EPA-815-Z-98-005
Tuesday
March 31, 1998
= i
Part IV
Environmental
Protection Agency
40 CFR Parts 141 and 142
National Primary Drinking Water
Regulations: Disinfectants and
Disinfection Byproducts Notice of Data
Availability; Proposed Rule
-------
-------
15674
Federal Register/VoI. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR Parts 141 and 142
[WH-FRL-5988-7]
National Primary Drinking Water
Regulations: Disinfectants and
Disinfection Byproducts Notice of Data
Availability
AGENCY: U.S. Environmental Protection
Agency (USEPA).
ACTION: Notice of data availability;
request for comments.
SUMMARY: In 1994 USEPA proposed a
Stage 1 Disinfectants/Disinfection
Byproducts Rule (D/DBP) to reduce the
level of exposure from disinfectants and
disinfection byproducts (DBFs) in
drinking water (USEPA, 1994a). This
Notice of Data Availability summarizes
the 1994 proposal and a subsequent
Notice of Data Availability in 1997
(USEPA, 1997a); describes new data that
the Agency has obtained and analyses
that have been completed since the 1997
Notice of Data Availability; requests
comments on the regulatory
implications that flow from the new
data and analyses; and requests
comments on several issues related to
the simultaneous compliance with the
Stage 1 DBF Rule and the Lead and
Copper Rule. USEPA solicits comment
on all aspects of this Notice and the
supporting record. The Agency also
solicits additional data and information
that may be relevant to the issues
discussed in the Notice.
The Stage 1 D/DBP rule would apply
to community water systems and
nontransient noncommunity water
systems that treat their water with a
chemical disinfectant for either primary
or residual treatment. In addition,
certain requirements for chlorine
dioxide would apply to transient
noncommunity water systems because
of the short-term health effects from
high levels of chlorine dioxide.
Key issues related to the Stage 1 D/
DBF rule that are addressed in this
Notice include the establishment of
Maximum Contaminant Level Goals for
chloroform, dichloroacetic acid,
chlorite, and bromate and the Maximum
Residual Disinfectant Level Goal for
chlorine dioxide.
DATES: Comments should be postmarked
or delivered by hand on or before April
30, 1998. Comments must be received or
post-marked by midnight April 30,
1998.
ADDRESSES: Send written comments to
DBF NODA Docket Clerk, Water Docket
(MC-4101); U.S. Environmental
Protection Agency; 401 M Street, SW.,
Washington, DC 20460. Comments may
be hand-delivered to the Water Docket,
U.S. Environmental Protection Agency;
401 M Street, SW., East Tower
Basement, Washington, DC 20460.
Comments may be submitted
electronically to
owdocket@epamail.epa.gov.
FOR FURTHER INFORMATION CONTACT: For
general information contact, the Safe
Drinking Water Hotline, Telephone
(800) 426-4791. The Safe Drinking
Water Hotline is open Monday through
Friday, excluding Federal holidays,
from 9:00 a.m. to 5:30 p.m. Eastern
Time. For technical inquiries, contact
Dr. Vicki Dellarco, Office of Science and
Technology (MC 4304) or Mike Cox,
Office of Ground Water and Drinking
Water (MC 4607), U.S. Environmental
Protection Agency, 401 M Street SW.,
Washington DC 20460; telephone (202)
260-7336 (Dellarco) or (202) 260-1445
(Cox).
SUPPLEMENTARY INFORMATION:
Regulated entities. Entities potentially
regulated by the Stage 1 D/DBP rule are
public water systems that add a
disinfectant or oxidant. Regulated
categories and entities include:
Category
Public
Water
System.
State Gov-
ern-
ments.
Examples of regulated entities
Community and nontransient
noncommunity water systems
that add a disinfectant or oxi-
dant.
State government offices that
regulate drinking water.
This table is not intended to be
exhaustive, but rather provides a guide
for readers regarding entities likely to be
regulated by this action. This table lists
the types of entities that EPA is now
aware could potentially be regulated by
this action. Other types of entities not
listed in this table could also be
regulated. To determine whether your
facility may be regulated by this action,
you should carefully examine the
applicability criteria in § 141.130 of the
proposed rule (USEPA, 1994a). If you
have questions regarding the
applicability of this action to a
particular entity, contact one of the
persons listed in the preceding FOR
FURTHER INFORMATION CONTACT section.
Additional Information for
Commenters. Please submit an original
and three copies of your comments and
enclosures (including references). The
Agency requests that commenters follow
the following format: Type or print
comments in ink, and cite, where
possible, the paragraph® in this Notice
to which each comment refers.
Commenters should use a separate
paragraph for each method or issue
discussed. Electronic comments must be
submitted as a WP5.1 or WP6.1 file or
as an ASCII file avoiding the use of
special characters. Comments and data
will also be accepted on disks in
WordPerfect in 5.1 or WP6.1 or ASCII
file format. Electronic comments on this
Notice may be filed online at many
Federal Depository Libraries.
Commenters who want EPA to
acknowledge receipt of their comments
should include a self-addressed,
stamped envelope. No facsimiles (faxes)
will be accepted.
Availability of Record. The record for
this Notice, which includes supporting
documentation as well as printed, paper
versions of electronic comments, is
available for inspection from 9 to 4 p.m.
(Eastern Time), Monday through Friday,
excluding legal holidays, at the Water
Docket, U.S. EPA Headquarters, 401 M.
St., S.W., East Tower Basement,
Washington, D.C. 20460. For access to
docket materials, please call 202/260-
3027 to schedule an appointment.
Abbreviations Used in Tills Notice
AWWA: American Water Works
Association
AWWARF: AWWA Research
Foundation
BAT: Best Available Technology
BDCM: Bromodichloromethane
CMA: Chemical Manufacturers
Association
CWS: Community Water System
DBCM: Dibromochloromethane
DBF: Disinfection Byproducts
D/DBP: Disinfectants and Disinfection
Byproducts
DCA: Dichloroacetic Acid
EDi0: Maximum likelihood estimate on
a dose associated with 10% extra risk
EPA: United States Environmental
Protection Agency
ESWTR: Enhanced Surface Water
Treatment Rule
FACA: Federal Advisory Committee Act
GAC: Granular Activated Carbon
HAAS: Haloacetic Acids (five)
HAN: Haloacetonitrile
ICR: Information Collection Rule
ILSI: International Life Sciences
Institute
IESTWR: Interim Enhanced Surface
Water Treatment Rule
IRFA: Initial Regulatory Flexibility
Analysis
LCR: Lead and Cooper Rule
LEDio: Lower 95% confidence limit on
a dose associated with 10% extra risk
LMS: Linear Multistage Model
LOAEL: Lowest Observed Adverse
Effect Level
LTESTWR: Long-Term Enhanced
Surface Water Treatment Rule
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15675
MCL: Maximum Contaminant Level
MCLG: Maximum Contaminant Level
Goal
M-DBP: Microbial and Disinfectants/
Disinfection Byproducts
mg/L: Milligrams per liter
MoE: Margin of Exposure
MRDL: Maximum Residual Disinfectant
Level
MRDLG: Maximum Residual
Disinfectant Level Goal
MTD: Maximum Tolerated Dose
NIPDWR: National Interim Primary
Drinking Water Regulation
NOAEL: No Observed Adverse Effect
Level
NOD A: Notice of Data Availability
NPDWR: National Primary Drinking
Water Regulation
NTNCWS: Nontransient Noncommunity
Water System
NTP: National Toxicology Program
PAR: Population Attributable Risk
PQL: Practical Quantitation Limit
PWS: Public Water System
q 1 *: Cancer Potency Factor
RFA: Regulatory Flexibility Act
RfD: Reference Dose
RIA: Regulatory Impact Analysis
RSC: Relative Source Contribution
SAB: Science Advisory Board
SBREFA: Small Business Regulatory
Enforcement Fairness Act
SDWA: Safe Drinking Water Act, or the
"Act," as amended in 1986 and 1996
SWTR: Surface Water Treatment Rule
TCA: Trichloroacetic Acid
TOG: Total Organic Carbon
TTHM: Total Trlhalomethanes
TWG: Technical Working Group
Table of Contents
I. Introduction and Background
A. 1979 Total Trihalomethane MCL
B, Statutory Authority
C. Regulatory Negotiation Process
D, Overview of 1994 DBF Proposal
1. MCLGs/MCLs/MRDLGs/MRDLs
2. Best Available Technologies
3. Treatment Technique
4. PreoxidaUon (Predisinfection) Credit
5. Analytical Methods
6. Effect on Small Public Water Systems
E. Formation of 1997 Federal Advisory
Committee
H. Significant New Epidemiology Information
for the Stage 1 Disinfectants and
Disinfection Byproducts Rule
A. Epldemlological Associations Between
the Exposure to DBFs in Chlorinated
Water and Cancer
1. Assessment of the Morris etal. (1992)
Meta-Analysis
a. Poole Report
b. EPA's Evaluation of Poole Report
c. Peer Review of Poole Report and EPA's
Evaluation
2, New Cancer Epidemiology Studies
3. Quantitative Risk Estimation for Cancers
From Exposure to Chlorinated Water
4. Peer-Review of Quantitative Risk
Estimates
5. Summary of Key Observations
6. Requests for Comments
B. Epidemiological Associations Between
Exposure to DBFs in Chlorinated Water
and Adverse Reproductive and
Developmental Effects
1. EPA Panel Report and
Recommendations for Conducting
Epidemiological Research on Possible
Reproductive and Developmental Effects
of Exposure to Disinfected Drinking
Water
2. New Reproductive Epidemiology
Studies
3. Summary of Key Observations
4. Request ifor Comments
ffi. Significant New Toxicological
Information for the Stage 1 Disinfectants
and Disinfection Byproducts
A. Chlorite and Chlorine Dioxide
1. 1997 CMA Two-Generation
Reproduction Rat Study
2. External Peer Review of the CMA Study
3. MCLG for Chlorite: EPA's Reassessment
of the Noncancer Risk
4. MRDLG for Chlorine Dioxide: EPA's
Reassessment of the Noncancer Risk
5. External Peer Review of EPA's
Reassessment
6. Summary of Key Observations
7. Request for Comments
B. Trihalomethanes
1. 1997 International Life Sciences Institute
Expert Panel Conclusions for Chloroform
2. MCLG for Chloroform: EPA's
Reassessment of the Cancer Risk
a. Weight of the Evidence and
Understanding of the Mode of
Carcinogenic Action
b. Dose-Response Assessment
3. External Peer Review of EPA's
Reassessment
4. Summary of Key Observations
5. Requests for Comments
C. Haloacetic Acids
1.1997 International Life Sciences Institute
Expert Panel Conclusions for
Dichloroacetic Acid (DCA)
2. MCLG for DCA: EPA's Reassessment of
the Cancer Hazard
3. External Peer Review of EPA's
Reassessment
4. Summary of Key Observations
5. Requests for Comments
D. Bromate
1.1998 EPA Rodent Cancer Bioassay
2. MCLG for Bromate: EPA's Reassessment
of the Cancer Risk
3. External Peer Review of EPA's
Reassessment
4. Summary of Key Observations
5. Requests for Comments
IV. Simultaneous Compliance
Considerations: D/DBP Stage 1 Enhanced
Coagulation Requirements and the Lead
and Copper Rule
V. Compliance with Current Regulations
VI. Conclusions
VH. References
I. Introduction and Background
A. 1979 Total Trihalomethane MCL
USEPA set an interim maximum
contaminant level (MCL) for total
trihalomethanes (TTHMs) of 0.10 mg/L
as an annual average in November 1979
(USEPA, 1979). There are four
trihalomethanes (chloroform,
bromodichloromethane,
chlorodibromomethane, and
bromoform). The interim TTHM
standard applies to any PWS (surface
water and/or ground water) serving at
least 10,000 people that adds a
disinfectant to the drinking water
during any part of the treatment process.
At their discretion, States may extend
coverage to smaller PWSs. However,
most States have not exercised this
option. About 80 percent of the PWSs,
serving populations of less than 10,000,
are served by ground water that is
generally low in THM precursor content
(USEPA, 1979) and which would be
expected to have low TTHM levels even
if they disinfect.
B. Statutory Authority
In 1996, Congress reauthorized the
Safe Drinking Water Act. Several of the
1986 provisions were renumbered and
augmented with additional language,
while other sections mandate new
drinking water requirements. As part of
the 1996 amendments to the Safe
Drinking Water Act, USEPA's general
authority to set a Maximum
Contaminant Level Goal (MCLG) and a
National Primary Drinking Water
Regulation (NPDWR) was modified to
apply to contaminants that "may have
an adverse effect on the health of
persons", that are "known to occur or
there is a substantial likelihood that the
contaminant will occur in public water
systems with a frequency and at levels
of public health concern", and for
which "in the sole judgement of the
Administrator, regulation of such
contaminant presents a meaningful
opportunity for health risk reduction for
persons served by public water systems'
(1986 SDWA Section 1412 (b)(3)(A)
stricken and amended with
The Act also requires that at the same
time USEPA publishes an MCLG, which
is a non-enforceable health goal, it also
must publish a NPDWR that specifies
either a maximum contaminant level
(MCL) or treatment technique (Sections
1401(1), 1412(a)(3),and 1412 (b)(4)B)).
USEPA is authorized to promulgate a
NPDWR "that requires the use of a
treatment technique in lieu of
establishing a MCL," if the Agency finds
that "it is not economically or
technologically feasible to ascertain the
level of the contaminant"
The 1996 Amendments also require
USEPA to promulgate a Stage 1
disinfectants/disinfection byproducts
(D/DBP) rule by November 1998. In
-------
15676
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
addition, the 1996 Amendments require
USEPA to promulgate a Stage 2 D/DBP
rule by May 2002 (Section
C. Regulatory Negotiation Process
In 1992 USEPA initiated a negotiated
rulemaking to develop a D/DBP rule.
The negotiators included
representatives of State and local health
and regulatory agencies, public water
systems, elected officials, consumer
groups and environmental groups. The
Committee met from November 1992
through June 1993.
Early in the process, the negotiators
agreed that large amounts of information
necessary to understand how to
optimize the use of disinfectants to
concurrently minimize microbial and
DBF risk on a plant-specific basis were
unavailable. Nevertheless, the
Committee agreed that USEPA should
propose a D/DBP rule to extend
coverage to all community and
nontransient noncommunity water
systems that use disinfectants. This rule
proposed to reduce the current TTHM
MCL, regulate additional disinfection
byproducts, set limits for the use of
disinfectants, and reduce the level of
organic compounds from the source
water that may react with disinfectants
to form byproducts.
One of the major goals addressed by
the Committee was to develop an
approach that would reduce the level of
exposure from disinfectants and DBFs
without undermining the control of
microbial pathogens. The intention was
to ensure that drinking water is
microbiologically safe at the limits set
for disinfectants and DBFs and that
these chemicals do not pose an
unacceptable risk at these limits.
Following months of intensive
discussions and technical analysis, the
Committee recommended the
development of three sets of rules: a
staged D/DBP Rule (proposal: 59 FR
38668, July 29, 1994), an "interim"
Enhanced Surface Water Treatment Rule
(IESWTR) (proposal: 59 FR 38832, July
29, 1994), and an Information Collection
Rule (final 61 FR 24354, May 14, 1996).
The IESWTR would only apply to
systems serving 10,000 people or more.
The Committee agreed that a "long-
term" ESWTR (LTESWTR) would be
needed for systems serving fewer than
10,000 people when the results of more
research and water quality monitoring
became available. The LTESWTR could
also include additional refinements for
larger systems.
D. Overview of 1994 DBF Proposal
The proposed D/DBP Stage 1 rale
addressed a number of complex and
interrelated drinking water issues. The
proposal attempted to balance the
control of health risks from compounds
formed during drinking water
disinfection against the risks from
microbial organisms (such as Giardia
lamblia, Cryptosporidium, bacteria, and
viruses) to be controlled by the IESWTR.
The proposed Stage 1 D/DBP rule
applied to all community water systems
(CWSs) and nontransient
noncommunity water systems
(NTNCWSs) that treat their water with
a chemical disinfectant for either
primary or residual treatment. In
addition, certain requirements for
chlorine dioxide would apply to
transient noncommunity water systems
because of the short-term health effects
from high levels of chlorine dioxide.
Following is a summary of key
components of the 1994 proposed Stage
1 D/DBP rule.
1. MCLGs/MCLs/MRDLGs/MRDLs
EPA proposed MCLGs of zero for
chloroform, bromodichloromethane,
bromoform, bromate, and dichloroacetic
acid and MCLGs of 0.06 mg/L for
dibromochloromethane, 0.3 mg/L for
trichloroacetic acid, 0.04 mg/L for
chloral hydrate, and 0.08 mg/L for
chlorite. In addition, EPA proposed to
lower the MCL for TTHMs from 0.10 to
0.080 mg/L and added an MCL for five
haloacetic acids (i.e., the sum of the
concentrations of mono-, di-, and
trichloroacetic acids and mono-and
dibromoacetic acids) of 0.060 mg/L.
EPA also, for the first time, proposed
MCLs for two inorganic DBFs: 0.010 mg/
L for bromate and 1.0 mg/L for chlorite.
In addition to proposing MCLGs and
MCLs for several DBFs, EPA proposed
maximum residual disinfectant level
goals (MRDLGs) of 4 mg/L for chlorine
and chloramines and 0.3 mg/L for
chlorine dioxide. The Agency also
proposed maximum residual
disinfectant levels (MRDLs) for chlorine
and chloramines of 4.0 mg/L, and 0.8
mg/L for chlorine dioxide. MRDLs
protect public health by setting limits
on the level of residual disinfectants in
the distribution system. MRDLs are
similar in concept to MCLs—MCLs set
limits on contaminants and MRDLs set
limits on residual disinfectants in the
distribution system. MRDLs, like MCLs,
are enforceable, while MRDLGs, like
MCLGs, are not enforceable.
2. Best Available Technologies
EPA identified the best available
technology (BAT) for achieving
compliance with the MCLs for both
TTHMs and HAA5 as enhanced
coagulation or treatment with granular
activated carbon with a ten minute
empty bed contact time and 180 day
reactivation frequency (GAC10), with
chlorine as the primary and residual
disinfectant. The BAT for achieving
compliance with the MCL for bromate
was control of ozone treatment process
to reduce formation of bromate. The
BAT for achieving compliance with the
chlorite MCL was control of precursor
removal treatment processes to reduce
disinfectant demand, and control of
chlorine dioxide treatment processes to
reduce disinfectant levels. EPA
identified BAT for achieving
compliance with the MRDL for chlorine,
chloramine, and chlorine dioxide as
control of precursor removal treatment
processes to reduce disinfectant
demand, and control of disinfection
treatment processes to reduce
disinfectant levels.
3. Treatment Technique
EPA proposed a treatment technique
that would require surface water
systems and groundwater systems under
the direct influence of surface water that
use conventional treatment or
precipitative softening to remove DBF
precursors by enhanced coagulation or
enhanced softening. A system would be
required to remove a certain percentage
of total organic carbon (TOC) (based on
raw water quality) prior to the point of
continuous disinfection. EPA also
proposed a procedure to be used by a
PWS not able to meet the percent
reduction, to allow them to comply with
an alternative minimum TOC removal
level. Compliance for systems required
to operate with enhanced coagulation or
enhanced softening was based on a
running annual average, computed
quarterly, of normalized monthly TOC
percent reductions.
4. Preoxidation (Predisinfection) Credit
The proposed rule did not allow
PWSs to take credit for compliance with
disinfection requirements in the SWTR/
IESWTR prior to removing required
levels of precursors unless they met
specified criteria. This provision was
modified by the 1997 Federal Advisory
Committee (see below).
5. Analytical Methods
EPA proposed nine analytical
methods (some of which can be used for
multiple analyses) to ensure compliance
with proposed MRDLs for chlorine,
chloramines, and chlorine dioxide. EPA
proposed methods for the analysis of
TTHMs, HAAS, chlorite, bromate and
total organic carbon.
6. Effect on Small Public Water Systems
The Regulatory Flexibility Act (RFA),
as amended by the Small Business
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15677
Regulatory Enforcement Fairness Act
(SBREFA), requires federal agencies, in
certain circumstances, to consider the
economic effect of proposed regulations
on small entities. The agency must
assess the economic impact of a
proposed rule on small entities if the
proposal will have a significant
economic impact on a substantial
number of small entitles. Under the
RFA, 5 U.S.C. 601 etseq., an agency
must prepare an initial regulatory
flexibility analysis (IRFA) describing the
economic impact of a rule on small
entitles unless the agency certifies that
the rule will not have a significant
Impact.
In the 1994 D/DBP and IESWTR
proposals, EPA defined small entities as
small PWSs—serving 10,000 or fewer
persons—for purposes of its regulatory
flexibility assessments under the RFA.
EPA certified that the IESWTR will not
have a significant impact on a
substantial number of small entities,
and prepared an IRFA for the DBP
proposed rule. EPA did not, however,
specifically solicit comment on that
definition. EPA will use this same
definition of small PWSs in preparing
the final RFA for the Stage 1 DBP rule.
Further, EPA plans to define small
entitles in the same way in all of its
future drinking water rulemakings. The
Agency solicited public comment on
this definition in the proposed National
Primary Drinking Water Regulations:
Consumer Confidence Reports, 63 FR
7606, at 7620-21, February 13, 1998.
E. Formation of 1997 Federal Advisory
Committee
In May 1996, the Agency initiated a
series of public Informational meetings
to exchange Information on issues
related to mlcrobial and D/DBP
regulations. To help meet the deadlines
for the IESWTR and Stage 1 D/DBP rule
established by Congress in the 1996
SDWA Amendments and to maximize
stakeholder participation, the Agency
established the Microbial and
Disinfectants/Disinfection Byproducts
(M-DBP) Advisory Committee under the
Federal Advisory Committee Act
(FACA) on February 12, 1997, to collect,
share, and analyze new information and
data, as well as to build consensus on
the regulatory Implications of this new
information. The Committee consists of
17 members representing USEPA, State
and local public health and regulatory
agencies, local elected officials, drinking
water suppliers, chemical and
equipment manufacturers, and public
interest groups.
The Committee met five times, in
March through July 1997, to discuss
issues related to the IESWTR and Stage
1 D/DBP rule. Technical support for
these discussions was provided by a
Technical Work Group (TWG)
established by the Committee at its first
meeting in March 1997. The
Committee's activities resulted in the
collection, development, evaluation,
and presentation of substantial new data
and information related to key elements
of both proposed rules. The Committee
reached agreement on the following
major issues that were discussed in the
1997 NODA (USEPA, 1997a): (1)
Maintaining the proposed MCLs for
TTHMs, HAAS and bromate; (2)
modifying the enhanced coagulation
requirements as part of DBP control; (3)
including a microbial bench marking/
profiling to provide a methodology and
process by which a PWS and the State,
working together, assure that there will
be no significant reduction in microbial
protection as the result of modifying
disinfection practices in order to meet
MCLs for TTHM and HAAS; (4) credit
for compliance with applicable
disinfection requirements should
continue to be allowed for disinfection
applied at any point prior to the first
customer, consistent with the existing
Surface Water Treatment Rule; (5)
modification of the turbidity
performance requirements and
requirements for individual filters; (6)
issues related to the MCLG for
Cryptosporidiunr, (7) requirements for
removal of Cryptosporidiunr, and (8)
provision for conducting sanitary
surveys.
n. Significant New Epidemiology
Information for the Stage 1 Disinfectant
and Disinfection Byproducts Rule
The preamble to the 1994 proposed
rule provided a summary of the health
criteria documents for the following
DBFs: Bromate; chloramines; haloacetic
acids and chloral hydrate; chlorine;
chlorine dioxide, chlorite, and chlorate;
and trihalomethanes (USEPA, 1994a).
The information presented in 1994 was
used to establish MCLGs and MRDLGs.
On November 3, 1997, the EPA
published a Notice of Data Availability
(NODA) summarizing new information
that the Agency has obtained since the
1994 proposed rule (USEPA, 1997a).
The following sections briefly discuss
additional information received and
analyzed since the November 1997
NODA. This new information concerns
the following: (1) Recently published
epidemiology studies examining the
relationship between exposure to
contaminants in chlorinated surface
water and adverse health outcomes; (2)
an assessment of the Morris et. al. (1992)
meta-analysis of the epidemiology
studies published prior to 1996; (3)
recommendations made by an
International Life Science Institute
(ILSI) expert panel on the application of
the USEPA Proposed Guidelines for
Carcinogen Assessment (USEPA, 1996b)
to data sets for chloroform and
dichloroacetic acid; and (4) new
laboratory animal studies on bromate
and chlorite (also applicable to chlorine
dioxide risk). This Notice presents the
conclusions of these supplemental
analyses as well as their implications for
MCLGs, MCLs, MRDLGs, and MRDLs.
The new documents are included in the
Docket for this action.
As a result of this new information,
the EPA requests comment on the
following: (1) Revisions to estimates of
potential cancer cases that can be
attributed to exposure from DBFs in
chlorinated surface water (USEPA,
1998a); (2) revisions to the noncancer
assessment for chlorite and chlorine
dioxide (USEPA, 1998b); (3) revisions to
the cancer quantitative risks for
chloroform (USEPA, 1998c); (4) updates
on the cancer assessment for bromate
(USEPA, 1998d); and (5) updates on the
hazard characterization for
dichloroacetic acid (USEPA, 1998e).
As in 1994, the assessment of public
health risks from chlorination of
drinking water currently relies on
inherently difficult and incomplete
empirical analysis. On one hand,
epidemio logic studies of the general
population are hampered by difficulties
of design, scope, and sensitivity. On the
other hand, uncertainty is involved in
using the results of high dose animal
toxicological studies of a few of the
numerous byproducts that occur in
disinfected drinking water to estimate
the risk to humans from chronic
exposure to low doses of these and other
byproducts. In addition, such studies of
individual byproducts cannot
characterize the entire mixture of
disinfection byproducts in drinking
water. Nevertheless, while recognizing
the uncertainties of basing quantitative
risk estimates on less than
comprehensive information regarding
overall hazard, EPA believes that the
weight-of-evidence represented by the
available epidemiological and
toxicological studies on DBFs and
chlorinated surface water continues to
support a hazard concern and a
protective public health approach to
regulation.
A. Epidemiologic Associations Between
Exposure to DBFs in Chlorinated Water
and Cancer
The preamble to the 1994 proposed
rule discussed several cancer
epidemiology studies that had been
conducted over the past 20 years to
-------
15678
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
examine the association between
exposure to chlorinated water and
cancer (USEPA, 1994a). At the time of
the 1994 proposed rule, there was
disagreement among the members of the
Negotiating Committee on the
conclusions that could be drawn from
these studies. Some members of the
Committee felt that the cancer
epidemiology data, taken in conjunction
with the results from toxicological
studies, provided ample and sufficient
weight of evidence to conclude that
exposure to DBFs in drinking water
could result in increased cancer risk at
levels encountered in some public water
supplies. Other members of the
Committee concluded that the cancer
epidemiology studies on the
consumption of chlorinated drinking
water to date were insufficient to
provide definitive information for the
regulation. As a response, EPA agreed to
pursue additional research to reduce the
uncertainties associated with these data
and to better characterize and project
the potential human cancer risks
associated with the exposure to
chlorinated water. To implement this
commitment, EPA sponsored an expert
panel to review the state of cancer
epidemiology research (USEPA, 1994b).
As discussed in the 1997 NOD A, EPA
has implemented several of the panel's
recommendations for short-and long-
term research to improve the assessment
of risks, using the results from cancer
epidemiology studies.
The 1994 proposed rule also
presented the results of a meta-analysis
that pooled the relative risks from ten
cancer epidemiology studies in which
there was a presumed exposure to
chlorinated water and its byproducts
(Morris et al., 1992). A conclusion of
this meta-analysis was a calculated
upper bound estimate of approximately
10,000 cases of rectal and bladder
cancer cases per year that could be
associated with exposure to chlorinated
water and its byproducts in the United
States. The ten studies included in the
meta-analysis had methodological
issues and significant design
differences. There was considerable
debate among the members of the
Negotiating Committee on the extent to
which the results of this meta-analysis
should be considered in developing
benefit estimates associated with the
proposed rule. Negotiators agreed that
the range of possible risks attributed to
chlorinated water should consider both
toxicological data and epidemiological
data, including the Morris eta/. (1992)
estimates. No consensus, however,
could be reached on a single likely risk
estimate.
For purposes of estimating the
potential benefits from the proposed
rule, EPA used a range of estimated
cancer cases that could be attributed to
exposure to chlorinated waters of less
than 1 cancer case per year up to 10,000
cases per year. The less than 1 cancer
case per year was based on toxicology
(the maximum likelihood cancer risk
estimate calculated from animal assay
data for THMs). The 10,000 cases per
year was based on epidemiology
(estimates from the Morris et al. (1992)
meta-analysis).
1. Assessment of the Morris et al. (1992)
Meta- Analysis
Based on the recommendations from
the 1994 expert panel on cancer
epidemiology, EPA completed an
assessment of the Morris et al. (1992)
meta-analysis which comprises three
reports: (1) A Report completed for EPA
which evaluated the Morris et al. (1992)
meta-analysis (Poole, 1997); (2) EPA's
assessment of the Poole report (USEPA,
1998f); and (3) a peer review of the
Poole report and EPA's assessment of
the Poole report (USEPA, 1998g). Each
of these documents is briefly discussed
below. The full reports together with Dr.
Morris's comments on the Poole Review
(Morris, 1997) can be found in the
docket for this Notice.
a. Poole Report. A report was
prepared for EPA which made
recommendations regarding whether the
data used by Morris et al. (1992) should
be aggregated into a single summary
estimate of risk. The report also
commented on the utility of the
aggregated estimates for risk assessment
purposes (Poole, 1997). This report was
limited to the studies available to Morris
et al. (1992) plus four additional studies
that EPA requested to be included
(Ijsselmuiden et al., 1992; McGeehin et
al., 1993; Vena et al., 1993; and King
and Marrett, 1996). Poole observed that
there was considerable heterogeneity
among the data and that there was
evidence of publication bias within the
body of literature. When there is
significant heterogeneity among studies,
aggregation of the results into a single
summary estimate may not be
appropriate. Publication bias refers to
the situation where the literature search
and inclusion criteria for studies used
for the meta-analysis indicate that the
sample of studies used is not
representative of all the research
(published and unpublished) that has
been done on a topic. In addition, Poole
found that the aggregate estimates
reported by Morris et al. (1992) were
sensitive to small changes in the
analysis (e.g., addition or deletion of a
single study). Based on these
observations, Poole recommended that
the cancer epidemiology data
considered in his evaluation should not
be combined into a single summary
estimate and that the data had limited
utility for risk assessment purposes.
Many of the reasons cited by Poole for
why it was not appropriate to combine
the studies into a single point estimate
of risk were noted in the 1994 proposal
(Farland and Gibb, 1993; Murphy, 1993;
and Craun, 1993).
b. EPA's Evaluation of Poole Report.
EPA reviewed the conclusions from the
Poole report and generally concurred
with Poole's recommendations (USEPA,
1998f). EPA concluded that Poole
presented reasonable and supportable
evidence to suggest that the work of
Morris et al. (1992) should not be used
for risk assessment purposes without
further study and review because of the
sensitivity of the results to analytical
choices and to the addition or deletion
of a single study. EPA agreed that the
studies were highly heterogeneous, thus
undermining the ability to combine the
data into a single summary estimate of
risk.
c. Peer Review of Poole Report and
EPA's Evaluation. The Poole report and
EPA's evaluation were reviewed by five
epidemiologic experts from academia,
government, and industry (EPA, 1998g).
Overall, these reviewers agreed that the
Poole report was of high quality and
that he had used defensible assumptions
and techniques during his analysis.
Most of the reviewers concluded that
the report was correct in its assessment
that these data should not be combined
into a single summary estimate of risk.
2. New Cancer Epidemiology Studies
Several cancer epidemiological
studies examining the association
between exposure to chlorinated surface
water and cancer have been published
subsequent to the 1994 proposed rule
and the Morris etal. (1992) meta-
analysis (McGeehin etal., 1993; Vena et
al. 1993; King and Marrett, 1996; Doyle
etal., 1997; Freedman eta7., 1997;
Cantor et al, 1998; and Hildesheim et
al., 1998). These studies, with the
exception of Freedman etal. (1997),
were described in the "Summaries of
New Health Effects Data" (USEPA,
1997b) that was included in the docket
for the 1997 NOD A.
In general, the new studies cited
above are better designed than the
studies published prior to the 1994
proposal. The newer studies generally
include incidence cases of disease,
interviews with the study subjects and
better exposure assessments. Based on
the entire cancer epidemiology
database, bladder cancer studies provide
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15679
better evidence than other types of
cancer for an association between
exposure to chlorinated surface water
and cancer. EPA believes the association
between exposure to chlorinated surface
water and colon and rectal cancer
cannot be determined at this time
because of the limited data available for
these cancer sites (USEPA, 1998a).
3. Quantitative Risk Estimation for
Cancers From Exposure to Chlorinated
Water
Under Executive Order 12866 (58 FR
51735, October 4, 1993), the EPA must
conduct a regulatory impact analysis
(RIA). In the 1994 proposal, EPA used
the Morris et al. (1992) meta-analysis in
the RIA to provide an upper-bound
estimate of 10,000 possible cancer cases
per year that could be attributed to
exposure to chlorinated water and its
associated byproducts. EPA also
estimated that an upper bound of 1200-
3300 of these cancer cases per year
could be avoided if the requirements for
the Stage 1 DBF rule were implemented
(USEPA, 1994a). EPA acknowledged the
uncertainty In these estimates, but
believed they were the best that could
be developed at the time.
Based on the evaluations cited above,
EPA does not believe it is appropriate to
use the Morris et al. (1992) study as the
basis for estimating the potential cancer
cases that could be attributed to
exposure to DBPs in chlorinated surface
water. Instead, EPA is providing for
comment an analysis based on a more
traditional approach for estimating the
potential cancer risks from exposure to
DBPs in chlorinated surface water that
does not rely on pooling or aggregating
the epidemiologic data into a single
summary estimate. Based on a narrower
set of improved studies, this approach
utilizes the population attributable risk
(PAR) concept and presents a range of
potential risks and not a single point
estimate. As discussed below, there are
a number of uncertainties associated
with the use of this approach for
estimating potential risks. Therefore,
EPA requests comments on both the
PAR methodology as well as on the
assumptions upon which it is based.
Epidemiologists use PAR to quantify
the fraction of the disease burden in a
population (e.g., cancer) that could be
eliminated if the exposure was absent
(e.g., DBPs in chlorinated water)
(Rockhlll, et al.. 1998). PARs provide a
perspective on the potential magnitude
of risk associated with various
exposures. The concept of PAR is
known by many names (e.g. attributable
fraction, attributable proportion,
etiologlc fraction). For this Notice, the
term PAR will be used to avoid
confusion. A range of PARs better
captures the heterogeneity of the risk
estimates than a single point estimate.
In the PAR analysis of the cancer
epidemiology data and the development
of the range of potential cancer cases
attributable to exposure to DBPs in
chlorinated surface water, EPA
recognizes that a causal relationship
between chlorinated surface water and
bladder cancer has not yet been
demonstrated by epidemiology studies.
However, several studies have suggested
a weak association in various
subgroups. EPA presents potential
cancer case estimates as upper bounds
of suggested risk as part of the Agency's
analysis of potential costs and benefits
associated with this rule. EPA focused
its current evaluation on bladder cancer
because the number of quality studies
that are available for other cancer sites
such as colon and rectal cancers are
very limited.
EPA estimated PARs for the best
bladder cancer studies that provided
enough information to calculate a PAR
(USEPA, 1998a). In addition, EPA
selected studies for inclusion in the
quantitative analysis if they met all
three of the following criteria: (1) The
study was a population based case-
control or cohort study conducted to
evaluate the relationship between
exposure to chlorinated drinking water
and incidence cancer cases, based on
personal interviews (no cohort studies
were found that met all 3 criteria); (2)
the study was of high quality and well
designed (e.g., good sample size, high
response rate, and adjusted for
confounding factors); and (3) the study
had adequate exposure assessments
(e.g., residential histories, actual THM
data). Based on the above selection
criteria, five bladder cancer studies were
selected for estimating PARs: Cantor et
al., 1985; McGeehin et al. 1993; King
and Marrett, 1996; Freedman et al.,
1997; and Cantor et a/., 1998. PARs were
derived for two exposure categories:
years of exposure to chlorinated surface
water; and THM levels and years of
chlorinated surface water exposure.
The PARs from the five bladder
cancer studies for the two exposure
categories ranged from 2-17%. The
uncertainties associated with these PAR
estimates are large as expected, due to
the common prevalence of both the
disease (bladder cancer) and exposure
(chlorinated drinking water). Based on
54,500 expected new bladder cancer
cases in the U.S., as projected by NCI
(1998) for 1997, the upper bound
estimate of the number of potential
bladder cancer cases per year
potentially associated with exposure to
DBPs in chlorinated surface water was
estimated to be 1100-9300.
EPA made several important
assumptions when evaluating the
application of the PAR range of
estimated bladder cancer cases from
these studies to the U.S. population.
They include the following: (i) The
study population selected for each of
the cancer epidemiology studies are
reflective of the entire U.S. population
that develops bladder cancer; (ii) the
percentage of bladder cancer cases
exposed to DBPs in the reported studies
are reflective of the bladder cancer cases
exposed to DBPs in the U.S. population;
(Hi) the levels of DBF exposure in the
bladder studies are reflective of the DBF
exposure in the U.S. population; and
(iv) the possible relationship between
exposure to DBPs in chlorinated surface
water and bladder cancer is causal.
EPA believes that these assumptions
would not be appropriate for estimating
the potential upper bound cancer risk
for the U.S. population based on a single
study. However, the Agency believes
that these assumptions are appropriate
given the number of studies used in the
PAR analysis and for gaining a
perspective on the range of possible
upper bound risks that can be used in
establishing a framework for further
cost-benefit analysis. In addition, EPA
believes these assumptions are
appropriate given the SDWA mandate
that "drinking water regulations be
established if the contaminant may have
an adverse effect on the health of
persons" (SDWA—Section 1412(b)(l)A).
Because of this mandate, EPA believes
that when the scientific data indicates
there may be causality, such an
analytical approach is appropriate. EPA
believes the assumption of a potential
causal relationship is supported by the
weight-of-evidence from toxicology and
epidemiology. Toxicology studies have
shown several individual DBPs to be
carcinogenic and mutagenic, while the
epidemiology data have shown weak
associations between several cancer
sites and exposure to chlorinated
surface water.
EPA notes and requests comment on
the following additional issues
associated with basing an estimate of
the potential bladder cancer cases that
can be attributed to DBPs in chlorinated
surface water from the five studies
selected for this analysis. The results
generally showed weak statistical
significance and were not always
consistent among the studies. For
example, some reviewers believe that
two studies showed statistically
significant effects only for male
smokers, while two other studies
showed higher effects for non-smokers.
-------
15680
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
One study showed a significant
association with exposure to chlorinated
surface water but not chlorinated
ground water, while another showed the
opposite result. Furthermore, two
studies which examined the effects of
exposure to higher levels of THMs failed
to find a significant association between
level of THMs and cancer. The Agency
notes that it is not necessary that
statistical significance be shown in
order to conduct a PAR analysis as was
stated by peer-reviewers of this analysis.
4. Peer-Review of Quantitative Risk
Estimates
The quantitative cancer risks
estimated from the five epidemiology
studies derived through the calculation
of individual PARs has undergone
external peer review by three expert
epidemiologists (USEPA, 1998a). Two
peer reviewers concurred with the
decision to derive a PAR range. This
approach was deemed more appropriate
than the selection of a single study or
aggregation of study results. One
reviewer indicated significant
reservations with this approach based
not on the method, but the
inconclusivity of the epidemiology
database and stated that it was
premature to perform a PAR analysis
because it would suggest that the
epidemiological information is more
consistent and complete than it actually
is. To better present the degree of
variability, this reviewer suggested an
alternative approach that involves a
graphical presentation of the individual
odds ratios and their corresponding
confidence intervals. Two reviewers
agreed that there is not enough
information to present an estimate of the
PAR for colon and rectal cancer.
EPA understands the issues raised
regarding the use of PARs and
recognizes there may be controversy on
using this approach with the available
epidemiology data. However, as stated
above, EPA believes the PAR approach
is a useful tool for estimating the
potential upper bound risk for use in
developing the regulatory impact
analysis. EPA agrees with two of the
reviewers that there is not enough
information to present an estimate of
colon and rectal risk at this time using
a PAR approach.
5. Summary of Key Observations
The 1994 proposal included a meta-
analysis of 10 cancer epidemiology
studies that provided an estimate of the
number of bladder and rectal cancer
cases per year that could be attributed
to consumption of chlorinated water
and its associated byproducts (Morris et
al., 1992). Based on the evaluations
previously described, EPA does not
believe it is appropriate to use the
Morris et al. (1992) study as the basis for
estimating the potential cancer cases
that could be attributed to exposure to
DBFs in chlorinated surface water.
Instead, EPA has focused on a smaller
set of higher quality studies and
performed a PAR analysis to estimate
the potential cancer risks from exposure
to DBFs in chlorinated surface water
that does not rely on pooling or
aggregating the data into a single
summary estimate, as was done by
Morris et al. (1992). EPA focused the
current evaluation on bladder cancer
because there are more appropriate
studies of higher quality available upon
which to base this assessment than for
other cancer sites. It was decided to
present the potential number of cancer
cases as a range instead of a single point
estimate because this would better
represent the uncertainties in the risk
estimates. The number of potential
bladder cancer cases per year that could
be associated with exposure to DBFs in
chlorinated surface water is estimated to
be an upper bound range of 1100-9300
per year.
In the PAR analysis of the cancer
epidemiology data and the development
of the range of potential cancer cases
attributable to exposure to DBFs in
chlorinated surface water, EPA presents
the estimates as upper bounds of any
suggested risk. As was debated during
the 1992-1993 M/DBP Regulatory
Negotiation process, EPA believes that
there are insufficient data to
conclusively demonstrate a causal
association between exposure to DBFs
in chlorinated surface water and cancer.
EPA recognizes the uncertainties of
basing quantitative estimates using the
current health database on chlorinated
surface waters and has identified a
number of issues that must be
considered in interpreting the results of
this analysis. Nonetheless, the Agency
believes that the overall weight-of-
evidence from available epidemiologic
and toxicologic studies on DBFs and
chlorinated surface water continues to
support a hazard concern and thus, a
prudent public health protective
approach for regulation.
6. Requests for Comments
EPA is not considering any changes to
the recommended regulatory approach
contained in the 1994 proposal, and
discussed further in the 1997 NODA,
based on the upper bound risk analysis
issues discussed above. Nonetheless,
EPA requests comments on the
conclusions from the Poole report
(Poole, 1997), EPA's assessment of the
Poole report (EPA, 1998f), the peer-
review of the Poole report and EPA's
assessment of the Poole report (EPA,
1998g); and Dr. Morris comments on the
Poole review (Morris, 1997). EPA also
requests comments on its quantitative
analysis (PAR approach) to estimate the
upper bound risks from exposure to
DBFs in chlorinated surface water, the
methodology for estimating the number
of cancer cases per year that could be
attributed to exposure to DBFs in
chlorinated surface water, and any
alternative approaches for estimating
the upper bound estimates of risk. In
particular, EPA requests comment on
the extent to which the approach used
in the PAR analysis addresses the ,
concerns identified by Poole and others
regarding the earlier Morris meta-
analysis. EPA also requests comments
on the peer review of the PAR analysis.
B. Epidemiologic Associations Between
Exposure to DBFs in Chlorinated Water
and Adverse Reproductive and
Developmental Effects
The 1994 proposed rule discussed
several reproductive epidemiology
studies. At the time of the proposal, it
was concluded that there was no
compelling evidence to indicate a
reproductive and developmental hazard
due to exposure to chlorinated water
because the epidemiologic evidence was
inadequate and the lexicological data
were limited. In 1993, an expert panel
of scientists was convened by the
International Life Sciences Institute to
review the available human studies for
developmental and reproductive
outcomes and to provide research
recommendations (USEPA/ILSI, 1993).
The expert panel concluded that the
epidemiologic results should be
considered preliminary given that the
research was at a very early stage
(USEPA/ILSI, 1993; Reif era/., 1996).
The 1997 NODA and the "Summaries of
New Health Effects Data" (USEPA,
1997b) presented several new studies
(Savitz era/., 1995; Kanitz et al. 1996;
and Bove et al., 1996) that had been
published since the 1994 proposed rule
and the 1993ILSI panel review. Based
on the new studies presented in the
1997 NODA, EPA stated that the results
were inconclusive with regard to the
association between exposure to
chlorinated waters and adverse
reproductive and developmental effects
(62 FR 59395)
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15681
1. EPA Panel Report and
Recommendations for Conducting
Epidemiological Research on Possible
Reproductive and Developmental
Effects of Exposure to Disinfected
Drinking Water
EPA convened an expert panel in July
1997 to evaluate epidemiologic studies
of adverse reproductive or
developmental outcomes that may be
associated with the consumption of
disinfected drinking water published
since the 1993ILSI panel review. A
report was prepared entitled "EPA
Panel Report and Recommendations for
Conducting Epidemiological Research
on Possible Reproductive and
Developmental Effects of Exposure to
Disinfected Drinking Water" (USEPA,
1998h). The 1997 expert panel was also
charged to develop an agenda for further
epidemlological research. The 1997
panel concluded that the results of
several studies suggest that an increased
relative risk of certain adverse outcomes
may be associated with the type of water
source, disinfection practice, or THM
levels. The panel emphasized, however,
that most relative risks are moderate or
small and were found in studies with
limitations of their design or conduct.
The small magnitude of the relative risk
found may be due to one or more
sources of bias, as well as to residual
confounding (factors not identified and
controlled). Additional research is
needed to assess whether the observed
associations can be confirmed. The
panel considers a recent study by Waller
et al. (1998), discussed below, to
provide a strong basis for further
research. This study was funded in part
by EPA as an element of the research
program agreed to as part of the 1992/
1993 negotiated M/DBP rulemaking.
2. New Reproductive Epidemiology
Studies
Three new reproductive epidemiology
studies have been published since the
1997 NODA. The first study (Klotz and
Pyrch, 1998) examined the potential
association between neural tube defects
and certain drinking water
contaminants, including some DBPs. In
this case-control study, births with
neural tube defects reported to New
Jersey's Birth Defects Registry in 1993
and 1994 were matched against control
births chosen randomly from across the
State. Birth certificates were examined
for all subjects, as was drinking water
data corresponding to the mother's
residence in early pregnancy. The
authors reported elevated odds ratios
(ORs), generally between 1.5 and 2.1, for
the association of neural tube defects
with TTHMs. However, the only
statistically significant results were seen
when the analysis was isolated to those
subjects with the highest THM
exposures (greater than 40 ppb) and
limited to those subjects with neural
tube defects in which there were no
other malformations (odds ratio 2.1;
95% confidence interval 1.1-4.0).
Neither HAAs or haloacetonitriles
(HANs) showed a clear relationship to
neural tube defects but monitoring data
on these DBPs were more limited than
for THMs. Nitrates were not observed to
be associated with neural tube defects.
Certain chlorinated solvent
contaminants were also studied but
occurrence levels were too low to assess
any relationship to neural tube defects.
This study is available in the docket for
this NODA. Although EPA has not
completed its review of the study, the
Agency is proceeding on the premise
that this study will add to the weight-
of-evidence concerning the potential
adverse reproductive health effects from
DBPs, but will not by itself provide
sufficient evidence for further regulatory
actions.
Two studies looked at early term
miscarriage risk factors. The first of
these studies (Waller et al., 1998)
examined the potential association
between early term miscarriage and
exposure to THMs. The second study
(Swan etal., 1998) examined the
potential association between early term
miscarriage and tap water consumption.
Both studies used the same group of
pregnant women (5,144) living in three
areas of California. They were recruited
from the Santa Clara area, the Fontana
area in southern California, or the
Walnut Creek area. The women were all
members of the Kaiser Permanente
Medical Care Program and were offered
a chance to participate in the study
when they called to arrange their first
prenatal visit. In the Waller et al. (1998)
study, additional water quality
information from the women's drinking
water utilities were obtained so that
THM levels could be determine. The
Swan et al. (1998) study provided no
quantitative measurements of THMs (or
DBPs), and thus, provided no additional
information on the risk from
chlorination byproducts. Because of
this, only the Waller et al. (1998) study
is summarized below.
In the Waller et al. (1998) study,
utilities that served the women in this
study were identified. Utilities'
provided THM measurements taken
during the time period participants were
pregnant. The TTHM level in a
participant's home tap water was
estimated by averaging water
distribution system TTHM
measurements taken during a
participants first three months of
pregnancy. This "first trimester TTHM
level" was combined with self reported
tap water consumption to create a
TTHM exposure level. Exposure levels
of the individual THMs (e.g.,
chloroform, bromoform, etc.) were
estimated in the same manner. Actual
THM levels in the home tap water were
not measured.
Women with high TTHM exposure in
home tap water (drinking five or more
glasses per day of cold home tap water
containing at least 75 ug per liter of
TTHM) had an early term miscarriage
rate of 15.7%, compared with a rate of
9.5% among women with low TTHM
exposure (drinking less than 5 glasses
per day of cold home tap water or
drinking any amount of tap water
containing less than 75 ug per liter of
TTHM). An adjusted odds ratio for early
term miscarriage of 1.8 (95% confidence
interval 1.1-3.0) was determined.
When the four individual
trihalomethanes were studied, only high
bromodichloromethane (BDCM)
exposure, defined as drinking five or
more glasses per day of cold home tap
water containing 5:18 ug/L
bromodichloromethane, was associated
with early term miscarriage. An
adjusted odds ratio for early term
miscarriage of 3.0 (95% confidence
interval 1.4-6.6) was determined.
3. Summary of Key Observations
The Waller et al. (1998) study reports
that consumption of tapwater
containing high concentrations of
THMs, particularly BDCM, is associated
with an increased risk of early term
miscarriage. EPA believes that while
this study does not prove that exposure
to THMs causes early term miscarriages,
it does provide important new
information that needs to be pursued
and that the study adds to the weight-
of-evidence which suggests that
exposure to DBPs may have an adverse
effect on humans.
EPA has an epidemiology and
toxicology research program that is
examining the relationship between
DBPs and adverse reproductive and
developmental effects. In addition to
conducting scientifically appropriate
follow-up studies to see if the observed
association in the Waller etal. (1998)
study can be replicated elsewhere, EPA
will be working with the California
Department of Health Services to
improve estimates of exposure to DBPs
in the existing study population. A more
complete DBF exposure data base is
being developed by asking water
utilities in the study area to provide
additional information, including levels
of other types of DBPs (e.g., haloacetic
-------
15682
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
acids). These efforts will help further
assess the significance of the Waller et
al. (1998) study, associated concerns,
and how further follow-up work can
best be implemented. EPA will
collaborate with the Centers for Disease
Control and Prevention (CDC) in a series
of studies to evaluate if there is an
association between exposure to DBFs
in drinking water and birth defects. The
Agency is also involved in a
collaborative testing program with the
National Toxicology Program (NTP)
under which several individual DBFs
have been selected for reproductive and
developmental screening tests. Finally,
EPA is conducting several toxicology
studies on DBFs other endpoints of
concern including examining the
potential effects of BDCM on male
reproductive endpoints. This
information will be used in developing
the Stage 2 DBF rule. In the meantime,
the Agency plans to proceed with the
1994 D/DBP proposal for tightening the
control for DBFs.
4. Requests for Comments
EPA is not considering any changes to
the recommended regulatory approach
contained in the 1994 proposal, and
discussed further in the 1997 NODA,
based on the new reproductive
epidemiology studies discussed above.
Nonetheless, EPA requests comments on
the findings from the Klotz, et al. (1998)
and Waller et ail. (1998) study and EPA's
conclusions regarding the studies.
III. Significant New lexicological
Information for the Stage 1
Disinfectants and Disinfection
Byproducts
The 1997 NODA reviewed new
lexicological information that became
available for several of the DBFs after
the 1994 proposal (USEPA, 1997a and
b). In that Notice, it was pointed out that
several forthcoming reports were not
available in time for consideration
during the 1997 FACA process. Reports
now available include a two-generation
reproductive rat study of sodium
chlorite sponsored by the Chemical
Manufacturer Association (CMA, 1996);
an EPA two-year cancer rodent study of
bromate (DeAngelo et al., 1998); and the
International Life Sciences Institute
(ILSI) expert panel report of chloroform
and dichloroacetic acid ([LSI, 1997).
These reports are discussed below, as
well as EPA's analyses and conclusions
based on this new information.
A. Chlorite and Chlorine Dioxide
The 1994 proposal included an MCLG
of 0.08 mg/L and an MCL of 1.0 mg/L
for chlorite. The proposed MCLG was
based on an RfD of 3 mg/kg/d estimated
from a lowest-obseryed-adverse-effect-
level (LOAEL) for neurodevelopmental
effects identified in a rat study by
Mobley et al. (1990). This determination
was based on a weight of evidence
evaluation of all the available data at
that time (USEPA, 1994a). An
uncertainty factor of 1000 was used to
account for inter- and intra-species
differences in response to toxicity (a
factor of 100) and a factor of 10 for use
of a LOAEL. The EPA proposed rule
also included an MRDLG of 0.3 mg/L
and an MRDL of 0.8 mg/L for chlorine
dioxide. The proposed MRDLG was
based on a RfD of 3 mg/kg/d estimated
from a no-observed-adverse-effect-level
(NOAEL) for developmental
neurotoxicity identified from a rat study
(Orme etal., 1985; see USEPA, 1994a).
This determination was based on a
weight of evidence evaluation of all the
available data at that time (USEPA,
1994a). An uncertainty factor of 300 was
applied that was composed of a factor
of 100 to account for inter- and intra-
species differences in response to
toxicity and a factor of 3 for lack of a
two-generation reproductive study
necessary to evaluate potential toxicity
associated with lifetime exposure. To
fill this important data gap, the
Chemical Manufacturers Associations
(CMA) agreed to conduct a two-
generation reproductive study in rats.
Sodium chlorite was used as the test
compound. It should be noted that data
on chlorite are relevant to assessing the
risks of chlorine dioxide because
chlorine dioxide rapidly disassociates to
chlorite (and chloride) (USEPA, 1998b).
Therefore, the new CMA two-generation
reproductive chlorite study will be
considered in assessing the risks for
both chlorite and chlorine dioxide.
Since the 1994 proposal, CMA has
completed the two-generation
reproductive rat study (CMA, 1996).
EPA has reviewed the CMA study and
has completed an external peer review
of the study (EPA, 1997c). In addition,
EPA has reassessed the noncancer
health risk for chlorite and chlorine
dioxide considering the new CMA study
(USEPA, 1998b). This reassessment has
been peer reviewed (USEPA, 1998b).
Based on this reassessment, EPA is
considering changing the proposed
MCLG for chlorite from 0.08 mg/L to 0.8
mg/L based on the NOAEL identified
from the new CMA study. Since data on
chlorite are considered relevant to
chlorine dioxide risks and the two
generation reproduction data gap has
been filled, EPA is also considering
changing the proposed MRDLG for
chlorine dioxide from 0.3 mg/L to 0.8
mg/L. The basis for these changes are,
discussed below.
1. 1997 CM A Two-Generation
Reproduction Rat Study
The CMA two-generation
reproductive rat study was designed to
evaluate the effects of chlorite (sodium
salt) on reproduction and pre- and post-
natal development when administered
orally via drinking water for two
successive generations (CMA, 1996).
Developmental neurotoxicity,
hematological, and clinical effects were
also evaluated in this study.
Sodium chlorite was administered at
0, 35, 70, and 300 ppm in drinking
water to male and female Sprague
Dawley rats (F0 generation) for ten
weeks prior to mating. Dosing continued
during the mating period, pregnancy
and lactation. Reproduction, fertility,
clinical signs, and histopathology were
evaluated in F0 and FI (offspring from
the first generation of mating) males and
females. FI and F2 (offspring from the
second generation of mating) pups were
evaluated for growth and development,
clinical signs, and histopathology. In
addition, F] animals from each dose
group were assessed for neurotoxicity
(e.g., neurohistopathology, motor
activity, learning ability and memory
retention, functional observations,
auditory startle response). Limited
neurotoxicological evaluations were
conducted on F2 pups.
The CMA report concluded that there
were no treatment related effects at any
dose level for systemic, reproductive/
developmental, and developmental
neurological end points. The report
indicates that there were small
statistically significant decreases in the
maximum response to auditory startle
response in the FI animals at the mid
and high dose (70 and 300 ppm); this
neurological effect was not considered
to be toxicologically significant. A
reduction in pup weight and decreased
body weight gain through lactation in
the FI and F2 animals and a decrease in
body weight gain in the F2 males at 300
ppm were noted. Decreases in liver
weight in F0 and FI animals, as well as
reductions in red blood cell indices in
FI animate at 300 ppm and 70 ppm were
noted. Minor hematological effects were
found in FI females at 35 ppm. CMA
concluded that the effects noted above
were not clinically or toxicologically
significant. A NOAEL of 300 ppm was
identified in the CMA report for
reproductive toxicity and for
developmental neurotoxic effects, and a
NOAEL of 70 ppm for hematological
effects. EPA disagrees with the CMA
conclusions regarding the NOAEL of
300 ppm for the reproductive and
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15683
developmental neurological effects for
this study as discussed below.
2. External Peer Review of the CMA
Study
EPA has evaluated the CMA 2-
generatlon reproductive study and
concluded that the study design was
consistent with EPA testing guidelines
(USEPA. 1992). Additionally, an expert
peer review of the CMA study was
conducted and Indicated that the study
design and analyses were adequate
(USEPA, 1997c). Although the study
design was considered adequate and
consistent with EPA guidelines, the peer
review pointed out some limitations in
the study (USEPA, 1997c). For example,
developmental neurotoxicity
evaluations were conducted after
exposure ended at weaning. This is
consistent with EPA testing guidelines
and should potentially detect effects on
the developing central nervous system.
Nevertheless, the opportunity to detect
neurological effects due to continuous
or lifetime exposure may be reduced.
The peer review generally questioned
the CMA conclusions regarding the
NOAELs for this study and indicated
that the NOAEL should be lower than
300 ppm. The majority of peer reviews
recommended that the NOAEL for
reproductive/developmental toxicity be
reduced to 70 ppm given the treatment
related effects found at 300 ppm, and
that the NOAEL for neurotoxicity be
reduced to 35 ppm based on significant
changes in the maximum responses in
startle amplitude and absolute brain
weight at 70 and 300 ppm. The
reviewers indicated that a NOAEL was
not observed for hematological effects
and noted that the CMA conclusion for
selecting the 70 ppm NOAEL for the
hematology variables needs to be
explained further.
3. MCLG for Chlorite: EPA's
Reassessment of the Noncancer Risk
EPA has determined that the NOAEL
for chlorite should be 35 ppm (3 mg/kg/
d chlorite ion, rounded) based on a
weight of evidence approach. The data
considered to support this NOAEL are
summarized in USEPA (1998b) and
included the CMA study as well as
previous reports on developmental
neurotoxicity (USEPA, 1998b). The
NOAEL of 35 ppm (3 mg/kg/d chlorite
ion) is based on the following effects
observed in the CMA study at 70 and
300 ppm chlorite: Decreases in absolute
brain and liver weight, and lowered
auditory startle amplitude. Decreases in
pup weight were found at the 300 ppm
and thus a NOAEL of 70 ppm for
reproductive effects is considered
appropriate (USEPA, 1998b). Although
70 ppm appears to be the NOAEL for
hemolytic effects, the NOAEL and
LOAEL are difficult to discern for this
endpoint given that minor changes were
reported at 70 and 35 ppm. EPA
considers the basis of the NOAELs to be
consistent with EPA risk assessment
guidelines (USEPA, 1991, 1998i, 1996a).
Furthermore, a NOAEL of 35 ppm is
supported by effects (particularly
neurodevelopmental effects) found in
previously conducted studies of chlorite
and chlorine dioxide (USEPA, 1998b).
An RfD of 0.03 mg/kg/d is estimated
using a NOAEL of 3 mg/kg/d and an
uncertainty factor of 100 to account for
inter- and intra-species differences. The
revised MCLG for chlorite is calculated
to be 0.8 mg/L by assuming an adult tap
water consumption of 2 L per day for a
70 kg adult and using a relative source
contribution of 80% (because most
exposure to chlorite is likely to come
from drinking water):
MCLG for chlorite =
0.03 mg/kg/d X 70 kg X 0.8
2L/day
MCLG for chlorite = 0.8 mg/L (Rounded)
= 0.84 mg/L
Therefore, EPA is considering an
increase in the proposed MCLG for
chlorite from 0.08 mg/L to 0.8 mg/L. A
more detailed discussion of this
assessment is included in the docket for
this Notice (USEPA, 1998b).
4. MRDLG for Chlorine Dioxide: EPA's
Reassessment of the Noncancer Risk
EPA believes that data on chlorite are
relevant to assessing the risk of chlorine
dioxide because chlorine dioxide
rapidly disassociates to chlorite (and
chloride) (USEPA, 1998b). Therefore,
the findings from the 1997 CMA two-
generation reproductive study on
sodium chlorite should be considered in
a weight of evidence approach for
establishing the MRDLG for chlorine
dioxide. Based on all the available data,
including the CMA study, a dose of 3
mg/kg/d remains as the NOAEL for
chlorine dioxide (USEPA, 1998b). The
MRDLG for chlorine dioxide is
increased 3 fold from the 1994 proposal
since the CMA 1997 study was a two-
generation reproduction study. Using a
NOAEL of 3 mg/kg/d and applying an
uncertainty factor of 100 to account for
inter- and intra-species differences in
response to toxicity, the revised MRDLG
for chlorine dioxide is calculated to be
0.8 mg/L. This MRDLG takes into
account an adult tap water consumption
of 2 L per day for a 70 kg adult and
applies a relative source contribution of
80% (because most exposure to chlorine
dioxide is likely to come from drinking
water):
MRDLG for Chlorine dioxide =
0.03 mg/kg/d X 70 kg X 0.8
2L/day
MRDLG for Chlorine dioxide = 0.8 mg/L (Rounded)
= 0.84 mg/L
EPA is considering revising the MRDLG
for chlorine dioxide from 0.3 mg/L to
0.8 mg/L. A more detailed discussion of
this assessment can be found in the
docket for this Notice (USEPA, 1998b).
5. External Peer Review of EPA's
Reassessment
Three external experts have reviewed
the EPA reassessment for chlorite and
chlorine dioxide (see USEPA, 1998b).
Two of the three reviewers generally
agreed with EPA conclusions regarding
the identified NOAEL of 35 ppm for
neurodevelopmental toxicity. The other
reviewer indicated that the
developmental neurological results from
the CMA study were transient, too
inconsistent, and equivocal to identify a
NOAEL. EPA believes that although
different responses were found for
startle response (as indicated by
measures of amplitude, latency, and
habituation), this is not unexpected
given that these measures examine
different aspects of the nervous system,
and thus can be differentially affected.
Although no neuropathology was
observed in the CMA study,
neurofunctional (or neurochemical)
-------
15684
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
changes such as startle responses can
indicate potential neurotoxicity without
neuropathological effects. Furthermore,
transient effects are considered an
important indicator of neurotoxicity as
indicated in EPA guidelines (USEPA,
19981). EPA maintains that the NOAEL
is 35 ppm (3 mg/kg/d) from the CM A
chlorite study based on
neurodevelopmental effects as well as
changes in brain and liver weight. This
conclusion is supported by previous
studies on chlorite and chlorine dioxide
(USEPA, 1998b). Other comments raised
by the peer reviewers concerning
improved clarity and completeness of
the assessment were considered by EPA
in revising the assessment document on
chlorite and chlorine dioxide.
6. Summary of Key Observations
EPA continues to believe that chlorite
and chlorine dioxide may have an
adverse effect on the public health. EPA
identified a NOAEL of 35 ppm for
chlorite based on neurodevelopmental
effects from the 1997 CMA two-
generation reproductive study, which is
supported by previous studies on
chlorite and chlorine dioxide. In
addition, EPA identified a NOAEL of 70
ppm for reproductive/developmental
effects and hemolytic effects. EPA
considers this study relevant to
assessing the risk to chlorine dioxide.
Based on the EPA reassessment, EPA is
considering adjusting the MCLG for
chlorite from 0.08 mg/L to 0.8 mg/L.
Because data on chlorite are considered
relevant to chlorine dioxide risks, EPA
is considering adjusting the MRDLG for
chlorine dioxide from 0.3 mg/L to 0.8
mg/L. The MRDL for chlorine dioxide
would remain at 0.8 mg/L. The MCL for
chlorite would remain at 1.0 mg/L
because as noted in the 1994 proposal,
1.0 mg/L for chlorite is the lowest level
achievable by typical systems using
chlorine dioxide and taking into
consideration the monitoring
requirements to determine compliance.
In addition, given the margin of safety
that is factored into the estimation of the
MCLG, EPA believes that 1.0 mg/L will
be protective of public health. It should
be noted that the MCLG and MRDLG
presented for chlorite and chlorine
dioxide are considered to be protective
of susceptible groups, including
children given that the RfD is based on
a NOAEL derived from developmental
testing, which includes a two-generation
reproductive study. A two-generation
reproductive study evaluates the effects
of chemicals on the entire
developmental and reproductive life of
the organism. Additionally, current
methods for developing RfDs are
designed to be protective for sensitive
populations. In the case of chlorite and
chlorine dioxide a factor of 10 was used
to account for variability between the
average human response and the
response of more sensitive individuals.
7. Requests for Comments
Based on the recent two-generation
reproductive rat study for chlorite
(CMA, 1996), EPA is considering
revising the MCLG for chlorite from 0.08
mg/L to 0.8 mg/L and the MRDLG for
chlorine dioxide from 0.3 mg/L to 0.8
mg/L. EPA requests comments on these
possible changes in the MCLGs and on
EPA's assessment of the CMA report.
B. Trihalomethanes
The 1994 proposed rule included an
MCL for TTHM of 0.08 mg/L. MCLGs of
zero for chloroform, BDCM and
bromoform were based on sufficient
evidence of carcinogenicity in animals.
The MCLG of 0.06 mg/L for
dibromochloromethane (DBCM) was
based on observed liver toxicity from a
subchronic study and limited animal
evidence for carcinogenicity. As stated
in the 1997 NOD A, several new studies
have been published on bromoform,
BDCM, and chloroform since the 1994
proposal. The 1997 NODA concluded
that the new studies on THMs
contribute to the weight-of-evidence
conclusions reached in the 1994
proposed rule, and that the new studies
are not anticipated to change the
proposed MCLGs for BDCM, DBCM, and
bromoform. Since the 1997 NODA, the
EPA has evaluated the significance of an
ILSI panel report on the cancer risk
assessment for chloroform. EPA has
conducted a reassessment of chloroform
(USEPA, 1998c), considering the ILSI
report. The EPA reassessment of
chloroform has been peer reviewed
(USEPA, 1998c). Based on EPA's
reassessment, the Agency is considering
changing the proposed MCLG for
chloroform from zero to 0.3 mg/L.
1. 1997 International Life Sciences
Institute Expert Panel Conclusions for
Chloroform
In 1996, EPA co-sponsored an ILSI
project in which an expert panel was
convened and charged with the
following objectives: (i) Review the
available database relevant to the
carcinogenicity of chloroform and DC A,
excluding exposure and epidemiology
data; (ii) consider how end points
related to the mode of carcinogenic
action can be applied in the hazard and
dose-response assessment; (iii) use
guidance provided by the 1996 EPA
Proposed Guidelines for Carcinogen
Assessment to develop
recommendations for appropriate
approaches for risk assessment; and (iv)
provide a critique of the risk assessment
process and comment on issues
encountered in applying the proposed
EPA Guidelines (ILSI, 1997). The panel
was made up of 10 expert scientists
from acadefnia, industry, government,
and the private sector. It should be
emphasized that the ILSI report does not
represent a risk assessment, per se, for
chloroform (orDCA) but, rather,
provides recommendations oh how to
proceed with a risk assessment for these
two chemicals.
To facilitate an understanding of the
ILSI panel recommendations for the
dose-response characterization of
chloroform, the EPA 1996 Proposed
Guidelines for Carcinogen Risk
Assessment must be briefly described.
For a more detail discussion of these
guidelines, refer to USEPA (1996b).
The EPA 1996 Proposed Guidelines
for Carcinogen Risk Assessment
describes a two-step process to
quantifying cancer risk (USEPA, 1996b).
The first step involves modeling
response data in the empirical range of
observation to derive a point of
departure. The second step is to
extrapolate from this point of departure
to lower levels that are within the range
of human exposure. A standard point of
departure was proposed as the lower
95% confidence limit on a dose
associated with 10% extra risk (LEDJO).
Based on comments from the public and
the EPA's Science Advisory Board, the
central or maximum likelihood estimate
(i.e., EDio) is also being considered as a
point of departure. Once the point of
departure is identified, a straight-line
extrapolation to the origin (i.e., zero
dose, zero extra risk) is conducted as the
linear default approach. The linear
default approach would be considered
for chemicals in which the mode of
carcinogenic action understanding is
consistent with low dose linearity or as
a science policy choice for those
chemicals for which the mode of action
is not understood.
The EPA 1996 Proposed Guidelines
for Carcinogen Risk Assessment are
different from the 1986 guidelines
approach that applied the linearized
multi-stage model (LMS) to extrapolate
low dose risk. The LMS approach under
the 1986 guidelines was the only default
for low dose extrapolation. Under the
1996 proposed guidelines both linear
and nonlinear default approaches are
available. The nonlinear approach
applies a margin of exposure (MoE)
analysis rather than estimating the
probability of effects at low doses. In
order to use the nonlinear default, the
agent's mode of action in causing
tumors must be reasonably understood.
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15685
The MoE analysis is used to compare
the point of departure with the human
exposure levels of interest (i.e., MoE =
point of departure divided by the
environmental exposure of interest).
The key objective of the MoE analysis is
to describe for the risk manager how
rapidly responses may decline with
dose. A shallow slope suggests less risk
reduction at decreasing exposure than
does a steep one. Information on factors
such as the nature of response being
used for the point of departure (i.e.,
tumor data or a more sensitive precursor
response) and biopersistence of the
agent are important considerations in
the MoE analysis. A numerical default
factor of no less than 10-fold each may
be used to account for human variability
and for Interspecles differences in
sensitivity when humans may be more
sensitive than animals.
The ILSI expert panel considered a
wide range of information on
chloroform including rodent tumor data,
metabolism/toxicokinetic information,
cytotoxlclty, genotoxicity, and cell
proliferation data. Based on its analysis
of the data, the panel concluded that the
weight of evidence for the mode of
action understanding indicated that
chloroform was not acting through a
direct DNA reactive mechanism. The
evidence suggested, instead, that
exposure to chloroform resulted in
recurrent or sustained toxicity as a
consequence of oxidative generation of
highly tissue reactive and toxic
metabolites (i.e., phosgene and
hydrochloric acid (HC1)), which in turn
would lead to regenerative cell
proliferation. Oxidative metabolism was
considered by the panel to be the
predominant pathway of metabolism for
chloroform. This mode of action was
considered to be the key influence of
chloroform on the carcinogenic process.
The ILSI report noted that the weight-
of-evldence for the mode of action was
stronger for the mouse kidney and liver
responses and more limited, but still
supportive, for the rat kidney tumor
responses.
The panel viewed chloroform as a
likely carcinogen to humans above a
certain dose range, but considered it
unlikely to be carcinogenic below a
certain dose range. The panel indicated
that "This mechanism is expected to
involve a dose-response relationship
which Is nonlinear and probably
exhibits an exposure threshold." The
panel, therefore, recommended the
nonlinear default or margin of exposure
approach as the appropriate one for
quantifying the cancer risk associated
with exposure to chloroform.
2. MCLG for Chloroform: EPA's
Reassessment of the Cancer Risk
In the 1994 proposed rule, EPA
classified chloroform under the 1986
EPA Guidelines for Carcinogen Risk
Assessment as a Group B2, probable
human carcinogen. This classification
was based on sufficient evidence of
carcinogenicity in animals. Kidney
tumor data in male Osborne-Mendel rats
reported by Jorgenson et al. (1985) was
used to estimate the carcinogenic risk.
An MCLG of zero was proposed.
Because the mode of carcinogenic action
was not understood at that time, EPA
used the linearized multistage model
and derived an upper bound
carcinogenicity potency factor for
chloroform of 6 x 10-3 mg/kg/d. The
lifetime cancer risk levels of 10-«, I0~s,
and lO"4 were determined to be
associated with concentrations of
chloroform in drinking water of 6, 60,
and 600 ug/L.
Since the 1994 rule, several new
studies have provided insight into the
mode of carcinogenic action for
chloroform. EPA has reassessed the
cancer risk associated with chloroform
exposure (USEPA, 1998c) by
considering the new information, as
well as the 1997 ILSI panel report. This
reassessment used the principles of the
1996 EPA Proposed Guidelines for
Carcinogen Risk Assessment (USEPA,
1996b), which are considered
scientifically consistent with the
Agency's 1986 guidelines (USEPA,
1986). Based on the current evidence for
the mode of action by which chloroform
may cause tumorgenesis, EPA has
concluded that a nonlinear approach is
more appropriate for extrapolating low
dose cancer risk rather than the low
dose linear approach used in the 1994
proposed rule. Because tissue toxicity is
key to chloroform's mode of action, EPA
has also considered noncancer toxicities
in determining the basis for the MCLG.
After evaluating both cancer risk and
noncancer toxicities as the basis for the
MCLG, EPA concluded that the RfD for
hepatoxicity should be used.
Hepatotoxicity, thus, serves as the basis
for the MCLG given that this is the
primary effect of chloroform and the
more sensitive endpoint. Therefore, EPA
is considering changing the proposed
MCLG for chloroform from zero to 0.3
mg/L based on the RfD for hepatoxicity.
The basis for these conclusions are
discussed below.
a. Weight of the Evidence and
Understanding of Mode of Carcinogenic
Action. EPA has fully considered the
1997 ILSI report and the new science
that has emerged on chloroform since
the 1994 proposed rule. Based on this
new information, EPA considers
chloroform to be a likely human
carcinogen by all routes of exposure
(USEPA, 1998c). Chloroform's
carcinogenic potential is indicated by
animal tumor evidence (liver tumors in
mice and renal tumors in both mice and
rats) from inhalation and oral exposures,
as well as metabolism, toxicity,
mutagenicity and cellular proliferation
data that contribute to an understanding
of mode of carcinogenic action.
Although the precise mechanism of
chloroform carcinogenicity is not
established, EPA agrees with the ILSI
panel that a DNA reactive mutagenic
mechanism is not likely to be the
predominant influence of chloroform on
the carcinogenic process. EPA believes
that there is a reasonable scientific basis
to support a mode of carcinogenic
action involving cytotoxicity produced
by the oxidative generation of highly
reactive metabolites, phosgene and HC1,
followed by regenerative cell
proliferation as the predominant
influence of chloroform on the
carcinogenic process (USEPA, 1998c).
EPA, therefore, agrees with the ILSI
report that the chloroform dose-
response should be considered
nonlinear.
A recent article by Melnick et al.
(1998) was published after the 1997ISLI
panel report and concludes that
cytotoxicity and regenerative
hyperplasia alone are not sufficient to
explain the liver carcinogenesis in
female B6C3F1 mice exposed to
trihalomethanes, including chloroform.
Although this article raises some
interesting issues, EPA views the results
for chloroform supportive of the role
that toxicity and compensatory
proliferation may play in chloroform
carcinogenicity because statistically
significant increases (p<0.05) in
hepatoxicity and cell proliferation are
found for chloroform in this study.
b. Dose-Response Assessment. EPA
has used several different approaches
for estimating the MCLG for chloroform:
the LEDio for tumor response; the EDio
for tumor response; and the RfD for
hepatoxicity. Each of these approaches
are described below. EPA believes the
RfD based on hepatotoxicity serves as
the most appropriate basis for the MCLG
for the reasons discussed below.
EPA has presented the linear and
nonlinear default approaches to
estimating the cancer risk associated
with drinking water exposure to
chloroform (USEPA, 1998c). EPA
considered the linear default approach
because of remaining uncertainties
associated with the understanding of
chloroform's mode of carcinogenic
action: for example, lack of data on
-------
15686
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
cytotoxicity and cell proliferation
responses in Osborne-Mendel rats, lack
of mutagenicity data on chloroform
metabolites, and the lack of comparative
metabolic data between humans and
rodents. Although these data
deficiencies raise some uncertainty
about how chloroform may influence
tumor development at low doses, EPA
views the linear dose-response
extrapolation approach as overly
conservative in estimating low-dose
risk.
EPA concludes that the nonlinear
default or margin of exposure approach
is the preferred approach to quantifying
the cancer risk associated with
chloroform exposure because the
evidence is stronger for a nonlinear
mode of carcinogenic action. The tumor
kidney response data in Osborne-
Mendel rats fromJorgenson eta/. (1985)
are used as the basis for the point of
departure (i.e., LED10 and EDi0) because
a relevant route of human exposure (i.e.,
drinking water) and multiple doses of
chloroform (i.e., 5 doses including zero)
were used in this study (USEPA, 1998c).
The animal data were adjusted to
equivalent human doses using .body
weight raised to the % power as the
interspecies scaling factor, as proposed
in the 1996 EPA Proposed Guidelines
for Carcinogen Risk Assessment. The
EDio and LEDio were estimated to be 37
and 23 mg/kg/d, respectively.
As part of the margin of exposure
analysis, a 100 fold factor was applied
to account for the variability and
uncertainty associated with intra- and
interspecies differences in the absence
of data specific to chloroform. An
additional factor of 10 was applied to
account for the remaining uncertainties
associated with the mode of
carcinogenic action understanding and
the nature of the tumor dose response
relationship being relatively shallow.
EPA believes 1000 fold represents an
adequate margin of exposure that
addresses inter- and intra-species
differences and uncertainties in the
database. Other factors considered in
determining the adequacy of the margin
of exposure include the size of the
human population exposed, duration
and magnitude of human exposure, and
persistence in the environment. Taking
these factors into consideration, a MoE
of 1000 is still regarded as adequate.
Although a large population is
chronically exposed to chlorinated
drinking water, chloroform is not
biopersistent and humans are exposed
to relatively low levels of chloroform in
the drinking water (generally under 100
Hg/L), which are below exposures
needed to induce a cytotoxic response.
Furthermore, EPA believes that a MoE
of 1000 is protective of susceptible
groups, including children. The mode of
action understanding for chloroform's
cytotoxic and carcinogenic effects
involves a generalized mechanism of
toxicity that is seen consistently across
different species. Furthermore, the
activity of the enzyme (i.e., CYP2E1)
involved in generating metabolites key
to chloroform's mode of action is not
greater in children than in adults, and
probably less (USEPA, 1998c).
Therefore, the ED10 of 37 mg/kg-d and
the LED™ of 23 mg/kg-d is divided by
a MoE of 1000 giving dose estimates of
0.037 and 0.023 mg/kg/d for
carcinogenicity, respectively. These
estimates would translate into MCLGs of
1.0 mg/L and 0.6 mg/L, respectively.
The underlying basis for chloroform's
carcinogenic effects involve oxidative
generation of reactive and toxic
metabolites (phosgene and HC1) and
thus are related to its noncancer
toxicities (e.g., liver or kidney
toxicities). It is important, therefore, to
consider noncancer outcomes in the risk
assessment (USEPA, 1998c). The
electrophilic metabolite phosgene
would react with macromoleeules such
as phosphbtidyl inositols or tyrosine
kinases which in turn could potentially
lead to interference with signal
transduction pathways (i.e., chemical
messages controlling cell division),
thus, leading to carcinogenesis.
Likewise, it is also plausible that
phosgene reacts with cellular
phospholipids, peptides, and proteins
resulting in generalized tissue injury.
Glutathipne, free cysteine, histidine,
methionine, and tyrosine are all
potential reactants for electrophilic
agents. Hepatoxicity is the primary
effect observed following chloroform
exposure, and among tissues studied for
chloroform-oxidative metabolism, the
liver was found to be the most active
(ILSI, 1997). In the 1994 proposed rule,
data from a chronic oral study in dogs
(Heywood et al., 1979) were used to
derive the RfD of 0.01 mg/kg/d (USEPA,
1994a). This RfD is based on a LOAEL
for hepatotoxicity and application of an
uncertainly factor of 1000 (100 was used
to account for inter-and intra-species
differences and a factor of 10 for use of
a LOAEL). The MCLG is calculated to be
0.3 mg/L by assuming an adult tap water
consumption of 2 L of tap water per day
for a 70 kg adult, arid by applying a
relative source contribution of 80%
(EPA assumes most exposure is likely to
come from drinking water):
MCLG for Chloroform Based on RfD for Hepatoxicity = °-01 m8/kS^X 70kg x 0.8 = Q ^ (rounded)
2L/day
Therefore, 0.3 mg/L based on
hepatoxicity in dogs (USEPA, 1994a) is
being considered as the MCLG because
liver toxicity is a more sensitive effect
of chloroform than the induction of
tumors. Even if low dose linearity is
assumed, as it was in the 1994 proposed
rule, a MCLG of 0.3 mg/L would be
equivalent to a 5 x 10 -« cancer risk
level. EPA concludes that an MCLG
based on protection against liver
toxicity should be protective against
carcinogenicity given that the putative
mode of action understanding for
chloroform involves cytotoxicity as a
key event preceding tumor
development. Therefore, the
recommended MCLG for chloroform is
0.3 mg/L. The assessment that forms the
basis for this conclusion can be found
in the docket for this Notice (USEPA,
1998c).
3. External Peer Review of EPA's
Reassessment
Three external experts reviewed the
EPA reassessment of chloroform
(USEPA, 1998c). The peer review
generally indicated that the nonlinear
approach used for estimating the
carcinogenic risk associated with
exposure to chloroform was reasonable
and appropriate and that the role of a
direct DNA reactive mechanism
unlikely. Other comments concerning
improved clarity and completeness of
the assessment were considered by EPA
in revising the chloroform assessment
document.
4. Summary of Key Observations
Based on the available evidence, EPA
concludes that a nonlinear approach
should be considered for estimating the
carcinogenic risk associated with
lifetime exposure to chloroform via
drinking water. It should be noted that
the margin of exposure approach taken
for chloroform carcinogenicity is
consistent with conclusions reached in
a recent report by the World Health
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15687
Organization for Chloroform (WHO,
1997). The 1994 proposed MCLG was
zero for chloroform. EPA believes it
should now be 0.3 mg/L given that
hepatic Injury is the primary effect
following chloroform exposure, which
is consistent with the mode of action
understanding for chloroform. Thus, the
RfD based on hepatoxicity is considered
a reasonable basis for the chloroform
MCLG. EPA believes that the RfD used
for chloroform is protective of sensitive
groups, including children. Current
methods for developing RfDs are
designed to be protective for sensitive
populations. In the case of chloroform a
factor of 10 was used to account for
variability between the average human
response and the response of more
sensitive individuals. Furthermore, the
mode of action understanding for
chloroform does not indicate a uniquely
sensitive subgroup or an increased
sensitivity in children.
EPA continues to conclude that
exposure to chloroform may have an
adverse effect on the public health. EPA
also continues tobelieve the MCL of
0.080 mg/L for TTHMs is appropriate
despite the increase in the MCLG for
chloroform. EPA believes that the
benefits of the 1994 proposed MCL of
0.080 mg/L for TTHMs will result in
reduced exposure to chlorinated DBFs
in general, not solely THMs. EPA
considers this a reasonable assumption
at this time given the uncertainties
existing in the current health and
exposure databases for DBFs in general.
Moreover, the MCLGs for BDCM and
bromoform remain at zero and thus, a
TTHM MCL of 0.080 mg/L is
appropriate to assure that levels of these
two THMs are kept as low as possible.
In addition, the MCL for TTHMs is used
as an indicator for the potential
occurrence of other DBFs in high pH
waters. The MCL of 0.080 mg/L for
TTHMs to control DBFs in high pH
waters (in conjunction with the MCL of
0.060 mg/L for HAAS to control DBFs in
lower pH waters) and enhanced
coagulation treatment technique
remains a reasonable approach at this
time for controlling chlorinated DBFs in
general and protecting the public health.
There is ongoing research being
sponsored by the EPA, NTP, and the
American Water Works Research
Foundation to better characterize the
health risks associated with DBFs.
5. Requests for Comments
Based on the information presented
above, EPA is considering revising the
MCLG for chloroform from zero to 0.30
mg/L. EPA requests comments on this
possible change in the MCLG and on
EPA's cancer assessment for chloroform
based on the results from the ILSI report
(1997) and new data.
C. Haloacetic Acids
The 1994 proposed rule included an
MCL of 0.060 mg/L for the haloacetic
acids (five HAAs-monobromoacetic
acid, dibromoacetic acid,
monochloroacetic acid, dichloroacetic
acid, and trichloroacetic acid). An
MCLG of zero was proposed for
dichloroacetlc acid (DCA) based on
sufficient evidence of carcinogenicity in
animals, and an MCLG of 0.3 mg/L for
trichloroacetic acid (TCA) was based on
developmental toxiciry and possible
carcinogenicity. As pointed out in the
1997 NODA, several toxicological
studies have been identified for the
haloacetic acids since the 1994 proposal
(also see USEPA, 1997b).
Since the 1997 NODA, the EPA has
evaluated the significance of the 1997
ILSI panel report on the cancer
assessment for DCA. EPA has conducted
a reassessment of DCA (USEPA, 1998e)
using the principles of the EPA 1996
Guidelines for Carcinogen Risk
Assessment (USEPA, 1996b), which are
considered scientifically consistent with
the Agency's 1986 guidelines (USEPA,
1986). This reassessment has been peer
reviewed (USEPA, 1998e). Based on the
scope of the ILSI report, EPA's own
assessment and comments from peer
reviewers, the Agency believes that the
MCLG for DCA should remain as
proposed at zero. This conclusion is
discussed in more detail below.
1. 1997 International Life Sciences
Institute Expert Panel Conclusions for
Dichloroacetlc Acid (DCA)
ILSI convened an expert panel in
1996 (ILSI, 1997) to explore the
application of the USEPA's 1996
Proposed Guidelines for Carcinogen
Risk Assessment (USEPA, 1996a) to the
available data on the potential
carcinogenicity of chloroform and
dichloroacetlc acid (as described under
the chloroform section). The panel
considered data on DCA which
included chronic rodent bioassay data
and information on mutagenicity, tissue
toxicity, toxicokinetics, and other mode
of action information.
The ILSI panel concluded that the
tumor dose-response (liver tumors only)
observed in rats and mice was nonlinear
(ILSI, 1997). The panel noted that the
liver was the only tissue consistently
examined for histopathology. It further
concluded that all the experimental
doses that produced tumors in mice also
produce hepatoxicity (i.e., most doses
used exceeded the maximally tolerated
dose). Although the mode of
carcinogenic action for DCA was
unclear, the ISLI panel concluded that
DCA does not directly interact with
DNA. It speculated that the
hepatocarcinogenicity was related to
hepatotoxicity, cell proliferation, and
inhibition of program cell death
(apoptosis). The panel concluded that
the potential human carcinogenicity of
DCA "cannot be determined" given the
lack of adequate rodent bioassay data, as
well as human data. This conclusion is
in contrast to the 1994 EPA proposal in
which it was concluded that DCA was
a Group 82 probable human carcinogen.
In its current reevaluation of DCA
carcinogenicity, EPA disagrees with the
panel's conclusion that the human
carcinogenic potential of DCA can not
be determined. EPA's more recent
assessment of DCA data includes
published information not available at
the time of the ILSI panel assessment.
Based on the current weight of the
evidence EPA concludes that DCA is a
likely human carcinogen as it did in the
1994 proposed rule for the reasons
discussed below.
2. MCLG for DCA: EPA's Reassessment
of the Cancer Hazard
In the 1994 proposed rule, DCA was
classified as a Group B2, probable
human carcinogen in accordance with
the 1986 EPA Guidelines for Carcinogen
Risk Assessment (USEPA, 1986). The
DCA categorization was based primarily
on findings of liver tumors in rats and
mice, which was regarded as
"sufficient" evidence in animals. No
lifetime risk calculation was conducted
at that time; EPA proposed an MCLG of
zero (USEPA, 1994a).
EPA has prepared a new hazard
characterization regarding the potential
carcinogenicity of DCA in humans
(USEPA, 1998e). The objective of this
report was to develop a weight-of-
evidence characterization using the
principles of the EPA's 1996 Proposed
Guidelines for Carcinogen Risk
Assessment (USEPA, 1996), as well as to
consider the issues raised by the 1997
ILSI panel report. The EPA hazard
characterization relies on information
available in existing peer-reviewed
source documents. Moreover, this
hazard characterization considers
published information not available to
the ILSI panel (e.g., mutagenicity
studies). This new characterization
addresses issues important to
interpretation of rodent cancer bioassay
data, in particular, mechanistic
information pertinent to the etiology of
DCA-induced rodent liver tumors and
their relevance to humans.
Based on the hepatocarcinogenic
effects of DCA in both rats and mice in
multiple studies, and mode of action
-------
15688
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
related effects (e.g., mutational spectra
in oncogenes, elevated serum
glucocorticoid levels, alterations in cell
replication and death), EPA concludes
that DCA should be considered as a
"likely" cancer hazard to humans
(USEPA, 1998e). This is similar to the
1994 view of a B2, probable human
carcinogen, in the proposed rule.
DCA concentrations as low as 0.5 g/
L have been observed to cause a tumor
incidence in mice of about 80% and in
rats of about 20% in a lifetime
bioassays, as well as inducing multiple
tumors per animal (USEPA, 1998e).
Higher doses of DCA are associated with
up to 100% tumor incidence and as
many as four tumors per animal in a
number of studies. Time-to-tumor
development in mice is relatively short
and decreases with increased dose. The
ILSI panel concluded that doses of 1 g/
L or greater in mice produced severe
hepatotoxicity, and thus exceeded the
MTD. They further indicated that there
was marked hepatoxicity at 0.5 g/L of
DCA, (albeit not as severe as the higher
doses). EPA agrees that there was
hepatoxicity at all the doses wherein
there was a tumor response in mice. It
should be noted that the MTD selected
for the DeAngelo et al. (1991) mouse
study was a dose that results in a 10%
inhibition of body weight gain when
compared to controls. This is within the
limits designated in EPA guidelines
(USEPA, 1998e). Furthermore, no
hepatoxicity was seen in the rat studies,
where DCA induced liver tumors of
approximately 20% at the lowest dose,
0.5g/L (USEPA, 1998e). It appears that
the ILSI report did not give full
consideration to the rat tumor results as
part of the overall weight-of-the-
evidence for potential human
carcinogenicity. EPA agrees with the
ILSI panel, that the rodent assay data are
not complete for DCA; for example, full
histopathology is lacking for both sexes
in two rodent species. This deficiency
results in uncertainty concerning the
potential of DCA to be tumorgenic at
lower doses and at tissue sites other
than liver. Nevertheless, the finding of
increased tumor incidences as well as
multiplicity at DCA exposure levels (0.5
g/L) in both rats and mice where
minimal hepatotoxicity and no
compensatory replication was seen
supports the belief that observed tumors
are related to chemical treatment.
Although DCA has been found to be
mutagenic and clastogenic, responses
generally occur at relatively high
exposure levels (USEPA, 1998e). EPA
acknowledges that a mutagenic
mechanism may not be as important
influence of DCA on the carcinogenic
process at lower exposure levels as it
might be at higher exposures. Evidence
Is still accumulating that suggests a
mode of carcinogenic action for DCA
through modification of cell signaling
systems, with down-regulation of
control mechanisms in normal cells
giving a growth advantage to altered or
initiated cells (USEPA, 1998e). The
tumor findings in rodents and the mode
of action information contributes to the
weight-of-evidence concern for DCA
(USEPA, 1998e; ILSI, 1997). EPA
considers that a contribution of
cytotoxicity and compensatory
proliferation at high doses cannot be
ruled out at this time; however, these
effects were inconsistently observed in
mice at lower exposure levels, and not
at all in mice at 0.5 g/L, or in rats, at
all exposure doses. Although the shape
of the tumor dose responses are
nonlinear, there is, however, an
insufficient basis for understanding the
possible mechanisms that might
contribute to DCA tumorigenesis at low
doses, as well as the shape of the dose
response below the observable range of
tumor responses.
In summary, EPA considers the mode
of action through which DCA induces
liver tumors in both rats and mice to be
unclear. As discussed above, EPA
considers the overall weight of the
evidence to support placing DCA in the
"likely" group for human
carcinogenicity potential. This hazard
potential is indicated by tumor findings
in mice and rats, and other mode of
action data using the 1996 guideline
weight-of-evidence process. The
remaining uncertainties in the data base
include incomplete bioassay studies for
full histopathology and information on
an understanding of DCA's mode of
carcinogenic action. The likelihood of
human hazard associated with low
levels of DCA usually encountered in
the environment or in drinking water is
not understood. Although DCA tumor
effects are associated with high doses
used in the rodent bioassays, reasonable
doubt exists that the mode of
tumorgenesis is solely through
mechanisms that are operative only at
high doses. Therefore, as in the 1994
proposed rule, EPA believes that the
MCLG for DCA should remain as zero to
assure public health protection. NTP is
implementing a new two year rodent
bioassay that will include full
histopathology at lower doses than
those previously studied. Additionally,
studies on the mode of carcinogenic
action are being done by various
investigators including the EPA health
research laboratory.
3. External Peer Review of EPA's
Reassessment
Three external experts reviewed the
EPA reassessment of DCA (USEPA,
1998e). The review comments were
generally favorable. There was a range
of opinion on the issue of whether DCA
should be considered a likely human
cancer hazard. One reviewer agreed that
the current data supported a human
cancer concern for DCA, while another
reviewer believed that it was premature
to judge the human hazard potential.
The third reviewer did not specifically
agree or disagree with EPA's conclusion
of "likely" human hazard. Other issues
raised by the peer review concerned the
dose response for DCA carcinogenicity.
The peer review generally concluded on
the one hand that the mode of action
was incomplete to support a nonlinear
approach, but on the other hand, the
mutagenicity data did not support low
dose linearity. One reviewer believed
that the possibility of a low dose risk
could not be dismissed. Other
comments concerning improved clarity
and completeness of the assessment
were considered by EPA in revising the
DCA assessment document.
4. Summary of Key Observations
EPA continues to believe that
exposure to DCA may have an adverse
effect on the public health. Based on the
above discussion, EPA considers DCA to
be a "likely" cancer hazard to humans.
This conclusion is similar to the
conclusion reached in the 1994
proposed rule that DCA was a probable
human carcinogen (i.e., Group 62
Carcinogen). EPA considers the DCA
data inadequate for dose-response
assessment, which was the view in the
1994 proposed rule. EPA, therefore,
believes at this time that the MCLG
should remain at zero to assure public
health protection. The assessment that
this conclusion is based on can be found
in the docket for this Notice (USEPA,
1998e).
5. Requests for Comments
Based on the information presented
above, EPA is considering maintaining
the MCLG of zero for DCA. EPA requests
comments on maintaining the zero
MCLG for DCA and on EPA's cancer
assessment for DCA in light of
conclusions from the ILSI report (1997)
and new data.
D. Bromate
The 1994 proposed rule included an
MCL of 0.010 mg/L and an MCLG of
zero for bromate. Since the 1994
proposed rule, EPA has completed and
analyzed a new chronic cancer study in
male rats and mice for bromate
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15689
(DeAngelo et al., 1998). EPA has
reassessed the cancer risk associated
with bromate exposure and had this
reassessment peer reviewed (USEPA,
1998d). Based on this reassessment,
EPA believes that the MCLG for bromate
should remain as zero.
1. 1998 EPA Rodent Cancer Bioassay
In the cancer bioassay by DeAngelo et
al. (1998), 78 male F344 rats were
administered 0,20,100, 200,400 mg/L
potassium bromate (KBrds) in the
drinking water, and 78 male B6C3F1
mice were administered 0,80,400,800
mg/L KBrOa. Exposure was continued
through week 100. Although a slight
increase in kidney tumors was observed
in mice, there was not a dose-response
trend. In rats, dose-dependent increases
in tumors were found at several sites
(kidney, testicular mesothelioma, and
thyroid). This study confirms the
findings of Kurokawa et al. (1986a and
b) in which potassium bromate was
found to be a multi-site carcinogen in
rats.
2. MCLG for Bromate: EPA's
Reassessment of the Cancer Risk
In the 1994 proposal, EPA concluded
that bromate was a probable human
carcinogen (Group B2) under the 1986
EPA Guidelines for Carcinogen Risk
Assessment weight of evidence
classification approach. Combining the
incidence of rat kidney tumors reported
in two rodent studies by Kurokawa et al.
(1986a), lifetime risks of 10-» 10-«. and
10-« were determined to be associated
with bromate concentrations in water at
5, 0.5, and 0.05 ug/L, respectively.
The new rodent cancer study by
DeAngelo eta/. (1998) contributes to the
weight of the evidence for the potential
human carcinogenicity of KBrOs and
confirms the study by Kurokawa et al.
(1986a, b). Under the principles of the
1996 EPA Proposed Guidelines for
Carcinogen Risk Assessment weight of
evidence approach, bromate is
considered to be a likely human
carcinogen. This weight of evidence
conclusion is based on sufficient
experimental findings that include the
following: Tumors at multiple sites in
rats; tumor responses in both sexes; and
evidence for mutageniclty including
point mutations and chromosomal
aberrations in vitro. It has been
suggested that bromate causes DNA
damage indirectly via lipid
peroxldation, which generates oxygen
radicals which in turn induce DNA
damage. There is insufficient evidence,
however, to establish lipid peroxidation
and free radical production as key
events responsible for the Induction of
the multiple tumor responses seen in
the bromate rodent bioassays. The
assumption of low dose linearity is
considered to be a reasonable public
health protective approach for
extrapolating the potential risk for
bromate because of limited data on its
mode of action.
Cancer risk estimates were derived
from the DeAngelo et al. (1998) study by
applying the one stage Weibull model
for the low dose linear extrapolation
(EPA, 1998d). The Weibull model,
which is a time-to-tumor model, was
considered to be the preferred approach
to account for the reduction in animals
at risk that may be due to the decreased
survival observed in the high dose
group toward the end of the study.
However, mortality did not compromise
the results of this study (USEPA,
1998d). The animal doses were adjusted
to equivalent human doses using body
weight raised to the % power as the
interspecies scaling factor as proposed
in the 1996 EPA cancer guidelines
(USEPA, 1996b). The incidence of
kidney, thyroid, and mesotheliomas in
rats were modeled separately and then
the risk estimates were combined to
represent the total potential risk to
tumor induction. The upper bound
cancer potency (q1+) for bromate ion is
estimated to be 0.7 per mg/kg/d
(USEPA, 1998d). Assuming a daily
water consumption of 2 liters for a 70
kg adult, lifetime risks of 10-", 10-*
and 10~6 are associated with bromate
concentrations in water of 5, 0.5 and
0.05 ug/L, respectively. This estimate of
cancer risk from the DeAngelo et al.
study is similar with the risk estimate
derived from the Kurokawa et al.
(1986a) study presented in the 1994
proposed rule. The cancer risk
estimation presented for bromate is
considered to be protective of
susceptible groups, including exposures
during childhood given that the low
dose linear default approach was used
as a public health conservative
approach.
3. External Peer Review of the EPA's
Reassessment
Three external expert reviewers
commented on the EPA assessment
report for bromate (USEPA, 1998d).The
reviewers generally agreed with the key
conclusions in the document. The peer
review indicated that it is a reasonable
default to use the rat tumor data to
estimate the potential human cancer
risk. The peer review also indicated that
the mode of carcinogenic action for
bromate is not understood at this time,
and thus it is reasonable to use a low
dose linear extrapolation as a default.
One reviewer indicated that it was not
appropriate to combine tumor data from
different sites unless it is shown that
similar mechanisms are involved. EPA
modeled the three tumor sites separately
to derive the cancer potencies, and thus
did not assume a similar mode of action.
The slope factors from the different
tumor response were combined in order
to express the total potential tumor risk
of bromate. Other comments raised by
the peer reviewers concerning improved
clarity and completeness of the
assessment were considered by EPA in
revising this document.
4. Summary of Key Observations
EPA continues to believe that
exposure to bromate may have an
adverse effect on the public health. The
DeAngelo etal. (1998) study confirms
the tumor findings reported in the study
by Kurokawa etal. (1986a) and
contributes to the weight of the
carcinogenicity evidence for bromate.
EPA believes that the an MCL of 0.010
mg/L and an MCLG of zero should
remain for bromate as proposed in 1994.
The assessment that this conclusion is
based on can be found in the docket for
this Notice (USEPA, 1998d).
5. Requests for Comments
Based on the recent two-year cancer
bioassay on bromate by DeAngelo et al.
(1998), EPA is considering maintaining
the MCLG of zero for bromate. EPA
requests comments on maintaining the
zero MCLG for bromate and on EPA's
cancer assessment for bromate.
IV. Simultaneous Compliance
Considerations: D/DBP Stage 1
Enhanced Coagulation Requirements
and the Lead and Copper Rule
EPA received comment on the
Novembers, 1997 Federal Register
Stage 1 D/DBP Notice of Data
Availability that expressed concern
regarding utilities' ability to comply
with the Stage 1 D/DBP enhanced
coagulation requirements and Lead and
Copper Rule (LCR) requirements
simultaneously. Commentors stated that
enhanced coagulation will lower the pH
and alkalinity of the water during
treatment. They indicated concern that
the lower pH and alkalinity levels may
place utilities in noncompliance with
the LCR by causing violations of optimal
water quality control parameters and/or
an exceedence of the lead or copper
action levels. EPA is not aware of data
that suggests that low pH and alkalinity
levels cannot be adjusted upward
following enhanced coagulation to meet
LCR compliance requirements.
However, as discussed below, the
Agency solicits further comment and
data on this issue.
The LCR separates public water
systems into three categories: large
-------
15690
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
(>50,000), medium (550,000 but >3,300)
and small (<3,300). Small and medium
systems that do not exceed the lead and
copper action levels (90th percentile
levels of 0.015 mg/L and 1.3 mg/L,
respectively) during the required
monitoring are deemed to have
optimized corrosion control. These
systems do not have to operate under
optimal water quality control
parameters. Optimal water quality
control parameters consist of pH,
alkalinity, calcium concentration, and
phosphate and silicate corrosion
inhibitors. They are designated by the
State. Small and medium systems
exceeding the action limits must operate
under State specified optimal water
quality parameters. Large systems must
operate under optimal water quality
parameters specified by the State unless
the difference in lead levels between the
source and tap water samples is less
than the Practical Quantification Limit
(PQL) of the prescribed method (0.005
mg/L).
Maintenance of each optimal water
quality control parameter mentioned
above (except for calcium
concentration) is directly related to
meeting specified pH and alkalinity
levels at the entry point to the
distribution system and in tap samples
to establish LCR compliance. In
treatment trains that EPA is aware of,
utilities have the technological
capability to raise the pH (by adding
caustic—NaOH, Ca(OH)2) and alkalinity
(by adding Na2CO3 or NaHCO3) of the
water following enhanced coagulation
and before it enters the distribution
•system. Although certain utilities may
need to add chemical feed points to
provide chemical adjustment, pH and
alkalinity can be maintained at the
values used prior to the implementation
of enhanced coagulation. Systems that
operate with pH and alkalinity optimal
water quality control parameters should
be able to meet the State-prescribed
values by providing pH and alkalinity
adjustment prior to entry to the
distribution system. Systems that
operate without pH and alkalinity
optimal water quality control
parameters can raise the pH and
alkalinity to the levels they were at
before enhanced coagulation by
providing chemical adjustment prior to
distribution system entry.
The goal of calcium carbonate
stabilization is to precipitate a layer of
CaCO3 scale on the pipe wall to protect
it from corrosion. As the pH of a water
decreases, the concentration of
bicarbonate increases and the
concentration of carbonate, which
combines with calcium to form the
desired CaCOs, decreases. At the lower
pH used during enhanced coagulation,
it will generally be more difficult to
form calcium carbonate. However,
post—coagulation pH adjustment will
increase the pH and hence the
concentration of carbonate available to
form calcium carbonate scale. Systems
that must meet a specific calcium
concentration to remain in compliance
with optimal water quality control
parameters should not experience an
increase in LCR violations due to the
practice of enhanced coagulation
provided the pH is adjusted prior to
distribution system entry and the
calcium level in the water prior to and
after implementation of enhanced
coagulation remains the same.
EPA recognizes that the inorganic
composition of the water may change
slightly due to enhanced coagulation.
For example, small amounts of anions
and compounds that can affect
corrosion rates (Cl—, SO,t~2) may be
removed or added to the water. The
effect of these constituents is difficult to
predict, but EPA believes they should be
minimal for the great majority of
systems due to the generally modest
changes in the water's inorganic
composition and because alkalinity and
pH levels have a greater influence on
corrosion rates. Increases in sulfate
concentration due to increased alum
addition during enhanced coagulation
can actually lower the corrosion rates of
lead pipe. EPA requests comment on
whether changes in the inorganic matrix
can be quantified to allow States to
easily assess potential impacts to
corrosion control.
EPA requests comment on how
lowering the pH and alkalinity during
enhanced coagulation may cause LCR
compliance problems, given that both
pH and alkalinity levels can be adjusted
to meet optimal water quality
parameters prior to entry to the
distribution system. EPA also requests
comment on whether decreasing the pH
and alkalinity during enhanced
coagulation, and then increasing it prior
to distribution system entry, may
increase exceedences of lead and copper
action levels.
EPA is currently developing a
simultaneous compliance guidance
document working with stakeholders.
The document will provide guidance to
States and systems on maintaining
compliance with other regulatory
requirements (including the LCR)
during and after the implementation of
the Stage 1 D/DBP rule and the Interim
Enhanced Surface Water Treatment
Rule. EPA requests comment on what
issues should be addressed in the
guidance to mitigate concerns about
simultaneous compliance with
enhanced coagulation and LCR
requirements. Further, the Agency
requests comment on whether the
proposed enhanced coagulation
requirements and the existing LCR
provisions that allow adjustment of
corrosion control plans are flexible
enough to address simultaneous
compliance issues. Is additional
regulatory language necessary to address
this issue, or is guidance sufficient to
mitigate potential compliance
problems?
V. Compliance With Current
Regulations
EPA reaffirms its commitment to the
current Safe Drinking Water Act
regulations, including those related to
microbial pathogen control and
disinfection. Each public water system
must continue to comply with the
current regulations while new microbial
and D/DBP rules are being developed.
VI. Conclusions
This Notice summarizes new health
information received and analyzed for
DBFs since the November 3, 1997
NODA and requests comments on
several issues related to the
simultaneous compliance with'the Stage
1 D/DBP Rule and the Lead and Copper
Rule. Based on this new information,
EPA has developed several new
documents. EPA is requesting
comments on this new information and
EPA's evaluation of the information
included in the new documents. Based
on an assessment of the new toxicology
information, EPA believes the MCLs and
MRDLs in the 1994 proposal, and
confirmed in the 1997 FACA process,
will not change. Based on the new
information, EPA is considering
increasing the proposed MCLG of zero
for chloroform to 0.30 mg/L and the
proposed MCLG for chlorite from 0.080
mg/L to 0.80 mg/L. EPA is also
considering increasing the MRDLG for
chlorine dioxide from 0.3 mg/L to 0.8
mg/L.
VII. References
1. Bove, F.J., etal. 1995. Public
Drinking Water Contamination and
Birth Outcomes. Amer. J. EpidemioL,
141(9), 850-862.
2. Cantor KP, Hoover R, Hartge P, et
al. 1985. Drinking water source and
bladder cancer: a case-control study. In
Jolley RL, Bull RJ, Davis WP, et al. (eds),
Water chlorination: chemistry,
environmental Impact and health
effects, vol. 5. Lewis Publishers, Inc.,
Chelsea, MI pp 145-152
3. Cantor KP, Hoover R, Hartge P. et
al. 1987. Bladder cancer, drinking water
-------
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
15691
source and tap water consumption: a
case control study. JNCI; 79:1269-79.
4. Cantor KP, Lunch CF, Hildesheim
M, Dosemeci M, Lubln J, Alavanja M,
Craun GF. 1998. Drinking water source
and chlorination byproducts. I. Risk of
bladder cancer. Epidemiology; 9:21-28.
5. CMA. 1997. Sodium Chlorite:
Drinking Water Rat Two-Generation
Reproductive Toxicity Study. Chemical
Manufacturers Association. Quintiles
Report Ref. CMA/17/96.
6. Craun, G.F. 1993. Epidemiology
studies of water disinfection and
disinfection byproducts. In:
Proceedings: Safety of Water
Disinfection: Balancing Chemical and
Microbial Risk. pp. 277-303,
International Life Sciences Institute
Press, Washington, D.C.
7. Deangelo, A.B., Daniel, F.B, Stober,
J.A., and Olson, G.R. 1991. The
Carclnogenicity of Dlchloroacetic Acid
In the Male B6C3F1 mouse. Fundam.
Appli.Toxicol. 16:337-347.
8. DeAngelo AB, George MH, Kilburn
SR, Moore TM. Wolf DC. 1998.
Carcinogenicity of Potassium Bromate
Administered in the Drinking Water to
Make B6C3F1 Mice and F344/N Rats,
Toxlcologic Pathology vol. 26, No. 4 (in
press).
9. Doyle TJ. Sheng W, Cerhan JR,
Hong CP, Sellers TA, Kushi, LH, Folsom
AR. 1997. The association of drinking
water source and chlorination by-
products with cancer incidence among
postmenopausal women in Iowa: a
prospective cohort study. American
journal of Public Health. 87:7.
10. Farland, WH and HJ Gibb. 1993.
U.S. perspective on balancing chemical
and microblal risks of disinfection. In:
Proceedings: Safety of Water
Disinfection: Balancing Chemical and
Microbial Risk. pp. 3-10, International
Life Sciences Institute Press,
Washington, D.C.
11. Freedman M, Cantor KP, Lee NL,
Chen LS. Lei HH. Ruhl CE, and Wang
SS. 1997. Bladder cancer and drinking
water: a population-based case-control
study in Washington County, Maryland
(United States). Cancer Causes and
Control. 8, pp 738-744.
12. Heywood R, et al. 1979. Safety
Evaluation of Toothpaste Containing
Chloroform. III. Long-Term Study in
Beagle Dogs. J. Environ. Pathol.
Toxicolo. 2:835-851.
13. Hildesheim ME, Cantor KP, Lynch
CF, Dosemeci M, Lubin J, Alavanja M,
and Craun GF. 1998. Drinking water
source and chlorination byproducts:
Risk of colon and rectal cancers.
Epidemiology. 9:1, pp: 29-35.
14. Ijsselmuiden CB, et al. 1992.
Cancer of the Pancreas and Drinking
Water: A Population-Based Case-Control
Study in Washington County, Maryland.
Am. J. Epidemtol. 136:836-842.
15. DLSI. 1997. An Evaluation of EPA's
Proposed Guidelines for Carcinogen
Risk Assessment Using Chloroform and
Dichloroacetate as Case Studies: Report
of an Expert Panel. International Life
Sciences Institute, Health and
Environmental Sciences Institute
November, 1997.
16. Jorgenson, TA, EF Meier henry, CJ
Rushbrrok, RJ Bull, and M. Robinson.
1985. Carcinogenicity of chloroform in
drinking water to male Osborne-Mendal
rats and female B6C3F1 mice. Fundam.
Appl. Toxicol. 5:760-769.
17. Kanitz, S. et al. 1996. Association
Between Drinking Water Disinfection
and Somatic Parameters at Birth.
Environ. Health Perspectives, 104(5),
516-520.
18. King, W. D. and L. D. Marrett.
1996. Case-Control Study of Water
Source and Bladder Cancer. Cancer
Causes and Control, 7:596-604.
19. Klotz, JB and Pyrch, LA. 1998. A
Case-Control Study of Neural Tube
Defects and Drinking Water
Contaminants. New Jersey Department
of Health and Senior Services.
Sponsored by Agency for Toxic
Substances and Disease Registry.
January 1998.
20. Kurokawa et al. 1986a Long-term
in vivo Carcinogenicity tests of
potassium bromate, sodium
hypochlorite, and sodium chlorite
conducted in Japan. Environ Health
Perspect 69:221-235.
21. Kurokawa etal. 1986b. Dose
response studies on the Carcinogenicity
of potassium bromate in F344 rats after
long-term oral administration. J Natl
Cancer Inst 77:977-982.
22. Melnick, R., M. Kohn, J.K.
Dunnick, and J.R. Leininger. 1998.
Regenerative Hyperplasia Is Not
Required for Liver Tumor Induction in
Female B6C3F1 Mice Exposed to
Trihalomethanes. Tox. And Applied
Pharm. 148: 137-147.
23. McGeehin, M. A. et al. 1993. Case-
Control Study of Bladder Cancer and
Water Disinfection Methods in
Colorado. Am. J. Epidemiology,
138:492-501.
24. Mobley, S.A, D.H. Taylor, R.D.
Laurie, and RJ. Pfohl. 1990. Chlorine
dioxide depresses T3 uptake and delays
development of locomotor activity in
young rats. In: Water Chlorination:
Chemistry, Environmental Impact and
Health Effects. Vol 6. Lolley, Condie,
Johnson, Katz, Mattice and Jacobs, ed.
Lewis Publ., Inc. Chelsea MI., pp. 347-
360.
25. Morris, R.D. etal. 1992.
Chlorination, Chlorination By-products,
and Cancer: A Meta-Analyis. American
Journal of Public Health, 82(7): 955-963.
26. Morris, RD. 1997. Letter from Dr.
RD Morris to Patricia Murphy on
response to Poole Critique. December
11, 1997.
27. Murphy, PA. 1993. Quantifying
chemical risk form epidemiological
studies: application to the disinfectant
byproduct issues. In: Proceedings:
Safety of Water Disinfection: Balancing
Chemical and Microbial Risk. pp. 373-
389, International Life Sciences Institute
Press, Washington, D.C.
28. NCI. 1998. Cancer Facts, National
Cancer Institute, National Institutes of
Health, http://www.meb.uni-bonn.de/
cancer net/600314.html
29. Orme, J. D.H. Taylor, R.D. Laurie,
and RJ. Bull. 1985. Effects of Chlorine
Dioxide on Thyroid Function in
Neonatal Rats. J. Tox. and Environ.
Health. 15:315-322.
30. Poole, C. 1997. Analytical Meta-
Analysis of Epidemiological Studies of
Chlorinated Drinking Water and Cancer:
Quantitative Review and Reanalysis of
the Work Published by Morris et al.. Am
J Public Health 1992:82:955-963.
National Center for Environmental
Assessment, Office of Research and
Development, September 30, 1997.
31. Reif, J. S. et al. 1996. Reproductive
and Developmental Effects of
Disinfection By-products in Drinking
Water. Environmental Health
Prospectives. 104(10):1056-1061.
32. Rockhill, B, B. Newman, and C.
Weinberg. 1998. Use and Misuse of
Population Attributable Fraction. Am. J.
Public Health. 88(1): 15-19.
33. Savitz, D. A., Andrews, K. W. and
L. M. Pastore. 1995. Drinking Water and
Preganancy Outcome in Central North
Carolina: Source, Amount, and
Trihalomethane levels. Environ. Health
Perspectives. 103(6), 592-596.
34. Swan SH, Waller K, Hopkins B,
Windham G, Fenster L, Schaefer C,
Neutra R., 1998. A prospective study of
spontaneous abortion: Relation to
amount and source of drinking water .
consumed in early pregnancy,
Epidemiology 9(2):126-133.
35. U.S. EPA. 1979. National Interim
Primary Drinking Water Regulations;
Control of Trihalomethanes in Drinking
Water. Vol. 44, No. 231. November 29,
1979. Pp. 68624-68707.
36. U.S. EPA. 1986. Guidelines for
carcinogen risk assessment, FR
51(185):33992-34003.
37. U.S. EPA. 1991. Guidelines for
developmental toxicity risk assessment
(Notice), FR 56(234):63798-63826.
38. U.S. EPA. 1992. Guidelines for
reproductive testing. CFR 798.4700. July
1, 1992.
-------
15692
Federal Register/Vol. 63, No. 61/Tuesday, March 31, 1998/Proposed Rules
39. U.S. EPA/ILSI. 1993. A Review of
Evidence on Reproductive and
Developmental Effects of Disinfection
By-Products in Drinking Water.
Washington: U.S. Environmental
Protection Agency and International
Life Sciences Institute.
40. U.S. EPA. 1994a. National Primary
Drinking Water Regulations;
Disinfectants and Disinfection
Byproducts; Proposed Rule. FR,
59:145:38668. (July 29, 1994).
41. U.S. EPA. 1994b. Workshop
Report and Recommendations for
Conducting Epidemiologic Research on
Cancer and Exposure to Chlorinated
Drinking Water. U.S. EPA, July 19-21,
1994.
42. U.S. EPA. 1994c. U.S.
Environmental Protection Agency.
Regulatory Impact Analysis of Proposed
Disinfectant/Disinfection By-Products
Regulations. Washington, D.C.
43. U.S. EPA. 1996a. Reproductive
toxicity risk assessment guidelines, FR
61(212):56274-56322.
44. U.S. EPA. 1996b. Proposed
guidelines for carcinogen risk
assessment, FR 61(79):17960-18011.
45. U.S. EPA. 1997a. National Primary
Drinking Water Regulations;
Disinfectants and Disinfection
Byproducts; Notice of Data Availability;
Proposed Rule. Fed. Reg., 62 (No.
212):59388-59484. (Novembers, 1997).
46. U.S. EPA. 1997b. Summaries of
New Health Effects Data. Office of
Science and Technology, Office of
Water. October 1997.
47. U.S. EPA. 1997c. External Peer
Review of CMA Study -2- Generation,
EPA Contract No. 68-C7-0002, Work
Assignment B-14, The Cadmus Group,
Inc., October 9, 1997.
48. U.S. EPA. 1998a. Quantification of
Cancer Risk from Exposure to
Chlorinated Water. Office of Science
and Technology, Office of Water. March
13, 1998.
49. U.S. EPA. 1998b. Health Risk
Assessment/Characterization of the
Drinking Water Disinfection Byproduct
Chlorine Dioxide and the Degradation
Byproduct Chlorite. Office of Science
and Technology, Office of Water. March
13, 1998.
50. U.S. EPA. 1998c. Health Risk
Assessment/Characterization of the
Drinking Water Disinfection Byproduct
Chloroform. Office of Science and
Technology, Office of Water. March 13,
1998.
51. U.S. EPA. 1998d. Health Risk
Assessment/Characterization of the
Drinking Water Disinfection Byproduct
Bromate. Office of Science and
Technology, Office of Water. March 13,
1998.
52. U.S. EPA. 1998e. Dichloroacetic
acid: Carcinogenicity Identification
Characterization Summary. National
Center for Environmental Assessment—
Washington Office. Office of Research
and Development. March 1998.
53. U.S. EPA. 1998f. NCEA Position
Paper Regarding Risk Assessment Use of
the Results from the Published Study:
Morris et al. Am J Public Health
1992:82:955-963. National Center for
Environmental Assessment, Office of
Research and Development, October 7,
1997.
54. U.S. EPA. 1998g. Synthesis of the
Peer-Review of Meta-analysis of
Epidemiologic Data on Risks of Cancer
from Chlorinated Drinking Water.
National Center for Environmental
Assessment, Office of Research and
Development, February 16, 1998.
55. U.S. EPA. 1998h. EPA Panel
Report and Recommendation for
Conducting Epidemiological Research
on Possible Reproductive and
Developmental Effects of Exposure to
Disinfected Drinking Water. Office of
Research and Development.
56. U.S. EPA. 1998L Final guidelines
for neurotoxicity risk assessment.
57. Vena JE, Graham S, Freudenheim
JO, Marshall J, Sielezny M, Swanson M,
SufrinG. 1993. Drinking water, fluid
intake, and bladder cancer in western
New York. Archives of Environmental
Health. 48:(3)
58. Waller K, Swan SH, DeLorenze G,
Hopkins B., 1998. Trihalomethanes in
drinking water and spontaneous
abortion. Epidemiology. 9(2): 134-140.
59. WHO. 1997. Rolling Revision of
WHO Guidelines for Drinking-Water
Quality; Report of Working Group
Meeting on Chemical Substances for the
Updating of WHO Guidelines for
Drinking-Water Quality. Geneva,
Switzerland, 22-26 April 1997.
National Primary Drinking Water
Regulations: Disinfectants and
Disinfection Byproducts Notice of Data
Availability page 86 of 86.
Dated: March 24,1998.
Robert Perciasepe,
Assistant Administrator for Office of Water.
[FR Doc. 98-8215 Filed 3-30-98; 8:45 am]
BILLING CODE 6560-60-U
-------
-------
-------
•60- TJ
CO CD
2 3
O JB
o'
TO
I-
CD
CO
0
CD
CO
3
CD
------- |