v>EPA
United States
Environmental Protection
Agency
Office of Water
(4305)
EPA/823-R-93-003
September 1993
Proceedings of the U.S.
Environmental Protection Agency's
National Technical Workshop
11 PCBs in Fish Tissue"
May 10-11, 1993
Washington, DC
Recycled/Recyclable
Printed on paper that contains
at least 60% recycled fiber
-------
-------
USft
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
OFFICE OF
WATER
Dear Requester:
We are pleased to transmit a copy of the document titled Proceedings of the U.S.
Environmental Protection Agency's "National Technical Workshop on PCBs in Fish Tissues,
EPA/823-R-93-003, September 1993." The workshop was held on May 10-11, 1993 in
Washington D.C.
The primary purpose of the workshop was to transfer current information about PCBs
to states and other parties involved in risk assessment and fish consumption advisories. The
workshop was structured to encourage an exchange of information between the users of the
fish tissue data (such as risk assessors) and the generators of the PCB data (such as
laboratory personnel). This exchange is important because the analysis of PCBs in fish
tissues involves a complex set of considerations including PCB toxicity information,
laboratory analytical techniques, exposure data, etc. Using case studies, the workshop also
illustrated how human health assessments may be affected by the assumptions and analytical
complexities associated with PCBs in fish tissues.
Additional copies of this document can be obtained by calling the Office of Water
Resource Center at (202) 260-7786 or writing to the U.S. Environmental Protection Agency,
Water Resource Center (RC4100), 401 M. Street S.W., Washington D.C. 20460. Please
include the document title and publication number noted above.
We appreciate your interest in the workshop's Proceedings and in other EPA activities
relating to fish contamination issues.
Sincerely,
Rick Hoffmann
Workshop Organizer
Risk Assessment and Management Branch
Office of Science and Technology
R*cycltd/R«cycliblt
Printed on p«p*r that contains
at toast 75% recycled fibor
-------
-------
PROCEEDINGS
U.S. ENVIRONMENTAL PROTECTION AGENCY'S
NATIONAL TECHNICAL WORKSHOP
"PCBs IN FISH TISSUE"
May 10-11, 1993
Washington, DC
Office of Water
Office of Science and Technology
Standards and Applied Science Division
U.S. Environmental Protection Agency
Washington, DC 20460
September 1993
-------
This document is based entirely on presentations at a workshop sponsored
by the U.S, Environmental Protection Agency (EPA) as a forum to share
information concerning PCB contamination of fish tissue. The material in
this document has been subject to Agency technical and policy review and
approved as an EPA report. The views expressed by individual authors,
however, are their own and do not necessarily reflect those of the U.S.
Environmental Protection Agency. Mention of trade names or commercial
names in no way constitutes endorsement or recommendation for use.
Printed on recycled paper.
-------
TABLE OF CONTENTS
EXECUTIVE SUMMARY i
PART ONE - INTRODUCTION TO PCBs 1-1
1.1 Welcome and Introduction 1-1
1.2 Introduction to PCBs and Analytical Methods 1-4
1.3 Temporal Trends of PCBs in the Environment 1-12
1.4 PCB Trends in Great Lakes Fish 1-22
1.5 Overview of PCB Toxicology 1-32
1.6 PCB Criteria for Water 1-48
1.7 Summary of Questions and Responses 1-54
PART TWO - PCB TOXICITY AND HEALTH EFFECTS 2-1
2.1 Regulatory Update: Human Carcinogenicity Effects . • 2-1
2.2 Regulatory Update: Non-carcinogenic Effects 2-6
2.3 Update: Toxicity Equivalents for PCBs 2-14
2.4 Effects of Occupational Exposure 2-18
2.5 Animal/Human Health Connection 2-27
2.6 Summary of Questions and Responses < 2-36
PART THREE - ANALYTICAL METHODS 3-1
3.1 PCB Analyses—An Overview 3-1
3.2 Recent PCB Research 3-3
3.3 PCB Analysis 3-20
3.4 Performance-Based Methods 3-29
3.5 EPA's Green Bay Study—Congener Analyses 3-38
3.6 State Lab Experience 3-47
3.7 Summary of Questions and Responses 3-49
PART FOUR - CASE STUDIES: HUMAN HEALTH/RISK ASSESSMENT 4-1
4.1 California 4-1
4.2 Tennessee Valley Authority 4-10
4.3 Delaware 4-18
4.4 Michigan 4-37
4.5 Summary of Questions and Responses 4-41
APPENDDOES A-l
A.I Speaker's Biographies A-l
A.2 Speaker's Addresses A-7
A.3 Workshop Agenda A-9
A.4 List of Attendees A-13
A.5 PCB Workshop Report Summaries from
EPA's Risk Assessment Forum A-23
-------
-------
EXECUTIVE SUMMARY
On May 10-11, 1993, the U.S. Environmental Protection Agency (EPA) sponsored a "National
Technical Workshop on PCBs in Fish Tissues." Polychlorinated biphenyls (PCBs) are a family
of human-made chemicals that are widely distributed throughout the environment. The analysis
of PCBs in fish tissues involves a complex set of considerations regarding PCB toxicity
information, laboratory analytical techniques, exposure data, etc. The national technical
workshop examined how human health assessments may be affected by current PCB analytical
issues for fish tissues.
The primary purpose of the workshop was to transfer current information about PCBs
to states and other parties involved with risk assessment and fish consumption advisories. The
workshop was structured to provide for an exchange of information between the users of the
PCB fish data (such as risk assessors) and the generators of the PCB data (such as laboratory
personnel).
This document summarizes the proceedings of the workshop. The workshop was divided
into four main parts:
Part One—Introduction to PCBs
Part Two—PCB Toxicity and Health Effects
Part Three—Analytical Methods
Part Four—Case Studies: Human Health/ Risk Assessment
Within each topic area, there were a series of individual presentations followed by
questions from the audience and responses by the speakers. The Proceedings document contains
a summary of each speakers presentation, a selection of key graphics, and a summary of
audience questions and responses.
Part One - Introduction to PCBs
Dr. Southerland and Mr. Hoffmann of EPA welcomed the participants to the workshop and
explained the contents of the workshop.
The morning session of the first day began with an overview of the major risk assessment
and analytical issues for PCBs. Dr. Erickson of the Argonne National Laboratory provided
background information on the chemical nature of PCBs and set the stage for subsequent
discussions about PCB analytical methods. Dr. Craddock of Craddock Associates discussed
temporal trends of PCBs in the environment. His talk, based on a comprehensive literature
review agency studies, described PCB occurrence in various environmental compartments such
as foods, human adipose tissue, shellfish, fish, and human blood sera. Mr. De Vault of EPA
focused on PCB trends in Great Lake Fish. Dr. Bolger of the U.S. Food and Drug
Administration reviewed the toxicology, hazards, and risks of PCBs with a focus on
111
-------
pharmacoMnetics. Ms. Orme-Zavaleta of EPA explained EPA's PCB criteria for water. A
question and response session concluded this session.
Part Two - PCB Toxicity and Health Effects
The afternoon session elaborated on the human health effects of PCBs. Dr. Cogliano of EPA
provided a regulatory update on the evidence of PCB's carcinogenicity. Dr. Cicmanec of EPA
described the agency's recent review of the non-carcinogenic effects of PCBs. Dr. Barnes of
EPA followed with a discussion of the use of toxicity equivalents for PCBs. Dr. Brown of
General Electric reviewed the effects of occupational exposure to PCBs. He focused on work
that GE has conducted on its capacitor workers who were previously exposed. Dr. Colborn of
the W. Alton Jones Foundation concluded the afternoon presentations by talking about adverse
effects upon wildlife and possible implications for human health. Questions and responses
followed.
Part 3 - Analytical Method
The morning of the second day was devoted to PCB analytical methods and associated issues.
Dr. Erickson introduced the topic and moderated the session. Mr. Schwartz of the US Fish and
Wildlife Service summarized some of the challenges that PCBs present to analytical chemists,
particularly for particular congeners. He also described work that FWS has been doing using
principal components analysis. Mr. Sawyer of FDA described the Aroclor-based method that
FDA uses to analyze its samples and compared FDA's method with other analytical approaches.
Dr. Krahn of the National Marine Fisheries Service discussed the "performance-based" approach
that NMFS currently employs for its congener-specific analyses. She also described a new
screening method for rapidly identifying and quantifying coplanar PCBs. Dr. Swackhammer of
the University of Minnesota described quality assurance and quality control aspects of EPA's
Green Bay PCB study. This study collected a PCB calibration set containing 80-90 congeners
and involved S.different laboratories, federal agencies, state agencies, and academic institutions.
Dr. Bush of the NY Department of Health concluded the morning session by discussing the
congener-specific analyses that the Wadsworth Laboratory has conducted.
Part 4 - Case Studies: Human Health/ Risk Assessment
Four case studies were presented to illustrate different risk assessment approaches that have been
used to assess the human health effects of PCBs. The studies were drawn from different
geographic areas. Dr. Pollock of the California EPA described a study of chemical
contaminants in fish conducted for the southern California area and the resulting fish advisories.
Ms. Cox of the Tennessee Valley Authority provided an example of how to assess risks using
a nomograph to display aggregate risks; she applied the risk assessment approach to screening-
level fish tissue data from the TVA area. Mr. Greene from the Delaware-Department of Natural
Resources described a comparative risk assessment study in which his agency evaluated risks of
PCBs in striped bass from the Delaware estuary. His study showed how the calculated risks
might vary under four different hazard/potency scenarios. Mr. Hesse from the Michigan
IV
-------
Department of Health completed the presentation of case studies by explaining the efforts of the
Great Lakes Sport Fish Advisory Task Force to reach agreement on a protocol for Uniform
Sport Fish Consumption Advisories through the Great Lakes' region.
Dr. Southerland concluded with a summary of the workshop.
-------
-------
PART ONE
INTRODUCTION TO PCBs
1.1 WELCOME AND INTRODUCTION
1.1.1 Elizabeth Southerland,
Chief, Risk Assessment and Management Branch, Office of Science and Technology,
EPA, Washington, DC
The Chief of the Risk Assessment and Management Branch, Betsy Southerland, welcomed
attendees to the national technical workshop. She noted that a diverse group had assembled for
the workshop. For example, more than a third of the attendees were from state agencies. Also,
EPA and a number of Federal agencies were represented as well as people from environmental
groups and industry organizations. The backgrounds of attendees were also diverse:
toxieologists, chemists, biologists, managers, lawyers, etc. This demonstrated that PCBs
continue to be an important topic throughout the country.
Dr. Southerland provided some historical background on the PCB workshop. The Risk
Assessment Branch first became interested in PCBs from the standpoint of sediment
contamination. Preliminary inventories were compiled listing contaminants found in sediments
and that posed human health and ecological problems. Not surprisingly, PCBs were one of most
frequently documented chemicals. It became obvious that, where there were high levels of PCBs
in the sediments, the states had been forced to issue a fish consumption advisory or ban in many
cases. That led to a focus on PCBs in our discussions with the states on fish consumption
advisories. Since fish are an extremely valuable source of protein, Dr. Southerland emphasized
that it is critical to carefully evaluate human health concerns before issuing restrictions on fish
consumption.
An increasing number of States are conducting risk assessments as part of their fish
consumption advisory process. However, current practice is quite divergent. In 1989, our
group provided a grant to the American Fisheries Society to do an inventory of all the state
government policies regarding fish consumption advisories. We asked them how they sample
and analyze fish tissues, how they set the limits for consumption advisories, and how they
communicated the risks. It was a very comprehensive survey. Although the FDA has explained
that assumptions underlying their action levels may not be appropriate for recreational and
subsistence fishermen, 34 states responded that they were using the FDA action level to trigger
their advisories. Ten states said that they use the EPA cancer potency factor to do a limited risk
assessment and to define the level at which they want to set fish consumption advisories. And
13 other states said that they used their own methodology without defining that any further, but
were not using the EPA cancer potency, or the FDA action level. The survey also confirmed
-------
that responsibilities for fish advisories are often spread throughout several agencies. This
multiple agency involvement makes it even more difficult to arrive at a consensus about when
to issue a fish advisory.
1.1.2 Rick Hoffmann,
Fish Contamination Section, Risk Assessment and Management Branch, Office of
Science and Technology, EPA
As the organizer of the workshop, Mr. Hoffmann elaborated on the purpose of the workshop
and some of the planning considerations for the workshop. He noted that the immediate
objective of the workshop is to give a "snapshot" of key risk assessment and analytical issues
related to PCBs. In prior discussions with state agency representatives, EPA was encouraged
to design a workshop that would allow an exchange of PCS information across the boundaries
that typically separate various programs, agencies, and academics. This information assumes
a greater importance as states move towards risk-based fish advisories. A longer term objective
for the Fish Contamination Section is to assess where EPA can provide further technical
assistance to the states.
Mr. Hoffmann mentioned that the levels of PCBs in fish tissues are generally declining.
However, there is an continuing focus on the potential health effects of PCBs, even at low
concentrations. In addition, agencies need to be aware of the appropriate analytical methods and
associated costs. Technical evaluations of PCB risks are often complicated by a contentious
atmosphere, scientific disputes, cost-benefit debates, and other intangibles.
The workshop was designed to start with general information about PCBs and then move
to specifics. Mr. Hoffmann explained that the workshop would talk about health effects,
approaches to analytical methods, and then illustrate how several agencies have applied this
information, using a series of case studies. He clarified what the workshop would not do. The
workshop's primary focus is on human health effects, not on ecological effects or status and
trends information. Secondly, the workshop would not duplicate some of the in-depth, single-
focus PCB workshops that EPA has held previously. For example, EPA's Risk Assessment
Forum in 1989 held a workshop on toxic equivalency factors for PCBs. In September 1992,
EPA held a workshop on the neurotoxic effects of PCBs. Finally, the workshop was not
designed to provide step-by-step guidance or policy judgments for site-specific risk assessments.
1-2
-------
1.2 INTRODUCTION TO PCBS AND ANALYTICAL METHODS
Mitchell D. Erickson, Group Leader, Environmental Research Division, Argonne
National Laboratory
Polychlorinated biphenyls (PCBs) are a class of 209 discrete chemical compounds, called
congeners, in which one to ten chlorine atoms are attached to biphenyl.
PCBs were commonly produced as complex mixtures for a variety of uses, including
dielectric fluids in capacitors and transformers. The major producer, Monsanto Corporation,
marketed PCBs under the trade name Aroclor® from 1930 to 1977. Aroclors were marketed for
use in transformers, capacitors, printing inks, paints, dusting agents, pesticides, and many other
applications. Their chemical and physical stabilities and their electrical insulating properties led
to the commercial utility of PCBs. Their chemical and physical stability also has been
responsible for the PCB contamination in the environment. Because PCBs do not readily
degrade in the environment after disposal or dissemination and are lipophilic, they are persistent
and tend to bioaccumulate. PCBs have been shown to be nearly ubiquitous environmental
pollutants, occurring in most human and animal adipose samples, milk, sediment, and numerous
other matrices.
As early as 1936, occupational exposure was reported to cause toxic effects and
workplace threshold limit values were subsequently set. Animal studies with both commercial
mixtures and individual congeners have shown a variety of chronic toxic effects. PCB-
contaminated cooking oil caused a total of 1,291 cases of "Yusho" in 1968 in western Japan.
The clinical manifestations include various somatic complaints, low birth weights, chloracne, and
pigmentation. The animal toxicology data have intended to indicate that PCBs are toxic.
However, contamination of the commercial PCB mixtures with more toxic compounds such as
polychlorinated dibenzofurans (PCDFs) confounds the data. For example, it is unclear whether
the PCBs or other contaminants in the Yusho oil are responsible for the observed health effects.
PCB toxicity is dependent not only on the degree of chlorination but also on the isomer. For
instance, having no ortho-substitution but heavy substitution at the meta and para positions can
assume a planar conformation that can interact with the same receptor as 2,3,7,8-
terachlorodibenzo-p-dioxin (TCDD). Examples include 3,3',4,4'-tetrachlorobiphenyl,
3,3',4,4',5-pentachlorobiphenyl, and 3,3',4,4',5,5'-hexachlorobiphenyl.
The public, legal, and scientific concerns about PCBs arose from the findings that PCBs
were toxic and therefore undesirable as commercial products or environmental contaminants.
The evidence for this toxicity was sufficient for special citation by the U.S. Congress in the
Toxic Substances Control Act as well as similar actions by other governments. However, the
degree of toxicity and the nature of the effects on humans and other organisms have been and
continue to be highly debatable.
Most PCB analyses follow a prescribed procedure, often issued by a regulatory agency.
Most analyses consist of extraction, cleanup, determination, data reduction, and quality control.
Sampling is an important component of the overall procedure, but it is omitted from this
1-3
-------
presentation because of space limitations. Reliable trace organic analysis begins with the
quantitative extraction of the analytes from the sample matrix. The general objective of an
extraction technique is to separate the analyte (e.g., PCBs) from the sample into a matrix that
is more compatible with the rest of the analytical procedure. The cleanup takes advantage of
the difference in physical or chemical properties of PCBs and interferences to remove unwanted
constituents. The cleanup process may be expressed in terms of enrichment, where the ratio of
PCBs to interferents is increased. Ideally, a cleanup reproducibility achieves 100 percent
recovery of PCBs in one fraction, with the interfering compounds relegated to other fractions.
All analytical methods are designed to determine whether the analyte is present, how
much analyte is in the sample, or both. The identification and quantitation are generally
accomplished in the same step. This determination step is the foundation of any method, around
which all other steps (cleanup, data reduction, quality control, etc.) are centered. With PCBs,
a gas chromatograph (GC) separation is almost always an integral part of the determination
technique. The separation is accomplished by using either packed column (PGC) or high-
resolution (capillary) (HRGC) techniques. The GC effluent is detected by using electron capture
(BCD), Hall electrolytic conductivity (HECD), mass spectrometry (MS), and other detectors.
In an analysis where "total PCB" is the desired output, packed column GC may provide
sufficient resolution. On the other hand, if congener-specific analysis is required for a
metabolism study, HRGC would be the technique of choice.
The qualitative discrimination power of a detector is a major factor in the selection of
a determination technique. The discrimination is especially relevant because of the variety of
PCB mixtures giving rise to complex chromatographic patterns.
Data reduction is a key element in sample analysis. In this step, the analyst converts the
instrumental output into information for the user. Specifically, any PCBs present in the samples
are identified and quantitated, and these results are reported to the user. Depending on the
detection and output system, data may be presented to the analyst as analog chromatograms,
numerical tabulations, MS-extracted ion current plots, etc. Computers and integrators can
reduce the analyst's work in data acquisition and reduction; however, the judgement of a
qualified analyst is critical to reliable data reduction.
The importance of data reduction cannot be overemphasized, especially with PCBs. The
first task in data reduction is to qualitatively identify the analyte, confirming its presence.
Quantitation can be attempted only after a compound has been identified, although the required
level of confidence in the identification can vary widely. Quantitation (or, less accurately,
quantification) is the final step in a chemical analysis sequence. Some measure of signal
intensity (peak height or peak area) is converted into concentration. For most detectors used in
PCB analysis, the mass of analyte in a peak is proportional to the signal for the analyte, and
concentrations are calculated on the basis of this relationship. With most organic compounds,
quantitation is relatively straightforward. The instrumental response is calibrated by using
standard solutions of the compound. The amount of unknown is measured by comparison of the
signal it generates with the calibration factor or curve. Quantitation of PCBs is not nearly so
1-4
-------
simple because the analyte is not a single compound but rather a complex mixture of 209
possible congeners. In addition, standards of all 209 congeners are not readily available for
calibration. Given these problems, analysts have devised alternate quantitation methods, often
based on the similarity of the sample PCB mixture to a commercial product (e.g., Aroclor).
Aroclor-rbased quantitation schemes may be appropriate if the sample and standard "fingerprints"
are similar. As the similarity diverges, the quantitative confidence diminishes.
The data report must be formatted to fulfill the analytical objectives. If individual
congener concentrations are needed, the report will be complex, while if "total PCB" is
sufficient, the data report may consist of a single value. The report must specify the reporting
units. The analyst should include on the data report some measure of the qualitative confidence.
This is often done in the text, or as a footnote.
Emphasis on quality assurance in chemical analysis has increased dramatically in the past
few years with the realization that data of unknown quality are virtually useless. Since PCBs
generally occur as complex mixtures of analytes, special quality control measures must be
considered. The PCBs used for calibration of the analytical instrument may be a mixture (e.g.,
Aroclor 1254) similar to that found in the samples or a group of individual congeners. Any
realistic option is a compromise from calibration with all 209 congeners. Thus, an estimate of
the error induced by the compromise should accompany the data. Because of the complexity
of the data, special precautions should be taken to assure both the qualitative and quantitative
aspects. Many quantitation techniques involve summing the calculated responses or
concentrations for many individual PCB peaks to yield a total PCB value. Any systematic error
replicated through several quantitations could result in a magnified error in the reported result.
The complexity of calculations also increases the chance of calculation and transcription
mistakes.
1-5
-------
o>
o
CO
o>
(0
CD
o
Q.
CO
o
in
o
00
o
CM
VVWVS^VV^W.VVSV\IW^V.V.WVVVW^A'AVrt»«V.WlV^.W.V^
o
o
o
o
CO
o
o
10
o
o
CO
o
o
CM
O
O
o>
OS
o
f
g
o
CO
1-6
-------
W
E
c
o
(k.
c
LU
15
u
• MMK
i^HM
CL
1-
sz,
c
'3
DC
£
CO
,-»
4*
O)
5
c
"8
o
o
0 o 0
T- ID O LO
• T- in CM
CO i i i
0 0 i- 0
O T- T-
0
Q in c\J ° CM O 'o
, , , co <^> '
O" • w} CD .^j* i u>
• OO/-%n— ^^-* O
o v 2 T~
V
^^
CO JO "cO
^ CO CO o
/IN CO CO CO -43
J3 -J CD co
§g3"co § § 2-g=3
*1^* ^Tj *^^ ^^^^ ^H^ ^*^V jSrf ^^^F
^1^ ^^^ *^*T "^1*'^ ^^^^J ^^^^J tl» f ^
a: o a: ±) JE JE 2o'O
1-7
-------
C
"x
O
5
0)
(0
CQ
O
Q.
CO
15
Q.
O
O
O I O O J. O
O O
CO
•<
o
1-8
-------
1-9
-------
1.3 TEMPORAL TRENDS OF PCBS IN THE ENVIRONMENT1
Written by Robert J. Fensterheim, RegNet Environmental Services, Washington,
DC. Presented by John Craddock, Principal, Craddock Associates Incorporated,
St Louis, MO
Based on a comprehensive literature review, data for several environmental compartments,
including air, precipitation, drinking water, sediment, human milk, soil, waste, and general
disposal, were considered too poorly documented to characterize temporal trends. However,
adequate information was available to discuss temporal trends for the five following
compartments: foods, human adipose tissue, shellfish, fish, and human blood sera. Review of
the available information for these compartments reveals that there is essentially no pattern of
studies that suggests an increase in the concentration of PCBs. On the contrary, a considerable
number of studies document a significant decrease in PCB levels. Higher confidence levels in
a trend assessment is typically reserved for studies or groups of studies where there is
consistency in sampling and analysis. Although all of the studies examined are not of such
quality, evidence presented in this report clearly point to an overall decrease in PCB levels
throughout the environment. The overall studies show that, through 1985, there were sharp
decreases in PCB concentrations in the environment as its uses were phased out. Since the mid
1980s, the rate of decline has slowed as PCB contamination levels approach trace levels in most
environmental compartments.
Foods
The National Research Council in 1979 identified persistence and lipophilicity as the "most
important properties of a chemical that relate to potential hazards to health and the
environment." Since PCBs are highly lipophilic and will slowly biotransform, they tend to
concentrate in tissues with high fat content. Thus, the temporal trends of PCBs in foods,
particularly those with high fat content, and in human tissues, are critical in assessing public
health implications.
PCB levels in food are regulated by the Food and Drug Administration. The FDA has
set tolerances for PCBs ranging from 0.2 ppm in infant and junior foods and feed for food-
producing animals to 10 ppm for paper food-packaging materials intended for use with human
food. The tolerance level for fish and shellfish is 2.0 ppm. High levels of PCB contamination
of food during the 1970s and the early 1980s frequently occurred because of contamination in
food packaging. The use of PCBs in carbonless copy paper was cited by the National Academy
as leading to contamination of paper products and foods by the recycling processes. Sawhney
and Hanldn note that these sources of contamination have been "essentially eliminated" since the
Federal government regulated the use of PCBs. The main dietary route of human exposure to
PCBs is now through consumption of PCB-contaminated fish. Thus, the majority of studies
Ed. note: This paper summarizes a report, prepared by RegNet for several trade associations, which
complies and interprets PCB trends information from a variety of federal and state agencies.
1-10
-------
identified focus on PCB levels in fish. Fish are analyzed as part of the FDA's Total Diet Study,
the principal source of information for documenting trends of PCBs in the U.S. food supply.
The results of this program indicate a substantial decline in PCB residues to near zero in recent
years.
FDA's Total Diet Study, also known as the Market Basket Survey, began in 1961,
initially to determine dietary intake of radionuclides resulting from atmospheric testing of nuclear
weapons. In 1971, the program was expanded to include analyses of selected nutrients and
pesticides, including PCBs. By 1982, this study included 234 food items representing 100
percent of the diet of the U.S. population. The primary purposes of the study, as currently
described, are to estimate the dietary intakes of pesticides, industrial chemicals, and other toxic
elements and radionuclides and to compare these intakes with established safe or recommended
dietary levels. Food items are collected from retail outlets simultaneously in three cities in each
of four regions in the U.S. and shipped to the Total Diet Lab in Kansas City for analysis.
Results on PCB contamination from the Total Diet Study are periodically reported in published
articles by the FDA researchers. The last report presenting data on PCBs was published in 1988
and includes data for the years 1982 to 1984. A more current profile of PCB trends documented
in the Total Diet Study show a significant decrease in PCB intake from 6.9 ug/day in 1971 to
0.05 ug/day for the years 1987 to 1990. From 1977 to 1980, there was a slight bobble in the
curve. The PCB intake rose slightly to 1.9 ug/day before dropping off again to the 0.2 level.
Human Adipose Tissue
The lipophilic properties of PCBs means they tend to collect in adipose tissue, making adipose
tissue a useful method for monitoring human exposure to PCBs. Adipose is especially preferred
in detecting low level or chronic exposure. Adipose tissue studies, however, are more difficult
to undertake than other environmental surveys because of the intrusive nature of collecting the
samples, necessitating collection from mainly autopsied cadavers and surgical patients. The
NAS in 1979 identified a mean concentration of 1.2 mg/kg of PCBs in adipose tissue of the U.S.
population. Robinson and others reviewed available data documenting trends of PCBs,
hexachlorobenzene, and benzene hexachloride in adipose tissue in the U.S. and concluded that
"while nearly the entire population has detectable levels of these chemicals, the actual
concentration levels are steadily decreasing."
The?principal source of information for documenting trends of PCBs and other substances
in human adipose tissue is the National Human Adipose Tissue Survey known as NHATS. The
primary purpose of the NHATS program is to establish baseline levels for the presence of toxic
chemicals in human tissue, adipose tissue, and to make statistical comparisons of the residue
levels for the population and sub-populations, including trends over time. Samples collected
during 1982 were analyzed using high resolution gas chromatography and mass spectrometry,
a technique which more sensitively detects volatile organic compounds. Samples taken during
1984 were split and analyzed using both of these methods to allow comparison of the two
different analytical techniques. Thus, two sets of data exist for the 1984 samples. The high
1-11
-------
resolution gas chromatographic method allowed for detection of each homolog rather than
measuring PCBs as if PCS is one compound. The high resolution gas chromatographic method,
furthermore, has an average detection limit for each homolog ranging from 0.03 ug/g of lipid
compared to 0.33 hi the electron capture method. The data reveals an upward trend in the
percentage of population having a detectable level of PCBs. The 1972 data show approximately
90 percent of adults in this country having detectable levels. This level increased to 100 percent
in 1981 and 1983. Adipose tissue levels were obtained from two Canadian cities between 1979
and 1981, and 100 percent of the samples analyzed in the Canadian cities were found to have
quantifiable levels, confirming what was found in this country. However, the results are
somewhat disputed when you look at the more sensitive high resolution gas chromatographic
method used in 1982, 1984, and 1986 composites. For 1982 samples, PCBs were detected in
only 38 out of 44 composites, about 83 percent, with all of the non-detectable samples coming
from persons in the 0 to 14 year old age group. A higher but still less than 100 percent
detection was also seen with the 1984 composites using the high resolution gas chromatographic
method. These results showed 98 percent had detectable PCB levels, while the electron capture
method reported 100 percent detection. More notable than the level of detection is that the
actual levels of PCB contamination reported by the electron capture method appear to be greater
determined using the more sensitive high resolution gas chromatographic technique.
Another significant finding from the NHATS program is that PCB levels in adipose tissue
over time show a downward trend in the percentage of the population with PCB levels greater
than 1 ppm or greater than 3 ppm. The percentage of persons having greater than 1 ppm PCBs
in adipose tissue showed a significant decreasing trend from a high of 62 percent in 1972 to 2
percent in 1984. The percentage of persons having greater than 3 ppm of PCBs also declined
from a high of near 10 percent in '78 to 0 in the '83-84 time frame.
Blood Sera
Analysis of PCB levels in blood sera is a valuable means to study PCB levels in humans also.
Although blood samples are less sensitive than other human tissue samples such as adipose tissue
for detecting small exposures to PCBs, the analysis of blood sera gives a direct indication of the
level at which the internal body organs are exposed. According to ATSDR, PCBs adipose levels
correlate with blood sera levels readily, although low level exposure over the long term is better
detected in adipose tissue. For those studies where multiple years of sampling were conducted,
downward trends were seen, suggesting that PCBs are not stored in blood as they are in adipose
tissue but tend to readily dissipate from blood sera after elimination of the exposure.
A study of Michigan residents found a decrease in serum PCB levels in persons
consuming greater than 10 kg of fish per year, from 56 ng/ml in 1974 to 21 in 1980. The study
authors note that this decline followed a reduction in the amount of PCB contaminated fish
consumed per year. For non-fish eaters, PCB levels declined also from a median of 15 ng/ml
in 1973 to 6.6 ng/ml in 1983. In Japan, a study of persons accidentally exposed to PCBs
concluded that, within five years after exposure, PCB levels in the blood were recorded at
almost the same levels as those of unexposed individuals. This decrease in blood sera PCB
1-12
-------
levels was also found, in pccupationally exposed Japanese workers. The results cited are
consistent with results obtained from the Greater New Bedford Harbor PCB health effects study.
This study was prompted by the discovery of high levels of PCBs in seafood taken from New
Bedford Harbor, Massachusetts.
Shellfish and Fish
The National Academy of Sciences in 1979 estimated that 50 to 80 percent of PCBs present in
the environment are contained in North Atlantic waters. Because of PCBs' high octanol/water
partition coefficient and low water solubility, PCBs readily adsorb to suspended particulates and
bottom sediment. PCB levels in sediments represent a time-averaged indication of contamination
at the sampling site or an indicator of historical PCB inputs rather than a representation of
contamination in the mobile aquatic environment.
The analysis of organisms is considered the most accurate measure of PCB contamination
in the aquatic environment. Ideal bioindicators generally have the following characteristics: they
accumulate PCBs in proportion to the average levels present in the ambient waters; they are
sedentary in order to be representative of the area from which they are collected: they are
.abundant in the study area; they tolerate the presence of high levels of the pollutant without
being lexicologically affected; they are of reasonable size, giving adequate tissue for analysis;
they should be easy to sample and be hardy enough to survive in the laboratory; and they are
long lived enough to permit time integration of the pollutants over several months or years.
Shellfish and, in particular, bivalves such as oysters, mussels, and clams are useful in evaluating
contaminant trends, because contaminant levels in tissues of a single organism change fairly
rapidly, that is, within months, in response to changes in water quality. Mussels provide an
especially good indication of spatial and temporal contaminant trends, because they are sessile
and abundant in many geographical areas. These characteristics also allow researchers to assess
environmental contaminant levels for a specific locality and time frame.
In 1988, NOAA conducted a historical assessment of trends in PCB contamination in
shellfish and fish, which examined available data from studies on fish and shellfish conducted
in the U.S. in an attempt to identify long-term trends in PCB contamination. The survey was
based on a synthesis of approximately 35,000 pieces of historical data from many different
studies, covering over 540 species collected between 1940 and 1985. Data on PCB levels in fish
and shellfish were contained in over 11,000 records. The study found that PCBs have been
detected in all estuaries sampled, including remote, non-industrial areas. The highest
concentrations were found to occur in fish samples near the urban areas. The results of data
compiled for NOAA '88 did not allow the study authors to make conclusions regarding a
national trend in PCB contamination in the U.S. among fish and shellfish as a whole. However,
for specific species such as the striped bass and menhaden, NOAA's 1988 historical study was
able to develop 15 to 20 histories of PCB contamination. Within geographical areas such as
Chesapeake Bay and San Francisco Bay, long-term declines in PCB contamination were
documented. In fact, site-specific sampling has yielded very few instances where PCB levels
increased in the short or long term.
1-13
-------
In conclusion, PCBs concentration in the environment have undergone a significant
decline over the past 15 years. The most dramatic declines appear to occur in the late 1970s
and the mid 1980s time frame which corresponds to the period following the regulatory controls
imposed by the Toxic Substances Control Act. Declines in PCB levels were documented in
environmental compartments of public health significance, most notably foods for human
consumption, human adipose tissue, shellfish and various fish species. Most of the reports that
we have screened suggest that these declines will continue through natural degradation processes.
1-14
-------
Trends in PCB Contamination
Lower Hudson River Striped Bass
180
CO
D)
3 140-
T3
CD
CO
CO
CD
j6
!9- 100
CO
CD
O
Q_
I
co
CD
80
20-
^8 79 80 8?
~Q2 83 84
Year
85 86
"S7 88 90^
1-15
-------
PCB Levels in Flatfish
Various Sampling Sites
JC
CO
I
U-
CO
03
O
OL
3
8-
7-
6-
5-
4-
3-
2
1
0
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85
Year
Elliott Bay
Orange Cty
S Monica
New Bedford
S Pedro
1-16
-------
1986-88 Mussel Watch Data
Sites in Washington State
1200
(0
"I 1000
-CJ?
D)
^
0)
QQ
O
CL
1
800
600
400
200-
1988
Grays Hrbr
Elliott Bay
Sinclair Inlet
Pt Roberts
1-17
-------
1986 - 88 Mussel Watch Data
Sites in California
1988
— "— Pt. Loma —
-e- Pt S Barbara -
-»— S. Catalina Is -**-
x— Pt Conception -^t-
- Santa Cruz Is
- Pt Delgada
1-18
-------
PCB Levels Greater Than 3 ppm
by Age Group, 1972 - 1983
25-1
20-
P
C
ft
c
£
N
r
c
r
p
p
M
15-
10-
5-
0-
Legend
0-H yrs
15-44 yrs
,45* yrs
7« 77 78
FISCAL YEAR
••••!•
79
80
(Reprinted from EPA, Office of Toxic Substances, Baseline Estimates and Time Trends for Beta-
benzene hexachloride, Hexachlorobenzene. and Polvchlorinated Biphenvls in Human Arfi T
1970-1983. Washington, DC, September 1985.)
1-19
-------
1986 - 88 Mussel Watch Data
Sites With Significant PCB Decreases
Boston Harbor
Hudson Estuary
Delaware Bay
1-20
-------
1.4 PCB TRENDS IN GREAT LAKES FISH
David De Vault, U.S. EPA's Great Lakes Program Office, Chicago, IL
PCB contamination is ubiquitous across the Great Lakes and residues occur in fish tissue from
all areas of the Basin (Figure 2). Lakes Superior and Michigan represent the least contaminated
and one of the most contaminated of the Lakes, respectively. These two lakes may also be used
to illustrate the overall success of PCB regulation (Lake Superior), and an area where we need
to do additional work (Lake Michigan).
PCB concentrations across the Great Lakes have declined significantly as a result of bans
on manufacture and use implemented in the mid 1970's. Concentrations in the Lake Superior
water column (Figure 5) and lake trout have declined from around 1 ng/1 in 1980 to less than
0.2 ng/1 in 1992. Concentrations in whole lake trout similarly declined from nearly 2 ug/g in
1980 to 0.5 ug/g in 1990 (figure 6). In both cases, the declines are consistent with first order
loss kinetics, and the decline continues through the most recent data available. This is supported
by a PCB mass budget for Lake Superior (Figure 7), which estimates inputs to the Lake at
around 200 kg/yr and losses, primarily through volatilization, at around 800 to 900 kg/yr. PCB
loading to Lake Superior appear to be below that required to maintain existing concentrations
and further declines in concentrations are expected.
PCB concentrations in several Lake Michigan fish species have declined in response to
regulatory actions (Figure 8). However, these declines have slowed or stopped in recent years
in several species. Lake Michigan lake trout (Figure 9) declined significantly from 1974 through
1982. However, from 1986 through 1990 concentrations declined only slightly, if at all. During
this period, concentrations were above the loss rate calculated from the 1974-1982 data, further
suggesting a leveling off of the downward trend. The Lake trout data represent fish that are
approximately 6 years old and concentrations have been relatively stable since 1986, suggesting
(using an very simplistic load-exposure scenario) that the PCB loadings to southern Lake
Michigan have been relatively constant since the early 1980s. Fillets from fall run coho salmon
(Figure 10) show a similar situation, with a definite leveling of concentrations beginning in
1983. Coho salmon are stocked in Lake Michigan and are in the lake for approximately 18
months before they return to the tributaries to spawn. They are taken in the Great Lakes Fish
Monitoring Program during the fall spawning run. Using the same simplistic relationship
between loads and exposure that was used for lake trout, the coho data would also suggest that
PCB loadings have been relatively stable since the early 1980s.
Dated sediment cores provide a more direct measure of relative net loadings than fish,
which are subject to food chain and other exposure variations. Core data from the southern
basin of Lake Michigan (Figure 11) indicate that concentrations in sediments increased rapidly
through the 1940s, 1950s, and 1960s, reflecting PCB manufacture and use. Concentrations then
declined from the mid 1970s, when use and production bans were implemented, through the
early 1980s. Concentrations in sediments from about 1982 through 1989 were virtually constant,
reflecting constant loading rates. Thus three independent data sets indicate that PCB loading to
southern Lake Michigan decreased beginning in the mid 1970s through the early 1980s, and that
1-21
-------
loadings have been relatively constant since then. Unfortunately, those loadings are sufficient
to maintain tissue residues at unacceptably high concentrations, as illustrated by Sport Fish
Consumption Advisories for several Lake Michigan species.
PCB congener data (normalized as a percentage of congener 153) shows that congener
profiles for lake trout from Lakes Superior, Michigan, Huron, and Ontario, as well as, walleye
from Lake Erie are quite similar (Figure 12,13). The similarity in congener profiles from open
lake samples across the Great Lakes suggests a similar source, such as the atmosphere. That
the source of PCBs is reflected in the congener profile in fish tissue is illustrated by the very
different profiles seen in Green Bay walleye, where the source is contaminated sediments in the
Fox River (Figure 14).
The indication that there is an atmospheric source, or sources, to southern Lake Michigan
is further suggested by atmospheric samples collected over the Great Lakes (Figure 7). These
data indicate relatively little variation across the basin, with the exception of Lake Michigan,
where there is a strong north or south gradient. Concentrations over the southern portion of the
Lake are nearly an order of magnitude above background and are twice the levels found in the
next most contaminated area (Detroit/St. Clair River).
Thus there is substantial evidence suggesting that PCB loadings to southern Lake
Michigan are being maintained at unacceptably high levels due to atmospheric sources. We are
attempting to identify the ultimate source of these loadings, and suspect the highly industrialized
Chicago-Northwestern Indiana region.
Some concern has been expressed that, although total PCBs have decreased,
concentrations of the planar PCBs may not be decreasing at equivalent rates. Figure 17
illustrates trends in mean concentrations of several planar congeners in Lake Michigan Lake
trout. Figure 18 presents the log fit of these data to time and illustrates that, with the exception
of congener 77, the declines in lake trout are similar to that observed for total PCB.
Data Acknowledgements
Steven Eisenreich, University of Minnesota, Gray Freshwater Biological Institute, Navarre,
Minnesota: Figure 5 - PCBs in The Lake Superior Water Column; Figure 7 - PCB Budget
for Lake Superior-1992; and Figure 11 - Lake Michigan Sediments-Site 18.
Robert Hesselberg, U.S. Fish and Wildlife Service, Ann Arbor, Michigan:
Figures 16 and 17 - Lake Michigan Lake Trout, Planar PCBs.
1-22
-------
1-23
-------
3
o
O
S
0
Q.
CO
cd
J
0)
CO
CQ
O
Q.
1-24
-------
CO
0)
tu
03
-------
i
CO
Q.
-------
PCBUDGET FOR LAKE SUPERIOR • 1992 J
Atmospheric Deposition
Wet 125
Dry 32
Total 167 kg/yr
Atmosphere
80kg
Volatilization
-800 kg/yr
Particle
Settling
•3000 kg/yr
Water Column
-3600 kg
Recycling
-2900 kg/yr
Water Column-1980 -12000 kg
-1992 -4000kg
Water Column Loss Rate - 700 kg/yr
Water Column Loss - Volatilization
Outflow
40 kg/yr
Burial I
50-100 kg/yr *
1-27
-------
§
bo
2f 0.
(0
-I
OJ
*- O?s
1-28
-------
D
O
CO
cm
OJO
D)OL
= .0
5
CO
¥
o g
o>
o
o>
0)
CO
CO
CO
N
CO
CO
CO
to
CO
CO
<0
CO
CM
« i.
T» CO
CO fl)
o ^"
CO
o
N
CO
N
N
N
to
N
IO
N
"fr
N
CM
N
O
CO
IO
CM
U>
IO
10
1-29
-------
D)
OP
o
O
0>
o>
o
o>
0)
O)
CO
o>
CO
00
CO
0)
I
o
00
O)
04
00
CD
CO
o>
CO
0>
ol
to
CM
1-30
-------
00
9t 9tG*0\Ok9* 9t 00 00
_< »4 *4 *•* _• i_C _* »M «M
o r* *
-------
8 § 8 8
1-32
-------
1-33
-------
r
*
§
1-34
-------
0)
m
O
03
.2
00.
CO
u>
r*.
u>
CO
CM
h-
h-
2
I
CO
li.
0)
.Q
O
CO
CO
«
I
1-35
-------
CO
T5
C
CD
O
CL
CO -j=
OL .2
T"
I
o>
1-36
-------
1.5 OVERVIEW OF PCB TOXICOLOGY
Michael Bolger, Chief, Contaminants Standards Monitoring and Program Branch,
Center for Food Safety and Applied Nutrition, U.S. FDA, Washington, DC
The following is a brief presentation of the toxicology and hazard and risk of PCBs, with a focus
on pharmacokinetics. PCBs, as a class, are highly lipophilic. Absorption of PCB congeners
is determined by their relative lipophilicity and is dependent on the ability of PCBs to cross fatty
membranes. Lesser chlorinated congeners are more readily absorbed. With increasing
chlorination, congeners are more lipophilic; however, absorption decreases because viscosity
increases, and solubility decreases, thus the congeners literally become stuck in membrane
structures. Absorption of PCBs is greater orally than by the dermal route because of size and
lipophilicity considerations. In terms of the absorption of PCBs via inhalation, very little work
has been done. However, one would surmise that the same generalizations presented above
would apply somewhat in the lungs.
After absorption, PCBs are distributed throughout the body by lipoprotein carrier
molecules. Partitioning from the blood to the tissues occurs rapidly. Chronic exposure to PCBs
results in concentrations in fatty tissues of the body. The most abundant storage site is the
adipose tissue, followed by the liver and skin. The distribution of PCBs is modulated by body
composition. That is, mobilization of storage depots occurs when fat stores are metabolized as
is seen with reduction in body weight and lactation. Elimination of PCBs can occur both
actively and passively. The major route of PCB elimination is dependent on metabolic changes
that effectively increase water solubility and increase elimination. Direct oxidation, introduction
of an oxygen molecule into the ring structure of PCBs, can also occur. Metabolism also occurs
by the mixed function oxygenase (MFO) system. For example, the action of the MFO system
results in the formation of an arene oxide intermediate which can be converted into dihydrodiol
by epoxide hydrase. The arene oxide intermediate can also be converted into a phenol, which
in turn can be conjugated with glucuronic acid. Further action by MFO can result in the
formation of diols. The arene oxide intermediate can be converted via reduced gastrointestinal
into a thioether, which is particularly important in terms of metabolism in the lungs. Finally,
dechlorination, that is removal of the chlorine molecule from the ring structure, does occur. In
passive elimination, diffusion occurs across the gastrointestinal wall, and partitioning into the
sebum, sweat, or other bodily secretions including milk occurs with active elimination, molecule
weight determines the route. In other words, smaller molecules will be eliminated via the urine,
larger molecules via the feces. Metabolism is most effective in congeners with two adjacent'
unsubstituted carbon atoms. Less chlorinated congeners are metabolized more rapidly than
higher chlorinated congeners. Also, congeners with chlorine substituted on one ring are
metabolized more rapidly than those with chlorines substituted on both rings. In active
elimination, the most important step is the initial oxidative step by the MFO system which
results in the formation of an epoxide intermediate. Dechlorination and direct insertion of
hydroxyl groups occurs at the 3 and 3' position. One MFO enzyme is CYP1A, which is
primarily induced by dioxin and related compounds that includes the coplanar PCBs, prefers the
2 or 2' ortho position. The other is CYP2B which is induced primarily by phenobarbital and
related compounds and prefers the 4 or 4' para position. In terms of metabolic efficiency,
1-37
-------
mammals are more efficient than birds, birds more efficient than fish, and within mammals, the
dog is particularly efficient, more so than other mammals including humans.
With acute exposure to PCBs, a range of systemic effects can occur including effects on
the liver, kidneys and gastrointestinal, or neurological system. The latter includes decreases in
brain dopamine levels and changes in behavior. In rodents, adverse effects on male fertility and
decreases in implantation and fetal weight and survival and decreases in weaning survival have
been noted. With intermediate exposure—5 weeks to 8 months—such effects as thymic atrophy,
decrease in natural killer cells, effects on lymph nodes, increase in rates of infection, and
decreases in antibody levels have been reported. In the monkey, PCS exposure results in
decreases in brain and hypothalamic dopamine, levels. Intermediate exposure in ;the rat results
in thyroid, hepatic, skeletal, dermal, and cardiac effects and decreases in body weight.
Developmental toxicity has been reported in the rat, rabbit, guinea pig, monkey, and mink, with
increases in neonatal death, decreases in Utter size, increases in fetal and neonatal deaths, and
increase in resorption/absorption. For reproductive effects, several mammalian species show
changes in the estrus cycle and decreases in rate of conception, reproductive rates, and litter
size. With chronic exposure, the major effects include decreased survival, body weight gain,
and spermatogenesis and dermal, hematological, gastrointestinal, and thyroid effects.
Studies of PCB structure-activity relationships (SAR), generally, have focused on the
dioxin-like congeners that are non-ortho substituted. They bind with the AH receptor, which
is an index of dioxin reactivity. The molecular structure is relatively flat with little to no
rotation, resulting in a stacking type of interaction in the receptor binding domain. The
interaction with this receptor eventually results in the expression of "a messenger RNA" which
includes the activation of the MFO system. Thus you have a molecule that actually activates its
own metabolism. The induction of various types of proteins are supposedly involved in the
pleiotropic and/or the toxicological responses seen with dioxins and dioxin-related compounds.
Chlorine substitutions in ortho substituted congeners tend to twist and bend the molecule making
it less "dioxin like" and as a result binding with the AH receptor is poor, if at all. The binding
interaction is known as "cleft-type" such as seen with the thyroxine carrier protein, prealbumin.
T4, which is structurally similar to the hormonal-like PCB congeners, has two ring structures
which are juxtaposed 90 degrees to each other. The relevance of prealbumin binding is that this
may prove useful in terms of a description of relative reactivity of the different PCB congeners,
particularly the ortho-substituted congeners. This in turn could prove useful in the development
of a toxicity equivalency factor (TEF) approach for "non-dioxin- like" PCB congeners which is
important in that it is likely that the most persistent PCB residues found in human tissue will
bind to prealbumin.
The SAR studies conducted on prealbumin binding indicate that laterally substituted PCB
congeners have the greatest activity. In addition, at least one lateral substitution will result in
prealbumin binding and ortho substitution does not diminish binding appreciably. Neither
biphenyls or fully substituted congeners demonstrate any binding activity.
1-38
-------
Recent work with coplanar PCBs suggests that the dioxin-like TEFs developed to date
tend to overestimate the actual potency of the coplanar congeners, depending on the target tissue.
It appears that the TEFs may be tissue-specific. The available SAR information on
developmental neurotoxicity, suggests that a thyroid hormone antagonistic property is important
and may prove useful in explaining the mechanisms of action of specific ortho substituted
congeners. This anti-thyroidal activity, whether it occurs directly at the receptor through an
alteration of thyroid metabolism or by changing the activity of the thyroid (T3 or T4) system,
results is an alteration in neurotransmitter concentrations. There is even some speculation that
a biochemical hypothyroid status at the cellular level occurs.
The SAR studies of the neurotoxicity of PCBs, clearly indicate the importance of specific
molecular properties, specifically ortho or ortho-para substitutions. These studies show that
activity is not related to cellular accumulation, hydrophobicity, or metabolism and that activity
is related to prealbumin binding. The 2,2' substitution is more active than the 2,2',4,6,6' or the
3,5 substitutions where there is more lateral substitution and, with no activity seen with planar
PCBs. It appears that these PCB congeners prefer the T3 family of proteins and that they are
thyroid hormone antagonists.
One of the mechanisms that has been suggested for the action of ortho substitution PCB
congeners is an effect, on dopamine metabolism, which would explain some of the neuro-
developmental effects seen in the rat or primate studies and possibly those observed in human
studies. What has been determined to this point is that ortho substituted congeners (2,2' or the
2,2',4,6 congeners), which are non-planar, exhibit the greatest effects on dopamine content.
Planar "dioxin-like" congeners exhibit no activity. Chlorination in the para position increases
potency, whereas inversed congener chlorination is not correlated with reduced potency and
complete chlorination substitution does reduce potency.
A review of what is known of the toxicity of PCBs indicates that doses, ranging from
0.01 to 100 mg/kg b.w./day result in a variety of adverse effects including, lethality, and
gastrointestinal, hematological, developmental, immunological effects, and systemic cancer. The
lowest observed effect level has been reported for an immunological end point which suggests
that the immunological effect is a more sensitive end-point than cancer. However, this is
somewhat misleading in that this could be an artifact of the quality of the studies. If additional
carcinogenic bioassays of PCB mixtures were conducted, lower dose-response carcinogenic
effects would probably be observed depending on the size of the study and particularly on the
number of animals used in each study group. In addition, the methodology used to extrapolate
the dose response information from animal studies to humans for cancer is inherently more
conservative that the methodology used for non-cancer endpoints. This results in a lower risk
number for cancer that it does for other end-points.
A range of reference doses (RfDs) for PCBs have been developed on the basis of several
lexicological end-points which have been noted in animal studies. In terms of reproductive
effects, decreased conception rate, and infant survival, the corresponding reference doses range
from 2 x 10~2 to 1.4 x 10^ mg/kg b.w./day. In terms of developmental toxicity, an RfD of 1.4
1-39
-------
x 10-s mg/kg b.w./day has been developed. The corresponding RfD for cancer risk at an upper
bound estimate of 1 in a million is 1.3 x 10'7 mg/kg b.w./day. Thus, there is at least several
orders of magnitude difference between the corresponding developmental and cancer risk
reference doses. In terms of human studies, the RfDs developed from the no observed effect
levels or estimated no observed effect levels and an uncertainty factor are about ten-fold greater
than the corresponding RfD for the one in a million cancer risk.
In summary, based on the spectrum of toxicity of PCB mixtures and congeners, their
pharmacokinetics, and structure-activity information, it appears that there are at least two general
classes of PCBs, "dioxin-like" and "hormonal-like". A spectrum of RfDs has been developed
that range from 10"2 to 10"7 mg/kg/day, depending on whether systemic, developmental,
reproductive, immunological, or carcinogenic effects are the end-point of concern.
1-40
-------
Dechlorinat'on
Products
Oxidation
COOH
MFO
lUKcrobial degradation
products
Thioethers
GSH
Oiols
MFO NAOPH
NAOPH 0
Rearamatizatian
Arene Oxide
Intermediate
EpoxkJa
hydrasa
H20
Oihydrodiols
Phenols
Gkicuronidttton,
SuKaflcn
Phenol-conjugates
and derivatives
•Sourc«: Safe 1984
1-41
-------
^H
ti
^•81
^
fff
i|
»»•<*.
B?ai
o
o
o
co
0
to
h-
*•*•*
O)
73
O
CD
O)
T5
O>
CD
C3)
T3
•oilH
T3
O5
•^u3
"D
CD
o o o o
.CD LO m LO
1-42
.
co ">
-------
.«. ,,§^
.1 - .h-
"^f: CO '•
'•-*••' - m •
o -T- : o o
v o o O
I E E £
] CD CD (M CO
LO CO
t5- .JQ
co O
O
•>> >- ,£•
-P. -c- -• >.
1-43
-------
? 8- 2
? i 3 d
5 O -
m o>
T- - c\| CVI O>
o ~*~ oi> •
IT5 OJ
co co od
U= »
T3 T3
O) C
Q-
1-44
-------
i co
10 co
f o o
d, E- E
OJ ;
"O T
CO •"•
CM
!^K*^BI^fl
- !co: ^
• ^f .
E
"'CD.
:E
CD
§Q.- Q.
CD
Q __ '= ~£Z ,
— ~s CL OS ". O
» . .AMU. -J™*
; © 'O ® .
'*- m «*--'.
,-; «_. :; CD ^ *-
1-45
-------
1-46
-------
' ^J)i t*2
O CB
E £ E
o" Jg
•4=; If-*
tS J§ "
Q. o
Q> Q, "D
-
1-47
-------
> CO
CO
i -. n
«J o
O o
cs tn
•& A
,'-.u ;
' B -»~i
•*^Ls i
£ "co
- Q.
J^.' (D
C3 (B
CO T3
1-48
-------
Accessible
Planar
-. Face
' 2 ^ Lateral
•CU-I Substituents
Accessible
, Planar ••'
f-sce
Accessible
Planar
Substituenfs
^ Lateral
I Substituents
1-49
-------
1-50
-------
,*••,
1-51
-------
1.6 PCB CRITERIA FOR WATER
Jennifer Orme Zavaleta, Drinking Water Health Assessment Section, Office of
Science and Technology, Office of Water, U.S. EPA, Washington, DC
The Office of Science and Technology develops risk assessments for water contaminants.
Within the Office of Science and Technology, there are several divisions, including the
Engineering and Analysis Division, Health and Ecological Criteria Division, and the Standards
and Applied Science Division. Within the Standards and Applied Science Division, the Risk
Assessment and Management Branch evaluates risks associated with contaminated sediments and
chemical contaminants in fish. Within the Health and Ecological Criteria Division, risk
assessments for drinking water, ambient water, and sediment contaminants are developed. The
Sludge Risk Assessment Branch, evaluates the disposal of municipal sewage sludge on land. In
the Ecological Risk Assessment Branch, criteria, including those for PCBs, are being developed
based on aquatic life and wildlife effects.
In the drinking water program, drinking water maximum contaminant level goals
(MCLGs) and maximum contaminant levels (MCLs) are developed as well as some of the
ambient water quality criteria (AWQC), specifically for the water and organisms and the
organisms only. In developing the criteria, we regulate contaminants for drinking water that
may have an adverse human health effect and that are known or anticipated to occur in water.
This is done through a two-part process. The first part is developing an MCLG that represents
a non-enforceable health assessment that would not be expected to have an adverse effect, and
incorporates a margin of safety. This is similar to the ambient water quality criteria. The
MCL, which represents the enforceable standard, is set as close to the MCLG as feasible, taking
into consideration other aspects such as analytical methods, available treatment technologies, and
costs. If a contaminant in water cannot be adequately monitored, then a treatment technique is
established that would, more or less, remove the contaminant as much as possible from the water
supply.
In comparison, AWQC are derived from an assessment that is based on the available
scientific knowledge and reflects the identifiable health effects as well as the effects to ambient
organisms found in water. In essence, this process involves an assessment of the entire
ecosystem of which humans are considered a part of the system. The AWQC, like the MCLGs,
are not federally enforceable, but unlike the drinking water program, the actual enforceable
standard is developed by the states. They may adopt the AWQC values that EPA recommends,
or they may develop their own criteria based on site-specific considerations. AWQCs that are
developed as part of this program are developed for aquatic life and also for human health
protection. In some of the comparisons with the surface water methodology for developing
human health criteria, the process begins with a quantitative risk assessments for either the
reference dose looking at non-cancer health effects or for the q!*, which represents a cancer
potency factor looking at a quantitative estimate for excess cancer risk. These assessments, are
then adjusted for a 70 kg adult and then divided by a number of other factors looking at water
consumption, bioconcentration factor, and fish consumption.
1-54
-------
In general, the water consumption rate used is 2 liters of water per day. This rate is used
in the drinking water program in developing the MCLGs. One difference, though, is in looking
at MCLGs, EPA assumes that water is being treated, which represents water-that people actually
drink. In the surface water program, the same value is used, but the 2-liter consumption rule
may not be appropriate since people do not tend to drink 2 liters of untreated water per day.
Thus, EPA will be reviewing the 2 liter water consumption factor for the AWQC.
EPA also will be looking at the bioconcentration factor, because currently, the
bioconcentration factor only looks at the accumulation in a water medium. However, PCBs and
other types of bioaccumulatives can accumulate in food materials. In addition, EPA also will
be revisiting the assumed fish consumption rate of 6.5 grams per day, which is considered
somewhat of a national average. However, in some areas, such as the Great Lakes or some of
the coastal areas, people may have higher consumption, or there may be areas of the country
where the consumption is actually lower. To give you an idea of where we might be going with
some of these methodology changes, EPA recently sponsored a workshop last September (1992)
and asked a number of experts to provide some indication of what the Agency should consider
in revising these standards. One of the biggest changes focuses on the issue of how to allocate
sources of exposure. In the drinking water program, sources are allocated to represent what
stems from drinking water versus other sources. Similar in the surface water program, the
bioaccumulation factor for drinking water will be reviewed. Specifically, EPA is developing
what is called a reference residue concentration, which is a value that represents a reference
dose, but subtracted or adjusted for sources of exposure other than fish consumption. Thus, the
reference dose would be adjusted for the weight of the protected individual, generally considered
to be an adult. However, for some contaminants, the protected individual also may be an infant
or a child. Other sources of exposure would be subtracted such as dietary intake from foods
other than fish. Inhalation may be a factor in some cases, particularly for volatile contaminants.
In this general methodology, there are no established numbers yet, e.g., the standard weight and
level of fish consumption have not been determined yet. Thus, EPA is looking into additional
information. Other factors, such as the dietary intake and the water mass, will all be specific
to the individual chemicals that will be reviewed. Once the residue concentration has been
developed, the water quality criteria would be developed by dividing with a bioaccumulation
factor.
For PCBs in drinking water, an MCLG of zero has been established. For both the
surface water and drinking water program, there are clauses within the Clean Water Act or the
Safe Drinking Water Act that indicate that, for contaminants that do not show a threshold of
effect or that may be carcinogenic, the level is set at zero. Ideally, the public does not want
carcinogenic substances in its drinking water, in fish, or in other materials that may be
consumed. This, however, is a somewhat unrealistic goal, because it may not be possible to
measure values that low or treat down that low. Thus, the MCL is set as close to zero as
feasible and is usually restricted to analytical methods availability or achievability. In the case
of PCBs, the MCL was set at 0.5 ug/L, and this was largely restricted to the analytical method
capability. In this particular standard, it represents the whole mixture of 209 possible congeners.
Again, this was restricted by available analytical methods, looking at somewhat of a fingerprint
1-55
-------
assessment in monitoring PCBs in drinking water. ;
i
On the ambient water side, ideally, the criteria for human health would be set at zero.
However, this would not be very helpful to the state standards program developing their
standards. For that reason, the cancer potency factor was used in trying to determine the criteria
for PCBs. Using the old 1980 methodology, the values that were derived for human health are
specific to the Aroclors 1016, 1021, 1232, 42, 48, 54, and 60, resulting in a value of 0.000044
ug/L, which includes consumption of water and organisms. In the organisms only value, also
based on the cancer potency factor, the value resulted in a slightly different value of 0.45 ug/L.
The aquatic life criteria for PCBs were developed for both fresh water and marine water,
although they were experimentally determined, the values have been reported for an acute
response in fresh water as 2 ug/L and a value of 0.014 ug/L for a chronic value. In the marine
systems, the risks do not appear to be quite as sensitive with an acute risk of 10 ug/L and a
chronic risk of 0.03 ug/L.
Back in 1980 when the original human health criteria for PCBs was determined, EPA
used a bioconcentration factor that was estimated from the log KQW, which resulted in a value
of 31,200. Since then, a bioaccumulation factor that focuses not just on water but also considers
accumulation in food materials and has been normalized to represent a 5 percent lipid content
in the trophic level remains to be determined. This will result in a considerable difference from
what was used in 1980 and results in a BAF value of roughly 1,700,000. For the wildlife
values, a slightly higher lipid content is used, and will result in a bioaccumulation value of
969,000 in trophic level 3 versus 2,800,000 in trophic level number 4. The Great Lakes
initiative package was just released for public comment not too long ago, and these are some of
the issues EPA will be accepting comments on to see how the Agency is going to change the
criteria and whether this particular methodology makes sense for the future. ,
Some of the other criteria that have been developed include recommendations from the
National Academy of Sciences. They were looking at water exposures only and developed what
is called suggested-no-adverse-response levels, which focus on the non-cancer health effects.
For a 1-day acute exposure for humans, they recommended that a value of 350 tjg/L would not
present an adverse health effect. A slightly lower value of 50 ug/L would be acceptable over
a 7-day exposure. NIOSH and OSHA also have developed some short-term values looking at
the inhalation route of exposure, again focusing on carcinogenicity as the main end point for the
NIOSH value. These agencies arrived at a reliable exposure level of 0.001 rhg/m2. OSHA
focused mainly on skin toxicity as their end point in determining their short-term effects and
ended up with values of 1 mg/m2 for chlorodiphenyls containing 42 percent chlorine, and for
those containing 54 percent chlorine, recommended a value of 0.5 ug/m2.
1-56
-------
Human Health Methodology
Current Surface Water Methodology
(RfD) or '(qt*) x 70 kg
(2 L/d) + BCF) x (6.5 g fish)
Current Drinking Water Methodology
RfD x 70 kg
MCLG
2 L/d
xRSC
= zero for non-threshold contaminants
Develop fish tissues residue and water column concentration criteria
RCC = (RfD x WT) - (DT + IN + WM)(WT)
FC
where: RRC
RfD
WT
DT
IN
WM
FC
WQC = RRC
BAF
where:
WQC
RRC
BAF
reference fish tissue residue concentration
reference dose
average human body weight
dietary intake other than fish
inhalation
water mass intake
fish consumption
water quality criteria
reference fish tissue concentration
bioaccumulation factor
1-57
-------
Drinking Water Criteria
For PCBs, EPA set an MCLG of zero based on carcinogenicity in animals.
MCL of 0.0005 mg/1 was set considering analytical methods.
10"4 to lO'6 excess cancer risk = 0.0005 to 0.000005 mg/1. ;
This regulation treats PCBs as a mixture of 209 possible congeners.
Ambient Water Quality Criteria (continued)
Aquatic Life Criteria for PCBs
Acute Fresh Water
Chronic Fresh Water
Acute Marine
Chronic Marine
Human Health Criteria
2ug/l
0.014 ug/1
10 ug/1
0.03 ug/1
(Aroclor 1016, 1021, 1232, 1242, 1248, 1254, 1260)
Water and Organisms 0.000044 ug/1
Organisms 0.000045 ug/1
1-58
-------
Bioconcentration/Bioaccumulation Factors for PCBs
Water Quality Criterion BCF for PCBs (U.S. EPA 1980)
Based on a predicted BCF estimated,from a log Kovv
Human Health BCF: 31,200
Great Lakes Initiative BAF for PCBs (U.S. EPA 1993)
Based on a measured BAF derived from field data for "total PCBs"
Human Health BAF at 5% lipid for Trophic Level 4: 1,776,860
Wildlife BAF at 7.6% lipid for Trophic Level 3: 969,239
Trophic Level 4: 2,807,439
Other PCB Criteria
National Academy of Sciences:
One Day SNARL = 350 ug/1;
Seven Day SNARL = 50 ug/1
NIOSH RELs: Chlorodiphenyl = 0.001 mg/m3 (carcinogenicity)
OSHA PELs and ACGIH TLVs:
Chlorodiphenyl (42% chlorine) = 1 mg/m3
(54% chlorine) = 0.5 mg/m3
(based on skin toxicity)
1-59
-------
1.7 SUMMARY OF QUESTIONS AND RESPONSES2
1.7.1 Dan Thomas of the Great Lakes Sport Fishing Council commented on the release of a
recent study that was conducted by the University of Wisconsin by Darr, Keneric, and
Anderson. He questioned why some of the information in this study was not given by the
presenters. :
Dr. Bolger indicated that the report had just been released and that there was not enough
time to incorporate the report into the slides. He described the study which observed the
reproductive outcome in women who were consumers of Great Lakes' fish. Dr. Bolger stated
that, contrary to the negative correlations that had been found in other studies, this study actually
showed a positive correlation between women who had consumed Great Lakes' fish and their
reproductive outcome. In other words, the study found that the higher the PCB body burden
in the mother, the heavier the child was. Dr. Bolger stated that this finding is therefore,
contrary to what has been reported in other epidemiological studies of populations in the Great
Lakes.
j
Dave De Vault stated that he believed that the results were actually not contrary to what
has been reported in other studies. He stated that in comparing the exposures in the recent study
and in some of the earlier Michigan studies, the "high exposures" in the recent study are
probably close to or perhaps even lower than the what was considered a "low exposure" dosage
in the previous Michigan studies. Mr. De Vault referred to the tissue concentrations in the fish
tissues. |
i
Dr. Bolger responded that the way fish consumption was estimated for some studies (such
as in the Jacobson study) made accurate correlations difficult. He stated that the most accurate
method to develop correlations is to compare the PCB levels in the blood with the levels in
maternal cord levels.
1.7.2 Dr. HeraUne Hicks of the Agency for Toxic Substances and Disease Registry (ATSDR)
in Atlanta, Georgia, asked Mr. De Vault whether he knew what source was contributing to the
increase in PCB atmospheric deposition in the Great Lakes region.
Mr. De Vault stated that EPA does not know the source at this point. He stated that the
apparent increase is in the southern Lake Michigan area. Since this is one of the most highly
industrialized areas in the country, there are numerous potential sources. EPA is attempting to
locate the source(s). !
Ed. note: The question and response portion includes summaries derived from transcribed conversations.
The summaries have been carefully edited to present the discussion as accurately as possible. However, these
question and response summaries have not been reviewed by the speakers—unlike the proceeding [abstracts.
1-60
-------
1.7.3 Dr. Heraline Hicks of the Agency for Toxic Substances and Disease Registry (ATSDR)
offered a comment to the panel and audience. She stated that ATSDR has recently initiated a
research program, mandated by Congress under the Great Lakes Critical Programs Act of 1990,
to look at human health impacts from pollutants in the Great Lakes area. For FY 92, ATSDR
received $2 million for the research program. For FY 93, ATSDR received $3 million and it
anticipated an additional $3 million in funding for FY 94. Thus far ATSDR has funded nine
research proposals in the Great Lakes area, which examine human epidemiological effects from
consumption of Great Lakes fish. She stated that seven of the eight Great Lakes States were
involved in this research program; five of those seven states represent a consortium of Great
Lakes State health departments. ATSDR studies encompass several issues (e.g., reproductive,
developmental end points, immunological end points, neuro-behavioral, and also long-term
health effects such as cancer). Dr. Hicks concluded by inviting other participants to see her for
additional Great Lakes information.
1.7.4 Jennifer Orme Zavaleta of EPA asked Dr. Hicks whether the ATSDR-sponsored studies
were measuring blood serum levels and adipose tissue in the subject cohorts.
Dr. Hicks confirmed that the studies would be examining adipose tissues as well as blood
serum levels in the individuals. She stated that the studies will be looking both established and
new population cohorts in the Great Lakes area.
1.7.5 Dr. Bolger asked Dr. Hicks if total PCBs or individual congeners will be studied in the
ATSDR-sponsored studies.
Dr. Hicks stated that ten persistent toxic substances of concern in the Great Lakes had
been identified. All nine grantees for FY 92 are examining PCBs as well as other persistent
toxic substances (e.g., mercury, lead, dioxin, furans, mirex, and others). She stated that the
grantees are looking at total PCBs as well as coplanar and non-coplanar PCBs.
1.7.6 Ram Tripathi from the Virginia Department of Health asked Jennifer Orme Zavaleta why
there are such discrepancies between the drinking water maximum contaminant level (MCL) and
the surface water ambient water quality criteria for PCBs.
Ms. Zavaleta stated that the difference can be attributed to the fish consumption factor
and the bioconcentration factor that were used in deriving the surface water value. For the
drinking water MCL value, although the goal for PCBs was set at zero, the standard was set
after considering analytical methodology. Thus, she continued, the main difference is because
of two different methodologies. For example, the surface water program incorporates
bioconcentration and fish consumption factors, whereas in the drinking water program, only
water consumption was examined and then adjusted for sources of exposure other than drinking
water.
1-61
-------
-------
PART TWO
PCB TOXICITY AND HEALTH EFFECTS
2.1 REGULATORY UPDATE: HUMAN CARCINOGENICITY EFFECTS
Jim Cogliano, Chief, Carcinogen Assessment Statistics and Epidemiology Branch,
Office of Research and Development, U.S. EPA*, Washington, DC
Studies on the potential carcinogenicity of PCBs are easier to describe than to interpret.
Aroclor 1260 causes liver carcinomas in three strains of rats. Aroclor 1254 causes liver tumors
in mice and rats; the tumor incidences appear to be lower than with Aroclor 1260. Less
chlorinated mixtures have not, in general, been adequately tested. Worker exposure to mixtures
with between 42 and 54 percent chlorine have been associated with some elevations in cancer,
but there has been no consistent overall pattern.
At the root of the difference of opinion about the potential carcinogenicity of PCBs is a
discussion of how to characterize the PCB mixtures that humans are exposed to. Alternative
ways of characterizing PCBs include: total PCBs; characterizing mixtures as being similar to
particular Aroclors; using percent chlorine as an index of toxicity; focusing on individual
congeners; considering classes of congeners, using toxicity equivalence factors based on dioxin-
like activity; or using toxicity equivalence factors based on other modes of action. Adequate
toxicity information does not exist for any of these characterizations of PCB exposure; hence,
substantial uncertainty will remain whatever measure is chosen.
In recent years new information is available to refine the cancer risk characterization of
PCBs. The rat liver pathology was reviewed by a pathology working group; its findings have
reduced the observed tumor incidences somewhat. A partial lifetime study in rats showed that
lifetime exposure is not necessary for the induction of precancerous lesions. Initiation-promotion
studies have been conducted for a few specific congeners; four congeners have shown promoting
activity, and one of these has also shown weak initiating activity. New epidemiologic studies
are available and have led to differing interpretations. EPA's Risk Assessment Forum
considered the question of developing toxicity equivalence factors for PCB congeners, but
concluded that there was more than one mechanism involved and that development of toxicity
equivalence factors would be less straightforward for PCBs than for dioxins and furans.
In recent years there has been a greater realization that the transformation of PCBs in the
environment is important to an accurate characterization of the risk to human health. The
congeners are affected differently, depending on the number and the position of the chlorines
* The views expressed in this paper are those of the author and do not necessarily reflect the views or policies of
the U.S. Environmental Protection Agency.
-------
in a PCB molecule. The transformation of PCBs in the environment has the consequence of
partitioning the original mixture so that substantially different fractions appear in air, water, soil,
and sediment. These fractions may be more or less chlorinated than the original mixture, with
a different distribution of congeners. Bioconcentration of PCBs in living organisms results in
the retention of higher-chlorinated fractions.
Thus, the uncertainty in cancer risk characterizations is likely to remain in the foreseeable
future. While toxicity information exists on some commercial PCB mixtures and on a few
specific congeners, human exposure is not to these mixtures, but to residual mixtures that are
weathered by years of environmental transformation and bioaccumulation.
2-2
-------
Cancer Data on PCBs
Aroclor 1260 causes liver carcinomas in 3 rat strains
Aroclor 1254 causes liver tumors in mice and rats, apparently is less potent
Clophen A30 (a 42% mixture) causes liver tumors in rats
Worker exposure to 42-54% mixtures associated with some elevations in
cancer, but no consistent pattern
Carcinogenicity depends somehow on congeners
Current PCB Risk Assessment
PCBs are classified as probable human carcinogens (Group B2) - available
on IRIS
7.7 per mg/kg-d average lifetime exposure is a plausible upper bound for the
increased cancer risk
2-3
-------
Criticisms of the Current PCB Risk Assessment
PCBs differ widely, but one assessment applies to all
Common criticisms: choice of data set, inclusion of benign tuinors, use of
surface-area scaling, upper bounds, ...
Some prefer a cancer assessment for each Aroclor
"Only 1260 is carcinogenic, others have no risk"
"Unit risk from less-chlorinated mixtures < 7.7"
"All studies together show unit risk from 1260 <7.7"
Some have advocated a TEF approach, based on congeners i
Some assessments are based on what had been released; others, on what is
present now i
What New Carcinogenicity Data are Available?
GE/IEHR review of rat liver pathology
Tumor incidences reduced slightly
Statistical significance for 60% mixtures only :
New initiation-promotion studies on specific congeners
Promoting activity with 2,2',4,5'-TCB, 3,3',4,4'-TCB; 2,3,4,4',5-
PCB, and 2,2',4,4',5,5'-HCB
Weak initiating activity with 2,2',4,5'-TCB
Partial-life study of Aroclor 1260 in rats '
New epidemiologic studies
2-4
-------
Uncertainty in Quantitative Risk Estimates
• Toxicity data are on commercial PCBs and some congeners
Less-chlorinated mixtures appear to be less potent
Less-chlorinated congeners can be strong promoters
• Exposure is to residual PCBs weathered by years of environmental
transformation and bioaccumulation
Not what had been released
Not what has been tested
Some congeners are enhanced 100-fold
• PCB risk are expected to vary, but we can't say by how much
2-5
-------
2.2 REGULATORY UPDATE: NON-CARCINOGENIC EFFECTS
John L. Cicmanec, Veterinary Medical Officer, Systemic Toxicants Branch,
Environmental Criteria Assessment Office, Office of Research and Development,
U.S. EPA, Washington, DC
The health effects of Aroclors, commercial PCB mixtures that were produced! in the United
states, have been extensively tested in a wide variety of laboratory animals species. The
evaluation of the health effects is complicated by the fact that each commercial mixture is made
up of many congeners and ultimately the adverse health effects of a mixture depends upon the
toxicity of the individual congeners of that mixture. To date, only limited information is
available for toxicity studies of individual congeners. I have chosen to present this summary of
the laboratory studies on the basis of individual target organs with some comparisons being made
between species and between various commercial mixtures. More than 50 reports of animal
studies of PCBs are added to the literature each year so this might provide an indication of the
volume of material that is being summarized. ;
General Acute Toxicity
This information is presented only to give a comparative overview for different commercial PCB
mixtures and for different animal species for which LD-50 data are available.
Rat
Rat
Mink
Rhesus monkeys
1,010 mg/kg
4,250 mg/kg
750-1,000 mg/kg
4 mg/kg
60 days
Aroclor 1254
Aroclor 1242
Aroclor 1221
Aroclor 1248
Gastrointestinal Effects
The ability of PCB mixtures to induce gastric ulcers is fairly well known. Studies in domestic
swine by Hansen indicate that doses of 100 mg/kg-day of Aroclor 1254 caused gastric ulcers
within 11 days of treatment. Studies with rhesus monkeys and pig-tailed macaques have shown
that 4 mg/kg-day of Aroclor 1248 caused gastric ulcers within 2 months in four subjects and
Becker reported induction of gastric ulcers with Aroclor 1242 at doses of 10 and|30 mg/kg-day.
Studies with mink have shown consistent induction of gastric ulcers with Aroclor 1016, 1242,
and 1254 within 90 days at doses lower than 50 mg/kg-day.
Hematopoietic Effects
I
I
Regarding adverse lexicological effects upon the blood-forming organs, consistent changes have
been observed in macaques but not in rodents or mink. Anemia induced by Aroclor 1248 and
2-6
-------
Aroclor 1254 is reported to occur at doses as low as 4 mg/kg-day and abnormal changes are
observed as early as 60 days. These changes are characterized by up to 20% reduction in
hemoglobin concentrations and packed cell volumes. Increases in reticulocyte counts are also
observed. In a chronic study with rhesus monkeys, anemia was observed at a dose level of 0.2
mg/kg-day with Aroclor 1254. In this study, reported by Arnold, changes were first seen at this
low dose 12 months into the study and these changes persisted to the end of the study at 28
months. In addition to anemia, hematologic changes that have been observed in long-term
studies in rats and monkeys generally show an increase in total neutrophil counts and a decrease
in lymphocyte counts.
Hepatic Effects
Perhaps the most consistent, and surely most thoroughly documented, toxic manifestations of
PCB administration are for the liver. One of the most characteristic features of PCBs is their
ability to induce hepatic microsomal enzymes (cytochrome P-450 of various groups) and it is
likely that the development of more serious liver changes are an extension of this process. It
has been postulated that the sequence of cellular change progresses from microsomal enzyme
induction to hepatocellular damage to grossly observed hepatic enlargement to lipid deposition
to fibrosis and finally focal hepatic necrosis. These microscopic changes are accompanied by
changes in serum chemistry. Alterations in blood cholesterol levels (both increases and
decreases) and vitamin A metabolism have been documented. Examples of significant liver
changes involve both short-term and chronic studies. In a study reported by Buckner, Aroclor
1254 administered at the dose of 1 mg/kg-day induced elevated serum cholesterol levels within
4 days. When studies were extended to 3-8 weeks duration, an increase in total hepatic lipids
with a decrease in hepatic cholesterol and a decrease in hepatic vitamin A concentration was
observed. The lowest concentration that significant hepatic effects have been noted in rats for
PCBs is 0.3 mg/kg. A similar pattern of microscopic liver lesions have been observed in rabbits
dosed with Aroclor 1254 for 8 weeks at 2.1 mg/kg.
A pattern that will be emphasized throughout this presentation is that non-human primates
are more sensitive to the toxic effects of PCBs that rodents and other conventional laboratory
species. Effects upon the liver provide ample evidence of this. Perhaps most striking are the
results reported by Allen and Barsotti in which two female rhesus monkeys that were dosed with
Aroclor 1248 at 0.1 mg/kg died after 173 days and 310 days of the study. At necropsy, the
most significant lesions noted were fatty accumulation within the liver and extensive focal
necrosis. Similar hepatic effects were reported by Tryphonas in which rhesus monkeys showed
fatty degeneration of the liver and focal necrosis as well as hypertrophic and hyperplastic
changes.
Dermal and Ocular Effects
Changes noted in the skin and structures adjacent to the eye are among the most significant
changes associated with PCBs. These, lesions are of particular importance because the effects
noted in animals, particularly non-human primates, have been directly associated with effects
2-7
-------
directly attributable to PCBs noted in occupational exposures of humans. Studies with Aroclor
1248 and 1254 with rhesus monkeys provide the most detailed reports of the dermal changes.
Specifically, facial edema most commonly observed in periorbital regions is accompanied by a
purulent ocular discharge, swelling of the Meibomian glands, regional alopecia, and folliculitis
of dermal sebaceous glands. If exposure persists, loss of fingernails occurs as well as gingival
hyperplasia and dermal effects can progress to regional necrosis of the skin. These changes are
consistently observed in Old World monkeys dosed with Aroclor 1248, 1242, and; 1254 at levels
as low as 0.1 mg/kg. The changes are observed as early as 2 months. Once dosing stops these
effects are reversible. «
Although dermal effects are not as thoroughly documented in rodent studies, erythema
and altered sebaceous gland function has been observed in mice treated with Aroclor 1254 at
doses of 26 mg/kg. Dermal and ocular effects observed in rats treated with Aroclor 1254
include alopecia, facial edema and exophthalmos.
Thyroid Effects :
Initial reports of the effect of PCBs upon the rat thyroid were made by Byrne in which SD rats
receiving 2.5 mg/kg (a relatively low "rodent" dose) of Aroclor 1254 exhibited reduced T-4
serum levels accompanied by enlargement of the thyroid, folHcular cell hyperplasia, and the
accumulation of colloid droplets. These changes were observed as early as 7 days of treatment
and a subsequent study in rats noted similar changes at a dose of 0.09 mg/kg after 35 days of
treatment. Similar changes have been noted for rhesus monkeys that received 0.2 mg/kg of
Aroclor 1254 that were examined after 28 months of treatment. The authors believe that these
effects may be reversible.
Adrenal Effects
The most extensive amount of data regarding the effects of PCBs upon the adrenal gland are
available from studies with rats. Both Aroclor 1248 and 1254 cause an increase in
corticosterone levels, the most biologically active adrenal steroid in rats. These changes were
noted after 20 to 70 days at relatively low doses of 15 and 35 mg/kg. The effects upon the
adrenal gland are felt to be particularly significant because chronic studies in which much lower
doses (0.05 mg/kg) of two lesser chlorinated PCBs Aroclor 1242, and 1221 caused similar
changes. No adrenal gland effects were observed in studies with rhesus monkeys that were
dosed with Aroclor 1254 for periods exceeding 22 months. Adrenal gland histology was normal,
and normal blood steroid levels were noted during the study.
Generalized Effects
A pronounced reduction in body weight for long term studies has been observed in rats, mink,
pigs, and rhesus monkeys treated with PCBs. Some investigators have described the effect as
a wasting syndrome in which for studies lasting more than 1 year the lack of body weight gain
when compared to control animals ranged from 15 to 20%.
2-8
-------
Immunologic Effects
The initial reports of immunologic effects of PCBs were made by Loose (in mice) and Thomas
and Hinsdale (for mice and rhesus monkeys) in 1978. The reports by Loose described a
reduction in immunoglobulins of BALB/c mice that had received 22 mg/kg of Aroclor 1242 for
3-6 weeks. Perhaps greater biological significance can be given to the results of increased
mortality among PCB-treated mice that were challenged with Salmonella endotoxin and the
malarial parasite, Plasmodium berghei. Earlier studies in rhesus monkeys showed a reduction
in the antibody response to sheep RBCs when the monkeys had been treated with Aroclor 1248
(0.2 mg/kg) for 11 months. Interestingly, the response to tetanus toxoid was not altered by PCB
treatment. A consistent response has been observed in all rat studies in which immunologic end
points were observed. There is a significant reduction in the weight of the thymus gland when
rats are treated with Aroclors 1248, 1254, or 1260 for 6 weeks. The most extensive study of
the immunologic effects of Aroclor 1254 has been recently reported by Tryphonas, et al. For
monkeys receiving 5 to 80 micrograms/kg after 23 to 55 weeks a reduction in IgG and Ig M
specific for Sheep RBCs was noted as well as a reduction in T-helper lymphocytes. However,
there was not a significant reduction in titers to pneumococcus vaccine in treated monkeys when
compared to the control group and the monkeys did not show evidence of microbial infection
at any time during the study. The authors describe these changes collectively as
immunomodulation.
Neurologic Effects
Through quite important I will discuss only briefly the neurologic effects in monkeys described
by Seegal's group in New York in which reduced concentrations of dopamine, an important
neurotransmitter in selected regions of the cerebrum. It is quite interesting that these changes
were associated with less-chlorinated congeners of PCBs and the changes were noted a long time
after dosing had been completed.
Developmental Effects
Perhaps some of the most significant adverse effects of PCBs are developmental changes
reported in rhesus monkeys. Barsotti, Allen, and others have reported the occurrence of reduced
birth weights in infants born to dams treated with Aroclor 1248. In addition to the smaller body
size, the infants also exhibited facial acne, swollen eyelids, the loss of eyelashes, and skin
hyperpigmentation. Rhesus monkey dams that received Aroclor 1016, a lesser chlorinated PCB,
also gave birth to smaller infants but the only clinical signs these infants exhibited was a skin
discoloration at the hairline of the face. Once the infants were weaned, these clinical signs
disappeared. The Lowest Observed Adverse Effect doses administered to the dams were 0.03
mg/kg-day of Ar 1016 and 0.1 mg/kg of Aroclor 1248. Some of the affected infants died at 44
to 329 days, of age and lesions noted at post mortem examination were small spleens,
rudimentary thymuses, and hypocellular bone marrow. To demonstrate the effect of residual
PCBs in the tissue of rhesus monkey dams, the females were rebred 6 months after PCB dosing
had ceased and the second group of infants also exhibited reduced birth weights and skin
2-9
-------
t
discoloration as well as thymic and splenic atrophy and hyperplastic gastritis. ,
i
Realizing the significance of the general toxicity exhibited in the infants, investigators at
the University of Wisconsin performed neurobehavioral tests upon the infants at the age of 14
months and 4 and 6 years. The infants from mothers treated with Aroclor 1248 showed
decreased performance in discrimination learning tasks. These deficits are indicative of residual
toxicity since tissue concentration of PCBs had returned to concentrations similar to control
monkeys when the testing was performed. Infants from mothers treated with Aroclor 1016
showed normal performance on spatial learning and memory tasks but exhibited impaired
learning of spatial discrimination tasks.
Developmental effects have also been reported for rats and mink.
i
Reproductive Effects I
Effects of Aroclor 1254 upon male reproductive capacity in rats has been described by Sager in
which young males exposed during lactation had a significant reduced ability to fertilize
untreated female rats once they reached maturity. There was no detectable change noted in the
sperm except the impaired ability to fertilize ova. One male rhesus monkey of 4 that received
4 mg/kg of Aroclor 1248 showed abnormal testicular histology, had hypoactive seminiferous
tubules, and exhibited reduced libido. i
The effects of PCBs upon female reproductive capacity have been more thoroughly
documented. Studies in mink show that 0.4 mg/kg Aroclor 1254 and 0.9 mg/kg of Aroclor
1248 had a significant reduction on female reproductive performance. In rhesus monkeys 0.1
mg/kg of Ar 1248 cause increased menstrual duration and bleeding as well as reduced
conception rates. The dose of 0.2 mg/kg of Aroclor 1254 caused erratic menstrual cycles and
abortions in female rhesus monkeys.
2-10
-------
Comparative Effects other Organs
Liver, Stomach, and Thyroid
Organ
Liver
Stomach
ulcers
Thyroid
Comparison
Rat
0.3 mg/kg 2 mo
Pig
100 mg/kg
Rat
2.5 mg/kg
Rhesus monkey
0.1 mg/kg 7 m
Rhesus monkey
4 mg/kg
Rhesus monkey
0.2 mg/kg
Salient Clinical Effects
Long Term Monkey Studies
Facial edema
Ocular discharge
Swelling Meibomian glands
Loss of fingernails
Gingival hyperplasia
Regional necrosis
Alopecia
Weight loss 15%
2-11
-------
Comparative LOAELs for Reproductive Effects
Mink vs. Rhesus Monkeys
Mixture
Aroclor 1016
Aroclor 1248
Aroclor 1254
Mink
0.9 mg/kg
0.4 mg/kg
Rhesus Monkeys
0.03 mg/kg
0.1 mg/kg :
0.025 mg/kg
Female Reproductive Alterations
Rhesus Monkeys Dosed with Aroclor 1248
Alterations
Anovulation
Decreased estrogen
Altered Function of
Corpus Luteum
Altered Length of
Menstrual Cycle
Altered Day of Ovulation
0.1 mg/kg
2/8
5/8
2/8
2/8
3/8
0.2 mg/kg
5/7
3/7
5/7
2/7
1/7
2-12
-------
Summary
Effects of PCBs in Laboratory Animals
PCBs produce a toxicity in many organ systems including reproductive,
immunologic, integument, CNS, thyroid, adrenal, and liver
Many species are affected with varying degrees of severity
In general, more highly chlorinated PCB mixtures cause more severe
toxicity
Lesions in the CNS are associated with lesser chlorinated congeners
2-13
-------
2.3 TOXICITY EQUIVALENTS FOR PCBS
Donald G. Barnes, Staff Director, Science Advisory Board, Washington, DC*
i
Over the past twenty years, analytical chemistry has provided toxicologists with a wealth of
information about the molecular identity of compounds in the environment. The wealth of
information has proved to be a significant interpretative challenge as toxicologists have tried to
use this information to make better, more defensible decisions about the likely risks posed by
these compounds—individually and as mixtures.
i
In the 1980s, an increasing number of laboratories were able to analyze environmental
mixtures of chlorinated dibenzo-p-dioxins and chlorinated dibenzofurans (CCDs/CDFs) on a
homologue-specific, isomer-specific, and, hence, congener-specific basis.3 Environmental
samples of CCD/CDF mixtures seemed to be popping up all over the place; e.g., residues in fish
tissues, emissions from municipal waste incinerators, and smoke smudges from electrical
equipment fires. At that time, considerable toxicological attention had been focused on the
2,3,7,8-TCDD congener.4 However, relatively little was known about the toxic properties of
the other 209 members of the structurally related CDD/CDF family. |
Given this situation—congener-specific information about CDDs/CDFs in the
environment—little toxicological information was left with two equally unappealing positions:
%
1. Ignore the toxic potential of all congeners in the CDDs/CDFs family other | than 2,3,7,8-
TCDD: the "Ignorance is Bliss" strategy, or I
2. Assume that all of the congeners in the CDDs/CDFs family are equally toxic to 2,3,7,8-
TCDD: The "It Could Be" strategy.
* The views expressed in this paper are those of the author and do not necessarily reflect the views or policies of
the U.S. Environmental Protection Agency.
3 Homologue: The group of CDDs (or CDFs) with the same level of chlorination. There are 8 homologous
groups of CDDs (or CDFs), ranging from 1 chlorine substitution to 8 chlorine substitutions.
I
Isomen The group compounds withina single homologous group (i.e., having the same degree of chlorination)
that differ in the spacial distribution of those chlorines around the CDD (or CDF) three-ring structure of carbon and
oxygen atoms. For example, there are 22 isomers in the homologous group of tetrachlorodibenzo-p-dioxins
(TCDDs) and 27 isomers in the homologous group of tetrachlorodibenzofurans (TCDFs). The most (in) famous
and most toxicologically potent of these is 2,3,7,8-TCDD. \
Congener: A specific isomer of a specific homologous group. In all, there are 72 CDDs congeners; i.e., the
sum of all the isomers of each of the 8 homologous groups.
I
4 A Federal workgroup estimated that, by 1983, nearly $1 billion of research had been directed |at the toxicology
of 2,3,7,8-TCDD.
2-14
-------
Dr. Judith Bellin, a toxicologist in the Office of Solid Waste at the time, and I tabled out
the limited toxicity information on the CDDs/CDFs and saw arguably consistent patterns of
relative toxicity, among the different congeners, across the different toxic endpoints. This
observation—expanded (by the Risk Assessment Forum), critiqued (by everybody and his/her
brother/sister), and peer reviewed (by the SAB)—led to the Agency's adopting the TEF (relative
toxicity) scheme as an interim procedure5 for assessing the risks of exposures to mixtures of
CDDs/CDFs.
Two years after its initial adoption by the Agency, the TEF scheme was reassessed, and
modest changes made. The new scheme was subsequently adopted by most of the Western
nations through a workgroup of the North Atlantic Treaty Organization (NATO).6
The TEF scheme for CDDs/CDFs has been employed to good effect in a variety of
situations over the past four years. While there is some interest is making additional
modifications to the scheme, most of those involved in using TEFs in decisionmaking find
greater benefit from a stable, widely accepted set of TEFs than in a marginally more "scientific"
set of TEFs that are changing on an irregular, uncoordinated basis in different countries.
Criteria for TEFs
In the late 1980s, congener-specific analyses for PCBs became a viable possibility in a number
of labs around the world. In addition, there was a growing body of lexicological information
linking the toxicity of PCBs to molecular structure. Consequently, the stage seemed set for the
appearance of another set of TEFs—this time for PCBs—to provide a rational interim procedure
for assessing risks posed by another set of chemically related compounds.
Again, Dr. Judy Bellin was involved in the initial stages of the effort. But this time the
data were even less complete than in the case of the CDDs/CDFs, and the story they told was
less clear even to a sympathetic reader.
Before we launched into the nitty-gritty of trying to generate TEFs for PCBs we decided
upon a set of criteria to guide our thinking and to keep us honest, especially for when our blood
rushed hot from the thrill of the hunt. Those criteria are found in Table I. Based on that initial
exploration of the issue, we decided that the PCBs fell short of meeting the requisite criteria.
5 The preferred approach was identified as- a bioassay that would test the toxicity of the environmental mixture,
or extract therefrom.
6 It can now be disclosed that the TEFs for CDDs/CDFs were not the West's secret weapon that lead to the fall
of the Berlin wall. However, it is true that OEM is seeking additional public comment on the TEF approach, in
the context of some rules that EPA is planning to issue.
2-15
-------
However, by December of 1990, additional data had been developed inside and outside
of EPA laboratories that suggested revisiting the TEFs for PCBs issue. Consequently, the Risk
Assessment Forum (RAF) sponsored a public workshop that drew participants from as far away
as FDA (5 blocks), Canada, and Sweden. That group was more sanguine about the possibility
of TEFs, particularly for the coplanar PCBs that exhibited "dioxin"-like toxicity, albeit at higher
doses (RAF purple book; peer reviewed article in Quality Assurance); cf. Table II.
Current Status and the Future 1
f
The Agency is nearing the end of a multi-year effort to reassess the toxicity of 2,3,7,8-TCDD.
This is a very complex and controversial project. It includes reassessing old endpoints (cancer
and reproductive effects), and introducing new mathematical risk models. This stout stew will
be further fortified by consideration of the toxicity of "dioxin"-like PCBs and the possibility of
associated TEFs. New data—both health and ecological—are being generated in this enterprise,
and new policies are likely to emerge as those data are examined. The data and the policy will
be subjected to scrutiny by the public and to peer review by the SAB later this year.
In short, the use of TEFs for PCBs is not a "gimme". The concept will have to prove
itself in the crucible of scientific review. In any event, the risk management implications are
likely to be significant. 1993 should prove to be an interesting year.
2-16
-------
1.
2.
3.
4.
5.
6.
7.
TABLE I
Criteria for a TEF Scheme
A need (e.g., an interim regulatory need) for such an approach should be demonstrated.
The set of chemicals to which the scheme will be applied should be well defined.
A broad base of toxicity data should be available, covering many endpoints for many congeners.
Relative toxicity among the different congeners should be generally consistent across many
different endpoints (in vivo and in vitro),
General additivity should be demonstrated in the response to simple mixtures of congeners.
A common mechanism should rationalize the observed SAR results.
A mechanism should be available for gaining widespread consensus on the TEF values.
TABLE H
TEFsfor CDDs/CDFs and PCBs Evaluated by Applicability Criteria (as Defined in Text)
CDDs/CDFs
PCBs
1. Need for TEFs
2. Well-defined group
3. Broad database
4. Consistency across toxic endpoints
5. Additivity of toxic response
6. Common mechanism
7. Mechanism for consensus
+ +
+ + +
+ +, improving
+ +
+/-, improving
+ +, for "dioxin"-like
endpoints
+ +, for "dioxin"-like
congeners?, for other
congeners
+ (potentially)
Note. +, generally meets criterion, based on available data; + + meets criterion well, based on available data; -
does not meet criterion, based on available data; ?, cannot evaluate, based on available data.
Source: Quality Assurance, 1(1991), 70-81, "TEFs for PCBs?," Barnes, D.G., et al.
2-17
-------
2.4 EFFECTS OF OCCUPATIONAL EXPOSURE
John F. Brown, Jr., Manager of Health Research, General Electric Corporate
Research and Development, Schenectady, NY7
During the period 1929-1978 about 1.3 billion pounds of PCBs were produced arid used in the
United States. Some 10-15 percent of this total is estimated to have reached the soils,
sediments, and waters of the general environment, resulting in widespread, though low level,
exposure of fish, other wildlife, and human populations. In addition, direct human exposure to
undiluted PCBs occurred in many PCB-using operations. During the 1970s and 1980s about 20
different clinical or epidemiological studies of the dozen-odd more easily identified PCB-exposed
worker populations, including our own longitudinal clinical study of GE capacitor workers [1],
were undertaken. These studies have resulted in the publication of several dozen reports in the
scientific literature, and several critical reviews [2]. The most recent, and most detailed, of
these reviews is that by James et al. [2], which addressed, in turn, the questions of possible
effects on the liver, lungs, skin, cardiovascular system, nervous system, certain endocrine
systems, the blood/immune system, and the GI and urinary tracts. It was concluded, in line with
the findings of previous reviews, that there was no evidence for adverse PCS effects on any
organ system other than the skin, for which there was ambiguous evidence of possible effects
at the highest exposure levels. (Parenthetically, we may note that the ambiguity arises because
none of the original investigators who reported occasional observations of "chloracrie" undertook
bistological or statistical examinations to distinguish the observed lesions from those of common
acne. No cases of chloracne were observed in the GE capacitor worker population [1] that we
studied, despite the presence of many individuals with serum PCB levels over 1000 jug/kg). The
reports reviewed by James et al. [2] included ten mortality analyses, covering about 1000
different deaths, which generally showed no increased mortality due to cardiovascular disease,
pulmonary disease, or cancer.
The extraordinary contrast between the absence of clinically- or epidemiologically-
demonstrable health effects in heavily-exposed worker groups and the numerous reports of
lexicological and carcinogenic effects in laboratory animals has been noted by several reviewers;
however, there has remained some uncertainty as to whether these differences arise from
interspecies differences in susceptibility to PCBs or to differences in dosage or accumulation [2].
Also unresolved has been the question of why statistical correlations between health
abnormalities and PCB levels may exist even in population groups carrying PCBs at only
background levels. :
In hopes of shedding light on such issues, we have been examining PCB metabolism and
pharmacokinetics in our capacitor worker study group, and will summarize some recent findings
here.
Richard W. Lawton was a coauthor of this paper.
2-18
-------
PCB Metabolism
The congener distribution in any partially metabolized PCB residue provides a highly
characteristic indicator of the PCB-metabolizing system(s) present in the host organism or
culture. The PCB residue in all occupational exposed workers thus far examined (> 200) show
only a congener distribution pattern like that produced by cytochrome P4502B (i.e.,
phenobarbital-induced) isozymes [3-4]. This is the pattern also exhibited by the chromatograms
of PCB-dosed sheep and mice, as well as those of many wild birds, a few fish species, and some
crustaceans.
Conversely, the pattern seen in the Yusho and Yucheng chloracne patients, who were
poisoned by PCDF + PCB mixtures, indicates PCB metabolism by agents with selectivities like
those of cytochrome P4501A (i.e., PCDF-induced) as well as by P4502B isozymes [3]. This
is the pattern also seen in most PCB-dosed rats and probably in adult monkeys, as well as in
many species of wild birds and fish. The rats and monkeys in which indications of P4501A
activity appeared carried tissue PCB concentrations that were the same or lower than those of
our capacitor workers, showing that the absence of a dioxin-like response to PCB in the human
arises from a physiological difference in the species, not one of lower dose.
PCB Kinetics
Following the 1977 cessation of PCB use in capacitor manufacturing, serum PCB levels in our
study group dropped rapidly. The decay in PCB levels occurred at rates that could be correlated
with various physiological characteristics of the individuals involved, including age, sex, body
fat, and serum iron, and followed approximately second order kinetics, indicating that the levels
of the PCB-metabolizing P4502B-like isozymes must have declined roughly proportionally with
those of the PCBs. Earlier indications of PCB-mediated induction of P4502B-like isozymes in
capacitor workers were suggested by data on antipyrine clearance times. PCBs have also been
observed to increase P4502B levels in rats, mice, and winter flounder, although apparently not
in the rhesus monkey. Thus, the human would appear to be at least qualitatively similar to the
rodents in this P4502B-induction response to PCBs.
In rodents, however, P4502B-inducing agents, such as the more heavily chlorinated
PCBs, DDT, other chlorinated pesticides, and the barbiturate and hydantoin drugs, all appear
to be either hepatotumorigenic or liver tumor promoters, whereas none of these agents have been
found to be human carcinogens. Thus, there would appear to be a clear interspecies difference
between rat and man in the sequelae of hepatic P4502B induction.
2-19
-------
PCB Levels in Chronically Exposed Individuals
I
Most PCS congeners are so rapidly metabolized by human P4502B as to be undetectable in the
serum. A few are scarcely metabolized at all, and hence can serve as permanent records of PCB
exposure events [4]. In between are the dozen or so congeners that account for most of a
measured human PCB level, and which have half-lives in the 1-15 year range [3]. In adults who
have been exposed for several years to a single PCB source, whether environmental or
occupational, body PCB levels of those dominant congeners will have approached a steady state
level, where uptake is approximately balanced by metabolism. Simple calculation shows that
in such individuals the serum PCB level will be as indicated by eq. (1) i
[PCB]S = [L]s(D/EF)1/n
(1)
where [PCB], and [L]s respectively represent the levels of PCBs and neutral lipids in the blood
serum; D the PCB intake rate; E, the inducibility of P4502B; F, the total level of neutral lipid
(i.e., fat) in the body; and n, a number ranging between 1 (in the lightly exposed background
population) and 2 (in heavily PCB-exposed and P4502B-induced individuals). !
What this relationship shows is that among ordinary, background-exposed individuals,
where n = 1, and D values are probably similar to those for everyone else in the same region,
observable variations in serum PCB will arise primarily from variations in E, F, and [L]g, and
hence that serum PCB levels will be covariant with any health conditions that are covariant with
P450 inducibility, body fat, or lipidemia. Early failures to recognize such relationships in
occupationally exposed workers led to several reports that serum PCBs were causally associated
with elevated serum lipids and the associated serum enzymes. The statistical associations
disappeared, however, when the serum PCB levels were corrected for variations in [L]s [1,2].
More recently, much weaker statistical associations with fetal neurodevelopmental deficiencies
have been reported for PCB levels in background-exposed promoters; however? no corrections
of the PCB levels for variations in E, F, and [L]s have yet been carried out, so the hypothesis
that the developing fetus may represent an organ system that is orders of magnitude more
sensitive to PCBs than any other in the human body (which is clearly had not the case for the
dioxin-like PCDFs) remains unproved.
Relative Risks of Different PCB Compositions
Long before the completion and review of clinical and epidemiological studies indicating absence
of significant health effects in occupationally exposed human populations, it was recognized by
the FDA that there existed no scientifically valid basis for the quantitative assessment of PCB
health risks, and hence that any determination of tolerance levels would have to be made instead
on the basis of administrative authority. This, however, leaves unresolved the question of how
the relative "risks" (if any) of different PCB compositions should be assigned. Currently, the
assessed risks of all PCB compositions are regarded as equal, a presumption widely regarded
as scientifically implausible, albeit administratively convenient. '
2-20
-------
One briefly considered alternative, discussed elsewhere in this document has been to scale
presumptions as to health risk on the basis of dioxin equivalency, a measure that may indeed be
a plausible indicator of relative risk to some species of wildlife. As a measure of cancer risk
to humans, however, it suffers the drawbacks (a) that PCBs do not seem to have appreciable
dioxin-like activity in humans, as we have just seen, and (b), even if rats, where PCB
compositions do have dioxin-like activity, that activity does not correlate with tumorigenic
response.
An alternative suggested by the recent availability of metabolic rate data for the PCB
congeners that are commonly detected in humans [3,4] would be to use the persistence, or
accumulability, of the PCB composition as a measure of relative risk. This parameter, which
is readily calculated from the rate data, does seem to track reported tumorigenicity in rats and
immunotoxicity in mice, and probably also P4502B induction, which would currently appear to
be the only demonstrable human response to elevated PCB loadings. Calculations show that the
relative accumulations of the various commercial PCB compositions in a human exposed over
a 70-year lifetime, would be: for Aroclor 1016, 0.026; for Aroclor 1242, 0.049; Aroclor 1248,
0.10; Aroclor 1254, 0.31; Aroclor 1260, 1.000; Aroclor 1262, 1.26; and Aroclor 1268, 2.32.
Thus, if present regulatory presumptions as to the theoretical human health risk posed by
Aroclor 1260 (which is calculated from the hepatocarcinogenicity of Aroclor 1260 in rats) were
to remain unchanged, but those of the other Aroclors assessed on the basis of relative
accumulation tendency, their measured levels would have to be multiplied by these factors in
order to convert them to Aroclor 1260 equivalents. Application of such Aroclor 1260
equivalency factors for the various individual congeners or gas chromatographic peaks to
available data on the PCB congener distributions in fish suggests that such an approach would
often result in a two- or four-fold relaxation of PCB tolerance levels. Not so affected, however,
would be fish those from sites contaminated with Aroclors 1260, 1262, or 1268, or those of
species such as eels, lobsters, and blue crabs, where considerable metabolism of the PCB
residues has already occurred.
Conclusion
The metabolic and pharmacokinetic behavior of PCBs in capacitor workers supports earlier
conclusions that humans differ from rats and monkeys in their response to PCBs, just as various
animal species differ from each other. At present, the only unequivocally demonstrable
pharmacological response of humans to PCBs at levels produced by direct occupational
exposure, which are 10-100 times greater than those produced by fish consumption, has been
induction of metabolic enzymes having activity profiles like that of cytochrome P4502B. There
is no evidence that such enzyme induction would occur at lower levels of exposure, nor is there
any clinical or epidemiological data to indicate that this or any other pharmacological response
has had deleterious effects on the health of the occupationally exposed individuals.
2-21
-------
However, if concerns remain that there may still be real, though immeasurable, health
risks associated with any agent having a pharmacological activity in the human., then it would
seem appropriate to regulate different PCB compositions according to their ability to accumulate
to levels that would produce such a response. Institution of such an approach (e.g., regulation
of fish on their content of "Aroclor 1260 equivalents" rather than that of "total PCBs") would
not require any new analytical methods, but merely the incorporation of available kinetic data
into the mathematical procedures used for computing a reportable parameters from the same raw
data.
References
i
1. Lawton, R.W., Ross, M.R., Feingold, J., and Brown, J.F. Jr. 1985. "Effects of PCB
exposure on biochemical and hematological findings in capacitor workers." Environ.
Health Perspect. 50:165-184. I
2. James, R.C., Busch, H., Tamburro, C.H., Roberts, S.M., Schell, J.D.j and Harbison,
R.D. 1993. "Polychlorinated biphenyl exposure and human disease." /. Occup. Med.
55:136-148. :
3. Brown, J.F. Jr., Lawton, R.W., Ross, M.R., Wagner, R.E., and Hamilton, S.B. 1989.
"Persistence of PCB congeners in capacitor workers and Yusho patients." Chemosphere
JP:829-834. !
4. Brown, J.F. Jr., J^awton, R.W., Ross, M.R., and Wagner, R.E. 1990. "Serum PCB
as a permanent record of PCB exposure and responses." In Addis, G. ed., Proceedings:
1989 EPRI PCB Seminar, Electric Power Research Institute, Palo Alto, CA, pp. 9-49-1
through 9-49-5. i
2-22
-------
U)
CO
5
o
DC
s
Q
HI
UJ
III
OC
CO
2
UJ
fe
CO
o
cc
o
(0
3
n.
CO
o
=
Jtf
co
o
0)
CO
ffi
s
I?
s .1 g
O
E (O O
.i § -g
^ .O U
§11
m o uj
Q
jj
LU
QC
o
c
I
J
-a
co
QC
O
Si
1
D.
1
o
•—
1 i
•
•— •- oe «r.
.
«—:?_«>
8 "
Ew G.
W5 ^™"
.
•-"
I
1
I
I
5 £
S S
I 1
I r
11
i
c
-J
Sfc
B S
? I
?!
§sg "'ssif SsffljS s
Sii
•t*!f-;*i!i |S
istiff IfI
jif illflMIlI
Jlil Illilliflll
„. 2
f
•ss
il
II
li
I'* J U
if - 5, | ,„•
2 2
s i
i i
ii i
2-23
-------
S
UJ
ol| -•
CC <=•— ?
X 3T"35 ~
O - g. — °
R Iff 1 §
> T ^ £ E
Q *•.£, a. *
> c « S
m Is •§ §
CO 0 g. £ 7=
ffi Ml!
2« > 5 ° .S
u S « | <
0 g CD ° 5
Z »| v !?
O C P) Q.
0 8 § | fe
CO 55 ,» "§ "^
O g g ^2
n o > •=* 3
"*• c o c 7S
UL g E spl ° S
0 - £ i | ~
•^ *o .^* ^ . ,S
CO m *S. ^ ^ 2 ?
5 g cc w 5 5 c?
O in i • in , o
92 2 <
1
^*
3
CO
CM
i
Si
rr
in
Si
•*
*
_c
CD
CO
1
"O
~
LU
i
^.
CM
*3-
S
Si"
$
CM
I
2
CD
.S
"5
UJ
O
2
^
PCB ACCUMULATION AND CLEAR
CO
UJ
Q
CO
UJ
CC
m
£
UJ
3
CO
CO
1-
tr
O
o
o
S
m
•5
UJ
CO
CO
CC
UJ
t=
Q.
g
o
m
m
o
Q.
_e
m
CD
"S
o
onkey Aro 1254 steady state accumulation 1st
P4502B activity ~ [PCB]<>
jman Aro 1242 clearance 2nd order in PCB
P4502B activity ~ [PCB]1
5
^
ke P4502B
-1
1 x '
CM
CO
2
- Background-exposed humans
- PCB-exposed workers
- PCB-dosed mice; lightly dosed
uman variables that may affect clearance rate:
x
CO
a>
"I
u
CD
S
CD
05
CD
O
of
CD
HI
1
,
C g>
ill
J= CO CO
CD CD >-
co co o
CD X T3
CO Q) O
< CO CQ
I
ke P4502B + P4501A
13
CO
"c
- PCDF-exposed chloracne patie
y
S
•a
ts
elative rates of PCB congener metabolism esti
CC
-t
£
c
"H
£
3
X
1
CQ
o
a.
— W >•
18 0. S"
• 55 congeners from human data
- 67 congeners from biochemical or animal dat
- 68 congeners by structure-activity relationshi]
- PCB-dosed rats, monkeys (?), r
- Dogs, polar bears, some birds
ike P4501A
-1
1
i
1
K3
I
1
- Some porpoises
- Data permits calculating accumulability for ai
- Many species of fish
Hi
2-24
-------
CO
CO
O u."
0^ -
J'LL
LU >-
^^ C3
LLj 0
mm.
o -J
^Q
o —
a: -J
LU -!
CO <
K t
2 LU
CO 2
> 2
O H5
NCY OF STEA
RATE, D, SER
INDUCTION
LU m O
Q co "2
Z jo 2
H.Qa
LU Z Z
Q O <
I ^j
*""' «t "°
Z s i
P| 1
§D DC 1
Q O a-
Z CO
z <
0 1 T,
z ID .9-
DC I "3
o ^ a
+ w
E FOR INDUCTION (
ACTIVITIES BY PCE
0 g
•y LO
LU 2
9 LU
LLI 0
u. S
o x
> o
EC O
|o
Z> LL
CO O
ef
At low PCB level!
,j
m
CM
<
*~
Q.
Q.
ffi'
£
1
I
O
•5
o
2"
03
I
CO
£
jr
5
+
i
&
S
*~
.
242, 1016 occup., 2-35 yr
*~
TT
in
CM
o
CVj"
"""*
c
co
m
f
i
I
X
ix
_co
II
CO
5"
£
+
1
o
ccupational, obs. yr. 25
o
O
CD
CM
•>-
—
C
ra
>.
1
o
CO
s
^*
s
-HI
+
V
o
1
u."
O
+
oo
Si
co"
o
?^_
I
tn
At high PCB leve
cvi
-H
1
CD
CO
CM
00 ppm/diet, 8 mo.
T~
(O
o
Tf
"—
•5
tr
CD
ffi
1
+
+
CO
00 ppm/diet, 8 mo.
•*~
CM"
Si
•»
—
cc
«
C
S
CO
1
£
ii
,
i
CO
g;
u
CO
sr
£
+
+
o
o
'
•§
J2
O
"o
§
^
^r
CM
co"
-*
o
cc
5j
5*
£
s
o
in
^
Q.
"o
+
1
|
'
00 mg/kg, obs. wk. 4-55
in
^r
CM
CM"
• — •
ffi
CO
j
(S
a
r
$
t
1
o
Approximate gen
CO
c-
•H
CO
^
CO
s
p
Q.
•p-
ca~
o»
CD
O
to"
— •
>1
s
5
S
Q.
1
g
II
Ul
j:
aft
2
it
CO
15"
O
IL,
i
•H
5
-80 ug'kg/da, 37 mo.
to
V
in
3?
— '
1
>• CM
5
1 Vl
1
I
,_
f
0
(15) indicate exposure to
>ably Aroclor 1248
c °
C o_
C0 "*
raco
|i
o o
-= s
0 S
"O ^
1 »
Is
-O CO
3 CD
CS
C EQUIVALENCY DEBATE
X
O
H
LU
H
mental PCDF/PCDD mixtures being
dioxin-like toxic equivalency (TEQ)
al toxic equivalency factor (TEF)
DD congener
IIII
C g T3 O
n
>i*O *Jr
c « o o
Q) CO (0 <0
*— O ^j .S
O CB .O S
congeners also have measurable
is; could be assigned TEF's
u/
to E
o .>
^ 0
o
~ _^
E =
Q) .C
p x
0 °
CO T3
relate with some indicators of
o
0
>»^
"><
U. 0
UJ *•
u^
S T3
• §
ate dioxins, PCDFs, PCBs as a
determined TEF Values
= >*
6 '5
O CO
— CO
CO "^
CO •*
o a.
Q. =
E 2
i— ^.
D. TO
©
0
'E
FEQ's don't correlate with carcinoge
r~~
^
o
«
t t
£
.
C
OF A PCB COMPOSITION
ADMINISTRATIVE AUTHORI
Lu >-
> to
-t LU
LU Z
2 «S
^f ^L
LU *~J
0 °
t- LU
F ?
H ^
< Z
sx rf
LU O
STIONED
LU
COMPOSITIONS BEING QU
K-
LU
cr
LU
LL
LL
Q
LL
O
CO
^£
CO
LU
>
r™*
5
LU
CC
o-
CB
3
C?
CO
CO
"5
"3
l±
1
dioxin equivalency (TEQ)?
"5
£>
C
0
CO
CO
CD
CO
<
1
relative accumulability?
"o
.2
ja
c
o
W
o
CO
<
1
2-25
-------
CO
o
CO
ID
O
O
o
sjouni qiiH siDiaiuo
I
i
CO
o
CD"
o
1
*2
co"
CO
C
o
i
CD
CO
O
Q.
CO
u
"co
o
ied pharmacol
i only identif
_g
£5
o
=i
P4502B ind
CO
CD
E
o
o
o
B exposures 1
0
Q.
To
o
I
u
0
o
r\
response U
.
_g
%•*
Q.
E
Z3
(0
8
CO
M=
&
e resulting f
8
*-
c
CO
£
>_
£
CO
V«
O
1
c
"o
1
TJ
^3
^^
s
*o
S
5
CD
1
s
jg
O
•o
CD
§
0.
X
;cupationally e
u
O
•5
CO
'•S
tz
CO
"co
o
CO
epidemiolo
w
CD
JC
sureable healt
CO
CD
O
C
CO
CD
>
workers ha
CO
£
o
CO
CD
to
i
£
CD
*••
1
•*-•
C
'5
o
±m
If concerns
CO
O
Q.
T3
CO
CO
o
Q.
.£2
.c
^5
CD
f
•s
CD
1
i
CD
CO
O
£
measu
lie appropriate
TJ
O
.Q
CD
CO
c
o
accumulati
CD
O
CD
to
1
CD
Q.
umulability, or
relative ace
CO
CO
*lM
.1
4-*
£
"5
i- CM
CO
U II
esuodsoa SAiiB[9U pSBipeJd
2-26
-------
2.5 ANIMAL/HEALTH CONNECTION
Theo Colborn, Senior Fellow, W. Alton Jones Foundation, World Wildlife Fund,
Washington, DC
PCBs are reported in fish tissue from the equator to the Arctic. They are persistent,
hydrophobia, and vaporize readily, atmospherically transported long distances. Recent studies
demonstrate that PCBs are deposited at the same rate in the Arctic as they are in the Great
Lakes. Atmospheric deposition contributes 90 percent of the PCBs to Lake Superior, a lake with
little industrial activity on its shoreline. By comparison, Lake Ontario, with a highly
industrialized shoreline, receives 6 percent of its PCBs from the atmosphere.
Effects Reported in Wildlife Dependent Upon Fish
A comprehensive examination of the literature about wildlife from the Northern Hemisphere
reveals that many species dependent upon fish are having difficulty maintaining stable
populations. Population instability is associated with effects in offspring leading to embryo or
early mortality while the animals directly affected, the adults, exhibit little or no overt effects.
The transgenerational problems in embryos, fetuses, and newborn animals have been associated
with PCB concentrations and dioxin toxicity equivalents in their tissue. The endocrine and
reproductive systems of the progeny are affected, measured in terms of thyroid and other
endocrine tissue malfunction, or sex reversal, and/or reduced fertility in birds, fishes, and
marine mammals; less quantitative and observational evidence has been reported on behavioral
changes in birds; and more recently, emerging evidence suggests that the immune; system is
involved in premature mortality in birds and marine mammals.
Some of the anomalies reported in Great Lakes wildlife that are associated with endocrine
disruption include:
Female herring gulls sharing a nest and males in the same population forming
"fraternities" and not carrying out their parental responsibilities
All double-crested cormorant chicks with crossed bills have been female
One hundred percent of fish examined by a team of wildlife biologists from Guelph
University have hypertrophic and hyperplastic thyroids and recently the thyroids of Lake
Erie fish are rupturing because of their extremely large size
Herring gulls also have enlarged thyroids
Fish and herring gull thyroid hormone ratios are skewed (T3:T4)
Top predator male fish are precocially developing sexually but never achieving full
sexual maturity
2-27
-------
Most top predator fish species and other bottom fish exhibit various stages of
hermaphroditism
A case study is provided to demonstrate that the endpoints (measurable health effects) as
a result of exposure to PCBs and other contaminants may be very sensitive and easy to miss.
The endpoints shift over time, differ among species, and are dependent upon timing of exposure
and dose. In 1983 an association was made between total PCB concentration and delay in
incubation time, poor hatchability, poor chick weight gain and loss of weight leading to early
mortality ("wasting"), decreased fledging success, and overall mortality in a breeding population
of Forster's terns on an island in Green Bay. In addition to chemical assay to measure the
concentration of total PCBs in the chicks and abandoned eggs, the HII4E rat liver hepatoma
assay was used to determine toxicity. By day 17, almost all the chicks were dead and the
parents had abandoned the area. The study was repeated 5 years later in 1988 after remedial
action for PCBs in the Green Bay area. Although everything appeared normal throughout
incubation, hatching, and among the offspring up to day 17, on day 18 the chicks began to
"waste" and by day 31, the same mortality was recorded as in 1983. If the biologists had
returned to their laboratories on day 18 instead of staying in the field, it would have been
assumed that the PCB concentrations reported in the offspring in 1988 were "safe".
The importance of congener specific analyses in order to make cause and effect
associations is borne out by a study comparing lake trout egg mortality with PCB #77 (3,4,3',4'-
tetrachlorobiphenyl) and total PCBs. Eggs were stripped from ripe females from 5 locations
around the Great Lakes. No association was found between egg mortality and total PCB
concentration in the eggs. However, a correlation of 99.9 was discovered with congener #77.
Almost no large-scale marine mammal die-offs occurred before 1987. Since then animals
have been found beached and dying from new strains of viruses specific for seals, dolphins, and
porpoises. In each event, elevated concentrations of PCBs and other organochlorine chemicals
were found in the tissues of these animals and the animals' immune systems appeared to be
compromised. It is yet to be determined whether immune suppression was the result of chemical
exposure or the result of the viral infections. These events took place only among toothed
mammals dependent upon fish. Baleen whales have not been affected similarly.!
i
Effects in Laboratory Animals Fed Contaminated Fish
Rats fed a 30 percent diet of Lake Ontario salmon for 20 days no longer respond well to
stressful events, such as a change of scenery in their cage, a mild shock, or a reduction in food
pellets. The same effect is reported after a 10 percent diet for 60 days. The offspring of
females fed Lake Ontario salmon from the day they were bred and during the first seven days
of lactation cannot cope with stress either. These fish were contaminated with approximately
8 ppm PCB and a number of other organochlorine chemicals. I
2-28
-------
The Wildlife/Human Parallel
A number of models have been generated to assign a biomagnification or bioaccumulation factor
to PCBs when determining risk to humans. However, the unpredictability of these models is
borne out by tracing the course of PCBs in the Arctic food web from the arctic cod to the ringed
seal, and ultimately to the polar bear. .The cod hold primarily the tri- and tetra-chlorinated
PCBs; the seals, the tetra- and penta-; and the polar bears the hexa- and hepta-chlorinated PCBs.
The anomalous enrichment of the higher chlorinated PCBs in the animals in this simple food
chain is reflected in the pattern of PCB congeners found in the tissue of native Americans
dependent upon marine and freshwater fish and mammals. Breast milk of native Americans
from eastern Arctic Canada holds 3 to 5 times more PCBs than breast milk from women in
temperate zones and the congener composition is quite different. This raises questions
concerning the reliability of models when dealing with chemicals that are comprised of multiple
compounds (PCBs are comprised of 209 different compounds called congeners) in which each
congener appears to have a distinct toxicity pattern; and which have become distributed
worldwide and whose fate, therefore, is almost impossible to determine.
Thirty two of the 43 areas of concern around the Great Lakes designated by the Canada
and US International Joint Commission and EPA were chosen because of elevated PCB
concentrations in abiotic and biotic samples from the areas. Advisories are currently issued in
these areas warning people not to eat certain species and sizes of fish depending upon the
consumer's age and sex. These advisories are based on a cancer model using data that are over
20 years old. Despite health advisories people are still fishing and eating their catch.
A number of studies suggest that PCBs when present in the womb have the potential to
affect the developing nervous system of the embryo, fetus, and newborn. Neurotoxic effects
have been reported in offspring at birth and later at age four of mothers who ate Lake Michigan
fish two to three times a month for at least six years preceding their pregnancies. Effects were
seen in children whose mothers' breast milk fat PCB concentrations exceeded 1 ppm. Similar
effects were reported at birth in a cohort of children borne of women from North Carolina.
However, no effects were detected later in follow-up examinations of these children. The
median PCB concentration in the women from North Carolina was higher than those from
Michigan who had consumed Lake Michigan fish. Both studies suggest that the neurotoxic
effects in the children were of prenatal origin, not from breast feeding. A study involving the
cross-fostering of Chlophen-A30-exposed and unexposed rat pups (a German analog of PCB),
confirmed that the neurotoxicological effect was initiated in the womb, not postnatally. It is
important to note, that when the children v/hose mothers ate Lake Michigan fish were tested at
age four, 17 of the children refused to cooperate and take some or all of the tests. These were
children whose mothers had the highest PCB breast milk concentrations in the study.
A close examination of the literature on laboratory animal and human neurotoxic effects
as a result of exposure to PCBs reveals that the human is 10,000 times more sensitive to PCBs
than the rat, using the traditional rat endpoint for neurotoxicity.
2-29
-------
An examination of the literature on PCBs and male fertility revealed that the:
Sperm of mature male rats exposed postnatally to PCBs through their mothers' breast
milk were unable to penetrate the ova in healthy, PCB-free females ;
I
Male pups of female rats fed one low-dose meal (0.064, 0.160, 0.4, and 1.0 ug/kg/body
weight) of dioxin (2,3,7,8-TCDD) on day 15 of gestation (the approximate day sexual
differentiation commences in rats) were demasculinized and feminized as they matured
and their sperm count was reduced by 75 percent
i
Motility of sperm from men visiting a fertility clinic was inversely related to the
concentration of three PCB congeners, one of which is 2,4,5,2',4',5'-hexachlorobiphenyl
which comprises about 20 percent of total PCBs in humans in temperate zones and 40
percent to 45 percent of total PCBs in humans in the Arctic
'
Male sperm count declined worldwide by 45 percent between 1938 and 1991. Taking
ejaculate volume into consideration, which decreased over the same time period by 25
percent, the decrease in sperm count is equivalent to a 50 percent reduction
I
» Reproductive biologists, endocrinologists, and lexicologists agree that exposure in the
womb to PCB and other estrogen-like substances in the womb can irreversibly affect
sperm production in adult males
Human health assessments can no longer be based on the probability of developing cancer
alone. Multigenerational loss of function must become a part of the human health assessment
process. Parallels in wildlife, laboratory animals, and humans in response to exposure by PCBs
and other chemicals found in fish tissue cannot be ignored. To date it appears that the exposed
adults are not overtly affected. Instead effects result from parental transfer of this chemical to
the offspring. In human offspring the effects may not be manifested as clinically relevant
endpoints and may not be recognized at birth. In humans and wildlife the effects could be easily
missed and not expressed until adulthood—making a causal link almost impossible.
2-30
-------
0
o
w-
(A
O
'o
O
Q.
cn
X
X
X
X
X
X
X
0)
O)
CO
111
15
m
X
X
X
X
X
X
1
X
X
"to
.c
CO
O)
"o>
ffl
X
X
X
X
_
-^
| Black-Crowned
X
X
X
X
X
X
c
h-
c
en
'5.
(A
en
O
X
X
X
1
X
1
c
O
£
CO
en
| Chinook/Coho
X
X
X
X
X
X
Common Tern
X
X
X
X
X
X
X
X
D.C. Cormorani
X
X
X
X
X
X
X
1 Forster's Tern
X
X
X
X
X
X
X
X
X
X
"5
0
en
c
I
X
X
X
X
1
X
X
*-<
o
o
X
X
X
1
X
X
c
i
X
X
X
2
Q.
(A
0
|
X
O
33
o
X
X
X
X
^^
=5
o
•o
5
O)
c
E
X
X
X
X
X
X
fli
Snapping Turtl
0)
3
03
O
*-
O
£
*->
_c
•o
o
Q.
C
O
o
jQ
(1)
(0
o
u—
H—
0)
•c
o
2-31
-------
2
o
X
-£o3Q
X X
X X X X X X X
m
_ en
c ro
ro c
O) E
ir ro
OQ
X X
X
XXX
UJ
E
c E
= e2
E S Q.
c1 ** C
= S? 5
E>-o
X X
X
o
UJ
u_
u.
at
Q
U!
o:
UJ
U)
CD
O
Is-
C~ C
= §E
•§•§1
i: c m
Q.UJ a.
LU
U.
UL
UJ
X
I
X
X X
X X X X
X X
X X X X X X X
Q>
M
o
V
o
JQ
T3
o>
to
a
ra
O
I *
CD C)
j
(5 c "3
'
Q.
<
cc
<
.0.
o
Q
03 C
-------
h-
cc
o
Q_
UJ
CE
CO
UJ
Q
Z)
CC
<
LU
CC
<
u_
o
2
o
h-
<
O
O
halogens
tion decline
lered
g^cf
co — ' ro
v_ O C
0 0.03
cncn
coco
cn cn
'"I" *"|~
ocnoo
cor- oo
cn cn cn
& li-
ra > co
JC > -C
§ CO >
s=£
— 03 CO
O.JD ~
u.
CO
+_i
W
foundland
.awrence E
of Fundy
> __1
03 +J CO
z:cn CQ
w
CO
• c
U— 4« 4— H— *O
s — H — 4— H— r*
O O O O co
_03 _03 _Q3 _03 g
'~O ~O "O "D W
T- r- r- r- CO
cn cn cn cn co
cn cn co cn cn
i t i i i
r*» r- r^ o co
coco co cn co
cn cn cn cn cn
WWW
CO CO CO
03 03 03
WWW
0 O O
JO JD JD
CO CO CO .£
>•>>.£- "TO
03 03 03 O JC
CO CO CO §
03 03 03 2 E
CO CO CO.— Q3
c c c j= c.
'uT ~ 'tZ W W
w
•(— *
CO W
03 CO
CO C
c °
m CO C
03 <" o3 .5
CO CO CO = _ C3
03 03 c 2 ^ O
0 JC"O.+± O. Q
•+3 t "° "C O ^
- ^ roo -2
CQ 2 > ^ UJ
H—
H —
0
.35
TJ
CO
CO
cn .
*"j~
r^
CO
cn
^ i
"5
03
W
"5
"co
« CQ
"co
"co
CQ
03
CO
H— **—
H« H—
O O
_03 _03
•OT3
CO
CO
CO
*"]"
r-O
co cn
cncn
.E .E
JC JC
"o "o
T3 T3
03 03
W W
O O
C C
0 0
jO X3
h Atlantic
of Mexico
+-• H—
0 3
~Z_ CD
g
<
o
o
CO
T3
CO
C
CO
O
03
0.
O
z>
LU
_co
"w
w
a
CC
en
.D
w w w w .w w
• c c c c c c
'03 03 03 03 03 0)
CO CO CO CO CO CO
_o_o _g _g _g _o
CO CO CO CO CO CO
JC JC JC JC JC JC
o o o o o o
c c c c c c
CO CO CO ' CO CO CO
CO CO CD CO CD CO
o o o o o o
CO
CO
cn
CN r- O T- 10 LO
CO CO CO CO CO CO
cn cn cn cn co cn
_Q3
CO
-§t<
O W 13
03
w
03 03 03 W i,
05 " '£ ^ ^
= t -i -E ^ "5
_M W > 5 JP CQ
o
CO
a.
c
CO
a.
CO
-3
2-33
-------
en
CD
O
a.
CO
CD
"cB
"cS
E
2
to
o
z:
-a
CD
co
CO
-------
Table V. Biomagnification Factors for the Food Chain
Arctic Cod/Ringed Seal/Polar Bear0
fish to seal
seal to bear
compound
S-PCB
PCB homolosue
C14
C15
CI6
Ck
Cla
Clo
S-DDT
S-CHLOR
S-HCH
S-CBz
dieldrin
p,p'-DDE
male
seals
8.8
2.5
11.3
23.0
17.0
>3
6
20.2
7.3
1.5
0.2
2.4
62.2
female
seals
3.7
1.2
5.1
9.7
6.9
>1.4
>109
8.2
4.7
1.6
0.2
2.2
17.6
male
seals
7.4
<0.2
1.9
9.3
19.8
24.6
6
0.3
6.6
9.3
15.55
4.88
0.3
female
seals
13.9
<0.4
3.1
18.3
38.6
47.2
>109
0.5
9.5
1.7
17.4
5.5
0.7
Fish
to bear
49.2
<0.5
13.9
170.6
263.4
>67
b
3.6
44.2
.2.7
3.4
11.4
10.7
a Ratio of tissue concentrations on a lipid weight basis. 6C19
PCBs not detected in fish muscle (<2 ^g/kg of lipid wt) or seal
blubber (<0.5 ^g/kg of lipid wt).
Source: Muir et al. , 1988. Environmental Science and
Technology, pp. 1071-1079.
2-35
-------
2.6 SUMMARY OF QUESTIONS AND RESPONSES8
2.6.1 Gordon Blaylock of the Oak Ridge National Laboratory questioned Dr. Cogliano
regarding the uncertainty that should be attached to the carcinogenic potency ("slope factor") for
PCBs. ;
I
\
Dr. Cogliano stated that when applying the slope factor to an Aroclor 1260 mixture, the
uncertainty factor should probably be no more than half an order of magnitude, a factor of 2 or
3 at most. He stated that when applying the slope factor to other mixtures, one to two orders
of magnitude is acceptable, although the uncertainty increases for dissimilar mixtures. He noted
that for mixtures that do produce cancer, one of the striking features is that the toxicity values
are fairly close. ;
2.6.2 Brian Toal of the Connecticut Health Department asked Dr. Cogliano whether the
contamination of PCB mixtures by furans is still considered an issue, as it was in some of the
older studies. :
•
Dr. Cogliano stated that he did not believe that furans were an issue with the mixtures
that were tested on animals. He noted that one of the studies reported that their mixtures were
free of furans. Dr. Cogliano stated that a sensitivity analysis was conducted in 1989 (based on
the toxic equivalency factors of some of the dibenzofuran contaminants of the PCB mixtures).
He noted that based upon this analysis, it was found that the furans would have contributed less
than approximately ten or twenty percent of the carcinogenicity seen in animals. However, he
stated, when evaluating actual human exposures to PCB mixtures, this may not be the case. He
noted that, in the Yusho incident for example, the rice oil was heated, which facilitates the
transformation of PCBs into dibenzofurans; therefore, he concluded, there may have been higher
concentrations of furans in those mixtures than in any commercial mixture.
2.6.3 Mitch Erickson from the Argonne National Lab asked Dr. Cogliano whether there were
any documented human cancers as a result of PCB exposure.
I
Dr. Cogliano stated that there are several epidemiological studies that increased cancer
rates. Other studies do not show an elevation in cancer. He stated that some of the elevations
are statistically significant. However, EPA considers the human data inconclusive. He stated
that the assessment on PCBs, as with most environmental chemicals, is currently based on
animal studies. '
Ed. note: The question and response portion includes summaries derived from transcribed
conversations. The summaries have been carefully edited to present the discussion as accurately as possible.
However, these question and response summaries have not been reviewed by the speakers—unlike the proceeding
abstracts. |
2-36 :
-------
2.6.4 Peter Somani of Columbus, Ohio, asked Dr. Cicmanec whether the doses used in non-
cancer studies were comparable to those used in cancer studies. Did the dosages differ between
the monkey studies and the rodent studies? Also, how is the information derived from rodent
and monkey studies used to develop coherent policy involving PCBs in humans.
Dr. Cicmanec did not recall carcinogenicity studies in monkeys where comparisons could
be made. He stated that in the rodent studies, doses of 100 ppm were generally used, although
doses as low as 50 ppm were sometimes used. In rodents, the effects were generally seen at 100
ppm. However, the Rhesus monkey studies sometimes showed effects at dosages 100-fold lower
than the 100 ppm dose.
Concerning the question on applying animal data to humans, he stated that although a
wide range of studies for the non-cancer effects are available for EPA to choose from, the
Agency has focused on Rhesus monkey studies, because of physiological and other similarities.
Mr. Cicmanec also noted that, in the case for non-cancer effects, his group had chosen
to use a number of commercial mixtures. He acknowledged that these may not be exactly the
same as the Aroclor mixtures found in the environment, but they are currently the best we have
available.
2.6.5 Dr. Bolger commented that although there is a vast body of information on PCBs,
regulatory agencies tend to focus on a key study (or studies) and the associated dose level. As
a result, he stated, other data may be discussed, but not used quantitatively.
Dr. Cicmanec agreed and added that the other studies may be used to characterize the
uncertainties associated with a risk assessment, although this information may not receive enough
attention. He stated that risk managers who make crucial decisions would have better
information upon which to base their decisions, rather than just using a single number.
2.6.6 Gerald Pollock of the California EPA questioned Dr. Barnes about whether it might be
appropriate to develop a Toxicity Equivalent for dioxin-like PCB congeners and add that to the
total TEQ for dioxin, thus deriving an overall Toxicity Equivalency calculation for dioxins in
fish tissues.
Dr. Barnes agreed that this was a reasonable goal but there was no official EPA position
at present. He noted that this was discussed at an earlier PCB workshop.
2.6.7 Jack Moore of IEHR commented that he believed it was going to be several years before
TEQs for PCBs are likely to be established. He cited two examples to illustrate the difficulties.
One was a study by Nickels and Peterson (who conducted a study last year using fish fry
2-37
-------
mortality with PCS congeners). The study indicated that the numerical toxicity Values that had
been historically assigned to particular congeners seemed to be in error. Mr. Moore also cited
another study which examined results of PCB rat tests, assigned TEQs on a congener-specific
basis, and then used the TEQs to predict what the results of a two-year bioassay might be. The
predicted results were the opposite to what had actually been observed. j
Dr. Barnes acknowledged that the interpretation can be complex. He noted that Steve
Safe has demonstrated antagonism within Aroclor mixtures in which non-dioxin like PCBs can
reduce the apparent toxicity of the mixture, despite the presence of other highly toxic congeners.
2.6.8 David Pierce of the Massachusetts Division of Marine Fisheries commented that, from
his review of recent literature in scientific journals, it appears that many areas of the country
have already begun to use the TEFs for PCBs. He stated that this suggests that the carcinogenic
potency factor (CPF) for dioxin could become a very significant factor in evaluating PCB
toxicities. Pierce asked if any work was being conducted within EPA or elsewhere to verify the
CPF for dioxin.
i
Dr. Barnes responded that EPA's reassessment of dioxin is looking at the current CPF,,
He stated that the Agency is seriously considering a different model for deriving the cancel-
potency factor. This could in turn affect the CPF for PCBs. ,
2.6.9 Brian Toal asked that, if TEFs are adopted for PCBs, whether the TEFs would likely be
based on dioxin carcinogenicity or on the most carcinogenic PCB congener(s).
Dr. Barnes responded that it is unclear at this point. TEFs could be kept within the
sphere of PCBs. However because the current focus is on the coplanar, dioxin-like PCBs, the
TEFs will probably be related to 2,3,7,8-TCDD.
2.6.10 Peter Somani from Columbus, Ohio, asked Dr. Brown whether there was information
about the type of PCBs that workers in his longitudinal studies were exposed to. And, based
upon these studies, Mr. Somani asked if problems had been noted with the liver enzymes or with
incidences of cancers or tumors. i
Dr. Brown responded that the workers in his studies were exposed initially to Aroclor
1254 in the 1940s and 1950s, to Aroclor 1242 in the 1960s, and to Aroclor 1016 in the last few
years of PCB usage. He noted that in G.E.'s studies, they tracked individual congeners. They
had pharmacokinetics data for approximately 40 individual congeners for the workers study. Dr.
Brown also responded to Mr. Somani by noting that there were no indications;of liver disease
in those individuals studies; he added that it was the absence of such clinical findings that led
to an examination of the correlation between PCB levels and serum, lipid levels where the
covariance was identified. Dr. Brown stated that he did not conduct a study on cancer
2-38
-------
epidemiology. However, he noted that two other groups had: one study was conducted by
NIOSH, and the other was conducted by New York State. The conclusions from these studies
were that there were no elevations of cancer in that study group.
2.6.11 Dr. Bolger of the FDA questioned Dr. Brown regarding the population of those studied.
Dr. Brown responded that the longitudinal study started with 194 workers in 1976. He
stated that the workers had been reexamined in 1979, 1983, and 1988. He noted that for the
last examination, 138 of the subjects participated. He added that, although many had left the
company, his research team completed health questionnaires from 30 additional persons. He
stated that the epidemiological studies have been based on all 6,300 people who have ever
worked in that plant.
Dr. Bolger pointed out that those numbers represent a fairly small population, and that
to see an effect in that sized population would have been unusual.
Dr. Brown responded that one would not expect to see a one in a million effect.
However, he noted that for a cancer mortality effect, 6,000 people comprised a substantial
group.
2.6.12 Dr. Bolger asked Dr. Brown to verify a previous remark during his presentation
regarding the toxicity difference observed in humans and in mammals.
Dr. Brown stated that the different animals responded differently to the PCBs in terms
of measurable parameters, such as enzyme induction or toxic manifestations. He stated that
there were some commonalities among the animals, but that a total commonality was not seen.
2.6.13 Dr. Heraline Hicks of ATSDR asked Dr. Brown how many female workers were in the
study cohort, and whether any of the females were observed to experience any unusual medical
effects.
Dr. Brown responded that approximately 25 percent of the cohort was female and that
no unusual effects were seen. He stated that metabolic clearance is slightly slower in the
females, but it had little effect.
He also mentioned one of the parallel studies that was conducted by Bill Taylor (who was
then with the New York State Department of Health). That study observed reproductive
outcome in the plant population as a whole; Dr. Taylor observed approximately 200 pregnancies
in the more-heavily-exposed women versus 200 less-heavily-exposed women. The study showed
that in the more-heavily-exposed women there was a small shortening of gestation time and a
slight reduction in average birth weight. However, there were fewer below-normal birth weights
2-39
-------
recorded. Dr. Brown stated that he did not have a comparable study for this group of very
heavily exposed people.
2.6.14 Peter Somani of the Ohio Department of Health described a policy dilemma that faces
agencies that issue fish advisories. He explained that it is difficult to convince people to reduce
their fish consumption when they are also being told that the chemical concentrations in the
tissues are generally declining and that the water quality is improving. i
Dr. Colborn agreed that fish are part of the protein food base that we need desperately,
and until we can definitely show that all the dietary inputs come from one food source, we will
continue to have a difficult time setting standards. In the interim a continued emphasis must be
placed on the prevention of these problems, through increased testing and other ways to prevent
the release of harmful chemicals. She emphasized the need for more developmental testing,
including transgenerational effects.
Dr. Southerland stated that one of the main reasons for the workshop was to assist the
states with fish consumption advisories. EPA recognized that the conference would not give
participants an explicit "cookbook" approach for handling PCBs in their fish advisory program.
However, the conference should give participants different information about PGBs that can be
used if states adopt a risk assessment approach to fish consumption advisories.
i •
2.6.15 Mr. Schwartz asked Dr. Brown if the participants are missing important data by not
analyzing for neutral lipids; also, Mr. Schwartz asked whether testing for "neutral lipids" would
clear up certain ambiguities associated with the total lipid measurement (as it rebates to organic
compounds). ;
Dr. Brown defined neutral lipids as "the sum of cholesterol, cholesterol esters, and
triglycerides with the exclusion of polar lipids, meaning phospholipids." He slated that neutral
lipid testing is conducted because lipophilic organic compounds are distributed among body
compartments in proportion to their content of fat. A possible exception, however, is the lipids
that are prominent in brain tissue, which does not pick up PCBs and DDT (unlike ordinary body
fat). . |
2.6.16 Dr. Gerald Pollock returned to the policy problem raised earlier by Mn Somani. Dr.
Pollock did not regard the setting of revised exposure limits as a major policy [dilemma, even
if a new, lower value is needed. If a new, more sensitive end point is recognized, they will be
begin using the new information. A problem that may arise, however, is how to balance the
risks of eating fish against the benefits and how to communicate that.
2-40
-------
Dr. Southerland noted that people may question the validity of exposure assumptions that
are frequently used in risk assessments, particularly for cancer assessments (e.g., assuming
maximum exposure for an entire 70-year life span). Because of this question, people are
becoming more interested in effects that manifest themselves after shorter exposure periods.
2.6.17 Mr. Thomas Fikslin of the Delaware River Basin Commission then asked if the Agency
was planning to move toward congener-specific risk assessments. He also wanted to know the
implications for state monitoring programs that currently monitor Aroclors, and level-of-
chlorination, rather than specific congeners.
Dr. Cogliano stated that he believed it was unrealistic to expect a congener-specific PCB
assessment in the next several years, because the toxicity data for individual congeners is lacking
and because of the difficulty establishing toxicity equivalency factors for most of the congeners.
Therefore, Dr. Cogliano stated that he anticipated the use of total mixture basis in the meantime.
Dr. Barnes added that his personal belief is that the EPA is moving toward congener-
specific analysis, and that it is basically a question of time.
Mr. Fikslin stated that there is a need for a long-range plan as to what to implement in
the interim (while waiting for the congener data). He stated that there appears to be a long lead
time in terms of getting laboratories capable of conducting a proper analysis.
Mr. Hoffmann stated that he believed that different parts of EPA were monitoring
different congeners (e.g., the EMAP program). He noted that certain programs within the EPA
are confronting the same issues that states are regarding analytical methods and interpretation
of the data.
2.6.18 Ms. Amy Owen of the Inter-Tribal Fisheries & Assessment Program then stated that she
was concerned about the absence of proof for developmental effects in humans. She stated that
lifestyle habits (e.g., smoking, drinking) had an effect on development rather than the fish
consumption in the study.
Dr. Colborn noted that she was aware of this factor, and stated that the Jacobson studies
had adjusted for lifestyle habits.
Ms. Owen stated that another study (Dr. Hovinga's) had not accounted for lifestyle
habits. Dr. Colborn stated that she would have to look at the study.
Dr. Colborn also noted that in regards to the study where women who had consumed
contaminated fish and who had delivered higher weight babies, the women had actually
consumed white fish, which is a far less fatty fish. These women also had consumed less fish
than the cohort of women in the earlier study. She also explained that the changes noted in the
2-41
-------
Jacobson study were very slight in regards to breast milk fat and PCB level of the mother. She
noted that in a small cohort, changes would be difficult to observe. ;
Ms. Owen then questioned Dr. Colborn about the population decline and adverse effects
that Dr. Colborn had described for various wildlife species. Specifically, she wished to know
if the population decline could have been caused by factors other than PCBs.
Dr. Colborn responded that the population declines were not necessarily the result of
PCBs; rather, the declines were reported in populations that were exposed to higher
concentrations of organochlorine-type chemicals in the Great Lakes system since the 1950s.
Factors other than PCBs could be responsible. She reiterated the need to prevent the release of
detrimental chemicals and to test for developmental endpoints.
2.6.19 Mr. David Pierce questioned Dr. Southerland regarding whether the EPA! will continue
to provide advice to the states regarding fish consumption, specifically regarding the level of
acceptable risk as IQ6 (as it was provided in the past), or the newer level of
lO'5. ;
Dr. Southerland answered that the Agency would provide technical guidance, when
requested by a state, regarding fish consumption. She emphasized that the guidance is strictly
non-regulatory. She noted that all 50 states have the Agency's current draft of how to sample
and analyze fish to support state fish consumption advisories. The Agency is also working on
its next guidance document for risk assessment. The risk assessment guidance will recommend
procedures for assessing cancer risks, estimating exposures, IRIS information, etc. It also will
include reproductive toxicity data, including appropriate end points, exposure information, etc.
Dr. Southerland stated that, following risk assessment, states have to make risk
management decisions about whether to issue fish consumption advisories; EPA has no
regulatory authority for fish consumption advisories. She also noted the difficulties that
bordering states would experience in attempting to reach a consensus (individual states may
reach different decisions with regards to the same body of water).
2.6.20 Dr. John Brown then returned to Dr. Colborn's earlier topic about multigenerational
effects in the children of PCB-exposed workers. He reiterated that his study group had
identified populations of pregnant women who had PCB levels above the background level.
Therefore, if he could obtain specific research questions and could identify quantitative indicators
related multigenerational effects, his group would be willing to explore them.
Dr. Colborn felt that Dr. Brown's cohort of women aged 15 to 35 years of age was a
perfect cohort to find out whether such effects are taking place. She cited the case of DES-
exposed women as an example of a subpopulation in which chemically induced alterations in
sexual development resulted in transgenerational effects.
2-42
-------
Dr. Southerland asked for clarification on Dr. Brown's statement, specifically, regarding
whether it was the cohort with the high blood serum levels or the 6,000 that had the more
variable exposures just slightly above background levels. Dr. Brown stated that it was the
roughly 400 women who had varying levels of PCB exposure while pregnant (400 out of the
6,000). Dr. Brown stated that 200 were in a higher exposure group versus 200 in a lower
exposure group.
2.6.21 Mr. Roy Martin of the National Fisheries Institute made two comments relating to how
exposures to chemical contaminants are addressed in fish advisories. He recommended that,
when states issue fish advisories, they explain to the public that research is often based on whole
tissues and organs, not necessarily the fish fillets that people actually eat. Specific
recommendations on how to reduce fat levels using different cooking methods is also useful
information in an advisory.
Dr. Southerland agreed. She noted that these were two of the suggestions that the
Agency would cover in its upcoming guidance. She also described how some states incorporate
fish size limits into their fish advisories. This approach can allow a state to minimize exposure
without depriving people of a valuable source of protein. It also helps avoid the banning of
whole fisheries because of PCBs, which are ubiquitous.
2.6.22 Mr. Daniel Thomas of the Great Lakes Fishing Council then asked Dr. Brown when his
reports were conducted and when the reports were made available to the scientific community.
Dr. Brown answered that the study was initiated in 1976 and is still continuing. He
stated that the major results appeared after the first two series of examinations, and that his
group had several major papers out in the 1985-86 period, and even smaller papers after that
time.
Mr. Thomas then brought up a point regarding an earlier slide that Dr. Colborn had
shown regarding animals that were at risk, specifically, the lake trout. Mr. Thomas wanted to
clarify that the lake trout had never been placed at risk or extirpated because of chemicals.
Instead, he noted, the lake trout was extirpated in the 1950s due primarily to habitat losses,
over-commercialization and the sea lamprey. He noted that since that time, 60 million lake trout
have been planted in the Great Lakes by the Fish and Wildlife Service, and overall, the
population* has not been re-established. He attributed this to the different genetic strains,
degraded habitat, etc. He also discussed factors that affected the Coho.
Dr. Colborn responded that several other species disappeared (the bald eagle, osprey,
other species) at the same time the lake trout disappeared during the 1950s. She noted that there
are toxicologists that would disagree with Mr. Thomas; they believe that by now, if the lakes
had been cleaned properly, the fish would have come back. Dr. Colborn acknowledged that she
is presenting the issue from a toxicological perspective, particularly since the parallels are quite
2-43
-------
convincing. She agreed that the lakes have been devastated for many years; however, in many
areas the habitat has been restored and the fish are not coming back.
Mr. Thomas commented that Dr. Terry Bills conducted extensive
lampreycide that is used for the control of lamprey, TMF, which show little or
other species.
studies on the
no effect on
2.6.23 Dr. Gerald Pollock addressed Dr. Southerland's earlier comments regarding the fish
consumption advisory and the fish size relationship. He agreed the Great Lakes states are a
good source of information with regards to fish size, body burdens, age, etc. However, he
explained that these size relationship principles may not hold true for ocean fish. In a limited
study of selected fish populations, he found that the contaminant levels did not necessarily go
up as the fish size increased. He speculated that as ocean and/or estuarine fish grpw, they may
move to different, less-contaminated niches and reduce their body burden. Thus; although the
general principal is valid, site specific information may be necessary.
2-44
-------
PART THREE
ANALYTICAL METHODS
3.1 PCB ANALYSES-AN OVERVIEW
Mitchell Erickson, Group Leader, Environmental Research Division, Argonne
National Laboratory
Communication with the laboratory is key to good data quality. It is necessary to understand
what the laboratory is doing to ensure there is sufficient information concerning the sample
request. Often there is insufficient communication resulting in that are not appropriate for the
data quality objectives.
Every analytical method basically consists of a flow scheme that starts with the laboratory
receipt of the sample and includes steps for sample preparation, instrumental determination, and
ultimately, data generation. Although there are various ways of conducting the analysis, the
important thing to understand is that there needs to be sufficient information concerning the
laboratory's requirements to ensure that they have a proper sample, conduct the appropriate
procedures, and provide appropriate documentation.
Although the defensibility and documentation issue sounds like contract laboratory
program legalities, even a scientific study for peer review publication needs to have its methods
defended and documented.
Method 8080 (a GC/ECD method) is a method found in SW-846, which is the RCRA
bible. This method is used by thousands of laboratories conducting tens, maybe hundreds of
thousands of analyses. A verbatim quote from the method on how to conduct qualitative analysis
is: "A choice must be made as to which Aroclor or mixtures of Aroclors will produce a
chromatogram most similar to that of the residue." This suggests that there is a lot of latitude
in analyst's decisions. Another quote is: "This may involve a judgment about what proportions
of different Aroclors to combine to produce the appropriate reference material." It says nothing
about non-Aroclor PCBs, i.e., weathered samples that no longer look like an Aroclor.
Quantitative analyses often are not much better: "Measure the total area or the height
response from a common baseline." Analysts who have seen chromatograms know the difficulty
in defining what is a "common baseline." This is the entire set of instructions on how to
quantitate the results.
-------
Descriptive methods are traditionally what are found in the scientific journals. This is
the text description in the experimental section that describes how the person conducted the
analysis, and to adapt it in your laboratory, you have to take certain liberties. This is the
traditional professorial type of an approach to analytical methods. j
There has been considerable evolution over the years to prescriptive methods, such as
those found in the AOAC manual, the ASTM manual, and many EPA methods. This approach
prescribes the steps that must be taken. The premise is that if all of the steps are followed
precisely, the correct answer will be achieved. There are limitations to this approach. Quality
control is not integrated into the methodology. Inflexible methods discourage adaptations to fit
circumstances, local laboratory preferences, or professional judgement.
Performance-based methods, which have become increasingly popular over the last
several years, involve a strong attempt on the part of the methodology to provide soime flexibility
while providing a measure of the method's performance on each individual sample. The premise
is that it is not as important how one arrives at the answer provided the correct answer is
obtained. This is where isotope dilution, addition of surrogates, internal standards, and other
similar actions will indicate on every sample whether or not decent recoveries were; obtained and
whether or not interfering compounds or other factors that may have given a wrong' answer were
present.
PCB analysis is a complex and challenging area. Despite more than two decades of
development, methods are still lacking in several aspects as discussed above. The challenges
of providing routine methods that can provide good data to customers in a timely fashion at a
reasonable cost continue to face analytical chemists and research is ongoing on several fronts as
discussed elsewhere in this document. ;
3-2
-------
3.2 RECENT PCB RESEARCH
Ted R. Schwartz, Chief Chemist, National Fisheries Contaminant Research Center,
U.S. FWS, Columbia, MO
Polychlorinated biphenyls (PCBs) constitute a complex heterogeneous group having 209 possible
isomers distributed among Cl^o homologues. In spite of the concern about contamination with
PCBs since their discovery as environmental pollutants, much remains to be determined about
their ultimate effects and fates in the environment. This lack of knowledge is due in part to the
complexity of the chromatographic profile and the associated problems that must be overcome
in data reduction and interpretation.
The interpretation of analytical results from PCB residue analyses is challenging from
several perspectives: (1) data obtained from a single analysis are numerous (e.g., 100-150 PCB
congeners are often encountered in a single environmental sample), (2) source profiles of PCB
input into the environment are poorly characterized, (3) PCB congeners in polluting materials
mix with congeners from other sources, and (4) a PCB mixture can undergo alteration due to
metabolism and become partitioned into multiple environmental compartments that may be
further changed by weathering or degradation. A thorough understanding of these processes and
correlation of residue profiles with specific toxic responses requires congener specific methods
of analysis and increased use of multivariate statistical tools.
Development of an congener specific method that can provide detailed information on
environmental samples has been a goal of many scientists. However, after data acquisition and
quantitation, a most important step remains, data must be examined for quality control and
information content. The U.S. Fish and Wildlife Service (FWS) approached the problem of data
reduction and interpretation from a chemometric perspective. The SIMCA (Soft Independent
Modeling of Class Analogy) pattern recognition technique developed by Albano, Wold, and their
coworkers is based on derivation of disjoint principal component models. These models can be
used for graphical representation and classification of new samples. In the chemometric
evaluation of complex profiles, data from a single analysis are viewed as a point in multi-
dimensional space. Data from the analysis of many samples form a data cluster that may have
structure related to such factors as exposure or distance from discharge. A distinct advantage
of principal components modeling of multivariate data, such as those encountered in Aroclors
and PCB residues, is that data is presented graphically rather than in a tabular format which is
difficult to visually interpret. Provided that the data generated can be reduced and interpreted
the most common question asked of PCB residue analyses is, what is the lexicological
significance of the data? The approach FWS took to answering this question is to evaluate PCB
residues in terms of "dioxin equivalents." The analysis and interpretation of PCB residues in
fish provide a good example of using a chemometrics approach to data analysis by pattern
recognition.
3-3
-------
NATIONAL CONTAMINANT BIOMONTTORING PROGRAM (NCBP)
i
Since 1967, the FWS's NCBP has measured and reported PCB levels in freshwater fish as
mixtures of four common Aroclors. These measurements indicate that PCB residues in fish
continue to decline in those areas of the U.S. where levels have historically been the highest.
An inherent assumption in these measurements is that the PCBs in fish closely resemble Aroclors
with respect to their congener compositions and toxicity(s). To determine if PCB congener
residue distributions are compositionally similar to commercial Aroclors, individual PCB
congeners were measured in a subset of the NCBP fish collected in 1988. Principal component
analysis9 was used to compare PCB congener distributions in fish .to the PCB congener
distributions of four reference Aroclors.
Lake Michigan Lake Trout
In efforts to describe the role of persistent organic contaminants in the failure of hatchery-planted
lake trout (Salvelinus namaycusK) to become self-sustaining throughout the Great Lakes, a series
of studies were conducted between 1979 and 1988 to evaluate the quality of eggs taken from
spawning adults. During this same period, observations were made on a number of other
species found in the Great Lakes basin that suffered reproductive impairment due tb the presence
of organic contaminants. Some of these observations were detailed in a Workshop on Cause-
Effect Linkages (1991). Based on the observations presented in that workshop and the growing
laboratory evidence that PCBs, particularly the dioxin-like PCB congeners, interfere with
reproduction, archived egg samples were analyzed for PCB congeners. The composition of
congeners present in eggs and some adult tissue were examined to see how this composition
changes over time, among lakes and from adult to egg. Eggs from southeastern Lake Michigan
comprised the majority of samples tested, but eggs from other sites were used for comparisons.
Methods i
i
Twenty-eight composite fish samples from the 1988 NCBP collection, representing 26
monitoring stations (Tables 1 and 2), were ground whole and analyzed for 115 congeners. Lake
trout (LKT) eggs and sperm were taken from spawning adults gill-netted in Lake Michigan near
Saugatuck, MI in the falls of 1979, 1981, 1983, 1985, 1986, 1987, and 1988 and near Sturgeon
Bay, WI in the fall of 1987. Collections made from 1979 through 1984 were pooled samples
of eggs from several females while eggs collected from 1985 through 1988 were from individual
females. In addition, pooled samples of lake trout eggs were collected from Lake Superior near
Marquette, MI in 1987 and 1988 and from Lake Huron near Alpena, MI and Lake Ontario off
Yorkshire Bar in 1988.
Principal component models are bilinear projection models obtained by decomposing a class data matrix
X into a score matrix T (n x F), a loading matrix P (F x p), and a residual matrix E: i
X = Iox + T o P + E [Eq. 1] i
3-4
-------
Carbon column fractionation and high resolution mass spectrometry was used in the
analysis of those PCS congeners in LKT and LKT eggs that induce a number of enzymes
through the Ah receptor. These are referred to as AHH-active, referring to one of these
enzymes, aryl hydrocarbon hydroxylase. A column (1.0 x 23 cm) was prepared with a mixture
of 350 mg carbon (sized 2-10 /xm) and 4.5 g of Whatman GF/D filter material. The column was
characterized with analytical standards to determine elution profiles for AHH-active PCB
congeners, then the samples and control materials were processed with tissue samples to assure
quality control. After the chromatographic analysis, the data were arranged into a matrix suitable
for pattern recognition.
RESULTS
National Contaminant Biomonitoring Program
Total PCBs were lowest at NCBP station 46 (Columbia R.) and highest at station 109 (L.
Ontario). Relationships between technical: Aroclors and PCB residues in fish are displayed in
PC-plots (Figure 2). The dominant feature in Figure 2 are the Aroclbr samples, however, Lake
Superior (sample numbers 22, 37, 41 and, 62) and Columbia River (number 25) samples are
clearly grouped together, apart from the main cluster, farthest from the Aroclor standards.
Removal of Aroclors 1242 and 1248 (Figure 3) and then removal of 1254 and 1260 (Figure 4)
from the PC-model emphasizes the relationship between groups of samples. In Figure 3, all
samples lie above the line defined by a linear mixture of Aroclors 1254 and 1260. Fish samples
from Lakes Michigan (14, 43, 47 and 61), Lake Huron (49), and Lake Ontario (52) group
together and are separated from Lake Erie samples (12 and 51) and Lake Superior samples (22,
37, 41 and, 62). The distinction between samples from different lakes was difficult to
quantitatively assess because sample sizes are limited. However, it appears that qualitative
differences in the PCB profiles exist between fish from Lake Superior, Lake Erie and fish from
the other Great Lakes (Figure 3). It also appears that fish from the east coast sampling stations
have remarkably different residue patterns (Figure 4). These data suggests that fish populations
are exposed to different sources of PCBs.
These data indicate that clustering of Aroclor samples 1242, 1248, 1254 and 1260,
showed little similarity to tissue residue profiles. To report data from these fish as combinations
of technical Aroclors would misrepresent the true composition of the environmental residues.
The PCB residue profiles were examined further to determine the relevance of modeling a
reduced data set containing 18 PCB congeners which are measured in EPA's EMAP program
(PCB IUPAC numbers 8, 18, 28, 44, 52, 66, 101, 105, 118, 128, 138, 153, 170, 180, 187,
195, 206 and, 209). After normalization and extraction of principal components as described
above, plots analogous to those presented in Figures 2-4 were generated. The first plot shows
the relationship between the fish samples and the technical Aroclor samples, which dominate the
explained variance in the data set (Figure 5). Reduction of the number of variables did not alter
the conclusion that the fish samples are not similar to Aroclors 1242 and 1248. However, the
EMAP data set suggests most of the fish samples lie along a line described by a linear
combination of 1254 and 1260. Removal of Aroclors 1242 and 1248 from the PC-model fails
3-5
-------
to emphasize the relationship between groups of samples (Figure 6). Most samples appear to
lie on the line defined by a linear mixture of Aroclors 1254 and 1260 with the exceptions of
sample 28 (Merrimack River) and 5 (Hudson River). The distinction between Great Lakes fish
samples are much less clear (or lost) with the EMAP 18 measured data set even after removal
of Aroclors 1254 and 1260 (Figures 7). These data suggests that the fish populations were
exposed to similar sources of PCBs, which is contradictory to the 115 congener data set. The
information loss resulting from the reduction in measured variables is potentially very large.
This loss of information must be considered, if comparisons of sample composition are an
important objective of the problem under study.
Lake Michigan Lake Trout
'
Chemical concentration profiles in tissue and eggs from Lake Michigan, Lake Huron, and Lake
Ontario form a distinct cluster which differ from Lake Superior samples and the commercial
Aroclor mixtures (Figure 8). The cluster representing samples from Lake Michigan indicate no
significant change in PCB composition over the 9-year sampling period for LKT eggs and the
2-year period for adult female LKT. The distinction between samples from different lakes was
difficult to quantitatively assess because sample sizes were limited. However it appeared that
qualitative differences in the PCB profiles existed between LKT eggs from Lake;Superior and
LKT eggs from the other Great Lakes, which suggests that the two populations were exposed
to different sources of PCBs. Also, the clustering of Aroclor samples 1242, 1248, and 1260,
showed little similarity to tissue and egg residue profiles. A second PCA was restricted to s
subset of the samples (Figure 9) which included adult tissue samples for which we had
corresponding egg samples and Aroclor 1254 as a reference point due to its close, proximity to
the tissue and egg cluster in Figure 8. Projections of the samples for the first two PC indicated
markedly different clusters for adult LKT and their eggs, and Aroclor 1254 (Figure 9). These
two figures indicate that the chemical profile in adult LKT tissue and their corresponding eggs
is different, and is unlike any technical mixtures of Aroclors. The difference in chemical residue
profile suggests some selective deposition of PCB congeners from adult to egg. Examination
of gas chromatographs revealed a greater abundance of lower chlorinated PCB congeners in the
LKT eggs. |
Examination of AHH-active PCB residues were unable to distinguish between adult fish
and the LKT egg groups of samples. This indicates that, even though there is a greater
concentration of AHH- congeners in the adult LKT, the composition is the same as in the LKT
eggs. It appears from the concentration data and TCDD-EQ data in Table 3 that there is no
relative enrichment of these PCB congeners from adult fish to egg. Even after lipid
normalization (Table 4) no clear indication of enrichment of AHH-active PCB congeners is
apparent.
3-6
-------
CONCLUSION
The PCB congener distributions in all NCBP samples analyzed are significantly different from
those of common Aroclor mixtures. Samples from stations 24 and 111 have PCB congener
distribution that are most similar to those of common Aroclors. PCB congener distributions for
samples from station 3 (Hudson River) appear to be similar to those of Aroclors only when
samples and Aroclors are analyzed together.
Samples from Lake Superior stations 22, 102, 103, and the Columbia River station 46
are the least similar to Aroclors. Others have suggested that Lake Superior fish receive PCB
contamination mostly from atmospheric sources to the lake.
Application of pattern recognition by SIMCA characterized the profiles of PCBs in a typical
environmental situation and statistically showed that residues in LKT tissue and LKT eggs from
various locations differ in PCB congener composition. No difference in congener composition
in eggs over time were observed, indicating that the more toxic AHH-active congeners are not
being selectively accumulated or enriched in lake trout. Differences in the composition of all
115 congeners between females and their eggs were observed. Egg deposition appears to select
for lower chlorinated congeners, but does not influence composition of AHH-active congeners.
.Because the residue profiles cannot accurately be represented as Aroclors, and total PCBs was
the best correlate of biological effects in eggs, PCBs should be expressed in terms of total PCB
concentration and calculated by summing all individual congeners.
3-7
-------
Table 1. Total PCB Concentrations in Selected NCBP Fish Samples Collected in 1988.
Station1
2
3
4
8
18
19
20
21
22
23
24
46
52
52
53
54
66
68
69
70
70
102
103
104
105
106
107
108
109
111
Species2
CHC •
WSU
CHC
CHC
WSU
WSU
C
LT
LT
CHC
C
BRB
WSU
NP
WSU
WSU
WSU
C
C
C
CHC
LT
LT
LT
LT
LT
C
C
LT
C
River or Lake
Connecticut R.
Hudson R.
Delaware R.
Cape Fear R.
L. Ontario
L. Erie
Saginaw Bay
L. Michigan
L. Superior
Kanawha R.
Ohio R.
Columbia R.
L. Champlain
L. Champlain
Merrimack R.
Rarrtan R.
St. Lawrence R.
Wabash R.
Ohio R.
Ohio R.
Ohio R.
L. Superior
L. Superior
L. Michigan
L. Michigan
L. Huron
L. St. Clair
L. Erie
L Ontario
Mississippi R.
Locations
Windsor Locks, CT
Poughkeepsie, NY
Trenton, NJ
Elizabethtown, NC
Port Ontario, NY
Erie, PA
Bay Port, Ml
Sheboygan, Wl
Bayfield, Wl
Winfield, WV
Marietta, OH
Cascade Lock, OR
Burlington, VT
Burlington, VT
Lowell, MA
Highland Park, NJ
Massena, NY
New Harmony, IN
Cincinnati, OH
Metropolis, IL'
Metropolis, IL
Keeweenaw, Ml
Whitefish Point, Ml
Beaver Island, Ml
Saugatuck, Ml
Alpena, Ml
Mt. Clemens, Ml
Port Clinton, OH
Roosevelt Beach, NY
Lake City, MN
Total PCB 3(ng/g)
1170
2669
1166
', 480
1292
444
743
I
I 2263
300
2119
3577
37
I
69
174
608
1458
212
623
2545
1342
! 1547
3696
546
I
1103
2174
I
1720
3939
624
4624
1969
1. Stations as designated by NCBP and shown in Figure 1.
2. CHC, Channel catfish; WSU, white sucker, C, Cannon carp; LT,
3. The sum of individual PCB congener concentrations.
lake trout; BRB, brown bullhead; NP, northern pike.
3-8
-------
Table 2. List of selected 1988 NCBP fish samples and Aroclors used for Principal Components Analysis.
Object
No.
1-4
5-8
9
10
11
12
13
14-21
22
23
24
25-26
27
28
29
30
31
32
33-36
37-40
41-42
4346
47-48
Station
No.
2
3
4
8
18
19
20
21
22
23
24
46
52
53
54
66
68
70
70
102
103
104
105
Description Object
No.
Connecticut R.
Hudson R.
Delaware R.
Cape Fear R
L. Ontario
LErie
Saginaw Bay
L. Michigan
L. Superior
Kanawha R.
Ohio R.
Columbia R.
L Champlain
Merrimack R.
Raritan R.
St. Lawrence R.
Wabash R.
Ohio R.
Ohio R.
L. Superior
L. Superior
L. Michigan
L. Michigan
49
50
51
52-53
54
55
56-57
58-61
62-65
66-69
70-73
74-77
78-82
83-86
87-90
91-92
93-95
96-97
98-100
101-102
103-105
106-107
108-110
Station Deccription
No.
106 L Huron
107 L St. Clair
108 L. Erie
109 L. Ontario
111 Mississippi R.
52 L. Camplain
69 Ohio R.
L. Mich. Composit
L. Superior Composit
A1111 (0.1)
A1111 (0.25)
A1 111(0.5)
A1111 (1.0)
A1111 (2.0)
A1111 (5.0)
1242 (0.8)
1242 Matrix Spike
1248 (0.8)
1248 Matrix Spike
1254 (0.8)
1254 Matrix Spike
1260 (0.8)
1260 Matrix Spike
3-9
-------
co
SH
0
o o
.rt • 0
n °* iJ 1
0) „, < !
& "5
4^ 0 »iH ^
C/> £}
£ V£>
CN . 0 o •
CN *° J O *"* '
CM i. 0 ^
TJ« r~ ^ «-•
ro cs ;
n o\ &
in
CN 'H
04
&4
W "m
m m *° S
vu ^ m to
rH rH -H
.Q H m m H • «
"S w « m H *
'H rH CM M f. S
in CN
K., oo Tt< o i o
_. i-l CT> CN CN m
8 ° * S t
* 0 «
rH in
LO t-H CN
H n H
^ s^
ro O
J2 t)
W P
•H S ' <
fa
in
CO
CO ^
C\ oo
t~» ^* *
CN CN 04
& M H CJ
CQ H HO
O o to
fe J-l rH *O
S 0 CJ 3
r-1 0 W
U H
o «;
CN H
(D 4, vo
}-l en
en o!
-H
IV. , — — — — — —
in
a\
M
•
co
CO
04
CQ
U
4J
®$
O •*
Qj
i3 *rj
0 0)
u a
•H
H Id
Id rH
•H X
U W
a
•H (I)
M U
04 a
nt
1i -H
1™ 1 ^;
id
^
(%8T
3-10
-------
o
•H
0)
3
CO
OJ 03
03 if •* VO
rH
to
d)
rH
•8 a
•d 1
n$ &
>Q
j_l
PQ 5
u \
04 -a *
in -H
H S
H .
* H
* vo
-ES 0
03 ro
fa 21 5!
*3* f* i
oo in
oo m
H
CU
PQ
U
rH
ro
.
ro
® oo
y 03 co
P rH
tn
•H
CL4
p^
CO
o
vo
OJ
in rt
OJ 5^
O
rH
0
a
rg *
03 rH (j) ^
•*> -H o
* *-• rH
W
OA
*f
ro H 01 ro rH
•* in ro in
00 rH VO
"S* c- H in
t^
OJ
o ^
ro in oj
O3
o
H cr>
O3
s <*
in
OJ
rH
H }H
O *- O
H rH
O
o
-------
er>
03
oo
CO
CM
CQ
U
53
CQ
0
tH
,Q
Ctf
•H
PQ
U
in
H
H
CQ
•H
in
eg
04
0)
^
OS
0)
D
CO
in
i-i
o\
o
ro
H CO
in
in
n
in
vo
in
4J
a ^
0) oV
•0
O Q)
u a
•H
r-l Rt
rt i-i
•H X
U W
-H 0)
}4 U
o< a
04
CQ
CO
CO
en
H
0)
O
nS
-H
M
i-l
0)
CO
CJ
(U
d
0
CO
•o
3
o
in
OJ
o
H
in
•H
i-l
tn
•H
fa
If)
3-12
-------
CQ
0)
rH
a
•rl
(C
!>
co
rH
CQ
Cn
§
W
0
EH
in
rH
£
0
fci
TJ
0)
0
3
Tj ^3*
0) m
OS f
H
CQ
X 0
Xl •-»
ns o
•H O
Si rl
(C <
CO
00
0\
rH
0)
CJ
s
.
in
0)
w
Dl
-H
PL.
eg
M
0> H
CN] LJ
H O
n "o
O o
1 a
'j Jf
t H
vo
en
.
S
-
tti
O
ns
e
•H
^1
^1
>, . 1
m DS
"• c *
• S O co
K rtf CQ ro
C -0 "
•H -H 3
a 01 EC
di re
•H CO J,
CQ v
CQ J, m
m 2
ro rH O
•H n H
4> eg o JH
in O
^ ro "fH erf
in ca ri
vo a) re
r> m a -H
CM »* m rH CO 6
<* <»1 rH H 3
rH in rH • rH O
CM J O VD
ro c~ in o U CN
•^ ^ ro i
m o\ A 1
Hcom CNOJ »» ^
u> '* ro H cs in o
C- «3 CM r-H
m u
rH VD O
•< o <- vi
H <
CM
eg
EH
X
cu
Ed
•u,-*
Si **
0) 00
w If)
Pi-O
S co
u ^
^*
IU
rt *o
Oi x
•rl H
o
« * * **
"^
3-13
-------
0}
i CO
ns to
CQ *I~1
to
S 03
(0 -rH
c s
•H
Cn
fd J/
CO
* s
rH
in
t H
m
Crf
| 0
TJ ro
w ^ N
CSl
en en
O i* CM
in
P> CO rH
VO ^J1 **}*
in rH
rH VD in
CM rH in
in ro
rH POO
in ro n
H
en
CM
CM CO
0 rH
U)
H
H
M l> O
0 n -H
rH H
o < H CM «- W
CM
in
rH
M
0
rH
O
O
M
f£
I
rH
O
H
Pi o\
O in
O 0)
u a
H rt
(Tj rH
U
ti
•rl 0)
k U
4J
to
3-14
-------
CO
n)
-rl
OQ
K
CO
H fi
m co TO
og r«
§ t •?
ft! .
W 4
0 oi in
E-<
^*
if) O
rH to
-H
O fc
to 0)
TV. *S^
tJ
0)
0
Qi
0)
0)
r-H
ffl
•H
8
g£
»G ' (0
in -H
-H Q
3
CO rH
CO O
a\ U
CQ M
U O in
S3 -H oj
(U M
Ql •» rH VO
r- 3 •*
03
0)
3 »4
tJ) 03
-H OJ
(0
c
•H
S1
TO
^
H
ro
^*
in
o
H
ro
o oj
a\ in
og
OJ Of\ CO
VD ^*
in <
rH H i-H
VO rH
o in
ro co
in in
H in
ro
cr\
OJ
ro
og
H
r-
ro
S
X
t
X
o<
^
m
4J
Old?
rt o
0 to
0)
u a
"ri
rH tS
id •-»
Pi g.
•n| p^
U Ed
5«
M U
s
15 -H
H «
>
(%ET
3-15
-------
m S
°> in in in vo
r- i- r-
co co co
*-. CO CO CO CO CO
OS ^^ ff\ r!P in *J* t*"
d . £i JL !ii •**. at
tH »?C «+C ^+J •**• ^^
•GI *H IJfc
l2l ** >
b*l ,*. 01 oi oi tn 01 2*
Ol 3 3 3 3 3 3
."I CO CQ CQ CQ CO CO
CO
i« in vo c- co o>
£3 m n n n *"* "
to
Dl
D)
w
___.. *** CF\ CF\ ^ r*{ *^ __
%3 *^ C^ t^ CO CO CD «
c ^
rtO«SS«S ^
* & ~ ~ _ t
•*-* 7^ S ta B « m <5 3|
3t».COtQWCQCQCO CQ
^*^ iH c$ ff) ^i in vo ^*"
J£
« t-
^j*
a *
•rl "
(3 ™
0) CJ
S H
T3
•f^ ^
S °
§>»>§'§>§> Soo
nSfd(d(dt0njm iii]^iM
CQCQCQCQCQCQCQ < <
OiHCJP)rj«mVD P-CO
^•^•^•^^•^TII ••a'Ti"
cococococococococococo
o
CJn^inp-cocniHHcjn
OiOiOiO)DiOiO)O)tnO)Ot
33333333333
(u(Qf9(on3fon)tTt(Qn5nl
CQCQCQCQCQCQCQCnCOCOCO
OHcjrOTi<
CJ
o ^<
iH f>
n
N
CO CO ^j* ^* O O
^* ^t in in vo vo
£} cj cj cj cj cj
^i ^i j^ j^ ^ ^
^ 0 0 0 O O
0 0 O O 0 0
p o o o o o
£ H M H Vj M
^ rij *3! raj rfj ^
**> ' o H ca ro 'J'
t in in in in in
co mcococomcocococo
CJ
| B'B'B'B'B1!'!1!'!1
CQ COCQCQCQCQCOCOCOCO
<" OHdro-vinvoir-co
••* CJCJCJdCNCJCJCICI
o
a\ m
^ i
CO
CJ
H
O
o
o
a
i
'
CO
CJ
vo en
CJ
^* p* ro r*
ro
ro co
m H in
CJ
CO
H H O
CJ CJ
CJ d
<* ^ cn
H
in cj vo
H 1C «*
VO H
n in
«H ca
0 CJ VO
^* n
in
VO P)
**• co m co
CO CO CO CO
J^ (,
O o O
•H ,f\ .|-(
| 8 Bi3
§• §• ^
CQ w a o ^.
^ j j
^ O
OJ 0 HCJ m
*^ co rn n
•o
0)
•H
in
,_!
rH
O
S
<,
CJ
in H
in
(0
o iH
S g.
* IfJ
fer PU
m
H Q)
£ °
ti
§
-r4
(4
Ml
TO
t*
•^
4.)
§
0
ft
g
HI
frt
a
•H
U
rt
'^
C i
PI
u
a
H
(%6T
3-16
-------
ffl *-
>-r\ CO
5J 10 vo r- p- p- r-
o> ~ co co co co co — .
^ TO|^,r~H<
c
•H
(Q
r-
cococo. — — *cococo ^^*
«-<««*» in in
S c2 S. t, N ^
•ft- ?ft. ^J; f\J "™* f>J ^f. -4». ^4- O
* * * 0 0 0
3tnCnOitnJ,tjiGiOicn *joo
nS333§13333 < O O
comoH^m^inu) rHw
^c in
n
5 S 5 5
•* <
o\
<0
I
•H
o
en
H
H
O
I
4J
(j
a) «->
vo
•0
0)
a
•H
RS
O
O
a
•H
LJ
rt
01
*>
a
H
W
•
ri
in CN
in
3-17
-------
1
o
I
i
ra
4~»
i
CO
_J
.S
a
i
o
m
o
X
I
o
I
CO*
$** O3
1 2iS
S ^^
S "i£
•» Ol
8 2 a
S LU
Ssi
ri III
n
o ns
§ ""* LU
«? j-
8 ^iT
n
o. a
§ S iS
CO
^ ?i
cn
I*J
CJ <•? .«
e*i t * ^^
i/"l \O fe
S -iB
~ -C
co *«j* «••
r>
Q *»
CO
c^ r*^ c^ c^ r^ in v~* ^M c^ f™ in in co 03 TJ*
t— »— ^*"co^t~"^ ^^"in^coo^o
d v
co »-* o o o o oo cooo^—p^oic}.
-10 <- CO t-CD •*- v
C^ CO 00 CO CO CO 00 ^^ ^^ *^ ^^ CO CO Ii5 ^J
h T-: N CM <° «-: 0. "° ro co ^ in d eo' d
o o o v
CM C^ CO CO C^ CO C5 C^ C5 lO lO O5 *™" UC5 "***•
* * ^O ^^ CO / _^* ^^ OO T«* _.J ^M * ip" ^^
CMCO ^OOj^^tCM T-'t °° *~
coco CM^CDCM'I^ t-incD^rvCM-*^
do d dd^dd..
V
mo o T" o o> co ooor^inr^oco.
OCO ^}'__'~O r>j *i" '"l^
V
CMOO comT--Or'^O
o o v
r^r-* ocoo»-o oocMt'*cocM'*
**»C OCMf^^_^ ^J'O^'CMCO-iI^'O
T~t" coco^ coco ~
cor- cMoot-int^ coincoo>r-cMCO'<*
6^ *~ dn d^~. '~'*cMto. co'". CMO
o o o o v
V
COr— CDOOCMOO COOO«"~*— f^-CO.
•_; r~s«co^; • COCMT-,.;CM»^— : o
T-TJ- 'CD~COT— ~j COOCD"'
v
cawfi ^3*OTt"cM« ^cMcoiolnScocoffl
O Or-r-T-T-T- oT-T-T-T-r-T-T-T-
o. a. a.
II 1
00 0
CM •* O
- m. CM
0
•<±
CM *-
O 00
CM !>.
CM »-
CM U)
* T-
O CO
*"
r-~ f~«
"t CM
r- r-
m T—
t- 0
CO O
T- CO
O) CO
CO T-
t u>
r~- CO
CO r-
CM CD
CM «-
CM
T- en
o **
co in
CO CM
CJ
$ cc
S %
o &
1 1
t* t*
•o
'C
o
to
to
•a
to
CO
•a
0}
w
£
CO
5.
0
§ S
•* "5
CD 4^
~
£ £
CO 00
o c
— T3
1 .§
** to
1 |
c E
I 3
>£ fl)
O
o c
o) **"* £
11 |
as a> *-
W Q} "S
2s S
.co to §
n CD
^ o ^j
O •£ r-
B S §
» E ,8
2^ CD . __
CO ""ti CO +J
u. .2 co .5
"" E cn^J
"3=0
II "Q CO
c .2 S ti
— • **Z CO C
£ o -a -2
3 03 c 5
1 s-10-!
olSl
O> c R>
2 to 3 Q
O5 5- ^S
o 21 £ u
^c!«"
CO A U
CO
o
CO
to
to
ra
to
in
tu
DC
as
£
a
N
t:
R)
U
o
R)
3-18
-------
g>
O>
CD
0
!c
"D
C
(0
D
£
G)
3
_c
'"*
^
n
CD
C
CD
O)
C
o
o
CO
o
0-
CD
1
i
X
H—
0
c
0
*5
CO
i
CD
U
0
TJ
.§
15
O
Z
!•
.
^j*
.2
.Q
CO
1
in M S
S "* UJ
N
<- *T j-
oo en *"
R **•
in
•"0§
CO
"i"i
«n
(N
§ c
*** a
r- OO _
i f_
•<*
in QO ;
isiS"
to
&^ K
CO
1-1
^J.
"iti
en "**
•f
<3- r» co co •* i-
CM 00
T- CM
oo oo oo T— in co ^f
r- O CM 60 T-
CO !••
co co in r^ co in co
^- <* CM «- T-
CO 03
'in •r- r*. ;: m 't co co to •.:
CO cok ffl O^-T-CMCM iS
a. a. a.
tj- 1ft SO
00 0
en r^ Tj- T— cocoon
to 10 om t*. o o
'—CM y
§CM O CM T- CD O "^
^- CO OO r- t- tO O
o t— co r-
t- in V
coooinTt ocoeo**
Oi^tt>'" ^ .*" ^ C>
"~ ^" V
«•• co co CM o r* co .
^* co o> in r^o co o
'"CM *" V
O O oo r^» in CM CM "^
in CM "~ ^ °
V
«-»-OOCO COtOCM'*.
in co O) CM co •* o
* 2 V
SoScM CM^S"*'
oo r>- T- °
*- v
cocor^r- o in T— ^t
en r- co in co ao o
in co T-
CM V
IO IO 00' ^^ CJ CO ^^ ^J*
00 CO CM in *" CM °
CM (O V
«-coeocM oooooo"^
*t to •<- i- eo *~ »- . .
T- in V
OOUJIs.,- t^gCOQ
«- V
V
co in ID eo in co CN "t
oo in Tt CM co "4* O
cM ^^
«- V
cooocor«.oocoi^eo
CMCoininincoco® QJ
T"~ T"" ^~ • ^J
a.
r*
O
m r^
«sj-
CM
•* 00
in »-
in tn
CO
r—
•<- O
CN
O •*
CM
1™ *
o in
in
to in
CD
^*
00 CO
CM
to «-
in CM
CO
00 O
O CM
tn
CO CO
i- CO
«—
O eo
r^ oo
CM
00
in
CO
00
CO
CO
0
CM
O
CO
CO
CO
00
00
in
£
CO
_£
1
M
O
61
1
O
00
CO
CM
O
CM
CM
^
CM
CM
cS
-
fj
,_
T"
in
m
<~
en
00
CO
o
CM
CM
CM
O
00
CO
w
-a
5
w
ci
w
H
1^
t_
CM
^
O>
CM
CO
in
CM
*
CM
*
00
CO
to
£
1
&
u
d
w
p
•D
CD
J3
-§
•a
a>
ca
3
CD
O
CA
ca
J s
CO
= s
ca
(0
c
™ g
.1
CO
I
w
_ CD
CO CD *~
* C 1
f I I
•g« c
3 CD C
tS E °
ra m *•
Hrt
i*Sl
i
c 5 *
CO
ro
o>
I
a
c
CO
co
£
cc
m
CO
-------
3.3 FDA PCB ANALYSIS
Leon Sawyer, Branch Chief, Methods Research Branch, Division of Pesticides and
Industrial Chemicals, Center for Food Safety and Applied Nutrition, U.S. FDA,
Washington, DC
The Food & Drug Administration (FDA) monitors for pesticides and PCB residues in food and
feeds for food-producing animals in approximately 15-20,000 samples a year. In FY'92, this
included 1148 samples of seafood, the commodity in which the highest incidence Of PCB finding
occur. In addition to the monitoring activities, FDA conducts a total diet study program, where
food is collected from around the country, prepared "table-ready" and analyzed for pesticides
and PCBs. Historically, the FDA determinative step for PCB analysis is the packed column gas
chromatography using halogen-specific electrolytic conductivity detectors or the selective
electron capture detectors. Results are reported as "total PCBs" by pattern matching with
specific lots of Aroclor(s) as reference standard materials. The quantitative procedures used
have gone through rigorous interlaboratory testing by the AOAC.
The AOAC has had four PCB collaborative studies related to quantitative procedure. The
first study in 1973 was conducted to show that PCBs could be accurately quantitated by "pattern
matching" with Aroclors and to show that PCB/DDT combinations could be dealt with. The
second collaborative study in 1974 addressed both the extraction and quantitation of PCB
residues in paperboard. The third collaborative study demonstrated that "highly altered" residues
could be analyzed by packed column chromatography using a Webb & McCall'approach that
uses individual peaks and calibrated Aroclor standards. The last study in 1989 was conducted
to demonstrate that PCBs could be accurately analyzed in blood serum using a variation of the
Web & McCall approach. ;
Commonalities of all the quantitation methods in the AOAC are they all use packed
methyl silicone columns, electron capture detection, and some method summing the "total" areas
or peak heights measured from the residue against similar responses with matching retention
times obtained with specific lots of Aroclor reference materials. In the Webb & McCall
procedure, "total" PCB is arrived at by summing individually calculated ppm values.
Demonstration of differing toxicity among congeners and development of high resolution
capillary chromatography procedures that can establish congener identity, provided a legitimate
rationale for monitoring PCBs on an individual congener basis. FDA has not done so and has
been criticized for not using this advanced technology. FDA rationale for following "archaic"
procedures is: (1) FDA is a regulatory agency and all published tolerances are on "total" PCBs.
(2) "Official" methods are available which, for regulatory purposes, is very important. (3) The
"Official" methods are compatible with FDA's commonly used pesticide monitoring procedures.
(4) Resource considerations do not deem it practical to equip several laboratories throughout the
country to conduct individual congener analysis when most residues that are encountered, mainly
in fish, actually have a chromatographic profile similar to some type Aroclor(s). The exceptions
would be a residue extracted from an internal organ such as lobster tomalley or a residue
isolated from a product like milk that has been altered by the digestive pathway. (5) From a
3-20
-------
public safety viewpoint, it seems prudent to report "total" PCBs than individual congener
concentrations since the question remains on which congeners are of concern—all observed
toxicological endpoints cannot be attributed to those that demonstrate AHH activity.
A point that should be clarified about FDA PCB findings is that "total" PCBs are
reported, contrary to an existing perception that results are "Aroclor" findings. Aroclors are
used for reference standards and the identity of the reference standard(s) is/are reported to
provide information that may be useful in either; (1) tracking the source of the contamination,
or (2) providing some hint if higher or lower chlorinated congeners are predominant in the
residue. Aroclor reference materials are used for analyses of PCB residues when it is known
an Aroclor was not the source of contamination, e.g., residues isolated from samples of foreign
origin.
Many approaches to individual congener analysis are being used and several have been
published. However, most of them target just a select few congeners and do not attempt to
measure or identify them all. In fact, some laboratories use as few as 2 congeners to report
quantitative results. Use of these various procedures leads to confusion when comparing
interlaboratory results from identical or similar samples. As an illustration, in a laboratory
comparison study of identical lobster tissues, results varied almost by a factor of 2 when
different methods of quantitation were used.
Criticism of using archaic methodology and generating poor risk assessment numbers for
PCBs, prompted FDA to initiate an individual congener analytical study. The method chosen
to study was that described by Dr. Mike Mullin of EPA's Large Lakes Research Station in
Grosse He, MI. Dr. Mullin, along with Dr. Stephen Safe, synthesized all 209 possible PCB
congeners and then used them to characterize the congener content of several Aroclors. The
Aroclors, once characterized, could be used individually or in mixtures to serve as secondary
standards for determining congener identity and content from real world samples.
Interest in obtaining comparative "total PCB" values using the Mullin approach and
traditional packed column procedures was not limited to FDA. Laura Maack and William
Sonzogni, published comparative PCB results for 18 Wisconsin fish samples calculated by both
procedures. The overall grand averages of the 18 samples were 1.18 ppm by summing the
individual congener values and 1.05 ppm by the traditional packed column procedure. The
authors' found the 2 data sets were correlated within a linear coefficient of 0.9854. FDA
confirmed the close agreement between the 2 procedures with fish samples that had been
previously analyzed in the field and reanalyzed at headquarters. The FDA headquarters study
was limited to 6 samples and the overall grand averages were 1.58 ppm by individual congener
summation and 1.50 ppm by packed column.
In conclusion, the Mullin procedure for individual congener analysis appears to be both
a good qualitative and quantitative analytical approach that could be used for "regulatory"
purposes where "total" PCBs are a concern, and for future risk calculations where specific
congeners may be of concern. However, considering the resources involved and the declining
3-21
-------
numbers and levels of PCB residues, FDA is encountering in its pesticide monitoring and total
diet programs, there does not appear to be a justifiable reason to change current procedures.
This does not rule out the possibility of doing limited congener investigations in cases where
"high" residues are found or, resources permitting, doing limited surveys.
3-22
-------
FDA — FY'92 Seafood PCB Findings
Total Samples 1148
No. Positive
209 (18.2%)
Violations
Range"
< 0.1 to 3.6 ppm
'21.3 ppm in one lobster tomalley70.76 ppm in flesh.
3-23
-------
AOAC/PCB Collaborative Studies
• Collaborative Study of the Recovery and GC Quantitation of PCBs in
Chicken Fat and PCB-DDT Combinations in Fish \
JAOAC 56, (1973) 1015-1023
i
• Collaborative Study of the Determination of PCBs in Paperboard
.
JAOAC 57, (1974) 518-520
i
• Quantitation of PCS Residues by Electron Capture Gas-Liquid
Chromatography: Collaborative Study
JAOAC £1, (1978) 272-281
* Gas Chromatographic Determination of PCBs (as Aroclor 1254) in
Serum: Collaborative Study
JAOAC 72, (1989) 649-659
AOAC/PCB Quantitation Methods
• Packed methyl silicone columns
* Electron capture detection
• Select reference Aroclor(s) by "pattern matching"
• Comparisons for quantitation - sample residue against Aroclor(s)
.
total area, total peak height, individual areas - summed, and
relative responses vs. DCB
• Measure "Total PCB"
3-24
-------
FDA Rationale for Method Choice
• Tolerances are on total PCBs
• "Official" methods are available
• Compatible with pesticide methods
• Resources
• Most residues similar to Aroclor(s)
• Aroclors are references—not findings!
• Safety
Laboratory
PCBs (ppm) in Lobster
[Aroclor 1254 Reference]
Tissue
Tomalley
Labi
Lab 2
Lab 3
1.9
2.2
2.0
11.6
10.2
8.0
1.2
1.0
1.1
5.7
5.7
4.6
1.2
1.2
1.2
7.1
7.4
"•"
Capillary GC; 2 congeners compared
Packed column; AOAC
Packed column; 4 maioi
"total area"
r peaks compared
3-25
-------
Comprehensive Congener Analysis
• Synthesis of all 209 congeners (primary stds)
• Determine congener content of Aroclors
• Aroclor(s) used as secondary stds
• Internal stds for response factors
• Compare response factors for quantitation
• Results: "Individual" or "Total" basis
Analysis of Aroclor 1254
EPA Reference Aroclors:
1221:1016:1254:1262
(20:10:7:6)
Internal Standards:
PCB congeners 30 & 204
Results:
EPA 1254 97% of formulation
FDA 1254 106% of formulation
60 m x 0.25 SPB-1 temperature programmed: 3 hr. chromatographic run
3-26
-------
PCBs (ppm) in Wisconsin Fish
L. Michigan
L. Michigan
Wise. River
Whitefish
Chub
Chub
Chub
Chub
Chub
Chub
Chub
Chub
Chub
Chub
Chub
Sturgeon
Sturgeon
Sturgeon
Sturgeon
Sturgeon
Sturgeon
Average (18)
Capillary
0.52
0.66
0.78
0.92
0.95
0.96
1.1
1.1
1.2
1.2
1.4
1.4
0.57
1.0
1.1
1.4
1.7
2.4
1 11
Packed
0.52
0.48
0.59
0.70
0.80
0.70
0.83
0.70
0.83
1.0
1.2
1.3
0.67
1.2
1.2
1.7
1.9
2.5
I.IK
Reference: Maack, L., Sonzogni, W.C. (198.8) "Analysis of Polychlorobiphenyl Congeners
in Wisconsin Fish" Arch. Environ. Contam. Toxicol. 17, 711-719.
"In general, good agreement was found between the two methods (the data
were linearly correlated with a correlation coefficient of 0.9854)"
"These data suggest that total PCBs obtained by summing the individual
congener concentrations in fish may be comparable to values obtained by
traditional packed column techniques. "
3-27
-------
FDA Comparative Analysis
(ppm)
Packed
Capillary
Field
CFSAN SPB-1 SB-Octyl 50
Bluefish
Chin. Salmon
Coho Salmon
Coho Salmon
Chin. Salmon
L. Tomalley
Average (6)
0.19
0.77*
0.89*
1.0*
1.3*
5.5
1.61
0.16
0.57
0.84
0.92
1.3
5.2
1.50
0.17
0.63
0.92
0.94
1.3
5.5
1.58
OJ23
0.65
0.96
0.75
1.5
taj»~
i
* Widebore capillary operated in "packed" mode by total area.
3-28
-------
3.4 PERFORMANCE-BASED METHODS
Margaret M. Krahn, Sin-Lam Chan, and Usha Varanasi, Environmental
Conservation Division, Environmental Chemistry Program, Northwest Fisheries
Science Center, National Marine Fisheries Service, NOAA, Seattle, WA
This presentation discusses the methods used by the Environmental Conservation Division of
NMFS/NOAA to analyze for PCBs in fish tissue. In addition, the reasons these methods
were selected for use in many of NOAA's research and monitoring programs are outlined.
Before an analytical method can be selected, the particular analytes to be determined
must be established, as well as the end-use for the analytical data. For example, two of the
major uses for PCS analytical data are for regulatory and research purposes. The regulatory
category is used in a general sense to include seafood safety. Several federal agencies
require PCB analyses for a variety of end uses. For example, the EPA uses PCB data to
regulate environmental quality and the FDA and NOAA to regulate seafood safety. NOAA
also uses contaminant data to assess damages to natural resources and to support litigation for
subsequent restoration of marine habitats. In addition, a number of groups conducting
research study the relationships between concentrations of PCBs in fish and shellfish and
environmental quality (NOAA, EPA), the health of marine organisms (NOAA), seafood
safety (FDA, NOAA) and human health (various agencies).
The regulatory agencies often have different needs for data than do the research
groups. For example, some questions that may be of interest to the regulators are: (1)
Which Aroclors are being released into aquatic environment? and (2) Can a source of
contamination be determined from the Aroclor patterns in fish or shellfish? To answer these
questions, PCB concentrations are determined and Aroclor patterns are identified in water,
sediment, or fish. These data can then provide information necessary to identify for
regulatory agencies the amounts and the sources of the Aroclors and possibly those
responsible for the contamination.
Researchers may ask different sorts of questions: (1) Are particular congeners
preferentially bioavailable to fish? or (2) Which individual PCB congeners accumulate in the
largest concentrations in fish? These question may be answered by analyzing for individual
congeners in fish for comparisons to sediment concentrations or to indices of human or
animal health. With the data obtained, links between PCB contaminants and various
deleterious biological effects may be pursued.
NOAA established the National Status and Trends Program (NS&T) in 1984 to
assess and document the status of and the long-term changes in the environmental quality of
the Nation's coastal and estuarine waters. Two major projects were included: Mussel
Watch, which is conducted by contractors outside NOAA, and the National Benthic
Surveillance Project (NBSP) which is conducted by NMFS. The original objectives of the
NBSP were to measure organic and heavy metal contaminants in sediments and in tissues of
bottom-dwelling species at selected sites in US coastal waters and to determine the
3-29
-------
prevalences of diseases as related to chemical contaminants in these fish.
In 1984, at the inception of NS&T, methods that analyzed for PCBs and other
contaminants in tissue samples were excessively laborious and time consuming. Gravity-flow
columns used large quantities of solvent and other materials. As a result, we soon launched
an effort to develop rapid, automated methods to replace these laborious methods (Figure 1).
Procedures for tissue extraction and instrumental analyses were modified only slightly. In
contrast, two gravity-flow columns that separated analytes from biogenic interferences were
replaced by HPLC using preparatory size-exclusion columns. As a result, the cleanup time
was reduced by 75 percent, solvent consumption was cut 50-70 percent, and the cleanup was
automated. This advance in methodology allowed us to increase our sample throughput
substantially without an increase in staff. Furthermore, this automated cleanup has been
adopted of use by government, academic, and private laboratories.
Without assuring and documenting the quality of the data from measuring
contaminants, the data obtained are of little value. For example, those who assess and
manage risks need to base their risk calculations on data of known quality. We use a
performance-based quality assurance program in which a laboratory documents analytical
methods and quality assurance/quality control (QA/QC) practices. The QA manager for the
project (e.g., NS&T) selects the performance evaluation materials to be analyzed by the
laboratories participating in the intercomparison exercises. The manager then evaluates the
performance of the laboratories on the intercomparison exercises and accepts into the
program those laboratories that have demonstrated comparability (within certain set limits)
with the other laboratories. To continue in the program, laboratories must maintain
performance-based QA standards.
In monitoring for NBSP, several species of bottom-dwelling fish were collected
yearly at 45 sites along the Atlantic, Gulf, and Pacific Coasts. The fish were necropsied and
samples of the internal organs were collected for chemical and histopathological analyses
(Figure 2). Chemical analyses for PCBs, PAHs, pesticides and metals were conducted on
sediments, fish livers, and fish stomach contents. In addition, bile of the fish was analyzed
for PAH metabolites because these contaminants, unlike PCBs, are metabolized! in the liver
and excreted in bile for elimination. Several organs were analyzed for diseases by
histopathology. The original analytes chosen for NBSP were those on the EPA "priority
pollutant" list, which included PCBs (Figure 3). First, concentrations of all the isomers of a
given chlorination level were summed and reported, and the sum of all these levels was then
reported as "total PCBs." In addition, 8 individual congeners were determined initially, but
that number was increased to 18 for the QA intercomparison exercises.
In the NBSP, results of the chemical and histopathological analyses were examined
statistically to find any correlations that would link contaminants to diseases in the fish
(Figure 4). Toxicologists then examined the data in terms of assessing the risk of the
contaminants for causing health problems, such as liver and kidney lesions, in individual fish
or the fish population. These analyses of the data showed correlations between
3-30
-------
concentrations of PCBs in sediments and in fish livers, establishing the bioavailability of the
PCBs and bioaccumulation of these contaminants by fish. In addition, correlations were
found between concentrations of PCBs and lesions in fish livers, but because many groups of
chemicals tend to co-occur (PAHs, PCBs, and pesticides), the relative contribution of each
type is difficult to determine. Regional differences also were found in PCB patterns in
sediments, possibly because of differences in the input sources. Furthermore, PCB patterns
in the fish did not match any particular Aroclor or mixture of Aroclors, possibly due to
differential bioavailability/bioaccumulation of the PCBs or to metabolism.
As a result of this research, questions have been raised: (1) Should we be
measuring lexicologically important compounds, e.g., the coplanar PCBs? (2) How do the
proportions of certain toxic contaminants (e.g., coplanar PCBs) found in marine species
change as they are transferred up the food chain to predators at the top? and (3) Are the
most important toxic endpoints being measured, i.e., should there be a change from the
emphasis of cancer/tumors/disease to looking at reproductive, immune, and developmental
disorders? The objectives of the NBSP program have been altered and other projects, such
as the Coastal Ocean Program, have been initiated to attempt to answer these questions by
evaluating links between tissue concentrations of contaminants and bioeffects, such as
reproductive or immune system disorders.
The well-established NBSP continues in 1993 and last year, NMFS added a research
effort in analytical methods development to its existing Seafood Product Quality and Safety
(PQ&S) Program (Figure 5). Two other projects have been initiated recently: habitat
research for NOAA's Coastal Ocean Program (COP) and research on contaminant levels and
sources for the Marine Mammal Health and Stranding Response Program. These programs
demonstrate how research objectives have changed since the inception of NBSP in 1984. For
example, the NBSP has broadened its definitions of analytes and of effects—investigating
those analytes with toxicological importance for inclusion in the list of analytes. In addition,
the endpoints used to define effects in NBSP and COP have been expanded to include
reproductive, immune, biochemical, and other disorders. The PQ&S Program includes
research into developing better analytical methods for determining those contaminants (1)
postulated by toxicologists as potentially harmful to human health and (2) that can be
transferred to the human consumer of seafood products. For the first time, toxicology is
driving which chemical and biological parameters are to be measured, as evidenced by
inclusion of toxicologically important analytes and biologically significant endpoints in our
contaminant programs (Figure 6). The results of the chemical and biological analyses are
examined statistically to find any links between contaminants in the fish and endpoints that
determine effects. Toxicologists can then evaluate the data in terms of assessing the risk of
contaminants for producing health problems in the human or fish populations.
Currently, detailed analytical methods available for determining eoplanar PCBs are
time-consuming and laborious (Figure 7). The fish tissue is extracted and the extract must is
cleaned up to eliminate lipids and some interfering analytes (PAHs). However, the coplanar
PCBs are only a small fraction of the total PCBs and several coplanars coelute with other
3-31
-------
PCBs in GC analyses. Thus, an initial separation on a gravity-flow carbon column is
necessary to isolate the coplanar PCBs and improve the accuracy of the determination.
Finally, the analytes are quantitated by HRGC/HRMS. Therefore, in programs requiring
large numbers of analyses, e.g., monitoring or seafood safety, cost-efficiency caii be
increased through the initial screening of samples, i.e., rapidly identifying those samples with
high concentrations of the coplanar PCBs that require further analysis by detailed methods.
I
We have developed a screening method for rapidly identifying and quantitating
coplanar PCBs (Figure 8). PCBs in the tissue extract are separated from lipid and interfering
contaminants (PAHs) on a small-size gravity-flow column of acidic silica gel. The cleaned-
up extract is then chromatographed on HPLC (Cosmosil PYE column) and the PCBs are
detected by photodiode array (ultraviolet) (Figure 9). The retention times and integrated
areas can be calculated at any ultraviolet wavelength in the range selected for analysis. In
addition, ultraviolet spectra are taken twice each second and a software algorithm compares
the spectra to each other to determine peak purity and to the spectra of standard in a library
to determine peak identity. Finally, PCS concentrations can be confirmed in selected
samples by HRGC/HRMS or other methods.
The HPLC/PDA method provides excellent quantitative results for many of the PCBs
and DDTs. A comparison of coplanar PCB concentrations obtained for MIST SRM 1588
(cod liver oil) by HPLC/PDA, NIST, and GC/ECD showed good agreement (Figure 10).
Advantages of the HPLC/PDA method are low cost, the ability to confirm compound identity
and purity, and the ability to select samples for detailed analyses by HRGC/HRMS. In
addition, the method provides semiquantitative concentrations for those PCB congeners that
coelute with other analytes and is nearly as sensitive as some of the detailed methods.
In conclusion, the objectives and methods for determining PCBs and other
contaminants have evolved over the past 10 years in a number of ways. Labor-intensive
procedures have been replaced by automated, cost-effective methods. The use of initial
screening analyses, followed by confirmation of selected samples by detailed analyses, has
lowered analytical costs and provided for rapid dissemination of results. Furthermore,
toxicologically and environmentally important analytes and endpoints of effects have been
chosen. '
3-32
-------
Figure 1
Analysis for PCBs in fish tissue
FIshtlMU*
Ttam*
•xtnctlon
Swnpte pnctowup
_/
Automated Cleanup
GC/MS Analysis
Figure 2
Chemical/histopathological
analyses for NBSP
PCBs,
pesticides,
metals
3-33
-------
Figure 3
Original PCB analytes (NBSP)
Ch I or i nation level
trichlorobiphenyis
tetrachlorobiphenyls
pentachlorobiphenyls
hexachlorobiphenyls
heptachlorobiphenyls
octachlorobiphenyls
nonachlorobiphenyls
decachlorobiphenyls
E levels = total PCBs
Individual PCBs
(congener number)
8
28
52
101
118
138
170
187
206
18
44
66
105
128
153
180
195
209
Figure 4
Original objectives of NBSP
Field Studies
Contaminant
Effects
Analytical
Chemistry
3-34
-------
Figure 5
NMFS's Contaminant Programs-
1993
NMFS's Contaminant
. Programs
Seafood
Product Quality and
Safety (PQ&S)
National Benthic
Surveillance Project (N3SP)
Marine Mammal Health and
Stranding Response Program
Coastal Ocean Program
Figure 6
Objectives of the NMFS
Contaminant Programs—1993
Pathology
Reproduction
Immunology
Fish health
(NBSP)
Bioindicators of
contaminant effects
Human health
(PQ&S)
Analytical
Chemistry
3-35
-------
Figure?
Standard analyses for coplanar PCBs
Rlh t!**u«
• Coplanar PCBs/dioxins
O Nonplanar PCBs
• Lipids/biogenic material
O PAHs
1 r ' ^
S; r' —T
-^- Q I carbon *-^-
O acid L. O clean"P
sample cleanup O column
ntraction .
column
HRGC/HRMS
determination
New
cost-effective
i methods
needed
Figures
HPLC/photodiode array analysis
Automated Analyses
Autosampler PYE
column Photodlodo
Array Detector
Confirmation by HRGC/HRMS
3-36
-------
Figure 9
HPLC/PDA Chromatogram
White croaker—West Harbor Is
Figure 10
Method comparison—NIST SRM 1588
Concentrations (ng/g)
Coplanar PCBs
(congener*)
118
105
156
77
126
169
HPLC/PDA
n=6
190±3
68±7
19±2
<1.4
<1.2
<1.2
NIST
certified
177±3
61±3
28±1
ND
ND
ND
GC/ECD
n=1
170
44
13
1.4
1.9
0.5
ND - not determined
3-37
-------
3.5 EPA'S GREEN BAY PCB STUDY—CONGENER ANALYSIS
Deborah L. Swackhamer, Associate Professor, University of Minnesota
1
This presentation reviews a study conducted by the U.S. EPA Great Lakes National Program
Office (Chicago, IL) that involved conducting a monitoring-level study using research-level
technologies, both for sample collection and analysis. The Green Bay Mass Balance Study
(GBMBS) (De Vault and Harris, 1989) was a pilot study to determine the feasibility of using
a mass balance modeling approach to make management decisions regarding regulatory,
remediation, and source reduction strategies. The goal was to construct a state-of-the-art fate
and exposure assessment model coupled to a state-of-the-art foodchain model to be able to
predict contaminant levels in fish with a low uncertainty. The need for such a model arises
from lake-wide management plans of the Great Lakes and the Water Quality Agreement of
1987 between Canada and the United States. This seven-year study is in its finishing stages.
One of the contaminants of interest was polychlorinated biphenyls (PCBs). To calibrate and
test the validity of the Green Bay Mass Balance Model, a calibration data set of PCBs was
needed for all relevant media and of sufficient temporal and spatial complexity. Green Bay,
a major bay of Lake Michigan, has received PCBs historically from industries along the Fox
River, which flows north through Wisconsin and empties into the southern end of Green Bay.
The lower 65 miles of the Fox River at one time had 15 pulp and paper mills, 4 paper
recycling plants, and 6 sewage treatment plants all contributing PCBs to the river in the form
of Aroclor 1242. The interest in implementing remediation strategies in the Fox River or in
the Bay made this site an ideal pilot study choice for model development and calibration.
We chose to collect a PCB calibration data set that was congener specific for a
number of reasons. The primary reason was that a data set containing 80-90 congeners
would provide the greatest flexibility for the model, compared to 3-4 Aroclors or simply total
PCBs. A state-of-the-art model deserved state-of-the-art data. It was clear that such data
would be useful to the research community far beyond the modeling effort. Having data for
a wide range of compounds would also permit correlation of distribution and fate processes
to compound physical-chemical properties that could be used to predict the behavior of other
chemicals.
The complexity and scale of study is underscored by the number of investigators and
support staff who participated in the study, and the number of analyses undertaken. Four
federal agencies, two state agencies, and 14 academic institutions were involved in the study.
Overall, several thousand PCB analyses were done by 8 different laboratories, including 2
federal contract laboratories, 1 federal laboratory, 1 state laboratory, and 14 academic
research laboratories. Media included air (vapor and paniculate), precipitation, bay water
and tributary waters (dissolved and particulate), sediments, phytoplankton, zooplankton, and
fish.
: i
A study of this size required an immense planning effort. As part of this planning, a
rigorous quality assurance (QA) program was designed at the onset to ensure that the data
generated in the study were of sufficient and comparable quality to be used in the model
3-38
-------
(Swackhamer, 1988). The number of sample analyses for PCBs required the participation of
eight laboratories. While analytical procedures for sample handling, extraction, and cleanup
were performance-based, the instrumental quantisation of PCBs was done by a specified
method. The method used was an adaption of one by Mullin (1985), who characterized the
weight-percent composition of each peak of a 25:18:18 mixture of Aroclors 1332, 1248,
1262 for both DB-5 and DB-1 columns by capillary column gas chromatography and electron
capture detection. Investigators were provided the Aroclors needed to make this standard
mixture and the weight-percent information, thus allowing the generation of congener-specific
response factors. Laboratories reported data in terms of 85 components by the internal
standard method using congeners #30 and #204. All data were corrected to the recoveries of
surrogate standards #14, #65, and #166.
Laboratories had to demonstrate their ability to meet certain QA criteria regarding
resolution, surrogate recoveries, matrix spike recoveries, precision, and detection limits. All
labs were also required to participate in a "round robin" quantitation exercise to demonstrate
successful application of the congener-specific method. The samples distributed consisted of
vials containing different mixtures of Aroclors of concentrations ranging from 500 ng/mL to
50 ng/mL. Several also contained toxaphene, a probable interference for PCBs in the Green
Bay samples. Controls also were included. In general, laboratories performed well on the
intercomparison, with errors of approximately 20 percent. The presence of toxaphene did
not have a significant impact on performance of the method. One laboratory had large errors
at low concentrations (error = 50-100 percent). This was largely due to the problem of co-
eluting interferences in reagent blanks, and underscored the need for strict control of
laboratory contamination when doing congener-specific analyses at trace levels.
This study demonstrated that congener-specific analyses can be used for large scale
operations such as monitoring programs. The implementation of these methods requires
certain capital investments including chromatographic data processing software, significant
time investments, and greater attention to field, reagent, and procedural blanks. In return,
laboratories obtained analytical capability that allows for greater flexibility in their projects
and contracts and in the use of data. In the case of Green Bay, we generated a "platinum"
self-consistent data set that will be used to understand environmental processes for years to
come, and introduced several analytical laboratories to state-of-the-art PCB quantitation
techniques.
References
De Vault, D.S. and H. Harris, 1989. Green Bay/Fox River Mass Balance River Study, U.S.
EPA Great Lakes National Program Office, Report #06-89, Chicago, IL.
Mullin, M, 1985. PCB Workshop, U.S. EPA Large Lakes Research Station, Grosse He, MI.
Swackhamer, D.L., 1988. Quality Assurance Document for Green Bay Mass Balance Study:
PCBs and Dieldrin, U.S. EPA Great Lakes National Program Office, Chicago, IL.
3-39
-------
WHY CONGENER-SPECIFIC
ANALYSES?
Greatest flexibility for model
i
Provide valuable data set for future
efforts
i
Allow measures to be correlated to pchem
properties which enables extrapolation to
other chemicals
To retain involvement of research
community
3-40
-------
QUALITY ASSURANCE PROGRAM
To ensure that all data generated in study
were of sufficient and comparable quality to
be used in the mass balance model.
• Coordinated 8 laboratories for PCB
analyses
• Approved all sampling SOPs
• Approved all analytical SOPs
• Required congener-specific PCB analysis
of all air, water, sediment, plankton, and
fish samples (thousands!)
•• Conducted Round Robin for PCB
analyses
• Visited each lab to aid in troubleshooting
problems
3-41
-------
METHOD OF QUANTITATION
(Adapted from Mullin, 1985)
Quantitation Standard: 25:18:18 mixture
of Aroclors 1232, 1248, 1262 which
contains all environmentally-significant
congeners
Weight-percent of each component
characterized for DB-5, DB-1 resolution
i
i
Internal standard method using IUPAC
#30, #204
Surrogate standards required to monitor
recovery efficiencies: IUPAC #14, #65,
#166
All data corrected to surrogate recoveries
Each lab reported 85 congeners (common
denominator) using DB-5 column
3-42
-------
QA CRITERIA
• Resolution of congeners #17 and #18
• Surrogate recoveries of 50% < R < 120%
• Spike recoveries of 50% < R < 120%
• Duplicate precision ±50%
Monitor ratio of #30/#204
• Blanks: < 10% of samples (peak basis)
• Congener-specific LODs, LOQs
3-43
-------
CO
ZJ
CO
CD
C
n
o
en
o
Q-
"5
-*-»
o
c
3
o
*•*-'••'
O => ^ Q Z)
ID01QI
_V . ."W- . • ">« .
-I 1 1 \ 1 1
o
O
m
CM
CM
o
o
CM
lO
o to
m CM
o
o
LO O
i^ in
LO
CM
3-44
-------
CO
-»_>
13
CO
G> ~>
CX CD
.E £
-O —
5 5
£
•o i—
c
13
O
ce
c
•si
c
•TH
fTnrtT-if3'_m 0 C
=*
=*=»'"«' « -M-J-J.-A- J. A JL, 1 ,,fr J
•••
•^
3=z=a
Ml
grrc
3=1
25 0)CL
=« — ^ (^ CL ^ Q
0 ID ^ 0 Z> ^
P^I i i n K3 n B
= M M Rl KJ •
=" kd LJ U Sj LJ •
=5
I » i t
1 I 1 -H 1
5 O O O O O
> O O O O O
) CM T$- CD 00 O
> O O C) CD ^
I i i 1 i
O
LJ
O
JOJJ3
3-45
-------
WHAT DID IT TAKE?
Required computerized chromatographic
acquisition and processing software
Required greater attention to blanks
i
Required extensive implementation time
WHAT DID WE GET?
"Platinum" data set
Ability to model congeners, homologs,
total PCB
Upgraded several labs to state-of-the-art
3-46
-------
3.6 STATE LABORATORY EXPERIENCE
Brian Bush, Research Scientist V, New York State Department of Health,
Wadsworth Laboratories, Albany, NY.
A congener specific analysis for PCBs is described that has been used for the analysis of
several matrices, including fish, for the past 10 years in the Wadsworth Laboratories. A 5 g
sample of homogenized fish flesh is ground 3 times with anhydrous sodium sulfate and
hexane using a Tissuemaster. The hexane extract is evaporated to 2 mL and purified on a
calibrated 10 g column of 4 percent deactivated Florisil. A 1:1:1:1 mixture of Aroclors
1221, 1016, 1254, and 1260 (200 ng/mL of each) is used to calibrate the electron capture
detector and the gas chromatograph (Aroclor 1221 is incorporated in the mixture because
many of our samples are from the Hudson River, which is polluted by mono- and
dichlorobiphenyls). The calibration mixture has been carefully characterized using 56
synthetic PCB congeners and with reference to the work of Mullin et al and Schultz et al.
p,p'-DDE, mirex, and hexachlorobenzene also are added to the mixture (10 ng/mL of each).
In all on 5 percent phenylmethylsilicone columns, 68 PCB containing peaks are calibrated
plus the three other pesticides. A confirmatory column of Apiezon L also is employed.
Altogether, 85 PCB congeners, which represent the major compounds distributed in the
environment, may be quantified using combined data from each column.
The resultant massive quantity of data requires electronic data processing. We rely on
the microprocessor of the Hewlett Packard chromatographs we employ to carry out the
primary quantitation, sending the quantitative data to IBM compatible personal computers
where they are labelled and formatted in a Lotus 1,2,3 spread sheet (see Laboratory Lotus, A
Complete Guide to Instrument Interfacing, L.M. Mezei, Prentice Hall, Englewood Cliffs,
NJ, 1989). Checking the performance of the system is then simplified and editing is easily
carried out by reference to the hard copy chromatogram, all changes being indelibly recorded
on the chromatogram, to produce what has been described as "platinum" data. The finished
Lotus spread sheet is then easily accessible for data evaluation by investigators in
epidemiology, toxicology, and environmental studies.
The minimum detection limit (MDL) at which the probability is <0.01 that the
compound is not present and > 0.988 that it is present (alpha and beta) is determined by
doing seven replicate determinations at a level near disappearance of the signals of the
compounds of interest (relative standard deviation >20 percent). The standard deviation of
the measurements is multiplied by the Student's t for 6 degrees of freedom and p=0.01, the
result is defined as the MDL. To obtain a similar value for "total PCB" that is the sum of
all PCB, the synthesis of variance theorem is employed. The square of the standard
deviation of each PCB congener measured is summed, the square root of this sum is
multiplied by Student's t for 6 degrees of freedom, p=0.01.
3-47
-------
We analyzed 60 striped bass samples from the Hudson River estuary, the NY Bight, and
Long Island Sound by the above method ("Polychlorinatedbiphenyl (PCS) Congeners in
Striped Bass (Morone saxatiltis) from Marine and Estuarine Waters of New York State
Determined by Capillary Gas Chromatography," Bush, B., Streeter, R.W., and RJ. Sloan,
1989, Arch Environ Contain Toxicol 19:49-61) to determine whether the quantity of less toxic
PCB derived from Aroclor 1242 would change the estimated risk for human consumption
purposes. The samples had been analyzed by a contract laboratory using a packed column
method similar to the method of Webb and McCall (1973). The correlation coefficient
between our data and the contract laboratory's was 0.91. Congener specific Apiezon L
analysis separated 2,2'- and 2,6-dichlorobiphenyl from each other and allowed a typical
Hudson River signature to be observed in some fish from as far removed from the river as
Mantouk Point (150 miles).
In conclusion, it is possible to undertake accurate multi-component analysis and to make
significant discoveries in studies of the environment, toxicology, and environmental health.
To do this, electronic data transfer to user friendly PC-based software is mandatory.
3-48
-------
3.7 SUMMARY OF QUESTIONS AND RESPONSES10
3.7.1. Dr. John Brown of General Electric asked Ted Schwartz about the problem of
reporting results in terms of principal components. For example, Dr. Brown stated that there
never appears to be a unit describing what the principal component is. Dr. Brown continued,
stating that the results are simply a mathematical artifact indicating that a given composition
is described in terms of correlation with component 1, component 2, etc.
Ted Schwartz responded that his presentation plotted samples in Score terms, which are
related to the composition in the parent samples. He indicated that his group has loading
terms that do not display any of the particular variables that would influence where the
congeners fit in the data space. For example, he stated that if loading terms from the EMAP
data set were shown, one could see that the separation was driven by primarily by just two
congeners. He stated that when analyzing the congeners in the EMAP program, perhaps one
should question whether the most appropriate congeners are being measured.
3.7.2. Jack Moore from IEHR questioned whether the current scientific focus on AHH-active
congeners is appropriate; perhaps the wrong congeners are being analyzed and evaluated.
Perhaps not enough attention is being given to other congeners (e.g., non-AHH congeners
which show tremendous accumulation potential).
Ted Schwartz responded that congeners are measured according to the established
benchmark (e.g., the 2,3,7,8-TCDD dioxiri). He noted that based upon this benchmark, the
congeners are measured for relative importance and influence.
3.7.3. Rich Pruell from EPA's Narragansett Laboratory commented to Mr. Sawyer of FDA
that measuring of total PCBs [as Aroclors] may not actually be a "conservative" approach.
Mr. Pruell pointed out that in highly altered residues (e.g., lobster tomalley), there can
actually be selective enrichment of the AHH-inducing compounds. Thus one may be
underestimating the lexicological potential of the mixture.
Mr. Sawyer's response was that a total PCB analysis should include the AHH-active
congeners. John Brown of General Electric agreed with the FDA in their total PCB
interpretation. He stated that, if the measure of human risk is in terms of the levels of the
persistent congeners (as the toxicological data appears to indicate), then the classical
approach of examining the major peaks and matching them to an Aroclor standard is actually
measuring those peaks that will be persistent and will ignore those peaks that will not
Ed. note: The question and response portion includes summaries derived from transcribed
conversations. The summaries have been carefully edited to present the discussion as accurately as possible.
However, these question and response summaries have not been reviewed by the speakers—unlike the proceeding
abstracts.
3-49
-------
accumulate in humans.
3.7.4 John Brown noted that Dr. Bush measured PCB congeners in his fish tissue studies and
that the congeners varied tremendously from fish to fish, depending on the depuration that
has occurred. He asked how Dr. Bush could infer that the source of these congeners was a
capacitor-type PCBs, such as Aroclor 1242—especially since Aroclor 1242 represented 50%
of all PCB sales and was widely used throughout the country.
. i
Dr. Bush replied that there were several indications. One was the occurrence pattern of
the 2,2,2,6 congener. Another was the presence of all the congeners that belong to 1242 in
macroinvertebrate samples. He described a mathematical procedure he used that indicated a
strong correlation.
3-50
-------
PART FOUR
CASE STUDIES:
HUMAN HEALTH/RISK ASSESSMENT
4.1 CALIFORNIA
Gerald Pollock, Acting Chief, Fish and Sediment Contamination Unit, Pesticide and
Environmental Toxicology Section, Office of Environmental Health Hazard
Assessment, California EPA, Sacramento, CA
The Department of Health Services (DHS) issued an interim health advisory in 1985 for
consumption of contaminated fish in the southern California area based on elevated DDT and
PCB levels; The California State Legislature then mandated that DHS conduct a study of
chemical contamination of marine fish and conduct a health evaluation.
DHS initiated a series of studies of chemical contamination of marine fish in southern
California and the Office of Environmental Health Hazard Assessment (OEHHA; previously
in the Department of Health Services) conducted a risk assessment based on the results.
Four studies were proposed; an initial pilot study, a comprehensive study and risk
assessment, an investigation of commercial fishing in the area, and a study of human breast
milk.
The first study, the Pilot Study, was designed to identify the chemicals of concern in fish
tissues. A few indicator fish species were sampled at highly contaminated locations and
analyzed for the priority pollutants and selected additional pesticides. The results of the Pilot
Study identified DDT (and metabolites) and the PCBs as the main chemicals of concern.
The Comprehensive Study followed the Pilot Study and examined the levels of the
chemicals of concern in tissues of several fish species from multiple sites. Twenty-four sites
were selected for sampling and between 5-10 different species of fish were sampled at a
single site. In all, fifteen different species (or groups) of fish were sampled in the study.
Sites were selected to represent areas frequented by party boat, private boat, and pier
anglers. Fish species were selected to represent those species frequently caught and
consumed by anglers.
Tissue samples were obtained from the edible flesh of each fish and composite samples
were homogenized and extracted for analysis of DDT and PCB levels. The results indicated
significant contamination of some species of fish at several sites. In general, white croaker
were the most contaminated species at a given site. This was especially true at highly
contaminated sites. The data were used in the OEHHA risk assessment and site/species
specific consumption advisories were issued based on the risk assessment.
-------
The risk assessment included evaluation of risks due to consumption of fish Contaminated
with PCBs. PCB residues were quantitated using an Aroclor 1260 equivalent concentration
and the cancer potency factor for Aroclor 1260, However, the evaluation noted problems in
addressing risks due to PCB residues. Specifically, the limit of detection for PCBs (50 ppb)
was above a human health based level of concern. In addition, it is known that the residues
in fish tissues represent a mixture of Aroclor 1242, 1252, and 1260 congeners and are not
identical to the original mixtures (e.g., weathered residues).
The Santa Monica Bay Restoration Project conducted a small scale follow-up study in
1991 of chemical contaminants in white croaker (10 sites) and yellow crab (2 sites) in the
same area. The SMBRP study quantitated PCB levels using congener specific analyses and
summed total Aroclor equivalents. These data will be used in a future risk assessment.
Risk assessment of PCBs in fish tissues presents several problems for the regulator.
These problems are scientific and public (communication). The scientific problems involve
interpreting the toxicological significance and potential carcinogenic potential of PCB
congeners. The use of toxicity equivalent factors has been proposed for use in risk
assessment of PCB congeners (congeners with "dioxin-like" activity) to improve this process.
However, consensus within the scientific community has not been reached regarding this
approach. Still, most agree that improvements in risk assessment of PCB congeners is
desirable.
Public problems are more difficult to identify because they involve interpretation of
scientific information by a variety of non-scientific sources. Frequently, one "expert" will
criticize an evaluation of PCB residues by citing the results as "not state-of-the-art." Such
allegations by experts (frequently academics) raise questions of credibility on the part of the
public and undermine public confidence in an assessment. Also, it is difficult to explain to a
non-scientific population the chemical complexity of PCBs and even more difficult the
toxicological uncertainty associated with PCBs. Contradictory statements by experts add to
the public's confusion.
The scientific community should determine where we are in the process of interpreting
the health significance of PCB residues and chart a course for future changes in the process.
We need to agree on a process for conducting risk assessments in the meantime (Do we still
use the q* approach using the Aroclor 1260 potency factor?).
We then need to identify how we will move (improve) the risk assessment process and
answer key questions in charting this movement. Are we just about ready to use the TEF
approach? How do we handle the potential effects caused by "non dioxin-like" congeners?
It is worthwhile for regulatory agencies to invest effort into studying ways to improve
the presentation of PCBs to the public. I foresee that we will be going back to previous
assessments and doing re-assessments once we have decided on consensus (TEF?) approach
and then either issuing new advisories or lifting existing advisories. These actions may
4-2
-------
confuse the public who may then feel betrayed.
And the point here, is that it is all in the name of progress. We are moving forward.
PCBs are a huge problem and we need to find a way to maintain the public credibility in our
process, as much as possible, as we move toward improved methods of risk assessment for
the PCBs. Coordination and cooperation by scientists will serve the regulatory community
and the public. Workshops like this one serve to facilitate the communication process and
coordination among involved scientists during these dynamic times.
4-3
-------
INTRODUCTION
STUDY OF MARINE FISH IN SOUTHERN CALIFORNIA
Focus on risk assessment of PCBS
DESCRIBE A FOLLOW-UP STUDY
Focus on differences in PCBs analyses
HIGHLIGHT SCIENTIFIC PROBLEMS IN RA OF PCBS
Focus on reality
HIGHLIGHT PUBLIC PERCEPTION PROBLEMS
Focus
4-4
-------
Study of Chemical Contamination
of Marine Fish from Southern California
I. Pilot Study (January 1991)
* Identification of chemicals of concern
II. Comprehensive Study (September 1991)
* Collection of -4000 fish
* 15 different species; 24 locations
* -1000 analyses of composite samples
* risk assessment of DDTs and PCBs
III. Data Resources on Commercial Fish
(in, preparation)
* Completion of existing data
IV. Contamination of Human Breastmilk
(in preparation)
* Epidemiological study
* Collection and analysis of human milk
4-5
-------
"O
C
O
D)
O
O
^•J
• !••§
CO
re
1
re
• Twin Harbor C
CO
o
3
Q
<*•«
'5
Q.
i
T-
• Point Vicnete
^
.Q
"re
1
CM
1=
"o
Q.
re
c
re
Q
1
10
§L
"35
•o
re
"C
re
I
CO
• Venice Beach
•
CO
c
re
m
-e
o
CO
1
Tf
5
75
• LA/LB Breakw
•
Palos Verdes
i
to
0)
E
2
75
CO
JZ
Redondo Beac
i
CO
(2
,22
Q.
O
O
^^
o
'E
0
1
re
CO
o
]S
'o
Q.
J»
i
i
i-
• Redondo Piers
o
c
re
m
Fourteen Mile 1
1
00
• CabrilloPier
,»
j-
0
re
Huntington Be
i
CD
_x-t_
re
S
c
Q)
CU
o
•o
l_
o
•MM
a.
CM
0
re
0)
CQ
re
c
3
O)
re
i
o
• Belmont Pier
•
CO
,0
w
E
Emma/Eva Oil
i
v
• Newport Pier
^
Q.
Horseshoe Kel
i
CM
4-6
-------
Matrix of Sites and Fish Species Sampled
in the Comprehensive Study of Southern California
Site Total //
Party Boat Private of Sites
Fish Areas Boat Piers Sampled
£
'i
Q.
c«
£
R
e
06
V>
2
3
a
s
e
00
-
*
">
*
«
"
-
S
f— «
CK
so
2
ts
S
t-
•n
OS
s
S
>O
«
ts
-
«
p-
o\
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
«-
VO
«-
*>
VO
0
^
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
O
t-
X
X
X
X
X
X
X
X
X
X
X
f
X
r*
X
X
s
X
X
X
X
i
X
X
X
X
X
X
(Birred sand bass
X
X
X
X
X
X
X
X
X
X
X
X
1
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
e
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1
X
X
X
X
X
[lloliblit
X
X
X
01
•O
1
X
X
X
X
1
X
X
X
X
X
X
X
X
X
X
X
X
[White croaker
X
X
X
X
X
X
X
X
X
X
X
X
1
X
X
X
X
X
X
X
X
X
jSurfperches
X
X
I
X
X
§
X
jllalfhioon
X
X
X
X
X
[Black croaker
0
00
r-
.«
00
-
0V
Wl
O
•—i
oo
OS
VO
0
oo
00
Total H of Species
Collected Per Site
4-7
-------
"Setting a health-based trigger level for PCBs, however, is problematic. Using
potential risks of 10"4, 10~5 or 10'6, the trigger levels for PCBs would be 40 ppb, 4 ppb,
and 0.4 ppb, repsectively. Concentrations of 4 ppb and 0.4 ppb are much lower than the
MDL in this study...."
"Ultimately, PCBs should be analyzed using congener specific analysis (which can
now be achieved analytically) and the results interpreted for each congener.
Unfortunately, a generally accepted method for toxicological interpretation of cogener data
has not yet been established...."
"Since the health-based trigger levels for PCBs are below the MDL, then the trigger
levels for PCBs could be set at the MDL which is the lowest practical analytical value.
However, analytical accuracy and reprodueibility is generally not good even at the
MDL...The trigger level for PCBs must be etablished which take into consideration
potential toxicity and analytical accuracy."
"...Therefore, one could propose using the LOQ or PQL as the trigger level (over
200 ppb) for issuing guidance for the PCBs."
"Unfortunately, the theoretical excess cancer risks estimated at these levels are high
enough to be of concern as are the theoretical excess risks estimated at the MDL (e.g.,
1 x 10-4)."
"Obviously, providing a health-based trigger level for PCBs is problematic...."
"It is proposed that a trigger level of 100 ppb for PCBs be used for providing
guidance for fish consumption based on the results of this study...It is important to note
that establishing this trigger level does not signify that OEHHA considers this level to be
acceptable. OEHHA supports the use of the health-based levels and recommends that
methods (analytical and toxicological) be developed which allow for providing consumers
with health-based (and, therefore, consistent) guidance."
"It is noteworthy that setting a trigger level for PCBs that takes into consideration
the MDL and laboratory performance would limit the use of the level to the particular
study being evaluated...That is, the trigger level applies only to the study for which it is
being applied—"
"However, immediate and practical needs demand that guidance be provided
presently rather than waiting until analytical/toxicological methods for PCBs are
developed and/or improved. When newer health-based methods for PCBs evaluation are
available, OEHHA will be able to provide improved health-based guidance."
Excerpts from: Pollock, G.A., I.J. Uhaa, A.M. Fan, J.A. Wisniewski, and I. Witherell. 1991. A Study of
Chemical Contamination of Marine Fish from Southern California. OEHHA; Cal/EPA.
Sacramento, CA.
4-8
-------
Establishing "Trigger Levels"
for
PCBs in Southern California
Criteria
Level
ppb
Excess Cancer
Risk
Health-Based
4
40
1 x 10-5
1x10-4
MDL
50
-10-4
Trigger Level
100
2x10-4
LOQ
120
3 x 10-4
PQL
>200
5x10-4
MDL: Minimum detection limit
LOQ: limit of quantitation
PQC: Practical quantitation limit
4-9
-------
SITE-SPECIFIC CONSUMPTION RECOMMENDATIONS
SITE
Marina del Rey
Huntington Beach
Fourteen Mite Bank
Laguna Beach
Redondo Beach
Emma/Eva oil platforms
Catalina (Twin Harbor)
Santa Monica Pier
Venice Pier
Venice Beach
Dana Point
FISH SPECIES
All species
RECO M M EN DATION
No restrictions
Newport Pier
Redondo Pier
Belrnont Pier
Pier J
Corbina
n
Surfperches
One meal every
two weeks
Malibu Pier
Queenfish
One meal a month
Short Bank
Malibu
Point Dume
Point Vicente
Palos Verdes - Northwest
White's Point
Los Angeles/Long
Beach Harbors (esp.
Cabrillo Pier)
White croaker
two weeks
White croaker
White croaker
White croaker
Sculpin
Rockfishes
Kelp bass
White croaker
Queenfish
Black croaker
Surfperches
One meal every
Do not consume
Do not consume
Do not consume
One meal every
two weeks +
Do not consume
One meal every
two weeks +
Los Angeles/Long
Beach Breakwater
(ocean side)
Horseshoe Kelp
White croaker
Queenfish
Surfperches
Black croaker
Sculpin
White croaker
One meal a
month +
One meal a
month +
* One meal is about six ounces.
•f* Consumption recommendation is for all the listed species combined.
4-10
-------
SCIENTIFIC PROBLEMS
Okay
Not Okay
Logic - Okay
Analytical Chemistry
Toxicology
More congener specific data
More endpoints
dioxin-like activity (co-planar)
other toxicities
(basic toxicity mechanisms)
Risk Assessment
Not Okay
TEFfor dioxin-like activity?
TEFsfor other toxicities? (applied to RfD?)
4-11
-------
PUBLIC PROBLEMS
+ Conflicting experts
"not state-of-the-art"
"obsolete"
''don ft know what they
i
are doing"
+ Uncertainty
+ PCBs are complicated
+ Re-issuing or lifting
advisories based on
changing RA method
(TEFs).
CONFUSION
4-12
-------
Risk
q* x dose
, Dose .
(exposure)
[X] [consumption]
rate
body weight
Consumption
Rate > 23glday
(about 1 meal/week)
[X]
> By analysis
Body weight > 70 kg adult
4-13
-------
03
'c
o
«*—
"co
O
O) 3
-Q
Q.
(D
o
C 3
O «
~ CO
OS '*=
§
§ 1
c °
E ^5
O O
*]-— •»—»
o> Ic
E 5
O
0.
I
o
in
Q
§
en
tn
o>
>
i
to
f
0>
cn
"m
I
4-14
-------
gj
'c
J5
15
O
O3
'CD
05
5
T3
0}
O
Q.
co
ra
CD
I
CD
•o
2
to
t
1
(0
o
•a
S
8
CO
s
CD
4-15
-------
CO
"c
i_
o
H—
"co
O
.
x:
O) 3
I
s
05
O
CO •*-
s «
o to
o S
o ^
c ®
S "«
CD o
O
"S
E
Q
Q
O
^—»
M—
O
CD
4-16
-------
4.2 TENNESSEE VALLEY AUTHORITY
Janice Cox, Water Management, Tennessee Valley Authority
As a federal resource management agency, the Tennessee Valley Authority (TVA) monitors
contaminant concentrations in fish throughout the 7-state Tennessee River basin. Because
TVA does not have the authority to issue consumption advisories, it defers to the
determinations of the state regulatory agencies when issuing advice on the safety of eating
fish from TVA reservoirs. However, differences in the consumption advisory policies of the
seven Valley states result in an unlevel playing field with respect to both protection of public
health and economic impact on sportfishing interests. These differences in state policies are
most apparent when dealing with PCB contamination because PCBs drive the cancer and
noncancer risks associated with eating fish from most TVA reservoirs.
To foster consistency in the evaluation of the fish contaminant data it collects, TVA
developed a graphic method of reporting risk assessment results to the states. The impetus
behind developing a graphic approach was to provide consistency in toxicity assessment
while retaining flexibility for the states to choose their assumptions about fish ingestion rate,
exposure duration, acceptable risk level, and whether to focus on cancer or noncancer
endpoints.
The graphic method presents conclusions from each fish tissue analysis in two
nomographs. The first nomograph, based on the aggregate impact of all carcinogens in the
sample, shows upper-bound incremental lifetime cancer risk as a function of fish
consumption rate. Different curves on the nomograph are used to illustrate the impact of
varying exposure duration assumptions. The second nomograph, based on the sum of hazard
quotients for each contaminant in the sample, shows the hazard index as a function of fish
consumption rate. Various lines on the nomograph can be used to represent the total hazard
index and the portion of the total hazard index associated with different classes of endpoints
(developmental impacts, hepatic toxicity, etc.). The second nomograph can easily be
modified to evaluate potential impact on consumers other than the standard 70-kg adult male.
For communicating with risk managers, this graphic approach is preferable to simply
reporting PCB concentrations or point estimates of risk:
The method evaluates aggregate risks from multiple contaminants. While PCBs are
the single largest contributor to the risk from consuming fish from Tennessee Valley
waters, DDT, chlordane, lead, heptachlor epoxide, dieldrin, endrin, and mercury are
also significant contributors to risk in fish samples from some waters.
The method moves beyond the black-or-white thinking inherent in defining
contaminant concentrations as either "safe" or "unsafe" by illustrating potential risk as
a function of exposure; that is, the method makes it easy to determine how much fish
consumption is acceptable given a target maximum acceptable risk level.
4-17
-------
• The method facilitates identification of the most sensitive adverse health effects, by
sample location, so that special "at risk" subpopulations (such as children and
pregnant or lactating women) can be targeted for risk management.
i
The method helps risk managers put the results of fish tissue studies in perspective by
facilitating comparison of the aggregate risks presented by fish consumption with the
risks attributable to contaminants in the general food supply. It also provides the
technical underpinning for effective risk communication with the public by permitting
qualitative and quantitative comparisons with health risks from other sources.
This risk assessment technique was used to evaluate screening-level data collected by
TVA in 1990. The results of the risk assessment depended more strongly on PCBs than on
any other single contaminant. Using an oral slope factor of 7.7 (mg/kg/day)"1, PCBs
accounted for most (average 89 percent) of the potential cancer risk from fish consumption.
For the recreational fisherman eating one meal of channel catfish per week (30 grams/day),
the calculated upper-bound incremental lifetime cancer risks varied by location from 2E-04 to
7E-03. For comparison, the background cancer risk from xenobiotic contaminants in the
general food supply, based on FDA market basket studies, is approximately 2E-04.
Using an estimated RfD of 5E-05 mg/kg/day, PCBs also accounted for more than 50
percent of the total hazard index in most of the samples. For the adult fisherman eating one
meal of channel catfish per week (30 grams/day), the calculated hazard indexes varied by
location from less than 1 to as much as 20. On a basis of "grams of fish per kilogram
body weight," the average 2-year old child eats about 50 percent more from the rneat-fish-
poultry food group than does the average adult. Therefore, the hazard index for small
children eating channel catfish on a regular basis may be as high as 30 for fish from some
locations. For comparison, the background hazard index for adult males from xenobiotic
contaminants in the general food supply, again based on FDA market basket studies, ranges
from 0.8 to 3 (depending on the assumptions one makes about lead intake and lead toxicity).
The background hazard index for the average 2-year old child ranges from 2 to 16 (again
depending upon assumptions about lead intake and lead toxicity).
With few exceptions, the sites with elevated (i.e., significantly higher thari
background food supply) hazard indexes corresponded with the sites that had an elevated
cancer risk as well. Incrementally varying assumptions about the estimated RfD for PCBs,
the amount of PCBs lost during fillet preparation and cooking, or the exposure duration
resulted in a relatively small change in the number of locations where the risk to
sportfishermen exceeded the background risk from the general food supply. However, the
total array of exposure assumptions taken together can markedly change the risk
characterization. Conservative assumptions corresponding to a "reasonable worst case"
resulted in 100 percent of sites exceeding the background food supply cancer risk and 73
percent of sites exceeding the background food supply hazard index for an adult consumer.
4-18
-------
More moderate assumptions corresponding to a "reasonable best case" resulted in only one
percent of sites exceeding the background food supply cancer risk and none of the sites
exceeding the background food supply hazard index for an adult consumer.
4-19
-------
0)
1
§
0)
o>
h
.1
c
o
.a
&
-
•s
C
O
fl
e
Chlord
Heptachlor
i
i
&
a
f
s
1
'O
i
CM
a
a>
I
IS
pfi
g
I
1
o*
-4-S
S
•S
e
o
03
O)
O)
IP
fe
I
i
a
§
H
i
.S
i i
S«u
J
s
4-20
-------
Comparison of Assumed Fish Consumption Rates
(Note: Typical "meal" size assumed to be 0.5 pounds)
1 meal/day
4 meals/week
3 meals/week
2 meals/week
1 meal/week
1 meal/month
grams/day
230"
210'
190'
170'
150'
130-
110"
90"
70"
50"
30'
10'
"Reasonable worst case" (EPA 1988a)
90th percentile for recreational fishermen (EPA 1989a)
95th percentile for Superfund RA (EPA 1989b);
default for subsistence fishermen (EPA 1991 a)
FDA's 90th percentile for subsistence fishermen
(Bolgeretal. 1990)
FDA's average for subsistence fisherman
(Bolgeretal. 1990)
default for recreational fishermen
in Superfund RA (EPA 1991 a)
50th percentile for Superfund RA (EPA 1989b)
50th percentile for recreational fishermen (EPA 1989a)
average recreational fisherman (USD11985)
U.S. average per capita (EPA 1988a)
4-21
-------
5
o
1
Q)
O
O
O
CQ
O
D_
O
CD
—•
CO
Q_
E
13
CO
O
O
r~ -4-!
Is
I
a
i|
£i
CD
O
c
o
O
CD
O
a.
o o o o o o o
tO ^J- CO CM T- O C5
oooooooo
h- 0> U3 "*• CO CVJ T-
'ejey uo^duinsuoo
4-22
-------
^
o
CD
•4—•
CO
a:
o
CL
E
.52
a:
E
CO '43
.Q —
E! «
11
§ &
O o
CQ —
J90UBQ
|e;u9ijuejou| punog
4-23
-------
0.05 0.1
Fish Consumption Rate (kg/day)
70-year
exposure
30-year
exposure
0.15
I
T3
0.05 0.1
Fish Consumption Rate (kg/day)
Risk Characterization for Elk River Mile 41
0.15
4-24
-------
4.3 DELAWARE
Richard W. Greene, Environmental Engineer, Watershed Assessment Branch,
Delaware Department of Natural Resources and Environmental Control11
Recreational and commercial fishing are important water-dependent activities in the Delaware
Estuary. There are growing concerns, however, regarding chemical contamination of
resident arid migratory fish in the system and potential health effects to anglers who may
consume their catch. New Jersey and Pennsylvania have both issued health advisories on the
consumption of catfish and white perch from the Estuary based on exceedence of the FDA
Action Level for PCBs. Furthermore, the United States Fish and Wildlife Service detected
elevated levels of PCBs in whole body striped bass taken from the system. In response to
these and similar findings, the State of Delaware conducted a pilot study of PCB
contamination in striped bass taken from the spawning ground adjacent to Wilmington,
Delaware in May of 1991. The results of that work lead state officials to further investigate
the problem. ,A more detailed study was undertaken in two areas of the Estuary in 1992 that
targeted the size ranges of striped bass most likely to be eaten by Delaware fishermen. The
two size categories selected for study were fish that were of a size legal in Delaware's
commercial fishery (those between 18 and 28 inches TL) and fish legal for recreational
harvest (minimum size 28 inches TL).
A total of 79 fish were obtained for the study. Forty-nine fish (25 commercial size
and 24 recreational size) were obtained from mid-Delaware Bay in February and March of
1992 as a part of commercial American shad by-catch. The remaining thirty fish (25
commercial size and 5 recreational size) were obtained from the spawning ground area in
May of 1992 by the Division of Fish and Wildlife. The 25 commercial size fish from the
Bay were grouped into 5 composite samples of 5 fish each. Equal mass aliquotes of muscle
tissue were taken from each fish. This same compositing scheme was used for the 25
commercial size fish from the spawning ground. The 24 recreational size fish from the Bay
were also grouped into 5 composite samples, this time, however, one of the samples only
contained 4 fish while the others contained 5. The remaining 5 recreational size fish from
the spawning ground were treated as individual samples. The decision was made not to
sacrifice 20 additional recreational size fish from the spawning ground because of concerns
over preservation of existing spawning stock. In all, 20 samples were prepared for analysis,
14 five-fish composites, 1 four-fish composite, and 5 individual fish.
The specific objective of this study was to fully characterize the PCB content of
striped bass to help support a credible human health risk assessment. The chemical
characterization is expected to be coupled with a consumption survey to further refine the
risk projections. The available literature suggested that a congener-specific and homolog-
specific analysis would provide the information of greatest utility for this study. Decision
11
Roy W. Miller of the Division of Fish and Wildlife, Delaware Department of Natural Resources and
Environmental Control was a coauthor of the paper.
4-25
-------
criteria used in specifying which congeners to analyze were: 1) coplanar structure, 2)
whether the congener is a principal component of commercial Aroclor mixtures, 3) whether
the congener had been reported in related studies involving fish of the same or similar
species and, 4) whether the congener had been detected in human blood serum, adipose
tissue, or mother's milk. In addition, at least two congeners were selected from each
chlorination level to allow for homolog determination. Use of these criteria resulted in the
selection of 47 congeners out of a possible 209.
I
All twenty samples were analyzed for mono-ortho and di-ortho PCB congeners,
additional congeners that met the selection criteria, PCB homologs, DDT and its metabolites,
dieldrin, chlordane, and percent lipid. Because of budget constraints, only four of the
samples, one from each sample area and size category, were analyzed for the non-ortho
substituted PCB congeners.
Total PCB varied among the 20 samples from a low of 0.463 ppm to a high of 2.253
ppm. Although mean values on the spawning grounds were nominally higher (1.07 ppm)
than from Delaware Bay (0.732 ppm), ANOVA and nonparametric tests revealed that these
differences were not statistically significant (alpha=0.05), nor was there a statistically
significant difference in mean PCB concentrations of commercial size fish versus recreational
size fish. Although total PCB content did not differ between size or location, the level of
chlorination in the recreational size fish from the spawning grounds (58 percent) was
statistically higher than any of the other categories (between 55 and 56 percent); This
finding is believed to have important implications to risk projections. Namely, some
evidence exists (BEHR, 1991) to suggest that only PCB mixtures with an overall level of
chlorination of approximately 60% represent a threat of cancer in animals. Other evidence
suggests that the cancer effects of PCBs are determined not so much by level of chlorination,
but more fundamentally by the extent to which individual congeners attain coplanar
conformation and hence, structural similarity to 2,3,7,8-TCDD. These issues are explored
as a part of the risk assessment portion of the Delaware Estuary Striped Bass Study.
Total coplanar PCB was found to increase with increasing total PCB. Toxicity
equivalents for the 4 samples selected for full PCB characterization ranged from 61 pptr to
95 pptr. The mono-ortho congeners were found to represent roughly two-thirds to three-
quarters of the computed toxicity equivalents. Of this percentage, congeners 126, 105, 118,
and 167 predominated.
Excess lifetime cancer risk was estimated by combining mean contaminant
concentrations, standard exposure factors, and potency slopes. The linearized, multistaged
model of carcinogenesis was used. Four separate fish consumption scenarios were
considered within each of four separate hazard/potency scenarios. Fish consumption
scenarios considered included one 8 oz meal/yr, two 8 oz meals/yr, one 8 oz meal/mo, and
one 8 oz meal/wk. The four hazard/potency scenarios considered included:
4-26
-------
A: q* = 7.7 applied to total PCB for all samples,
regardless of level of chlorination.
(traditional EPA policy on PCBs)
B: q* = 7.7 applied to total PCB, but only for samples
with level of chlorination approximately
equal to 60 percent
C: q* = 1.9 applied to total PCB, but only for samples
with level of chlorination approximately
equal to 60 percent (IEHR recommendation)
D: Toxicity Equivalence approach. TEFs from Safe combined
with q* of 2,3,7,8 TCDD of 1.56E+05
Method A and Method D yielded similar risk estimates. Because neither is dependent
on level of chlorination, they would both apply to the entire study area. Both methods A and
D project cancer risk in excess of l-in-100,000 assuming as little as one 8 ounce meal of
striped bass per year. At a top end consumption of one 8 ounce meal per week, risk
projections increase to in excess of l-in-1000 using both of these methods.
Method B and C are based on the premise that only PCB mixtures with approximately
60% chlorination represent a cancer hazard. The recreational sized striped bass from the
spawning ground was the only category which met this criterion. Using Method B for those
fish, it only takes one 8 ounce meal per year to exceed a l-in-100,000 risk, whereas it takes
two 8 ounce meals per year to exceed this same risk level using Method C. If Method B or
C are used, the assessor must still consider the cancer risk associated with chlorinated
pesticides in the stripers taken in the Delaware Bay. Due to chlorinated pesticides alone, a
person would have to eat one 8 ounce meal of striped bass per week taken from the Bay in
order for their cancer risk to exceed in l-in-100,000 level.
Although Delaware has not taken an official position on this matter, management
options being considered include the following:
Limited consumption advisory for entire geographic region (based on Method
A or D).
Limited consumption advisory for spawning ground only (based on Method B
or C).
No advisory - "wait it out" until fish consumption survey is complete and
consensus reached on PCB risk assessment.
4-27
-------
Upon completion of the fish consumption survey, Delaware intends to further refine
its current risk estimates through a probabilistic risk analysis. Such an approach allows the
assessor to construct a more complete picture of the risk continuum based upon assumed
probability distributions of the various exposure factors. An illustrative example was
prepared as a part of this study to suggest potential future refinements.
4-28
-------
w
KU
QC
H
LJJ
LJL
UJ
o
CO
UJ
Q
O
TE SAMPLIN
CO
O
Q.
^
O
o
UJ
5
O
CL
HI
QC
*
CO
2 LOCATIOr
Q
<
CLASSES
UJ
Nl
CO
CM
*
O
LL
O
UJ
CL
CO
O
O
O
X
Q
Z
O
LL.
O
UJ
CL CO
WCO
QC >:
UJ J
Z <
UJ *s
0 <
0 0
0 CL
*
I
QC
UJ
O
Z
0
o
z
Q
UJ <
C H
UJ <
Q Q
CO 2
og
°t
£2
b =>
3 CO
UJ O
QC O
*
CO
CO
z
CO
QC
O
1
m
CD
O
QC
CL
*
4-29
-------
RITERIA
O
O
HMM
O
UJ
UJ
CO
cc
UJ
z
CONGE
00
o
Q.
a
UJ
H
3
h;
CO
00
Q-CO
ui O
^ x
•V (••
H 0
mo
^^
CO Q
' iM
O H
DC|=
O CO
1 CQ
'LANARS (NOI
JO-ORTHO S
81
*
h-
LU
CO
O
O
CC
3
JCIPLE AROC
4mm
CC
Q.
CO
LL.
z
s
DC
O
VIOUSLY REP
UJ
cc
QL
CO
cc
UJ
X
i«
~ a:
DC ^
UJ ¥>
ff\ f^
Z o
<^
§1
UJ 3-
CC LL.
00
VIOUSLY REP
< OR TISSUE
UJ j
CC g
*
i LEVEL
o
S
o
X
o
o
UJ
o
cc
LL.
CO
cc
III
LU
UJ
•'
o
O
CM
CO
UJ
j:
O)
o
CM
LU
•••i
CO
CO
CO
o
a.
LL.
O
CO
cc
UJ
47 CONGEN
5
CO
LU
QC
4-30
-------
Q
O
I
LU
•E
.
<£
O
>I
mmJt •
^f
Z
1 ,
DETECT
LEVEI
o
O
*3L
LU
LU
Z
^
Q.
Q.
O
7
CO
cc
^•^
O
O
oc
"O
« «
CD a
O oT
Q. CD
0 0
£Q- w
*r O <1>
? •= 2
o o -2
c i "JS
o .1 S
Son.
Q.
a.
CN
co .2
*sz *"•
1^
••• p^an
I T3
o §•
C^ ^MVf
I g
CO
CD
O
D_
O
•c
o
1
. c
o
&
a.
Q.
^Lm
O
T
o
CO
CD
g
LU
€0
O)
O
o
o
CO
a
a.
4-31
-------
RESULTS
PCB CHARACTERIZATION
*NO SIGNIFICANT DIFFERENCE IN TOTAL PCB
BETWEEN SIZE CLASSES OR LOCATION.
*RECREATIONAL SIZE STRIPERS FROM SPAWNING
GROUND EXHIBIT STATISTICALLY HIGHER
LEVEL OF CHLORINATION THAN OTHER
CATEGORIES.
INCREASES IN COPLANAR PCB CONTENT
ARE CORRELATED WITH INCREASES IN
TOTAL PCB.
*MONO-ORTHOs REPRESENT ROUGHLY 66% TO
75% OF THE TOXICITY EQUIVALENTS; NON-
ORTHOs REPRESENT ROUGHLY 20 TO 30%;
DI-ORTHOs GENERALLY COMPRISE <5%.
*CONGENERS 126, 105, 118, AND 167 MAKE
UP THE MAJORITY OF THE TOXICITY
EQUIVALENTS.
4-32
-------
LLJ O
CD Q
°
Q_
_ CO
2 CO
is
— LLJ
CO
IO ::::*:;*
0 C
N
10 "
iq
T-
">
T"
M Ll
T
IO
00
3"
CO
o"
CO I
q I
N
q
T-
1'
IO
h-
r<
q
D • i- L
C
CO
00
IO
o"
D C
a
D
(
<
*
i
I
i
\
1
i
t
i
1
i
1
<
1
1
I
D
O
O
CD
O
LU
DC
CD
O
O
O
CO
o
LLJ
DC
O
CO
uudd ONOO
4-33
-------
CO
O
^1
o
^
o
CD
CL
Z
JZ CD
O x
^••^ ^^L
^^^y
LU °|X|l
DC
<
cr """
2 m
JC CL
LL <
oco
HIH^
LU —
' §
0
0
T—
X
• ^MBP
LLI
CL
^
CO
z
CD
0
o
cj
O
o
II
••••
1
BIPHENYI
0
-!
O
v£>
II
-<
LU
__i
CL
•^
CO
CD
0
CL
<
O
a
^
CO
?
LU
z
1
O
LL
O
•
•BC
«*?
X
II
LL
O
Z
g
-------
o
EQ
'S
o
LU
O co
DC CO
HJ <
0-0
Z LU
— N
Z CO
DC
D C
0 L
CO
T-
00
w
& •
» iiii
bo PPI
10 *
J) C
O L
CM
•
r-s.
to
•
00
lii^illl! IO
!!H!!!!!! h-
UD
o r
O L
•*
CO
CO*
10
(D
00
IO
LO
in
(0
in
«- C
Q u
•^r
w"
in
CM
CVJ
in
in
CO
10
10
0 L
0 L
_ CM
CO
•
^r
LO
CO
•
_ -^f
LO
O *
O U
(
<
«
t
•
i
(
\
1
1
1
<'
(
(
(
{
1
1
t
r>
O
o
DQ
O
LU
CC
CQ
O
o
CO
o
LU
DC
CD
CO
%
4-35
-------
(0
CQ
O
Q.
CC
Q.
O
O
s
O
CO
CO
QC
LLJ
(0
CQ
O
Q.
O
^
•
o
CO
o
en
CM
•
O
Q.
O
o
o
CM T-
uidd
CO
•
O
CO ^- CM
• • • •
O O O
1V1O1
4-3«
-------
CO
o
Q_
LU
O
u_
O
z
LJ
O
HI
O
X
o
CM
h*.
q
CO
in
to
CO
CQ
O
CD
CO
en
o
0)
^
o
o
CO
CM
or
O
o
o
CO
o
CO
CM
ONOO
tr
LLJ
(D
HI
Q.
<
^~ ^
O
o
o
u.
LLJ
H
UJ
4-37
-------
CO
CO
O
°-co
DC H-
< Z
Li!
o
o
Q_
R5
O o
U_ LJU
Eo
i- i-
en
o
o
N.
c\i
CO
O
I
CO
^t
o
LU
D.
<
CO
cr
O
CD
cvi
0
CO
CM
oc
111
Q.
CO
CD
CO
DC
LU
<
CO
4-38
-------
CO
CO Q
CO
LU
CO
CO
CO O
< o
^
o
LL
Q
X
Q
LU
X
DC
O
x
O
II
LU
ID
LL
X
Q3
CO
DC
CO "-
co j^J
LU O
^<
Luc§
O)
o _
Q.CO
Q-E
-r LL
coO
E LU
Is
O ^
00
^-co
ID Q to
CO LU ^
O > ~
6ffi i
DC
il
z {—
s%
< ^
HID
Z CO
°S
°8
H £ H-, 3
~7 \^> ?>* LU
? < LU o^-~
/*; LU
-si
pi
^ LL LU
o:0^
^ZS
> f~\ •**
LL
o
•
o
z
o
o
II
HUMAN
O)
DC
LU
CO
z
o
o
X
CO
LL
CO
>o^
Ogee
zc5^
/-^k -^- _
DC
O O
EIGHT
>• -^
Q F
II II II
Q LL
LU Q
II
LU
DL
,3
rCO
£gS
b z
< LU
g§
*£. \u _j
Q g LU cc
iE£
§^5
Q Q S
*•
CE
O LU Z
§y=<
_J
O
II
4-39
-------
CO
DC
O
I
O
^^
u_
LU
DC
CO
O
a.
X
LU
Q
LLJ ^
CO >
CO
<
cc
o
^ ^ O ^
^^^ ^"^^ ^™ 5*
co 75 co co
CD Q CD CD
E £ E E
N N N N
0 o 0 0
CO co 00 00
0 Q 0 0
0 j| § 0
DC
0
(0
o
CO
D)
J£
O
£2
>*
in
Q
LLJ
LJL
D
CD
4-40
-------
CO
z
nr
Q_
CC CO
cc >
LU CC
O <
O CO
cc
<
CO <
|8
O
O
E | O *
5 tl!
< < <
LU UJ LU
CM
EA
LLI
y//////.
v//////////,
V//////Y///////,
w/////////////,
y//////
O
y//////
y//////
D
pi
a. a.
a. co
08 u_
CD UJ
Q ^
O W
•^ "C
I- O
LU UJ
2D=
• a
LU
O
Q
O
LU
CO
UJ
o
z
Z)
o
oo
4-41
-------
4-42
-------
DC
CO
CO
H-
Z
CC LU
CO
O
Q.
CO
Q
Z
LU
O
DC £
CC DC
CO &
(D Q-
>: LLI
H CL
Z
HI
DC
CC
D
O
z en
~ LU
f- °
£iz
z ^ p
D -*- <
LU ^
CO ^ D
CC !^ 1
LU ^ ^
£co8
WLU°
Z CO Q
O co z
o < <
z
f
c:
LU
o
X
CO
O
a.
a.
O
z
<
Z CO
D D
LU O
LLI
DC
CO
LU
111
LU
CO
X
0
D
CO Q O
or. !£
X
H
CO
-~£
cc
5; ^ co
z
o
§t
^"Sl^
^?com
r5 •*• -y ^
CO
O ^
LU
co
CO
LLI
CO
CO
o
9 <
o
DC
0.
O
X
5 £
gz
co P
co
LU
CO
z z o
"12 H
Q b $
LU < Z
111
DC U_ 2
I LU X
< O CO
020:
0 I LU
o o |±!
LU D O
O S DL
X
o
<
o
a:
o_
CL
<
HI
h-
<
LU
O
HI
cc
LU
D.
4-43
-------
4.4 MICHIGAN
John L. Hesse, Chief, Site Assessment Section, Michigan Department of Health
The Great Lakes have received considerable attention and publicity in terms of fish
contaminant problems, perhaps more than any other region of the nation. Part of this
attention is because the Great Lakes represent the largest freshwater resource in the world.
They have also been more thoroughly studied than other areas.
Some of the nine Great Lakes jurisdictions initiated fish consumption advisory
programs more than 20 years ago, dating back to the discovery of mercury and PCBs in fish
in 1970. This presentation will focus on how Michigan and other Great Lakes states have
issued advisories in the past and will provide a preview of a proposed advisory protocol
currently being considered for uniform application by all states in the region.
While a few isolated advisories have been issued because of DDT, chlordane, toxaphene,
mercury, and dioxins in the Great Lakes over the years, PCBs continue to be the primary
chemical group responsible for most consumption advisories.
Trend monitoring conducted by EPA and other agencies provides clear evidence of
rather dramatic declines of PCB levels since the mid-1970's, approximating a 90 percent
drop in concentrations found in many popular Great Lakes sport fish species. During this
same time interval, however, epidemiologic and toxicologic research into the potential health
effects of PCBs on animals and humans has increased our level of understanding about fish
consumption patterns, specific congener toxicities, possible modes of action, and most
sensitive endpoints of concern. It is appropriate that we continue to use this new information
in refining our "risk management" criteria and strategies (i.e., fish consumption advisories).
In the 1970's, when pesticide and PCB levels were the highest in Great Lake's fish,
most of the states which had advisory programs in place depended largely upon FDA Action
Levels as the basis for advice to anglers. It seemed to make sense that advice to people who
were eating sport-caught fish should be comparable to protection being provided through
regulation of fish being sold commercially.
By 1978, with the expanded use of risk assessment in our regulatory programs,
Michigan started to more critically evaluate the trigger levels for our advisories. When FDA
was ordered by the courts in 1979 to raise their action level for mercury from 0.5 ppm to
1.0 ppm due to economic impacts upon marine fisheries, Michigan decided not to adopt the
new standard for fish advisories because we felt that the toxicology of mercury supported
continued use of 0.5 ppm as a trigger level.
In 1979, when FDA issued notice of an intent to lower the PCB action level from 5
ppm to 2 ppm, Michigan began our own evaluation of the PCB trigger level and decided to
adopt the 2 ppm value in 1981 for fish advisory purposes. FDA's decision to lower the
standard to 2 ppm didn't occur until 1983.
4-44
-------
As part of an initiative to establish new guidelines for surface water quality discharge
limits, Michigan had adopted a standardize cancer risk estimation procedure and a level of
acceptable risk of 1 X 10"5. Using assumptions and cancer risk models which were fairly
well accepted in 1981, we concluded that 2 ppm for PCB and 0.3 ppm for chlordane each
approximated a 10"5 cancer risk and that they were appropriate trigger values for fish
consumption advisories. Similarly, we adopted a 10 part per trillion advisory level for
dioxin based upon a 10"5 estimated cancer risk. We recognize, however, that the application
of more conservative models and assumptions since the early 1980's have increased the
estimated risk at these levels. As I will discuss later in my presentation, this may be
irrelevant because of a tentative decision by the Great Lakes states to use adverse
reproductive outcomes rather than cancer risk estimates as the principal basis for fish
consumption advisories.
Because differing approaches to fish advisories by neighboring jurisdictions causes
unnecessary confusion for the public, all of the Lake Michigan states (Michigan, Indiana,
Illinois, and Wisconsin), began an effort in the early 1980's to reach consensus on issues
associated with sampling, analysis, and advisory criteria. The U.S. EPA Region V office
provided assistance, and, by 1985, we had actually reached agreement on all major Lake
Michigan fish monitoring and consumption advisory issues.
The Lake Michigan states knew that the FDA Action Levels were being criticized as
being out-dated and perhaps not adequately protective of the developing fetus and of anglers
who tend to eat more fish than the general public. On the other hand, we wanted to retain
some association with how fish from the Great Lakes or other area waters were being
regulated for commercial sale.
To partially take into account the extra sensitivity of the fetus and higher average
consumption rates of anglers compared to the general public, we decided to initiate a "no
consumption" advisory for women and kids when only 11-49 percent of the samples for a
species exceeded any of the FDA action levels. At this frequency, the general population
was advised to eat no more than 1 meal per week. When 50 percent of the samples
exceeded an FDA action level, a "no consumption" advisory was issued that applied to
everyone.
While this lacked a true scientific basis, we felt that it would be less confusing to the
public than completely divorcing ourselves from the FDA regulatory numbers. A survey in
Wisconsin conducted in 1985 showed that about 90 percent of anglers surveyed were aware
of the consumption advisories and 60 percent were modifying their consumption patterns
accordingly. We feared that we would lose voluntary compliance with the advisories if we
changed to a system that treated anglers grossly different from consumers of commercially
harvested fish from the same waters.
The Lake Michigan states also began to annually pool monitoring data and coordinate
sampling plans. This all worked out so well in Lake Michigan that the Governors of all the
4-45
-------
Great Lakes states signed an agreement (Governors' Great Lakes Toxic Substances Control
Agreement, 1986) mandating common fish consumption advisories on each of the Great
Lakes. While the jurisdictions were able to comply with the spirit of the agreement by the
1987 fishing season with advisories that were essentially uniform, we have not been able to
reach agreement on common criteria, even after several years of effort.
For several years, we have essentially maintained a moratorium on significant changes
to the Great Lakes advisories pending final development of uniform criteria. We have added
species to the advisory as necessary but have hesitated to delist species as contaminant levels
have declined (because of the likelihood of new criteria being more conservative).
In 1993, the Great Lakes Sport Fish Advisory Task Force, currently co-chaired by the
Wisconsin Department of Health and Social Services and the Wisconsin Department of
Natural Resources, has tentatively reached agreement on a Protocol for a Uniform Sport Fish
Consumption Advisory for the Great Lakes' region (GLSFATF, 1993). The proposed
protocol will likely be undergoing peer review in next few months and hopefully will be
ready for implementation for the 1994 advisories.
The proposed protocol involves use of a Health Protection Value (HPV) of 0.05
ug/kg/day maximum ingestion of PCBs. The goal of the advisories will be to keep the PCB
ingestion via sport fish consumption below 3.5 ug PCB per day for a 70 kg person. The
Health Protection Value approach primarily focuses on protection against neuro-
developmental effects in infants born to exposed mothers. The selected value is from an
aggregate of several human and animal studies showing similar thresholds of effects. The
protocol assumes an average meal size of 227 gms (1/2 Ib) and an average adult weight of 70
kgs.
Perhaps unique to our proposal, we would be using a 50 percent estimated reduction
factor for residues in the untrimmed raw fillet due to losses through trimming and cooking.
Recent research at Michigan State University (Zabik, M.E., et al., 1993), coupled with other
research on contaminant reductions, support at least a 50 percent reduction as a conservative
estimate. The draft EPA Sampling and Guidance Manual (US EPA, 1993) cites 60-90
percent reductions possible through trimming and cooking but does not propose an
adjustment to the screening values.
The protocol establishes 5 advisory categories for different rates of consumption
(unrestricted; 1 meal/week; 1 meal/month; 6 meals/year, and; no consumption). Fish are
placed into these groupings according to concentration ranges that would not result in
ingestion of more than 3.5 ug PCBs/day. Provisions are made for adjustments in the
advisory if PCBs are not the predominant contaminant in fish from a specific location.
Due to the high level of uncertainly associated with cancer risk projections, the Great
Lakes Task Force has rejected use of such extrapolations as the primary basis for PCB or
other organochlorine compound fish consumption advisories . However, the advisory
4-46
-------
information provided to the public will include general statements about potential cancer risks
from fish consumption. As presented earlier, emphasis will be given to protection against
adverse reproductive impacts and other non-cancer endpoints, and by doing so, other
potential adverse effects will be minimized also.
Persons interested in obtaining a copy of the Proposed "Protocol for a Uniform Great
Lakes Sport Fish Consumption Advisory" should write to James F. Amrhrein, Wisconsin
Department of Natural Resources, Bureau of Water Resources Management, 101 S. Webster
Street, Box 7921, Madison, WI 53707-7921.
References Cited
Great Lakes Council of Governors. "The Great Lakes Toxic Substances Control Agreement,"
May 21, 1986.
Great Lakes Sport Fish Advisory Task Force. "Proposed protocol for a uniform sport fish
consumption advisory," Draft, March 1993.
US EPA. "Fish sampling and analysis: A guidance document for issuing fish advisories,"
Draft, Office of Science and Technology, February, 1993.
Zabik, M.E., M.J. Zabik, and H. Humphrey. 1993. "Assessment of contaminants in five
species of Great Lakes fish at the dinner table," Part I, Final Report to the Great Lakes
Protection Fund, March, 1993.
4-47
-------
4.5 SUMMARY OF QUESTIONS AND RESPONSES12
4.5.1. A representative from the Massachusetts Division of Marine Fisheries asked Dr.
Pollock how he deals with the issue of risk comparisons. How are the viewpoints of
colleagues who favor the use of risk comparisons factored into risk assessments and risk
communication work?
Dr. Pollock stated that there have been suggestions about comparing exposures due to
fish consumption to other sources of exposure; however, at this point, such comparisons have
not been used in California. In part, this is because a total diet approach might be needed
and that is not being done.
4.5.2. Dr. Bolger then commented to the panel about several difficulties facing people
involved with risk assessment. Some methodologies, developed years earlier, are not well
suited for evaluating environmental contaminants where we are evaluating degrees of risk,
rather than a simple question of "safe or unsafe." Other approaches (e.g., probability
distributions) may require further education of those who must use the assessment.
The second comment focused on the use of a nomogram, which describes all the data
in one pictorial representation. Nonetheless it is difficult to interpret the resulting
nomograph because it is based on aggregate risk.
Ms. Cox acknowledged that there may be difficulties associated with adding all of the
hazard quotients to obtain a hazard index. However, she believed that this is a standard
protocol used in the Superfund program.
12
Ed. note: The question and response portion includes summaries derived from transcribed
conversations. The summaries have been carefully edited to present the discussion as accurately as possible.
However, these question and response summaries have not been reviewed by the speakers—unlike the proceeding
abstracts.
4-48
-------
PART FIVE
APPENDICES
A.I Speakers Biographies
Elizabeth Souther land, Ph.D.
Elizabeth Southerland currently directs EPA's Risk Assessment and Management Branch within
the Office of Water's Office of Science and Technology. The Branch is responsible for directing
sediment contamination programs and evaluating risks associated with chemical contaminants in
fish.
Dr. Southerland graduated with a Ph.D. in Environmental Engineering from Virginia
Polytechnic Institute and State University in 1980. She worked in State government and in
consulting engineering prior to joining EPA.
Rick Hoffmann
Rick Hoffmann organized the PCS workshop. Mr. Hoffmann is an environmental scientist in
EPA's Risk Assessment and Management Branch. The Branch is located in the Office of
Science and Technology within the Office of Water. The Branch is responsible for directing
sediment contamination programs and evaluating risks associated with chemical contaminants in
fish. Mr. Hoffmann works on fish contamination issues.
Prior to that, Mr. Hoffmann worked in EPA's San Francisco region where he held
various positions relating to water quality planning and pollution control as well as overall
environmental impact assessments. He has also worked for the Hawaii State Department of
Health. Mr. Hoffmann received a B.A. in Zoology from California State University at San
Diego and an MPH from the University of Hawaii's School of Public Health, with an emphasis
in environmental/occupational health.
Mitchell D. Erickson, Ph.D.
Mitchell D. Erickson is a Group Leader for Environmental Chemistry in the Environmental Research
Division at Argonne National Laboratory, Argonne, EL. His current research interests focus on the
improvement of radio analytical methods (with an emphasis on faster and less expensive routine methods)
and the evaluation, design, and testing of sensors and samplers for use in subsurface monitoring
(specifically for interfacing to cone penetrometers). He also provides technical contributions on PCBs,
-------
PCDDs, PCDFs, and other environmental pollutants through presentations, publications, and consulting.
He is the author of Analytical Chemistry of PCBs, first published in 1986, and is currently
working on the second edition for Lewis Publishers. He also authored the book Remediation ofPCB
Spills (Lewis, 1993).
John H. Craddock, Ph.D.
John H. Craddock is founder and principal of Craddock Associates, Incorporated Regulatory and
Environmental Consultants based in St. Louis, MO. Dr. Craddock also is a Senior Consultant with
RegNet Environmental Services in Washington, DC. Dr. Craddock has more than 30 years experience
in the chemical industry, including wide-ranging involvement with state and federal regulatory and
compliance issues. For the past 13 years, he has worked closely with federal and State agencies
responsible for regulating toxic chemicals, particularly PCBs. For more than 12 years, he has managed
PCB regulatory and compliance issues for Monsanto Company in St. Louis.
Dr. Craddock was an active member of the Chemical Manufacturers Association PCB Panel since
its inception in 1980 and its chairman from 1984 to 1992. Working closely with the EPA in the
development of all major federal PCB regulations since 1981, he has led industry advocacy efforts to
develop cost-effective, technically achievable rules. He received his B.S. in Chemistry from Memphis
State University and his Ph.D. in Chemistry from Vanderbilt University.
P. Michael Bolger, Ph.D., D.A.B.T.
P. Michael Bolger is the Chief of the Contaminants Standards Monitoring and Program Branch in the
Center for Food Safety and Applied Nutrition of the U.S. Food and Drug Administration (FDA) in
Washington, DC.
Dr. Bolger received his B.S. hi Biology from Villanova University and his Ph.D. in Physiology
and Biophysics from Georgetown University. After a two year postdoctoral position with the Department
of Physiology in the Georgetown University Medical Center, Dr. Bolger became a staff fellow in
toxicology with the Bureau of Foods hi the FDA. Upon completion of the staff fellowship, he accepted
a position as a lexicologist with the contaminants branch at FDA. Over the last decade, Dr. Bolger has
been involved in a number of hazard/risk assessments of food contaminants, including PCBs. Dr. Bolger
is a board certified toxicologist by the American Board of Toxicology. Dr. Bolger is currently Chief of
the Contaminants Standards Monitoring and Programs Branch, which is responsible for the monitoring
and hazard/risk assessment of environmental contaminants in the food supply.
Vincent James Cogliano, Ph.D.
Vincent James Cogliano is the Chief of EPA's Carcinogen Assessment Statistics & Epidemiology Branch.
In this capacity, he is responsible for evaluating human studies and using human and animal studies to
estimate the risks to human health from exposure to environmental pollutants.
A-2
-------
Dr. Cogliano received his Ph.D. in Operations Research from Cornell University in 1982. From
1981 to 1983, he worked as a manufacturing engineer for the IBM Corporation in New York. In 1983,
he moved to the Washington area and began working for the U.S. Environmental Protection Agency,
where he has held positions in the Office of Pesticide Programs; the Office of Policy, Planning, and
Evaluation; and the Office of Health and Environmental Assessment.
John L. Cicmanec, D.V.M.
John L. Cicmanec is a Veterinary Medical Officer of the Systemic Toxicants Assessment Branch in the
Environmental Criteria Assessment Office of EPA's Office of Research and Development in Cincinnati,
Ohio.
Dr. Cicmanec is a Research Veterinarian who presently works as a risk assessor in the Cincinnati
office of EPA. Prior to joining the staff of the Environmental Criteria Assessment Office he directed the
operation of the research animal facility of EPA in Cincinnati. Prior to the 8 years that he has spent with
EPA, he spent 16 years at a clinical veterinarian and study director for private animal research contract
firms in the Washington, DC area. During this tune, as a study veterinarian, he conducted a subchronic
reproductive research study involving the effects of Aroclor 1254 on a large group of rhesus monkeys.
Dr. Cicmanec is a Diplomate of the American College of Laboratory Animal Medicine. In addition to
his veterinary training, he received a M.S. from the University of Michigan Medical School.
Donald G. Barnes, Ph.D.
Donald G. Barnes has held numerous senior science advisory positions with the U.S. Environmental
Protection Agency in Washington, DC, including EPA's Science Advisory Board (SAB). The SAB,
which are organized into nine committees, is composed of 16 staff members and a Board of more than
300 distinguished scientists and engineers. The SAB committees meet 50 to 60 times a year and produce
nearly 40 reports annually. Dr. Barnes currently serves as a member of the EPA Risk Assessment Forum
and the Risk Assessment Council. He also participated in the development and adoption of EPA risk
assessment guidelines for cancer, exposure assessment, and complex mixtures.
Dr. Barnes is an international expert in the toxicology, exposure, and risk assessment of 2,3,7,8-
tetrachlorodibenzo-p-dioxin, and has published more than a dozen papers on the subject. He served as
a consultant to the World Health Organization on "dioxin" hi municipal solid waste combustion and as
a contaminant in human milk. He also led a successful effort to develop an international consensus on
"Toxicity Equivalency Factors" (TEFs) for conducting risk assessments for 210 "dioxins and furans."
Dr. Barnes represented EPA for more than a decade on an interagency science panel established by the
White House to address Agent Orange.
He received his B.A. hi chemistry from the College of Wooster and his Ph.D. from Florida State
University with a major hi chemistry and a minor in physics.
A-3
-------
Theodora (Theo) Colborn, Ph.D.
Theo Colborn, a Senior Fellow with the W. Alton Jones Foundation and the World Wildlife Fund,
manages their Toxics and Wildlife Program. In 1985, she was awarded a Fellowship at the Congressional
Office of Technology Assessment. In 1987, she moved to the Conservation Foundation where she
provided scientific guidance for the book, Great Lakes, Great Legacy?, released in 1990 in collaboration
with the Institute for Research and Public Policy, Ottawa, Canada. Recently, Dr. Colborn edited a book,
Qieinically Induced Alterations in Sexual and Functional Development: The Wildlife/Human Connection.
Dr. Colborn has testified before the U.S. House and Senate, lectured extensively, and served in an
advisory capacity to state, Federal, and international groups concerning the non-cancer hazards of
exposure to toxic chemicals. Currently, Dr. Colborn serves on the Science Advisory Committee of
UNEP's Marine Mammal Plan Task Force; she also serves on the Science Advisory Consultant Group
authorized under the Great Lakes Critical Program Act of 1990 for the Agency for Toxic Substances and
Disease Registry. She holds an adjunct faculty position with George Mason University and is a member
of the Rocky Mountain Biological Laboratory, the Colorado Field Ornithologists, and the Society of
Environmental Toxicology and Chemistry.
Dr. Colborn earned a Ph.D. hi Zoology (distributed minors hi epidemiology, toxicology, and
water chemistry) at the University of Wisconsin-Madison; an M.A. in Science at Western State College
of Colorado (fresh water ecology); and a B.S. hi Pharmacy from Rutgers University College of
Pharmacy.
John F. Brown, Jr., Ph.D.
John F. Brown Jr. is the Manager of Health Research for General Electric Corporate R&D in
Schenectady, NY. Dr. Brown received his B.S. from Brown University and Ph.D. from MIT, both in
chemistry, and later took postdoctoral training hi clinical pathology at the SUNY Upstate Medical School
in Syracuse. Except for his post-doctoral training, Dr. Brown's entire professional career has been spent
in various research or research management positions at General Electric, where he was initially
concerned with chemical synthesis and molecular structure, and more recently with medical diagnostics,
and health risk assessment and environmental biodegradation, especially as it relates to PCBs.
Dr. Brown has served as a member of an NHLBI study of blood-compatible materials, the MIT
SCEP Study of Critical Global Environmental Problems, and the recent IEHR Workshop on Hydrophobic
Organic Chemicals Bioaccumulation. He also has served as Natural History Chairman for the Adirondack
Mountain Club, Chairman of the Schenectady Museum Nature Preserve Committee, and long-time
member of both the Niskayuna (Town) and Schenectady (County) Environmental Management Councils.
Ted R. Schwartz
Ted Schwartz is the Chief Chemist of the U.S. Department of the Interior Fish and Wildlife Service's
National Fisheries Contaminant Research Center (NFCRC) hi Columbia, Missouri. Mr. Schwartz is
responsible for the administration and management of the Chemistry Division of NFCRC. He evaluates
and determines the nature, extent, and limits of equipment, facilities, and data needed to accomplish the
Division's goals. He is responsible for staff development and direction of chemistry research. His
A-4
-------
personal research interest is in the interpretation of complex environmental patterns of organic residues
hi aquatic ecosystems using chemometric data analysis techniques.
Mr. Schwartz received a B.S. degree hi Chemistry from the California State University at
Humboldt in 1978, and a M.S. degree in Organic Chemistry at the University of Missouri in 1982.
Margaret M. Krahn, Ph.D.
Margaret Krahn serves as the Assistant Program Manager of the Environmental Conservation Division's
Environmental Chemistry Program. This program is located at the Northwest Fisheries Science Center
of the National Marine Fisheries Service, which is a part of the National Oceanic and Atmospheric
Administration in Seattle, Washington.
Dr. Krahn has developed state-of-the-art methods for determining trace organic contaminants,
such as polynuclear aromatic hydrocarbon metabolites and coplanar PCBs in marine samples. Among
the techniques she uses are high-performance liquid chromatography, fluorescence spectrometry,
photodiode array (ultraviolet) spectrometry, gas chromatography, and mass spectrometry. Dr. Krahn has
played a key role in developing new screening methods to rapidly determine contaminant levels in bile
and tissues from marine animals and in marine sediments. In addition, she has developed and automated
procedures for the cleanup of sediment and tissue extracts before analysis by gas chromatography/mass
spectrometry. She has published extensively and many of her published methods have been adopted for
use by government, academic, and private laboratories.
Prior to joining the Environmental Conservation Division in 1978, Dr. Krahn taught chemistry
at the University of Delaware. She earned her B.S. in Chemistry from the University of Minnesota and
her Ph.D. in Organic Chemistry from the University of Washington.
Leon D. Sawyer
Leon D. Sawyer is the Branch Chief of the Methods Research Branch, Division of Pesticides and
Industrial Chemicals at the Food and Drug Administration's Center for Food Safety and Applied Nutrition
in Washington, DC.
Mr. Sawyer has worked for the U.S. Food and Drug Administration for 29 years; specifically,
he has worked 25 years as a chemist, research chemist or scientific coordinator, and the last four years
as a supervisory chemist hi FDA's Division of Pesticides and Industrial Chemicals. He has been involved
with analytical and regulatory issues relating to PCBs since their identification as environmental
contaminants in 1969 and has actively participated hi numerous workshops and symposia relating to them.
He served as an Associate Referee hi the Association of Official Analytical Chemists (currently AOAC
International) on four different residue topic areas, and he has served for 18 years (1971-1989) as the
Associate Referee on PCBs. During this time, two AOAC Collaborative Studies on the analysis and
quantitation of PCBs were conducted which are recognized as "Official Methods of Analysis" by AOAC
International. Also during this tune period, investigations were initiated on an individual congener
approach to PCB identification and quantitation, which will be the subject of this presentation.
A-5
-------
Brian Bush, Ph.D.
Brian Bush is a Senior Research Scientist for organic analytical toxicology for the New York State
Department of Health, Wadsworth Laboratories hi Albany. Dr. Bush also is an associate professor at
the School of Public Health Sciences, State University of New York at Albany. Dr. Bush has published
nearly 70 professional articles on organic analysis topics including a number of articles on PCB analysis.
He received his B.S. (with honors) in Chemistry from the University of Leeds, the United Kingdom and
received his Ph.D. hi the Analysis of Steroids also from the University of Leeds hi the United Kingdom.
Dr. Bush is a Fellow of the Royal Society of Chemistry and a Member of the Society for Analytical
Chemistry.
Gerald Pollock, Ph.D.
Gerald Pollock is a Staff Toxicologist and Acting Unit Chief of the Fish and Sediment Contamination
Unit for the California EPA, Office of Environmental Health Hazard Assessment, Pesticide and
Environmental Toxicology Section in Sacramento. Presently, he is responsible for the evaluations of
human health hazards associated with consumption of chemically contaminated seafood. He also managed
the Department of Health Service's studies of the contamination of fish in southern California and
Monterey Bay. He has conducted risk assessments of consuming fish contaminated with DDT, PCBs,
and dioxins, and has served as a consultant to the Committee on Wastewater Management for Coastal
Urban Areas of the National Research Council. Previously, Dr. Pollock has held positions as both an
Assistant Professor of Toxicology in the Regional Program in Veterinary Medicine at the University of
Idaho, and as a Research Supervisor for Animal Metabolism at Diamond Shamrock Corporation.
Dr. Pollock received both his B.S. in Biochemistry and his Ph.D. in Pharmacology and
Toxicology from the University of California at Davis. He is a Diplomate of the American Board of
Toxicology.
Janice P. Cox
Janice P. Cox is currently a Project Leader for the Tennessee Valley Authority's Water Resource Issues
Analysis projects. She developed the issues analysis format to identify and evaluate — on a watershed-
wide basis - sources of impacts to water quality and their potential for impact on public health. Issues
analyses have been used for defining needs for monitoring, resource protection, and impact mitigation
projects.
Ms. Cox currently is a member of TVA's Hiwassee River Action Team, charged with identifying
water resource problems and building inter-agency coalitions to implement solutions on a watershed basis.
Previously, she was an Associate Editor for the International Journal of Childbirth Education, 1986-1988,
which focused on the potential impacts of environmental stresses on fetal development and birth
outcomes.
Ms. Cox received a B.A. in Chemistry from the University of Arkansas, an M.S. in Phycology
from the University of Arkansas, and a C.E. in Risk Assessment, Toxicology, and Risk Communication
from The Johns Hopkins University.
A-6
-------
Richard W. Greene
Richard W. Greene is an Environmental Engineer for the State of Delaware Department of Natural
Resources and Environmental Control's Watershed Assessment Branch. His principal focus is the
assessment of ecological and human health risks associated with toxics in surface water, sediment, and
biota. Mr. Greene has been instrumental in developing Delaware's Toxics in Biota Program and was
involved in all three fish consumption advisories issued for Delaware waters. Most recently, he was the
driving force behind the transition to a risk-based approach to evaluating fish contaminant data in
Delaware. As part of Delaware's transition to a risk-based approach, Mr. Greene identified the important
link between PCB analytical methods and subsequent risk characterization.
John L. Hesse, M.S.
John L. Hesse is an Environmental Health Administrator within the Michigan Department of Public
Health, currently serving as Chief of the Site Assessment Section. Mr. Hesse has been with the
Department of Public Health for 14 years. Prior to this, he worked for 11 years with the Michigan
Department of Natural Resources hi toxic chemical monitoring and control. He-was involved with the
identification of the PCB problem in the Great Lakes in 1970 and assisted in the passage of legislation
at the state and Federal level to ban the use of PCBs. As one of two primary program areas under his
supervision, Mr. Hesse is currently principal investigator of a cooperative agreement from the Federal
Agency for Toxic Substances and Disease Registry for Michigan to conduct health assessments at
Superfund contamination sites. He also coordinates the fish consumption advisory program for the State
of Michigan. Mr. Hesse works closely with other members of the Division of Health Risk Assessment
who are involved with ongoing investigations to determine human health effects of eating Great Lake fish.
Mr. Hesse received his B.S. from Utah State University and his M.S. in Aquatic Biology from
Michigan State University.
Deborah L. Swackhamer, Ph.D.
Deborah L. Swackhamer is an Associate Professor at the University of Minnesota. She directs the
Environmental Chemistry Program in the Division of Environmental and Occupational Health, School
of Public Health. Dr. Swackhamer's research interests include studying the chemical and biological
processes that control the fate of hydrophobic contaminants in aquatic systems. For the past 15 years,
one of her greatest emphases has been the processes controlling PCB behavior in the Great Lakes. Her
current research is on PCB accumulation by phytoplankton.
She received her B.A. in Chemistry from Grinnell College and her M.A. and Ph.D. degrees from
the Water Chemistry Program at the University of Wisconsin-Madison. After doing postdoctoral work
at Indiana University hi Bloomington, she joined the faculty at the University of Minnesota in 1986.
A-7
-------
A.2 - Speaker's Addresses
Don Barnes
Science Advisory Board
US EPA
(A-101)
401 M Street, SW
Washington DC
202-260-4126
John Brown
GE R&D Center
POBoxS
Schenectady NY 12301,
518-387-7987
John Cicmanec
US EPA, ORD
26 Martin Lather King Drive
Cincinnati OH 45268
513-569-7481
Theo Colborn
W. Alton Jones Foundation
1250 Twenty-Fourth Street, NW
Washington DC 20037
202-778-9643
Mike Bolger
US FDA
(HFF-156)
200 C Street, NW
Washington DC 20204
202-205-8705
Brian Bush
Wadsworth Laboratory
NY State Laboratory
Albany NY 12201
518-473-7582
Jim Cogliano
US EPA, ORD
401 M Street, SW (RD-689)
Washington DC 20460
202-260-2575
Janice Cox
Tennessee Valley Authority
Haney Building 2C
1101 Market Street
Chattanooga TN 37402
615-751-7337
C." v' I 1i'; •• i
A-8
-------
John Craddock
Regulatory Network, Inc.
1350 I Street, NW, Suite 200
Washington DC 20005
202-939-3444
Mitchell D. Erickson
Env. Chemistry Group Leader
Environmental Research Div
Argonne National Lab
9700 South Cass Ave.
Building 203
Argonne IL 60439
708-252-7772
John Hesse
Chief, Health Assessment Prog
MI Department of Public Health
PO Box 30195
Lansing MI 48909
517-335-8353
Peggy Krahn
Supervisory Research Chemist
NOAA-NMFS
2725 Montlake Boulevard East
Seattle WA 98112
206-553-1433
Dave Devault
Great Lakes Nat. Program Office
USEPA(G-9J)
77 West Jackson Boulevard
Chicago IL 60604
312-353-1375
Richard W. Greene
Environmental Engineer
Delaware Dept. of Natural Resources and
Environmental Control
89 Kings Highway
Dover DE 19903
302-739-4590
Rick Hoffmann
US EPA, OST
401 M Street, SW
Washington DC 20460
202-260-0642
Jennifer Orme Zavaleta
Drinking Water Health Assessment Section
Health and Ecological Criteria Section
US EPA, OST
•401 M Street, SW
Washington DC 20460
202-260-7586
Gerald Pollock
Cal-EPA OEHHA7PETS
601 North 7th Street
PO Box 94732
Sacramento CA 94234
916-327-7319
Ted Schwartz
US Fish & Wildlife Service
4200 New Haven Road
Columbia MO 65201
314-875-5399
Deborah L. Swackhamer
University of Minnesota
School of Public Health
PO Box 807 - UMHC
420 Delaware Street, SE
Minneapolis MN 55455
612-626-0435
Leon Sawyer
US FDA
(HFS-335)
200 C Street, NW
Washington DC 20204
202-205-4795
Betsy Southerland
US EPA, OST
401 M St., SW
Washington DC 20460
202-260-7157
A-9
-------
A.3 - Workshop Agenda
AGENDA; PCB WORKSHOP
PCBs in FISH TISSUES-
Exchanging Information Between Data Generators and Data Users
A forum to discuss current PCB issues regarding analytical methods for fish tissues and
considerations for human health assessments.
DAY 1 MONDAY MAY 10, 1993
PART I: INTRODUCTION To PCBs & FISH
1. Welcome & Introduction
9:00-9:20
{Dr. Elizabeth Southerland, EPA}
{Mr. Rick Hoffmann, EPA}
Topic: Welcome; Workshop purpose; EPA programmatic overview.
2. Introduction to PCBs and Analytical Methods
9:20-9:50
{Dr. Mitchell D. Erickson, Argonne National Lab}
Topic: Structure, chemistry, & analysis of PCBs.
3. Overview: Occurrence of PCBs in Fish Tissues
9:50-10:35
A. Temporal Trends of PCBs in the Environment
9:50-10:15
{Mr. John H. Craddock, Regulatory Network, Inc.}
B. PCB Trends in Great Lakes Fish
10:15-10:35
{Mr. David Devautt,
EPA's Great Lakes National Program Office}
A-10
-------
4. Overview of PCB toxicology
10:50-11:15
(Dr. Mike Bolger, Food and Drug Administration}
Topic: Introduction to PCB toxicology; FDA uses of PCB data
5. PCB Criteria for Water
11:15-11:40
(Ms. Jennifer Orme Zavaleta, EPA}
Topic: Overview of PCB regulatory criteria
11:40—12:00 Responses to Introductory Remarks
PART H: PCB TOXICITY & HEALTH EFFECTS
6. PCB Toxicity: Recent Evaluations of Human Health Effects
1:00-4:30
A. Regulatory Update: Human Carcinogenicity Effects
1:00-1:25
{Dr. Jim Cogtiano, EPA}
Topic: Current/Anticipated cancer value; recent information
B. Regulatory Update: Non-Carcinogenic Effects
1:25-1:50
(Dr. John Cicmanec, EPA}
Topics: proposed RfD for Aroclor 1016, other
C. Update: Toxicity Equivalents for PCBs
1:50-2:15
{Dr. Donald Barnes, EPA}
D. Animal/Human Health Connection
3:00-3:25
{Dr. Theo Colborn, W. Alton Jones Foundation}
E. Industry Research
2:35-3:00
{Dr. John Brown, GE R&D Center}
Topics: Longitudinal medical studies; PCB accumulation patterns; other
A-ll
-------
F. 3:30—4:30 Questions/ Discussions
DAY 2 TUESDAY MAY 11, 1993
PARTHI: ANALYTICAL METHODS
7. PCB Analyses—An Overview
8:00-8:15
(Dr. Mitch Erickson, Argonne National Lab)
Topic: Introduction of analytical methods panel
8. Methods Panel—Aroclor vs. Congener Methods vs. Other Methods
8:15-11:30
A. Recent PCB Research
8:15-8:40
(Mr. Ted Schwartz, U.S. Fish & Wildlife Service}
Topic: Further comparisons of aroclor versus congener-specific analyses in
fish; pattern recognition
B. FDA Method for Analyzing PCBs
8:40-9:05
{Mr. Leon Sawyer, U.S. Food & Drug Admin.)
Topic: FDA method for "total PCBs"— Aroclor-based; comparisons to
congener-specific analysis
C. "Performance-based" Methods
9:05-9:30
{Dr. Peggy Krahn, NMFS/NOAA}
Topic: Performance-based methods used by NMFS--total aroclor & congeners;
method advances
D. EPA's Green Bay PCB Study—Congener Analyses
9:30-9:55
{Dr. Deborah Swackhamer, University of Minnesota}
Topic: Measuring specific congeners—lessons from 7 lab. QA for EPA's Green
Bay PCB mass loading study
A-12
-------
E. State Lab Experience
10:15-10:35
{Dr. Brian Bush, NY State Wadsworth Lab}
Topic: Congener-specific analyses in New York
10:35—11:30 Methods Panel Discussion
PART IV: CASE STUDIES-Human Health/Risk Assessment
9. Risk Assessment Panel-
Case Studies of PCS Risk/ Health Assessments
1:00-3:30
A. California
1:00-1:25
{Dr. Gerald Pollock, Col EPA}
Topic: PCS risk assessment in Southern California
B. Tennessee Valley Authority
1:25-1:50
{Ms. Janice Cox, TVA}
Topic: Recent TVA fish risk assessment, PCB emphasis
C. Delaware
1:50-2:15
{Mr. Richard Greene, Delaware Water Resources}
Topic: Recent PCB assessments in Delaware
D. Michigan
2:15-2:35
(Mr. John Hesse, Michigan Dept. of Public Health}
Topic: Fish advisories for the Great Lakes: Past and Proposed Methodologies
2:35—3:45 Panel Discussion
A-13
-------
PART V: CONCLUDING REMARKS
3:45-4:15
(Dr. Elizabeth Southerland, EPA}
Topic: Summarize meeting; Follow-up activities
A.4 - Workshop Attendees
Darvene Adams
Environmental Scientist
Region 2
US - EPA
2890 Woodbridge Ave.
Edison NJ 08837
908-321-6700
John A. Arway
Chief
Div. of Environmental Services
PA Fish and Boat Commission
450 Robinson Lane
Bellefonte PA 16823
814-359-5147
Jay Auses
Technical Specialist
Alcoa Technical Center
100 Technical Drive
Alcoa Center PA 15069
417-337-2187
Mark Bevelhimer
Oak Ridge National laboratory
Environmental Division
P.O. Box 2008
Oak Ridge TN 37831
615-574-7977
Ann Alford-Stephens
US EPA
26 West Martin Luther King Junior Blvd.
Cincinnati OH 45268
513-569-7492
Sue Anne Assimon
CFSAN/FDA
Contaminant Branch
200 C Street, SW
Washington DC 20204
202-205-8705
David Bailey
Senior Attorney
Environmental Defense Fund
1875 Connecticut Ave., NW
Washington DC 20009
202-387-3500x11
Jeff Bigler
Standards and Applied Science Division
US EPA
OW, OST
401 M Street, SW
Washington DC 20460
202-260-1305
A-14
-------
Joan Blake
Biologist
OPPT
Operations Branch
US EPA
401 M Street, SW (TS-798)
Washington DC 20460
202-260-6236
Mark Briggs
MN Dept. of Natural Resources
5463 West Broadway
Forest Lake MN 55025
612-464-1247
Elizabeth Bush
Laboratory Scientist
VA Institute of Marine Science
P.O. Box 1346
Gloucester Point VA 23062
804-642-7236
Carrie Buswell
Env. Program Coordinator
Dyn Corp Viar, Inc.
300 North Lee Street
Suite 200
Alexandria VA 22314
703-519-1385
Helen Chernoff
Environmental Scientist
TAMS Consultants
655 Third Ave.
New York NY 10017
212-867-1777
Gregory Cramer
Chemist
US Food & Drug Administration
200 C St., SW
HFS-416
Washington DC 20204
202-254-3888
B. Gordon Blaylock
Senior Scientist
Environmental Sciences Division
Oak Ridge National Laboratory
P.O. Box 2008
Bldg. 1505, MS-6036
Oak Ridge TN 37831
615-574-7397
Kathy Brohawn
Natural Resource Biologist
Water Quality Program
Shellfish Certification Division
MD Department of the Environment
2500 Broening Highway, Room 1120
Baltimore MD 21224
410-631-3906
Hugo M. Bustos
Environmentalist
Panhandle Eastern Corporation
5444 Westhimer Court
WI766
Houston TX 77056
713-989-2322
Moses Chang
Water Management Division
US EPA, Region 2
26 Federal Plaza (Room 813)
New York NY 10278
212-264-1552
Emelise S. Cormier
Program Manager
Louisiana Department of
Environmental Quality
P.O. Box 82215
Baton Rouge LA 70884
504-765-0511
Roger Crawford
Director Env. Affairs
Outboard Marine Corp.
190 Seahorse Drive
Waukegan IL 60085
708-689-5219
A-15
-------
John Crellin
Environmental Health Scientist
Agency for Toxic Substances &
Disease Registry
1600 Clifton Road
Mail Stop E-32
Atlanta GA 30244
404-639-0635
James Curtis
Sanitarian
Water Quality Program
Biomonitoring Division
MD Dept. of the Environment
2500 Broening Hwy, Room 1120
Baltimore MD 21224
410-631-3906
Thomas Fazio
Special Assistant to Director
OSAS/CFSAN
Food & Drug Administration
200 C Street, S.W.
Washington DC 20204
202-205-5182
Jay Field
NOAA
7600 Sand Point Way
Seattle WA 98115
206-526-6404
Robert P. Fisher
Program Director
NCASI
P.O. Box 141020
Gainesville FL 32614
904-377-4708
Henry Folmar
Lab Director
MS Dept. of Env Quality
121 Fairmont Plaza
Pearl, MS 39208
601-961-5183
Philip Crocker
Aquatic Biologist
US EPA
6W-QT
1445 Ross Avenue
Dallas TX 75202
214-655-6644
Michael DiNovi
Chemist
Food and Drug Administration
HFS-247
200 C St., SW
Washington DC 20204
202-254-9537
Lynn C. Feldpausch
Environmental Scientist
EPAHQ
401 M St. SW
Washington DC 20460
202-260-8149
Thomas J. Fikslin
Director
Estuary Toxics Management Program
Delaware River Basin Commission
P. O. Box 7360
West Trenton NJ 08628
609-883-9500
Joseph E. Flotemersch
Graduate Research Assistant
Mississippi State University
Dept. of Wildlife & Fisheries
P.O. Drawer LW
Miss. State MS 39762
601-325-1835
Anthony Forti
Research Scientist JJ
N.Y. State Dept. of Health
2 University Place
Room 240
Albany NY 12203
518-458-6409
A-16
-------
Catherine Fox
Environmental Scientist
EPA-OST
WH-585
401 M Street, SW
Washington DC 20460
202-260-1327
Karen Gold
Environmental Health Scientist
Research Triangle Institute
P.O. Box 12194
RTF NC 27709
919-541-5840
Robert F. Frey
Water Pollution Biologist ,
PA Dept. of Environmental Resources
P.O. Box 8465
Tenth Floor
Harrisburg PA 17105
717-783-3638
John G. Haggard
Manager
Hudson River Projects Engineering
General Electric
1 Computer Drive South
Albany NY 12205
518-458-6619
Robert Hale
Assistant Professor
VA Institute of Marine Science
P.O. Box 1346
Gloucester Point VA 23062
804-642-7236
David Hohreiter
Associate
Blasland Bouck & Lee
6723 Towpath Road
Box 66
Syracuse NY 13214
315-446-9120
Cheng-Fei Huang
Fishery Attache
Taiwan Embassy
4301 Connecticut Ave., N.W.
Washington DC 20008
202-686-6400
Gerry loannides
Ohio EPA
1571 Perry Street
Columbus OH 43201
614-644-2782
Heraline E. Hicks
Senior Environmental Health Scientist
ATSDR
1600 Clifton Road, NE
Mail Stop E-29
Atlanta GA 30333
404-639-6306
Skip Houseknecht
Chief, Fish Contamination Section
Standards and Applied Science Division
US EPA
OW, OST
401 M Street, SW
Washington DC 20460
202-260-7055
Robert A. Hughes
Division Engineer
Panhandle Eastern Corporation
5444 Weshimer Court
WT770
Houston TX 77056
713-989-2331
John R. Jackson
Graduate Research Assistant
Mississippi State University
Dept. of Wildlife & Fisheries
P.O. Drawer LW
Miss. State MS 39762
601-325-1835
A-17
-------
B. Paul Jiapizian
Environmental Specialist
Water Quality Toxics Division
Water Quality Program
MD Dept of the Environment
2500 Broening Hwy, Room 1120
Baltimore MD 21224
410-631-3906
Billy Justus
Dioxin Coordinator
MS Dept. of Env. Quality
121 Fairmont Plaza
Pearl MS 39208
601-961-5183
Hamid Karimi
Chief
Water Qual. Monitoring Branch
Water Resources Mgmt Div
Env. Regulation Admin.
D.C. Government
2100 MLK Ave. SE
Suite 203
Washington DC 20020
202-404-1120
Susan Keith
Manager
ENVIRON Corporation
4350 North Fairfax Drive
Arlington VA 22203
703-516-2300
Roman Khidekel
Environmental Manager
Ohio EPA
1571 Perry Street
Columbus OH 43201
614-644-4247
Kathy A. Knowles
Environmental Chemist
Delaware DNREC
89 Kings Hwy
P.O. Box 1401
Dover DE 19903
302-739-4771
Brenda P. Julius
Chemist I
Environmental Management
State of Alabama
1890-A Dickinson Dr., Bldg. M.
Montgomery AL 36109
205-260-2770
Charles Kanetsky
Water Quality Monitoring Coordinator
US EPA Region HI (3ES11)
841 Chestnut Bldg.
Philadelphia PA 19107
215-597-8176
Russell E. Keenan
National Director
ChemRisk Division, McLaren/Hart
Environmental Engineering
1685 Congress Street
Portland ME 04102
207-774-0012
Jackie Key
Environmental Scientist
MS Dept. of Environmental Quality
121 Fairmont Plaza
Pearl MS 39208
601-961-5183
Carole A. Kimmel
Senior Developmental lexicologist
US EPA (RD-689)
401 M St., SW
Washington DC 20460
202-260-7331
Rosanna Kroll
Environmental Specialist
Water Quality Toxics Division
Water Quality Program
MD Department of the Environment
2500 Broening Highway, Room 1120
Baltimore MD 21224
410-631-3906
A-18
-------
Peter Lanahan
Manager
Hudson River Project
GE-CEP
1 Computer Drive South
Albany NY 12205
518-458-6648
Fred Leutner
US EPA, OST
401 M Street, SW
Washington DC 20460
202-260-1542
Robert LeHew
Scientist
Oak Ridge National Laboratory
Environmental Science Division
PO Box 2008
Oak Ridge TN 37831
615-576-8448
C. A Manen
Senior Chemist
Damage Assessment
NOAA/NOS
6001 Executive Boulevard
Room 323
Rockville MD 20852
301-713-3034
Randall O. Manning
Toxicologist
GA Dept. of Natural Resources
Floyd Towers East, Suite 1152
205 Butler Street
Atlanta GA 30334
404-656-4713
Roy Martin
Vice President
National Fisheries Institute
1525 Wilson Boulevard
Suite 500
Arlington VA 22209
703-524-8880
Malcom Meaburn
Physical Science Administrator
Charleston Laboratory
USDA
NMFS
PO Box 12607
Charleston SC 29422
803-762-1200
John Moore
President
IEHR
1101 Vermont Ave., NW
Suite 608
Washington DC 20005
202-289-8721
Gary Mappes
Manager
Product and Environmental Safety
Monsanto
800 North Lindbergh
St. Louis MO 63167
314-694-3344
Pamela A. Matthes, Chief
Damage Assessment and Restoration Branch
US Fish and Wildlife Service
Division of Environmental Contaminants
4401 North Fairfax Drive #330
Arlington VA 22203
703-358-2148 ;
Penny Miller
Chemist
The Bionetics Corporation
445 First Street
Arnold Air Force Base TN 37389
Jacqueline Moya
Environmental Engineer
ORD/OHEA/EAG
401 M Street SW
Washington DC 20460
202-260-2385
A-19
-------
Deirdre Murphy
Head
Water Quality Program
Water Quality Toxics Division
MD Dept of the Environment
2500 Broening Hwy, Room 1120
Baltimore MD 21224
410-631-3906
Dan Olson
US Fish and Wildlife Service
Division of Env. Contaminants
4401 N. Fairfax Drive
Room 330
Arlington VA 22203
703-358-2148
Adrien L. Partridge
Analytical Chemist
Tennessee Valley Authority
401 Chestnut Street
Suite 150
Chattanooga TN 37401
615-751-3713
Mark J. Peterson
Research Associate
Oak Ridge National Laboratory
Bldg. 1506, Env. Sciences Div.
Box 2008, O.R.N.L.
Oak Ridge TN 37831
615-576-3461
Jack Pingree
DE Dept. of Health
P.O. Box 637
Dover DE 19903
302-739-5410
ReshaM. Putzrath
Principal
StepS
1101 17th St., N.W.
Washington DC 20036
202-429-8761
Paul Noe
Ropes and Gray
1001 Pennsylvania Avenue, NW
Suite 12005
Washington DC 20004
202-626-3915
Amy Owen
Environmental Scientist
Inter-Tribal Fisheries &
Assessment Program
186 East Three Mile Road
Sault Ste. Marie MI 49783
906-632-0072
Ken Paxton
Administrative Assistant
Fisheries
Ohio Division of Wildlife
1840 Belcher Dr.
Fountain Sq., Bldg. G
Columbus OH 43224
614-265-6343
David E. Pierce
Marine Fisheries Biologist/Manager
Massachusetts Division of
Marine Fisheries
100 Cambridge St.
Boston MA 02202
617-727-3193x366
Richard J. Pruell
Research Chemist
US EPA
27 Tarzwell Drive
Narragansett RI 02882
401-782-3091
Gilberto Quintero
Manager
Environmental Chemistry
Tennessee Valley Authority
401 Chestnut Street
Suite 150
Chattanooga TN 37401
615-751-3705
A-20
-------
Mark Reimer
Attorney
Fort Howard Corp.
1919 S. Broadway St.
P.O. Box 19130
Green Bay WI 54307
414-435-8821 x2406
Connie G. Ritzert
PCB Program Manager
Alcoa
1501 Aloca Building
Pittsburgh PA 15219
417-553-3734
Jack Schwartz
Laboratory Supervisor
Massachusetts Division of
Marine Fisheries
92 Fort Avenue
Salem MA 01970
508-745-3107
Peter C. Sherertz, Ph.D.
Toxicologist
VA Dept. of Health
Bureau of Toxic Substances
P.O. Box 2448
Room 124
Richmond VA 23218
804-786-1763
Kathy Snider
National Fisheries Institute
1525 Wilson Blvd, Suite 500
Arlington VA 22209
703-524-8881
Stephen K. Sorenson
Regional Biologist
NAWQA Program
US Geological Survey
433 National Center
Reston VA 22091
703-648-5113
Jeffrey L. Riback
Senior Attorney
Consolidated Edison of New York, Inc.
4 Irving Place
Room 1820
New York NY 10003
212-460-6677
Debbie Rouse
DE Dept. of Health
P.O. Box 637
Dover DE 19903
302-739-5410
Gayle Shelton
Chemist
Bionetics
445 First Street
Arnold Air Force Base TN 37389
615-454-7340
Mohsin R. Siddique
Chief
Water Quality Control Branch
Water Resources Management Division
Environmental Regulation Administration
D.C. Government
2100 MLK Ave., SE. Suite 203
Washington DC 20020
202-404-1120
Peter Somani
Executive Director
Ohio Department of Health
246 North High Street
P.O. Box 118
Columbus OH 43266
614-466-2253
Karl Strause
Project Scientist
Blasland & Bouck Engineers
6723 Towpath Road
Box 66
Syracuse NY 13214
315-446-9120
A-21
-------
Daniel M. Thomas
President
Great Lakes Fishing Council
293 Berteau
Elmhurst IL 60126
708-941-1351
Douglas Tomchuk
Environmental Engineer
US EPA, Region 2
26 Federal Plaza
New York NY 10278
212-264-7500
Michael Unger
Assistant Research Scientist
VA Institute of Marine Science
P.O. Box 1346
Gloucester Point VA 23062
804-642-7236
Ronald Wicks
Sanitarian
Biomonitoring Division
Water Quality Program
MD Dept of the Environment
2500 Broening Hwy, Room 1120
Baltimore MD 21224
410-631-3906
Durwood H. Willis
Environmental Engineer
Dept of Environmental Quality
Commonwealth of Virginia
4900 Cox Road
Glen Allen VA 23060
804-527-5104
Brian Toal
Assistant Division Director
Division of Environmental Epidemiology &
Occupational Health, CT Health Department
150 Washington Street
Hartford CT 06106
203-566-8167 ;
Ram K. Tripathi
Toxicologist
VA Department of Health
1500 East Main Street, Room 124
Richmond VA 23219
804-786-1763
David Velinsky
Environmental Geochemist
Interstate Commission on the
Potomac River Basin
6110 Executive Boulevard, Suite 300
RockvilleMD 20852
301-984-1908
Lisa Williams
Toxicologist/Epidemiologist III
TX Department of Health
1100 West 49th Street
Austin TX 78756
512-458-7510
Travis Wagner
Environmental Scientist
SAIC
7600-A Leesburg Pike
Falls Church VA 22043
703-821-4625
A-22
-------
Appendix A.5 - PCB Workshop Report Summaries from EPA's Risk Assessment Forum
Neurotoxic Effects Summary
On September 14 and 15, 1992, EPA's Risk Assessment Forum sponsored a workshop on
the developmental neurotoxic effects of PCBs (57 FR 39200, August 28, 1992). The
meeting was held in Research Triangle Park, NC, and was chaired by Linda Birnbaum and
Carole Kimmel of EPA. Participants from academia, industry, and state and federal
government brought expertise from a wide range of disciplines to the discussion. Members
of the public and EPA scientific staff attended the workshop as observers.
The report collects workshop papers and discussion on principles and methods for
evaluating data from animal and human studies. The report also summarizes data from other
information discussed at the workshop for characterizing risk to human development, growth,
survival, and function following exposure to PCBs prenatally or to infants and children.
EPA complied several issues papers on various aspects of PCB toxicity and especially on
developmental neurotoxicity, as a framework for workshop discussion. These issue papers
were distributed to all invited participants, who then submitted pre-meeting comments.
As outlined in the issue papers, the purpose of this workshop was to arrive at a
general "sense of the meeting" regarding the current state of the science on neurotoxic effects
associated with prenatal and perinatal exposure to PCBs. EPA did not expect participants to
reach a common position on all of the issues before the group. Because PCBs are present in
air, water, and food, information developed at the workshop will assist EPA in evaluating the
effects of PCBs in these media and will serve as a basis for protecting public health from
PCBs occurring in these media.
Recent studies defined the issues selected for workshop analysis. Data from rodents
and monkeys have demonstrated that prenatal and perinatal PCB exposure results in
neurotoxicity in the offspring. Related effects have been reported in human studies. For
example, human poisonings (Yusho and Yucheng) have led to developmental delays and
impairment in neurobehavioral indices in offspring of exposed women. Also, relatively low
levels of exposure to PCBs in cohorts in Michigan and North Carolina have suggested
neurobehavioral deficits in infants and younger children. Thus, the public health
consequences of exposure to developmental neurotoxicants such as PCBs are potentially
significant.
These observations pose several questions regarding the use of PCB data for assessing
risk of neurotoxic effects because of prenatal or perinatal exposure:
• Are all PCBS alike in these effects or, if not, are any useful structure/activity
relationships discernable?
• What are the dose/response relationships?
A-23
-------
• Are there populations at special risk due to elevated exposure or to inherent
sensitivity?
• What are the endpoints of greatest concern and greatest sensitivity?
These questions are important because of the persistence of PCBs in environmental
media such as water and air, and the nature of the data available on PCBs and developmental
neurotoxicity; i.e., many studies are available on various mixtures but little or no information
is available regarding specific congener effects on the developing organism, or the
mechanism of action of PCBs.
Availability: The full report, Workshop Report on Developmental Neurotoxic Effects
Associated with Exposure to PCBs (EPA/630/R-92/004, May 1993), is available from EPA's
Center for Environmental Information in Cincinnati, Ohio at (513 569-7562.
Toxicity Equivalency Summary
On December 11 and 12, 1990, EPA's Risk Assessment Forum held a workshop on toxicity
equivalency factors for PCB congeners. The purpose of the workshop was to examine the
existing toxicity and exposure data base on PCBs to ascertain the feasibility of developing
toxicity equivalency factors (TEFs) for PCB congeners. Given the widespread acceptance
and acknowledged utility of the TEF method for assessing risks associated with exposures to
complex mixtures of chlorinated dibenzo-p-dioxins and dibenzofurans, some experts have
urged development of comparable TEF schemes for other structurally related chemicals, such
as PCBs. Information from the workshop will contribute to Risk Assessment Forum
recommendations on whether to pursue development of a TEF scheme for PCBs.
EPA's Risk Assessment Forum assembled approximately 30 experts in the field of
PCB toxicity and mechanisms of action, environmental exposure, and analytical methods for
measuring PCBs in human and environmental samples. Dr. Donald Barnes chaired the
workshop. After presentations by Dr. Barnes and Dr. Stephen Safe, the participants divided
into two work groups: the Work Group on Exposure/Analytical Issues, chaired by Ms. Ann
Alford-Stevens; and the Work Group on Toxicity/Mechanisms of Action Issues, chaired by
Dr. Linda Birnbaum. The groups discussed the following questions:
• Is the existing data base on toxicity and mechanisms of action sufficient to support a
TEF scheme for PCBs?
•1
• What is known about environmental exposures to specific PCB congeners?
• What analytical methods are available to identify and quantify individual congeners in
environmental matrices?
A-24
-------
• What are the important data gaps and what research is needed to fill them?
On the second day of the workshop, all participants reconvened and the Work Group
chairs led the discussion of each group's rinding and recommendations. Dr. Barnes closed
the meeting with a summary of the workshop's conclusions and recommendations, which are
contained in the full report.
Availability: The full report, Workshop Report on Toxicity Equivalency Factors for
Polychlorinated Biphenyl Congeners (EPA/625/3-91/020, NTIS Order No. PB 982-114529,
June 1991), is available from the National Technical Information Service at 1-800-553-6847.
A-25
-------
------- |