
(1917) observed "positive reactions" (not defined) in brown bullheads and
white suckers in 214 mg/1 of chloroform, and "a few negative reactions
... in somewhat higher concentrations." Jones (1974) recorded an
avoidance reaction by the ten-spined stickleback in 100 mg/1. Fifty
percent of a group of goldfish were anesthetized by a 167 mg/1 solution
of chloroform (Gherkin and Catchpool 1964). This effect was observed to
decrease with increasing temperature.
More recent research on the acute toxicity of chloroform has yielded
lower overall effects levels than the studies described above. The
lowest reported 96-hour LC5Q value for a marine organism is 28 mg/1, for
the dab (Pearson and McConnell 197S). Equally susceptible is Daphnia
magna. with a 48-hour LC5Q value of 28.9 mg/1 chloroform (U.S. EPA 1978a).
The "chronic value" (as defined by U.S. EPA) for this species is 2.5 mg/1.
The 96-hour LC5Q for the pink shrimp, a marine invertebrate, is 81.5 mg/1,
as reported by Bentley.et.al. (1975).
Clayberg (1917) observed mortality in the orange-spotted sunfish
after 1 hour of exposure to concentrations ranging from 107 mg/1 to
153 ag/1.
Anderson £t al. (1979) conducted 96-hour acute toxicity bioassays
with largemouth bass, rainbow trout, and bluegill sunfish. The respec-
tive LCjQ values for chlorofora vere 45-56 oig/1, 15-22 ag/1, and 13-22
aig/1. In a comparative scudy, Bencley ec al. (1975) found rainbow crouc
-------
to be more sensitive to chloroform than bluegill sunfish, with respec-
tive 96-hour LCso'sof 66.8 mg/1 and 115.0 mg/1 in water 35 mg/1 hardness.
When the hardness was increased to 200 mg/1 and the bioassays repeated,
96-hour LCso's values decreased to 43.8 mg/1 and 110.0 mg/1, respectively,
a reduction indicating significantly greater toxicity in hard water for
the rainbow trout. The data available regarding the toxicity of chloro-
form to aquatic organisms are summarized in Table 25.
3. Bromoform
The data for bromoform are limited and are all based on static
bioassays. The bluegill sunfish was the only freshwater finfish tested
for sensitivity to bromoform. The 96-hour LC$Q for this species was
29.3 mg/1. Daphnia magna experienced median lethality after 48 hours of
exposure to 46.5 mg/1 bromoform. For the alga (Selenastrum capricornutum).
the 96-hour EC5Q (effects observed in 50% of test organisms) level was
172 mg/1 with respect to chlorophyll-^ activity, and 116 mg/1 as deter-
mined by cell number (U.S. EPA 1978b).
The effects of bromoform were also measured for three marine spe-
cies. The 96-hour LC$Q for the sheepshead minnow was 17.9 mg/1 bromo-
form, while the. chronic value was 9.2- mg/1. The mysid shrimp was
apparently somewhat less susceptible to bromoform, with a 96-hour LC5Q
value of 24.4 mg/1. The alga, Skeletonema costatum, was tested for
bromoform effects in terms of chlorophyll-^ activity and proportion of
cells affected. The respective 96-hour £€50 concentrations were 12.3
mg/1 and 11.5 mg/1, respectively (U.S. EPA 1978b).
In a study by Gibson _et_ al. (1979), five marine species were tested
for sensitivity to bromoform in continuous-flow bioassays. An LCjQ
value was not obtained for the littleneck clam (Protothaca staminea)
because, in high concentrations of bromoform, the clams closed up and
did not pump water. However, mortality was observed in 5 mg/1 and
10 mg/1 bromoform. At concentrations above 10 mg/1, the Atlantic oyster
(Crassostrea virginica) and the quahog (Mereenaria mereenaria) also
closed and did not circulate water, and this resulted in no mortalities
during the 96-hour bioassay. However, when exposure was continued for
an additional 3 days, several organisms died. Ninety-six-hour LC5Q
concentrations were then estimated at 40 mg/1 for the oyster, and 140
mg/1 for the quahog. For the brown shrimp (Penaeus aztecus) and the
menhaden (Brevoortia tyrannus). 96-hour LC5Q values were given as 26 mg/1
and 12 mg/1, respectively. These data are summarized in Table 26.
4. Summary
The LC5Q values for the bluegill included the lowest and highest
values recorded, from 13 mg/1 to 115 mg/1 chloroform. Daphnia magna and
rainbow trout were found to be comparatively sensitive, with LC5Q values
for chloroform of 28.9 mg/1 and 15 mg/1, respectively." The limited data
do not indicate whether marine or freshwater species are generally more
sensitive to chloroform. In one bioassay, 'rainbow trout were signifi-
cantly more susceptible in hard water than in soft water. No data were
available on the toxicity of chloroform to aquatic plants.
92
-------
TAULE 2). EFFECTS Of CHLOROFORM OH AQUATIC ORGANISMS
Ciiiiruiit rail tin
28
2B.-J
41. H
6b.H
Ul.i
IOII
lOfc.B'J-IV,!. 7
IIII.O
115.0
167
Simcieu
(l.lmanda up.)
Daphnla
Ualnbow trout
(Sal mo Kalrdncrl
Rainbow trout
(Salmo Kalrdnerl)
fink shrimp
(I'mtaciia duorarum)
fun-up 1 nud bllckleback
(PyKosteus
Oiange-spotted sunflah
(l.tipoials huinllls)
BluuBlll sunfish
(Lepoiala m.icroclilrus)
Illueglll sun Mali
(Lepomla macrochlma)
Culddah
(Carasalua amatua)
Brown bullhead
(Icculiiriiii nebulosus)
*zn.U White uuckcr
j^ouiiib I'ooiimirboul)
Hardness (ug/1)
Salinity o/oo pll Temperature Teat Type Test Duration Effects
-------
TABLE 26. EFFECTS OF BHOMOFUKM ON AQUATIC OKCANISMS
Cuncuntratlon
(uig/l)
5
9.2
11.5
12. J
12
17.9
24.4
26
29.1
40
46.5
112
116
Water
l.lttleneck clan
(Protothaca atamtnea) Marine
Shecpsliead minnow
(Cyprlnodon varlegatiis)
Alga
(Skctutonema costacuai)
cell number
chlorophyll-ji
Menhaden
(Brevoortla tyrannuu) Marine
Sheepshead nlnnuw
(Cyprlnodon varleaatiia) Marine
Mysld shrimp
(Hyuldopsls bahla) Marine
Brown shrimp
(Penacus aztecus) Marine
Blueglll sun f I ah
(I.epouia macrochlrua) Freshwater
Atlantic oyster
(Craaaostrea vlrelnlca) Marine
Paphnla aa^na • Freshwater
Algu
(Solcnantruin capr Iconiutuia)
chlorophyll-a
cell number
Teat Duration
96 hr
-
96 hr
96 hr
96 hr
96 hr
96 hr
96 hr
96 hr
96 hr
48 hr
96 hr
96 hr
Effects
Mortality
Chronic Value
ESo
ESO
LC
'So
^50
'•So
LCSO
'•So
LC50
ESo
EC50
Heiercncub
et .il. (1979)
U.S. EPA (1978)
U.S. EHA (197U)
U.S. EPA (1978)
Gibson £t al. (1979)
U.S. EPA (1978)
U.S. EPA (1978)
Gibson e_t a_l. (1979)
U.S. EPA (1978)
Glbuon et al^. (1979)
U.S. EPA (1978)
U.S EPA (1978)
U.S. EPA (1978)
-------
Although information on bromoform toxicity was equally scarce,
bromoform appears to be somewhat more toxic to aquatic life than chloro-
form. The LC5Q values ranged between 12 mg/1 for the menhaden and
46.5 mg/1 for Daphnia magna (the one species tested that exhibited less
sensitivity to bromoform than to chloroform). For the marine alga
species tested, EC^o concentrations were 11.5- mg/1 and 12.3 mg/1,
approximately one order of magnitude less than for the freshwater alga.
B. EXPOSURE OF AQUATIC BIOTA TO TRIHALOMETHANES
Monitoring data for the trihalomethanes considered here are
extremely sparse; consequently, any generalizations made with regard
to exposure levels must be strictly qualified. Since the levels
observed were in many cases at or below the detection limits of the
analytic procedure, future improvements in detection ability could
conceivably lower the estimates made below.
Ambient concentrations of chloroform in water normally fell between
0.1 ug/1 and 10 ug/1, with a small proportion of measurements exceeding
10 ug/1 (see Chapter IV). In a few instances the levels exceeded 100
ug/L Nearly all (97%) of the bromoform measurements were between 1.0
ug/1 and 10 ug/1. Roughly two-thifds of observations concerning both
bromodichloromethane and dibromochloromethane fell into the 0.1-1.0
ug/1 category, with approximately one-third exceeding 1.0 ug/1. The
Pacific Northwest and California basins were the most intensively sampled
watersheds for the latter two trihalomethanes. Perhaps coincidentally,
the concentrations in the Pacific Northwest for both were usually between
1.0 ug/1 and 10 ug/1, while California had similar levels of dibromo-
chloromethane .
In addition, no fish kills attributed to chloroform have been
reported in the data files of the Monitoring and Data Support Division,
Office of Water Regulations and Standards, U.S. EPA.
Most of the data for bromoform, bromodichloromethane, and dibromo-
chloromethane were remarked, which probably means the chemical was
detected but the concentration was not quantified. However, the fact
that the levels of all of the trihalomethanes were usually so low as to
approach the detection limits may suggest that ambient exposure levels
in most areas of the U.S. are negligible. Factors that modify the
toxicity of trihalomethanes for aquatic biota have not been identified •
or studied sufficiently to merit a discussion of geographical variations
in such factors.
95
-------
REFERENCES
Anderson, D.R. _et_aj. Progress Reporc Covering Period January 1 through
March 31, 1979: Biocide by-products in aquatic environments. PNL-2988.
U.S. Nuclear Regulatory Commission; 1979.
Bentley, R.E.; Heitmuller, F.; Sleight, B.H.; Parrish, P.R. Acute coxicity
of chloroform to bluegill (Lepomis macrochirus), rainbow trout (Salmo
gairdneri). and pink shrimp (Penaeus durorarum). Contract No. WA-6-1414-B.
Washington, DC: U.S. Environmental Protection Agency; 1975.
Gherkin, A.; Catchpool, J.F. Temperature dependence of anesthesia in
goldfish. Science 144:1460-1462; 1964.
Clayberg, H.D. The effect of ether and chloroform on certain fishes.
Biol. Bull. 32:234; 1917.
Gibson, C.I.; Tone, F.C.; Wilkinson, P.; Blaylock, J.W. Toxicity and
effects of bromoform on five marine species. PNL-3023. U.S. Nuclear
Regulatory Commission; 1979. Available from NTIS, Springfield, VA;
NUREGICR-0835.
Jones, J.R.E. The reactions of Pygosteus pungitius L. to toxic solutions.
J. Exp. Biol. 24:110; 1974.
Pearson, C.R.; McConnell, G. Chlorinated C^ and G£ hydrocarbons in the
marine environment. Proc.* R. Soc. Lond. B 189:305-322; 1975.
U.S. Environmental Protection Agency (U.S.EPA). Ambient water quality
criteria. Criterion document — chloroform. Washington DC: Criteria
and Standards Division, Office of Water Planning and Standards, U.S.
EPA; 1978a. Available from: NTIS, Springfield, VA; PB 292 427.
U.S. Environmental Protection Agency (U.S. EPA). Ambient water quality
criteria. Criterion document — halomethanes. Washington, DC: Criteria
and Standards Division, Office of Water Planning and Standards, U.S. EPA;
1978b. Available from: NTIS, Springfield, VA; PB 296 797.
96
-------
CHAPTER VII. RISK CONSIDERATIONS
A. INTRODUCTION
This chapter assesses risk to both humans and non-humans associated
wich exposure to chloroform, bromoform, dibromochloromethane, and
bromodichloromethane. However, only the risks associated with exposure
to chloroform can actually be quantified, since sufficient information
on exposure and effects was not available for the other three trihalo-
methanes.
Several different dose-response models were applied to the available
data concerning the carcinogenic effects of chloroform, and the range of
potential human risk was estimated for various possible exposure levels.
These risk estimates have been compared against other risk estimates
derived from the literature, although the latter usually focus upon
drinking water alone as the pathway of exposure.
•
With respect to non-human biota, .little information is available on
the effects of environmental exposure to trihalomethanes; therefore risks
to non-human biota were addressed only in qualitative terms.
B. RISK TO HUMANS
1. Effects of Trihalomethanes
Chapter V discusses the effects of trihalomethanes in laboratory
animals and humans in some detail. Tables 27-30 summarize these effects.
Chloroform has been shown to be carcinogenic in rats and mice. However,
no evidence indicates teratogenicity or mutatagenicity. Acute effects
attributable to chloroform include liver necrosis and kidney damage.
It is likely that these effects would not be observed in the general
population at the present exposure levels.
The other three trihalomethanes have not been tested for carcino-
genicity; however, they all are potential mutagens based on the results
of the Ames bioassay.
As a result of the available data, the only risks that can be
quantified at this time are the potential cancer risks due to chloroform
exposure, which are considered below.
2. Carcinogenic!ty of Chloroform
The potential carcinogenic effects of chloroform upon humans can be
quantitatively estimated through extrapolation of ia vivo laboratory
results. The available data concerning mammalian effaces were discussed
previously in Chapter V and summarized in Table 27. For extrapola-
tion purposes, the NCI data-shown in Table 31 have been selected.
-------
TABLE 27. ADVERSE EFFECTS OF CHLOROFORM ON MAMMALS
Adverse Effect
No Apparent
Species Lowest Reported Effect Level Effect Level
Hepatocellular
Carcinoma
Mouse
138 mg/kg (diec)
Renal Epithelial
Tumors
Rat
90 mg/kg (diet)
Teratogenesis
Mouse
Rat
Rat
Rabbit
60 mg/kg (gavage)
490 mg/m3/hr inhal
17 mg/kg
126 mg/kg
(orally)
50 mg/kg
(orally)
Hepatic and Renal Rat
Necrosis Mouse
250 mg/kg (orally)
Oral LD
50
Mouse
119 mg/kg
Source: Chapter V.
-------
TABLE 28. ADVERSE EFFECTS OF BROMOFORM ON MAMMALS
No Apparent
Adverse Effect Species Lowest Reported Effect Level Effect Level
Pulmonary Adenoma Mouse 48 mg/kg ip x 23 4 mg/kg ip x 18
Mutagenicity Salmonella 50 ul (vapor)
typhimurium
TA 100
Teratogenicity — No data available
Chronic Toxic No data available
Effects
Median Oral Mouse 1400 mg/kg
Lethal Dose
Source: Chapter V.
-------
TABLE 29. ADVERSE EFFECTS OF DIBROMOCHLOROMETHANE ON MAMMALS
No Apparent
Adverse Effect Species Lowest Reported Effect Level Effect Level
Carcinogenicity
No data available
Mutagenicity Salmonella
typhimurlum
TA 100
10 ul (vapor)
Teratogenicity —
Chronic Toxic
Effects
No data available
No data available
Median Oral
Lethal .Dose
Mouse
800 mg/kg
Source: Chapter V.
100
-------
TABLE 30. ADVERSE EFFECTS OF DICHLOROBROMOMETHANE ON MAMMALS
Adverse Effect
Species
No Apparent
Lowest Reoorted Effect Level Effect Level
Pulmonary Adenoma Mouse
100 mg/kg i.p. x 24 40 mg/kg i.p. x 24
Mutagenicity
Salmonella
typhimurium
TA 100
50 ul (vapor)
Teratogenlcity
Chronic Toxic
Effects
No data available
No data available
Median Oral
Lethal Dose
Mouse
450 mg/kg
Source: Chapter V.
101
-------
TABLE 31. CARCINOGENIC EFFECTS OF CHLOROFORM IN RODENTS
Soecies
Dosage(mg/kg/day) Response (%) Tumor Site
NCI Study
male rat
male mouse
female mouse
ICI Study
male mouse
0
90
180
0
138
277
0
238
477
0
17
60
0
12
26
6
36
98
0
80
95
0
0
22
kidney
liver
liver
renal cortex
Source: Reuber (1979).
102
-------
demonstracing increased renal cumors in mice, and increased hepatic
tumors in male rats. The ICI data also indicate carcinogenic activity
at a lower dose, but the NCI data form a larger and more consistent
data base for extrapolation. It must be noted that interpretation of
these results for human risk assessment is subject to a number of impor-
tant qualifications and assumptions:
• Though positive carcinogenic findings exist, there have
also been negative findings in tests with several species.
Thus the carcinogeniclty of chloroform to humans is far
from certain.
• Assuming that the positive findings indeed provide a
basis for extrapolation to humans, the estimation of
equivalent human doses involves considerable uncertainty.
• Due to inadequate understanding of the mechanisms of
carcinogenesis, there is no scientific basis for selecting
among several alternate dose-response models, which yield
widely differing results.
• Apart from the risk via ingestion, there may be rieks due
to either dermal absorption or inhalation of chloroform.
However, toxicological data corresponding to these routes
are not available for extrapolation purposes.
In order to deal with the large uncertainties inherent in extrapolation
to humans, three commonly-used dose-response models have been applied
in order to establish a range of potential human risk.
The first step in extrapolation- of laboratory data is to determine
the equivalent human dose. Reitz £t al. (1978) showed that considera-
tion of metabolic rates in rodents led to a more consistent interpreta-
tion of laboratory dose-response data than did the U.S. EPA method of
correcting for surface area. They argue that humans are less sensitive
to chloroform than rodents due to their slower metabolic rate (they
maintain that the effects of chloroform are due to a .toxic metabolite),
and that past efforts at extrapolation have over-estimated the human
risk. Accordingly, the rodent doses have been used as a basis for
extrapolation in this analyses, without an attempt to convert them to
a human equivalent dose. If a surface area conversion factor were
introduced, it would decrease the equivalent human dose by a factor of
6 for rat data or 14 for mouse data; thus the conversion would imply
considerably higher human risk.
The three dose-response models used to extrapolate human risk were
the linear "one-hit" model, the log-probit model, and the multistage
model. The latter is actually a generalization of the one-hit model,
in which the hazard rate was permitted to be a quadratic rather than a
linear function of the dose. All of these models are well known in the
103
-------
literature, and a theoretical discussion may be found in Arthur 0.
Little (1980). The one-hit models assume that the probability of a
carcinogenic response is described by
P (response at dose X) » 1 - e ""
where h(x) is the "hazard rate" function. The log-probit model assumes
that human susceptibility varies with dose, according to a log-normal
distribution. Due to their differing assumptions, these dose-response
models usually give widely differing results when effects data are
extrapolated from high laboratory doses to the low doses typical of
environmental exposure.
The results of these extrapolations are shown in Table 32 for human
exposures ranging from 0.01 mg/day to 10 mg/day. The linear model may
be applied to several different groups of data, yielding a range of risk
estimates. The best, and most conservative, fit was obtained using the
female mouse data for liver cancer, which indicate a coefficient of
about 8.6 x 10"5 increase in risk per mg/day increase in human dose.
The log-probit extrapolation was performed using a unit slope with
respect to the log-dose, which is generally accepted as a conservative
procedure. Again, a wide range of estimates can be derived by selecting
different groups of data. In order to provide a less conservative esti-
mate of the risk, the male rat data were also utilized for kidney cancer,
which predicted risks considerably lower than did the female mouse data.
Finally, the multistage model was fit to the combined data for liver
cancer in male and female mice.
3. Literature Review of Carcinogenic Risks of Chloroform
Since chloroform was first identified in drinking water in 1974,
there has been a great deal of interest in the assessment of carcino-
genic risk from trihalomethane exposure (Stokinger 1977, Tardiff 1977,
NAS 1977, MAS 1978, Reitz &t al. 1978, U.S. EPA 1979). These analyses
contain widely varying results, which are discussed briefly below.
Tardiff (1977) used four different models for extrapolation of
carcinogenic effects observed in laboratory animals at high doses to
effects that may be anticipated at low doses, but did not convert data
from one species to another. His analysis was based upon a maximum
human dosage of 0.01 mg/kg chloroform per day, and utilized the NCI data
on the carcinogenicity of this chemical. His estimates, based on four
different extrapolation models, range from 0 to 0.84 additional cancers
per million population each year. This is equivalent to a maximum of
60 lifetime tumors per million population at a dose of 0.7 mg/day, which
agrees with the estimates in Table 32. Tardiff concludes that between
0% and 1.6% of cancer incidence in liver or kidneys may be attributable
to chloroform in drinking water.
The National Academy of Sciences (1978) ristc estimation was based
on che multistage model of Guess and Crump, and a species conversion of
104
-------
TABLE 32. PROBABLE UPPER BOUNDS ON EXPECTED
EXCESS TUMORS PER MILLION POPULATION
DUE TO CHLOROFORM EXPOSURE
Exposure level
Extrapolation Model (me/day)
0.01 0.1 1 10
Linear Model l 0.86 8.6 86 860
Log-probit Model l ' N 0.3 32 1300
Log-probit Model2 N N 0.8 72
Multi-stage Model3 .06 .65 6.5 65
Note: Use of a dose conversion factor would increase all
estimates by an order of magnitude
N • negligible
Based on NCI data for female mouse.
Based on NCI data for male rat (renal tumors).
Based on NCI data for male and female mouse.
105
-------
carcinogenicicy data based on surface area. Using Che NCI carcino-
genicicy data, the authors of Che MS study found that an oral dose
of 1 ug chloroform per day resulted in an estimated lifetime risk of
cancer in a 70-kg man of as much as 2 x 10"^. This is considerably
greater than the estimate of Tardiff, when one considers the extremely
low dose assumption. However the majority of this difference is due to
the surface area conversion factor, which Tardiff did not include. The
estimates in Table 32 would be increased by about an order of magnitude
if that conversion factor were used.
4. Human Exposure Scenarios
Table 33 estimates chloroform exposures for five hypothetical
exposure situations: a rural area with and without a chlorinated
drinking water supply, an urban setting, an industrial setting, and
maximum exposures for an adult and child. A number of assumptions have
been used to construct the total exposure estimates. First, 100% absorp-
tion has been assumed for oral and inhalation routes. Second, the
carcinogenic activity of chloroform has been assumed to be similar for
these routes, thus allowing the summation of exposures for purposes of
risk evaluation. There are no data to support the validity of these
assumptions, but they permit consideration of the relative contributions
of various exposure routes to total risk. For the rural, urban, and
industrial settings, the concentrations chosen are meant to represent
typical levels for these areas. The maximum values represent maximum
concentrations for all exposure routes.
As can be seen in Table 33, in areas where drinking water is
chlorinated, drinking water is the predominant route of exposure.
However, in urban areas, inhalation represents an Important additional
exposure route. Swimming in chlorinated pools may represent another
important exposure route. What portion of the U.S. population may fall
into each of the five categories is unknown, and there are certainly
other situations that are not represented in Table 33. Nevertheless,
this table is meant to illustrate a range of typical exposures.
The maximum risks of cancer were estimated for the five exposure
conditions by use of the multistage model results shown in Table 32.
5. Other Trihalomethanes
Table 34 summarizes estimates for known routes of exposure to other
trihalomethanes. Risk estimates of the type made for chloroform cannot
be made for the other trihalomethanes since information on exposure and
effects are lacking. The typical levels of exposure are much lower than
for chloroform; however, exposure in a few locations is high relative to
chloroform. Very little information is available on the effects of bromo-
form, dibromochloromethane, and bromodichloromethane for man. The only
basis for comparison among the trihalomethanes is the oral LDso for mice,
which shows chloroform to be the least toxic (Bowman .at. al_. 1978). Until
che carcinogenesis bioassays for these other chemicals are completed, no
conclusive statements can be tnade about: risk to man.
106
-------
TABLE 33. ESTIMATED EXPOSURE OF HAN TO CHLOROFORM
Probable Maximum
Associated
Risks
Exposure in nig/day (Z of total) (Lifetime
lix|»osure Dermal ~Total Cancer risk
Si illation Drinking Water Food Inhalation Contact (max) per capita)
KuraJ area 0 (0) ] 0.0006 - 0.04 (100) .002 (0) 0 (0) 0.04 0.26 x 10~6
no clilorlnation
Uural area _ ,
,:l, I urination 0.02 - 0.13 (60) 0.0006 - 0.04 (18) .002 (9) 0.03 (13) 0.2 1.3 x 10~
i- Urban .03 - 0.2 (67) 3 0.0006 - 0.04 (13) 0.02 (6) O.OS'(14) 0.3 2.0 x 10~6
o
Industrial * 0.03 - 0.2 (25) 0.0006 - 0.04 (5) 0.5 (60) 0.05 (6) 0.8 5.2 x 10~6
Maximum 1.2 (46) 0.04 (2) 1 (38) 0.3 (12) 2.6 16.9 x 10~6
5 fi 6
Cl.il.l - maximum 0.5 (19) 0.04 (2) 1 (38) 1.1 (41) 2.6 47.3 x 10
I'ercent of total (using maximum values) percentages may not add to due to rounding.
Assumes median level of chloroform In rural areas.
^Assumes chloroform concentration of O.lmg/1. ,
('Assumes concentration of 50 ug/m^ 24 hrs/day.
Includes swimming.
a 25 kg child.
Source: See Chapter V.
-------
TABLE 34. ESTIMATED HUMAN EXPOSURE TO BROMOFORM,
DIBROMOCHLOROMETHANE, AND BROMODICHLOROMETHANE
Exposure (nig/day)
Dermal
Drinking Water Contact Total
Median Max. Typical Max. Typical Max.
Bromoform 0.007 • 0.6 0.001 0.8 0.008 1.4
Dibcomochloromethane 0.007 0.6 0.001 0.07 0.008 0.7
Bromodichloromethane 0.002 0.4 0.003 0.04 0.005 0.4
This is for child swimming in saline pool 2 hrs/day.
Source: Chapter V.
108
-------
C. AQUATIC BIOTA
The complete lack of effects data for bromodlchloromethane and
dibromochloromethane renders a risk analysis for aquatic life impossible.
Although monitoring and effects information on chloroform and bromoform
are extremely limited, some tentative conclusions can be drawn.
On a nationwide scale, no aquatic populations appear to be exposed
consistently to harmful levels of either chloroform or bromoform.
Anderson .et al. (1979) state that the acute toxicity levels found in
their freshwater fish bioassays ."are orders of magnitude above the
maximum level (24.7 mg/1) found in chlorinated water samples taken
across the U.S." However, ambient levels of chloroform between 100 mg/1
and 1,000 mg/1 have been measured, and the occurrence of spills and dis-
charges from industries may result in temporary and localized high
concentrations that could be hazardous to aquatic life. Unfortunately,
monitoring data for all of the trihalomethanes are so scarce that it is
not possible to identify any specific regions that have a potential for
.high levels of exposure to biota.
109
-------
REFERENCES
Anderson, D.R. et al. Progress Report Covering Period January 1 through
March 31, 1979: Biocide by-products in aquatic environments. PNL-2988.
U.S. Nuclear Regulatory Commission; 1979.
Arthur D. Little, Inc. An approach to exposure and risk assessments for
priority pollutants. Draft. Contract 68-01-3857. Washington, DC:
Monitoring and Data Support Division, U.S. Environmental Protection
Agency; 1980.
Bowman, M.F.; Borzelleca, J.F.; Munson, A.E. The toxicity of some halo-
methanes in mice. Toxicol. Appl. Pharmacol. 44(1):213-216; 1978.
National Academy of Sciences (NAS). Drinking water and health.
Washington, DC: NAS; 1977.
National Academy of Sciences (NAS). Chloroform, carbon tetrachloride,
and other halomethanes: an environmental assessment. Washington, DC:
NAS; 1978. 294 p.
Reitz, R.H.; Gehring, P.J.; Park, C.N. Carcinogenic risk estimation for
chloroform - an alternative to EPA's procedures. J. Food & Cosmet.
Toxicol. 16:511-514; 1978.
Reuber, M.D. Carcinogenicity of chloroform. Environ. Health Persp.
31:171-182; 1979.
Stokinger, H.E. Toxicology and drinking water contaminants. J. Amer.
Water Works Assoc. 69(7):399-402; 1977.
Tardiff, R.G. Health effects of organics: risk and hazard assessment
of ingested chloroform. J. Amer. Water Works Assoc. 69(12):658-660;
1977.
U.S. Environmental Protection Agency (U.S. EPA). Ambient water quality
criteria. Criterion document — chloroform. Washington, DC: Criteria
and Standards Division, Office of Water Planning and Standards, U.S. EPA;
1979. Available from: NTIS, Springfield, VA; PB 292 427.
110
-------