A REPORT
ASSESSMENT OF HEALTH RISK
FROM
ORGANICS IN DRINKING WATER
BY AN
AD HOC STUDY GROUP
TO THE
HAZARDOUS MATERIALS ADVISORY COMMITTEE
SCIENCE ADVISORY BOARD
ENVIRONMENTAL PROTECTION AGENCY
April 30, 1975
-------
\
-------
EPA NOTICE
This report has been written as a part of the activities of the
Agency Science Advisory Board, a public advisory group providing extramural
scientific information to the Administrator and other officials of the
Environmental Protection Agency. The Board is structured to provide a
balanced expert assessment of scientific matters related to problems facing
the Agency. This report has not been reviewed for approval by the Agency
and hence its contents do not represent the views and policies of the
Environmental Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
-------
HARVARD UNIVERSITY
SCHOOL OF PUBLIC HEALTH
DEPARTMENT OF PHYSIOLOGY
663 HUNTINGTON AVENUE
BOSTON, MASSACHUSETTS O2I13
April 30, 1975
Dr. Emil M. Mrak
Chairman, Hazardous Materials Advisory Committee,
Science Advisory Board, EPA
University House
University of California
Davis, California 95616
Dear Dr. Mrak:
I transmit, herewith, the report of the Ad Hoc Study Group to Consider
Organics in Drinking Water. The Study Group has attempted to address itself
to the issues contained in your letter of charge of March 12, 1975. Although
the report must be considered a limited, first assessment of health risk from
consuming certain contaminants in drinking water, T believe it addresses the
issues in as objective and comprehensive a manner as was possible under the
time constraints necessary to meet a May 1 due date.
The formulation and preparation of this report was made possible only
by the spirit of cooperation and diligence of the members of the Study Group,
and I commend and thank them for their efforts. I wish also to acknowledge with
thanks the continuous support and responsive assistance provided by Dr. Thomas
Bath, Dr. J. Frances Allen, Mr. Wade Talbot and their secretarial staff in the
Science Advisory Board of the EPA and the assistance of Mr. William Coniglio
of the Office of Water and Hazardous Materials in providing reference material.
The Study Group hopes that the attached report will be useful to you and
to the Administrator of the EPA in developing policy relative to the Safe
Drinking Water Act.
Respectfully submitted on behalf of the Study Group,
Sincerely yours,
SDMrmm
cc: Members of Study Group
Sheldon D. Murphy, Ph.D.
Chairman, Ad Hoc Study Group
ill
-------
-------
A REPORT
ASSESSMENT OF HEALTH RISK
FROM
ORGANICS IN DRINKING WATER
BY AN
AD HOC STUDY GROUP
OF THE
SCIENCE ADVISORY BOARD - HAZARDOUS MATERIALS ADVISORY COMMITTEE
Members:
Dr. Sheldon D. Murphy, Chairman
School of Public Health
Harvard University
Dr. Harris n. Hartzler
Dupont Experimental Station
Dupont Company
Dr. David G. Hoel
National Institute Environmental
Health Sciences
Dr. George B. Hutchison
School of Public Health
Harvard University
Mr. Gregor A. Junk
Ames Laboratories/ERDA
Iowa State University
Dr. Benjamin L. Van Duuren
New York University Medical Center
Dr. Gerald X. Wogan
Massachusetts Institute of Technology
Dr. Philippe Shubik, Eppley Cancer Institute, University of Nebraska
Medical Center, and Mr. Henry J. Ongerth, State of California, Department
of Public Health, were selected and agreed to serve on the Ad Hoc
Study Group. Because of unforeseen circumstances, however, they were unable
to attend the Study Group meetings or participate in the preparation of this
report.
April 30, 1975
-------
-------
SUMMARY
.Any assessment of possible human health risk associated
with consumption of drinking water contaminated with low concen-
trations of organic chemicals depends on at least three factors:
the adequacy of analytic methods for identifying and measuring
the contaminants and the scope of their application, the existence
and adequacy of toxicological data on the contaminants, and the
extent to which appropriate epidemiologic studies have been conducted
to test a hypothesis of association derived from the water quality
data and the toxicological data. The Study Group addressed itself
to these issues, with carcinogenesis as the toxic effect of primary
concern. It is recognized that a complete assessment of the possible
risk should include those risks associated with exposure to other
contaminants— such as pesticides, asbestos, and inorganic chemicals —
which were explicitly excluded from the charge to the Study Group.
It was the judgment of the Study Group that current methods
of extraction, identification and measurement are available and
adequate for field surveys of the identified drinking water
contaminants of major toxicologic concern. It is highly unlikely,
however, that the majority of drinking water purveyors would have
available sophisticated equipment and trained personnel sufficient
to provide monitoring of the individual contaminants on a routine
basis. It would be desirable if procedures could be developed
to permit routine monitoring of groups of potentially harmful
chemicals. .Any trends in contamination of a water supply suggested
by chemical-group monitoring might then be subjected to more detailed
study.
The Study Group also expressed concern that the chemicals
which have been measured account for only a few percent of the total
organic content of drinking water. Thus, attempts to evaluate the
health risk of contamination may be mistakenly directed toward
identified, potentially toxic compounds while other groups of
compounds perhaps of equal or greater toxicologic significance
go undetected. Furthermore, attention has been focused on the
concentrations of contaminants in drinking water itself, while a
complete analysis of the problen would also require analytical
chemistry data on exposure to these chemicals by ingestion of
foods and beverages processed with the contaminated water and on
other possible exposures resulting indirectly from environmental
redistribution and possible biomagnification of the chemicals in
food organisms, which also consume the water in question. On the
other hand, the Study Group felt that other unrelated sources of
exposure to some of the chemicals in question would likely contribute
a much greater potential intake than the consumption of drinking water.
VI1
-------
With respect to assessment of health risk associated with
exposure to the specific contaminants identified in the charge
to the Study Group, it was concluded that some human health risk
exists. This conclusion was reached on the basis of evidence that
some of the compounds, particularly chloroform, are widespread
contaminants of U.S. drinking water supplies, and that studies in
laboratory animals indicate that chloroform produces hepatomas.
It should be emphasized that experimental carcinogenesis data for
chloroform are extremely limited, although support for its tumorigen-
icity is reinforced by more extensive studies demonstrating carcinogenic
action of the related compound, carbon tetrachloride. These two
compounds probably act by a similar mechanism to produce hepatomas.
Carbon tetrachloride, although occasionally identified as a contaminant
of drinking water, occurs generally at much lower concentrations
and is much less widespread as a contaminant than chloroform and
related trihalogenated compounds. Benzene has not been clearly
established to be carcinogenic in experimental animals; although
epidemiologic and clinical studies, largely of occupational exposures,
suggest its possible carcinogenicity. Certain haloethers, chloro-
olefins, and polynuclear hydrocarbons have been demonstrated to be
carcinogenic in laboratory animals and have been identified in drinking
water. To the very limited extent that they have been measured,
the data available to the Study Group indicate that the potential
human dosage of these compounds from ingestion of drinking water will
generally be considerably Less, in absolute quantities as well as
relative to experimentally carcinogenic doses in laboratory animals,
than for chloroform. However, the Study Group notes the existence
of local situations in which this generalization would not apply.
The Study Group felt that for all the compounds reviewed, the
carcinogenicity data and experimental designs were generally either
inappropriate or below the standard of current toxicological practice
and protocols for carcinogenicity testing. Additional well-designed
experimental studies to determine the carcinogenicity of lifetime
exposures by ingestion are sorely needed.
Data from epidemiologic studies on the contaminants of primary
concern to the Study Group are very limited and the designs of the
studies are generally inadequate for a conclusive assessment of
health risk. The recent studies alleging an association of high
cancer incidence in New Orleans with consumption of contaminated
drinking water are considered by the Study Group to be hypothesis-
formulating studies, but should not be interpreted to have established
a causal relationship. Numerous other variables might explain the
apparent associations. Indeed, the experimental toxicology studies
suggest that, if there were a carcinogenic risk, increased liver cancer
would be a probable finding. In fact, however, this was not revealed
by the epidemiologic studies. Recent studies of 79 cities in the
National Organics Reconnaissance Survey have identified many water
Vlll
-------
supplies in which some suspect halogenated organic compounds occur
in higher concentrations than in New Orleans. It should be possible,
therefore, to test further the hypothesis formulated for New Orleans
water and cancer in other cities that have a completely different
set of variables from those of New Orleans.
In summary: Based upon recent, reasonably extensive, water
quality data for many U.S. water supplies and on extremely limited
data from experimental carcinogenesis studies, the Study Group
concludes that there may be some cancer risk associated with consumption
of chloroform in drinking water. The level of risk, estimated
from consideration of the worst case and for the expected cancer
site for chloroform (the liver) might be extrapolated to account
for up to 40% of the observed liver cancer incidence rate. A more
reasonable assumption, based upon current water quality data which
show much lower levels than the worst case in the majority of U.S.
drinking water supplies, would place the risk of hepatic cancer
much lower and possibly nil. Further, it is emphasized that both
the experimental carcinogenicity data and the mathematical and
biological extrapolation principles used to arrive at the upper
estimate of risk are extremely tenuous. Epidemiologic studies
do not, thus far, support the conclusion of an increased risk of
liver cancer; although hypothesisformulating studies in southern
Louisiana suggest the possibility of an association with contaminated
water and overall high cancer incidence. Critical, definitive
tests of this hypothesis have not been conducted. Although some
other organic contaminants contained in the charge to the Study
Group have carcinogenic potential, the cancer risk to man is judged to
be minor because of their low concentration and/or infrequent occurrences
in drinking water.
IX
-------
-------
CONTENTS
I. INTRODUCTION 1
II. ASSESSMENT OF CONTAMINANT MONITORING 5
A. Adequacy of Analytical Methods 5
B. Amounts and Distribution of Contaminants 9
III. ASSESSMENT OF EPIDEMIOLOGIC STUDIES 15
IV. .ASSESSMENT OF EXPERIMENTAL CARCINOGENICITY AND OTHER
TOXICITY STUDIES 19
A. High Priority Compounds Identified in the Charge
to the Study Group 19
1. Chloroform 19
2. Carbon Tetrachloride 22
3. Chloroethers 24
4. Benzene 25
B. Other Potentially Hazardous Compounds 27
1. Phthalic Anhydride and Phthalate Esters 27
2. Octadecane and Cn-C^ Hydrocarbons 28
3. Polynuclear Aromatic Hydrocarbons and Heterocyclics .. 30
4. Halogenated Methanes 31
5. Chloro-Olefins 33
V. RISK ESTIMATION 35
VI. APPENDICES 45
A. Summaries of Epidemiologic Studies Evaluated 45
B. Phthalic Anhydride and Phthalate Esters 51
C. Chloro-01efins--Toxicity Summaries 57
-------
I. INTRODUCTION
On February 3, 1975, the Assistant Administrator for Water and
Hazardous Materials (EPA) asked the Chairman of the Science Advisory
Board (SAB) of the EPA to give a best judgment of the degree of health
risk posed by exposure to certain organic compounds that had been
identified as contaminants of drinking water supplies in certain areas
of the U.S. The purpose of this assessment was to provide information
useful to the Administrator of the EPA with respect to the promulgation
of standards or other actions as provided for in the Safe Drinking Water
Act (Public Law 93-523, December 16, 1974). In response to this
request the Chairman of the Science Advisory Board established an
Ad Hoc Study Group on Organics in Drinking Water under the auspices
of" "the Hazardous Materials Advisory Committee of the SAB. Members
of the Ad Hoc Study Group were appointed in late February and early March
of 1975, and, in a memo dated March 12, 1975, the Chairman of the
Hazardous Materials Advisory Committee charged the Ad Hoc Study Group
to consider the potential carcinogenic or other health risk from
ingestion of those chemicals that are present in drinking water
(Attachment A).
Four compounds were identified in the charge as deserving of
primary consideration: benzene, carbon tetrachloride, bis (2-chloroethyl)
ether, and chloroform. Three other compounds were cited in the charge
for consideration: B-chloroethylmethylether, octadecane, and phthalic
anhydride. The first four compounds had been identified in drinking
water of New Orleans and were known, suspected, or alleged to have
carcinogenic action in man or experimental animals. Additionally,
the Study Group was asked to review a much larger list of organic
chemicals that had been found in drinking water, usually at parts
per billion (yg/1) or lower levels, and to comment as to whether any
of them represented a greater health hazard than the seven compounds
cited above. The importance of supplying a report by a due date of
May 1, 1975, was stressed in the charge to the Study Group.
The Study Group held its first meeting March 24 and 25. Informational
materials were distributed by Agency staff and presentations were made
by several Agency scientists and representatives, representatives of the
Environmental Defense Fund, and other interested persons.
In order to meet the due date, within little more than a month's
time, the Study Group necessarily had to limit and to focus its
consideration of the problem. It should be recognized, therefore,
-------
that the report that follows cannot be considered complete and
comprehensive in all respects. It is a first assessment and
represents the best judgments that could be reached by the Study
Group under the constraints of time and information available for
review.
The Study Group took note of the fact that the Safe Drinking
Water Act calls for the Administrator to transmit to the Congress,
not later than six months after the date of enactment, the initial
results of a study relating particularly to contamination of water
supplies by chemicals or other substances suspected of being
carcinogenic (Sec 1442 (a)(9) p. 23). For this reason, and because
recent intense public interest in chemical contamination of drinking
water had centered around alleged carcinogenic risk, the Study Group
felt that their focus should relate primarily to assessment of
information related to the possible health risk of cancer. Although,
the Study Group recognized that: other types of toxic action might
be equally or even more important than carcinogenicity, time did
not permit a comprehensive review of all the toxicological literature.
The Study Group acknowledges the inherent weakness of such a limited
approach, but it also noted that the Safe Drinking Water Act calls
for an intensive review of potential adverse effects of contaminants
to be conducted by the National Academy of Sciences over a two year
period (Sec. 1412 (a) (1-6) pp. 4 and 5).
At its first meeting, the Study Group considered the compounds
identified specifically in the charge, and concluded that there was
no compelling reason that octadecane and phthalic anhydride should
be singled out for priority consideration as potential health risks,
particularly in view of the absence of any quantitative data on
their occurrence in drinking water and because there was no apparent
evidence to place them in the suspect carcinogen class. A list of
162 "Organic Compounds Identified in Drinking Water in the United
States (as of 11/25/74)" was reviewed and discussed by the Study
Group. No quantitative data with respect to concentrations or
frequencies of occurrence were available, and the Study Group found
any attempt to assign priorities for consideration of potential
health risks extremely difficult. (Subsequently, an updated (March
15, 1975) list containing 187 compounds with notations regarding
concentrations and numbers of locations where compounds were
identified was supplied by the Water Supply Research Laboratory, EPA,
Cincinnati, Ohio.) Nevertheless, certain compounds or groups of
compounds (in addition to those cited in the charge) were identified
as deserving of some attention and various members of the Study
Group agreed to give them consideration. These included the
following classes of chemicals: chloro-olefins, chloroethers,
halogenated methanes, C8-C30 hydrocarbons, phthalates, and polynuclear
aromatics. Concern for these compounds arose from members' knowledge of
their ubiquitous distribution or because of their potential for serious
-------
toxic action. The Study Group took note of the fact that, as
stated in the charge, pesticides, asbestos, and inorganic chemicals
were being evaluated separately. The Study Group did not address in
detail the question of sources of organic contaminants.
It was the conclusion of the Study Group at its first meeting
that a first assessment of health risk associated with contamination
of drinking water could best be met by evaluating the data and
information in three primary areas: contaminant chemistry, epidemi-
ologic studies, and carcinogenic!ty (or other toxic actions) of
individual compounds or groups of compounds. Individual members
agreed to undertake related reviews and writing assignments to be
considered subsequently by the Study Group for their report to the
Hazardous Materials Advisory Committee. Discussion and integration
of a report occurred at a meeting on April 17 and 18, 1975. At
that meeting a memorandum from the Director of the Water Supply
Research Laboratoiy, EPA (dated April 15) with an attached table
showing the concentrations of six volatile halogenated organic
compounds in the drinking water of 79 of the 80 cities in the
National Organics Reconnaissance Survey was made available to
the Study Group. These data were also considered in the preparation
of the final report, which was reviewed and approved at a meeting
on April 25, 1975.
-------
-------
II. ASSESSMENT OF CONTAMINANT MONITORING
A. ADEQUACY OF ANALYTICAL METHODS
Two general comments concerning the current status of analytical
procedures preface this discussion of the adequacy of analytical
methods for the determination of trace levels of organic chemicals
in water. Firstly, no single analytical scheme, even disregarding
the cost and complexity factors, is available for the complete
determination of all the organics which might be expected to be
present in drinking waters. Secondly, the combination of the various
extraction, separation, and detection procedures fails to present a
complete profile of the organic materials which are present in any
water environment.
The extensive research over the past decade using a variety
of analytical approaches has been only partially successful in
establishing the desired profile. In a recent report based on a
literature survey (1) one hundred sixty two different organic
chemicals have been identified from various drinking waters.
Realistically, these represent at best 101 of the total weight of
organics present in any one drinking water supply. The more probable
figure, based on a comparison of the dissolved organic carbon
(2,3,12,13) and the summation of the concentrations of the identified
components from various drinking waters, is 21 or less. Most of the
unidentified components are probably of natural origin and, as such,
they may not represent a controllable pollution problem. However,
the presence of these materials, most of which are probably soluble
humic substances, tends to complicate the development of simple and
accurate analytical schemes. In addition, some of these unknown
materials may either be toxic or play a role, favorable or unfavorable,
in the toxicity of the controllable synthetic contaminants.
The continuing evolution of analytical methodology for the
characterization of all the organics in water is evident. Many
decades of basic research will be necessary before a reasonably
complete profile is established. The vacuum of knowledge in this
area is understandable but discouraging from both the analytical and
toxicological viewpoints.
Yet much of the recent progress is cause for encouragement.
For example, many organic chemicals present in drinking water can
be accurately assayed using sophisticated, and in some cases, even
routine and relatively simple monitor schemes. Indeed, the
sensitivity and the reliability of methods for some classes of
compounds, such as the chlorinated hydrocarbon pesticides, far exceed
-------
the requirements of the current drinking water standards. The
profile of the synthetic organics found in water is growing. Reliable
quantification data are emerging. Simpler and less expensive
analytical schemes are evolving. More efficient and selective
extraction schemes are being investigated. Highly effective separation
schemes are being studied.
A review of these developments suggests that existing analytical
methodology is satisfactory for the accurate (see Accuracy section, page 8)
monitoring of the high priority contaminants discussed in Section TV
of this report. Most of these contaminants are volatile so that the
inert gas stripping procedures (4-8) are applicable for extracting
the organics from the water. Sorption and desorption of the stripped
components is not unduly complicated. Separation is achieved by
low resolution gas chromatography. Detection using halogen-specific
detectors is both sensitive and reproducible for the volatile
halogenated materials. For other organic components, the flame
ionization detector is applicable and sufficiently sensitive. In
some cases, charcoal adsorption, solvent extraction or resin sorption
procedures must be used. Specific compounds or groups of compounds
reviewed by this Study Group are discussed below.
Carbon Tetrachloride - This material can be assayed in water at
less than one yg/1 using the Bellar (4) procedure of inert gas
stripping with sorption on a poly(p-2,6-diphenylphenylene) oxide
adsorbant. The sorbed CCl^ is then heat desorbed directly onto an
analytical gas chromatography column where separation from the other
sorbed and de-gassed components is accomplished. Detection is
routinely achieved by halogen-specific detectors. Other similar
and equally applicable techniques are also described in the literature
(4-8). All volatile halogenated components are readily measured using
this procedure provided adequate GC resolution is available.
Even direct aqueous injection which eliminates the need for the
gas stripping, sorption, and de-sorption is available and sufficently
sensitive for many applications (9). In many cases solvent extraction
with a solvent insensitive to halogen-specific detectors is also
useful.
Chloroform - Procedures are identical to those discussed for CCl^.
Comparable detection limits are achieved.
Other Halogenated Cj.-^ Hydrocarbons - Most all these components
can be measured simultaneously with the CCl^ and HCC13 (4).
Depending on the complexity of the mixture of halogenated compounds,
higher efficiency GC columns may be necessary to separate the
components. Detection limits of less than yg/1 are achieved.
-------
g-chloroethers - These materials can readily be incorporated
into the standard inert gas stripping procedure. Comparable
sensitivity should be attained. In addition, conventional solvent
extraction methods (10) are also available. Both charcoal
adsorption (5) and resin sorption methods (11) are useful for
these chloroethers.
Hexachlorobutadiene - This material can also be measured by
the inert gas stripping procedures. However, larger volumes of
stripping gas and elevated temperatures are necessary. Under these
conditions, a more efficient adsorbant (4) is necessary to retain
analytical accuracy for the more volatile constituents. A variety
of alternate methods such as solvent extraction, charcoal adsorption,
and resin sorption are available. In general, any procedure employed
for the chlorinated pesticides is applicable and nanogram per liter
detection limits are expected.
Benzene - The inert gas stripping procedure is adequate for benzene.
Detection is accomplished using a flame ionization unit. Less than
one yg/1 detection limits are attained.
Octadecane and other Cg~^30 Hydrocarbons - A modified Bellar (4)
procedure using flame ionization detection is satisfactory. For
octadecane and the higher molecular weight hydrocarbons up to C24
a larger water volume and more stripping gas is required. It has
been suggested that a recycle system (5) be used to prevent
artifacts. The charcoal adsorption and resin adsorption techniques
are also available for large water volumes. Solvent extraction
is useful at higher concentrations and for the less volatile hydro-
carbons. Less than one yg/1 limits are achieved.
Phthalate Esters and Phthalic Anhydride - Solvent extraction (21).
charcoal adsorption (20), and resin sorption (11) methods are well
established. Less than one ug/1 detection limits are achieved.
Polynuclear Aromatics - The polynuclear aromatics or poly-
aromatic hydrocarbons have usually not been measured in the
U.S.A. water assays, despite the fact that analytical methodology
is available. Thin layer chromatographic separation and ultraviolet
detection have been used successfully in Europe (17). High temperature
gas chromatography used in conjunction with solvent extraction or
resin sorption is also feasible. Less than one yg/1 detection limits
are achieved.
-------
In conclusion, .analytical methods for the accurate monitoring
of the high priority chemicals mentioned in this report are adequate.
This conclusion may be extrapolated to include most synthetic
organic chemicals which might be present in drinking water. However,
current schemes are generally too slow, too complicated, and too
expensive for routine applications by water purveyor laboratories.
Accuracy - Accuracy is the extent of agreement of the reported concentration
with the true concentration. For trace analyses of organics in
drinking water this agreement cannot be exactly established. It
is an estimated value based on the critical assessment of the tests
of the variables in the analytical scheme.
In the concentration range for the high priority chemicals
being considered in this study, the reported amounts are
estimated to be within a factor of ten of the true amount. This
analytical accuracy is considered to be adequate for use in evaluating
health risks. Other factors in the health risk evaluations exceed
the uncertainty in the amounts ingested by drinking contaminated
water.
-------
B. AMOUNTS AND DISTRIBUTION
Extensive information on contamination is available only for
the chlorinated hydrocarbon pesticides such as dieldrin, DDT, etc.
For other classes of chemicals and some of the specific high priority
chemicals mentioned in this report, sufficient data are not available.
Most of the research to date has been directed toward the development
of analytical schemes for organics in water and the establishment of
the profile of organics in water. Extensive quantitative information
is frequently missing or questionable. For many contaminants
sampling frequency is not adequate to establish probable fluctuations.
In addition, the number of sampling sites is very limited and often
includes only one water supply. Hence, definitive data on the
distribution of contaminants are not available.
The information in Table I includes judgments and extrapolations
based on very limited available information on distribution, production
levels and uses, and expected persistence. Selected references used
to arrive at the tabular assessments are included.
In summary, the concentrations in drinking water of all of the
priority chemicals that have been measured are less than one part
per million (1 ppm or mg/Iiter). Except for the halogenated C± and
C2 chemicals, the concentrations are generally less than one microgram
per liter.
-------
-------
o
J-l
o
.
en
•
|~~)
0
45
p
£l
•H
to
r*
03
d
•H
^
o3
p
d
o
U
13
0
P
O
0
rH
0
CO
L|_j
O
d
0
•H
^_j
^3
43
• H
5H
4_)
CO
• H
13
13 •
0 to
p 0
0 -H
0 rH
O . O t
CJ>
CO
ro 0
P
to o3
P S
S bO
O d
S T-l
B ,X
d
0 -H
PTJ
t
I — 1
3
PQ
§
0
U
0
0
<4H
(S
03
£j
O
•H
P P
0 43
U -H
0 P
C3-, co
•H
Q
^ — ^
PQ
rv
t/) <3
-P
S, — |
*-^
O bQ
S J-l
K?
d
•H
CO
P
CO
.s
p
d
o
U
^f-
(XI
»^J- «^J- «r^J-
(X] (XI "(X)
^
45 oj P
O d 0
03 CD O
5-i bC 5-i
P S 0 0
0 H i— i rH
£•8:3 f
O 5-i fn (xi
43 O 0 v — '
Si rH 45 CO
rt 45 P -H
u O o PP
T^-
(XI
rH rH CXI
*2 ^2 *2
O O O
i— 1 rH rH
«
^
PJ ^- to
(XI rH (XI
LO LO rH
^
bOCD >
•H LO O
45 rH
rH
rH rH •
V ( O
V
CO
d
0 0
^ O rrj
5n tO -H
03 5n 5-i
O 0 13
O P >^
13 0 d
X o3
45 0
P O
O OS -H
0~lrH rH
U a3 03
1 p p
03
[*•--
i-H
»v
[-~^
rH
*
rH
r^
bO
•H
r*j
rH
1
rH
O
O
•
0
T3
CO
U
•H
P
qrt
g
0
5-i
03
0
rH
U
^
i— 1
O
0
P 0
•H 03
0 CO
45 0
p p
•H
"4H CO
O
P
0 d
O 0
d -H
0 U
CO -H
0 MH
5-1 <4H
ft 3
to
0 d
p
0
0 CD
O !2
_ (-^
to
i — \ o\°
"^ O
CN! i — 1
p
to to
5-1 0
0 5-i
P 03
03 0
S d
00 0
•H P
^ ^
d o
•H p
"r3 *~d
0
rH 13
1-1 ^
0
MH 5n
O
0
CO r-<
-P 03
to
0 CO
P 0
P rH
o3 o3
45 >
P
P
to d
d 0
03 O
0 5-i
E 0
PH
d
O
•H
P ' — ^
43 ^
•H -P
p ro
to d
o\o ^_>
o d
0 0
rH U
ffl
•
0
13
K
d
0
0
Q
CD
^>
03
CO
0
p
rrt
^
• H
P
CO
0
bO
• H
45
5-i
0
3^
0
rH
13
0
rH
C5
CO
g
43
f-j
U
13
§3
0
13
•H
^_i
O
rH
f"{
U
03
^H
P
0
P
d
JB
^
03
U
P
ft
0
U
X
0
to
d
0
43
!H
03
0
0
5-H
13
^/^
13
0
P
03
d
0
bo
O
rH
03
^J^j
CN
C_J
13
03
rH
rH
rH
03
<4H
O
d
0
•H
P
to
to
1
rH
L)
d
•H
•
rr3 o i
o3 3
0
5n tO
ft 0
to d
CD 0
13 N
•H 0
5 d
0
rH
43 13
03 0
43 P
O 03
5-1 rH
i C-^
to i— I
•H 03
d 0
0 45
• H P
P
3 13
•H H
^_<
P 0
to d
•H 0
13 NI tO
d !-i
to 0 0
P 43 P
•H o3
O £
13 P
0 bO
P 5n d
fH 0 -H
O f i i ^/i
ft 0 R
0 5-i -H
0 13
X 5n
rH 0 X
c TO
0 CO g
3 d
tr 1 d
CD 3 -H
5-i rH
<-H O 13
U 0
0 d 5-1
0 O O
43 -H ft
p 0
P 3 5-i
O 43
d -H d
5-1 0
CO P 0
o3 to 43
45 -H
13 0
0 >
d 13 03
0 d 45
N O3
d 45
0 P U
S t *""i
oj &
5-i U
0 0
S H P
o
•
CO
13
r3
O
g
O
U
rH
^3
i3
• H
^>
•H
13
d
•H
<4H
O
to
d
0
•H
P
03
P
d
0
u
d
5
0
0
45
P
MH
0
d
5
•H
P
cH
§
to
oS
0
03
to
p
S
o
13
0
P
to
•H
rH
0
•0
11
-------
REFERENCES
1. USEPA, Water Supply Research Laboratory, Organic Compounds
Identified in U.S. Drinking Waters and Their Toxicity,
Cincinnati, Ohio. December 9, 1974.
2. Leenheer, J. A., et al., J. Res. U.S. Geol. Survey, 2,
361 (1974).
3. Symons, J. M. and Stevens, A. A., private communication, CCE-m
Results on finished Kater From Surf;^.- '-n^rvr--. ,7ur:o <:!'7J to
Present ;'19'74?). ^:,r: >n,;; i '•'•'•'">:<,
Mioure, J. i . and i.;;-. • sci., 'i. ;.„ . . . :
559 (1973). ~ "
8. Saunders, R. A., c< al, J, i'.'ji:1:^, . :,TNS_ ,-ijec(_rom., i;; r-re,1;^-.
(1974j.
9. Harris, L. E., et al., Aoai. r.^eM ,, 4! , 1912 (1974).
10. Oressman, R. C. and Moi-a.rren, .-;. F., Paper presented at 2nd
Annual Water Quality Yechnol. Conf. of the Amer. Water Works
.Association, Dallas', Oec. (1974).
11. Junk, G. A., et al., J. Chromatogr., 99, 745 (1974).
12. Malcolm, R, L. and Leenheer, J, A., The Usefulness of Org.
Carbon Parameters in Water Quality Investigations. USGS.
13. Stevens, A. A. and Symons, J. M., Paper presented at the Amer.
Water Works Technol. Conf., Dec. 1973. Cincinnati USEPA.
14. Junk, G. A., private communication, Ames Lab I.S.U., Ames,
Iowa.
15. Robeck, G., private communication, USEPA, Cincinnati, Ohio.
16. Symons, J. M., unpublished report, USEPA, Cincinnati, page 7 -
B.C.J. Zoeteman data for Rhide River, Oct. 29, 1974.
12
-------
17. Andelman, J. B, and Suess, M. J., Bull. World Health
Organization, 43, 479 (1970) - see p. 487 for summary of
Borneff and Kunte data.
17a. Andelman, J. B. and Snodgrass, J. E., CRC Critical Reviews in
Environmental Control, (No. 1), CRC Press, Cleveland (1974).
18. Dressman, R. C. and McFarren, E. F., 2nd Annual Water Quality
Conf. of the Amer. Water Works Assoc., Texas, Dec. 1974.
19. Kleopfer, R. D. and Fairless, B. J., Environ. Sci. Techno1. 6,
1063 (1972).
20. USEPA, Region VI, Draft Analytical Report New Orleans Area
Water Supply Study, Dallas, Texas, Nov. 1974.
21. Hites, R. A., Environ. Health Perspectives, !_, 17 (1973).
22. Saalfedd, F. A., private communication, Naval Res. Lab., Wash., D.C.
23. Junk, G. A., Environ. Sci. Technol., 8_, 1100 (1974).
24. USEPA, Water Supply Research Laboratoiy, Data from National
Organics Reconnaissance, attachment to memo from Director,
dated April 15, 1975.
13
-------
-------
III. ASSESSMENT OF EPIDEMIOLOGIC STUDIES
Six epidemiologic reports have been identified which permit quanti-
tative measures of associations between frequency of malignant disease
in humans and exposure to substances found in drinking water.
Two studies of occupational exposure to benzene were carried out to
investigate a long-standing clinical impression, supported by many studies
of individual patients with leukemia following such exposure, that the
exposure was leukemogenic. The study of Ishimaru et al. (1) gives sub-
stantial support to an hypothesis that certain occupations involving
benzene exposure are causally associated with leukemia. Usual confound-
ing factors of residence, age, sex, and race are appropriately considered.
The occupational exposure, however, was non-specific, with patients
exposed to a large variety of substances, many of which may be carcino-
genic. There is no quantitative estimate of the exposure level, though
it may be supposed that many of these patients were exposed to air con-
centrations of the order of 1 part per million over years to decades.
Skin exposure and ingestion may also be supposed to have been substantial.
A study by Vigliani and Saita (2) of occupational benzene exposure
is based on generally inadequate observations, with no actual enumera-
tion of the exposed population and no analytic control of probable con-
founding variables. As with the Ishimaru et al. study, no organic com-
pound was specifically incriminated. This study is consistent with the
reports of individual cases but adds little to such reports.
Four studies (3-6) were designed to investigate the relationships be-
tween the previous observations that (a) Louisiana, in total, and New
Orleans, in particular, have had high frequencies (mortality for both
state and city, incidence for New Orleans) of all cancer and certain
cancer sites in white males for many years, and (b) recent analytical
methods have made it possible to identify extremely low concentrations of
certain organic carcinogens and other compounds in certain drinking waters,
including waters supplied to certain Louisiana counties along the
Mississippi River. These two observations alone raise the possibility of
a causal association between water-borne carcinogens and increased cancer
frequency.
In the absence of experimental manipulation of a suspected causal
experience, a final causal interpretation must ultimately be made on a
judgmental basis, though certain types of observations greatly strength-
en the presumption. When a cause-effect relation is suggested by pre-
liminary observations, a classical approach is to propose a specific
quantitative hypothesis based on the preliminary observations and to
test this hypothesis in another body of data. In the investigations of
Louisiana counties no specific quantitative hypothesis was stated, but
15
-------
the general hypothesis derived from the observations might be that
populations served by water supplies containing demonstrated carcinogens
would experience higher total cancer mortality than would populations
with supplies in which carcinogens could not be demonstrated by the
same analytic methods. Since the preliminary observations were made
in New Orleans, an appropriate test of the hypothesis would be made
elsewhere. It will be a matter of judgment whether the recent studies
of mortality by county effectively involve a body of data different
from previously available rates for the state of Louisiana and the city
of New Orleans in which the hypothesis arose. Technically the analysis
shows positive associations eAren when New Orleans is excluded. The
water variable, however, appears to be highly correlated with distance
from New Orleans. Insofar as the final associations simply say that
cancer mortality decreases with distance i'roiu New Orleans, the new
analyses differ little from the preliminary observation.
In the absence of a specific quantitatjvc hypoLhesis, strong
evidence may still be obtained by test.if.,.-; ;. ju.vlitative hypothesis
specifying direction of association, it .1,-, iivc clear in the present
study what mortality rates wore £ pj~ior_j_ exported to vary in what
direction. There is a suggestion that livci cancv" and leukemia might
be increased by the carcinogens observed and that lung cancer might be
unaffected. These relationships were not supported by the findings.
On the other hand, hypotheses relating to increases in total cancer
might have been stated a priori, and this association \vas found. The
analysis by De Rouen and" Diem (6) demonstrates some positive and some
negative associations between water quality and a variety of cancer
sites. Without a clear statement of prior hypothesis, the possibility
exists of seeking sites that would tend to support a non-specific
hypothesis.
Individually weak tests of hypotheses may, under some circumstances,
reinforce one another. This is true when methodologies are sufficiently
different to make a common systematic bias unlikely. It will be clear
that the several studies of the New Orleans data do not reinforce one
another in this sense, since all studies involve analysis of the same
set of mortality data by county. There is reinforcement in the sense
that the original relationships persist, though weakened, after consi-
dering in analysis additional confounding variables of elevation and
latitude (both of vvhich measure distance from New Orleans). Further
details of the studies evaluated may be found in Appendix A.
Comment:
It is the judgment of the Study Group that presently available
epidemiologic data strengthen the clinical impression of a leukemogenic
effect of benzene or of substances commonly associated with benzene in
occupational exposures.
16
-------
It is the judgment of the Study Group that epidemiologic data
associating cancer mortality with drinking water quality can be inter-
preted only as hypothesis-formulating studies. It is emphasized, how-
ever, that statistically significant associations have been demonstrated
between cancer mortality and several variables, including source of
water, elevation, latitude, rural-urban characteristic, and income. It
may be assumed that other environmental factors, probably including com-
ponents of diet, occupational exposure, medication, and household exposures,
are responsible for these associations. It is possible that water and air
pollutants also contribute. Continuing epidemiologic, toxicologic, and
basic science studies will be required to clarify the associations.
REFERENCES
1. Ishimaru, T., Okada, H., Tomiyasu, T., Tuschimoto, T., Hoshino,
T., and Ichimaru, M., AM. J. Epid., 93, 157 (1971).
2. Vigliani, E.G. and Saita, G., N. Eng. J. Med., 271, 872 (1964).
3. Harris, R. H., The implications of cancer-causing substances in
Mississippi River water. Unpublished. 1974.
4. Page, T., Harris, R. H., and Epstein, S. S., Relation between
cancer mortality and drinking water in Louisiana, Unpublished.
1975.
5. Tarone, R. E. and Gart, J. J.. Review of "The implications of
cancer-causing substances in Mississippi River water" by Harris,
R. H., Unpublished. 1975.
6. De Rouen, T. A. and Diem, J. E., Ethnic, geographic differences
in cancer mortality in Louisana. Unpublished- 1975.
17
-------
IV. ASSESSMENT OF EXPERIMENTAL CARCINOGENICITY
OR OTHER TOXICITY STUDIES
A. HIGH PRIORITY COMPOUNDS IDENTIFIED IN CHARGE
1. CHLOROFORM
Information supplied to the Study Group by the Water Supply
Research Laboratory of the EPA indicates that chloroform (CHClj)
has been identified in drinking water supplies in 79 locations of
the National Organics Reconnaissance Survey in the U.S. (1). The
concentrations were stated to range between 0.3—311 yg/liter.
The most useful experimental study of possible tumorigenic action
of chloroform is that reported by Eschenbrenner and Miller in 1944 (2).
Groups of 5 male and 5 female A-strain mice were given various doses of
chloroform in olive oil by stomach tube every 4 days for a total of 30
doses, and were examined for hepatomas 1 month after the last dose, when
the animals were 8 months old. Twenty-four hours prior to necropsy,
animals in each dosage group were given an additional dose of chloroform
to determine the relation between dose and the occurrence of liver and
kidney necrosis and incidence of hepatomas. The results of this study
are summarized in the following table.
Dose (ml CHCl3/kg)a
Observation Sex 1.6 O.TTOT? 0.2 0.1 _0
Liver Necrosis F + + + 0 0 0
M + + + 0 0 0
Kidney Necrosis F 0 0 0 0 0 0
M + + + + +0
Deaths F 5/5b 1/5 2/5 0/5 0/5 0/5
M 5/5 5/5 5/5 2/5 0/5 0/5
Hepatomas in F 4/4 3/3 0/5 0/5 0/5
surviving animals M — — — 0/3 0/5 0/5
receiving 30 doses
a. Cubic centimeters chloroform per kg body weight per dose (30 doses).
b. Numerators=positive occurrences; denominators=animals observed.
19
-------
These investigators noted that the dose-end point for hepatoma induction
coincided with the end point for liver necrosis, with neither effect
being observed at the two lowest dosages. Special stains used to detect
cirrhosis also revealed moderately cirrhotic livers only in those mice
that received repeated doses of greater than 0.2 ml/kg of chloroform.
It is of interest that the most sensitive index of injury in male (but
not female) mice was the occurrence of kidney necrosis. This peculiar
sensitivity of male mice to chloroform nephrotoxicity has been noted by
others (3). Eschenbrenner and Miller (2) conclude from their experiments
that hepatoma induction by chloroform occurred only with dosages that
were sufficient to cause the acute response of hepatic necrosis, and
since male mice died from the nephrotoxic action of chloroform at less
than hepatotoxic doses, it was not possible to demonstrate hepatoma pro-
duction in the males. They suggest, however, that one cannot rule out
the possibility of hepatoma production with other dosage-time schedules.
It is worthy of note that a working group of the International Agency for
Research on Cancer (IARC) (4) concluded that for carbon tetrachloride
repeated liver necrosis and chronic regeneration were not necessary
for tumor induction by that compound. In one other study by Rudali (5),
hepatomas were observed in 3 of 5 surviving mice from a total of 24 that
were given 0.1 ml of an oily solution of chloroform intragastrically
twice weekly for 6 to 24 months. A Working Group of the IRAC concluded
that "An assessment of the carcinogenicity of chloroform awaits further
experimental evidence" (6).
Fetotoxicity and teratogenic effects of chloroform in rats exposed
by inhalation to 30—300 ppm have been reported (7); however, oral admini-
stration of 20—126 mg/kg/day to pregnant rats and 20—50 mg/kg/day to
pregnant rabbits during the period of embryonic organogenesis did not
result in fetotoxic or teratogenic action (8). The authors suggest that
differences in distribution with different routes of administration might
account for the differences in effects on the embryos, but blood and
tissue levels of chloroform were not determined.
Comment:
The experimental studies discussed above suggest a tumorigenic
action of chloroform; but because of the limited nature of these
studies with respect to numbers of animals, single species, duration
of exposure, etc., extrapolation of the data to the practical situation
of chloroform-contaminated drinking water and its implications to
human health is extremely tenuous. Certain aspects of problems inherent
in the experimental evaluation of carcinogenicity of chemicals are
apparent, however, and are worthy of note. Firstly, the failure to
demonstrate hepatoma production in male mice dosed with chloroform,
because of the overriding nephrotoxicity, is a clear example of the
limitations posed by carcinogenesis screening tests on a single species,
strain or sex of experimental animals. Secondly, the apparent pre-
requisite for production of necrosis and regeneration before hepatoma
production becomes msaiifest raises the question as to whether certain
20
-------
theories relating to chemical carcinogenesis (e.g., additivity of dose-
effect, specificity of molecular targets, one-hit concept) apply in
the case of chloroform. Furthermore, questions can be raised as to
whether the possibility of other types of chronic disease (e.g.,
cirrhosis) might not be of equal or greater concern than possible
cancer production, and whether epidemiological or clinical studies
utilizing sensitive tests for evidence of other types of liver
injury might be meaningful.
In summary, very limited experimental data place chloroform in
the category of a suspect carcinogen to man. Prudence demands
continued research on this problem in view of the fact that chloroform
appears to be an ubiquitous contaminant of drinking water in which
chlorination procedures are used (9) and because the experimental data
on the potency and mechanism of chloroform-induced hepatomas in animals
is currently inadequate to evaluate fully its implication to human health.
REFERENCES
1. Data from National Organics Reconnaissance Survey, April 15, 1975.
Water Supply Research Laboratory, EPA.
2. Eschenbrenner, A.B. and Miller, E., J. Nat. Cancer Insti., 5,
251 Q945). " ~
3. Shubik, P. and Ritchie, A.C., Sci., 117, 285 (1953).
4. IARC Monographs on the Evaluation of Carcinogenic Risk of
Chemicals to Man, 1, 53., International Agency for Research
on Cancer, Lyon, France (1972).
5. Rudali, G., UICC Monograph Series, 1_, 138 (1967).
6. IARC Monographs on the Evaluation of Carcinogenic Risk of
Chemicals to Man, 1, 61, International Agency for Research on
Cancer, Lyon, France (1972).
7. Schwetz, B.A., Leong, B.K.J., and Gehring, P.J., Toxicol.
Appl. Pharmacol. _28, 442 (1974).
8. Thompson, D.J., Warner, S.D., and Robinson, V.B., Toxicol. Appl.
Pharmacol. 2£, 348 (1974).
9. Bellar, T.A., Lichtenberg, J.J., and Kroner, R.C., Report No.
EPA-670/4-74-008, November 1974, also: in JAMA, Dec., 703 (1974)
21
-------
2. CARBON TETRACHLORIDE
Carbon tetrachloride is a strongly toxic chemical for the liver
of experimental animals and man. It has been extensively used as an
experimental tool in attempts to elucidate mechanisms underlying
hepatotoxic response. Although the precise cellular and subcellular
mechanisms remain only partially characterized, it seems well estab-
lished that the parent compound is not the toxicologically active
form. Metabolic: conversion to a form (or forms) that can bind coval-
ently to cellular macromolecules is a requisite process in activation
of the compound (1). This factor is thought to contribute to the high
degree of tissue specificity associated with toxicity. Liver possesses
the necessary enzymatic competence to convert carbon tetrachloride to
its active derivative(s) and therefore is affected to the greatest
extent in the tcxic response.
Several experiments have been adequately designed and conducted
to provide meaningful data on carcinogenicity (i.e.,, dosing regimen
has been appropriate to permit survival for sufficiently long periods
for tumors to develop).
Carcinogenicity of carbon tetrachloride has been demonstrated
convincingly in three rodent species (mouse, hamster, and rat), and
equivocally in the rainbow trout. Several features of the response
are of interest to the question of carcinogenic risk to humans in-
gesting contaminated drinking water. Tumors were not induced in any
species in tissues other than liver. Liver tumors were induced
following oral or subcutaneous administration, and by inhalation.
Multiple exposures (usually 1 or 2 doses per week) over prolonged
periods (30 to 70 weeks) were used in these experiments in which
tumors were induced.
For several of these experiments, the published descriptions of
experimental designs were sufficiently detailed to permit calculation
of total doses associated with tumor induction. These figures do not
necessarily suggest minimum effective doses, but give some suggestion
of comparative sensitivity among species.
Following oral dosing (2), mice receiving 30 doses, each of 0.1
ml/kg body weight, over a period of 90 days or more developed liver
tumors at a significant incidence. This regimen provided each animal
with a total dose of 4.77 grams CC1, per kg body weight.
22
-------
Hamsters receiving 30 weekly oral doses of 6.25 - 12.5 yl per
animal and surviving 10 weeks or longer after cessation of dosing
developed liver cell carcinomas (3). In this experiment, each animal
received a total of 0.188 to 0.375 ml or a total dose of 2.98 to
5.96 grams/kg body weight (assuming an average body weight of 100
grams).
Tumors were induced in rats in two experiments in which carbon
tetrachloride was administered by subcutaneous injection. A small
incidence of liver tumors was observed in animals dosed twice weekly
for about 25 weeks with a dose of 2 - 3 ml/kg body weight (4). The
higher level subjected each rat to a total dose of 150 ml (238.35
grams)/kg over that period. In another study (5) of similar design
using several rat strains, higher incidences of liver tumors were in-
duced by a total dose of 91 ml (144.60 grams) per kg body weight.
Two other published experiments report tumor induction by carbon
tetrachloride. Rainbow trout developed a small incidence of hepatomas
when fed diets containing 12,800 ppm CC14 for 20 months (6). Rats that
inhaled CCl^ at an unspecified dose and regimen for 7 months developed
liver carcinomas within 2-12 months later (7).
Collectively, these data indicate that carbon tetrachloride is a
carcinogen for the liver of several animal species.
REFERENCES
1. Recknagel, R.O. Pharmacol. Rev. 19, 145 (1967).
2. Eschenbrenner, A.B. and Miller, E.J., Nat. Cancer Inst.,
4_, 385 (1944).
3. Delia Porta, B., Terracini, B. and Shubik, P., J. Nat.
Cancer Inst., 26_, 855 (1961).
4. Kawasaki, H., Kurume Med. J., 12_, 37 (1965).
5. Reuber, M.D. and Glover, E.L., J. Nat. Cancer Inst.,
44_, 419 (1970).
6. Halver, J.E., U.S. Fish and Widlife Service Research
Report 70, 78 (1967).
7. Costa, A., Weber, G., Bartoloni St. Omer, F., and Campana,
G., Arch. De Vecchi Anat. Pat, 39, 303 (1963).
23
-------
3. CHLOROETHERS
The compounds in this group that should be considered in terms
of health effects and their concentrations in drinking water are:
Name Concentration yg/1
l,2-Bis(chloroethoxy) ethane ?
Bis(2-chloroethyl)ether 0.07-0.42
0-Chloroethylmethylether ?
Bis(2-chlcroisopropyl)ether 0.18-1.58
All four of these compounds are B-chloroethers (C1CR2CR2-0-R) and are
expected to hydrolyze much more slowly than their highly reactive
a-chloro analogs (C1-CR2-0-R). This accounts for their detection in
water and their persistence. Of these four chloroethers only one has
been tested for carcinogenic activity; i.e., bis(2-chloroethyl)ether.
This compound was administered to mice by intragastric feeding from
age seven days until age four weeks. After weaning, the mice were fed
the compound in the diet at a level of 300 ppm for a period of approxi-
mately 13 months. Four of 36 female and 23 of 26 male mice bore hepa-
tomas at the end of the experiment (1). These are combined results
using two strains of mice. In another experiment, the same compound
was given by subcutaneous injection (1 mg/0.05 ml tricaprylin, once
weekly) to female ICR/Ha Swiss mice. Two of 30 animals bore sarcomas
at the injection site. This test was run for the lifespans of the
animals (2). The more meaningful experiment, i.e., the feeding experi-
ment (1)» was carried out at a high dose, 300 ppm in the diet, but the
result suggests potential harmful effects to humans exposed to low
levels of this compound in drinking water. Based on structure-activity
relationships, it is expected that similar findings would be obtained
in mouse or rat feeding experiments with the other three bis(2-chloro)
ethers. To our knowlege these compounds have not yet been tested for
carcinogenic activity.
REFERENCES
1. Innes, J.R.M., Ulland, B.M., Valerio, M.G., Petrucelli, L.,
Fishbein, L., Hart, E.R., and Pallotta, A.J., Bates, R.R.,
Falk, H.L., Gart, J.J., Klein, M., Mitchell, I., and Peters, J.
J. Nat. Cancer Inst., 4_2, 1101 (1969).
2. Van Duuren, B.L., Katz, C., Goldschmidt, B.M., Frenkel, K.,
and Sivak, A., J. Nat. Cancer Inst., 48_, 1431 (1972).
24
-------
4. BENZENE
There is a little biological data relevant to the evaluation of the
carcinogenic and other health risks to man posed by the contamination of
drinking water by benzene.
Reported 11)50 values for benzene administered orally to rats vary
over the range of 0.93 to 5.6 grams/kg body weight (1). In acute
inhalation experiments, death occurred in rats exposed to 33 mg/1
(10,000 ppm) benzene in air for 12.5 to 30 minutes daily for 1 to 17
days (2). Administration of sublethal levels produces blood dyscrasias,
a prominent feature of which is leukopenia.
Benzene is metabolized by experimental animals and man by ring
hydroxylation, and the hydroxylated products are conjugated and excreted.
The patterns of metabolism and excretion vary in different species. In
man exposed to benzene by inhalation, 0.1—0.21 is excreted unchanged in
urine and the remainder as water-soluble metabolites (3). Subjects
inhaling concentrations of 0.35 mg/1 (110 ppm) benzene in air for 5 hours
excreted 29% as phenol, 3% as catechol, and II as hydroquinone, most as
ethereal sulfates (4).
Evidence regarding carcinogenicity of benzene in animals is very
limited. In all the published experiments on animals, observations are
of limited value either because alleged responses were equivocal or
experimental designs had characteristics that make interpretation diffi-
cult. Several investigations suggest the induction of leukemia in mice
by subcutaneous or intramuscular administration of benzene. In one such
experiment (5), albino mice were injected with 1 microliter of benzene
in olive oil weekly for 17-21 weeks (total dose 1 mg/kg body wt).
Leukemia occurred in 8/33 survivors between 4—11 months. However, no
controls were used, and therefore there is no indication of spontaneous
incidence of the malignancy in the mouse strain.
In another experiment (6) of similar design (0.001 ml benzene per
mouse per week) 30% (6/20) of treated animals and 141 29/212 of
controls developed leukemia; the difference in incidence was not statis-
tically significant. Another experiment at the same dose level using
120 mice of four inbred strains was continued for the lifespan of the
animals (7). There was no evidence of induction of leukemia or other
tumors by the treatment.
Induction of subcutaneous sarcomas at the injection site was
reported to result from repeated injections of mice with 0.1 ml benzene
weekly (8).
In man, it is well established that exposure to benzene may result
in damage to the hematopoietic system. Numerous case reports that have
25
-------
been made over the past 45 years suggest that long-term exposure to
benzene may be associated with the development of leukemia. These
consist almost exclusively of reports of occupational exposures in
which very little, if any, information is available on either the level
or duration of exposure. This suggested association is somewhat
strengthened by a recent case control study in Japan; again there is
no indication of level of exposure (see Section III).
Collectively, Che information from animal experiments is insuffi-
cient to demonstrate that benzene is a carcinogen for experimental
animals. The circumstantial evidence :•> m,v -r^-jgesis an
between prolonged exposure u: hpn.;.ri'^ -.- : r" " • •«• :
2. Pumas , B.IV. anc i'.re, . ,,' • .,•_ ,••,, . . • . .. -.f J Lh, lc-_, .;
(1958). ~ " " ........ ~~
5. Srbova,, J., Tei singer, ,•., .'"• : .•"'..:.•••..;'. •;';•', ^. „ __: '<<.. Indus I .
_., 2_, 1 (195C).
4. Teisinger, J., Bergerova-Fii.orova, V., a;.'- \u-Jn\a, J., Praco v .
Lek. £, 175 (1952). ' ~
5. Lignac, G . 0 . E . Krankheit b > br sen . 9_, kt; .;l')32).
6. Kirschbanm, A. and Strong, L.C., Cjncer Res_. 2_, 841 (1942).
7. Amiel, J.L., Rev, franc. StaJ. clin. biuj_. , _5, 198 (1960).
8. Hiraki, K. , Irino, S., arid Miyoshi, [., Ga-'.n, S4, 427 (1903).
26
-------
OTHER POTENTIAL HAZARDOUS COMPOUNDS
1. PHTHALIC ANHYDRIDE AND PHTHA1ATE ESTERS
Phthalic anhydride has been found in only one drinking water
supply in the U.S. Phthalate esters are widely distributed in the
aquatic environment and have been found in many drinking water
supplies.
Based on a survey of the available literature (see Appendix
B), it appears that neither the anhydride nor the esters at the
levels found in drinking waters pose a serious health hazard.
Apparently, the concern for phthalic anhydride is based
on the high production levels and the possibility that it will
appear in the drinking waters as a degradation product of the
phthalate esters. Available data on its occurrence suggest
that this is not the case.
There continues to be some concern for the possible health
effects of phthalate esters. This is apparently based on the follow-
ing information:
9
1) the yearly production levels which approach 1 X 10
pounds;
2) the widespread use and the eventual disposal of polymers
containing as high as 50% by weight of phthalate
esters;
3) the almost universal distribution of phthalates in
the ecosystem, including the aquatic environment;
4) the teratogenicity of phthalates at high dosage levels;
5) a lack of definitive data from chronic exposure studies;
6) the biomagnification and relatively slow biodegradability.
Of these factors, the still growing production records
seem to activate the most concern. The amounts dwarf the
production records for more hazardous chemicals such as the
chlorinated hydrocarbon pesticides. Eventually, this burden
of synthetic phthalates on the environment could pose a problem.
No evidence for carcinogenicity of phthalates has been reported.
Teratogenic effects have been reported for the phthalates but the
dosage levels were orders of magnitude above expected human
exposure. There is an absence of definitive data on subtle
effects from chronic low level exposure.
27
-------
2. OCTADECANE AND Cg -- C3Q HYDRO ARBONS
Octadecane along with many other long chain hydrocarbons, alcohols,
and acids (C^g " ^30^ have been found to accelerate the formation of
skin cancer in mice (1- 3). By themselves the chemicals do not produce
tumors, but when topically applied at the same site with, before, or
after polynuclear aromatic carcinogens, tumor development is speeded.
Sub-carcinogenic doese of polynuclear aromatic hydrocarbons can pro-
duce tumors when applied with hydrocarbons such as dodecane or octadecane.
The amounts of co-carcinogen used in the animal studies involve high
local concentration of the compounds. In experiments of sixty weeks'
duration, a small area of the mouse's skin would have been exposed to
nearly one ml of the co-carcinogen (3).
A possible mode of action of the co-carcinogens is by changing
rates of transport of chemicals through cell membranes (4). The CIQ - €39
hydrocarbons have only been known to function as co-carcinogens in ex-
periments where they were present at the same site and at nearly the
same time as the carcinogen.
Octadecane has occasionally been found in drinking waters at con-
centrations of ^ 0.1 ug/1. The entire range of Cg - C^™ hydrocarbons
which have been identified have been found at a total of< 1 ug/1 (5).
These hydrocarbons are distributed widely in U.S. drinking water -
coming from many sources including automobile exhaust and indigenous
biological materials. Polynuclear aromatic hydrocarbons, some of which
are carcinogenic, have been found in waters worldwide (6) and are also
found in U.S. drinking waters in concentrations of 0.001 to 1 ug/1.
The conditions and low concentrations in drinking waters are totally different
from those of the laboratory experiments. Co-carcinogenicity and proba-
bilities are difficult to judge. However, 'the need to concentrate two
materials present at part per billion levels to some active level at a
specific site seems to suggest minimal health risk from Cg -- C3Q hydro-
carbons present at part per billion levels in drinking waters.
REFERENCES
1. Horton A.W., Denhan, D.T., and Trosset, R.P., Cancer Res.,
17, 758 (1957).
2. Holsti P., Acta path, microbiol. Scand., 46>, 51 (1959).
28
-------
3. Sice J., Toxicol. and Appl. Phanmcol., £ 70 (1966)
4. Horton, A.W. and McClure, D.W., Biochim. Biophys. Acta,
225, 248 (1971).
5. Organic Compounds Identified in Drinking Water in the
United States (March 1s> 1975). (Supplied to the Study
Group by Water Supply Research Laboratory, EPA,
Cincinnati, Ohio, with notations on number of locations
where formed and concentrations).
6. C. E. Zobell, Proceedings of Convention on Prevention
and Control of Oil Spills, American Petroleum Institute,
Washington, B.C., June 1971 (quoted in H. F. Kraybill,
The Distribution of Chemical Carcinogens in Aquatic
Environments). (1974)
29
-------
3. POLYNUCLEAR AROMATIC HYDROCARBONS AND HETEROCYCLICS
Only one polynuclear aromatic hydrocarbon, benzo(a)pyrene, has been
reported in U.S. water supplies. However, these compounds are not routinely
measured in water supply analysis (see Section 11,A). This group of
compounds occurs widely as incomplete combustion products of other organic
materials, and it is expected that they will occur also in water. Many
compounds in this series have been tested for carcinogenic activity in a
variety of animal species and by various routes of administration (1).
Some of them have been shown to be carcinogenic; e.g., in the lung and
skin of test animals (1) and they are generally held responsible for the
occurrence of a higher than expected incidence of skin and lung cancer in
coke oven workers (2). Polynuclear aromatics are probably also responsible,
in part, for the carcinogenicity of cigarette smoke (3). Information on
the carcinogenic effects of polynuclears,resulting from their ingest ion,
implicates them as carcinogens for the stomach (4,5) but not for the liver
(1). Carcinogenesis experiments on mouse skin show that aromatic hydro-
carbons that are non-carcinogenic, e.g., fluoranthene, enhance markedly
the carcinogenic activity of a related hydrocarbon, such as benzo(a)pyrene
(6).
REFERENCES
1. Survey of Compounds Which Have Been Tested for Carcinogenic
Activity. 2nd Edition, U.S.P.H.A., (1951).
2. Doll, R., Fisher R.E.W., Gamman E.J., Gunn, W., Hughes, G.O.,
Tyrer F.H., and Wilson W.,. Brit. J. Ind. Med. 22_, 1 (1965).
3. Smoking and Health. Report to the U.S. Surgeon General,
U.S.P.H.S., Chapter 9 (1964).
4. Stewart, H.L., and Lorenz, E.J., Nat, Cancer. Inst. 2, 193
(1941).
5. Stewart, H.L., and Lorenz, E., J, Nat. Cancer. Inst. _3, 175
(1942).
6. Hoffmann, D., and Wynder, E.L., J. Air. Poll. Control. Assoc.
13, 322 (1963),
30
-------
4. HALOGENATED METHANES
Halogenated methane derivatives are among the most frequently
identified organic compounds in drinking water in the United States.
Members of this class that have been identified in drinking water
supplies by the Water Supply Research Laboratory, EPA (1,2) are
shown below:
No. of locations Concentrations
Compound detected/No, sampled (yg/liter)
Bromodichloromethane 76/79 0.8 -- 116
Bromoform 25/79 1.0 -- 92
Carbon Tetrachloride 10/79 2.0 -- 3
Chloroform 79/79 0.3 -- 311
Dibromochloromethane 70/79 0.4 -- 100
Methyl chloride la —
Methylene chloride 6a 5
Trichlorofluoromethane la —
Reference (1) total number of sampling sites not identified.
Studies conducted by the National Environmental Research Center,
EPA, Cincinnati, Ohio, (3) indicate that the concentrations of the tri-
halogenated methanes (chloroform, bromodichloromethane, and dibromochlor-
omethane) increase from traces or none-detectable in raw river water to
levels up to 100 yg/1 (for chloroform) in finished water. The concentra-
tions increased after each cblorination step in the water treatment
plant. The presence of bromine-containing trihalogenated methanes was
believed due to bromine inpurities in the chlorine. It has been suggested
that, in spite of the apparent formation of halogenated methanes the
total content of organics may be reduced by chlorination procedures (4).
Standard references provide very little information on the toxicity
of these compounds other than for carbon tetrachloride and chloroform
(5-7). Most available toxicological data has been obtained in inhala-
tion exposure studies of relatively short duration. Summaries of
surveys of the literature on the toxicity of methyl chloride, methylene
chloride, and trichlorofluoromethane (8) provided no indication that
these compounds were likely to pose a health risk from long-term ex-
posure at concentrations reported to occur in drinking water. However,
a thorough search and review of the literature on all of the halogenated
methanes listed as present in drinking water was not conducted by the
Study Group. Presently there is no compelling reason to suspect that
any of the compounds pose any greater health risk as water contaminants
31
-------
than carbon tetrachloride and chloroform which are considered separately
in this report. It is important that more thorough reviews of those
additional halogenated methanes that occur frequently and in relatively
high concentrations (e.g., bromoform, bromodichloromethane, and
dibromochloromethane) be conducted. It is also important to determine
experimentally if bromochloromethanes are as biologically active
as the chloromethanes.
A halogenated ethane, 1,2-dichloroethane, has occasionally been
identified in drinking water. Carcinogenesis studies on this compound
are currently in progress at the National Cancer Institute. The results
of those tests will be important to the assessment of health risk of
organics in drinking water, particularly since recent studies suggest
that 1,2-dibromoethane produces gastric carcinomas in rats and mice
(9).
REFERENCES
1. Organic Compounds Identified in Drinking Water in the United States
(March 15, 1975). (Supplied to the Study Group by Water Supply
Research Laboratory, EPA, Cincinnati, Ohio, with notations on number
of locations where formed and concentrations).
2. Data from National Organics Reconnaissance Survey, April 15, 1975
Water Supply Research Laboratory, EPA.
3. Bellar, T.A., Lichtenberg, J.J., and Kroner, R.C., Report No.
EPA-670/4-74-008, 703 (1974). Ibid. JAWWA, 703 December (1974).
4. Morris, J.C., Draft Report on Formation of Halogenated Organics by
Chlorination of Water Supplies. Water Supply Research, EPA
Contract No. P5-01-1805J, February 1, 1975.
5. Patty, F.A. (ed.) Industrial Hygiene and Toxicology. Volume II,
Interscience Publications, 1963.
6. Christenson, H.E.: The Toxic Substances List 1973 Editions, U.S.D.H.E.W.,
PHS, NIOSH. U.S. Gvt. Printing Office, Washington, B.C.
7. AGGIH: Documentation of Threshold Limit Values, Revised Edition.
American Conference of Governmental Industrial Hygienists (1966).
8. Du Pont Company Literature surveys, furnished by Dr. H.D. Hartzler
9. Powers, M.B., Voelker, R.W., Page N.P., Keisburger, E.K., and
Kraybill H.D., Abstracts, Fourteenth Annual Meeting Soc. Toxicol.,
99 (1975).
32
-------
5. CHLORO-OLEFINS
Several chloro-olefins have been reported in water supplies.
These compounds are: dichloroacetylene, hexachlorobutadiene,
tetrachloroethylene, and 1,1,2-trichloroethylene. There are no
published reports on the carcinogenicity of any of these compounds.
On the basis of their close similarity in chemical structure and
reactivity to the carcinogen, vinyl chloride, it is likely that some
of these compounds may exhibit carcinogenic activity (1) t and hence
they may pose a health hazard in drinking water.
1,1,2-Trichloroethylene is currently on test for carcinogenicity
in mice and rats by gastric intubation. Preliminary findings from this
experiment suggest that this compound is carcinogenic in several organs,
particularly in the liver where it results in hepatocellular carcinomas
(Personal communication from Dr. U. Saffiotti, National Cancer
Institute). The known acute and toxic effects of these compounds and
their metabolism in animals and man are briefly reviewed in Appendix C.
REFERENCES
1. Van Duuren, B.L., Ann. N.Y. Acad. Sciences, 246, 258 (1975).
33
-------
-------
V. RISK ESTIMATION
In previous sections of the report experimental animal carcinogenesis
studies have been discussed. Two priority compounds, carbon tetra-
chloride and bis(2-chlorethyl)ether, have been shown to be carcino-
genic in rodents by oral administration. In addition, a third (chloro-
form) is suspect. Because of the presence of these compounds in drinking
water it is reasonable to assume that man may well be exposed to some
carcinogenic risk. Therefore, there is a need to estimate the possible
risk to man from these compounds at the levels found in drinking water.
Before considering the individual compounds cited above it should
be stated that extrapolations from ve>ry high to very low doses and
from species to species are highly speculative. This is particularly
true in the present case since there is little information on the
biological mechanisms invoJvea and rigorous experimental dose-
response data are lacking. The following discussion is an illustrative
example of an application of these extrapolations to estimate human
risk.
Chloroform
The only useable dose response data for carcinogenesis available
on oral administration of chloroform is from a study by Eschenbrenner
and Miller (see Section IV,A,1). Relating this study to lifetime
exposure to man in drinking water has several drawbacks.
(1) Administration was by stomach intubation every 4 days
as contrasted with the continuous exposure that would
occur by ingestion in drinking water.
(2) The experimental protocol involved a single mouse strain
and only 5 animals per sex per dose.
(3) The entire study lasted only 150 days as opposed to a
more valid lifetime study.
Keeping these difficulties in mind, the first calculation relates
the lowest experimental dose for which a positive effect was observed
(4xlO~^ ml/gm body weight per 4 days) to drinking water levels for
man. The animals were dosed only every 4 days. The dose of 4xlO~4 ml/gm
is considered here to be equivalent to a daily dose of 1x10"^ ml/gm
or 150 mg/kg. Acute toxicity studies of several anti-cancer agents
in animals and man suggest that dose rates calculated per unit of
surface area produce equivalent effects in several animal species (1).
35
-------
.Since siiriaco area is approximately pi oport lonal to tno :./J> power of hociy
weight, conversion from a 25 gin mouse to a 70 kg man, on ,; mg/kg basis,
would be made by dividing the mouse dose by M. Accepting this approach
and not including any additional "species conversion factors," the daily
mouse dose of 150mg/kg/day translates to a human dose cf 10.7 mg/kg/day'.
Chloroform has been found at levels as high as 500 i g/1 in drnfkin.,
\\ater. bsing 4 liters of water as a maximum daily intake for a 70 kg
man, these assumptions translate to a possible Ja:ily chlci^fonu intaKi-
ef 0.0171 mg/kg/day. Therefore, the ratio
dose (13.7 mg/kg/day) to the calculi/red intake- uose •. • '!"i~7l
is 026, An anirral sr:.iuy iaee^'porat IT/. ::'":•• ' •'- •
c iloroforfii wouJd pro!a'',-- ' v -v-• ;
J:;in f,/ar>d bv t'if ifn, a./ • .• iy -,. , •
{ l-ci r->''or\ . resi i t [:. '• i „'.,."• ,
; • •; •••e '/'Vibi Jered t.: ."•"• '; . " '' ; • •
re'~;.u;r e ,icse tr- •,•:•: f.i - ;>• •
i ; i iiie;!i.' < -x* i •;. C... . : :
i 01.--- T ^ ^'..Lil\^ i I . ', >, , lig V ';..','• , ^
i.iit j den1.'1.' i ourii. a''." 'xi""t- ' . •'. . • , • ; . , , .,;. • ••"•.. r ,.
extrripoiation on the tv\x i '-; •-.
izpper 9V,1 confidence b,oi!n, !.
(considering ;?, confidence ..v.iits) in mice.
Assuming an equivalent cfie';t ir man for a gi\eii. cj.^e per iLri~ sui face
area, a lifetime inciden'-e o: i.vcr ttiiiio'> -. ''a'1 -sing ^ ranking water a"
tlie highest observe J cr.lorofor i concent rat i..;.: i^oil.' DC- :x:)ectcd to
lie in the range ol 0 to .OOi, or 0 to L;OxJf)"' ji a ufetiine. Tliis
rate may be compared \:nl\ the lifetime irvikv e e . A]O~J for
malignancy of liver derived ffon data of tr^ ure \ui;onal Cancer
Sun/ey :'2j . That is, i^n-Jcr t .;- assi.m^tion.1- j;]voivrod i)-: .linear non-
threshold extrapolation over i->./e "u,in .! i ) .."' - of ^nay.itueie of UCTC
uind f roi i species to spre r. , ; ir: .'•;,xi,lu[ ob^c vc-; ehl'-rol O:T:I
concentration in Cr.nkir '•'-<•• ,~ :, •" consi.j:ieu .. cguiurl) s voulu produce
u chloroform-induce- 'ru : Jence ^l nn I j i;nancv o llvei i o tu an:iosc
40',', of the cbserve-d united -hates ij.ciclence. it must be noted that
-------
this value is the upper limit of the confidence interval and an
incidence of zero is also compatable with the observations. A linear
non-threshold dose-effect model may be thought of as allowing an
estimate of the maximal risk associated with a dose below the levels
where a risk has actually been observed. Other models that might be
considered for this extrapolation, such as probit, threshold, or
dose-squared models, would all yield lower estimates of the risk.
Estimates much lower than the estimate given here would be reasonable,
and the observations do not exclude the possibility of a threshold
concentration higher than the observed levels in drinking water. It
has been noted in the discussion of epidemiologic data that no excess
of malignancies of liver has been observed in the Louisiana counties
using water in which chloroform has been found (though at a lower
concentration than that of Case 3) and that in one analysis there was
a significantly lower incidence of malignancies of the liver in these
counties.
37
-------
-------
TABLE 1
Case 1 Case 2 Case 3
Observed incidence of hepatomas
in mice 3/3 0/18
95% confidence limits of
incidence 0.37-1.00 0-0.15 0-0.001
Dose to 25 gm mouse ml/gm/day lxlO~4 0.25xlO~4 0.16xlO~6
Estimated equivalent dose to
70 kg man mg/kg/day 10.7 2.68 0.017
Concentration in water mg/1
required to give man dose 187 47 0.300
Relative dose 4 1 1/156
Constants and assumptions of_ the model:
1. Specific gravity of chloroform 1.5.
2. Incidence is proportional to dose in mass per unit surface
area per unit time for doses below that of Case 2 (linear
non-threshold dose-effect model)*
3. Constant of proportionality between dose and effect is the
same for mouse and man.
4. Surface area is proportional to the two-thirds power of
weight.
5. Upper 951 confidence limit for incidence at highest dose
at which incidence of 0 is observed is given by
p=l-exp (^ In .05). Lower 95% confidence limit for
incidence at lowest dose at which incidence of 1 is
observed is given by p=exp (^ In .05). n=number of
subjects experiencing incidence of 0 to 1, respectively.
6. Water intake of 70 kg man is 4 liters per day.
39
-------
1
^t
b
r—
x.
LU
in
D
0
z
<
;
!
cc
LU "
H i
< ;
5 ;
rn '
**^
2 l(
-------
CM
LU
o
o
in'
CM
CO
CD
CM"
rf
\h
V\ V\
\
CM
LU
CC
D
^2
u.
O
CC
X
01
DC
<
at
•z.
o
in
LU
O
Z
LU
Q
O
41
oH
x
<
LU Q
w oc
2 LU
O a.
SSE
o
oc
LU
O.
r»»
*~.
o
<
D
OC
LU
Z a-
< a
s ^
IT
LU
a.
CD
CO
ca'
CC1
jS'tt
I"
^
?£
^ Q-
Z e?
EC
O
-------
Bis(2-chloroethyl)ether
One lifetime carcinogenesis study using oral admir.isVration of
bis-:(chloroethyl)ether has been reported by Innes et a].(see Section
IV,A,3). In this study both sexes of hybrid mouse strains \\ere studied
vith 18 ar.inals per experimental group for a total of "."!. A single
dosage regimen was used (100 mg/kg within 7-28 days of ap<- followed h\
500 ppm in diet for the remainder of life) and a centre] croup were
studied. High observed incidences over background were found,
particularly in the males.
Using the same analytic procedure as :ir thi. prcv:\^
the experimental dose translates to -J.2P mg/j\.;/c!r>y in ,r,
level in drinking water of f\5 ;jg/l th.e doso in ]!.,'• . .-
ug/kg/dny. The ratio of these dose rales i.c a, \-\-^\ i:-;-; 'h
which is considerably larger than the (<2b found u; th '.'; ;
should be recognized, however, that there Is n.~ inlV1!''!:-
concerning how large a reduction in the- expert'onf;i 1 C.< ••<-
would continue to y eld a positive response. .rhcr
-------
REFERENCES
1. Freireich, E. J., Gehan, E. A., Rail, D. P., Schmidt, L. M.
and Skipper, H. E., Cancer Chemotherapy Rep. 50, 219-244 (1966).
2. Preliminary Report, Third National Cancer Survey, 1969 Incidence,
Nat. Cancer Institute, USDHEW (1971)
43
-------
-------
VI. APPENDICES
A. SUMMARIES OF EPIDEMIOLOGIC STUDIES EVALUATED
1. Ishimaru, T., Okada3 H., Tomiyasu, T., Tsuchimoto, T., Hoshino,
T., and Ichimaru, M. Occupational factors in the epidemiology of
leukemia in Hiroshima and Nagasaki. Am. J. Epid. 93: 157-165.
1971.
The essential data of the association of malignancy and exposure
in this study are enumerations of matched pairs of adult individuals,
each pair including one patient with leukemia and one non-affected
person. Each pair was characterized as to whether both, one, or
neither had been exposed occupationally to benzene
Number of Pairs
Leukemia Patient
Exposed
Non-Affected
Exposed
The risk ratio estimate is the ratio of the 2 discordant cells,
28/12 =2.3 and differs significantly from the null ratio of 1 with a
p value less than II.
Leukemia patients were classified as having definite or probable
leukemia diagnosed between 1945 and 1967 and resident in Hiroshima or
Nagasaki, Japan. Non-affected persons were selected from the Atomic
Bomb Casualty Commission Leukemia Registry sampling frame, matched to
the patient as to city, sex, age, distance from hypocenter of bomb
explosion, and alive at the date at which leukemia was diagnosed in
the patient. Exposure referred to holding any of 10 occupations
considered to involve exposure to benzene. It is noted that in all
the identified occupations there was also exposure to other organic
substances.
No estimate of the exposure dose is given.
If the incidence of leukemia in the non-exposed persons in this
population is assumed to be 10 per 100,000 per year, the benzene-
No
Yes
'AL
NO
261
12
273
YES
28
2
30
TOTAL
289
14
303
45
-------
induced incidence may be estimated as 13 per 100,000 per year
Vigliani, E.C. and Saita, G. Benzene and leukemia. N. Eng. J.
Med. 271: 872-876. 1964.
This stud}- involves an enumeration of leukemia judged to be due
to chronic benzene poisoning in persons covered for insurance by the
Xaticnal Institute for Insurance against Accidents and Occupational
Diseases in the provinces of Milan and Pavia5 Italy, This number c f
cases is related to the investigators' belief as to the number of
workers exposed to benzene, and the derive-,; .ncuxnce ^ate ',•- <• ••
with tlie reported incidence of lenKe:11 \- ",- . ',-• • \:\<.><.<\i ' ••
for Milan, .\\mterical vali:e> i'',"nd :; •
CI 1U , 'UP i >opuL';t i .;r- ] ' "
• . 2 to 2.1 ,: -.rt.-. per •.,''••• -
concentrate;/, cf !vu; ^ •, .;
of the pat .ents s',; 'live
From the ^bove, \i;o benzene- :'• rcco
11 per 5,0l:0 per 4 years — 1 per .ifjOiK1
per year associated witii an cxpOMac- of
air concentration.
Harris, X. H. Tlie implicate>;:--- of conce"-causing substances s;i
Mississippi River water. Unpublished.
4. Page, T. and Harris, R. H. Relation between cancer mortality
and drinking water in Louisiana. Ui.publishcd. 1975.
This stud}' involves a tabulation of mortalit>' rates by county
for all cancer and for selected cancer sites and site groups, sex and
race specific, for the State of Louisicxia for the total period ID SO
to 1969. Counties are categorized as to proportion of drinking
water derived from the Mississippi River, ;md a regression analysis
is used to determine the cancer risk attr^Liitable to the use of
Mississippi River water for drinking. Three known or suspected
carcinogens — chloroform, benzene, and carbon tetrachloricle — '.scic
identified in raw or treated water supplies of plants serving parts of
Louisiana. Chloroform was found in finished water supplies. The
analytic procedure included statistic;'! control for rural-urban
characteristic, median income, proportion of employed population i:,
petroleum industry, proportion of population in cnemical industry,
and proportion of population in mining industry.
46
-------
The following table lists regression coefficients for the
variable representing proportion of drinking water from Mississippi
River.
Regression
Coefficients
Cancer Sites White Male Nonwhite Male White Female Nonwhite Female
All sites
Lung
All other than lung
Genitourinary
Gas tro inte s t inal
Liver
32.5* 49
7.5
25.4*
3.6* 1
7.0* 19
-0.15
.5* 3.0 29.3*
.6 1.5 2.7*
.4* 4.9* 13.3*
Coefficient significantly different from 0 at 51 level
The regression coefficients have dimensions deaths per 100,000 per
year per percent of water from Mississippi River. Therefore, a
coefficient can be interpreted as an attributable risk in units of
deaths per 100,000 per year, attributable to 1001 of water from River
as compared with 0% from River. (Here the word "attributable" is
used, as is common in epidemiologic literature, to refer to a
difference in rates between two exposure categories, without
implication that the difference is causally related to the exposure
difference.)
Discussion sections of this paper seem to give a causal inter-
pretation to these attributable risks, though in oral discussion
between the investigators and this Study Group it was indicated that
this interpretation was not intended.
5. Tarone, R. E. and Gart, J.J. Review of "The Implications of
Cancer-Causing Substances in Mississippi River Water" by Harris,
R. H. Unpublished. 1975.
This study involves a further analysis of tabulations of cancer
mortality by county in Louisiana as related to proportion of drinking
water derived from the Mississippi River, expanding on the analysis
given by Harris, R. H., 1974. The new analysis includes the
additional variable elevation above sea level, some refinement of
the regression model, and expansion to race-sex groups not initially
studied by Harris.
47
-------
The following table lists the presence (*) or absence (NS) of
statistical significance at the 5% level for the variable representing
proportion of drinking water from Mississippi River.
Significance of Regression Coefficients
White Male Xonwhite Male White j'emale Nonwhi_t_e_ Female
Cancer Sites
All sites * *
c. Lie use oi: group vcc-;rit}; :
-issociat ions.
d, The inconsistency of ihc association among tie 4 race-sCA
sub-groups.
.)e Rouen, T. A. and :\eni, •'<. •",, ?thr;ic, geographic diffeiencc
in cancer mortalit}' in Louisiana, i Tpuo [ ished, 1975.
fhis is a third study involving analysis of v.mcer mortality n
county in Louisiana as relatet! to viKtncr or not ; art of the- di^j1'. i
water of the count)' is oi'tained fuw; t1 •;• -uss • ss i, pi 'River. TIMS
analysis includes the additional variu I-,' o(" latitude, JividJng l/o
State of Louisiana .into Xorth and South counties, a division
noted to be associated with major socio-cultural differences. Ot'.e
variables shown to tie significant con ;:OUP hng variables in the
other two studies wen not used tr_ tin- analysis. These omitUvi
variables are urban-rural charactC'ris". ic, nic-dian income,
emplo>iiient characteristics, and elevation above sea Level. The
48
-------
water variable is here treated as a simple dichotomy, none of
water obtained from River versus some or all from River. The
statistic studied is again the difference in cancer mortality
rates between the rates for counties using River water and those
for Southern counties using no River water. The difference,
however, is defined somewhat differently from that in the studies
by Harris and Page and the study by Tarone and Gart in that the
rate for counties using River water is a representative value
for the group of counties obtaining some or all of their water
from the River rather than the rate for 1001 use of River water.
Differences in mortality rates for Southern counties between
using and not using counties are the following -
Mortality Rate Differences
Cancer Sites White Male Nonwhite Male White Female Nonwhite Female
(1.) All sites 11.15
(2.) Lung 3.75
(3.) All other than 7.40
lung**
(4.) Kidney .50
(5.) Bladder l.SO*
(6.) Kidney plus 2.00
bladder**
(7.) Stomach -1.35
(8.) Rectum 1.80*
(9.) Large intestine 1.40
(10.) Stomach rectum,** 1.85
and large
intestine
(11.) Liver -1.85*
(12.) Breast
(13.) Cervix
(14.) Uterus
(15.) Ovary
(16.) Melanoma -0.80*
(17.) Brain -1.0
(18.) Pancreas -1.85
(19.) Multiple Myeloma 0
(20.) Leukemia 1.30
(21.) Prostate 0
13.30
5.40
7.90
.35
.20
.15
7.10
1.50
.20
8.80
-5.75
-2.05*
-3.70
.30
-.10
.20
-.30
.70
2.50*
2.90
18.00*
-.35
18.35
,60
33
,93
5.05*
.70
3.55*
9.30
-1.90
-2
.89
.60
0.4
-2.30
-1.15
.20
.80
.40*
.50*
.15
.10
.40
.65
.80*
.50
..20
..90
L85*
-.25
.90
-.75*
1.80
.20
.10
*Difference significantly different from 0 at 51 level
**Combination values not given by authors. Significances
of these combination values not known
49
-------
Discussion by these investigators is to the effect that a
difference of cancer rates exists within South Louisiana with slightly
higher rates in counties along the Mississippi River. They caution,
however, that the many potential causes in addition to water (quality)
make it difficult to identify any of these as true causes, fhef.e
comments evidently refer to the cancer categories 1 to 10, above,
identified as associated with the water variable in the studies by
Harris and Page, and not to the additional categories 11 to 21.
50
-------
B. PHTHALIC ANHYDRIDE AND PHTHALATE ESTERS
Phthalates in Water.--The distribution of phthalates over the
entire aquatic environment is evidenced from data in eleven reports.
Of these, five reports (2,6,7,73,56) deal with treated water supplies.
It appears as if phthalates are general contaminants in drinking
waters with amounts approximately 1-2 yg/1 or less. The total
phthalate concentration found in the New Orleans drinking waters
was 1.8 vig/1 distributed among eight different phthalate esters.
The amounts found in>w40 diffei'ent municipal drinking waters in the
U.S. were generally below 1 ;jg/l (7).
One probable source of this general phthalate contamination is
contact of the water with plasticized polymers (la). Other similar
sources associated with the use and disposal of plasticized polymers
seem highly probable. Industrial dumps of phthalate esters into water-
ways probably contributes very little to the total phthalate burden in
the environment.
Phthalate Production Levels.--The total production level of
phthalates (41,42) approaches one billion pounds per year and has been
increasing and generally shifting from the lower to the high molecular
weight esters where the vapor pressure, the water solubility, and the
acute toxicity is less. However, the intrinsic toxicity of these
heavier phthalates is reported to be much greater (20).
Phthalate Acute Toxicity.--The low acute toxicity of the
phthalate esters is well documented (14,20-25). 11)50 values for the
phthalates are all measured in the gram/kg range. No conceivable human
exposure at this lethal level is to be expected.
Phthalate Teratogenicity.--Fetal abnormalities are well documented
in the literature (14-20).However, the dosage levels are well above
the known or anticipated human exposure to phthalates especially
through oral ingestion of H20.
Phthalate Metabolism and Accumulation.--The accumulation of orally
fed phthalates and the metlEolism have been studied (43-48). The rela-
tively high levels in the liver and lung (47,48) as well as lesser
amounts in the heart (44,48) may be clues to future research efforts.
The occurrence of phthalates in the beef pineal gland (46) of cattle,
whose only apparent exposure to phthalates was from drinking water
transported through plastic pipe, is noteworthy.
51
-------
I'ae slow biodcgradabil ity and the documented biomatnification (43)
suggest that problems presently unknown may be manifested in the
future.
The known metabolic products are phthalic acid, the monoester,
alcohols, phthalic anhydride, and a variety of .^characterized polar
metabolites and conjugates (43,47a).
Phthaiate Chronic Toxicity and Health Threats.--The chronic and
sub-acute toxicity studies and the human health implications have been
recently reviewed by Autian (20). The conclusion that human exposure
to phthalates poses no imminent health hazard is generally shared by
other investigators (14,24,26-40).
Synergistic effects, either positive or negative, liave not [seen
studied and where these effects may be openstite, the exposure of
humans (39a) and aquatic organisms (39"), have been an ill-JcfuicJ mix-
ture of chemicals with the principal component'-', being phthalatc esters.
The moderately pronounced toxic polyneuritis observed as nn occu-
pational illness by the Russians (59a) is cause for some concern. L\~c-:\
though these results closely parallel the effects observed in animal
studies, the exposures, routes of administration, confirmations, syrier-
gistic effects, and extrapolations to expected human exposure are so
poorly defined that a reasonably accurate assessment of the effects from
chronic exposure is not possible. L. B. Tepper (23) in an overview of
phthaiate esters stated that we appear to be in the preferable position
of having an "etiology which is searching for a disease" rather than the
reverse situation where overt diseases stimulate intensive searches for
etiologies.
When the most incriminating summaries (8,9,14524,26-29) of the
potential phthaiate ester problem are objectively reviewed, one concludes
that the evidence fcr direct human health threats is primarily specula-
tive and non-definitive.
52
-------
REFERENCES
PHTHALATES IN WATER
1. Buadze, 0. B., et al., J. Polymer Sci., _33, 349 (1971). [test H20].
la. Junk, G. A., et al., Environ. Sci. Technol., 8_, 1100 (1974). [test
H20].
2. Burnham, A. K., et al., J. Amer. Water Works Assoc., 65_, 722 (1973).
[treated H20]
3. Hites, R. A., Environ. Health Perspectives, .1, 13 (1973). [River H20]
4. Hites, R. A., and Biemann, K., Science, 178, 158 (1972). [River H20]
5. Corcoran, E. F., Environ. Health Perspectives. I_, 13 (1973).
6. USEPA, Region VI, Draft Analytical Report--New Orleans Area Water
Supply, Dallas, Texas, Nov. (1974). [Treated H20]
7. Junk, G. A., private communication [Rivers, Wells, Treated H20]
7a. Grob, K., J. Chromatogr., 84_, 255 (1973). [Treated H20]
8. Pastorelli, L., et al., Ann. Chem., 61, 311 (1971). [Sea H20]
See also ref. 56.
PHTHALATES IN MILK
9. Cerbulis, J., and Ard, J. S., J. Ass. Offie. Anal. Chem., 50, 646
(1967). ~
10. Wilbrett, G., et al., Fette Serf en, Anstrichm., 71_, 330 (1969).
PHTHALATES IN SOIL
12. Ogner, G., and Schnitzer, y., Science, 170, 317 (1970).
13. iMatsuda, K., and Schnitzer, M., Bull. Environ. Contam. Toxicol.,
6, 200 (1971).
53
-------
14. Singh, A. K., et al., J_. Pharm. Sci., 61_ (1972).
15. Singh, A. R., et al., 10th Annual Meeting of the Society of
'lexicology, Washington, D. C. March (1971).
16. Bower, R. K., J_. Pharm. Exp_. Thera., 171, 314 (1970).
1". Mayer, F. L., 27th Midwest Regional Meeting of the ACS,
St. Louis, Oct. 1971.
18. Dillingham, E. 0. and Autian, J., Environ. Health Perspectives,
1, SI (1973).
19. Haberman, S., et al., Soc. Plastics Eng. J., 34, 62 (1968).
20. Autian, J., Environ. Health Perspectives, 1, 3 (1973)
PHIHALATE ACUTE TOXICITY
21. Gesier, R. M., Environ. Health Perspectives, !_, 73, (1973).
22. Smyth, H. F., Amer. Ind. Hyg. Assoc. J., 30, 470 (1969).
25. [lodge, H. C., Proc. Soc. Exper. Biol. Med., S3_, 20 (1943).
24. Shaffer, C. B., et al., J. Ind. Hyg. Toxicol., 27, 130 (1945).
25. Radeva, M. and Dineva, S., Khig. Zdraveopazvane, 9_, 510 (1966)
in Bulgarian - see CA 66_, 1153632 (1967).
See also references 20 and 14.
PHTHALATE CHRONIC TOXICITY AND HEALTH THREATS
26. Shea, K. P., Environment, 13, 2 (1971).
27. Eckardt, R. E. and Hindun, R., J. Occupat. Medicine, 15,
808 (1975).
54
-------
28. Tepper, L. B., Environ. Health Perspectives, 1_, 179 (1973).
29. Marx, J. L., Science, 178, 46 (1972).
30. Nematollahi, J., et al., J. Pharm. Sci., .56, 1446 (1967).
31. Anonymous, California Medicine, 112, 43 (1970). February.
32. Mayer, F. L., Chem. Eng. Mews, 49, 8 (1971).
33. Neerguard, J., et al., Scand. J. Urol. Nephrology, 5, 141 (1971).
34. Galley, D. J., et al., J. Pharm. Sci., 56, 240 (1967).
35. Galley, D., et al., J. Pharm, Sci., 55_, 158 (1966).
36. Guess, W. L., et al., Amer. J. Ftosp. Pharm., 2_4, 494 (1967).
37. Guess, W. L. and Haberman, S., J. Biomed. Mater Res., 2, 313 (1968)
38. Stalling, D. L., et al., Inviron. Health Perspectives, l_,
159 (1973).
39. Mayer, F. L., et al., Report from Fish-Pesticide Research
Laboratory, U. S. Dept. of Interior, Columbia, Missouri (1971).
39a. Milkov, L. E., et al., Einviron. Health Perspectives, 1, 175
(1973).
40. Meyler, F. L., et al., Circ. Res., 8_, 44 (1960).
See also references 14, 20, and 24.
PHTHALATE PRODUCTION LEVELS
41. Graham, P. R., Environ. Health Perspectives, l_, 3 (1973).
42. Mather, S. P., J. Environ. Quality, _3, 189 (1974).
PHTHALATE METABOLISM AND ACCUMULATION
43. Metcalf, R. L., et al., Environ. Health Perspectives, !_, 27 (1973)
44. Nazir, D. L., et al., Biochem., 10, 4228 (1971).
55
-------
45. .\azir, D. J., Fed. Proc., _26, 412 (1907).
46. Taborsky, R. G., J. Agr. Food Chem., 1_5_, 1073 fl%7).
47. Jaeger, R. J. and Rubin, R. J., Science, 170, 460 (1970).
47a. Stalling, D. L., et al., Environ. Health Perspectives, 1,
159 (1973). ~
48. Jaeger, R. J., and Rubin, R. J., Environ. Health Perspectives,
1_, 95 (1975).
«
PHTHALIC ANHYDRIDE TOXIC I TV
49. Baader, E. W., Arch. Gewerbepathol. Gewerbehyp., 13, 419
(1955).
50. Menschick, II., Arch. Gewerbepathol. Gewerbehyg., 13, 454 (1955)
51. Friebel, H., Arch_. Gewerbepathol. Gewerbehyg., 14_, 465 (1956).
52. Jacobs, J. L., Proc. Soc. Exp. Biol. Med. , 43_, 74 (1940).
53. Lefaux, R., Practical Toxicology o£ Plastics, CRC Press»
Cleveland, Ohio, 132 (1968).
54. Sax, N. I., Dangerous Properties of Industrial Materials,
3rd Edition, Van Xostrand-Reinhold, X. Y., 1026 (1968).
55. Patty, F. A., Industrial Hygiene and Toxicology Vol II,
Interscience Publishers, X. Y., 1823 (1968).
See also reference 20.
PHTHALIC ANHYDRIDE OCCURRENCE IN DRIXKIXG WATER
56. USEPA, Water Supply Research Laboratory, Organic Compounds
Identified in U.S. Drinking Waters and their Toxicity. Unpub-
lished Report. Dec. 9, 1974.
56
-------
C. CHLORO-OLEFINS - TOXICITY SUMMARIES
Dichloracetylene
Humans inadvertently exposed to dichloroacetylene in a sealed
system had severe nausea, vomiting, headaches, and facial sensory dis-
turbances (1). Rats exposed to 2.8, 9.8, and 15.5 ppm of this compound
for 6 hr/day, 5 day/wk, for 6 weeks showed pronounced morphological
changes in the kidneys (2,4).
Hexachlorobutadiene
A 1969 study of Russian vineyard workers exposed to HCBD and poly-
chlorobutane came to the following conclusions: the population suffered
from increased hypertension, myocardial dystrophy, respiratory diseases,
diseases of the nervous system, and hepatic disturbances (5). The legal
maximum air concentration in the Soviet Union was set at 0.01 pg/l (6).
This is based on the fact that rats exposed to vapors at this dose for
5 hr/day for 6 months showed no ill effects.
The 11)50 for HCBD by parenteral injection has been shown to be
90— 350 mg/kg for mice, rats, guinea pigs, and rabbits (8). In rats the
LDso given topically was 4.3 g/kg and given orally 165 mg/kg (7). A
large number of studies of the effect of HCBD on rats have been carried
out. Some of the findings are as follows:
(1) 20 mg/kg, orally, showed rapid (30,90, and 360 mins) degenera-
tion of body protein, fats, and carbohydrates (9);
(2) British workers found that rats exposed to vapor for three weeks
showed severe kidney damage (10);
(3) 8.5 — HOmg/kg, orally, showed tissue degeneration and ab-
normal changes in the brain, liver, and other organs (11);
(4) 300 mg/kg showed decreased glutathione and ascorbic acid in the
liver while the glutathione in the kidneys increased. The same study
showed the levels of succinimide oxidase and cytochrome oxidase in
internal organs to have been decreased (12).
Additional experiments have shown the toxicity of HCBD to aquatic
organisms: 3 mg/1 is toxic to Daphnia magna and Leucaspius delineatus
(13). One experiment showed the effect of administration of 20 mg/kg
to albino rats on their offspring. In three months all the offspring
were dead compared to ~20I of the control offspring (14). HCBD has been
found in drinking water at a concentration of 0.60 yg/1. (EPA document).
Its solubility in water is low, 0.5 mg/1. It is known to degrade "fairly"
rapidly in air and water. In air, it is degraded to hexachlorobutadiene
57
-------
epoxide and phosgene, llexachlorobutadiene is produced in the U.S. (7.3
million Ibs/year), is used as a solvent for polymers and as a heat
transfer liquid. Jts dispersion in the environment can be attributed to
this wide usage.
No carcinogenicity or mutagenicity data are available.
REFERENCES
1. Saunders R.A., Closed Atmosphere Contamination, Naval Research Lab.
Report, C.A., 66, 49,030m (1967).
2. Siegel J., et al_., Toxicol. Appl. Pharmacol., 1_8, 168 (1971).
4. Jackson M.A., et_ al_., Toxico.. Appl. Pharmacol., IS., 175 (1971).
5. Krashyuk E.P., et_ al., C.A., 71_, 94,514m (1969).
6. Poteryaeva G.E., C.A., 76_, 117 (1972).
7. Chernokan V.F., C.A., 74, 97,218r (1971).
8. Murzakaev F.G., C.A., 67_, 89,280r (1967).
9. Marzakaev F.G., C.A., _66_, 49,030iri (1967).
10. Gage, J.C., Brit. J. Ind. Med., 27_, 1 (1970).
12. Gudumak V.S., C.A., 71, 37,420d (1969).
13. Stroganov N.S. and Kolosova L.V., C.A., 72_, 97,852b (1970).
14. Poteryaeva G.E., C.A., 65_, 1281£ (1966).
Tetrachlorethylene
This compo'orid is widely used in dry cleaning, in degreasing metals.,
as a solvent, and as a vermifuge (1).
14
A study reported in 1961 of mice exposed to C-labelled tetraclilcro-
ethylene vapor showed that after four days ~70°l) of the label was expired
and ~20t, was in the urine. The major urinary metabclities were trichloro-
acetic acid (52?6), and oxalic acid (1U) and s trace of chloroacetic acid.
The author postulated an epoxide intermediate in the metabolism of
58
-------
tetrachloroethylene (2). Another study with rats stated that ethylene
glycol was the major metabolite; trichloroacetic acid and oxalic acid
were also found (3).
REFERENCES
1. Hawley G.G., ed., The Condensed Chemical Dictionary, 8th ed.,
Van Nostrand Reinhold Co., N.Y., (1971).
2. Yllner S., Nature, 191, 820 (1961).
3. Dmitrieva N.V., Gig. Tr. Prof. Zabol., 11, 54 (1967)
CA 66, 93,533b (1967).
1,1,2-Trichloroethylene (TCE)
This compound occurs in drinking water at a concentration of 1.0
ug/1. It is manufactured in the U.S. in large quantities (429 million
Ibs/year) and is used as a degreasing agent for metals, as a heat transfer
liquid, cleaner for raw wool, and as solvent for the extraction of residual
oils from vegetable oil cakes (e.g., soybeans).
The carcinogenic and mutagenic effects of TCE have not been explored.
Its metabolism and toxicology in animals and man have been examined (1-3).
Trichloroethylene is slowly oxidized in air, in presence of light,
to an epoxide. Other degradation products include phosgene and dichloro-
acetyl chloride (Cl CHCOC1J.
I*
REFERENCES
1. Powell J.F., Brit. J. Ind. Med., _2, 142 (1945).
2. Daniel S.W., Biochem. Pharmacol., 12_, 795 (1963).
3. Smith G.F., Brit. J. Ind. Med., 23. 249 (1966).
59
-------
VIi. ATTACHMENT A
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\SHINGTON, D.C.
March 12. 1975
^ WASHINGTON, D.C. 20460
OFFICE OF THE
ADMINISTRATOR
SUBJECT: Charge to the Ad Hoc Study Group to Consider Organics
in Drinking Water
FROM: Dr. Emil M. Mrak, Chainran ^XL^^-4^. ^££
Hazardous Materials Advisory Committee'(RD-6"f^f/
TO: Members of the Ad Hoc Study Group to Consider
Organics in Drinking Water
In recent months intense public interest has become apparent regarding
the observed occurrence of certain organic compounds in drinking water.
The reported levels are extremely low, detectable only by analytical
methods that have recently come into practice. The assertion has been
made by some scientists and many non-scientists that these observations
indicate the potential for a significant increase in the risk of cancer
in populations exposed to these waters.
The Safe Drinking Water Act provides the Administrator with several
options for dealing with the potential risk, ranging from use of emergency
powers contained in Section 1431 of the Act to a decision not to include
any such chemicals in the interim primary drinking water standards that,
under the law, must be promulgated by June 17, 1975.
The Agency has sought the advice of the Science Advisory Board as to
the significance of contamination in drinking water relative to potential
carcinogenicity or other effects in humans resulting from chronic exposure
to these compounds. The Hazardous Materials Advisory Committee will
respond to that request.
In order to provide the best possible advice to the Agency I am
asking that the Ad Hoc Study Group to Consider Organics in Drinking Water
consider the available information in detail and provide the Coirmittee
with a report including the following:
1. The Study Group's best assessment of the risk to people drinking
water contaminated with the following chemicals:
Benzene 5 ppb (micrograms/liter)
Carbon Tetrachloride 5 ppb
ciiloioforfii 100 to 200 ppb
VII-1
-------
The levels indicated above ho.ve been observed In one or more
drinking water supplies. Presumably, at times these chemicals
could occur at concentrations substantially higher. These levels,
therefore, should be considered as an indication of order of
magnitude only.
2. The Study Group's best assessment of the significance of try-
following compounds, suspected of careinogenicity, which, have
been found in drinking water, although concentrations have not
been determined:
beta-chloroethyl methyl ether
octadecane (?)
phthalic anhydride
Although the seven were selected with regard to possible carcinogen
risk as referenced in a recent petition by the Environmental
Defense Fund, the Agency would not wish to have other health effects
of these or the possible effects of other chemicals ignored., if
they present a greater hazard to public health.
3. The Study Group should review the list of well over 100 organic
chemicals found to occur in drinking water, for the most part, at
ppb or lower levels, provided by the Agency.
If the Group is of the opinion that one or more of these chemicals
presents as great or greater health hazards than the seven
identified above, the Agency would appreciate a staterert of how
you would describe the degree of health hazard for such chemicals.
Because of the size of the list an exhaustive review will hardly be
practical in the time allowed for the report. Therefore, I suggest
that the study group concern itself primarily with the seven
chemicals listed above and any others critical to this aspect of wate
supply in attempting to formulate a first report by the end of April.
Consideration can be given to a more comprehensive report at a later
time. Pesticides. <=isbpRtn=;, ^nd •ir\c>~c&^n~i<~' r
-------
/TTACHMENT B
Speakers and Participants, Meeting of March 24-25, 1975
Ad Hoc Study Group to Consider
Organics in Drinking Water
Mr. William A. Coniglio
Office of Toxic Substances
EPA Office of the Assistant Administrator for
Water and Hazardous Materials
Mr. George Coling
Environmental Defense Fund, Inc.
Dr. Timothy A. de Rouen
Department of Health Measurement Sciences
School of Public Health and Tropical Medicine
Tulane University
Dr. Hend Gorchev
Office of Environmental Sciences
EPA Office of the Assistant Administrator for
Research and Development
Dr. Edgar A. Jeffrey
Water Supply Division
EPA Office of the Assistant Administrator for
Water and Hazardous Materials
Mr. Leland J. McCabe
EPA Water Supply Research Laboratory
Dr. Talbot Page
Resources for the Future, Inc.
Dr. Michael J. Prival
Office of Toxic Substances
EPA Office of Water and Hazardous Materials
Mr. Gordon G. Robeck
EPA Water Supply Research Laboratory
Dr. Robert G. Tardiff
EPA Water Supply Research Laboratory
VII-3
-------
!ir. > iJllam '•'. > 'pholt
ITA Of! ice of the Assistant Administrate.'!-
Hazardous Materials Control
r)r. } 'd i th Vcrmaj ii
FnvLroranental Defense Fund, Inc.
Mr. Herbert Wiser
Office of I-nvircranental Science,-.
I:IY Office of the Assistant \,'•:•'," ~\-i' >
1'eserTch and "evelenr-f r '
VII-4
-------
ATTACHMENT C.I
Organic Compounds Identified
In
U. S. Drinking Waters
And
Their Toxicity
Water Supply Research Laboratory
Environmental Protection Agency
Cincinr. ati, Ohio
December 9, 1974
VII-S
-------
3 . OC<-n<.t!.h (I: V i >
2. ^.r-UluVI,.,,!.
3. ocr t i c .-.ici d
I u 'i in; UM i i.i) :>
(as of n/xw/'i) Ty.'T-Vr-' . -,-
f VL •' ' '
i r \ j ' / / •
'' '
5. occloph nor, ,
G . ;r.
bu ty 1 b.1 ,,:(•!;.,•
butyl brof.iiric
c-capro 1 cjctc/n
ca r bo n d i -.i' i f i do
car b o n t o t r n c h 1 o r i d c
ch loi-r.;r,n( •?)
cb 1 orc;L;.'ii : one
chlorodi L'!'O",oi,ic Ihcne '''
\, 2-bi s-c 1 1 orootli'">:;y otbonc
chl oro;,'Llu),\y cthc"'
bi s-2-ciil orooLhy 1 c t !u> r
/3-chl oru'jthy 1 iv.otiiyl cllior*
chl orof o r.ii *
ch 1 orohyd roxy bonxf/plionuno
b i s • c !i 1 o f o i r> o |; r o •:• y 1 c L !i c r
v ch loro:.ivl:!v/ 1 ct ho>'
ch 1 o roi'i'- Lhy ! othyl otfiof
in-ch 1 oi'on i 1 r cdjc-n.: c iic
3 ~ c h 1 o i c • p y r i (.1 i n o
*Subr,tancr'G in Cincinnati
Wntor Supply
VII-6
-------
1.
53.
56.
57.
58.
59.
GO.
Cl.
G2.
G3.
64.
G5.
66.
67.
6
G9.
70.
71.
72.
73.
74.
75,
7G.
77.
7ft.
79.
80.
€1.
02.
83.
84.
85.
06.
G7.
88.
89.
DDE
DOT
doGUno
d i bi'oii'.obe n>: uno
Ci i b roinOCM > O i'(_ |,!0 thv.no *
d i bromod i en 1 o roe I hi' no
d i - t- bit ty 1 - p~b(;MZo<-;u i no no
di butyl phthala to *
1, '; •• d i c h 1 or o I) c: n z c n o
d i ch 1 o rod i riuorocthano
1, 2 •- d i c h 1 (•> i' o o t IT a n c
d i ch 1 o roo thy 1 o the'
d i ch 1 o ro:,"c t hone
die]d r i n
di ethyl phtho la to *
di(2-ethyl hor.yl) phtholate*
di hcxy 1 pl'.thal ato
d i hydroc:;: rvone
d i - i GO b'J ty 1 CG r b i nc 1
di-iso butyl phtho1cte
1,, 2-d i nctiioxy benzene;
dimethyl benzene
1, 3 - d i me 1". h y 1 n 6 ph t h ':>} one
2, k-dimethy1 phenol
dimethyl phthai ate
di methy • su1foxi do
k. G-d i n i t i'o-2-arni no phenol
2/ 6 -d i n i '. i'ololuene
diocty1 ad i p?te
d i propy 1 phtiiol ate
docc ; me *
n-d^-decane
ei cosanc *
c n d r i n . .
e tli a no 1*
ethyl ami nc
ethyl benzene
2-ethy 1 -*n-hcxane
2 - e t h y 1 - !f - me t h y 1 - 1, 3 - d i o xo 1 o n e
k-c t h y1 - 2-me t hy1 -1/3-d i o xo1 a n c
o-ethy1f1uone
90. guaiacol
91. heptnchlor
92. heptachlor cpoxide
93. 1,2, 3,'(,5, 7,7-heptachloronorbornene
9<*. hcxoch lore benzene
*Substances in Cincinnati Water Supply
VII-7
-------
97 . ho/.aoh 1' >'o( thane
98. hor.ruJcc.'iir *
9 9 . 2 •• 11 y d i o.-. y a -,; i po n i t r M e
100. indeno
3.01 . i sodeoane
3 02 . i sopho rono
303. i r.obortiool
30'. } •- i r.opre,;-; nyl •-!(•-! sopropy Iben;: one
3.05. i s o p r o p y 1 b o n,'. o n (?
3 0 C . 1 i non one
107. in (: t h y 1 o L t f.' r of 1 i f, no eerie a c i c! *
10H. p-meiUh" 1 -CMV-S-Ol
109. mothano
310. nothano!
3.3.1. 2--ne'choyy bipiicnyl
1.12. niuthy 1 bf niioatc
113, 1110 I hyl boh,:o th i a;:o 1 o
114. rnothyl bifhoriyl
ll'S. 3~ met hyl but anal
3.3 (-, motahy 1 eh 1 or i de
117. n"i ethyl ethyl kotono
3.10. 2- i.iclhy 1 -5- ethyl -pyri di no
119. sue thy 1 i r.dene
.120. iT-'Mivl nr.-r.h \.ha 1 one
321. methyl p;: 1 MI la to *
322, ric-thy) pho.iyl carbinol (1- phony 1 e lhano 1)
323. 2--tr.cll-.yl p>-op,. pen tone
137. pentanol N
130. phony 1 benz or, to
139. ph 1. h a 1 i c a n' i y d r i d o *
3>'0. pi'op a no 1
*Substnncos in Cincinnati V7atcr Supply
'VII-8
-------
343. pr"py 1 tihi i no
142. o opy 11'On/.ono
143. 1" lor pi nc o (
144 . i / ] / 3 / 3- 1111. »'uch i oi-( >iicc tone *
14 !3 . 11' L :Vj ci; 1 or oh i phony i*
1'-'.'. I r l r^c.h 1 oi'o" tii y 1 one
1If. tot t'odccono
1/5, tti i oinothy 1 L'0n/:oth i ..20! o j
l-5(.. toluene
1 >. „ t i" i c h 1 o i" o I o P, 7. o n c1
lr>i . tr i ch ! o<"<.-L.> i phony 1*
1 j: . 1 , 1 , 2 - t f i o i) 1 o >' oo t h; n c
3 'j . I f 1, 2- I <' I ch 1 o roe th y 1 one
1:5' . I >'i ch 1 oro-n uor'oi.iGi i anc
15( » 2, '!, G- L c S ch } oi'opiioi
-------
-------
ATTACHMENT C.2
ORGANIC COMPOUNDS IDENTIFIED IN DRINKING WATER
IN THE UNITED STATES
' (MARCH 15, 1975)
Water Supply Research Laboratory
National Environmental Research Center, EPA
Cincinnati/ Ohio k5268
•j . acenaphthene
2. acenaphthylene
3. acetaldehyde
4. acetic acid
5. acetone •
6. acetophenone
7. acetylene dichloride
8. aldrin
9. afazlne
10. (deethyl ) atrazine
11. barbital
12. behenlc acid, methyl ester
13. benzaldehyde
14. benzene
15. benzene sulfonic acid
16. benzoic acid
17. benzopyene
18. benzothiazole
19. benzothiophene
20. benzyl butyl phthalate
21. bladex
22. borneol
23. bromobenzene
24. bromochlorobenzene
25. bromodichloromethane
26. bromoform
27. bromoform butanal
28. bromophenyl phenyl ether
29. butyl benzene
30. butyl bromide
31. camphor
32. e-caprolactam
33. carbon dioxide
34. carbon disulfide
35. carbon tetrach1 oride
36. chlordan(e)
37. chlordcne
38. chlorobenzene
VII-11
-------
39. It 2-bI s-ch'loroethoxy ethane
40. chloroethoxy ether
41. bi s-2-chloroethy1 ether
42. 2-chloroethyl methyl ether
43. chloroform
44. chlorohydroxybenzophenone
45. bis-chloroisopropyl ether
46. chloromothy 1 ether
47. chloromethy1 ethyl ether
48. ni-ch 1 oron i t robenzene
49. 1-chl orop^'ropene
50. 3-chloropyridine
51. o-cresol
r ') .- -.-•-, •-
J /- • • ' • <_
} , u - ri i , • \ . > "o' • '• i\ ' • • • -
65, die h 1 o r c cl i ,' 1 u or,, ; ! • • •
66. 1,2-dichlo-o-t1' •
G7. l,l-dichlo-o-2-:,". .--.T • = '
68. 2 / 't - d i c i; 1 o •• o p h e TI i j 1
69. d i ch 1 oropropanc1
70. 1,3-dichlo-opropcnc
71. d i e1d r i n
72. fii - (2-ethyl hexyl ) adipnto
73. diethyl benzone
74. die thy! phthai ate
75. cli (2-ethyl hexyl) phthalat
76. d i h e x y1 p h t h a 1 a t e
77. dihydrocarvone
78. di-isobutyl c a r b i n o1
79. di-isobutyl phthai ate
80. 1/2-dimethoxy benzene
81. 11 3-dimethy 1 naphtha]ene
82. 2,1*-dimethyl phenol
83, dimethyl phthalate •
84. dimethyl suIfoxide
85 . 1*, G-di n i t ro-2-ain i n op ho no I
VII-L2
-------
86. 1t 6-dfn!t^otoluene
87. dloctyl adipate
88. d}phenylhydrazine
89. dipropyl phthalate
90- docosane
91. n-dodecane
92. elcosane
93. endrin
94. ethanol
95. ethylamlne ' .
96. ethyl benzene
97. 2-ethyl~n-hexane
98. ds-2-ethyl -U-methyl-1,3-dioxol ane
99. trans-2-ethyl~ii-methyl-l,,3-dioxolane
100. o-ethyltoluene
101. m-ethylto!uene
102. p-ethylto!uene
•-103. geosmln
104. heptachlor
105. heptochlor epoxlde
106. l/2/3/U,5/7,7-heptachloronorbornene
107. hexachlorobenzene
108. hexachloro-l/3-butadiene
109. hexachlorocycl ohexane
110. hexachloroethane
HI. hexachlorophene
'l!2. hexadecane
113. 2-hydroxyadi poni t«*I le
'-114. Indene
115. Isoborneol
116. J sodecane
117. Isophorone
.'118. 1 -1 sopropenyl -U-I sopropylbenzene
119. Isopropy] benzene
120. llmonene
121. p-menth-1-en-8-ol
122. methane
123. methanol
124. 2-methoxy biphenyl
125. o-methoxyphcnol
126. methyl benzoate
127. methyl benzothfazole
VII-13
-------
128. methyl blphenyl
129. 3-methyl butanal
•130. methyl chloride
131. methylene chloride
132. methyl ethyl benzene
133. methyl ethyl ketone
134. 2-methy 1-5-ethy 1-pyridine
135. methylindene
136. methyl methacrylate
137. methyl naphthalene
138. methyl palmitate
139. methyl phenyl carbinol
140. 2-methy1propanal
141. methyl stearate
142. methyl tetracosanoato
143. naphthalene
144. nit roan i sole
.145. nl t robenzene
146. nonane
147. octadecane
148. octane
149. octyl chloride
150. pentachloroblphenyl
151. pentachlorophenol
152. pentachloropheny1 methyl ether
153. pentadecane
154. pentane
155. pentanol
156. phenyl benzoate
157. phthalic anhydride
158. pi per i dene
159. propanol
160. propaz i ne
161. propy1 ami ne
162. propylbenzene
163. sImaz i ne
"164. 1 / 1 , 3,3-tet rachloroacetone
165. tetrachlorobiphenyl
166. 1,1,\, 2-tetrachloroethane
167. tetrachloroethy1ene
168. tetradecane
169. tetramethy 1 benzene
170. thiomethy1benzothiazole
VII-14
-------
171. to!uene
172, tr|chlorobenzene
173. trfchlorobipheny1
174. 1 ,1 , 2-tr i chloroethane
175. 1,1,2-trichloroetiiylene
176. t«* f chlofof 1 uorome thane
177. 2,kf6-trichlorophenol
178. n-t»"Idecane
179. fimethy! benzene
180. 3,5,5-trimethyl-bicycle (U/i/0) heptene-2-one
181. trimethyl-tr!oxo-nexahydro-trl azlne Isomef
182. trlphenyl phosphate
•
183. n-undecane
184. vinyl benzene
185. • o-xylene
186. m~xylene
187. p-xylene
VII-15
-------
Oi
ATTACHMENT C.3
g
i_L^ ** ^j.
LD i
UO i £
0 0
*' — r~"~ 2
C T-
Q.
S- o rt
c > -? 'M
•— -^ J-j 4~>
OS- rt rj
•J3,S u H
•4—* CO
Z 1
e o
o o
<•£ *"*
' CM O
» -H
-1 Q
t5 °
b E
2: O
Q M
* m
s ^ 01
~* ^ so
m S ° ^
^ ~ ^ 0
^^ ,O *-H
2 ^ *H ,rC
2 hd Q u
2
•7- °
2 *-<
rr • o
U* O r-4
6 X
O 0
03 -O
C
rt
CxO
'M
0
C)
..-1 r-l
f-l --.
o to
u
O r-i
rt M
-C 3
o
E -i
r-t \
O t>C
<4-i 3
C
rt ' — (
4-) CiC
4) 3.
E
0
c
rt "™ *
,C ^
4_) d£
0 3
E
o
p '-i
rH -^
O 00
r— ( 3
^:
-i r\i v , v]- v j rxl ^ ,
•^
tL, CX LX ^U tL, CL, CM
Z v 2 2 2 2 v
*
CM CM
CL, O U. tL, tL, u, O
2 V 2 2 2 Z V
CL, i— i CL, <-< CL, CM CL,
2 V 2 2 2
vO O CTi
• IX
O O 2 ^O O vO CM
r-4 1—4
•
CO
G1 O O CM t^ CT) O
r-t ^f CM rH
, — i t^ -^t t-o CNJ r-^ CTS
Ol CTl O CM -^ LO
•"*
0
(/J
W M
4-> 4-> O
•*-» 3 to -o o *~5
CD O -M C O
W -H •*-> rt -Mi
3 4J - +-•
W C O 0) S-i X
(/) O rt "O 3 o Q
rt U "> O O 3 r-4
S w x: 2 a, r-.
« X 2 - >
tt> V< . ^ C
U 3 « 4J r-4 rt O
C J3 C ^4 O 3 -H
O rH O O >i ^5 rt
J-4 p_ 'O
3 -4-> w J 3 C :n
rt rt O <1> O rt (r3
J S 03 2 2 CO CL.
— i CM to r? in vo r-
o
v
tL,
2
tL,
2
tL,
2
to
OO
O
V
\O
,
O
,
^>^
^)
(/)
^
o
3
D
2
M
J_J
O
^>
• r-i
c;
10
E
O
H
00
f — I
CL,
2
CM
O
V
CL,
2
^j-
0
i— i
0
,r-H
•^
P_j
0
^^
3
O
2
•S
0
r— i
rt
(4-,
<— <
3
03
cn
r-4 r-1
U. CL,
Z 2
*
CM \D
u. tL,
Z 2
r-4 LO
r-l Cl
r-l
CTl \O
•* OO
rt
• H
pj
«
£»
I— 1
X
^ t/)
r-l C
0 C
0.
3
2 rt
•rt
•* S**
r^i P ,
O r-4
O O
.a -a
Z
•o
, — t
u.
2
1
O
V
(X
2
CM
00
f—4
^f
•
O
Q
•t
C
O
4-1
oo
c
•H
.c
Ul
rt
"•"
~
T~{
IX
2
LL,
2
CL,
2
CM
r-4
1— 1
CM
to
•o
rt
r-4
X
f-,
rt
s:
•,
O
f_4
o
e
•H
^>
r-4
rt
aa
tn
o
U,
2
tL,
2
U,
2
•«?•
.
O
CM
CO
fr
CJ
•H
j^
Cti
^
rH
X
c
c
o
D.
<»
rC
DO
L^
*J3
_r~^
t/;
4_j
4_i
•r-i
Cu
""O
O CM
U. CL.
CL, U.
Z Z
CL, tt.
Z Z
vO
CL,
Z 0
V
CL, \O
r-l
F
O '
V ^O
rt
•H
C
rt
^ rrt
r-4 .,_
x E
W -r-
C i.i
V -r-
a. >
•* 4ft
OO QJ
t* r-
^3 fO
Xl T3
tn c:
rt ns
!-4 C
w
M
i— 1
r-4
0)
3
O
o*
JD
—
VII-17
-------
"?
O
.— '
R
•r-l
O
U
v— • '
CM
T-H
r-l
£>
rt
H
i — <
CJ
O
H C
O
O J2
cd bo
,5 o e
O -H
> c
i rt
Ct £0
O 5-
Z 0
o
"O rH
C 1 TH ^
O rt r-l OO
JD r-l O P.
5H 4-> rH
rt o —
t_3 £— O
1
o
r-l
O O rH
rH C ^.
i f. rt oO
CM U ,C 3.
-•HP
rH Q O
O
e e --i
0 rH -^
r- O CO
cc
OJ
r-< rt ft
O 0
tn r^* vju i/> CM ^f" "^ ^" co
r-H\OCM i — li — li — IrH^-r-l
LL,Ci.CMP-.tL,tLitL,DL,tL,
•^f >~7- *-r i-r *-? •-->- •--r ~y
^-* t^-t *-• *--. •<-. «
-------
r— 1
cd
0 I
E— C :
0
0 J2
— ' CJ •— 1
C. W}
i— i O £
JD -H
•»•* C—
i cd
0 f-.'
Z 0
a>
•a rH
C 1 T-I •-«.
o cti f-i aoj
^3^03.
^ f-i r-i
CC O ,C
U E- 0
i
O
ft
O -H 4-1
i-H Q 0)
0
E E *-~*
O f-< *"^.
ft o oc
co <+-i a.
O 1
,O •— I *-> ^
• H X O
a o E
1
0
rH 0>
1 O C •-<
O r-l Cd \
e ,c -c c-o
O O 4-J 3.
f-l • r-l O
03 -O E
o
O --v
r-* ClO
o a
,— <
u
NO r^ o
CM o O
CT , f\J f r
z: v ?:
CN
•
ex, O ex.
Z v Z
u. ex.
Z ^t Z
t^J-
.
O i— I •— f
v _i ,-,
oo co r~~-
rr oo LO
to
o
• H
rC
O
.,
tn
^v^
fM
O
2
S-i
O
4_>
cd
O 2
•H
X O r-l
O -r-1 .— I
-C -H
•> o in
t/)
3
J3 C C
E O fd
— t >. -a
o cd c
U Q "-<
O — t CN
OJ r- 1
•* to
u, ex.
z z
CN1
•
o ex.
v Z
c^ ex.
z z
tO rH
V
tO LO
r-l
•-! O
to oo
o
•H
x:
o
o
"t C.
o o
4_)
•> 1/1
cd ejj
" d
cr 3
• rH O
P- >-
to »*
f~
._,
ex.
CM
0
V
ex.
Z
to
•
r^
CTl
C
• H
I/)
C
o
u
to
•H
2£
K
Ci)
0
.X
rs
ca
2
f— H
•H
•*•
LO
ro
to
ex.
(N
•
O
V
ex.
z
<•
•
o
v
*3-
\o
CN
j^
• H
l/l
C
O
u
l/l
•H
3:
^
_r^
tTj
O
r^
CO
O
£
CM LO
tO rH
t-. f " i
z; z
CJ
o ex.
Z
f-H fT ,
v Z
oo r-.
to c
O- cd
^;
(U fn
C <
C
O
.G c.
0 0
!- -3
ft E
CO
H U
r- oo
•cl- "3-
LO
tO
ex.
pr ,
Z
to
^*-
CN
cn
to
00
CM
cd
C
TO
-H
i/)
•H
^3
O
, ~]
«,
4J
L,
O
ex
c
cd
DO
C
cr>
LO
o
o
v
ex.
Z
ex.
Z
to
CM
i-(
^j-
.
0
o
o
X
Q^
2
^
4>
Z
**
0)
^3
cr
fn
0)
3
cr
3
J3
*—4
<
0
LO
CO
CM
CM
V
"»
•
O
v
vO
O
CN
CO
CM
*^
"*
cd
E
O
•£-
CO
i— i
\st
O
•t
X
4_)
•it
u
cd
E
O
X
»— (
JxC
o
•— 1
LO
. — 1
to
Cx,
ex.
z
CM
CTl
O
o
rM
^
to
CM
r-i
tfi
cd
X
o
E-
•t
O
f-H
r— 1
•H
^>
t/1
pj
S
0
^
03
CM
LO
cn
CM
ex.
z
ex.
Z
ex.
Z
CM
V
•*
OO
1— 1
I/)
cd
X
a>
E-
^
to
CM
ex.
Z
tx.
ex.
z
ex.
z
fx.
Z
00
•
o
cd
3
O
t— t
M
cd
^
4j
4->
O
*
LO
IO
C,
fv
c;
t^-
•e
•c *—*
r-j M
LL. £7^
z 2:
ex. ex.
z 2
ex. ex.
^—» ^^
P_ "^*
py
^ 2
r-l °0
»
cd
a 3
3 0
O •— <
(—1
^
« cd
Cd *C3
3 C
E -1-1
3 >-<
*J cd
•»-> rH
0 U
LO V0
LO LO
<• CM O
*3" CM tO
ex, to CM
z
^
xf tO
• *
o ex, o
v 2:
ex, to ex.
z z
V0
•
O CT» CM
V ^
00 00 .-4
tO CM
oo oo \o
CO CO r-l
•—I
•H
ft
3
o
w
V)
•H
2
C^ •»
S w 3
O cd cd
i— i (/i ID
C "0
•> CO f-l
4-> u: cd
f-4 f-4
O - 'H4
fXi cd o
cs _^
cu o o
J> f^4 fl,
cd o cd
Q E- U
f^ oO Crt
LO LO tO
0
I-l
'-
ex.
Z
ex.
Z
ex.
z
CM
OO
^^
CM
•H
f-l
3
O
w
in
•rH
2]
A
^^
^J
•r1^
^^
W
cd
in
C
rt
0
VII-19
-------
Q
O
/•"*•-
•u
o
^3
J=
•H
4-1
C!
O
C_}
v — '
rj
1—1
o
i — i
-C
rt
£—•
o
H c
O j
O ,0
• H rt — '
«tl/
tJ-O
rH O £
O -H
> c
i rt
£ 00
0 rH
2: o
O
C. f T-H \
O rt rH bC
rO rH O ^L
SH -P rH
rt o f.
c_> f- u
o
»-
O 0
S rt -^
o X oi;
•rH 4~> 3.
O O
o
e e -<
o P \
rH O tlC
co U-i 3.
O 0)
E O C
O rH ri rH
rH O X -^
• H X O J.
Q U g
,
O
rH O
i 0 C
O -H rt rH
E rC X ^
0 0 H-J Si
(H -rH O 3.
oa -a E
E
VH
O
0 rH
'rH ^
o «
•—I ^L
JZ
u
vO **f CN t^~~ vO CN
CN rH LO •— 1 r-< CN
r— 1
ti. :i. tL, U. a, u.
X x z. z Z z:
£
f
U- LL. U- U. a.
o z z z z x
rH CM tL, ti, U, CO
V V Z Z Z
U-
(O ^3" X to CN C7i
V •*
tO O ' — ! O CN \O
rH rH rH
r-l
LO ^f tO TJ- CN CTl
LO r-l O
to
LU CM
'— ' r-l
LO
CT>QO^CT)CTitOCOvOCTlCOtO*3-rHl
OOrHCNr- ICNi— ICNIi— lOCNOOtOLO
V o
o
\/
f)
U,UL,H,ti,ti.U-LL.tUU.tl.U-,U-. J.L. i . 1
^-•1 «*^* -^-4 ^— -^-l X" , X-< ••- -1 -^ X" lr<— . If , f - • ,* - t
f ! f f- — ' i ' i '. " '1. ' " h •
j'* '''*/""' ' ' •• '"
I •!
iy, f~j^
'^ t'". -^- C i r ! T U,, fO '' U, P 1 , ^-L l~. (
X r-< V V X OX X X X X J-
V X
•trg' *~j
O
Cl-t CO tO fl , LO rH
cocNr--Loox''tocTirj-2;v 2; >
i— 1 f— 1 tO t-( ji,
~H
CM rH
to
OOCTiO' — '• — iCNCNr— ILOOOl
' ' r-l
1-t
O
V
VII-20
• H
3
o
10
•H
«y
•,
w
•H
O
,
1 t
co
^
rt
t/i
rt
rH
J3
O
z
^
P;
r — 1
O
o
p;
•r-i
^
s
rt
4_)
O
rX
rt
Q
4-J
rn
O
^r
A
C/l"
^V^
^
o
tt.
ID
C
CO
U
rO
O
-a
rt
^
o
^H
o
*
rH
o
^>
c
o
Q
s
0
•a
rt
rH
o
p — 1
o
u
^
o
rH
^r"i
o
-J
Cu
LO
ri
4J
o
rt
Q
^r"|
4_j
r3
o
in
«
c
p
i-*
a:
o
vO
X rt
re C
4-> O
:D txi
•rH
(H
O <
_v
rt
-J c
o
4-J O
I — 1 t/)
rt 3
co H
t^ CO
rt
C
O
N
.rH,
fa
<^
VI
X
•H
C
o
o
x
a.
CTi
rt
•H
£
o
SH
• H
, — 1
rt
C__J
-
rt
^0
ci;
• rH
,— H
.7^
o
u
o
rt
•H
c
rH
C
CTH
•H
. — i
rt
U
«N
"O
t-l
o
u
c
o
u
rH
rt
•H
c
r-l
0
*— t
O
CO
c
<^
'-0
0
ro
rt
•H
C
fH
O
<4H
•M
rH
rt
CJ
-.
O
00
O
•rH.
Q
C
rt
CO
rr
r-
cci
•H
ct
^
0
•H
rH
rt
U
•*
O
0
tn
•H
O
C
rt
rH
U-,
c
rt
co
LO
p:
o
bO
P:
•H
rC
rt
^i
•X
O
r — 1
4_)
•I-J
rt
o
co
MD
rt
-X
rt
, — i
<*H
•i
Ul
rt
• — i
00
^3
O
Q
p:
o
rt
•a
rH
•t
U1
i — 1
» — i
rt
u.
O
f—
rt
•a
> — i
CO
c
0
M
O
rH
O
«\
lA
•H
. — 1
i-~t
Cj
?*
SH
O
u
en
o
O
•H
ao
r^
•H
rx^
•yj
rt
3c
o
- so
O c
U tT3
C3 CiC
^
rH
K-<
O
CO
„
(-.
•^
ul
CJ
c
o
*
.
fe
-------
ATTACHMENT D
Unpublished Documents provided to the Ad Hoc Study Group
De Rouen, T. A. and J. E. Dieir«. The New Orleans Drinking Water
Controversy: A Statistical Perspective. 1975.
De Rouen, T. A. and J. E. Dien. Ethnic, Geographical Differences
in Cancer Mortality in Louisiana. 1975.
Dressman, R. C. and E. F. McFarren. The Detection and Measurement
of Bis (2 Chloro) ethers and Diedrin by Gas Chromatography.
December 1974.
E.P.A., Region IV. Draft Analytical Report - New Orleans Area
Water Supply Study. November 1974.
Harris, Robert H. The Implications of Cancer-Causing Substances
in Mississippi River Water. November 6, 1974.
Harris, Robert H. Results of Further Statistical Studies-Memo to
Ad Hoc Study Group to Consider Organics in Drinking Water.
April 17, 1975.
Kraybill, H. F. The Distribution of Chemical Carcinogens in Aquatic
Environments. October 1974.
Kraybill, H.F., Weisburger, E. K., and T. Page. Evaluation of
Biorefractories in the New Orleans Area Water Supply.
February 20, 1975.
Love, 0. Thomas Jr., Carswell, J. Keith, Stevens, Alan A., and
Symons., J. M., Evaluation of Activated Carbon as a Drinking
Water Treatment Process, ]'regress Report. March 1975.
Miller, Robert D. Evaluation of the report by Robert H. Harris, Ph.D.,
of the Environmental Defense Fund in Cancer Causing Substances
in Mississippi River Water. December 18, 1974.
Morris, J. Carrel. Formation of Halogenated Organics by
Chlorination of Water Supplies, Draft Report., February 1, 1975.
Page, T., Harris, R. H., and Epstein, S. S. Relation Between Cancer
Mortality and Drinking Wafer in Louisiana. Draft undated.
Page, Talbot and Harris, R. H. Implications of Cancer-Causing
Substances in Mississippi River Water: A reply to
reviews by R. W. Miller, I:. E. Tarone, and J. J. Gart. March 1975.
\II-21
-------
Tarone, R. E., and Gart, J. J "Review of the Implications of Cancer
Causing Substances in Mississippi River Water," by Robert J.
Harris of the Environmental Defense Fund, January 10, 1975.
Train, Russell E. Statement by Environmental Protection Agency
Administrator Russell E. Train at a Press Conference on
Results of the 80 City Drinking Water Survey, April 18, 1975.
r •-, -
VII-22
------- |