DEQ-CFW_00002018

J�KSo SP,gr�
Ank
UNITED STATES ENVIRONMENTAL PROTECTION
j;`,/j c� - -• AGENCY
t� OFFICE OF CRIMINAL ENFORCEMENT, FORENSICS AND TRAINING
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 25, BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
November 17, 2006
MEMORANDUM
TO: Walker Smith, Esq., Director
Office of Civil Enforcement -
THROUGH: Diana A. Love, Director ` L �'
National Enforcement estzgatvions Cent6r
FROM: Christopher P. Weis, Ph.D., D
Senior Toxicologist
SUBJECT: Hazard Evaluation and Revised Site -Specific Threshold for Perfluorooctanoate
(PFOA or C8; CAS #335-67-1) in drinking water near the DuPont Washington
Works facility, West Virginia. (Rev. 1)
The purpose of this memorandum is to present an overview of emergent and historical
toxicological data in light of site -specific human exposure information on perfluorooctanoate (PFOA
or C8; hereafter referred to as C8) and to provide recommendations relevant to these exposures in
the vicinity of the DuPont Teflon® manufacturing facility known as the Washington Works facility
near Parkersburg, WV. This memorandum is not intended to provide a comprehensive review of the
available literature on C8. Such a review may be found in the USEPA draft risk assessment for this
compound (1). Rather, it is intended to express NEIC's concerns about elevated human exposures
near the Washington Works facility and to provide precautionary recommendations for the
protection of communities affected by release of C8 from the Washington Works facility.
As part of the effort by the Office of Pollution Prevention and Toxics (OPPT) to understand
health and environmental issues presented by fluorochemicals in the wake of unexpected
toxicological and bioaccumulation discoveries with respect to perfluorooctane sulfonates (PFOS),
OPPT has been investigating C8 and its salts. The Director of OPPT initiated a priority review
(9/27/2002) of C8 and its salts which resulted in the development of an ongoing series of hazard and
risk assessments. In February 2005, the SAB reviewed., the draft risk assessment, and finalized their
comments in May 2606.
C8, a perfluorinated carboxylic acid with the general formula C8HFI5O2, is man-made and
does not occur naturally in the environment. C8 is an environmentally stable surfactant with an
extremely long elimination half life in humans; the current estimate is a mean of 3.8 years with a
range of 1.5 to 13.5 years (2). Because of this long elimination half life, C8 fiends to occur in human
tissues at concentrations greater than those found in the surrounding environment.
Recent human exposure studies were conducted by the University of Pennsylvania. The
study participants included 326 residents from four communities in southeastern Ohio near DuPont's
Washington Works facility who demonstrated average C8 concentrations in blood serum ranging
Page 1 of 6
DEQ-CFW 00002018

from 298 to 369 parts per billion (3, 4_(ppb; range 7 to 4,520 ppb). The highest exposures in this
study were found in children and in the elderly. These blood serum levels are substantially higher
than the 5 ppb nationwide average identified in other studies (5). The researchers in the University of
Pennsylvania study found no evidence of toxicity based upon limited assessment. However, in the
human exposure study, the investigators
did not examine the toxic endpoints that Figure 1: Age and C8 Levels in Blood Measured in the
Residential Population near the Washington Works
were of concern in animal studies
including increased liver weights, 1006 Facility (2006)
cytological changes in the liver, alterations
in gene expression, or developmental 800
effects. The available data do not provide
a definitive picture of the presence or ;, 600
absence of C-8 effects on human health. g
400
C8 can be absorbed through the
ingestion, inhalation, and dermal routes of
204
exposure. Possible exposure pathways for
C8 near the Washington Works facility
0
Mior•Q�rBta ren0a
Mean, •
Median...-.
include drinking water in affected areas C6 6.10 11.15 16.20 21.30 3140 41.50 61.60 >60
! and water used for cooking, washing, Age (years)
laundry, and showering. Limited data
from the Ohio study suggest that drinking water is the primary exposure route, and that eating home-
grown vegetables and fruits from this vicinity may increase exposures to C8. However, thorough
exposure pathway analyses for residents have not yet been completed and additional pathways may
be present. Ambient air emissions from the Washington Works facility may contribute to human
exposure in the vicinity of the plant through direct or indirect contamination of groundwater and
contributions to household dust. Workers at the facility may accidentally carry dust home on
clothing and indirectly expose household members.
Studies in rodents have shown that C8 toxicity is similar through both ingestion and inhalation
pathways (6). Since C8 is expected to cross the skin, similar toxic responses are expected from
dermal exposure.
C8 is carcinogenic in rodents by multiple mechanisms (7, 8) and in multiple organ systems (9-
11). Some tumors caused by C8 are associated with a biochemical process known as peroxisome
proliferation. There is an ongoing scientific debate regarding the relevance of these tumors for adult
humans. Nonetheless, C8 may cause tumors and non -cancerous toxicity by mechanisms
independent of peroxisome proliferation and these may be of concern to humans.
C8 displays moderate to severe non -cancer toxicity in short term dosing studies conducted in
primates. Exposure of Cynomolgus monkeys to C8 for six months demonstrated liver toxicity in all
dose groups tested. This toxicity was not related to peroxisome proliferation (12) though the exact
mechanism is unclear. One high dose animal in this study suffered severe treatment -related effects
and had to be euthanized. A second animal, in the low dose group (3 mg/kg-d), was euthanized on
day 127 of the study and, although the cause of morbidity was unclear, treatment -related effects
could not be ruled out. Rhesus monkeys (13) were exposed to C8 for 13 weeks at doses ranging from
3 to 30 mg/kg-day and displayed clinical signs of toxicity that included gastrointestinal distress,
reductions in body weight in all groups tested, and mortality in the high dose group (30 mg/kg-day).
Page 2 of 6
DEQ-CFW 00002019

A variety of scientific information has emerged since development of the 2002 drinking _ y
water action level was developed for the vicinity of the Washington Works plant (4, 14). Recent
developmental studies (15) in mice suggest toxicity findings for C8 that were not available for
consideration in the 2002 risk calculations conducted by the West Virginia Department of
Environmental Protection (16). These developmental studies suggest that in utero exposure of mice
to C8 during late pregnancy results in a striking dose -related perinatal mortality (similar findings
were observed for perfluorooctane sulfonate, PFOS).
A margin of exposure (MOE) approach can be used to describe the potential for human health
effects associated with exposure to a chemical. The MOE is usually based on external exposure
levels, and is calculated as the ratio of the No Observable Adverse Effect Level (NOAEL) or Lowest
Observable Adverse Effect Level (LOAEL) for a specific toxicological endpoint in an animal or
human study to the estimated human exposure level at the site of concern. The MOE does not
provide an estimate of population risk, but simply describes the relative "distance' between the site -
specific exposure level and the NOAEL or LOAEL. In general, a smaller MOE means greater
potential risk that humans may experience toxicity. As noted above, there are substantial species
differences in the elimination of C8 from the body (17, 18). This means that two different species
could have vastly different blood levels resulting from the same exposure level. Thus, for C8 it is
necessary to use an actual measure of C8 in the body (known as internal dose) when estimating a
margin of exposure between humans and animal test species. This approach was used in the draft
OPPT risk assessment and was supported by the SAB (19).
The studies conducted in Cynomolgus monkeys (12) and Rhesus monkeys (13) showed
toxicity in the low dose group (3 mg/kg-day) with steady-state serum C8 concentrations of 77,000
ppb. Recent human exposure studies conducted in 326 individuals near the Washington Works
facility (Figure 1) measured mean blood concentrations of C8 of 369 ppb (3, 4) the range of blood
concentrations was 7 to 4520 ppb. The average MOE for C8 in this population is therefore,
approximately 209 with a range of 17 to 11,000, This represents an extremely narrow MOE. Given
uncertainties in the toxicology of C8 and the severity of the effects (mortality) observed in the
monkey studies, this MOE may represent significant average risk to the exposed human population
and high risk to highly exposed individuals within the population.
As a result of the weight of scientific evidence regarding the toxicity and toxicokinetics of C8,
including emergent data such as: 1) evidence of high blood serum levels in human populations
adjacent to the Washington Works facility and the resultant narrowing MOE, 2) animal studies
demonstrate potential for developmental toxicity including mortality, 3) steep and unstable dose -
response relationships in non -human primates that often culminate in mortality and, 4) evidence of
carcinogenicity in laboratory animals in multiple organ systems, NEIC recommends that steps be
taken to eliminate or reduce human exposure to C8 in the vicinity of the Washington Works facility.
This includes a site -specific drinking water action levef of 0.5 ug/L for the communities surrounding
the Washington Works facility (see Appendix 1). We feel that it is particularly important to address
ongoing exposure to the residential population, especially pregnant women, neonates, and the elderly
by identification and elimination of multiple exposure pathways associated with C8 in water
distribution systems. This is a precautionary recommendation to reduce exposure to the population
living in the vicinity of the Washington Works facility.
Page 3 of 6
DEQ-CFW 00002020

Reference List
1. USEPA. Draft risk assessment of the potential human health effects associated with exposure
to perfluorooctanoic acid and its salts. 2005. Risk Assessment Division, Office of Pollution
Prevention and Toxics. 1-4-2005.
2. J. Burris. Determination of serum half-lives of several fluorochemicals. Interim Report #2.
Lundberg, JK Olsen GW Simpson:C Mandel J. 2002. 3M Medical Department.
3. Janet Raloff. Non stick pollution sticks in people.168(9). 2005. Science News.
4. E. A. Emmett et al., J. Occup. Environ. Med. 48, 759 (2006).
5. A. M. Calafat, Z. Kuklenyik, S. P. Caudill, J. A. Reidy, L. L. Needham, Environ. Sci. Technol.
. 40, 2128 (2006).
6. R. E. Staples, B. A. Burgess, W. D. Kerns, Fiwdam. Appl. Toxicol. 4,429 (1984).
7. L. Sibinski. Two-year oral (diet) toxicity/carcinogenicity study of fluorochemical FC-143
(perfluorooctane ammonium carboxylate) in rats. (3M No.0281 CRO01 2: 8EHQ-1087-0394).
1991. Report prepared for 3M, St. Paul, MN by Riker Laboratories Inc., Riker Laboratories,
Inc.
8. L. B. Biegel, M. E. Hurtt, S.R. Frame, J. C. O'Connor, J. C. Cook, Toxicol. Sci. 60,44 (2001).
9. J. C. Cook, M. E. Hurtt, Frame S.R., Biegel L.B. Mechanisms of extrahepatic tumor induction
by peroxisome proliferators in Crl:CDBR (CD) rats. Toxicologist 14(301). 1994.
10. A. G. Abdellatif, V. Preat, H. S. Taper, M. Roberfroid, Toxicol. Appl. Phannacol. 111, 530
(1991).
11. P. C. Mann. Two year oral toxicity-oncogenicity study in rats: Peer review of ovaries. Frame,
S.R. 2004. Dupont Project ID 15261.
12. J. Butenhoff et al., Toxicol. Sci. 69, 244 (2002).
13. E. I. Goldenthal. Final Report: Ninety day subacute Rhesus monkey toxicity study. 1978.
International Research and Development Corporation.
14: J. L. Butenhoff et al., Toxicol. Sci. 82, 394 (2004).
15. C. Lau et al., Toxicol. Sci. 90, 510 (2006).
16. Department of Environmental Protection. Final Ammonium Perfluorooctanoate (C8)
Assessment of Toxicity Team (CATT) Report. 2002.
17. J. P. Vanden Heuvei, B.1. Kuslikis, M. J. Van Rafelghem, R. E. Peterson, J Biochem. Toxicol.
6,83 (1991).
18. J. L. Butenhoff et al., Toxicol. Sci. 82, 394 (2004).
Page 4 of 6
DEQ-CFW 00002021

19. M. Morgan, D. Cory-Slechta. SAB Review of EPA's Draft Risk Assessment of Potential
Human Health Effects Associated with PFOA and Its Salts. 5-30-2006.
http://wwww.gpa..gov/sab/ dfp /sab 06 006.pdf.
20. Olsen G, J. Ehresman, J. Froehlich, Burris, Butenhoff. Evaluation of the half-life (tI/2) of
elimination of perfluorooctanesulfonate (PFOS), perfluorohexanesulfonate (PFHS) and
perfluorooctanoate (PFOA) from human serum G. 1-8-0005.3M Company, St. Paul, MN and
Pace Analytical, St. Paul, MN, Presented at the First International Conference on Perfluoros in
Toronto. Aug., 2005.
Page 5 of 6
DEQ-CFW 00002022

Appendix 1:
Generation of a recommended site -specific drinking water (DW) Action Level for protection of
infants and children exposed to C8 associated with the Washington Works facility,
Washington, WV.
LOAEL x BW x RSC
Recommended Site -Specific Action Level for DW (SAL) = OF x IR
Chemical Toxicity endpoint/ LOAEL OF RSC SAL for DW
Critical Study(12)
C8 Liver Disease 3 mg/kg-d 1.35E4 0.2 0.5 Ag/L*
CAS #335-67-1 and Clinical toxicity
LOAEL = Lowest Observable Adverse Effect Level (12)
BW = Body Weight (10 kilograms)
RSC = Relative Source Contribution — EPA typically estimates that 20% of exposure is derived from
drinking water. Other exposure sources include (but are not limited to) dermal exposure, ingestion of
garden vegetables, and inhalation of volatile C8.
(littp://www.epa.gov/waterscience/criteria/humanhealth/method/complete.pdf)
IR = Ingestion Rate (0.85 liters/day: http://www.epa.gov/ncea/efb/pdfs/efh-chapter03.pdf)
OF = Uncertainty Factor (as indicated below)
10X for extrapolation of a LOAEL to a NOAEL
1 OX for intraspecies variability
135X for interspecies variability ---- [3 for toxicodynamics - 45mos(=0)/lmos] (clearance in
humans)/(clearance in monkeys)
' This site -specific drinking water action level is associated with both uncertainty and variability. The
estimate, though intended to be protective of the general population (including infants, children, and
the elderly) in the vicinity of the Washington Works facility, is dependent upon assumptions regarding
both exposure and toxicity of PFOA. Actual risks may vary as a function of individual exposure (such
as drinking water intake) and biological susceptibility.
Page 6 of 6
DEQ-CFW 00002023