BY ELECTRONIC MAIL
October 1, 2015
Ms. Nancy Rice
Minnesota Department of Health
625 Robert Street North
P.O. Box 64975
St. Paul, MN 55164-0975
Re: Proposed Amendments to Rules Governing Health Risk Limits for Groundwater,
Minnesota Rules, Chapter 4717, Part 7500, Part 7850, Part 7860, and Part 7865
Dear Ms. Rice:
The American Chemistry Council (ACC) 1 appreciates the opportunity to submit
comments on the proposal to replace the current Health Risk Limit (HRL) of 5 micrograms per
liter (µg/L) for 1,1,2-trichloroethylene (TCE) with separate HRLs for short-term, subchronic, and
chronic non-cancer endpoints as well as an HRL for a cancer endpoint. Our comments focus on
concerns with the Department of Health’s (MDH) use of non-guideline studies reporting
immune and developmental (i.e., fetal cardiac malformations) effects as a basis for the
proposed non-cancer HRLs.
Several critical factors - including parental toxicity, coincident toxicity, control of litter
effects, and positive control data - were not included in the report by Peden-Adams et al.
(2006) of immune effects identified by the Department as “critical.” The lack of this
information makes interpretation challenging and prevents comparison with results from
1
ACC is a national trade association representing the leading companies engaged in the business of chemistry
ACC members apply the science of chemistry to make innovative products and services that make people's
lives better, healthier and safer. ACC is committed to improved environmental, health and safety
performance through Responsible Care®, common sense advocacy designed to address major public policy
issues, and health and environmental research and product testing. The business of chemistry is an $801
billion enterprise and a key element of the nation's economy. It is the nation’s largest exporter, accounting for
fourteen percent of U.S. exports. Chemistry companies are among the largest investors in research and
development. Safety and security have always been primary concerns of ACC members, and they have
intensified their efforts, working closely with government agencies to improve security and to defend against
any threat to the nation’s critical infrastructure.
americanchemistry.com®
700 Second St., NE | Washington, DC | 20002 | (202) 249-7000
Ms. Nancy Rice
October 1, 2015
Page 2
studies conducted according to US Environmental Protection Agency (EPA) guidelines. 2 The
developmental effects that MDH has identified as “co-critical,” moreover, are not consistent
with results from other well-conducted studies and have been the subject of considerable
controversy since they were first published by Johnson et al. (2003). These issues are described
in greater detail below. In addition, ACC is concerned about the Department’s use of default
values for relative source contribution (RSC) and intake rates used in calculating the non-cancer
HRLs.
Discussion of Results Reported by Peden Adams et al. (2006)
Although Peden-Adams et al. described their study as a developmental immunotoxicity
research, it was not designed to assess potential developmental effects. Exposure to 1.4 or 14
parts per million in drinking water occurred from gestation day 0 through sacrifice which
occurred at weaning (3 weeks) or at adulthood (8 weeks). In addition, the small number of
offspring analyzed made per-litter analysis impractical. As noted in EPA’s guidelines for
developmental toxicity, statistical analysis should be conducted on a per-litter basis because,
during gestation, the dam is the unit of treatment and exposure to the pups is dependent on
her. 3 Performing statistics on a per-fetus basis inflate the significance of the findings.
Among the challenges with interpreting the results reported by Peden-Adams et al.
study is the inability to determine the amount of water consumed by dams and pups. The
authors state that the water was changed twice per week, but no mention of the consumption
rates for each pup over the observation period is provided. Furthermore, the authors indicate
that an emulsifier was used to keep the TCE in solution and distributed in the water column;
however, it is unclear how the emulsifier affected the concentrations per drinking volume
consumed. The actual ingested dose of TCE may vary across the test animals with larger
amounts of TCE taken in by certain pups. This fact was not addressed in the study and could
have resulted in greater intakes and exposures to TCE over the course of the study by certain
dams and pups compared to others. Gastric lavage would have been the appropriate dosing
mechanism as this would allow a specific measured dose of TCE to be administered to each
animal.
As a result, the Peden-Adams et al. findings are based on potential exposure
concentrations rather an actual administrated dose. The actual daily dose of TCE to which the
mice were exposed is unknown, clearly calling into question the actual dosing and dosing group
2
EPA. Health effects guideline for immunotoxicity. OPPTS 870.7800. EPA 712–C–98–351 (August 1998). Available
at http://www.epa.gov/ocspp/pubs/frs/publications/Test_Guidelines/series870.htm.
3
EPA Risk Assessment Forum. Guidelines for developmental toxicity risk assessment. EPA/600/FR-91/001
(1991). Available at http://www2.epa.gov/sites/production/files/2014-11/documents/dev_tox.pdf.
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 3
results. In addition, no data are provided to verify TCE concentrations in the drinking water. 4
This means animals may have actually been exposed to much higher concentrations of TCE than
reported. These uncertainties significantly decrease the credibility of the study and call into
question the MDH dependence on a non-guideline drinking water study.
Among the reported findings, Peden-Adams et al. observed suppression of plaqueforming cell (PFC) response, modulation of T-cell numbers, and stimulation of the delayed type
hypersensitivity (DTH). Suppression of humoral immunity was observed at both ages in male
offspring, but was more pronounced in females at 8 weeks of age. Suppression of this response
did not appear to be related to B-cell numbers which were not consistently decreased. While
the authors suggest the response may be related to a reduction in B or T lympohocytes, T- or Bcell mitogen-induced proliferation was not altered, indicating that lymphocytic proliferative
responses were normal.
Transient alterations in splenic T-cell subpopulations were observed in the 3-week old
pups, but were not altered in offspring sacrificed after 8 weeks. At 3 weeks of age, a significant
decrease in thymic T-cells subpopulations (i.e., CD4-/CD8-) was observed in pups exposed to 14
ppm TCE. Numbers of all thymic T-cell populations were increased at 14 ppm in offspring at 8
weeks of age, but was only significantly increased in double positive cells at 1.4 ppm. DTH
response was only assessed in the offspring at 8 weeks of age.
Although it is considered a sensitive indicator of immunotoxicity, alteration in thymus
mass was not reported by Peden-Adams et al. at either dose at either age. This observation
conflicts with the results of a similar study conducted at the same laboratory after 30 weeks of
exposure (Keil et al. 2009). 5 No significant increase in anti-ds-DNA antibodies was reported in
blood samples after 8 weeks of exposure, moreover. Anti-ds-DNA antibodies are a sensitive
measure of autoantibody production and an important marker of lupus-like symptoms. Natural
killer cell activity also was not altered by exposure to TCE in either age group.
Discussion of Results Reported by Johnson et al. (2003)
The report of fetal heart malformations (FHM) by Johnson et al. (2003) is inconsistent
with results from well conducted, guideline studies in rats exposed to TCE by inhalation at
doses up to 600 ppm (Carney et al. 2006)6 and by oral gavage to 500 mg/kg/day of TCE (Fisher
4
Although the authors report that drinking water concentrations were analyzed through the course of the
experiment, no data are provided.
5
Keil DE et al. Assessment of trichloroethylene (TCE) exposure in murine strains genetically-prone and nonprone to develop autoimmune disease. Sci Health A Toc Hazard Subst Environ Eng 44(5):443-453 (2009).
6
Carney E et al. Developmental toxicity studies in Crl:CD (SD) rats following inhalation exposure to
trichloroethylene and perchloroethylene. Birth Defects Res B Dev Reprod Toxicol 77(5):405–412 (2006).
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 4
et al. 2001). 7 Neither of these studies reported exposure related developmental toxicity, even
in the presence of maternal toxicity. Furthermore, neither of these studies reported evidence
of specific cardiac teratogenicity, even when the microdissection technique employed by
Johnson et al. was used and a member of that research group was part of the study team
(Fisher et al. 2001).
In the original drinking water study conducted by the Johnson et al. group, water
containing 1.5 or 1100 ppm TCE 8 was provided before conception, during pregnancy, or both
before and during pregnancy (Dawson et al. 1993). 9 With drinking water exposure before
conception, no impact on mating success or intrauterine survival was observed, and
pregestational exposure alone had no influence on heart defects. The number of fetuses
reported with “abnormal hearts” was significantly increased for dams exposed to 1.5 ppm TCE
before conception and during pregnancy, but not for those exposed at that level only during
pregnancy. Abnormal hearts were significantly increased in fetuses in both groups
(before/during conception, during conception only) of dams exposed to 1100 ppm TCE.
In the Johnson et al. study, rats were given drinking water containing TCE at 0.0025
ppm, 0.25 ppm, 1.5 ppm, or 1100 ppm. Cardiac defects were reported to be significantly
increased at the 0.25 ppm and 1100 ppm exposure concentrations, but not at intermediate
level of 1.5 ppm. 10 Further evaluation of the 2003 study by Hardin et al. (2004) confirmed that
results from the two higher doses were the same as those reported in 1993 and that the
control group represented a combination of an unspecified number of historic controls rather
than one control group run concurrently with the two lower exposed groups (Johnson et al.
2004). In the absence of a clear dose-response relationship, it is difficult to conclude that the
observed effects were treatment-related.
Several possible explanations have been suggested for the consistent positive findings of
an association between TCE exposure and FHM incidence in rats reported by Johnson et al. –
higher TCE concentrations, the mode and timing of exposure, differences in detection
techniques, or the use of non-standard statistical evaluations (Watson et al. 2006).11 The
7
Fisher JW et al. Trichloroethylene, trichloroacetic acid, and dichloroacetic acid: do they affect fetal rat heart
development? Int J Toxicol 20(5):257–267 (2001).
8
The concentrations used in the drinking water studies correspond to the following daily doses: 2.5 ppb 0.00045 mg/kg/day; 250 ppb – 0.048 mg/kg/day; 1.5 ppm – 0.218 mg/kg/day; 1100 ppm – 129 mg/kg/day
(Johnson et al. 2003).
9
Dawson BV et al. Cardiac teratogenesis of halogenated hydrocarbon-contaminated drinking water. J Am Coll
Cardiol 21(6):1466–1472 (1993).
10
The 2.5 ppb exposure level showed no effect on FHM, despite the fact that 16.4% of control litters had a
cardiac defect.
11
Watson RE et al. Trichloroethylene-contaminated drinking water and congenital heart defects: a critical
analysis of the literature. Reprod Toxicol 21(2):117–147 (2006).
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 5
findings cannot be explained by the high concentrations since Fisher et al. also dosed dams at
500 mg/kg/day TCE by gavage. 12 Exposure of the dams throughout pregnancy in the study by
Johnson et al., rather than limiting exposure to the most sensitive period of organogenesis
(gestation days 6 through 15 or GD 6-15) as in Fisher et al., similarly is unlikely to explain the
difference since the heart is formed during GD 6-15 and any exposure before or after this
period would not increase FHM. Other studies in which dams were exposed for all or most of
the pregnancy also failed to observe a FHM increase.
Although Johnson et al. is the only study to investigate FHM in rats exposed to TCE
through drinking water, the difference in route of exposure cannot explain the positive results
reported. Oral gavage studies, such as that conducted by Fisher et al., will result in higher
blood concentrations than those using drinking water exposures (Watson et al. 2006).
Consequently, gavage studies would be expected to be more likely to cause developmental
effects than the drinking water study.
While Johnson et al. have suggested that the dissection technique used in their studies
may be more sensitive in detecting certain lesions, those lesions are not the predominant FHM
types reported in the study. Perhaps more significantly, Fisher et al. failed to observe an
increase in FHM despite collaborating with the Dr. Johnson and using the same dissection
method.
It appears, therefore, that one significant reason for the positive results reported by
Johnson et al. is that the statistics were performed differently than in traditional developmental
studies. Original statistics were performed on a per-fetus basis, rather than on a per-litter
basis, despite the fact that per-litter analysis is the accepted method for developmental effects
related to chemical exposure during pregnancy, as recommended by EPA. Performing statistics
in a per-fetus manner artificially inflates the significance of the findings. Had the correct
statistical unit been used in these studies, a positive correlation between TCE and FHM
probably would not have been reported in the original drinking water studies (Watson et al.
2006).
In the later publication, Johnson et al. re-published data from the earlier study (Dawson
et al. 1993), using pooled controls from all of their studies in their statistical evaluation. Pooling
of controls is not an appropriate statistical practice and is likely to have exaggerated the alleged
statistical significance (Hardin et al. 2004). 13,14 Fisher et al., moreover, express concern that
12
As noted, the highest dose in the Johnson et al. study was 129 mg/kg/day.
13
Hardin BD et al. Trichloroethylene and cardiac malformations [Letter]. Environ Health Perspect 112(11):A607–
A608 (2004).
14
USEPA’s benchmark dose guidance (USEPA 2012) requires concurrent controls in key studies to be used in the
calculation of an RfD or RfC.
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 6
“[t]he high background of fetal heart malformations on a per litter basis provides a challenge
for using these data in regulatory decisions relating to risk characterization of TCE, TCA, and
DCA.”
Fisher et al. also note that the lack of clear dose-related effects in the study by Dawson et al.
and their own study provide “data of questionable utility for risk assessment applications.” In
an attempt to provide support for a dose-response, Johnson et al. (2003) present a doseresponse curve, based on a probit analysis, at concentrations up to 4878 ppm. The
concentration of 4878 ppm is well above water solubility for TCE, however, and the authors fail
to explain how they could generate a curve using concentrations for which no data exist.
In considering the data from Johnson et al., and earlier studies from the same
laboratory, NRC noted that the rodent studies showing trichloroethylene-induced cardiac teratogenesis at
low doses were performed by investigators from a single institution. Also noted
were the unusually flat dose-response curves in the low-dose studies from these
investigators. . . Thus, the animal data are inconsistent, and the apparent species
differences have not been addressed. 15
The species differences remain unaddressed, despite the eight years that have passed since
NRC raised its concern. California’s Office of Environmental Health Hazard Assessment
(OEHHA), moreover, explained its decision to reject the findings accordingly Johnson et al. (2003) reported a dose-related increased incidence of abnormal
hearts in offspring of Sprague Dawley rats treated during pregnancy with 0, 2.5
ppb, 250 ppb, 1.5 ppm, and 1,100 ppm TCE in drinking water (0, 0.00045, 0.048,
0.218, and 128.52 mg/kg-day, respectively). The NOAEL for the Johnson study
was reported to be 2.5 ppb (0.00045 mg/kg-day) in this short exposure (22 days)
study. The percentage of abnormal hearts in the control group was 2.2 percent,
and in the treated groups was 0 percent (low dose), 4.5 percent (mid dose 1), 5.0
percent (mid dose 2), and 10.5 percent (high dose). The number of litters with
fetuses with abnormal hearts was 16.4 percent, 0 percent, 44 percent, 38
percent, and 67 percent for the control, low, mid 1, mid 2, and high dose,
respectively. The reported NOAEL is separated by 100-fold from the next higher
dose level. The data for this study were not used to calculate a public-health
protective concentration since a meaningful or interpretable dose-response
relationship was not observed. These results are also not consistent with earlier
15
NRC. Assessing the Human Health Risks of Trichloroethylene: Key Scientific Issues. Washington, DC: The
National Academies Press (2006), at 171. Available at http://www.nap.edu/catalog.php?record_id=11707.
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 7
developmental and reproductive toxicological studies done outside this lab in
mice, rats, and rabbits: The other studies did not find adverse effects on fertility
or embryonic development, aside from those associated with maternal toxicity
(Hardin et al., 2004). 16
Discussion of Default Values Used in the Derivation of the Non-Cancer HRLs
Notwithstanding our concerns about the critical and co-critical studies identified by
MDH, ACC is concerned about the default values used for the relative source contribution (RSC)
and intake rate in calculating the non-cancer HRLs for TCE. The RSC concept has been used for
conducting risk assessments of chemicals in drinking water for many years and is an estimate of
the proportion of the total daily exposure to a chemical that is attributed to or allocated to tap
water (accounting for multi-route exposures) in calculating acceptable levels of chemicals in
water. RSC values typically range from 0.2 to 0.8, with a default of 0.2. Use of a default RSC of
0.2 in the development of federal Drinking Water Standards and state regulatory program
ground water standards and screening levels is, however, based on tradition, not scientific data.
MDH’s use of an RSC factor of 0.2 suggests that 20% or less of daily exposure to TCE is
attributable to exposure from drinking water, with the other 80% of daily exposure attributed
to sources which have not been described in the document. Given the dramatic reduction in
the use of TCE (and many other volatile contaminants) in consumer and commercial products
and the significant reduction in ambient levels of TCE, 17 an RSC of 0.2 underestimates the
current contribution of drinking water to the total exposure to TCE. Rather than use a default
value developed nearly 20 years ago when TCE use was considerably higher, MDH should use a
value that is more relevant to today’s exposure.
In light of the various concerns with the studies that are the basis of MDH’s proposed
HRLs, calculation of short-term and sub-chronic HRLs lacks sufficient justification. This is
particularly true since these values are based on intake rates for fetal and/or infant exposure.
As described above, neither of the studies identified by MDH provide sufficient evidence for
developmental effects. If MDH prefers to continue to rely on the study by Peden–Adams et al.,
it should only be used as a basis for a chronic HRL since the bulk of the reported effects were
observed in animals at 8 weeks of age. However, I have enclosed a systematic review of the
studies that EPA identified as candidates for developing its reference values which can provide
far better basis for calculating an appropriate chronic HRL.
16
Office of Environmental Health Hazard Assessment (OEHHA). Public health goals for chemicals in drinking
water – trichloroethylene. OEHHA. Sacramento, CA (2009). Available at
http://www.oehha.ca.gov/water/phg/pdf/TCE_phg070909.pdf.
17
Agency for Toxic Substances and Disease Registry (ATSDR). Draft Toxicological Profile for Trichloroethylene
(October 2014). Available at http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=173&tid=30.
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000
Ms. Nancy Rice
October 1, 2015
Page 8
Please do not hesitate to contact me at [email protected] or at (202)
249-6727 if you have questions about the information outlined above.
Sincerely,
Steve Risotto
Stephen P. Risotto
Senior Director
Enclosure
americanchemistry.com®
700 Second St., NE | Washington, DC 20002 | (202) 249.7000