DEQ-CFW_00000021

ONE HUNDRED -AND -THIRTY NINETH MEETING OF THE
SCIENCE ADVISORY BOARD (NCSAB) ON TOXIC AIR POLLUTANTS
Proceedings of the February 25, 2009 Teleconference
Dr. Starr called the meeting to order at 2:14PM. NCSAB Members Drs. Thomas Starr, Wayne Spoo,
Elaina Kenyon, Man -Sung Yim and James Gibson were in attendance via teleconference. Dr.
Woodhall Stopford was not in attendance. Andrew Jacques of the AN Group, Sandra Moore, Connie
Brower and Nikki Remington of NCDWQ, Dr. Christopher Lau of USEPA and Dr. Robert Rickard and
Mike Johnson of DuPont participated in the teleconference.
PFOA Presentations
Drs. Christopher Lau, USEPA and Robert Rickard, DuPont presented updates on PFOA research.
Both Drs. Lau and Rickard made presentations to the NCSAB previously (123`d meeting, May 10,
2007 and 116'" meeting, Sept. 28, 2006 respectively).
Presentation: Toxicity of Perfluorooctanoic Acid in Animal Studies: An Update
Dr. Christopher Lau
USEPA, Developmental Biology Branch (MD-67),
Reproductive Toxicology Division, NHEERL
Research Triangle Park, NC
Dr. Lau, in his presentation:
• Reviewed the physical, chemical and pharmacokinetics of Perfluoroalkyl Acids (PFAAs)
• Reviewed the toxicity of PFOA
• Discussed risk characterization of PFOA in case studies
• Discussed historical and current research
• Summarized toxicity and pharmacokinetic data, and mode of action with respect to PFOA
• Summarized regulatory actions
Introduction
Dr. Lau began his presentation with a review of information on PFAAs (perfluoroalkyl acids). He
briefly described what they are, where they are used commercially, how they are produced, PFAAs
commonly found in the environment, why PFAAs are of toxicological and environmental concern, and
their fate and transport. He mentioned that perfluoroalkyl phosphonic acids had been found in the
environment recently, however they are rare and do not seem to have the same toxicological and
environmental concerns as the sulfonic and carboxylic acid forms. He declared that the list of PFAAs
found in the environment is growing and now includes C4, C6, C9, and C10 chemicals in addition to
the C8s. He discussed two CDC studies (NHANES I' and 112) that report blood levels of PFOS,
PFOA, PFHxS and PFNA in non -occupationally exposed individuals compared to occupationally
exposed humans and wildlife on a national scale. He noted that blood levels seem to be declining
between the two study periods and speculated that the difference was because one major
manufacturer in MN had phased -out production of these chemicals. He stated that insofar as children
and infants are concerned: (1) exposure to PFAAs is elevated compared to adults; (2) PFOA and
PFOS have been found in maternal and umbilical cord blood, so there is transplacental exposure; (3)
both PFOA and PFOS have been found in breast milk, so there is lactational transfer; and (4)
ingestion and absorption are modes of exposure that must be evaluated. PFOA and PFOS are both:
well absorbed orally; found mainly in the serum, liver and kidney. The highest levels are found in the
liver. The levels found in serum and the kidney are about the same; highly bound to proteins,
therefore not metabolized well; and poorly eliminated through fecal and urinary excretion. In the rat:
PFOS serum t1/2 = 35 days, PFOA serum tv2 = 6.5 days (males), 3 hrs (females). He mentioned that
there seems to be a correlation between chain length and length of time for elimination. This is most
' hftp://www.cdc.gov/nchs/abouVma4or/nhanes/nhanes99 OO.htm
2 http://www cdc gov/nchs/about/maior/nhanes/nhanes2003-2004/nhanes03 04.htm
DEQ-CFW 00000021

clearly observed in humans where the elimination ranges from days (for C4) to years (for C8). Dr.
Lau could not explain the results observed for C6 (PFHxS), noting that it had a longer elimination
period compared to chain length.
Dr. Lau said that there are no adverse health effects from exposures to PFOA or PFOS that have
been substantiated to date. He proposed using a pharmacokinetic model for extrapolating between
animal and humans to determine the level at which adverse heath effects occur in humans.
Carcinogenicity
PFOA is linked to a "tumor triad" in animal studies: liver adenomas, pancreatic acinar cell tumors, and
testicular Leydig cell adenomas mediated by PPARa. Ovarian tubular hyperplasia has also been
observed in animals. It is not common, but it is observed, and no explanation has been proffered as
yet for this. PFOS is linked to liver adenomas and thyroid adenomas and carcinomas in animals.
Hepatotoxicity
Most common is liver hypertrophy associated with peroxisome proliferation and vacuole formation.
Vacuoles are the organelles responsible for lipid and lipoprotein metabolism that alter lipid and
lipoprotein transport. PFOA down -regulates cholesterol and bile acid synthesis. At least one MOA is
by activation of nuclear receptors, one of which is PPARa. PFOA interferes with GJIC (gap junction
mediated intercellular communications) and also interferes with the binding of fatty acids to protein.
Dr. Lau hypothesized that because the body views these chemicals as fatty acids that the
homeostatic mechanism is unable to function optimally.
Recent Research
Rosen et. aL (2008a) (see Key References) reported a microarray analysis of PFOA in fetal mouse
liver tissue. Gene expression profiles were analyzed. The signaling pathway highlighted was PPARa.
The genes observed in this study are typical PPARa associated genes. In a similar study by Rosen
et. aL (2009), PFOS results were compared with PFOA results. The results were similar.
Wolf et. a/. (2008a) also used liver tissue to study the MOA in PFOA. A PPARa activator (WY-50)
was administered to a wild -type group of mice. PFOA was then administered at doses of 1 mg/kg, 3
mg/kg and 10 mg/kg to the wild type group and to a group of mice with the PPARa pathway knocked
out (PPARa-KO). The hypothesis was that if the PPARa pathway is knocked out then there should
be no response from the PFOA dose. Liver weight was ascertained for each mouse type at each
dosing level. The results indicate that there was little difference between liver weights of the wild -type
and knock -out mice when dosed with PFOA at levels below 10 mg/kg. Liver weights for mice dosed
at 10 mg/kg indicate that another MOA for PFOA maybe activated as the knock -out mice had higher
liver weights than the wild -type mice.
Rosen et. aL (2008b) analyzed the results from Wolf et. a/. (2008a) and reported that there are about
900 genes activated in the wild type mouse when the PPARa antagonist (WY14, 643) is administered
compared to 10 in the PPARa-KO. At the 3 mg/kg PFOA dose there are about the same number of
genes activated in the wild type (879 vs. 900) but only 176 in the PPARa-KO.
Corton (co-author in Rosen et, aL (2008b)) did a more in-depth investigation comparing wild type and
PPARa-KO mouse gene expressions using a heat map. 975 genes are PPARa-related and 156 are
not suggesting these genes are mediated by another mechanism. Corton also indicated which
mechanisms were up -regulated, like fatty acid oxidation and lipid metabolism, and those that were
down -regulated, like inflammation and coagulation. Corton also showed genes that were PPARa
independent are mostly responsible for lipid metabolism in the body. Lipid metabolism is being
mediated by PPARa, PPARP/5, and PPARy. Corton thinks that PPARP6y are backups to PPARa, so
when PPARa is knocked out they take over and their function is enhanced. Barbara Abbott is
conducting research on this now. With xenobiotic metabolism there are four nuclear receptor
candidates that can mediate for this function. Corton aligned the responses for CAR, PXR, AhR and
Nrf2 to see if any gene signatures matched. There were no results for Nrf2 and AhR has not been
investigated further. There is relatively good concordance with CAR; PXR has not been investigated
2
DEQ-CFW 00000022

yet. It is believed that CAR is a likely candidate for PFOA action and Corton is attempting to design a
knock -out model for CAR to see if the residual response can be knocked -out as well.
Barbara Abbott (co-author of Wolf et a/ 2008b), EPA NHEERL, established a neutral cell model for
PPARa. Cells were transfected with either a mouse or human PPARa receptor-luciferase reporter
plasmid. The cells were exposed to PFAAs, including PFOA (0.5-100 pM). After 24 h of exposure,
luciferase activity from the plasmid was measured. All PFAAs activated both mouse and human
PPARa in a concentration -dependent fashion. Measuring the luciferase signal associates with the
potency of the chemical. Results displayed in Figure 1 indicate a 20% maximum response for PFAA
activity on PPARa for the mouse and human. In general, increased carbon chain length increases
PPARa activity for both species, albeit the mouse is slightly more active than that for humans; the
carboxylate seems to be more potent than the sulfonate; and the mouse PPARa appears to be more
sensitive to PFAAs than the human PPARa in this model. The results indicate that C9 is the most
potent, requiring only 5 pM to activate a 20% response. C10 does not behave as predicted and there
is no explanation for this at this time.
1 18
Comparison of PFAA activities on PPARa
C20... (AM)
Compound
Mouse
_. Human
PFNA (C9)
5...
11
PFOA (C8)
6
16
PFDA (C10)
20
no activity
PFHxA (C6)
38
47
PFBA (C4)
51
75
PFHxS (C6)
76
81
PFOS (C8)
94
262
PFBS (C4)
317
206
Wolf atal., 2008 t1.1
FIGURE 1
Immunotoxicity
Bob Lucas (EPA NHEERL) is conducting research in this area. Reduction of thymus and spleen
weight has been observed to be associated with inhibition of cell synthesis. T cell and B cell activation
is attenuated, so PFOA is immunosuppressive. This activation is similar to the results seen in the
liver expression as well. Suppression of NK cell function and decrease of IgM production have been
observed after in utero exposure to PFOS. Suppression of inflammatory response in the animal liver
has also been observed; some think that this could be mediated by stress. The immunosuppressive
observations are consistent with the EPA gene array data. In a very new study (unpublished) using
mice that have had the adrenal glands removed immune suppression was observed, indicating part of
the response can be mediated by PPARa and another part is independent.
Neurotoxicity
Johansson et al (2008) tested both the behavior and biochemistry of the mouse model. Acute doses
of PFOA were administered to 10-day old neonates. Some behavioral changes were observed.
Attempts are being made trying to correlate these changes with changes in CNS proteins specific to
the hippocampus, related to neuronal growth and synaptogenesis. Responses observed appear to
be similar to other organochemicals, so there may be some overarching mechanisms occurring.
Slotkin et al (2008) conducted an in vitro study with PC12 cells. Changes in cell replication,
differentiation and oxidative stress were observed. It was observed that PFOS was the most potent
DEQ-CFW 00000023

chemical and PFOA was the least potent. This also indicates that the underlying mechanisms can be
different. In another study, Butenhoff (co-authored Chang et a/ 2009), did an extended developmental
neurotoxicity test using PFOS and found no significant adverse effects of PFOS in the rat.
Hormone imbalance
Exposure to PFOA (and other PFAAs) reduces T4 and T3 in rats, but there is no observed feedback
elevation of TSH (Lau et a/ 2007). These chemicals may not inhibit synthesis or metabolism of the
hormone, but may displace the hormone from its binding protein. There is a decrease in serum
testosterone and an increase in serum estradiol in male rats. These effects are associated with
induction of aromatase in the liver. It has been suggested that induction of hepatic aromatase may
lead to Leydig cell tumors in male rats. PFOA may have an estrogenic mechanism in fish. This has
been observed in rainbow trout.
Developmental toxicity (PFOA)
Developmental toxicity is unremarkable in the rat. This is likely related to the short half-life of PFOA in
the female rat (2 hours) as compared to that in the male (7 days). Using a mouse model to compare
serum levels of PFOA, Lau et a/ (2008) conducted further pharmacokinetic studies that show that the
half-life of PFOA in mice by sex and by dose is quite similar (FIGURE 2).
23
Pharmacokinetics of PFOA in mouse
Mouse (1 mg/kg) Mouse (Male)
0 10 20 30 40 50
Time (days)
Male Female
Half-life (days) 1 21.7 15.6
100
i 10
0 10 20 30 40 50
Time (days)
1 mg/kg 10 mg/kg
Half-life (days) 1 21.7 20.4
Lou et al., 2009
FIGURE 2
In the fetal mouse, PFOA can accumulate throughout pregnancy. Neonatal mortality is dose
dependent (FIGURE 3). This is similar to what was observed with PFOS (FIGURE 4). In the survival
group, mice have growth retardation and development deficit at a level that is dose -dependent
(FIGURE 5).
4
DEQ-CFW 00000024

25
Postnatal survival of Mice exposed to PFOA
100
80
60
C
N
a
40
20
0
0
5 10 I6 20 25
Postnatal Age (days)
FIGURE 3
-0- caxra
Z 1.4kg PFOA
-10- 3 n;gkg PFOA
41�5 n5x9 PFOA
10 mg9cg FFOA
-W 20 nog FFOA
'-' r / 26
Postnatal Survival of Mouse Exposed to PFOS
100
Control
60 1 mgrkg
ai T ♦T —�— 5 mgrkg
a 40 ,i 1 r — � -I -- --*_ 15 mgrkg
-a- 20 mg/kg
20
0 a a
0 5 10 15 20 25
Postnatal Age (days)
FIGURE 4
5
DEQ-CFW 00000025

Neonatal growth and development
Eye Opening
Body Weight
u
0 5 10 15 20 25
Postnatal Age (Oats)
FIGURE 5
n
Mani
Relative e + s s
veoa lma*al
Relative liver weight
u
�cs`a
� �rtaat.�ea
m KW i Epp
p`s
S �
t s
aa. IW�u
27
Sue Fenton, EPA NHEERL, has conducted research (co-authored White et al2007) on the effects of
PFOA exposure on mammary gland development. These results are shown in FIGURE 6. All treated
groups show stunted mammary gland development. Currently offspring are being tracked after in
vitro exposure to see if there are functional deficits in lactation. The latest results will be available
later in 2009.
r—' r 28
Mammary gland development
■ Mannnary glands stained with
carmine, scored on a 1-4
development -based scale.
■ All treated groups showed
stunted development.
PND 10
PND 20
Control
3.1 t 0.21
3.3 10.20
1.7t0.20t
1.4t0.101
GD 9.18
1.5 1 0.17t
1.5 1 0.12t
GD 2-18
1.6 1 0.16t
1.6 1 0.23t
White at al., 2007; 2009
FIGURE 6
J11 :1J
DEQ-CFW 00000026

The postnatal findings of PFOA exposure indicate that in contrast to the rat, neonatal survival of mice
is severely compromised. This is likely due to the females accumulating PFOA during pregnancy.
The PFOA profiles of neonatal mortality are similar to those observed with PFOS. There are
significant growth deficits and developmental delays among surviving pups. The liver weights were
significantly increased in neonates.
Barbara Abbott et al (2007) has investigated an MOA for developmental toxicity of PFOA/S using a
transgenic model to analyze whether knocking out PPARa would affect developmental toxicity. Wild -
type mice show a significant mortality to PFOA exposure while PPARa-KO mice exhibit no mortality
(FIGURE 7). Research indicates that the developmental process is also dependent upon PPARa. In
contrast to that with PFOS, the results indicate that there is mortality in the PPARa knock -out mouse
(FIGURE 8). There appears to be a different MOA because PFOA is mediated by PPARa where
PFOS may interfere with lung function.
/ 31
PPARa Involvement in PFOA Neonatal Mortality
Wildtype Mice
7
,1
7 ' e
1
Grd� Pb_-�♦-p
—o— WT control
2 WT0.1 mg
W T 0.3 mg
_ W T 0.0 mg
-d-- WT 1.0 mg
0 5 i0 95 20 A
PND
PPARa-null Mice
—� PPAR KO Contml
PPAR KO 0.1 mg
• PPAR KO 0.3 mg
-- PPAR KO 1.0 mg
—PPAR KO 3.0 mg
15 20 25
PND
Abbott et al-, 2008
FIGURE 7
7
DEQ-CFW 00000027

Wild Type
PFOS-induced Neonatal Mortality 32
is Independent of PPARa. Signal
�— control
-: 4.5 mg/kglday
- • 6.5 mg/kg/day
8.5mg/kg/day
• 10.5 mg/kg/day
2 3 4 5 6 7 8 9 10 11 12 13 14
100
PPAR KO v • .__: ,...,...,....,. , .....,...�, .., c
> `�—�'�—a_-�--a--s--a--F--a
50
Abbott at al., 2009
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
PND
FIGURE 8
Dr. Lau concluded his presentation by summarizing recent regulatory actions (FIGURE 9) for either
drinking or surface water regarding PFOA and PFOS.
34
Recent actions regarding PFOA
• Human health risk assessment of PFOA prepared by EPA OPPT in
2005 and reviewed by SAB in 2006; risk assessments conducted by
European countries
• In 2006 EPA Region 3 and Region 5 issued a consent order to
industry to lower level of PFOA in drinking water from 150 ppb to 0.5
ppb
• In 2007. MN Department of Health recommended to lower the health -
based value for PFOA from 7 ppb to 0.5 ppb; and for PFOS from 1
ppb to 0.3 ppb
• In 2007, NJ Department of Environmental Protection recommended a
preliminary health -based guidance of 0.04 ppb for PFOA in drinking
water
• In 2008, European Food Safety Authority issued a Tolerable Daily
Intake of 150 ng/kg b.w. for PFOS and of 1.5 pg/kg b.w. for PFOA
• In 2009, EPA OW issued provisional health advisories of 0.4 µg/L
(ppb) for PFOA and 0.2 Ltg/L (ppb) for PFOS in drinking water
FIGURE 9
DEQ-CFW 00000028

Key References:
Rosen MB, Abbott BD, Wolf DC, Corton C, Wood CR, Schmid JE, Das KP, Zehr RD, Blair ET, Lau C.
2008a. Gene Profiling in the Livers of Wild -type and PPARa-Null Mice Exposed to
Perfluorooctanoic Acid. Toxicologic Pathology 36[4]:592-607.
Rosen MB, Schmid JE, Das KP, Wood CR, Zehr RD, Lau C. 2009. Gene expression profiling in the
liver and lung of perfluorooctane sulfonate-exposed mouse fetuses: Comparison to changes
induced by exposure to perfluorooctanoic acid. Reproductive Toxicology Received 16 October
2008; revised 22 January 2009; accepted 23 January 2009. Available online 6 February 2009.
hftp://www.sciencedirect.com/science? ob=ArticleURL& udi=B6TC0-4VJBTP4-
1 & user-10& rdoc=1 & fmt=& orig=search& sort=d&view=c& acct=C000050221 & version=1 &
urlVersion=0& userid=10&md5=3a2ae32c897b3d78ba79d81419042d50.
Wolf DC, Moore T, Abbott BD, Rosen MB, Das KP, Zehr RD, Lindstrom AB, Strynar MJ, Lau C.
2008a. Comparative Hepatic Effects of Perfluorooctanoic Acid and WY 14,643 in PPARa
Knockout and Wild -type Mice. Toxicologic Pathology 36[4]:632-639.
Rosen MB, Lee JS, Ren H, Vallanat B, Liu J, Waalkes MP, Abbott BD, Lau C, Corton JC. 2008b.
Toxicogenomic dissection of the perfluorooctanoic acid transcript profile in mouse liver: evidence
for the involvement of nuclear receptors PPAR alpha and CAR. Toxicological Sciences
103[1 ]:46-56.
Wolf CJ, Takacs ML, Schmid JE, Lau C, Abbott BD. 2008b. Activation of Mouse and Human
Peroxisome Proliferator-Activated Receptor Alpha by Perfluoroalkyl Acids of Different Functional
Groups and Chain Lengths. Toxicological Sciences 106[1]:162-171.
Johansson N, Fredriksson A, Eriksson P. 2008. Neonatal exposure to perfluorooctane sulfonate
(PFOS) and perfluorooctanoic acid (PFOA) causes neurobehavioural defects in adult mice.
Neurotoxicology 29:160-169.
Slotkin, T.A., MacKillop, E.A., Melnick, R.L., et. a[. 2008. Developmental Neurotoxicity of
Perfluorinated Chemicals Modeled in Vitro. Environmental Health Perspectives 116[6]:716-
722.
Chang S, Ehresman DJ, Bjork JA, Wallace KB, Parker GA, Donald G. Stump DG, Butenhoff JL. 2009.
Gestational and lactational exposure to potassium perfluorooctanesulfonate (K+PFOS) in rats:
Toxicokinetics, thyroid hormone status, and related gene expression. Reproductive Toxicology
Received 3 September 2008; revised 29 December 2008; accepted 7 January 2009. Available
online 21 January 2009. hftp://www.sciencedirect.com/science? ob=ArticleURL& udi=B6TC0-
4VDY7YN-
_ rnnnA O
Lau C, Anitole K, Hodes C, Lai D, Pfahles-Hutchens A, Seed J. 2007. Perfluoroalkyl Acids: A Review
of Monitoring and Toxicological Findings. Toxicological Sciences 99[2]:366-394.
Lau C, Lau, Das K, Thibodeaux J, Grey B, Rogers J. 2008. Comparative Description of PFAA
Developmental Toxicity: An update. PFOA Days II: Recent Advances in Perfluoroalkyl Acid
(PFAA) Research. June 3-5, 2008. US EPA. Research Triangle Park, NC.
White SS, Calafat AM, Kuklenyik Z, Villanueva L, Zehr RD, Helfant L, Strynar MJ, Lindstrom AB,
Thibodeaux JR, Wood C, Fenton SE. 2007. Gestational PFOA Exposure of Mice is Associated
with Altered Mammary Gland Development in Dams and Female Offspring. Toxicological
Sciences 96[1]:133-144.
DEQ-CFW 00000029

Abbott BD, Wolf CJ, Schmid JE, Das KP, Zehr RD, Helfant L, Nakayama S, Lindstrom AB, Strynar
MJ, Lau C. 2007. Perfluorooctanoic Acid —Induced Developmental Toxicity in the Mouse is
Dependent on Expression of Peroxisome Proliferator—Activated Receptor -alpha. Toxicological
Sciences 98[2]:571-581.
Presentation: PFOA Update
Robert W. Rickard, Ph.D., D.A.B.T., Science Director
E.I. DuPont de Nemours and Company
DuPont Haskell Laboratory for Health and Environmental Sciences
Newark, Delaware
Dr. Rickard, in his presentation:
• Discussed the EPA Voluntary Stewardship Program
• Reviewed epidemiological studies
• Reviewed water guidance values
• Discussed future research
Dr. Rickard began his presentation saying that there had been about 137 papers published since his
last presentation to the NCSAB in September 2006. At least one paper a day is being published
regarding perfluoroalkyl acids.
Dr. Rickard added clarification to Dr. Lau's Slide #6 stating that the PFOA measurements for the
Cape Fear River were taken upstream of the manufacturing plant and that those measurements were
for all PFAAs and not just PFOA. PFOA has been around as long as PFOS — over 60 years. Dr. Lau
asked about the production levels. Dr. Rickard responded that the NC plant was opened in 2002 and
was designed to have low emissions of PFOA (less than 50 Ibs/yr in air) so production level was
lower. He noted that it is not appropriate to compare this site to previous manufacturing sites.
EPA Volunteer Stewardship Program (http://www.er)a.pov/oppt/r)foa_o
This program was developed for manufacturers of PFOA. Eight companies have agreed to emissions
reductions, product content reductions, and the possibility of phasing out PFOA. This program has
global stakeholders. There is an annual progress report due in October. In February 2008, EPA
reported that eight major companies have reported significant reductions in PFOA emissions. The
overall goal is 95 percent PFOA emissions reduction per company by 2010. Three US companies,
including DuPont, reported at least 98 percent reductions in emissions in PFOA. There has been
progress in the development of substitutes and alternatives to PFOA. In November 2007 the CDC
reported a 25 percent reduction in human serum PFOA and a 32 percent reduction in human serum
PFOS in the interim between the NHANES I and II surveys (FIGURE 9). These decreases may be
attributed to decreases in PFOA emissions. 3M conducted a similar survey indicating a 50 percent
reduction in PFOA and PFOS. Additionally a survey conducted in New York shows a decline in blood
serum levels. Reduced emissions and reductions observed in blood serum levels appear to be
correlated.
10
DEQ-CFW 00000030

PFCs in the U.S. Population: NHANES 2003-04
Comparisons to 1999-2000 (Calatat et al., 2007)
Mean Serum Concentration (Dpbl
1999-2000 2003-2004 % Change
PFOA 5.2 3.9 .25
PFOS 30.4 20.7 -32
PFHxS 2.1 1.9 -10
PFNA 0.5 1.0 +100
Conclusion:
Reductions related to discontinuation in 2002 of industrial
production by electrochemical fluorination of PFOS and related
compounds
(POD?
February 25, 2009
FIGURE 9
Epidemiological studies
Dr. Rickard summarized EPA comments regarding PFOA. EPA has found no definitive health link to
PFOA.
3M study
3M has investigated PFOA toxicology for over 50 years. They have three plants sites that have been
involved with this chemical, the Cottage Grove plant in MN, a plant in Decatur, AL and one in
Antwerp, Belgium. Occupational exposures ranged from 1-10 ppm at the Cottage Grove plant with
highest exposures in excess of 100 ppm; the other two sites had reported exposures at 1-2 ppm.
Emissions from the Decatur site are confounded by PFOS emissions. A number of variables have
been investigated: mortality incidence, liver function, lipid profiles, reproductive hormones, tracked
employee physician visits (how many and why), and birth weight. It was concluded that there were
no observable adverse health effects from exposures to PFOA. A 2007 paper (Olsen & Zobel 2007,
see Additional References) that examined correlated lipid, liver, and thyroid parameters with PFOA
concentrations at all three plants and it was concluded there was no evidence that employee serum
PFOA concentrations were associated with total cholesterol or LDL. There was slight negative
correlation with HDL that was explained by demographic differences across the three plants. This is
one of the few studies that investigated several plant sites.
Miteni Health Monitoring
Miteni is an Italian manufacturing facility with 30-50 employees. PFOA has been manufactured at this
facility for over 20 years. In 2005, the reported average PFOA blood level in workers was 16 ppm.
Currently PFOA blood levels are around 8 ppm. PFOS is also manufactured at this facility that has
confounded the blood measurements. The same physician has been tracking this group of
employees for over 20 years. It was concluded that there were no health effects associated with
PFOA exposure and although there was a statistical increase in LDL a causal relationship could not
be determined.
DuPont Employee Health Study — Phase I
Sakr et al. 2007. J. Occup. Environ. Med. 49, 1066-1096.
Over 1000 employees at the Washington Works (West Virginia) facility volunteered for this study that
examined 62 clinical endpoints in blood serum and urinary analysis and compared them with PFOA
11
DEQ-CFW 00000031

exposures. The PFOA blood serum concentrations at the facility ranged from 0.005-9.55 ppm
(average 0.5 ppm). No correlation was determined between PFOA exposure and most parameters
measured. There were slight increases observed in cholesterol fractions (total, LDL, VLDL) and one
liver enzyme (GGT), but no increase in HDL. The data are shown in FIGURES 10 and 11.
do to opene-UF We I
11 CHOL_TOTAL by PFOA Quantile
Q U i ol,� Sex=M, Heartneds=N, WN 2004
Mean 19&823 W.016 205628 190.867 196.306 204.1336 200.871 188.873 208.097 2AW7
Std Dev 32345 28.787 38.434 3t099 37.476 32836 30.472 1 3t022 38.740 XW
300
300 ❑
❑ ® ❑ ❑
2b0
JI
x
O 200 — — — — —
b0
❑ ❑
V0
RWR Deciles
9128, 06
FIGURE 10
12
DEQ-CFW 00000032

Lk_Chd_Total Vs LOG-PFOA : WW 2004, Terms= LogPFCA AL06
Where Sex=M, Heartmeds=B
S.8
•
SLOPE P—VALUE=O •
s.7
R—SQR=0A2 • • • • • •
• i %
5.6
Total Cholesterol • •
,240mgldl • .� • • • •♦ • • •
• • •
5.5
.. i'•
i• • • f • • • • •
........................................ R...�. .......�. ........... j..... •�...�.~......................
5.4
• , • i��Y s, 40 ! » •i • •
ft
S.3
• • i r •
5.2
•� j� • �,•�•• :+ • • i
• w •
•
S♦•� •� i
• • • i 4b•
4.9
of +• •4P '• we: i
4 8
• • 1011 docile of serum PFOA
• a 1000 ppb
4.7
2 3 4 5 6 7 8 9
AMR`VR LBG_PFBA
9 28106
FIGURE 11
Blue -colored data points indicate those subjects taking cholesterol -lowering drugs. There was no
correlation between total cholesterol measurements and PFOA exposures with respect to these
subjects. Carine Sakr, visiting scientist from Yale University, did a longitudinal study (Sakr et al
2007b) to see if cholesterol and PFOA concentrations could be modeled. The findings indicated that
1-ppm PFOA in blood correlates with a 1-5 unit increase in cholesterol. The model results also
indicated that this is not a dose -response. It was expected (given the Miteni data) that a significant
increase in cholesterol would be observed with elevated exposure, but the data did not show this.
This may well be a reverse causality because this result has been seen in different populations.
PFOA in human blood often acts like a lipid lowering drug - decreases in cholesterol would be
expected to be observed. If this is causal, it is biologically opposite of what would be expected.
DuPont Employee Health Study: Phase II
Leonard et al. 2007. J. Occup. Environ. Med. 8, 872-879.
This was a retrospective cohort mortality study at the Washington Works facility. It examined all
causes of death from 1948 to 2002 and compared the data to three different reference populations:
other DuPont workers, West Virginia (WV) residents, and the general US population. Significant
reductions in mortality were observed relative to US and WV populations. Pancreatic cancer
incidence did not appear significant relative to the other DuPont workers, the US population, or WV.
This is note -worthy given the tumors observed in animal models. Prostate cancer was reported in an
early 3M study, however, in an updated study it was discounted but thought to be important to
investigate. The incidence of prostate cancer is low, especially when compared to the US and WV
populations. Kidney cancer was elevated compared to the other DuPont workers population, but
compared to the US and WV populations, it is only slightly elevated. It is not thought that this is
correlated to PFOA exposure. Diabetes is elevated compared to the other DuPont workers but lower
than the US and WV populations; this was thought to be due to regional effects - especially given the
incidence of diabetes in WV. This is being followed up. The results indicate that there is no
convincing evidence of increased mortality associated with exposure to PFOA (FIGURE 12).
13
DEQ-CFW 00000033

FIGURE 12
A non -significant increase in kidney cancer mortality and a statistically significant increase in diabetes
was observed across the facility. There was no correlation with that in the PFOA work area; however,
there were too few incidents of either to make a definitive conclusion. Because of the increase in
cholesterol, ischemic heart disease (IHD) became the study focus. There was only one significant
increase in IHD at one lag time in one model. All the other models were negative. Mortality ratios
were also modeled with cumulative exposures and various lag times. The only significant increase
was in Model A of the 10-year lag; all of the others were negative. This is being followed up, but
there does not appear to be a strong association there. Results are shown in FIGURE 13. Research
is continuing in this area. An independent group is following up to see if there is any correlation with
IHD.
14
DEQ-CFW 00000034

Mortality rate ratios for (HD by cumulative exposure category,
increasing 5-year lags of exposure
Model A) exposure categories based on case distribution
Model B) exposure categories based on cohort distribution.
NO LAG
5-YEAR LAG
10-YEAR LAG
15-YEAR LAG
20-YEAR LAG
A)
\ = 4.460
\ = 4,440
\ = 3,989
N = 3,986
\ = 3,440
NIRR
MRR
MRR
MRR
MRR
(95 % CI)
(95 % C-I)
(95 % CI)
(95 % CI
Refereuce*
1
1
I
1
1
(1to exposure)
Category 1
1.046
0.864
0.996
0.935
0.647
(toss exposure)
(0.711-1.539)
(0.579-1.291)
(0.647-1.531)
(0.598-1.461)
(0.407-1.029)
Category 2
1.156
1.204
1.377
1.102
0.692
(ineditunexposure)
(0.745-1.793)
(0.757-1.914)
(0.829-2.287)
(0.644-1.887)
(0.388-1.233)
Category 3
1.110
1.077
1.610
1.089
0.764
(high exposure)
(0,698-1.767)
(0.657-1.764)
(0.942-2.753)
(0.597-1.984)
(0.393-1.489)
B)
Refereucet
1
1
1
1
1
(uo exposure)
Category 1
1.018
0.980
1.062
0.914
0.967
(101), exposure)
(0.515-2.010)
(0.607.1.582)
(0.662-1.680)
(0.557-1.502)
(0.568-1.647)
Category 2
1.132
1.014
0.825
0.860
0.931
(medium exposure)
(0.573-2 235)
(0.603-1.705)
(0.483-1.409)
(0.494-1.497)
(0.505-1.717)
Category 3
1.027
1.019
1.256
1.069
0.882
(high exposure)
(0.496-2.127)
(0.571-1.517) 1
(0.721-2.191)
(0.598-1.910)
(0.461-1.685
*rum
February 25, 2009
FIGURE 13
Community C8 Study Little Hocking Water Service Area
Emmett et a/. 2006. J. Occup. Environ. Med. 48, 759-770.
Emmett et al. 2006. J. Occup. Environ. Med. 48, 771-779.
This study, independently funded by NIH, was conducted about 2 years ago. C8 exposure and
potential health effects were examined in the Little Hocking area. This community was downwind from
a DuPont site and was the community with the highest exposure to PFOA in the area. The study
population was approximately 300 volunteers. The median C8 blood level was determined to be 340
ppb, highest in children and older adults. It was concluded that the major source of exposure was the
drinking water and not the air. DuPont conducted extensive air modeling, drinking water, and well
water monitoring. Those data were provided to Dr. Emmett. It was observed that a 1 ppb PFOA
water concentration corresponded to a 100 ppb PFOA concentration in human blood. No
relationships were observed between C8 blood levels and cholesterol, TSH, kidney function, liver
function or red cell, white cell and platelet counts.
Cord Serum Concentrations of PFOS and PFOA in Relation to Weight and Size at Birth
Apelberg et al. 2007. Environ. Health Perspect. 115, 1670-1676.
A study conducted at Johns Hopkins University examined cord serum concentrations of both PFOS
and PFOA relative to birth weight and head circumference. The CDC analyzed samples from 293
births for a suite (14-15) of perfluorinated chemicals. The median PFOS concentration was 5.0 ppb
(range <LOD - 34.8 ppb). The median PFOA concentration was 1.6 ppb (range 0.3 - 7.1 ppb). These
results indicate that cord blood levels tend to be lower (typically 1/3 to 1/) than the blood levels in the
mother. PFOS and PFOA concentrations in blood were highly correlated. Small negative
associations between PFOS and PFOA with birth weight, Ponderal index, and head circumference
(vaginal deliveries only) were observed. No associations were observed between either
PFOS or PFOA concentrations and newborn length or gestational age. This was the first study
investigating cord serum levels in the general population.
15
DEQ-CFW 00000035

Perfluorinated Chemicals and Fetal Growth: A Study within the Danish National Birth Cohort
Fei et al., 2007. Environ. Health Perspect. 115, 1677-1682.
This study used randomly selected women (n = 1400) and their infants from the Danish National Birth
Cohort to investigate the association between plasma levels of PFOS and PFOA in pregnant women
and infant birth weight and length of gestation. Blood was drawn from the women during the first and
second trimesters of pregnancy. The women also took part in four separate computer -assisted
telephone interviews. Cord blood samples were taken from a subset of infants (n = 146). The results
indicate that (1) average PFOS and PFOA levels in maternal plasma were 35.3 and 5.6 ng/mL,
respectively, (2) neither maternal PFOS nor PFOA levels were associated with the risk for preterm
birth or low birth weight, (3) no adverse effects on gestational age were observed related to maternal
PFOS or PFOA levels, and (4) only PFOA levels were inversely associated with birth weight.
Perfluorinated Chemicals and Developmental Milestones: A Study Within the Danish National
Birth Cohort
Fei et al., 2008. Environ. Health Perspect. 116,1391-1395.
Using the same Danish cohort, this study examined developmental milestones, including age
associated with sitting up, and other associated activities for infants between 6-18 months. It was
concluded that there were no convincing associations between developmental milestones in early
childhood and levels of PFOA or PFOS (mean = 5.6, 35.3 respectively) as measured in maternal
plasma early in pregnancy.
Perfluorinated Chemicals and Time to Pregnancy: A Study within the Danish National Birth
Cohort
Fei et al., 2009. Human Reproduction 1, 1-6.
This is the most recent study using this same Danish cohort. The relationship between the time to
pregnancy (TTP) and level of PFOA and PFOS was investigated. Women were surveyed on length
of time between trying to get pregnant and actual pregnancy. Infertility was defined as TPP over 12
months. All the women in the study had babies, so the definition of infertility is uncertain. Longer
TPP was found to be associated with higher maternal blood levels of PFOA and PFOS. The adjusted
fecundity odds ratio (FOR) was virtually identical for women in the three highest PFOS exposure
groups compared with that of the lowest group. A linear -like trend was observed for PFOA
(FOR = 0.70, 0.67, 0.60 for three highest exposure groups compared to the lowest group). A
summary of the results is shown in the table below. The findings suggest that PFOA and PFOS
exposure may reduce fecundity.
16
DEQ-CFW 00000036

Table ll Estimated OR for infertility (TTP > 12 months) and FOR according to plasma concentrations of PFOS or PFOA
(ngiml) In early pregnancy'
Exposure
No. of ptanned pregnancy
.
Inferdlity
.........................:.......................
.
FOR (95% Cif
Per cent
...........................
OR (95% CI}°
.._................... ...............................
........._....
_..........................................
PMS (nglml)
....................................... .......
6.4-26.0
293
10.6
1.00
1.00
26.1-33.3
305
15.4
1,70 (1,01, 2,86)
0.70 (0.56, 0.87)
33.4-43.2
317
19.5
2.34 (1.40, 3.89)
0.67 (0.53, 0,84)
n43.3
317
18.6
1.77 (1.06, 2.95)
0.74 (0.58, 0.93)
P-V" for txendd
0,025
0.002
PFOA (nglml)
cLL4Q -3.91
293
8.9
1.00
L00
3.91-5.20
308
18.2
2.06 (1.22, 3.51)
0.72 (0.57, 0.90)
5.21-6.96
315
15.5
1.60 (0,93. 2,78)
0.73 (0.58. 0,92)
a6.97
316
21 S
2.54 (1.47, 4.39)
0.60 (0.47. 0.76)
P-waste for trendd
0.006
<0.001
'Fire eamam were adiated for matemal age at defaery, ". we-I+V=qY SMI, mate+aral SES, AmW mnuxrPum badore pwwcl, PtaTW aye, and P&"nvl educatim seven
women (lam a total of 1240) were excluded because o(w&ft data on amriatec bLgsdc repemkwx `Cox dsaute�time modet dit AvAjes for Uw d tests d FOAs were 9~
for the fc uryuatile comparaon of PFOA and PF kn usn a Nc and rho test cl, canfi mm k avvd: FOR. fewrAiy adds ratios: LIOQ, b *w finis of quaft7icaloe.
Perfluorinated Chemicals and Fetal Growth: Japan
Washino et a/,, 2008. Environ. Health Perspect. Available online:
hftp://www.ehponline.org/docs/2008/11681/abstract.htmI
428 pregnant women from Hokkaido Japan were studied. Mean maternal serum levels were 5.6 ppb
PFOS (range 1.3 ppb -16.2 ppb) and 1.4 ppb PFOA (range ND - 5.3 ppb). These levels are
relatively low compared to those observed in similar populations around the world. Birth weight and
size were measured. The results indicate that PFOS is negatively associated with birth weight. PFOA
and birth weight were not correlated.
Perfluorinated Chemicals and Birth Weight: Canada
Monroy et al., 2008. Environ. Res. 108, 56-62.
The objective of this study was to determine if PFAA exposure to women adversely affects birth
weight in fetuses using blood serum measurements of PFAAs in mothers and umbilical cords at
delivery. Blood samples were taken from 101 pregnant women in their second trimester and 105
umbilical cords at delivery and analyzed for PFAAs including PFOA and PFOS. A regression
analysis was performed for birth weight, size and gestational age. The conclusion was that there is
no association between serum perfluorinated chemicals and birth weight.
The authors hypothesize that in the future there will be databases large enough to observe subtle
effects. Harvey Clewell, The Hamner Institutes, is doing some modeling to see whether known
physiological responses explain the results seen in studies like these.
Self -Reported Medical Conditions in PFOS Manufacturing Workers
Grice et al., 2007. J. Occup. Environ. Med. 49, 722-729.
This study's objective was to evaluate whether occupational exposures of PFOS were related to
some cancers, other health conditions and pregnancy outcomes. A survey of current and former
employees revealed health conditions and several cancers. The survey included a brief pregnancy
history from female cohorts. Medical records were examined to validate cancers reported. There
were 421 singleton live births reported. It was concluded that based on biomonitoring data from this
plant site, PFOA serum concentrations were similar but slightly lower than PFOS. Exposure level and
classifications are shown in TABLE 1. There was no association between PFOS exposure and
several cancers, common health conditions, and birth weight.
17
DEQ-CFW 00000037

TABLE 1
Population
PFOS serum concentrations b
"Nonex osed"
110 - 290
Low exposure
390 - 890
High exposure
1300 -1970
PFOA Water Guidance Levels
Dr. Rickard briefly reviewed U.S. and EU PFOA levels, as summarized in the following tables.
Euro can Union
Life Time Ex osure LTE b
concern b
immediate action b
Germany
0.3
0.5
5.0
UK
0.3
10
90
United States
US EPA Water
Office
0.4 ppb "Provisional Health Advisory" (multi -region)
NJ
0.04 ppb preliminary guidance value (NJDEP has recommended no change in
consumptionpatterns)
MN
0.5 ppb current health risk limit 0.3 ppb proposed)
WV
0.5 ppb temporary site -specific action level; SDWA Order on Consent
NC
2.0 ppb interim maximum allowable
Future Research — Perfluorinated Chemicals
Dr. Rickard concluded by discussing areas of future research, namely the ongoing epidemiological
studies by DuPont and the community health study being conducted by an independent science
panel.
• Human Biomonitoring — Longitudinal studies by CDC
• Environmental — Monitoring, Sources, Transport, Fate
• Toxicokinetics
o PFOS models exist
o PFOA models hopefully in development
• Mechanistic Toxicology
o 3M / DuPont
o Penn State
o Wildtype, knockout, and humanized mice studies
• Epidemiology
o DuPont - Occupational, birth weights from birth certificates
o Community (Independent Science Panel)
■ Panel Members:
• Tony Fletcher, PhD, London School of Hygiene and Tropical Medicine
• David Savitz, PhD, Mount Sinai School of Medicine
• Kyle Steenland, PhD, Rollins School of Public Health, Emory University
■ Charge: To conduct community health studies on PFOA in West Virginia and
Ohio counties near the Washington Works site to determine if there is any
probable link between C8 (PFOA) and any human disease
■ www.c8sciencepanel.org
■ Results to date:
• Correlation between PFOA and PFOS and cholesterol (PFOA levels in this
community elevated, PFOS same as background)
• Slight association between uric acid and PFOA and PFOS (levels within
normal however)
18
DEQ-CFW 00000038

• Dr. Fletcher presenting immuno endpoint results from this study at SETAC
next month
o Brookmore Project — parallel study to above
■ Blood serum levels
■ Clinical chemistry data (-65 endpoints)
■ 70K population in same community
■ Database available at the University of West Virginia website
Additional References:
Olsen GW, Zobel LR. 2007. Assessment of lipid, hepatic, and thyroid parameters with serum
perfluorooctanoate (PFOA) concentrations in fluorochemical production workers. Int Arch Occup
Environ Health 81[2]:231-46.
Sakr CJ, Leonard RC, Kreckmann KH, Slade MD, Cullen MR. 2007b. Longitudinal study of serum
lipids and liver enzymes in workers with occupational exposure to ammonium perfluorooctanoate.
J Occup Environ Med. 49[8]:872-879.
Discussion
Dr. Kenyon inquired about the approach taken with the toxicokinetic models that seem to drive some
of the differences seen in the water advisory levels. She wanted to know how various groups have
chosen to handle the differences. Dr. Rickard indicated that the West Virginia regional level of 0.5
ppb was reached using pharmacokinetics that employed a ratio of the half-life in the monkey and half-
life in man (a factor of 45) and a total uncertainty factor of 300. The USEPA Office of Water3 derived
the Provisional Health Advisory for PFOA using the following calculation:
Provisional Health Advisory = BMDLIO x BW x RSC
UFa x UFh x Extrapolation Factor x Water Intake
0.46 k 9 x 1000 mg x 10kg x 0.2 Pg
g g = 0.4--
10x3x81x1d
Where:
BMDL10 = 0.46 mg/kg/day (Lau et a/ 20064)
BW = Body Weight (10-kg child)
RSC = Relative Source Contribution (exposures from non -air sources)
UFa = Intraspecies Uncertainty Factor
UFh = Interspecies Uncertainty Factor
Extrapolation Factor = animal clearance level divided by human clearance level
Water Intake = 10-kg child consuming 1 L/day drinking water
The extrapolation factor was derived from converting the half-life retention time for PFOA in female
mice and humans (17 and 1387 days respectively) to a clearance level using the one -compartment
model foundation, assuming steady state.
Minnesota used pharmacokinetic modeling by Harvey Clewell and North Carolina Division of Water
Quality also used pharmacokinetic modeling.
3 USEPA Office of Water. Health and Drinking Water Advisories for Chemical Contaminants. Water Quality
Criteria. Provisional Health Advisory PFOA and PFOS. Accessed March 17, 2009.
http://www.epa.gov/waterscience/criteria/drinkinci/pha-PFOA PFOS.pdf
4 Lau C, Thibodeaux JR, Hanson RG, Narotsky MG, Rogers JM, Lindstrom AB, Strynar MJ. 2006. Effects of
perfluorooctanoic acid exposure during pregnancy in the mouse. Toxicol. Sci. 90[2]:510-518.
19
DEQ-CFW 00000039

DuPont and 3M contracted The Sapphire Group (Bob Tardiff) and The Hamner Institutes (Mel
Anderson) to perform an independent risk assessment and to derive a drinking water equivalent
concentration and a maximum contaminant level goal for PFOA. The assessment uses some of the
latest pharmacokinetic modeling available from the Hamner Institutes for the risk assessment. It's a
very comprehensive report reviewing both the epidemiology and toxicology of PFOA. The report
compares their approach to other levels that have been derived. The publication should be available
in the next week or so.
Minnesota also has a risk assessment posted on their website. This assessment was developed at
the request of the MN legislative body and compares their approach to that taken in NJ and NC.
Almost all risk assessments conducted to date have either used the monkey six-month study, some
of the latest mouse data that Dr. Lau reviewed, or the two-year rat study. These are the three studies
that seem to drive the risk assessments. Some assessments have used either NOAELs or LOAELs
and others have used benchmark doses and there are different results depending upon the study
data used. Most use an internal dose risk assessment. Other differences in the risk assessments
are because some use a 70 kg person with 2-liter intake correction factor and others use 10 kg with 1
liter; one assessment is a 10 kg with 0.85 liter. Minnesota used a 0.053 L/kg/day correction factor.
Then almost everyone has used the default relative source contribution of 20%. Some consistency is
observed in the drinking water standards with the average around 0.3 - 2 ppb. NJ has the lowest
standard (0.04 ppb).
Dr. Starr asked for clarification of the relative source contribution factor. Dr. Rickard responded
indicating that approximately 80 percent of the exposures come from air and other sources while only
20 percent of the exposure comes from drinking water. This is a default source contribution, and was
not observed in the Little Hocking River study where 90+ percent of the exposures came from
drinking water. Recent environmental studies indicate that most of the exposure comes from diet.
Occupational exposures however are due to inhalation and not ingestion. The source contribution
exposure is relative to the population being exposed.
Dr. Lau mentioned that future research regarding the transport mechanisms may result in resolutions
to the species differences being seen. Dr. Rickard hypothesized that perhaps re -absorption was
occurring in transport especially if the body is recognizing the chemical as a fatty acid and
transporting it that way.
Dr. Rickard offered to make the information from the Sapphire report available to the NCSAB as soon
as the authors release the publication.
Dr. Starr indicated that he would like both Drs. Tardiff and Anderson to present the results of the
report. He would especially like a discussion surrounding levels set from 0.3 — 2 ppb given that there
are no adverse health effects of note in observed in humans and that there are large interspecies
half-life differences. Pharmacokinetics must be taken into account when using an animal model. Dr.
Jordan remarked that North Carolina's 0.63 ppb health guideline is based on Harvey Clewell's
monkey model, and 2 ppb is an interim maximum allowable concentration. Dr. Starr asked if a level
of 0.3 or 0.5 ppb would produce exceedances in N.C. rivers or water supplies. Dr. Jordan will ask
DWQ for a range of measurements of PFOA and where the measurements were taken for the next
meeting.
20
DEQ-CFW 00000040

Other Business
The minutes of the 136th meeting were approved. The minutes of the 138th meeting were approved
as amended per Dr. Spoo's comments.
There will be a discussion regarding having a NCSAB person to be the chemical manager/point
person for PFOA. Dr. Jordan will discuss with Dr. Spoo.
The next meeting of the NCSAB will be held at 2:OOPM on WEDNESDAY, March 25'h, 2009 by
teleconference. The call -in number is [phone].
The meeting was adjourned at 4:25PM.
Respectfully submitted,
Reginald C. Jordan, Ph.D., CIH
Liaison, Science Advisory Board
These minutes were accepted as written at the 140th SAB meeting, March 25, 2009.
21
DEQ-CFW 00000041