DRAFT
DO NOT QUOTE OR CITE
EPA/600/P-96/001A
January 1996
External Review Draft
ni
«te»n
L!BP,
-------
DRAFT-DO NOT QUOTE OR CITE
1 DISCLAIMER
2 This document is an external draft for review purposes only and does not
3 constitute Agency policy. Mention of trade names or commercial products does
4 not constitute endorsement or recommendation for use.
DRAFT-DO NOT QUOTE OR CITE ii V24/96
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS
2 List of Tables ................ ..v
3 Preface ........... vi
4 Authors and Contributors . . . viii
5 1. INTRODUCTION /,..... . 1
6 1.1. PCB MIXTURES ...\. 1
7 . 1.2. CANCER POTENTIAL OF PCB MIXTURES 4
8 1.3. APPROACH TAKEN BY THIS ASSESSMENT ................ 5
9 2. SUMMARY OF STUDIES USED IN THE DOSE-RESPONSE ASSESSMENT . . 7
10 2.1. LIFETIME CANCER STUDIES IN ANIMALS 7
11 2.2. PARTIAL LIFETIME STUDIES IN ANIMALS 10
12 2.3. TUMOR INITIATING AND PROMOTING ACTIVITY ... ... 12
13 2.4. PHARMACOKINETICS AND METABOLISM .... ....... 14
14 3. DOSE-RESPONSE ASSESSMENT . ........................ 16
15 3.1. DATA SELECTION . 16
16 3.2. EMPIRICAL MODELING OF TUMOR INCIDENCE :............ 18
17 3.3. BIOLOGICALLY BASED MODELING OF TUMOR PROMOTION ... 23
18 3.4. ANALYSES OF CONGENER TOXICITY . 23
19 4. APPLICATION OF THE DOSE-RESPONSE ASSESSMENT . . 26
20 4.1. APPLICATION TO PCB MIXTURES IN THE ENVIRONMENT ..... 26
21 4.2. APPLICATION TO DIFFERENT ROUTES OF ENVIRONMENTAL
22 EXPOSURE 29
23 4.3. APPLICATION TO PARTIAL LIFETIME EXPOSURE 30
24 4.4. APPLICATION WITH DIOXIN TOXIC EQUIVALENCE FACTORS . . 32
.,*"'' '
25 <5. CHARACTERIZATION AND GUIDANCE FOR RISK ASSESSORS . . 34
26 5.1. DOSE-RESPONSE CHARACTERIZATION 34
27 5.2. INFLUENCE OF CANCER GUIDELINE REVISIONS ...... ?% 38
28 5.3. RESEARCH NEEDS, RESEARCH IN PROGRESS, AND
29 .PRELIMINARY RESULTS . . ... ^ . . . 39
30 5.4. SUMMARY OF GUIDANCE FOR RISK ASSESSORS ....'.- 44
DRAFT-DO NOT QUOTE OR CITE Hi 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 5.5. EXAMPLE . . . . 46
2 6. REFERENCES . . . . 48
3 APPENDIX: DOSE-RESPONSE ASSESSMENT RESULTS 56
DRAFT--DO NOT QUOTE OR CITE
IV
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 LIST OF TABLES
2 1-1 Typicai composition (%) of some commercial PCB mixtures . ....;.... 2
3 ,. 2-1 Original and revised rat liver tumor incidences ....-.' 8
4 2-2 Mixtures and congeners tested for tumor promoting activity ..... . . . . 13
5 3-1 Potency and slope estimates for humans from .lifetime cancer studies in
6 animals . ,. 22
7 3-2 PCB congeners of highest concern . 25
8 , 3-3 WHO interim TEFs for human intake of dioxin-like PCBs 26
9 5-1 Preliminary liver tumor incidences from the GE study . 40
10 5-2 Preliminary potency and slope estimates for humans from the GE study .41
11 5-3 Comparison of slope estimates from published studies and new study . . 42
12 5-4 Factors to consider in assessing risks from PCBs in the environment . . . 45
13 5-5 Sample lifetime average daily dose calculations 47
14 5-6 Sample risk calculations ..'..' 48
DRAFT-DO NOT QUOTE OR CITE v 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 PREFACE
»
2 This report updates the cancer dose-response assessment for RGBs and
3 shows how information on toxicity, disposition, and environmental processes can
4 be considered together to evaluate health risks from PCB mixtures in the
*
5 environment. It is intentionally brief, in order to focus on the analysis and
6 interpretation of available information rather than to provide a detailed compilation
7 of study results. Further information on PCB toxicity has been compiled by the
8 Agency for Toxic Substances and Disease Registry (ATSDR, 1993, 1995), Safe
9 (1994), Silberhorn et al. (1990), and the U.S. Environmental Protection Agency
10 (1988a).
11 Although not covered by this report, PCBs also have significant ecological
12 and human health effects other than cancer, including"neurotoxicity, reproductive
13 and developmental toxicity, immune system suppression, liver damage, skin
14 irritation, and endocrine disruption. Toxic effects have been observed from acute
15 and chronic exposures to PCB mixtures with varying chlorine content.
16 This report is to be used to support risk-based decisions within the general
17 policy framework provided by applicable EPA statutes and does not alter such
18 policies. It does not imply that one kind of information or another is a prerequisite
19 for action. Not every risk assessment based on this dose-response assessment will
20 have the same scope or depth; the level of detail of an assessment is a matter of
21 management policy.
22 This report is being made available to the public and the Congress of the
23 United States, responding to the report of the House of Representatives
24 Appropriations Committee, which specifies:
DRAFT-DO NOT QUOTE OR CITE vi 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 By December 31, 1995, the Administrator shall submit to the
2 Congress, and make available to the public, a draft report providing an
3 assessment of the risk of each of the polychlorinated biphenyl (PCS)
4 mixtures that has been the subject of laboratory animal cancer
5 bioassays, and a proposed methodology for assigning cancer risk
6 numbers to mixtures of RGB's found in the environment. By
7 September 1, 1996, the Committee directs that EPA shall have
8 completed, by a panel of independent experts on the earcinogenicity
9 of PCB's, a peer review of the draft report, and shall submit a final
10 report to the Congress and make it available to the public.
11 Although significant new cancer studies are nearing completion and new
12 information will soon become available, updating the dose-response assessment at
13 this time allows current decisions to reflect current science and provides a
14 framework for incorporating new information.
DRAFT-DO NOT QUOTE OR CITE vii 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 AUTHORS AND CONTRIBUTORS
2 This report was written by Dr. Jim Cogliano of EPA's National Center for
3 Environmental Assessment. The author gratefully acknowledges the helpful and
4 insightful comments from many scientists in EPA's program, regional, and research
5 organizations:
6 Air and Radiation:' Jane Caldwell.
7 Prevention, Pesticides and Toxic Substances: David Lai,
8 Elizabeth Margosches.
9 Solid Waste and Emergency Response: Dorothy Canter.
10 Water: Robert Cantilli, Michael Cox, Krishan Khanna, Edward Ohanian.
11 Region 1 (Boston): Mary Ballew, Andy Beliveau, Ann-Marie Burke,
12 Jui-Yu Hsieh, Ronnie Levin, Margaret McDonough, Marybeth Smuts.
13 Region 2 (New York): Marian Olsen, Anita Street, Doug Tomchuk. '
14 Region 3 (Philadelphia): Debra Forman.
15 Region 5 (Chicago): Carole Braverman, Milton Clark, Stephen Johnson.
16 Region 6 (Dallas): Young-Moo Kim, Maria Martinez, Jeffrey Yurk.
17 Region 7 (Kansas City): Dave Crawford.
18 Region 8 (Denver): Bob Benson, Suzanne Wuerthele.
19 Region 9 (San Francisco): Arnold Den, Daniel Stralka.
20 Region 10 (Seattle): Dana Davoli, Roseanne Lorenzana.
21 Research and Development: Linda Birnbaum, Ruth Bleyler, William Farland,
22 Charli Hiremath, James Holder, Prasad Kodavanti, Aparna Koppikar,
23 Jim Lake, Susan Norton, Charles Ris, Cheryl Siege! Scott,
24 Dharm Singh, Jeanette Wiltse.
25 The author also gratefully acknowledges Dr. Stephen Hamilton of the General
26 Electric Company for providing preliminary results of new cancer studies nearing
27 completion.
DRAFT-DO NOT QUOTE OR CITE viii 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 1. INTRODUCTION
2 1.1. PCB MIXTURES
3 PCBs (polychlorinated biphenyls) are chemical mixtures of variable
4 composition.1 Domestic manufacture of commercial mixtures stopped in 1977,
5 while existing PCBs continue in use. Although their properties vary widely,
6 different commercial mixtures can have many common components. Table 1-1
7 shows the overlapping composition of some commercial mixtures in terms of
8 congeners with 1 to 10 chlorines.
9 In the environment, PCBs also occur as mixtures of congeners, but their
10 composition differs from the commercial mixtures. This is because after release
11 into the environment, the composition of PCB mixtures changes over time, through
12 partitioning, chemical transformation, and preferential bioaccumulation.,
13 Partitioning refers to processes by which different fractions of a mixture
14 separate into air, water, sediment, and soil. PCBs adsorb to organic materials,
15 sediments, and soils; adsorption tends to increase with chlorine content of the
16 PCBs and organic content of the other material (Callahan et al., 1979). PCBs can
17 volatilize or disperse as aerosols, providing an effective means of transport in the
18 Some notes on chemical structure and nomenclature: Each PCB molecule consists of two 6-carbon
1 9 rings, with a chemical bond joining one carbon from each ring (imagine sunglasses with hexagonal
20 ; frames). Chlorine can attach to any of the other 10 carbons. There are 209 possible arrangements,
21 called congeners; congeners with the same number of chlorines .are called isomers. The number and
22 position of chlorines determine a molecule's physical and chemical properties. The 10 positions are
23 numbered 2-6 on one ring and 2'-6' on the other. For example, the congener 2,4,2',5'-
24 tetrachlbrobiphenyl has chlorines in positions 2 and 4 of one ring and 2 and 5 of the other. Positions 2,
25 6, 2', and 6', adjacent to the bond, are called ortho positions; 3, 5, 3','and 5', meta positions; 4 and
26 4', the outermost or para positions. The International Union of Pure and Applied Chemists (IUPAC) has
27 adopted an alternative system for numbering congeners sequentially from 1 to 209; nurfe^ers assigned
28 to congeners named in this assessment are listed in table 3-2. A molecule's two rings can twist on the
29 bond joining them; they are coplanar if aligned in the same-plane. Chlorine in ortho positions inhibits a
30 coplanar alignment. Coplanar molecules have dipxin-like properties (Safe, 1990, 1994; U.S. EPA,
31 1994d). .
DRAFT-DO NOT QUOTE OR CITE 1 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
Table 1-1. Typical composition (%) of some commercial PCB mixtures
Arodor Clophen Kanechior
10161242124812541260 A 30 A 60 300 400 500
Mono-CBs 21 _ ___
Di-CBs
Tri-CBs
Tetra-CBs
Penta-CBs
Hexa-CBs
Hepta-CBs
Octa-CBs
Nona-CBs
Deca-CBs
Notes:
Columns may
19
57
22
not
13
45
31
10
total
1
21
49
27
2
100%
1
15
53
26
4
due to
12
42
38
7
1
rounding; '
20
52
22
3
1
_
_
1
16
51
28
4
_
17
60
23
1
_
_
_
3
33
44
16
5
__
_
-
I
5
26
55
13
_-
«.
1 -" signifies less than 1 %.
Lot-to-lot variability exists but has not been quantified.
Impurities include
1993).
chlorinated.
dibenzofurans and
Sources: Adapted from Silberhorn et
chlorinated
naphthalenes
(WHO,
al. (1990), ATSDR (1995).
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20 environment (Callahan et al., 1979). Although congeners with low chlorine
21 content tend to be more volatile, actual volatilization in the environment is difficult
22 to predict, as volatilization competes with water solubility, and the more volatile
23 congeners also tend to be more soluble (Callahan et al., 1979). Vaporization rates
24 and water solubility of Aroclors and congeners vary over several orders of
25 magnitude (Hutzinger et al., 1974; Erickson, 1986).
26 Biodegradation transforms the chemical composition of PCB mixtures in the
27 environment. Anaerobic bacteria in sediments selectively remove chlorines from
28 meta and para positions, thus appearing to reduce the toxicity and bioaccumuiation
29 potential of residues; the occurrence and extent of these dechlorinations can be
30 limited by the concentrations of PCBs in the sediment (Abramowicz, 1990; Brown
31 and Wagner, 1990; Lake et al., 1992). Aerobic bacteria remove chlonrles from
32 PCBs with low chlorine content (1-4 chlorines^and break open the carbon rings
DRAFT-DO NOT QUOTE OR CITE 2 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 through oxidation (Abramowicz, 1990). PCBs with higher chlorine content are
2 extremely resistant to oxidation and hydrolysis (Callahan et al., 1979). Photolysis
3 can slowly break down congeners with high chlorine content (Callahan et al.,
4 1979). Overall, rates are slow and altered PCB mixtures persist in the environment
5 for many years.
6 PCBs selectively bioaccumulate in living organisms. PCBs are highly soluble
7 , in lipids and are absorbed by fish and other animals. Elimination from the body is
8 slow and varies by congener (Matthews and Anderson, 1975). Bioaccumulation
9 increases selectively with chlorine content and continues through the food chain
10 (Oliver and Niimi, 1988). In general, bioaccumulated PCBs appear to be more toxic
11 than commercial PCBs on a weight-to-weight basis (Aulerich et al., 1986;
12 Hornshawet al., 1983).
13 PCBs are widespread in the environment, and humans are exposed through
14 multiple pathways. Levels in air, water, sediment, soil, and foods vary over several
15 orders of magnitude, often depending on proximity to a source of release into the
16 environment (ATSDR, 1993; WHO, 1993). Average daily intake via ambient air is
17 about 100 ng, and about an order of magnitude higher if indoor concentrations are
18 considered (ATSDR, 1993). Average daily intake via drinking water is less than
19 200 ng (ATSDR, 1993). Estimates of average daily intake via diet vary widely
20 depending on geographic area, food habits, and sampling methodology; 5-15 ug is
21 considered a good estimate of average daily intake via diet in industrialized
22 countries (WHO, 1993). For nursing infants, average daily intake was estimated at
23 1.5-27 ug/kg (ATSDR, 1993); another study estimated 3-11 ug/kg (WHO, 1993).
f ' ~ - ... i
24 Using the narrower range, average daily intake for a 5-kg nursing infant would be
25 15-55 ug, about triple the average adult intake, and approximately 50-fold higher
26 when adjusted for body weight. Nursing infants are, therefore, an important highly
27 exposed population. Another is high-end consumers (above the 90th ^/ercentile) of
DRAFT-DO NOT QUOTE OR CITE 3 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 contaminated game fish, game animals, or products of animals (such as seabird
2 eggs) high on the food chain.
3 1.2. CANCER POTENTIAL OF PCB MIXTURES
4 Occupational studies show some increases in cancer mortality in workers
5 exposed to PCBs. Bertazzi et al. (1987) found significant excesses of cancer
6 mortality at all sites and in the gastrointestinal tract in workers exposed to PCBs
7 containing 54 and 42 percent chlorine. Brown (1987) found a significant excess
8 of liver, gall bladder, and biliary tract cancer mortality in capacitor manufacturing
9 workers exposed'to Aroclors 1254, 1242, and 1016. Sinks et al. (1992) found a
10 significant excess of malignant melanoma mortality in workers exposed to Aroclors
11 1242 and 1016. Several other studies, however, found no increases in cancer
12 attributable to PCB exposure (ATSDR, 1993). The lack of consistency overall
13 limits the ability to draw definitive conclusions from these studies. Incidents in
14 Japan and Taiwan where humans consumed rice oil contaminated with PCBs
15 showed some excesses of liver cancer, but this has been attributed, at least in
16 part, to heating of the PCBs and rice oil, causing formation of chlorinated
17 dibenzofurans (which have the same mode of action as some PCB congeners)
18 (ATSDR, 1993; Safe, 1994). Overall, the human studies have been considered to
19 provide limited (IARC, 1987) to inadequate (U.S. EPA, 1988a) evidence of
20 carcinogenicity.
21 Laboratory animal studies found high, statistically significant incidences of
22 liver tumors"in rats ingesting Aroclor 1260 or Clophen A 60 (Kimbrough et al.,
23 1975; Norback and Weitman, 1985; Schaeffer et al., 1984). Statistically
24 significant increases in gastric cancer and leukemia and lymphoma were found in
25, rats ingesting Aroclor 1254 (NCI, 1978). Nonsignificant increases in liver cancer
26 appeared in rats ingesting Aroclor 1254 (NCI, 1978) or Clophen A 30 fSchaeffer
27 et al., 1984). Partial lifetime studies found precancerous liver lesions in rats and
* *
DRAFT-DO NOT QUOTE OR CITE 4 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 mice ingesting PCB mixtures of high or low chlorine content: Overall, the animal
2 studies have been considered to provide sufficient evidence of carcinogenicity
3 (IARC, 1987; U.S. EPA, 1988a).
4 Several mixtures and congeners test positive for tumor promotion (Silberhorn
5 et al,, 1990). Toxicity of some PCB congeners is correlated with induction of
6 mixed-function oxidases; some congeners are phenobarbitai-type inducers, some
7 are 3-methylcholanthrene-type inducers, and some have mixed inducing properties
8 (McFarland and Clarke, 1989). The latter two groups most resemble 2,3,7,8-
9 tetrachlorodibenzo-p-dioxin in structure and toxicity.
10 Based on these findings, some commercial PCB mixtures have been
11 characterized as probably carcinogenic to humans (IARC, 1987; U.S. EPA, 1988a).
12 There is less controversy about this conclusion than about how it applies to PCB
13 mixtures found in the environment.
14 1.3. APPROACH TAKEN BY THIS ASSESSMENT
15 Previous assessments developed a single dose-response slope (7.7 per
16 mg/kg-d average lifetime exposure) for evaluating PCB cancer risks (U.S. EPA,
17 1988a). With no agreed-upon basis for reflecting differences among environmental
18 mixtures, this slope was used by default for any mixture. Different alternatives
19 have been suggested that would make some distinctions about cancer risks from
20 different PCB mixtures. One alternative would assume there is no cancer hazard
21 from environmental mixtures with less than 60 percent chlorine content (Delaware
22 Department of Natural Resources and Environmental Control, 1994). Another
' "V ' '
23 alternative would develop a separate assessment for each commercial mixture that
24 has been studied. These alternatives begin to distinguish among PCB mixtures, but
25 they do not address.how the environmental processes of partitioning,
- -, " ' - ~3t-~
26 transformation, and bioaccumulation diminish the similarity of environmental
27 mixtures to any of the commercial mixtures. ^
DRAFT-DO NOT QUOTE OR CITE 5 1/24/96
-------
DRAFT-DO NOT QUOTE OR GITE
1 This assessment adopts a related approach that distinguishes among PCB
2 mixtures by using information on environmental processes. Environmental
3 processes have profound effects that can decrease or increase toxicity, so toxicity
4 of an environmental mixture is only partly determined by the original commercial
5 mixture. This assessment, therefore, considers all cancer studies (which used
6 commercial mixtures only) to develop a range of dose-response slopes, then uses
7 information on environmental processes to provide guidance on choosing an
8 appropriate slope for representative classes of environmental mixtures and different
9 exposure pathways. -
10 Diverse kinds of information, many not typically considered in dose-response
11 assessments, are used in this approach. Other innovative features include:
12 A range of potency estimates for PCB mixtures, with guidance on
13 using exposure circumstances to choose estimates from this range.
14 Both upper-bound and central slope estimates, with guidance on when
15 each is appropriate.
16 Guidance on adjusting exposure duration to include internal exposure,
17 reflecting persistence of PCBs in the body.
18 Discussion of biologically based modeling results of tumor promotion
19 and cell dynamics.
20 Application of some principles from EPA's cancer guideline revisions
21 (U.S. EPA, 1994a, 1994b), including the interagency consensus
22 _ cross-species scaling factor (U.S. EPA, 1992b).
23 Section 2 briefly recounts the studies used in developing the dose-response
24 assessment and applying it to environmental mixtures. For a comprehensive
25 discussion of PCB toxicity, including other studies, see ATSDR (1993), Safe
26 (1994), Silberhorn et al. (1990), or U.S. EPA (1988a). Section 3 uses these
*
27 studies to develop a new dose-response assessment. Section 4 discusses
28 application of the dose-response assessment to" environmental mixtures, to
DRAFT-DO NOT QUOTE OR CITE 6 1 /24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 different exposure routes, to partial lifetime exposure, and in combination with
2 dioxin toxic equivalence factors. Section 5 characterizes the results of this
3 assessment, describes research needs and preliminary research results, and gives
4 specific guidance for risk assessors.
5 2. SUMMARY OF STUDIES USED IN THE DOSE-RESPONSE ASSESSMENT
6 2.1. LIFETIME GANGER STUDIES IN ANIMALS
7 Published lifetime cancer studies used a limited number of commercial
8 mixtures and dose levels. A new study of several commercial mixtures over a
9 wider range of dose levels is nearing completion; preliminary results are described
10 in section 5.
11 Kimbrouah at al. (19751.1 Groups of 200 female Sherman rats were fed diets
12 with 0 or 100 ppm Aroclor 1260 for about 21 months. Six weeks later the rats
13 were killed and their tissues were examined. Hepatocellular carcinomas and
14 neoplastic nodules were significantly increased in rats fed Aroclor 1260, Tumor
15 incidences are given in table 2-1.
16 National Cancer Institute (NCI. 1978). Groups of 24 male or female
17 Fischer 344 rats were fed diets with 6, 25, 50, or 100 ppm Aroclor 1254 for
18 104-105 weeks (24 months). Then the rats were killed and their tissues were
19 examined. The combined incidence of leukemia and lymphoma in males was
20 significantly increased by the Cochran-Armitage trend test; however, since Fisher
21 exact tests were not also significant, NCI did not consider this result clearly related
22 to Aroclor 1254. Hepatocellular adenomas and carcinomas appeared increased;
23 the result was not statistically significant but was considered related to
24 Aroclor 1254. Liver tumor incidences are given in table 2-1. ^ ' -.'
DRAFT-DO NOT QUOTE OR CITE 7 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
2
3
4
5
6
7
8
9
10
11
12
13
14
15
T6
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Table 2-1 . Original an
Study
Klmbrough, 1260,
females
NCI, 1254, males
NCI, 1254, females
Schaeffer, A 30, males
Schaeffer, A 60. males
Norback, 1260, males
Norback, 1260,
females
d revised rat liver tumor incidences
Original incidence
Carcinomas Carcinomas
Dose onlv or nodules
Control 1/173
100 ppm 26/184
Control 0/24
25 ppm 0/24
50 ppm 1/24
100 ppm 2/24
Control 0/23
25 ppm . 0/24
50 ppm 0/22
100 ppm 0/24
Control 1/131
100 ppm 4/130
Control 1/131
100 ppm 61/129
Control 0/32
100/50/0 ppm 2/46
Control 0/49
100/50/0 ppm 43/47
1/173
170/184
0/24
0/24
1/24
3/24'
0/23
0/24
1/22
2/24
6/131
42/130
6/131
123/129
0/32
7/46
1/49
45/47
Moore (1994) incidence
Carcinomas Carcinomas
only or adenomas
'" 1/187
21/189
0/24
0/24
0/24
2/23
0/23
0/24
0/24
0/24
2/1 20
2/1 28
2/120
67/125
0/31
1/40
0/45
19/46
1/187
138/189
0/24
1/24
1/24
3/23
0/23
1/24
2/24
1/24
8/120
16/128
8/1 20
114/125
0/31
5/40.
1/45
41/46
Note:
Decreases between original and revised denominators are due to lost slides;
.increases, to slides that were censored originally but could not be specifically
identified for exclusion in the reevaluation.
Source: Adapted from Moore et al. (1994).
Morgan et al. (1981) and Ward (1985) reevaluated gastric lesions from this
study and found 6 adenocarcinomas in 144 exposed rats. This result is
statistically significant, as gastric adenocarcinomas had occurred in on!^1 of 3548
control male and female Fischer 344 rats in the. NCI testing program. Intestinal
DRAFT-DO NOT QUOTE OR CITE
8
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 metaplasia in exposed rats differed morphologically from controls, suggesting
2 Aroclor 1254 can act as a tumor initiator.
3 Schaeffer at ai. (1984). Male weanling Wistar rats were fed a standard diet
4 for 8 weeks, then were divided into three groups. One group was fed the basic
5 diet; for the other groups 100 ppm Clophen A 30 or A 60 Was added. Rats were
6 killed at 801 -832 days (26.3-27.3 months) and were examined for lesions in the
7 liver and some other tissues. For both mixtures, preneoplastic liver lesions were
8 observed after 500 days (16.4 months) and hepatocellular carcinomas after
9 700 days (23 months) in rats dying before the end of the study. The investigators
10 concluded, "Clophen A 60 had a definite, and Clophen A 30 a weak, carcinogenic
11 effect on rat liver." Tumor incidences are given in table 2-1.
12 Norback and Weltman (1985). Groups of male and female Sprague-Dawley
13 rats were fed diets with 0 or 100 ppm Aroclor 1260 for 16 months; the latter dose
14 was reduced to 50 ppm for 8 more months. After 5 additional months on the
15 control diet, the rats were killed and their livers were examined. Partial
16 hepatectomy was performed on some rats at 1, 3, 6, 9, 12, 15, 18, and 24
17 months to evaluate sequential morphologic changes. In males and females fed
18 Aroclor 1260, liver foci appeared at 3 months, area lesions at 6 months, neoplastic
19 nodules at 12 months, trabecular carcinomas at 15 months, and adenocarcinomas
20 at 24 months, demonstrating progression of liver lesions to carcinomas. By
21 29 months, 91 percent of females had liver carcinomas and 95 percent had
22 carcinomas or neoplastic nodules. Incidences in males were smaller, 4 and
23 15 percent,'respectively. Tumor incidences are given in table 2-1.
24 Vater et al.~(1995) obtained individual animal results to determine whether
25 the partial hepatectomies, which exert a strong proliferative effect on the
26 remaining tissue, affected the incidence of liver tumors. They reported that the
27 hepatectomies did not increase the tumor incidence. Among females r«id
28 Aroclor 1260, liver tumors developed in 4 of 7;with hepatectomies and 37 of 39
.' ' ''".""
DRAFT-DO NOT QUOTE OR CITE 9 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 without hepatectomies; no liver tumors developed in controls or males with
2 hepatectomies.
3 Moore at al. (1994): Institute for Evaluating Health Risks (1991). The
4 preceding rat liver findings were reevaluated using criteria and nomenclature that
5 had changed to reflect new understanding of mechanisms of toxicity and
6 carcinogenesis. The reevaiuation found somewhat fewer tumors than did the
7 original investigators. The apparent increase for Clophen A 30 (Schaeffer et al.,
8 1984) is no longer statistically significant. Original and revised rat liver tumor
9 incidences are given in table 2-1.
10 2.2. PARTIAL LIFETIME STUDIES IN ANIMALS
11 Although lifetime studies are preferred for dose-response modeling, partial
12 lifetime studies often use experimental designs addressing specific issues in the
13 application of a dose-response assessment. Partial lifetime studies for PCBs have
14 compared different commercial mixtures and the relative sensitivity of the sexes.
15 Some studies examined early-life exposure, which is not covered by most lifetime
16 cancer studies, where exposure starts at age 2-3 months, when the animals are
17 mature.
18 Kimbrouqh et al. (1972). Groups of 10 male or female Sherman rats were
19 fed diets with 0, 20, 100, 500, or 1000 ppm Aroclor 1254 or 1260, beginning at
20 3-4 weeks of age and continuing for 8 months. Incidences of adenofibrosis
21 reached 2/10 in males and 4/7 in females fed 1000 ppm Aroclor 1260; in contrast,
22 for 100 and 500 ppm Aroclor 1254, incidences were 1/10 and 10/10 in males and
23 7/10 and 9/9 in females. With regard to differences between sexes, the
24 investigators concluded Aroclor 1260 is more toxic to female rats than males, but
25 such a difference could not be established for Aroclor 1254. With regard to
26 differences between mixtures, the investigators concluded the effect san the liver
DRAFT-DO NOT QUOTE OR CITE 10 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 "is more pronounced with Aroclor 1254 when all morphologic changes of
2 equivalent dietary levels of Aroclor 1254 and 1260 are compared."
3 Although adenofibromas are not carcinomas, these lesions, particularly in
4 short studies, are sometimes regarded as indicating a potential for tumor formation
5 over a longer duration. After 23 months, for example, most female rats of this
«*
6 strain fed 100 ppm Aroclor 1260 developed hepatocellular carcinomas or
7 neoplastic nodules (Kimbrough et al., 1975).
8 Kimbrouah and Under (1974). Groups of 5Q male BALB/cJ mice'wara f*H
9 diets with 300 ppm Aroclor 1254 for 11 months, or for 6 months followed by
10 5 months without exposure. Hepatomas were found in 9 of 22 surviving mice
11 exposed 11 months, in 1 of 24 exposed 6 months, and in none of 58 controls.
12 Adenofibrpsis was observed in all mice exposed 11 months, but in none of the
13 others.
14 Kimura and Baba (1973). Groups of 10 male or female Donryu rats were fed
15 diets that increased from 38 to 462 ppm (time-weighted average, 330 ppm)
16 Kanechlor 400, beginning at 10 weeks of age and continuing for up to 400 days
17 (13 months). Multiple adenomatous-liver nodules were found in the six females
18 exposed for the longest durations. No nodules were found in males or in five
19 controls of each sex.
20 Ito et al. (1973). Groups of 12 male dd mice were fed diets with 100, 250,
21 or 500 ppm Kanechlor 300, 400, or 500, beginning at 8 weeks of age and
22 continuing for 32 weeks (7.5 months). Among mice fed 500 ppm Kanechlor 500,
23 five had hep'atocetlular carcinomas and seven had nodular hyperplasia. No other
24 , groups, including six controls, showed these effects. .
25 ito et al. (1974). Male Wistar rats were fed diets with 0, 100, 500, or
26 1000 ppm Kanechlor 300, 400, or 500, beginning at 8 weeks of age and
27 continuing for 28-5,2 weeks (6.5-12 months). Nodular hyperplasia was seen with
28 all three mixtures, highest for Kanechlor 500 and lowest for Kanechlor 300, but
DRAFT-DO NOT QUOTE OR CITE 11 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 not in controls. Histologically, the nodular hyperplasia was similar to that induced
2 by other chemical carcinogens, suggesting the nodular hyperplasia is preneoplastic.
3 In contrast to their study in mice (Ito et al., 1973), where only Kanechlor 500
4 induced hepatocellular carcinomas after 32 weeks (7.5 months), the investigators
5 concluded, "Hepatocellular carcinomas could be induced by administration of
6 Kanechlor-500, -400, or -300 for a longer period."
7 Rao and Banerii (1988). Groups of 32 male Wistar rats were fed diets with
8 0, 50, or 100 ppm Aroclor 1260, beginning at 5 weeks of age and continuing for
9 120 days (4 months). Neoplastic nodules with adenofibrosis were found in 24 of
10 32 rats fed 50 ppm Aroclor 1260 and in 16 of 32 rats fed 100 ppm. None of 32
11 controls showed these changes. The investigators concluded Aroclor 1260
12 induces liver tumors when fed to young rats, for a short time.
13 2.3. TUMOR INITIATING AND PROMOTING ACTIVITY
* -/, ' ' ' ' i
14 Studies of tumor initiating and promoting activity are available for a few
15 commercial mixtures and congeners. The congener studies are beginning to
16 identify a subset of mixture components contributing to cancer induction. As
17 some of these congeners are present in environmental mixtures, these studies
18 provide information about the potential of environmental mixtures to cause cancer.
19 Several commercial PCB mixtures and congeners show tumor promoting
20 activity (Silberhorn et al., 1990). Aroclor 1254 and Kanechlors 400 and 500
21 promote liver tumors in initiation-promotion studies; Aroclor 1254 also promotes
22 lung tumors (Anderson et al., 1983, 1994; Beebe et al., 1992, 1993).
j* ' -
23 Aroclor 1254, Clophens A 30 and A 50, four tetrachlorobiphenyls, three
24 pentachlorobiphenyls, and one hexachlorobiphenyl showed promoting activity in
25 studies to identify alterations in adenosine triphosphatase (ATPase), gamma-
26 glutamyl transpeptidase (GGT), or placenta! glutathione S-transferase (PGST)
27 activity. One study found the interaction of 23,2',5'- and 3,4,3',4'-
DRAFT--DO NOT QUOTE OR CITE 12 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 tetrachlorobiphenyl to produce more alterations than either alone (Sargent et al.,
2 1991). One monochlorobiphenyl and one dichlorobiphenyl showed no promoting
3 activity. Lists-of mixtures and congeners tested for promoting activity (with either
4 positive or negative results) appear in table 2-2; references can be found in
5 Silberhorn et al. (1990) and later references cited in the table.
*
; . .''..', *
6 Table 2-2. Mixtures and congeners tested for tumor promoting activity
7 , Mixture Tumors Mixture or congener Altered foci
8 Aroclor 1254 Liver, lung Aroclor 1254 GGT +
9 Kanechlor 400 Liver Clophen A 30 Marker not reported
10 Kanechlor 500 Liver Clophen A 50 ATPase-, GGT +
11 4-MCB Negative
12 4,4'-DiCB Negative
13 2,4,2',4'-TeCB GGT +
14 2,4,2',5'-TeCB ATPase-, GOT +
15 2,5,2',5'-TeCB ATPase-, PGST +
16 3,4,3',4'-TeCB ATPase-, GGT + , POST+
17 2,3,4,3',4'-PeCB GGT + , PGST+
18 , 2,4,5,3',4'-PeCB ATPase-, GGT +
19 3,4,5,3',4'-PeCB GGT-I-, PGST +
20 2,4,5,2',4',5'-HxCB ATPase-, GOT+ , PGST +
21 Compiled from many studies; not all mixtures or congeners were tested in all systems.
22 Sources: Adapted from Silberhorn et al. (199O), Buchmann et al. (1991), Laib et al. (1991),
23 Sargent et al. (1991), Beebe et al. (1992, 1993), Hemming et al. (1993), Anderson et al.
24 (1994). ,
25 Although PCBs are not generally described as tumor initiators, in some
26 studies a small number of ATPase-deficient or GGT-positive foci were initiated by
27 treatment with Clophen A 50 alone .(Silberhorn et al., 1990). Weak initiating
28 activity was found with 2,4,2',5'-tetrachlorobiphenyl, which induced ATPase-
29 deficient, but not GGT-positive, foci (Rose et ah, 1985; Laib et al., 19%1).
30 Initiation potential had been suggested by the different intestinal metaplasia
31 morphology induced by Aroclor 1254 (Morgan et al., 1981; Ward, 1985). Many
DRAFT-DO NOT QUOTE OR CITE 13 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 other investigators, however, report negative results for tumor initiation by PCB
2 mixtures or congeners (Silberhorn et al., 1990).
3 The significance of these studies is apparent, as ail ortho-substituted
i
4 congeners in table 2-2 are abundant in commercial mixtures (Schulz et al., 1989)
5 and have been found in environmental samples (Lake et al., 1995; McFarland and
6 Clarke, 1989). Known for its bioaccumulation potential and abundance in
7 environmental samples, 2,4,5,2',4',5'-hexachlorobiphenyl has been found to
8 . comprise 21.5 and 12.0 percent, respectively, of PCB residues in human fat and
9 milk; 2,4,5,3',4'-pentachlorobiphenyl constitutes 5.4 and 6.5 percent,
10 respectively, of these residues (McFarland and Clarke, 1989). The coplanar
11 congeners 3,4,3',4'-tetrachlorobiphenyl and 3,4,5,3',4'-pentachlorobiphenyl have
12 lower abundance in commercial mixtures (Kannan et al., 1988) but have been
13 found in a variety of organisms; including humans (Safe, 1994).
14 2.4. PHARMACOKINETICS AND METABOLISM
15 Cancer studies of lifetime and partial lifetime PCB exposure have been by
16 ingestion only. Pharmacokinetic studies provide information about the potential for
17 absorption and a risk of cancer by other exposure routes. Other studies have
18 quantified the retention and persistence of PCBs in the body.
19 Humans absorb PCBs from ingestion, inhalation, and dermal exposure
20 (ATSDR, 1993). Once absorbed, PCBs enter the circulation and are transported to
21 all tissues. Initial distribution is to liver and muscle, which are highly perfused;
22 subsequently, PCBs, being highly lipophilic, accumulate in fat and skin (Matthews
23 and Anderson, 1975).
24 Inhalation can be a principal absorption route for occupational PCB exposure
25 (Wolff, 1985). In animals, an inhaled PCB aerosol was rapidly absorbed, although
'.?,-
26 rates could not be estimated (ATSDR, 1993).
.. fjQ* CXJ» ' °'4i ^,9""' .> "''!'' ' . -
356
DRAFT-DO NOT QUOTE OR CITE 14 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 . PCBs can cross human skin and increase the body burden. Dermal exposure
2 can contribute significantly to body burdens of workers (Wolff, 1985) and can be a
3 major route of environmentalexposure (ATSDR, 1993). In vivo dermal absorption
4 by rhesus monkeys exposed for 24 hours to soil with 44 ppm Aroclor 1242 or '
5 23 ppm Aroclor 1254 was 14 percent in each case (Wester et al., 1993). Earlier
6 studies found similar absorption rates for PCBs in mineral oil, trichlorobenzene, and
7 acetone (Wester et al., 1990). Subsequent washing did not remove all PCBs,
8 especially if time had elapsed after exposure (Wester et al., 1983). In vitro human
9 skin accumulation of Aroclor 1254 from water was 12 percent after a half hour
10 (Wester et al., 1987) and 44 percent after 24 hours (Wester et al., 1990),
11 suggesting absorption is rapid initially and continues at a slower rate with further
12 contact.
13 PCBs are eliminated through metabolism, which occurs primarily in the liver
14 (Matthews and Anderson, 1975). Metabolism rates are generally lower with high
15 chlorine content, but chlorine position is also important (Hutzinger et al., 1974;
16 Matthews and Anderson, 1975). Absence of chlorine at two adjacent positions
17 facilitates metabolism (Matthews and Anderson, 1975). Metabolism and
1.8 elimination can be quite slow; for example, the half-life of 2,4,5,2',4',5'-
19 hexachlorobiphenyl exceeds the lifespan of rats (Matthews and Anderson, 1975).
20 Persistent congeners can retain biological activity long after exposure stops;
21 residual liver enzyme induction was observed in mice 42 weeks after a single dose
22 of Aroclor 1254 (Anderson et al., 1991). The majority of the retained mixture
23 comprised 2,4,5,3'4'- and 2,3,4,3'>4'-:pentachlorobiphenyl and 2,4,5,2',4',5'- arid
24 2,3,4,2',4',5'-hexachlorobiphenyl (see tables 2-2, 3-2, and 3-3).
25 Analysis of 1977 and 1985 serum levels in 58 Indiana workers exposed to
26 PCBs yielded median half-lives of 2.6 years for Aroclor 1242 and 4.8 years for
27 Aroclor 1254 (Phillips et al., 1989). Among workers with lowest concentrations
v ''"_. *
28 (0-30 ppb), median half-lives were higher, 3."Tyears for Aroclor 1242 and
DRAFT-DO NOT QUOTE OR CITE 15 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 6.5 years for Aroclor 1254. In another study, from 1977 to 1984 serum levels in
2 five Indiana workers exposed to PCBs decreased 89-94 percent (median,
3 92 percent) for Aroclor 1242 and 14-53 percent (median, 16 percent) for
4 Aroclor 1260; among six others without current occupational exposure, decreases
5 were 23-71 percent (median, 39 percent) for total PCBs (Steele et al., 1986).
6 Analysis of serum levels in the exposed workers yields half-lives of 2 years for
7 Aroclor 1242 and 16 years for Aroclor 1260; in those without current occupational
8 exposure, a half-life of 8 years for total serum PCBs.2 In another study, from
9 1977 to 1985 mean serum levels from 111 Great Lakes fisheaters decreased only
10 slightly, from 20.5 to 19.0 ppb (Hovinga et al., 1992).
11 It is important to recognize that ascribing a half-life to a mixture is
12 problematic if half-lives of its components differ widely; moreover, half-life
13 estimates for a mixture can underestimate its long-term persistence.3
14 3. DOSE-RESPONSE ASSESSMENT
15 3.1. DATA SELECTION
16 EPA's mixture guidelines (U.S. EPA, 1986b) favor basing assessments on
17 the effects of the mixture of interest; the second choice is to use a sufficiently
18 similar mixture; next, to assess the components of the mixture. The guidelines
/ '
19 further advise,
20 2Serum concentration was modeled as an exponentially decreasing function of time: cf=c0exp(-6f),
21 where ct is concentration at time t. c0 is initial concentration, and b is the rate parameter, estimated by
22 linear regression of ln(CQ/cp on t. .
23 To illustrate, consider a mixture of two components in equal parts: one component has a half-life of
24 t year; the other, 100 years. If the mixture concentration is sampled after 10 years, the half-life of the
25 total mixture will appear to be approximately 10 years: virtually all the first component will be gone,
26 virtually none of the second, so about half the original mix*ure will remain. This half-life, however,
27 overestimates the slow rate of decrease in the more persistent mixture fraction that remains.
DRAFT-DO NOT QUOTE OR CITE 16 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Attention should also be given to the persistence of the mixture in the
2 ~- environment as well as to the variability of the mixture composition over
3 , time or from different sources of emissions, If the components of the
4 mixture are known to partition into different environmental compartments or
5 to degrade or transform at different rates in the environment, then those
6 factors must also be taken into account, or the confidence in and
7 applicability of the risk assessment is diminished.
. -,-=",. . , +
8 There are no cancer studies of PCS mixtures found in the environment.
9 Studies are available for some commercial mixtures. The similarity of an
10 environmental mixture to the commercial mixtures can be a matter of considerable
11 uncertainty, because environmental mixtures are altered by partitioning,
12 transformation, and bioaccumulation. Assessing components of a PCB mixture is
13 not now a viable alternative; toxieity testing has been done for only a few of the
14 209 congeners. Thus assessments of environmental mixtures must use
15 information on commercial mixtures.
16 Risk estimates can be derived from either human or animal studies; each has
17 strengths and limitations. Estimates based on human studies reflect an observed
18 association between human exposure and cancer; however, it is difficult to
19 reconstruct reliable estimates of past exposure and separate the effect of
20 confounding exposures to other carcinogens. Estimates based on animal studies
21 benefit from controlled exposures and absence of confounding factors; however,
22 there is uncertainty in extrapolating dose and response rates across species.
23 EPA's cancer guidelines (U.S. EPA, 1986a) favor basing dose-response
24 assessments on human studies. In the absence of adequate human information,
25 assessments use animal species responding most like humans If this cannot be
26 determined, assessments emphasize long-term animal studies showing the greatest
27 sensitivity, with due regard to biological relevance* exposure route, and statistical
28 considerations. This default approach is considered to be conservative, tending
;'-'. . ' . . -3f.
29 toward public health protection.
DRAFT-DO NOT QUOTE OR CITE 17 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 For PCBs, human studies included relatively few subjects or lacked
2 contemporaneous exposure estimates. Some studies reported air concentrations,
3 but because skin contact is a major route of occupational exposure, air
4 concentrations would be a poor measure of exposure (Bertazzi et al., 1987; Brown,
5 1987). Some studies reported blood levels, but for relatively few workers at the
6 end of exposure (Bertazzi et al., 1987; Brown, 1987; Taylor, 1988; Sinks et al.,
7 1992); reconstruction of past exposure is problematic because different mixtures
8 had been in use over the years, the distribution of exposure and absorption by
9 route and congener is unknown, and congener persistence in the body varies.
10 Because of their controlled exposures, animal studies will be used for dose-
11 response modeling.
12 3.2. EMPIRICAL MODELING OF TUMOR INCIDENCE
13 Dose-response analysis begins by considering the possibility of developing a
14 biologically based model, that is, a model whose mathematical structure reflects
15 the ascertained mode of action and whose parameters are measured in
16 experimental studies. Several analyses of PCB congener toxicity (see section 3.4)
17 have identified multiple modes of action for PCBs: an important subset of dioxin-
18 like congeners acts through binding to the aryl hydrocarbon receptor (Safe, 1990,
19 1994), some congeners promote tumors through nondioxin-like modes of action
20 (Safe, 1994; U.S. EPA, 1991), and other congeners of low chlorine content
21 produce mutagenic metabolites (U.S. EPA, 1991). A biologically based model has
22 been employed for two congeners, and relative rates of dioxin-like activity have
23 been estimated for some dioxin-like congeners; these are discussed in sections 3.3
24 and 3.4, respectively. Few congeners, however, have been tested in experimental
25 studies to measure the rate parameters that would be used in a biologically based
. ' ' o
26 model of PCB mixtures generally. Consequently, the information available at this
27 time is more suited to empirical modeling, where default models describing tumor
DRAFT-DO NOT QUOTE OR CITE 18 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 incidence as a function of dose are fitted to results of the available lifetime cancer
2 : studies.
3 Tne lifetime cancer studies, most limited to only one 100-ppm dose level,
4 provide no information about the shape of the dose-response curve. The partial
5 lifetime studies are mixed in this regard, suggesting either sublinearity in the
6 experimental range (Ito et al., 1973) or supralinearity (Rao and Banerji, 1988). As
7 a policy default, a flexible curve-allowing either linearity or nonlinearity-is fitted
8 in the experimental range, and extrapolation to lower doses uses a linear approach
9 if there is not sufficient information to support a sublinear model (U.S. EPA,
10 1986a, 1994a, 1994b). This policy rests, in part, on some general considerations.
11 Low-dose linear models are appropriate for extrapolation to lower doses when a
12 carcinogen acts in concert with other exposures and processes leading to a
13 background incidence of cancer (Crump.et al., 1976; Lutz, 1990). This applies to
1.4 PCBs, since environmental exposure, especially incremental exposure to dioxin-like
15 congeners (see sections 3.4 and 4.4), augments existing background exposures.
16 Further, even when the mode of action indicates a nonlinear dose-response curve
17 in homogeneous animal populations, the presence of genetic and lifestyle factors in
18 a heterogeneous human population tends to make the dose-response curve more
19 linear (Lutz, 1990). This is because genetic and lifestyle factors contribute to
20 widening the spread of human sensitivity, which tends to straighten the dose-
21 response curve over a wider range. Although these considerations provide a
22 reasonable argument for a model that is linear at low doses, the relation of the
23 low-dose slope to one from the experimental range is uncertain; this uncertainty
24 increases with th«- distance from the experimental range.
25 For each lifetime cancer study in animals, the multistage model (U.S. EPA,
26 1980; Howe et al., 1986) is fitted in the experimental range to data for each tumor
27 type and sex. The usual measure of dose used for modeling is cumulative dose
28 expressed as a lifetime daily average (U.S. EP&, 1986a, 1992a). Dosing was
DRAFT-DO NOT QUOTE OR CITE 19 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 decreased or discontinued during two studies; however, it is likely that tumor
2 development had already begun and internal exposure remained high with release
3 of PCBs stored in fat (Vater et al., 1995). Thus initial dose levels will be used
4 without averaging over the study duration; this reduces potency estimates by up to
5 one-third compared to the default of averaging dose over the study duration.
. '.
6 Administered doses are converted from ppm in the diet to mg/kg-d by assuming
7 rats consume food equal to 5 percent of body weight daily (U.S. EPA, 1980), then
8 scaling to humans using a factor based on the 3/4 power of body weight (U.S.
<
B EPA, 1992b).4
10 Response is taken as the incidence of hepatocellular carcinomas or
11 adenomas, as reevaluated by Moore et al. (1994). Including adenomas reflects
12 guidance of the National Toxicology Program (McConnell et al., 1986) and the
13 observed progression of hepatocellular adenomas to carcinomas (Norback and
14 Weltman, 1985). Because their trends are statistically significant, dose-response
15 curves are also fitted for gastric adenocarcinomas and male leukemia and
16 lymphoma in the NCI (1978) study.
17 Several measures of cancer potency are estimated. The ED 10 (estimated
18 dose associated with 10 percent increased incidence) has been used as a
19 benchmark for potency rankingcomparing other carcinogens or other severe toxic
20 effectsor as a starting point for low-dose extrapolation (Cogliano, 1986; U.S.
21 EPA, 1988b,1994c; National Research Council, 1993). The ED01 (estimated
22 dose associated with 1 percent increased incidence) is a similar benchmark that,
23 falling just below the experimental range, reflects any curvature observed in that
24 range. These measures are expressed as equivalent human doses.
.TOT;'*.
25 Equivalent human dose (mg/kg-d) - (ppm in diet) x 0.05° x (animal weight / 70 kg human .
26 weight)174.
DRAFT-DO NOT QUOTE OR CITE 20 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 For extrapolation to lower doses, an ED01 can be converted to a slope by
2 computing 0,01/ED01.5 Similarly, an upper-bound slope can be obtained by
3 dividing 0.01 by a lower bound for ED01. The usual upper-bound slope for low-
4 dose extrapolation comes from the linearized multistage procedure (U.S. EPA,
5 1980, 1986a); the ED01 method and the linearized multistage procedure give
6 similar upper bound slopes. Advantages of the ED01 method are that
7 extrapolation starts near the experimental range, is mostly independent of choice
8 of model, and is statistically stable without requiring use of upper bounds
9 (Cogliano, 1986; U.S. EPA, 1988b). (Note that slopes are inversely related to
10 ED01s and ED 10s; high potency is indicated by high slopes, but low ED01s and
11 ED10s.) For the NCI (1978) study, slopes for liver and gastric tumors and male
12 leukemia and lymphoma, which trends are statistically significant, are added to
13 reflect overall cancer risk, as recommended by the National Research Council
14 (1994). Potency and slope estimates are compiled in table 3-1; details of the
15 calculations appear in appendix tables A-1 through A-10.
16 5The slope is the change in response divided by the changa in dose. Relative to the origin (an increased
1 7 response of 0 at a dose of 0), the change in response is 0.01 at the dose EOO1; thus the slope is
18 0.01/ED01. :
DRAFT-DO NOT QUOTE OR CITE 21 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Table 3-1. Potency and slope estimates for humans from lifetime cancer studies in
2 animals
3 Bound on IMP
4 . Study ED10* ED01b Slope6 , Slope* Slope* See
5 (using Moore revaluation) (mg/kg-d) (mg/kg-d) (mg/kq-d)"1 (mg/ko-d)"1 (mg/ka-d)"1 tabie
6 Kimbrough, 1260, females 0.18 0.020 0.5 1.2 1.2 A-1
7 NCI, 1254, males 0.34f 0.051f 0.2fl 0.7° " O.?0 A-2,3,4
8 NCI, 1254, females 0.73f 0.070f 0.19 0.3« 0.39 A-4,6
9 Schaeffer, A 30, males 1.8 0.32 0.03 0.1 . 0.1 A-7
10 Schaeffer, A 60, mates 0.11 0.011 0.9 2.2 2.2 A-8
11 Norback, 1260, males 1.1 0.17 0.06 0.2 0.2 A-9
12 Norback, 126O, females 0.11 0.012 0.8 2.2 2.3 A-10
13 aEstimated dose associated with 10% increased incidence.
14 bEstimated dose associated with 1 % increased incidence.
15 cComputed as0.01/ED01. ;
1 6 ^Computed as 0.01 divided by a lower bound on ED01.
17 "95% upper bound slope from linearized multistage procedure (U.S. EPA, 1986a).
18 '"Harmonic sum" of ED 10s or EDO 1s for liver and gastric tumors and male leukemia and lymphoma;
19 for example, ED01con,bined=0.01/I0.01/EDO1 f, analogous to reciprocal of sum of slopes.
20 8Sum of slopes for liver and gastric tumors and male leukemia and lymphoma.
21 These estimates span ranges of about an order of magnitude each. Some
22 estimates come from, male rats, which appear to be less sensitive to PCB toxicity
23 (Norback and Weitman, 1985; Kimbrougrret al., 1972; Kimura and Baba, 1973).
24 The two data sets showing lowest potency may reflect the low sensitivity of male
25 rats; consequently, this assessment focuses on the remaining five data sets. The
26 remaining EDO 1s span a range of 0.1-0.7 mg/kg-d average lifetime exposure; the
27 ED01s, 0.01-0.07 mg/kg-d. The remaining central estimate slopes span a range
28 of 0.1-0.9 per mg/kg-d average lifetime exposure; their upper bounds, 0.3-2 per
29 mg/kg-d. r
30 These ranges reflect experimental uncertainty and variability of commercial
31 mixtures, but not human heterogeneity and differences between commercial and
32 environmental mixtures. Environmental processes have profound effects that can
33 increase or decrease toxicity, so toxicity of an environmental mixture is only partly
34 determined by the original commercial mixture.^ Potency estimates for an Aroclbr
DRAFT-DO NOT QUOTE OR CITE 22 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 may not be the best surrogate for assessing that Aroclor as altered in the
2 environment. Sections 4 and 5 develop specific guidance for applying these
3 potency ranges to environmental mixtures.
4 3.3. BIOLOGICALLY BASED MODELING OF TUMOR PROMOTION
5 Using a two-stage carcinogenesis model, Luebeck et ah. (1991) modeled
6 tumor promoting activity of 2,4,2',5'- and 3,4,3',4'-tetrachlorobiphenyl,- based on
7 the study of Buchmann et ah (1991). Female Wistar rats were initiated with 10
8 mg/kg-d diethylnitrosamine for 10 days, followed by eight weekly injections of 10
9 or 150 umol/kg of various compounds. The rats were killed 1 or 9 weeks later,
10 and preneoplastic activity was characterized by changes in ATPase and GGT
11 activity.
12 Because results are available for two times after dosing stopped, modeling
13 can assess persistence of promoting activity. There was little or no promoting
14 activity by 2,4,2',5'-tetrachlorobiphenyl after dosing stopped/ but 3,4,3',4'-
15 tetrachlorobiphenyl continued to promote vigorously (Luebeck et ah, 1991).
16 Modeling can also estimate the probability of altered foci becoming extinct;
17 that is, disappearing after dosing stops. After dosing stops, the probability of
18 extinction is high, as most altered foci do not develop into, observable tumors.
19 Those that become large, however, tend to persist, as the probability of extinction
20 decreases as size increases (Luebeck et ah, 1991).
21 3.4. ANALYSES^OF CONGENER TOXICITY
22 McFarland and Clarke (1989) discuss how toxicity of some PCB congeners is
23 correlated with induction of mixed-function oxidases. Some congeners are
24 characterized as phenobarbital-type inducers, a smaller number of congeners is
" - - - ' ' -.3." .
25 characterized as 3-methylcholanthrene-type inducers, and some congeners have
26 mixed inducing properties. The latter two groups most resemble 2,3,7,8-
''"."" /' " ' - "' ' F !
DRAFT-DO NOT QUOTE OR CITE 23 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 tetrachlorodibenzo-p-dioxin in structure and toxicity. Based on potential for
2 toxicity (some forms of toxicity, for example, neurotoxicity, may not be well
3 represented) and frequency of occurrence in environmental samples, 36 congeners
4 of highest concern were identified and classified. These are listed in table 3-2.
5 U.S. EPA (1991) examined toxic effects, including cancer, of four structural
6 classes: dioxin-like PCBs, ortho-substituted PCBs, hydroxylated metabolites, and
7 sulfonated metabolites. Dioxin-like PCBs were associated with cancer and
8 described as strong promoters; reference also was made to initiating properties of
9 some PCS congeners (Buchmann et al., 1991). Similarly, some ortho-substituted
10 PCBs were considered potent promoters. Hydroxylated metabolites of some
11 congeners with low chlorine content were described as having genotoxic or
12 carcinogenic potential. Dioxin-like and other PCBs would operate by different
13 mechanisms to induce cancer. It was concluded that congener toxicity cannot be
14 characterized by chlorine content alone. Before adopting toxic equivalence factors
15 (TEFs) for PCB congeners, it was recommended to define other classes of PCBs
16 and identify the mechanisms involved.
17 Safe (1990, 1994) characterized dioxin-like PCBs as eliciting a spectrum of
18 biochemical and toxic responses similar to chlorinated dibenzo-p-dioxins and
19 dibenzofurans, all acting through the aryl hydrocarbon receptor. Based on
20 quantitative structure-activity studies, the first conservative TEFs for dioxin-like
21 PCBs were proposed and refined. Use of these TEFs is limited to responses
22 mediated through the aryl hydrocarbon receptor. The Aroclor 1260 response,
23 however, appears attributable mostly to its nondioxin-like congeners, based on its
24 relatively low level of dioxin toxic equivalents (Safe, 1994). Aroclor 1260 has a
25 lower level of dioxin toxic equivalents than Aroclors 1254, 1248, and 1242 (Hong
26 etal., 1993; Safe, 1994).
DRAFT-DO NOT QUOTE OR CITE 24 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Table 3-2. PCB congeners of highest concern
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19
20
Highest toxicity
and abundance*
3-MC-type inducers:
77: 3,4,3',4'-TeCB
126: 3,4,5,3',4'-PeCB
169: 3,4,5,3',4',5'-HxCB
Mixed-type inducara:
105: 2,3,4,3',4'-PeCB
118: 2.4,5,3',4'-PeCB
128: 2,3,4,2',3',4'-HxCB
138: 2,3,4,2'.4',5'-HxC8
156: 2,3,4,5,3',4'-HxCB"
170: 2,3,4,5,2',3',4'-HpCB
High toxicity
and abundance1*
PB-type inducers:
87: 2,3,4,2',5'-PeCB
99: 2,4,5,2',4'-PeCB
101: 2,4,5,2',5'-PeCB
153:2.4,5,2'.4',5'-HxCB
180: 2,3,4,5,2',4',5'-HpCB
183: 2,3,4,6,2'.4',5'-HpCB
194: 2,3,4,5,2',3',4',5'-OCB
Abundant in
environment0
18: 2,5,2'-TrCB
44:, 2,3,2',5'-TeCB
49: 2,4,2',5'-TeCB
52: 2,5,2',5'-TeCB
70: 2,5,3',4'-TeCB
74: 2,4,5,4'-TaCB
151:2,3,5,6,2',5'-HxCB
177: 2,3,5,6,2',3',4'-HpCB
187: 2,3,5,6,2',4',5'-HpCB
201: 2,3,4,5,2',3',5',6'-OCB
Potential
for toxJcftvd
37: 3,4,4'-TrCB
81: 3,4,5,4'-TeCB
114:2,3,4,5,4'-PeCB
119:2,4,6,3',4'-PeCB
123:3,4,5,2',4'-PeCB
157:2,3,4,3',4',5'-HxCB
158:2,3,4,3',4',6'-HxCB
167:2,4,5,3',4',5'-HxCB
168:2,4,6,3',4',5'-HxCB.
189:2.3,4,5,3',4',5'-HpCB
Pure 3-methylchdlanthrene-typ« inducers and mixed-type inducers reported frequently in environmental samples.
°Phenobarbital-type inducers reported frequently in environmental samples.
cWeak inducers or noninducers reported frequently in environmental samples.
dMixed-type inducers not reported frequently in environmental samples, but lexicologically active.
Source: Adapted from McFariand and Clarke (1989).
21 Subsequently, WHO (World Health Organization) derived TEFs for dioxin-like
22 PCBs (Ahlborg et al., 1994). Included are congeners that show structural similarity
23 to chlorinated dibenzo-p-dioxins and dibenzofurans, bind to the aryl hydrocarbon
24. receptor, elicit dioxin-specific biochemical and toxic responses, and persist and
25 accumulate in the food chain. On the basis of these criteria, 13 PCB
26 congenersothers may have been omitted for lack of dataare assigned TEFs,
27 expressed as a fraction of the toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin.
28 These are given in table 3-3; use of TEFs to supplement the mixture-based
29 approach developed in this assessment is discussed in section 4.4.
DRAFT-DO NOT QUOTE OR CITE 25
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
3
4
5
6
8
9
10
11
Table 3-3. WHO interim TEFs for human intake of dioxin-like PCBs
Non-crtho conww
TEF
77: 3.4,3'.4'-TeCB 0.0005
126:3.4.5,3',4'-PtCB 0.1
169: 3,4,5,3',4',5'-HxC8.0.01
Mono-oftho cong«n«f
105: 2,3,4,3',4'-PaCB
114: 2,3,4,5,4'-PeCB
118: 2,4,5,3',4'-PeCB
123: 3,4,5,2',4'-PeCB
156: 2,3,4,5.3',4'-HxCB
157: 2,3,4.3',4',5'-HxCB
167: 2,4,5,3',4',5'-HxC8
189: 2,3,4,5,3',4',5'-HpCB
Source: Adapted from Ahlborg at al. (1994).
HE
0.0001
0.0005
0.0001
0.0001
0.0005
0.0005
O.OO001
0.0001
Di-ottho eonq«n«f
TEF
170: 2,3,4,5,2',3',4'-HpCB 0.0001
180: 2,3,4,5,2',4',5'-HpCB 0.0000,1
12
4. APPLICATION OF THE DOSE-RESPONSE ASSESSMENT
13 4.1. APPLICATION TO PCB MIXTURES IN THE ENVIRONMENT
14 After re/ease Into the environment, PCB mixtures change over time so their
15 composition differs from commercial mixtures. How can toxicity values for
16 commercial mixtures be applied to mixtures in the environment?
17 After release into the environment, PCB mixtures undergo partitioning,
18 transformation, and bioaccumulation (see section 1.1). Through partitioning,
19 different fractions of the original mixture appear in air, water, sediment, and soil
20 due to the congeners' different rates of volatilization, solubility, and adsorption
21 (Callahan et al., 1979; Hutzinger et al., 1974; Erickson, 1986). Chemical
22 transformation tends overall toward dechlorination and can result in substantial
23 changes in the refative distributions of PCB congeners (Abramowicz, 1990; Brown
24 and Wagner, 1990; Lake et al., 1992). Bioaccumulation and selective metabolism
25 in the food chain tend to concentrate congeners of higher chlorine content and can
?-
26 result in residues that are considerably different from the original Aroclors
27 (Schwartz et al., 1987; Oliver and Niimi, 1988; Lake et al., 1995).
DRAFT-DO NOT QUOTE OR CITE
26
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Although environmental mixtures are often characterized in terms of
2 Aroclors, this can be both imprecise and inappropriate. Complex judgments offer
3 occasion for qualitative and quantitative errors, and large differences, have been
4 found in results reported by laboratories analyzing the same sediment samples
5 (Alfofd-Stevensetal., 1985; Alford-Stevens, 1986). Schwartz et al. (1987)
6 analyzed PCB residues from fish and turtles and compared tfiem to Aroclor 1242,
7 1248, 1254, and 1260 standards; the Aroclor standards could not characterize the
',-'" >
8 fish and turtle residues, which had been substantially changed by environmental or
9 metabolic alteration. As described by McFarland and Clarke (1989), congener
10 distributions in several species, including humans, do not resemble any commercial
11 mixture. Safe (1994) wrote, "Regulatory agencies and environmental scientists
12 have recognized that the composition of PCBs in most environmental extracts does
13 not resemble the composition of the commercial products." Along similar lines,
14 .". ATSbR (1993) advised,
15 It is important to recognize that the PCBs to which people may be exposed
16 are likely to be different from the original PCB source because of changes in
17 congener and impurity composition resulting from differential partitioning and
18 transformation in the environment and differential biological metabolism and
19 retention. Because of this concern, current data are considered inadequate
20 to differentiate between the toxicity and carcinogenicity of environmental
21 PCB mixtures with any reasonable degree of confidence.
22 For these reasons, risks from environmental mixtures are not assessed by
23 reference to Aroclors. This does not mean all environmental mixtures are regarded
24 as equally potent; environmental mixtures differ from commercial mixtures and
25 from each other. To make distinctions about risks from environmental mixtures,
26 the range of potency observed for commercial mixtures (see section 3.2) can be
27 considered along with factors that increase or decrease risk.
28 Chlorine content appears to be associated with cancer risk. Studies of
29 Clophens A 30 and A 60 (Schaeffer et al., 1984) and Kanechlors 30O?400, and
30 500 (Ito et al., 1973, 1974) found more tumor^ with mixtures of higher chlorine
DRAFT-DO NOT QUOTE OR CITE 27 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 content. In contrast, more adenofibrosis occurred with Aroclor 1254 than with
2 1260 (Kimbrough et al., 1972). Overall, tumors seem to increase with chlorine
3 content, although a direct relationship is not apparent. Resistance to metabolism
4 and persistence in the body are loosely related to chlorine content; this partially
5 explains the greater experimental potency of commercial mixtures with high
6 chlorine content. Nonetheless, mixtures with low chlorine content are considered
7 to pose some cancer risk: their composition overlaps those of carcinogenic
8 commercial mixtures, partial lifetime studies in rats and mice show induction of
9 early stages of tumor development, and several mixtures and congeners with
10 40-50 percent chlorine content promote tumors. Further, the environmental
11 processes of partitioning, transformation, and bioaccumulation can increase
12 concentrations of some congeners of concern; thus dechlorination is not
13 synonymous with detoxification.
14 Also important are persistence and bioaccumulation through the food chain.
15 Each species, in turn, retains persistent congeners that prove resistant to
16 metabolism and elimination (Oliver and Niimi, 1988). Bioaccumulated PCBs appear
17 to be more toxic than commercial PCBs (Aulerich et al., 1986). Mink fed Great
18 Lakes fish contaminated with PCBs showed liver and reproductive toxicity
19 comparable to mink fed Aroclor 1254 at quantities three times greater (Hornshaw
20 et al., 1983). It is crucial to recognize that commercial PCBs tested in laboratory
21 animals were not subject to prior selective retention of persistent congeners
22 through the food chain. For exposure through the food chain, risks can be higher
23 than those estimated in this assessment.
24 These detefminants of toxicitychlorine content, persistence, and
25 bioaccumulationcan be related to exposure pathway. Evaporated or dissolved
26 congeners tend to be lower in chlorine content than the original mixture; they tend
j?--.
27 also to be more inclined to metabolism and elimination and lower in persistence
28 and toxicity. On the other hand, congeners adsorbed to sediment or soil tend to
' ' i .
DRAFT-DO NOT QUOTE OR CITE 28 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 be higher in chlorine content and persistence, and bioaccumulated congeners
2 ingested through the food chain tend to be highest of ail. Rates of these processes
3 vary over several orders of magnitude (Hutzinger et al., 1974; Erickson, 1986),
4 thus the effect of environmental processes can be greater than the order of
5 magnitude spread in potency or slope estimated from commercial mixtures (see
6 section 3.2). Consequently, the low end of these ranges is appropriate for drinking
7 water ingestion or vapor inhalation, where environmental processes are likely to
8 decrease risk, while the high end is appropriate for food chain exposure or
9 ingestion of contaminated sediment or soil, where environmental processes are
10 likely to increase risk. This use of exposure pathway as an indicator of lower or
11 higher potency allows risk assessments to make distinctions among environmental
12 mixtures as altered by environmental processes. Section 5 provides an example of
13 this approach.
14 4.2. APPLICATION TO DIFFERENT ROUTES OF ENVIRONMENTAL EXPOSURE
15 What inferences can be made about dermal or inhalation exposures, two
16 routes for which there are no lifetime cancer studies?
17 PCBs are absorbed through ingestion, inhalation, and dermal exposure, after
18 which they are transported similarly through the circulation (see section 2.4). This
19 provides a reasonable basis for expecting similar effects from different routes of
20 environmental exposure.
21 Dermal absorption through human skin is rapid initially and continues at a
22 slower rate'with further contact (Wester et al., 1987, 1990). Rhesus monkeys
23 exposed to Aroclor 1242 pr 1254 in soil for 24 hours absorbed 14 percent through
24 the skin (Wester et al., 1993). Absorption of PCBs from soil involves competition
25 between the lipophilic attraction of PCBs to skin and adsorption to organic soil
26 material. Absorption increases with duration of skin contact and moisture content
27 of soil, but these relationships have not been quantified. Overall, use of the low
DRAFT-DO NOT QUOTE OR CITE 29 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 end of the potency ranges (see section 3.2) for dermal exposure appears
2 appropriate in light of the substantial but incomplete absorption through the skin.
3 Inhaled PCBs can be rapidly absorbed, although rates have not been
4 quantified (ATSDR, 1993). Rapid absorption, however, suggests potency by
5 inhalation is comparable to potency by ingestion. Because PCBs are slowly
6 metabolized, little uncertainty results from the first-pass effect, where ingested
7 toxicants are subject to metabolism in the liver before entering the circulation,
8 while inhaled toxicants enter the circulation before reaching the liver. In addition,
9 the capacity of Aroclor 1254 to promote mouse lung tumors (Anderson et al.,
10 1983, 1994; Beebe et al., 1992, 1993) suggests a risk of lung cancer by
11 inhalation. As with ingested mixtures, the composition of an inhaled mixture
12 influences its toxicity. Evaporated mixtures tend to have low chlorine content and
13 persistence, mixtures adsorbed to dust and soil tend to be high in this regard, and
14 mixtures suspended in an aerosol can be more diverse. Section 5 provides an
15 example of assessing inhalation exposure.
16 4.3. APPLICATION TO PARTIAL LIFETIME EXPOSURE
17 In assessing cancer risks from partial lifetime exposure, the common practice
18 is to prorate cumulative exposure over the lifespan (U.S. EPA, 1986a, 1992a). For
19 example, exposure lasting 7 years of a 70-year lifespan would be assumed to have
20 one-tenth the effect of lifetime exposure. Does the information available for PCBs
21 support this default or suggest an alternative?
22 Partial lifetime studies in rats and mice (see section 2.2) suggest less-than-
23 lifetime exposure can quickly induce high incidences of early stages of tumor
24 development (Kimbrough et al., 1972; Ito et al., 1973, 1974; Rao and Banerji,
25 1988). With further exposure, these can progress to malignancy (Kimbrough
_" , '&! . .
26 et al., 1975; Norback and Weltman, 1985). Tumor incidences from some partial
27 lifetime exposures can be similar to those of full lifetime exposure (Rao and Banerji,
DRAFT-DO NOT QUOTE OR CITE 30 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 ,1988). Timing of exposure to PCBs, particularly early-life exposure, is an
2 important factor in carcinogenesis (Rao and Banerji, 1988).
3 Persistence in the body is another significant consideration. PCBs entering
4 the body are transported by the circulation to internal organs and fat, where they
5 are stored (Matthews and Anderson, 1975). Equilibrium is maintained among
6 external exposure levels, concentrations in blood, and concentrations in fat and
7 other tissues. When external exposure is reduced, to maintain equilibrium, stored
8 PCBs reenter the circulation and provide a continuing internal source of exposure
9 (Matthews and Anderson, 1975). Thus PCBs from short-term exposure can be
10 stored in the body and emerge as a source of exposure much later.
11 , Persistence in the body can enhance the opportunity for PCB congeners to
12 express tumor promoting activity (Safe, 1994). Persistent congeners can retain
13 biological activity long after exposure stops (Anderson et al., 1991); some
14 persistent congeners are tumor promoters. The congener 3,4,3',4'-
15 tetrachlorobiphenyl continues to promote tumors vigorously-after dosing stops, but
16 not 2,4,2', 5'-tetracnlorobiphenyl(Buchmann et al., 1991; Luebeck et al., 1991).
17 Although the probability of liver focus extinction is high after dosing stops, those
18 that become large tend to persist (Luebeck et al., 1991). This would allay concern
19 for short-term exposure but increase concern as exposure duration increases.
20 Taken together, these results indicate both external and internal exposure
21 are important. When lifetime average daily dose (LADD) Is calculated as the
22 product of concentration C, intake rate //?, and exposure duration ED divided by
23 body weight BWand lifetime LT (U.S. EPA, 1992a),
24 V LADD = C x IR x ED 1 (BW x LT)
25 exposure duration would ideally reflect both external exposure and persistence of
26 internal exposure after external exposure stops.
, , . -S3&- . '
27 Persistence of internal exposure is not directly measured; a related quantity
28 for which estimates are sometimes available is naif-life in the body. One approach
DRAFT-DO NOT QUOTE OR CITE 31 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 is to approximate persistence of .internal exposure by 1.4 times half-life6 and add
2 it to duration of external exposure to estimate overall exposure duration. Half-lives
3 of 3.1 years for Aroclor 1242 and 6.5 years for Aroclor 1254 were estimated for
4 workers with relatively low PCB concentrations (Phillips et al., 1989). From these
5 half-life estimates, persistence of internal exposure would be approximated by
6 4 years (1.4 times 3.1 years) for mixtures of lesser persistence to 9 years (1.4
7 times 6.5 years) for mixtures of greater persistence. Although other half-life
8 estimates exist, the higher estimates are based on smaller samples and the lower
9 estimates are based on higher occupational concentrations (see section 2.4).
10 Taking these half-lives to be representative of environmental PCBs in the
11 general population, this assessment recommends adding 4 years duration to less
12 persistent mixture exposures and 9 years duration to more persistent mixture
13 exposures. This recommendation is consistent with the dose measure used in
14 section 3, where no adjustment was made for decreased or discontinued dosing in
15 expectation that internal exposure levels would remain high. Other reasonable
16 approaches can be developed; this approach, which uses half-lives estimated for
17 different PCB mixtures, presents a practical alternative to the default of ignoring
18 persistent internal exposure. Section 5 provides an example of this approach.
19 4.4. APPLICATION WITH DIOXIN TOXIC EQUIVALENCE FACTORS
20 TEFs have been proposed for some dioxin-like PCB congeners, while this
21 assessment is based on studies of mixtures. How can the TEF approach
22 supplement"the mixture-based approach developed in this assessment?
23 Model internal dose as an exponentially decreasing function of time: rff=
-------
DRAFT-DO NOT QUOTE OR CITE
1 When assessing PCS mixtures, it is important to recognize that PCB
2 carcinogenesis likely arises by both dioxin-like and nondioxin-like modes of action
3 (Safe, 1994> McFarland and Clarke, 1989; Birnbaum and DeVito, in press).
4 Relatively few PCB congeners are dioxin-like (U.S. EPA, 1994d), hence dioxin toxic
5 equivalence can cover only part of a PCB mixture's toxicity. A mixture-based
6 approach is necessary to avoid underestimating cancer risks' (as well as risks of
7 neurotoxicity and endocrine disruption, Birnbaum and DeVito, in press).
8 When assessing mixtures of PCBs and other dioxin-like compounds, it is
9 important to determine whether concentrations of dioxin-like compounds are
10 increased over those in the Aroclors on which this assessment is based. For
11 example, cooking food contaminated with PCBs can cause formation of chlorinated
12 dibenzofurans. When dioxin-like compounds are enhanced, the mixture-based
13 approach can be supplemented by dioxin f EFs to evaluate dioxin-like toxicity above
14 that present in the Aroclors. To determine whether dioxin-like compounds are
15 enhanced, levels of chlorinated dibenzo-p-dioxins and dibenzofurans in
16 environmental samples can be compared to levels in Arodors. Concentrations of
17 chlorinated dibenzofurans are approximately 1-2 ppm in Aroclors, 8 ppm in
18 Cldpheh A 60, and up to 20 ppm in Kanechlors (WHO, 1993). Chlorinated
19 dibenzo-p-dioxins have not been reported in commercial PCB mixtures.
20 When assessing mixtures of dioxin and related compounds, it is important to
21 consider the contribution of dioxin-like PCBs to total dioxin equivalents (U.S. EPA,
22 1994d). TEFs from table 3^3 can be added to TEFs for other dioxin-like
23 compounds: In some situations, PCBs can contribute more dioxin-like toxicity than
24 chlorinated dibenzo-p-dioxins and dibenzofurans (Schecter et al., 1994; Oewailly
25 et al., 1991, 1994). The congener 2,4,5,3',4'-pentachlorobiphenyl, shown to
26 have tumor-promoting activity, is a major contributor to total dioxin equivalents in
27 the United States (Patterson et al., 1994) and maritime Quebec (Dewa^lyet al.,
28 1994).
DRAFT-DO NOT QUOTE OR CITE 33 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Exposure to dioxin-like PCBs adds to background exposure of dioxin-like
2 compounds and augments processes associated with dioxin toxicity. There is
3 support for using low-dose linear dose-response models for incremental doses that
4 add to existing background exposure (Crump et al., 1976; Lutz, 1990). Thus
5 confidence in this assessment's use of low-dose linear models is enhanced for the
<
6 dioxin-like portion of a PCB mixture.
7 5. CHARACTERIZATION AND GUIDANCE FOR RISK ASSESSORS
8 5.1. DOSE-RESPONSE CHARACTERIZATION
9 Joint consideration of cancer studies and environmental processes leads to a
10 conclusion that environmental PCB mixtures pose a risk of cancer to humans.
11 Although environmental mixtures have not been tested in cancer assays, this
12 conclusion is supported by several complementary sources of information. High
13 incidences of liver carcinomas and adenomas were induced in three rat strains by
14 Aroclor 1260 and Clophen A 60, mixtures composed mostly of penta-, hexa-, and
15 heptachlorobiphenyls. Significant increases in gastric adenocarcinomas and male
16 leukemia and iymphoma were induced by Aroclor 1254. The apparent increases in
17 liver cancer for Arpclor 1254 and Clophen A 30, though not statistically
18 significant, were considered dose related and are consistent with some cancer risk
19 for mixtures of lower chlorine content. Partial lifetime studies in rats and mice fed
20 Aroclor 1254 or Kanechlors 300, 400, and 500 suggest mixtures composed
r ' '
21 mostly of di- through hexachiorobiphenyls can induce cancer over a longer
22 duration. Studies of tumor promoting activity implicate several tetra-, penta-, and
23 hexachiorobiphenyls as tumor promoters. These congeners have been found in
24 environmental samples and in a variety of organisms, including humans, the result
DRAFT-DO NOT QUOTE OR CITE 34 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
*\, of the environmental processes of partitioning, chemical transformation, and
2 preferential bioaccumulation.
.3 Risks for environmental mixtures are inferred from the range of toxicity
4 observed for commercial mixtures and the action of environmental processes to
5 alter mixtures in the environment, thereby decreasing or increasing risk. The range
6 observed for commercial mixtures may underestimate the true range for
7 environmental mixtures, because only a limited number of commercial mixtures
8 have been tested. Use of this range for environmental mixtures reflects a choice
9 to base risk estimates on experimental results, however limited, rather than apply
10 safety factors to compensate for lack of information. In contrast, the
11 environmental processes of partitioning, transformation, and bioaccumulation have
12 been extensively studied, and exposure pathway is a reliable indicator of whether
13 toxicity has been decreased or increased by environmental processes. The use of
14 exposure pathway to represent environmental processes increases confidence in
15 the risks inferred for environmental mixtures,
16 Estimates of cancer potency are based on animal studies of commercial
17 mixtures. ED 10s (estimated dose associated with 10 percent increased incidence)
18 span a range of 0; 1-0.7 mg/kg-d average lifetime exposure; EDO 1s span a range
19 of 0.01 -0.07 mg/kg-d. These estimates are mostly independent of choice of
20 model. The ranges reflect experimental uncertainty and variability of commercial
21 mixtures, but not human heterogeneity or differences between commercial and
22 environmental mixtures.
23 Low-dose extrapolation is based on models that are linear at low doses. In
24 part, this reflects the inability of studies with one exposed group to provide
25 information on the shape of the dose-response curve. Low-dose linear models are
26 appropriate when a carcinogen acts in concert with other exposures and processes
27 leading to a background incidence of cancer. Even when the mode of action
28 indicates a nonlinear dose-response curve in homogeneous animal populations, the
DRAFT-DO NOT QUOTE OR CITE 35 1/24/96
-------
DRAFT-DO NOf QUOTE OR CITE
1 presence of genetic and lifestyle factors in a heterogeneous human population
2 tends to make the dose-response curve more linear. Slopes based on the ED01
3 span a range of 0.1-0.9 per mg/kg-d average lifetime exposure. Extrapolation
4 based on the ED01 starts just below the experimental range, is mostly independent
5 of choice of model, and is statistically stable without requiring use of upper
6 bounds. Upper-bound slopes, based on either the ED01 or the linearized multistage
7 procedure, span a range of 0.3-2 per mg/kg-d. When a linear model is used at low
8 doses (below the ED01), slopes are multiplied by exposure levels to estimate the
9 risk of cancer. '
10 Depending on the specific application, either central estimates or upper
11 bounds can be appropriate. When considering exposure standards to protect
12 public health, emphasis would be placed on upper bounds and sensitive groups.
13 There is no scientific basis for expecting less sensitive groups or an average of
14 exposed groups to be representative or protective of a heterogeneous human
15 population. On the other hand, when describing aggregate effects across a
16 population, central estimates or a distribution of estimates are preferred, as upper
17 bounds are unlikely to be descriptive of the tendencies of the whole. When
18 comparing or ranking environmental hazards, central estimates can be appropriate,
19 while upper bounds provide information about the precision of the comparison or
20 ranking. Comparing a central estimate with its upper bound indicates whether the
21 central estimate is precise enough to support credible risk estimates.
22 Uncertainty around these estimates extends in both directions. Estimates
23 based on animal studies benefit from controlled exposures and absence of
24 confounding factors; however, there is uncertainty in extrapolating dose and
25 response rates across species. Information is lacking to evaluate high-to-low-dose
26 differences. Studies suggest that differences in toxicity across different exposure
27 routes are small. One study suggests timing of exposure is important, particularly
DRAFT-DO NOT QUOTE OR CITE 36 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 early in life. The principal uncertainty, though, is using commercial mixtures to
2 make inferences about environmental mixtures.
3 ... When exposure involves the food chain, however, uncertainty extends
4 principally in one direction: through the food chain, living organisms selectively
5 bioaccumulate persistent congeners, but commercial mixtures tested in laboratory
6 animals were not subject to prior selective retention of persistent congeners.
7 Bioaccumulated PCBs appear to be more toxic than commercial PCBs. For
8 exposure through the food chain, risks can be higher than those estimated in this
9 assessment. Two highly exposed populations, nursing infants and consumers of
10 contaminated game animals, are exposed through the food chain.
'11 Regarding partial lifetime exposure, some PCBs persist in the body and retain
12 biological activity, including tumor promotion, long after exposure stops. It is,
13 therefore, important to consider both external and internal exposure. A new
14 approach recommends exposure assessments use half-life information to represent
15 persistence of internal exposure. Although half-life can underestimate a mixture's
16 long-term persistence, using half-life as part of the overall exposure duration would
17 improve the current default practice of ignoring internal exposure after external
18 exposure stops.
19 The dioxin-like nature of some PCBs raises a concern for cumulative
20 exposure, as dioxin-like congeners add to background exposure of other dioxin-iike
21 compounds and augment processes associated with dioxin toxicity. This weighs
22 against considering PCB exposure in isolation or as an increment to a background
23 exposure of zero. ^Confidence in this assessment's use of low-dose linear models
24 is enhanced when there is additivity to background exposures and processes.
25 To gauge the distance of an extrapolation, human exposures can be
26 compared with those in animal studies to estimate a "margin of exposure." This is
27 a comparison of relative exposure, not risk. Calculating a meaningful margin,
28 however, is not simple; Since PCBs partition, transform, and bioaccumulate in the
DRAFT-DO NOT QUOTE OR CITE 37 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 environment, some congeners increase in concentration while others decrease, and
2 no simple measure can capture this complexity. The congeners operate through
3 different modes of action and have potencies ranging over several orders of ,
4 magnitude. Additionally, since RGBs can operate with other compounds to cause
5 cancer through both dioxin-like and nondioxin-like modes of action, identifying and
6 quantifying these other compounds is necessary for a true measure of total
7 exposure. Despite these limitations, a crude margin of exposure using total PCBs
8 can be of some value: for example, the.distance of extrapolation is smaller, hence
9 confidence in a risk statement is greater, when exposure involves a fishrbased diet
10 contaminated at 2 ppm than for exposure from drinking water contaminated at
11 0.5 ppb.
12 5.2. INFLUENCE OF CANCER GUIDELINE REVISIONS
13 , This assessment has been influenced by ideas being discussed in revision of
14 EPA's cancer guidelines (U.S. EPA, 1994a, 1994b). Revised guidelines are
15 expected to be proposed in 1996.
16 Most prominent is development of a range of potency estimates, considering
17 all lifetime cancer studies, instead of focusing on a few studies showing greatest
18 sensitivity. Two approaches are used to derive potency estimates: the linearized
19 multistage procedure and the ED0.1 approach, which uses ED01 as both a
20 benchmark for potency ranking and a starting point for low-dose extrapolation.
21 The ED01 approach provides a statistically stable method for deriving central
22 estimates of low-dose slopes. Dose calculations use the interagency consensus
23 cross-species scaling factor, based on the 3/4 power of body weight (U.S. EPA,
24 1992b).
25 The new emphasis on discussing circumstances under which carcinogenesis
26 is likely, especially exposure route considerations, is found throughout
-------
DRAFT-DO NOT QUOTE OR CITE
1 the composition and toxicity. of PCS mixtures. Exposure circumstances are
'2 addressed in a framework that distinguishes different exposure pathways as lower
3 risk or higher risk. Another key factor is persistence of PCB mixtures in the body,
4 leading to a new approach for quantifying persistence of internal exposure to
5 replace the current default of ignoring internal exposure after external exposure
6 stops. ' -
7 None of these features, however, is inconsistent with the current guidelines
8 (U.S. EPA, 1986a). The intent of these guidelines is "to permit sufficient flexibility
9 to accommodate new knowledge arid new assessment methods as they emerge."
10 Each new feature of this assessment can be viewed in this spirit.
11 5.3. RESEARCH NEEDS, RESEARCH IN PROGRESS, AND PRELIMINARY RESULTS
12 Research to address the principal uncertainties in this assessment includes:
13 Cancer studies of environmental mixtures found in the food chain,
14 soil, and water.
15 Further investigation of tumor development after exposure stops.
16 Further investigation of early-life sensitivity.
17 Studies to quantify absorption and retention by ingestion, inhalation,
18 and dermal exposure, particularly when adsorbed to organic particles.
19 Cancer studies of critical congeners identified by McFarland and
20 Clarke (1989), including mechanistic studies to elucidate interactions.
21 New information will soon become available from 2-year feeding studies
22 being sponsored by the General Electric Company at Battelle Laboratories in
*" ' " ' ' .
23 Columbus, Ohio. Under study are Aroclor 1016 at 50, 100, and 200 ppm;
24 Aroclor 1242 at 50 and 100 ppm; Aroclor 1254 at 25, 50, and 100 ppm; and
25 Aroclor 1260 at 25, 50 and 100 ppm. Each dose group has 50 male and 50
26 female Sprague-Dawley Crl:CD rats; there are 100 controls of each sex! To
27 investigate tumor progression,, additional rats were killed at 13, 26, 39, 52, and
DRAFT-DO NOT QUOTE OR CITE 39 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 78 weeks; other female rats were dosed for 52 weeks and killed at 78 or
2 104 weeks.
3 Recently, the General Electric Company (1995) reported preliminary results
4 of statistically significant increased incidences of liver tumors in female rats for all
5 Aroclors and all doses except the lowest dose of Aroclor 1016 and in male rats for
6 Aroclor 1260. They later submitted the preliminary liver tumor incidences to U.S.
7 EPA for inclusion in this assessment; these are presented in table 5-1.
8
Table 5-1 . Preliminary liver tumor incidences from the
Commercial mixture Dose
Aroclor 1016 Control
50 ppm
100 ppm
200 ppm
Aroclor 1242 Control
50 ppm
1 00 ppm
Aroclor 1254 Control
25 ppm
50 ppm
1 00 ppm ,
Aroclor 1 260 Control
25 ppm
50 ppm
100 ppm
."incidence of adenomas or carcinomas
observed.
blncid«nce of adenomas or carcinomas
Females8
1/85
1/48
7/45
6/50
1/85
11/49
17/45
1/85
19/45
29/49
29/49
1/85
10/49
11/45
24/50
GE study
Male«b
7/98
2/47
, 2/49
4/49
7/98
1/50
5/46
7/98
4/48
4/49
. 6/47
7/98
3/50
7/48
10/47
in rats alive when the first tumor was
in rats alive at 1 year.
Source: Unpublished preliminary results submitted to U
.S. EPA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27.
28
29 Analysis of these preliminary results shows the anticipated effect of the new
30 studies on the potency and slope estimates. As done for the published studies in
*-A*
31 section 3, the multistage model (U.S. EPA, 1980; Howe et al., 1986) is fitted in
DRAFT-DO NOT QUOTE OR CITE 40 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 the experimental range to data for each mixture and sex. Administered doses are
2 converted from ppm in the diet to mg/kg-d by assuming rats consume food equal
3 to 5 percent of body weight daily (U.S. EPA, 1980), then scaling to humans using
4 a factor based on the 3/4 power of body weight (U.S. EPA, 1992b). Response is
5 taken as the incidence of rats with either hepatoceilular carcinomas or adenomas.
6 Preliminary potency and slope estimates are compiled in table 5-2 (the format is
7 similar to table 3-1); details of the calculations appear in appendix tables A-11
8 through A-18.
f * . - - , *
9 Table 5-2. Preliminary potency and slope estimates for humans from the GE study
10 Bound on LMP
11 ED10a ED01b Slope0 Sloped Slope9 See
,12 Mixture and sex (mo/kg-d) (mo/ka-dl (mq/ka-d)"1 (mq/ka-d)~1 (mq/ka-d)"1 table
13 Aroclor 1016, females 1.8 0.17 0.06 0.09 0.09 .-11
14 Aroclor 1242, females O.30 0.028 0.4 0.5 0.5 A-12
15 Aroclor 1254, females 0.07 0.007 1.4 1.8 1.8 A-13
16 Aroclor 1260, females 0.22 0.021 0.5 0.6 0.6 A-14
17 Aroclor 1016, malesf 3.4 2.3 0.004 0.03 0.03 A-15
18 " Aroclor 1242, malesf 1.5 0.99 0.01 0.06 0.06 A-16
19 Aroclor 1254, malesf 1.5 0.53 0.02 0.1 0.1 A-17
20 Aroclor 1260, males 0.97 0.20 0.05 0.2 0.2 A-18
21 aEstimated dose associated with 10% increased incidence.
22 Estimated dose associated with 1 % increased incidence.
23 cComputed as0.01/ED01.
24 dComputed as 0.01 divided by a lower bound on ED01.
25 e95% upper bound slope from linearized multistage procedure (U.S. EPA, 1986a).
26 fNo significant increase in liver tumors; quantities indicate sensitivity of study.
27 These estimates benefit from studies that include several dose levels,
28 providing information about the shape of the dose-response curve in the
29 experimental range. Basing low-dose extrapolation on the ED01, whie|i falls just
30 below the experimental range, allows extrapolations to fully reflect curvature
*&&
i '"', ' " ' ' t . -
DRAFT-DO NOT QUOTE OR CITE 41 1/24/96
-------
1
2
3
4
5
6
DRAFT-DO NOT QUOTE OR CITE
observed in the experimental range. The presence of several dose levels also
reduces uncertainty in the potency and slope estimates, resulting in upper bounds
that are closer to their central estimates, compared to those in table 3-1.
Table 5-3 compares slope estimates derived from published studies (see
table 3-1) with preliminary results of the new study (see table 5-2). Central
estimate slopes are used to facilitate comparisons.
Table 5-3. Comparison of slope estimates from published studies and new study
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Mixture and sex
Aroclor 1016, females
Aroclor 1242, females
Aroclor 1254, females
Aroclor 1260, females
Aroclor 1016, males
Aroclor 12428, males
Aroclor 1254, males
Aroclor 1260f, males
Slopes from published studies"
(mg/ko-d)
-1
No study
No study
0.1C (F344)
0.5 (Sherman), 0.8 (Sprague-Dawley)
No study
0.03 (Wistar)
0.2° (F344)
0.9 (Wistar), 0.06 (Sprague-Dawley)
Slopes from new study'1
(mg/kg-dr1
0.06 (Sprague-Dawley)
0.4 (Sprague-Dawley)
1.4 (Sprague-Dawley)
0.5 (Sprague-Dawley)
0.004d (Sprague-Dawley)
0.01d (Sprague-Dawley)
0.02d (Sprague-Dawley)
0.05 (Sprague-Dawley)
a0.01/ED01 from published studies (see table 3-1).
b0.01/ED01 from new GE study (see table 5-2).
cUver tumors only; ga'stric adenocarcinomas and male leukemia and lymphoma were
significantly increased.
dNo significant increase in liver tumors; quantities indicate sensitivity of study.
"Includes Clophen A 30 study in Wistar rats.
'includes Clophen A 60 study in Wistar rats.
25
26
27
28
29
30
31
made:
From-these results and analyses, several preliminary observations can be
* * ',
The results support this assessment's conclusion that all PCB mixtures
can pose a risk of cancer.
There is no evidence of sublinearity in the experimental rc/sge.
The lower potency of Aroclor 1016 suggests cancer risk is reduced
, ' j»&
for mixtures composed entirely of congeners with 1 -4 chlorines.
DRAFT-DO NOT QUOTE OR CITE
42
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 . The similar slopes for Arocfor 1242 and 1260 females, together with
2 the overlapping composition of these mixtures, casts doubt on
3 chlorine content of the original, mixture being a useful indicator of
4 cancer potency in this range of chlorine content.
5 There are significant differences among studies of the same mixture
6 and sex. These can be due to rat strain, laboratory differences, or lot-
7 to-lot differences of Aroclors.
8 The new study's pattern of liver cancer in females and weaker or
9 negative results in males parallels the dioxin results, suggesting a
10 dioxin^like mode of action may be important for many common PCB
11' mixtures.
12 .. . The significant results for Aroclor 1260 males indicate a nondioxin-like
13 mode of action is also operating.
14 If these preliminary results are confirmedand no other tumor sites are
15 foundpotency ranges could be augmented by values derived from female rats in
16 the new study. The composition of Aroclor 1016 could be used to refine the
17 criteria for using the low end of the potency range, specifying that a low-potency
18 mixture would contain only congeners with 1-4 chlorines.
19 Beyond these preliminary results, data on congener composition in the
20 Aroclors and rat tissues from the new study may make it possible to perform a
21 factor analysis, which can lead to identifying a subset of congeners most
22 responsible for tumor induction. Field analyses can then reduce uncertainty by
23 quantifying a small number of critical congeners. Another topic for further study is
24 to test the hypothesis that carcinogenic activity for Aroclor 1016 is due to
25 tetrachlorobiphenyls only, with no contribution from congeners with 1-3 chlorines.
26 Current information is inadequate to evaluate this hypothesis.
DRAFT-DO NOT QUOTE OR CITE 43 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 5.4. SUMMARY OF GUIDANCE FOR RISK ASSESSORS
2 Joint consideration of cancer studies and environmental processes leads to a
3 conclusion that environmental PCB mixtures pose a risk of cancer to humans. The
4 cancer potency of PCB mixtures is described by two ranges of estimates (see
5 table 3-1 and following text). Upper-bound slopes span a range of 0.3-2 per
6 , mg/kg-d average lifetime exposure to PCBs. Central estimates of these slopes
7 span a range of 0.1-0.9 per mg/kg-d. Depending on the purpose of a risk
8 assessment, either central estimates or upper bounds can be appropriate. Upper
9 bounds are useful for estimating risks or setting exposure standards to protect
10 public health.
11 Because PCB mixtures partition, transform, and bioaccumulate in the
12 environment, toxicity is likely to decrease by some exposure pathways and
13 increase by others. The effect of these environmental processes can be greater
14 than the order of magnitude spread in the slope estimates; consequently, the low
15 end of these ranges is used when environmental processes are likely to decrease
16 risk, and the high end is used when environmental processes are likely to increase
17 risk. When congener analyses are available from environmental samples,
18 uncertainty can be reduced by verifying the presence or absence of persistent
19 congeners, dioxin-Hke congeners, and tumor promoting congeners, (individual
20 congener concentrations would be most useful, as Aroclor-based characterizations
21 can be misleading.) In the absence of congener analyses, an instructive default is
^ .
22 to regard drinking water ingestion, vapor inhalation, and dermal exposure as lower
23 risk; sediment ingestion and dust inhalation as higher risk; and exposure through
24 the food chain, which contains bioaccumulated PCBs, as highest risk. For
25 , exposure through the food chain, risks, can be higher than those estimated in this
26 assessment.
DRAFT-DO NOT QUOTE OR CITE 44 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Table 5-4 lists some factors that are useful in evaluating risks from
2 environmental mixtures. These factors are discussed in section 4, summarized. in
3 section 5.1, and used in an example in section 5.5.
4
5
6
7
8
9
10
11
2
13
14
15
5~4- Facto" to consider in assessing risks from PCBs in the environment
Lower risk
> Low chlorine content in the
environment
-.-'.
> Vapor inhalation .
* Ingestion of water-soluble
congeners
* Dermal exposure
Higher risk and
+ ' Food chain exposure
> Persistent congeners
.
> Dioxin-like congeners
* Tumor-promoting congeners
* . High chlorine content in the
environment
>. Oust or aerosol inhalation
-
» Sediment or soil ingestion
> Early-life exposure
16 PCBs persist in the body, providing a continuing source of internal exposure
1 7 after external exposure stops. To reflect persistence of internal exposure, this
18 assessment recommends adding 4 years duration to less persistent exposures and
19 9 years duration to more persistent exposures, as indicated in table 5-4. One
20 study suggests timing of exposure is important, particularly early in life; this further
21 supports not prorating early-life exposures over a lifespan without adjustment.
22 Highly exposed populations include nursing infants and consumers of
23 contaminated game fishr game animals, or products of animals (such as seabird
24 eggs) high on the food chain.
25 Joint consideration of toxicity information and environmental processes
DRAFT-DO NOT QUOTE OR CITE
45
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 allows flexibility to make reasoned judgments on a case-by-case basis. The
2 specific guidance in this section, together with potency ranges not exceeding an
3 order of magnitude, encourages consistency in risk assessments. The choices and
4 judgments made in applying this assessment to environmental mixtures is thus
5 grounded in scientific information from a variety of sources.
6 5.5. EXAMPLE
7 Consider a release of PCBs onto the ground near a river or lake. Potential
8 pathways of human exposure include vapor inhalation, drinking water, fish
9 ingestion, and skin contact with ambient water and contaminated soil. The
"* ' "
10 population of interest includes anglers who consume an average of two 105-g
11 portions of local fish each week. They spend most of their time in the area,
12 breathing 20 m3 air and drinking 2 L water, on average, each day. Skin contact
13 with ambient water and soil is negligible for this population. A 30-year exposure
14 duration is to be considered, with a representative lifespan of 70 years and body
15 weight of 70 kg. Environmental samples indicate long-term average concentrations
16 of 10 ng/m3 in ambient air, 100 ng/L in drinking water, and 0.1 ppm in the edible
17 portion of local fish. Dust in ambient air and sediment in drinking water are
18 negligible.
19 Because of partitioning, transformation, and bioaccumulation, different
20 fractions of the original mixture are encountered through these pathways, hence
21 different potency values and persistence values are appropriate. Vapor inhalation
22 appears in the "lower risk" column of table 5-4 (evaporating congeners tend to
**
23 have low chlorine content and be inclined to metabolism and elimination), so the
24 low end of the range (upper-bound slope of 0.3 per mg/kg-d and persistence of
25 internal exposure of 4 years) is used for vapor inhalation. Similarly, ingestion of
Y?"
26 water-soluble congeners appears in the "lower risk" column (dissolved congeners
DRAFT--DO NOT QUOTE OR CITE 46 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 tend.to have low chlorine content and be inclined to metabolism and elimination),
2 so the low end is also used for drinking water. (If ambient air or drinking water
3 had container-significant amounts of contaminated dust or sediment, the high-end
4 potency and persistence values would be appropriate, as adsorbed congeners tend
5 to be of high chlorine content and persistence.) Food chain exposure appears in
6 the "higher risk" column (aquatic organisms and fish selectively accumulate
7 congeners of high chlorine content and persistence that are resistant to metabolism
8 and elimination), so the high end of the range (upper-bound slope of 2 per mg/kg-d
9 and persistence of internal exposure of 9 years) is used for fish ingestion.
10 The lifetime average daily dose (LADD) is calculated as the product of
11 concentration C, intake rate IR, and exposure duration ED divided by body weight
12 BW and lifetime Z.7"(U.S: EPA, 1992a),
13 LADD = C x //? x EDI (BW x IT)
14 To reflect continuing internal exposure after external exposure stops, the
15 persistence of internal exposure (either 4 or 9 years) is added to the 30-year
16 (external) exposure duration. The LADD is calculated as follows:
17 Table 5-5. Sample ifetime average daMy dose calculations
18 Pathway C !& ED BW LT LAPP
19
20
21
Pathway C IR ED BW LT LAPP
Vapor inhalation 10 ng/m3 20 m3/d 30+4 yr 70 kg 70 yr 1.4 x 10~6 mg/kg-d
Drinking water 10Ong/L 2 L/d 30+4yr70kg 70 yr 1.4 x 10~6 mg/kg-d
Fish ingestion 0.1 ppm 30 g/d 30 + 9yr 70kg 70 yr 2.4 x10~5 mg/kg-d
22 For each pathway, the LADD is multiplied by the appropriate slope to
23 estimate risk:
DRAFT-DO NOT QUOTE OR CITE 47 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1
2
3
4
5
Table 5-6. Sample risk calculations
Pathway LAPP
Vapor inhalation 1 .4 x 1 0~6 mg/kg-d
Drinking water 1.4x1 0~6 mg/kg-d
Fish ingestion 2.4x10~5 mg/kg-d
Slope
0.3 per mg/kg-d
0.3 per mg/kg-d
2 per mg/kg-d
Risk
4.2x10-7
4.2x10~7
4.8x10~5
6 Sum 2.7 x10~5 mg/kg-d 4.9x10~5
7 Thus, fish ingestion is the principal pathway contributing to risk, and vapor
8 inhalation and drinking water exposure are of lesser consequence. It would be
9 advisable to further examine variability in fish tissue concentrations and fish
10 consumption rates to determine whether some individuals are at much higher risk.
11 it is important to remember that this specific site exposure adds to a background
12 level of exposure from other sources.
13 6. REFERENCES
14 Abramowicz, D.A. (1990) Aerobic and anaerobic biodegradation of PCBs: a review. Biotechnology
15 10(3):241-251.
16 Ahlborg, U.G.; Becking, G.C.; Birnbaum, L.S.; Brouwer, A.; Oerks, H.J.G.M.; Feeley, M.; Golor, G.;
17 Hanberg, A.; Larsen, J.C.; Liem, A.K.D.; Safe, S.H.; Schlatter, C.; Waarn, F.; Younes, M.;
18 Yrjinheikki, E. (1994) Toxic equivalency factors for dioxin-like PCBs. Chemosphere,
19 28(6):1049-1067.
20 Alford-Stevens, A.L.; Budde, W.L.; Bellar, T.A. (1985) Interlaboratory study on determination of
21 polychlorinated biphenyls in environmentally contaminated sediments. Anal. Chem.
22 57:2452-2457. ,
23 Alford-Stevens, A.L. (1986) Analyzing PCBs. Environ. Sci. Technol. 20(12):1194-1199.
24 Anderson, L.M.; van Havere, K.; Budinger, J.M. (1983) Effects of polychlorinated biphenyls on lung
25 and liver tumors initiated in suckling mice by /V-nitrosodimethyjamine. J. Natl ?Cancer Inst.
26 71(11:157-163. '"'
DRAFT-DO NOT QUOTE OR CITE 48 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Anderson, L.M.; Fox. S.D.; Dixon, D.; Beebe, L.E.; Issaq, H.J. (1991) Long-term persistence of
£ , polychlorinated biphenyi congeners in blood and liver and elevation of liver aminopyrine
3 . demethylase activity after a single high dose of Aroclor 1254 to mice. Environ Toxicol
4 , . Chem. 10:681-690. '
r ' = ' - . - , , ' " =-
5 Anderson, L.M.; Logsdon, D.; Ruskie, S.; Fox, S.D.; Issaq, H.J.; Kovatch, R.M.; Riggs, C.M. (1994)
6 Promotion by polychlorinated biphenyls of lung and liver tumors in mice. Carcinogenesis
7 15(101:2245-2248.
8 ATSDR (Agency for Toxic Substances and Disease Registry) (1993) Toxicologies! profile for
9 polychlorinated biphenyls. Atlanta: ATSDR, TP-92/16, update.
10 ATSDR (Agency for Toxic Substances and Disease Registry) (1995) Toxicological profile for
11 polychlorinated biphenyls. Atlanta: ATSDR, draft for public comment.
12 Aulerich, R.J.; Ringer, R.K.; Safronoff, J. (1986) Assessment of primary vs. secondary toxicity of
13 Aroclor 1254 to mink. Arch. Environ. Contam. Toxicol. 15:393-399.
14 Beebe, L; Fox, S.D.; Riggs, C.W.; Park, S.S.; Gelboin, H.V.; Issaq, H.J.; Anderson, L.M. (1992)
15 Persistent effects of a single dose of Aroclor 1254 on cytochromes P450IA1 and IIB1 in
16 mouse lung. Toxicol. Appi. Pharmacof. 114:16-24.
17 Beebe, L.E.; Kim, Y.E.; Amin, S.; Riggs, C.W.; Kovatch, R.M.; Anderson, L.M. (1993) Comparison
18 of transplacental and neonatal initiation of mouse lung and liver tumors by
19 /V-nitrosodimethylamine (NDMA) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)
20 and promotability by a polychlorinated biphenyls mixture (Aroclor 1254)..Carcinogenesis
21 14(8): 1545-1548.
22 Bertazzi, P.A.; Riboldi, L.; Pesatori, A.; Radice, L.; Zocchetti, C. (1987) Cancer mortality of
23 capacitor manufacturing workers. Am. J. Ind. Mod. 11:165-176.
24 Birnbaum, L.S.; DeVito, M.J. (in press) Use of toxic equivalency factors for risk assessment for
25 dioxins and related compounds. Toxicology.
26 Brown, b.P. (1987) Mortality of workers exposed to polychlorinated biphenylsan update. Arch.
27 Environ. Health 42(6):333-339.
28 Brown, J.F., Jr.; Wagner, R.E. (1990) PCB movement; dechlorination, and detoxication in the
29 Acushhet Estuary. Environ. Toxicol. Chem. 9:1215-1233.
30 Buchmann, A.; Ziegler, S.; Wolf, A.; Robertson, L.W.; Durham, S.K.; Schwarz, M. (1991) Effects
31 P^ polychlorinated biphenyls in rat liver: correlation between primary subcellular effects and
32 promoting activity. Toxicol. Appl. Pharmacol. 111:454-468.
33 Callahan, M.A.; Slimak, M.W.; Gabel, N.W.; May, I.P.; Fowler, C.F.; Freed, J.R.; Jennings, P.;
34 . Durfee, R.L.; Whitmore, F.C.; Maestri, B.; Mabey, W.R.; Holt, B.R.; Gould, Cf7l979)
35 Water-related environmental fate of 129 priority pollutants, vol. I, ch. 36. Washington: U.S.
36 EPA, Report No. EPA-440/4-79-029a. '~
DRAFT-DO NOT QUOTE OR CITE 49 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Cogliano, V.J. (1986) The U.S. EPA's methodology for adjusting the reportable quantities of
2 potential carcinogens. Proceedings of the 7th National Conference on Management of
3 Uncontrolled Hazardous Wastes (Superfund '86). Washington: Hazardous Materials Control
4 ' Research Institute, pp. 182-185.
5 Crump, K.S.; Hoel, D.G.; Langley, C.H.; Peto, R. (1976) Fundamental carcinogenic processes and
6 their implications for low dose risk assessment. Cancer Res. 36:2973-2979.
7 Delaware Department of Natural Resources and Environmental Control (1994) Summary and
8 assessment of polychlorinated biphenyls and selected pesticides' in striped bass from the
9 Delaware Estuary. Delaware, Project No. AFC-5; Grant No. NA26FA0148-01.
10 Dewailly, £.; Weber, J.-P.; Gingras, S.; LalibertS, C. (1991) Coplanar PCBs in human milk in the
11 province of Quebec, Canada: are they more toxic than dioxin for breast fed infants? Bull.
12 Environ. Contam. Toxicol. 47:491-498.
13 Dewailly, £.; Ryan, J.J.; Laliberte", C.; Bruneau, S.; Weber, J.-P.; Gingras, S.; Carrier, G. (1994)
14 Exposure of remote maritime populations to coplanar PCBs. Environ. Health Perspect.
15 102{Suppl. 1):205-209.
16 Erickson, M.D. (1986) Analytical chemistry of PCBs. Boston: Butterworth Publishers.
17 General Electric Company (1995) Letter from Stephen B. Hamilton, Jr., to U.S. Environmental
18 Protection Agency Section 8(e) Coordinator, October 10, 1995.
19 Hemming, H.; Ftodstrom, S.; Wirngard, L; Bergman, A.; Kronevi, T.; Nordgren, I.; Ahlborg, U.
20 (1993) Relative tumour promoting activity of three polychlorinated biphenyls in rat liver.
21 Eur.J. Pharmacol. 248:163-174.
22 Hong, C.-S.; Bush, B.; Xiao, J.; Qiao, H. (1993) Toxic potential of non-ortho and mono-ortho
23 coplanar polychlorinated biphenyls in Aroclors, seals, and humans. Arch. Environ. Contam.
24 Toxicol. 25:118-123.
25 Hornshaw, T.C.; Aulerich, R.J.; Johnson, H.E. (T983) Feeding Great Lakes fish to mink: effects on
26 mink and accumulation and elimination of PCBs by mink. J. Toxicol. Environ. Health
27 11:933-946.
28 Hovinga, M.E.; Sowers, M.; Humphrey, H.E.B. (1992) Historical changes in serum PCS and DDT
29 ' levels in an environmentally-exposed cohort. Arch. Environ. Contam. Toxicol. 22:362-366.
30 Howe, R.B.; Crump, (C.S.; Van Landingham, C. (1986) Global 86: a computer program to
31 extrapolate quanta! animal toxicity data to low doses. Prepared for U.S. EPA under contract
32 68-01-6826. ,
33 Hutzinger, 0.; Safe, S.; Zitko, V. (1974) The chemistry of PCB's. Boca Raton, FL: CRC Press.
"!??*"
34 IARC (International Agency for Research on Cancer) (1987) (ARC Monographs on the Evaluation of
35 Carcinogenic Risks to Humans, Supplement 7, Overall Evaluations of Carcinogenicity: An
36 Updating of IARC Monographs Volumes 1-42. Ly'bn, France.
DRAFT-DO NOT QUOTE OR CITE 50 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 Institute for Evaluating Health Risks (1991) Reassessment of liver findings in five PCB studies in
2 rats. Washington, DC, dated July 1.. Report submitted to.U.S. EPA.
3 Ito, N.; Nagasaki, H.; Arai, M.; Makiura, S.; Sugihara, S.; Hirao, K. (1973) Histopathologic studies
4 on liver tumorigenesis induced in mice by technical polychlorinated biphenyls and its
5 promoting effect on liver tumors induced by benzene hexachloride. J. Natl. Cancer Inst
6 51 (5):1637-1646.
7 Ito, N.; Nagasaki, H.; Maktura, S.; Arai, M. (1974) Histopathological studies on liver tumorigenesis
8 in rats treated with polychlorinated biphenyls. Gann 65:545-549.
9 Kannan, N.; Tanabe, S.; Tatsukawa, R. (1988) Toxic potential of non-orthp and morio-ort/jo
10 coplanar PCBs in commercial PCB preparations: "2,3,7,8-T4CDD toxicity equivalence
11 factors approach." Bull. Environ. Contam. Toxicol. 41:267-276.
12 Kimbrough, R.D.; Under, R.E.; Gaines, T.B. (1972) Morphological changes in livers of rats fed
1 3 polychlorinated biphenyls: light microscopy and ultrastructure. Arch. Environ. Health
14- 25:354-364.
15 Kimbrough, R.D.; Under, R.E. (1974) Induction of adenofibrosis and hepatomas of the liver in
16 BALB/cJ mice by polychlorinated biphenyls (Aroclor 1254). J. Nad. Cancer Inst
17 53(2):547-552.
18 Kimbrough, R.D.; Squire, R.A.; Under, R.E'.; Strandberg, J.D.; Montali, R.J.; Burse, V.W. (1975)
19 Induction of liver tumors in Sherman strain female rats by polychlorinated biphenyls
20 Aroclor 1260. J. Natl. Cancer Inst. 55:1453-1459.
21 Kimura, N.T.; Baba, J. (1973) Neoplastic changes in the rat liver induced by polychlorinated
22 biphenyl. Gann 64:105-108.
23 Laib, R.J.; Rose, N.; Brunn, H. (1991) Hepatocarcinogenicity of PCB congeners. Toxicol. Environ.
24 Chem. 34:19-22.
25 Lake, J.L.; Pruell, R.J.; Osterman, F.A. (1992) An examination of dechlorination processes and
26 pathways in New Bedford Harbor sediments. Marine Environ. Res. 33:31-47.
27 Lake, J.L.; McKinney, R.; Lake, C.A.; Osterman, F.A.; Heltshe, J. (1995) Comparisons of patterns
28 of polychlorinated biphenyl congeners in water, sediment, and indigenous organisms from
29 New Bedford Harbor, Massachusetts. Arch. Contam. Toxicol. 29:207-220:
30 Luebeck, E.G.; Moolgavkar, S.H.; Buchmann, A.; Schwarz, M. (1991) Effects of polychlorinated
31 biphenyls in rat liver: quantitative analysis of enzyme-altered foci. Toxicol. Appl. Pharmacol.
32 111:469-484.
33 Lutz, W.K. (1990) Oose-response relationship and low dose extrapolation in chemical
34 carcinogenesis. Carcinogenesis 11(8): 1243-1247. ^
35 Matthews, H.B.; Anderson, M.W. (1975) Effect of chloelnation on the distribution and excretion of
36 polychlorinated biphenyls. Drug Metab. Oispos. 3(5):371-380.
DRAFT-DO NOT QUOTE OR CITE 51 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 McConnell, E.E.; Solleveld, H.A.; Swenberg, J.A.; Boorman, G.A, (1986) Guidelines for combining
2 neoplasms for evaluation of rodent carcinogenesis studies. J. Natl. Cancer Inst.
3 76{2):283-289.
4 McFarland, V.A.V Clarke, J.U. (1989) Environmental occurrence, abundance, and potential toxicity
5 of polychlorinated biphenyl congeners: considerations for a congener-specific analysis.
6 Environ. Health Perspect. 81:225-239.
7 Moore, J.A.; Hardisty, J.F.; Banas, D.A.; Smith, M.A. (1994) A comparison of liver tumor
8 diagnoses from seven PCS studies in rats. Regul. Toxicol. Pharniacol. 20:362-370.
9 Morgan, R.W.; Ward, J.M.; Hartman, P.E. (1981) Aroclor 1254-induced intestinal metaplasia and
10 adenocarcinoma in the glandular stomach of F344 rats. Cancer Res. 41:5052-5059.
11 National Research Council (1993) Issues in risk assessment. Washington: National Academy Press.
12 National Research Council (1994) Science and judgment in risk assessment. Washington: National
13 Academy Press.
14 NCI (National Cancer Institute) (1978) Bioassay of Aroclor 1254 for possible carcinogenicity.
15 Carcinogenesis Tech. Rep. Ser. No. 38.
\
16 Norback, D.H.; Weltman, R.H. (1985) Polychlorinated biphenyl induction of hepatoceltular
17 carcinoma in the Sprague-Dawley rat. Environ. Health Perspect. 60:97-105.
18 Oliver, B.C.; Niimi, A.J. (1988) Trophodynamic analysis of polychlorinated biphenyl congeners and
19 other chlorinated hydrocarbons in the Lake Ontario ecosystem. Environ. Sci. Techno).
20 22:388-397.
21 Patterson, D.G.; Todd, G.D.; Turner, W.E.; Maggio, V.; Alexander, L.R.; Needham, L.L. (1994)
22 Levels of non-ortho-substituted (coplanar), mono-, and di-ortho-substituted polychlorinated
23 biphenyls, dibenzo-p-dioxins, and dibenzofurans in human serum and adipose tissue.
24 Environ. Health Perspect. 102(Suppl. 1): 195-204.
25 Phillips, D.L.; Smith, A.B.; Burse, V.W.; Steele, G.K.; Needham, L.L.; Hannon, W.H. (1989) Half-life
26 of polychlorinated biphenyls in occupationally exposed workers. Arch. Environ. Health
27 44(6) :351-354.
28 Rao, C.V.; Baoerji, A.S. (1988) Induction of liver tumors in male Wistar rats by feeding
29 polychlorinated biphenyls (Aroclor 1260). Cancer Lett. 39:59-67.
j* '
30 Rose, N.; Laib, R.J.; Brunn, H.; Bolt, M.M. (1985) Biotransformation and toxicity of polychlorinated
31 biphenyls (PCBs): investigation of initiating and promoting activities of 2,2',4,5'-tetra- and
32 2,2',4,4',5,5'-hexachlorobiphenyl. Naunyn-Schmiedeberg's Arch. Pharmacol. Suppl. 330.
33 (Abstr. 87), R21..
'3,~
34 Safe, S. (1990) Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs), dibenzofurans
35 (PCDFs), and related compounds: environmental and mechanistic consideration which
DRAFT-DO NOT QUOTE OR CITE 52 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 support the development of toxic equivalency factors (TEFs) Grit Rev Toxicoi
2 . 21(11:51-88.' ' ' . . '
3 Safe, S. (1994) Polychlorinated biphenyls (PCBs): environmental impact, biochemical and toxic
4 responses, and implications for risk assessment. Grit. Rev. Toxicoi. 24(2):87-149.
5 Sargent, L; Dragan, Y.P..; Erickson, C.; Laufer, C.J.; Pitot, H.C.,(1991) Study of the separate and
6 combined effects of the non-planar 2,5,2',5'- and 3,4,3',4'-tetrachlorobiphenyl in liver and
7 lymphocytes in vivo. Carcinogenesis 12(5):793-800. /
8 Schaeffer, E.; Greim, H.; Gbessner, W, (1984) Pathology of chronic polychlorinated biphenyl (PCS)
9 feeding in rats. Toxicoi. Appl. Pharmacol. 75:278-288.
10 Schecter, A.; Stanley,; J'.; Boggess, K.; Masuda, Y.; Mes, J.; Wolff, M.; Furst, P.; Furst, C.;
11 Wilson-Yang, K.; Chisholm, B. (1994) Polychlorinated biphenyl levels in the tissues of
12 exposed and nonexposed humans. Environ. Health Perspect. 102(Suppl. 1):149-158.
13 Schulz, D.E.; Petrick, G.; Duinker, J.C. (1989) Complete characterization of polychlorinated
14 biphenyl congeners in commercial Aroclor and Clophen mixtures by multidimensional gas
15 chromatbgraphyelectron capture detection. Environ. Sci. Technol. 23(7):852-859.
16 Schwartz, T.R.; Stalling,'D.L; Rice, C.L. (1987) Are-polychlorinated biphenyl residues adequately
17 described by Aroclor mixture equivalents? Isomer-specific principal components analysis of
18 such residues in fish and turtles. Environ. Sci. Tecnnol. 21:72-76.
19 Silberhorn, E.M.; Glauert, H.P.; Robertson, L.W. (1990) Carcinogenicity of polyhalogenated
20 biphenyls: PCBs and PBBs. Crit. Rev. Toxicoi. 20(6):439-496.
21 Sinks, T.; Steele, G.; Smith, A.B.; Watkins, K.; Shults, R.A. (1992) Mortality among workers
22 exposed to polychlorinated biphenyls. Am. J. Epidemic). 136(4):389T398.
23 Steele, G.; Stehr-Green, P.; Welty, E. (1986) Estimates of the biologic half-life of,polychlorinated
24 biphenyls in human serum. New Engl. J. Med. 314(14):926-927.
25 Taylor, P. (1988) The health effects of .polychlorinated biphenyls. Boston: Harvard School of Public
26 Health, unpublished thesis.
27 U.S. Environmental Protection Agency (1980) Water quality criteria documents; availability. Federal
28 Register 45(231 ):79318-79379.^
29 U.S. Environmental^Protection Agency (1986a) Guidelines for carcinogen risk assessment. Federal
30 Register 51 (185):33992-34003.
31 U.S. Environmental Protection Agency (1986b) Guidelines for the health risk assessment of
32 chemical mixtures. Federal Register 51(185):34014-34025.
33 U.S. Environmental Protection Agency (1988a) Drinking water criteria document for polychlorinated
34 biphenyls (PCBs). Cincinnati: U.S. EPA, ECAO-CIN-414.
.i#.*
35 U.S. Environmental Protection Agency (1988b) Methodology for evaluating potential
36 carcinogenicity in support of reportable quantity-adjustments pursuant to CERCLA
37 section 102. Washington, DC. Report No. EPA/600/8-89/053.
DRAFT-DO NOT QUOTE OR CITE 53 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1 U.S. Environmental Protection Agency (1991) Workshop report on toxicity equivalency factors for
2 polychlorinated biphenyl congeners. Risk Assessment Forum, Washington, DC Report No
3 EPA/625/3-91/020.
4 U.S. Environmental Protection Agency (1992a) Guidelines for exposure assessment. Federal
5 Register 57(1041:22888-22938.
6 U.S. Environmental Protection Agency (1992b) Draft report: a cross-species scaling factor for
7 carcinogen risk assessment based on equivalence of mg/kg3/4/day; notice. Federal Register
8 57(109):24152-24173.
9 U.S. Environmental Protection Agency (1994a) Draft revisions to the guidelines for carcinogen risk
10 assessment. Prepared by the Office of Health and Environmental Assessment, Office of
11 Research and Development, Washington, DC. External Review Draft. Report No.
12 EPA/600/BP-92/003.
13 U.S. Environmental Protection Agency (1994b) Report on the workshop on cancer risk assessment
14 guidelines issues. Risk Assessment Forum, Washington, DC. Report No.
15 EPA/630/R-94/005a.
16 U.S. Environmental Protection Agency (1994c) Technical background document to support
17 rulemaking pursuant to the Clean Air Actsection 112(g): ranking of pollutants with
18 respect to hazard to human health. Research Triangle Park, NC. Report No.
19 EPA-450/3-92-010.
20 U.S. Environmental Protection Agency (1994d) Health assessment document for 2,3,7,8-
21 tetrachlorodibenzo-p-dioxin (TCDD) and related compounds. Prepared by the Office of
22 Health and Environmental Assessment, Office of Research and Development, Washington,
23 DC. External Review Draft, 3 vol. Report No. EPA/600/BP-92/001 c. Available from the
24 National Technical Information Service, Springfield, VA; PB 94-205457.
25 Vater, S.T.; Velazquez, S.F.; Cogliano, V.J. (1995) A case study of cancer data set combinations
26 for PCBs. Regul. Toxicol. Pharmacol. 22:2-10.
27 Ward, J.M. (1985) Proliferative lesions of the glandular stomach and liver in F344 rats fed diets
28 containing Arocior 1254. Environ. Health Perspe'ct. 60:89-95.
29 Wester, R.C.; Bucks, D.A.W.; Maibach, H.I.; Anderson, J. (1983) Polychlorinated biphenyls (PCBs):
30 dermal absorption, systemic elimination, and dermal wash efficiency. J. Toxicol. Environ.
31 Health 12:511-519.
32 Wester, R.C.; Mobayen, M.; Maibach, H.I. (1987) In vivo and in vitro absorption and binding to
33 powdered stratum corneum as methods to evaluate skin absorption of environmental
34 chemical contaminants from ground and surface water. J. Toxicol. Environ. Health
35 21:367-374.
36 Wester, R.C.; Maibach, H.I.; Bucks, D.A.W.; McMaster, J.; Mobayen, M. (1990) Percutaneous
37 absorption and skin decontamination of PCBs: in vitro studies with human skin and in vivo
38 studies in the rhesus monkey. J. Toxicol. Environ. Health 31:235-246. -^
39 Wester, R.C.; Maibach, H.I.; Sedik, L; Melendres, J.; Wade, M. (1993) Percutaneous absorption of
40 PCBs from soil: in vivo rhesus .monkey, in vitro hdman skin, and binding to powdered
41 human stratum corneum. J. Toxicol. Environ. Health 39:375-382.
DRAFT-DO NOT QUOTE OR CITE 54 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
s ' , . , .-..'. - .
i 1 ' WHO (World Health Organization) (1993) Polychlorinated biphenyls and terphenyis. Geneva: WHO,
2 Environmental'Health Criteria 140, second ed.
i- J3 , Wolff, M.S. (1985).Occupational exposure to polychlorinated biphenyls (PCBs). Environ Health
4 Perspect. 60:133-138.
DRAFT-DO NOT QUOTE OR CITE 55 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
APPENDIX: DOSE-RESPONSE ASSESSMENT RESULTS
2 Table A-1. Dose-response assessment for female Sherman rats fed Aroclor 1260
3 Tumors Liver carcinomas and adenomas
4 Animal Female Sherman rats
5 Routs Diet
6 Reference Kimbrough et al. (1975), Moore et at. (1994)
7 Exposure duration 21 mo.
8 Study duration 23 mo ' -
9 Animal Hfespan 23 mo (study duration) .
10 Animal weight 0.35 kg (assumed)
11 Administered dose 0 100 ppm
12, Equivalent human dose 0 1.3 mg/kg-d
13 Turner incidence 1/187 138/189
»
14 Model Risk(d) = ,1 - exp<-0.0054-0.50d-0.39cr*) in observed range
15 Table A-2. Dose-response assessment for leukemia and rymphome in mete Fischer 344 rats
16 fed Aroclor 1254
17 Tumors Leukemia and lymphoma
18 Animal Male Fischer 344 rat*
19 Route Diet
20 Reference NCI (1978)
21 Exposure duration 104-105 wk
22 Study duration 104-105 wk
23 Animal Hfespan 104-105 wk (study duration)
24 Animal weight 0.3 kg
25 Administered dose 0 25 50 10O ppm
26 Equivalent human dose 0 0.32 0.64 1.28 mg/kg-d
27 Tumor incidence 3/24 2/24 5/24 9/24
v
28 Model Risk(d) = 1 - exp<-0.11-0.23CT2) in observed range
DRAFT-DO NOT QUOTE OR CITE 56 1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1
2
3
4
5
6
7
8
9
10
if
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Table A-3. Dose-response
Aroclor 1254
Tumors
Animal
Route
. . ' Reference
Exposure duration
Study duratkxi
Animal Irfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
TflhlA A 4 Onftn-fAMvwwA
rats fedArodor 1254
Tumors
Animal
Route
Reference
Exposure duration
Study duration
. Animal Ofaspan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
assessment for liver tumors in male Fischer 344 rats fed
Liver carcinomas and adenomas
Male Fischer 344 rats
Diet
NCI (1978), Moore et al. (1994)
104^105 wk
104-105 wk
104-105 wk {study duration)
0.3kg
0 25 50 100 pom
. 0 0.32 0.64 1.28 mg/kg-d
0/24 1/24 1/24 3/23
Risk(d) = 1 - exp(-O.089o'-0.0058d6) in observed range
assessment f or gastric tumors hi male and female Fischer 344
Gastric adenocarcinomas
Male and female Fischer 344 rats
Diet
NCI (1978), Ward (1985)
104-105 wk
104-105,wk ,
104-105 wk (study duration)
0.25 kg (average of males and females)
0 25 50. 100 pom
0 0.305 0.61 1.22 mg/kg-d
0/47 1/48 3/48 2/48 . ,
Risk(d) = 1 - exp(-0.060d) in observed range
DRAFT-DO NOT QUOTE OR CITE 57
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Table A-5. Dose-response
rats fed Aroclor 1 254
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal Irfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
Table A-6. Dose-response
Aroclor 1254
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal Rfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
assessment for leukemia and lymphoma in female Fischer 344
Leukemia and lymphoma '
Female Fischer 344 rats , , .
Diet ' .
NCI (1978) .
104-105 wk
104-105 wk
104-105 wk (study duration)
0.2kg
0 25 50 100 ppm
0 . 0.29 0.58 1.16 mg/kg-d
4/24 6/24 6/24 6/24
Risk(d) = 1 - exp(-0.22-0.082e/) in observed range
, - - *
assessment for fiver tumors in female Fischer 344 rats fed
Liver carcinomas and adenomas
Female Fischer 344 rats ' .
Diet
NCI (1 978), Moore et al. (1 994)
104-105 wk
104-105 wk
104-105 wk (study duration)
0.2 kg
0 25 50 100 ppm
0 0.29 0.58 1.16 mg/kg-d
0/23 1/24 2/24 1/24
Risk(d) = 1 - exp(-0.084
-------
DRAFT-DO NOT QUOTE OR CITE
Table A-7. Dose-response assessment for male Wistar rats fed Clophen A 30
2
3
4
5
6
7
8
9
10
11
12
13
Tumors
Animal
Route
Reference.
Exposure duration
Study duration
Animal lifespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
Liver carcinomas and adenomas
Male Wistar rats
Diet
Schaeffer et al. (1984), Moore et al. (1994)
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed)
0 100 ppm
0 .1.3 mg/kg-d
8/120 16/128
RiskW) = 1 -exp(-0.069-0.025o'-0.019d2) in observed range
14
15
16
17
18
19
20
21
22
23
24
25
26
Table A-8. Dose-response assessment for male Wistar rats fed Clophen A 60
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal Ufespan
Animal weight
Administered dose
Equivalent human dps*
Tumor incidence
Model
Liver carcinomas and adenomas
Male Wistar rats
Diet
Schaeffer et al. (1984). Moore et al. < 1994)
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed)
0 100 ppm
0 1.3 mg/kg-d
8/120 114/125
RiskW) » 1 - exp(-0.069-0.91«/-0.70flr2) in observed range
Of
DRAFT-DO NOT QUOTE OR CITE 59
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1
2
3
4
5
6
7
S
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Table A-9. Dose-response assessment for male Sprague-Dawley rats fed Aroclor 1260
Tumors
Animal
Routa
Reference
Exposure duration
Study duration
Animal lifespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
Table A-10. Dose-response
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal lifespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
Liver carcinomas and adenomas
Male Sprague-Dawley rats
Diet f ' - ' '
Norback'and Weltman (1985), Moore et al. (1994)
24 mo
29 mo
29 mo (study duration) .
0.35 kg (assumed)
0 100 ppm
0 1 .3 mg/kg-d
0/31 5/40
RisMcO - 1 - exp(-0.051o'-0.040d2) in observed range
assessment for female Sprague-Oawley rats fed Aroclor 1 260
Liver carcinomas and adenomas
Female Sprague-Dawley rats
Diet
Norback and Weltman (1985), Moore et al. (1994)
24 mo
29 mo (
29 mo (study duration)
0.35 kg (assumed)
0 100 ppm
0 1.3 mg/kg-d
1/45 41/46
RiskW) = 1 - exp(-0.022-0.84o'-0.65tf2) in observed range
DRAFT-DO NOT QUOTE OR CITE
60
1/24/96-
-------
DRAFT-DO NOT QUOTE OR CITE
. 1.
2
3
4
5
6
7
8
9
10
11
12
13
14
Table A- 11. Preliminary
Aroclor 1016
Tumors
Animal
Routa ,w
Reference
Exposure duration
Study duration
Animal Ufespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Modal
dose-response assessment for female Sprague-Oawley rats fed
Liver carcinomas and adenomas
Female Sprague-Oawley rats -
Diet ' ......
Unpublished preliminary results submitted to U.S. EPA
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed) "
0 50 100 200 ppm .
0 0.66 1.33 2.66 mg/kg-d
, 1/85 1/48 7/45 6750
Risk(d) = 1 - exp(-0.01 1 -0.059d) in observed range
, , =
15
16
17
18
19
20
21
22
23
24
25
26
27
28
TableA-12. Preliminary
Aroclor 1242
Tumors
Animal
Route
Reference ..
Exposure duration
Study duration
Animal Hfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
dose-response assessment for female Sprague-Oawley rats fed
Liver carcinomas and adenomas
Female Sprague-Oawley rats
Diet , -,*
Unpublished preliminary results submitted to U.S. EPA
24 mo . -
24 mo ,-.
24 mo (study duration)
0.35 kg (assumed)
6 50 10O ppm
0 0^66 1.33 mg/kg-d
1/85 11/49 17/45
RisMcfl = 1 -exp(-0.01 2-0.354} in observed range
DRAFT-DO NOT QUOTE OR CITE 61
/24/96
-------
DRAFT-DO NQT QUOTE OR CITE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Table A-13. Preliminary
Aroclor 1254
Tumors
Animal
Routs
Reference
Exposure duration
Study duration
Animal Irfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
TaWeA-14. Preliminary
Aroclor 1260
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal Irfespan
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
dose-response assessment for female Sprague-Dawley rats fed
Liver carcinomas and adenomas
Female Sprague-Dawley rats
Diet
Unpublished preliminary results submitted to U.S. EPA
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed)
0 25 50 100 ppm
0' 0.33 0.66 1.33 mg/kg-d
1/85 19/45 29/49 29/49
Risk(d) 3 1 - exp(-0.012-1.4d) in observed range
dose-response assessment for female Sprague-Dawley rats fed
Liver carcinomas and adenomas
Female Sprague-Dawley rats
Diet
Unpublished preliminary results submitted to U.S. EPA
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed)
0 25 50 100 ppm
0 0.33 0.66 1.33 mg/kg-d
1/85 10/49 11/45 24/50
Risk(d) = 1 - exp(-0.013-p.49d) in observed range
DRAFT-DO NOT QUOTE OR CITE
62
1/24/96
-------
DRAFT-DO NOT QUOTE OR CITE
1
2
3
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19.
20
21
22
23
24
25
26
27
28
TableA-15. Preliminary
Aroclor 1016
Tumors
Animal '
Route .
Reference
Exposure duration
Study, duration
Animal Irfespan
Animal weight
Administered dose
Equivalent human do**
Tumor incidence
Model
Table A- 16. Preliminary
Aroclor 1242
Tumors
Animal
Rout*
, Reference
Exposure duration
Study duration
Animal Irfespan
Animal weight
Administered do**
Equivalent human do**
Tumor incidence
Model
dose-response assessment for male Sprague-Oawley rats fed
Liver carcinomas and adenomas
Male Sprague-Dawley rats
Diet
Unpublished preliminary results submitted to U.S. EPA
24 mo -
24 mo
' i 24 mo (study duration)
0.35 kg (assumed) ''
0 50 100 200 ppm
0 0.66 1.33 2.66 mg/kg-d
7/98 2/47 2/49 4/49
RisMoV = 1 - exp(-0.058-O.OOOO75
-------
: . 1-
DRAFT-DO NOT QUOTE OR CITE
T,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
TableA-17. Preliminary
Aroclor 1254
Tumors
Animal
Route
Reference1
Exposure duration
Study duration
Animal life span
Animal weight
Administered dose
Equivalent human dose
Tumor incidence
Model
TaWeA-18. Preliminary
Aroctor 1260
Tumors
Animal
Route
Reference
Exposure duration
Study duration
Animal lifespan
Animal weight
Administered dose
Equivalent human oose
Tumor Incidence
Model
dose-response assessment for male Sprague-Dawley rats fed
Liver carcinomas and adenomas
Male Sprague-Dawley rats
Diet
Unpublished preliminary results submitted to U.S. EPA
24 mo
24 mo
24 mo (study duration)
0.35 kg (assumed)
0 25 50 100 ppm ;
0 0.33 0.66 1 .33 mg/kg-d
7/98 4/48 4/49 6/47 .
Risk(d) = 1 - exp(-0.075-0.19o'-0.0066d6) in observed range
dose-response assessment for male Sprague-Dawley rats fed
Liver carcinomas and adenomas . . -
Male Sprague-Dawley rats
Diet
Unpublished preliminary results submitted to U.S. EPA
24 mo ,
24 mo .
24 mo (study duration)
0.35 kg (assumed)
0 25 50 100 ppm
0 0.33 0.66 1 .33 mg/kg-d ,
7/98 3/50 7/48 10/47
Risk(d) = 1 - exp(-0.070-0.035o'-0.076rf2) in observed range
DRAFT-DO NOT QUOTE OR CITE 64
1/24/96
-------
PCBs: Cancer Dose-Response Assessment
and Application to Environmental Mixtures
AGENCY: Environmental Protection Agency
ACTION: Notice of Availability ,
SUMMARY: This notice announces the availability of a report titled, PCBs: Cancer Dose-
Response Assessment and Application to Environmental Mixtures, External Review Draft,
(EPA/600/P-96/001A). The National Center for Environmental Assessment (NCEA) of the
Office of Research and Development developed this report, which is an external draft for review
purposes only and does not constitute U.S. Environmental Protection Agency (EPA) policy. The
report will not have official status or receive clearance as an EPA document until after peer .
review has taken place. The document is being made available at this time because of public
interest in PGBs.
ADDRESSES: The document will be available on the Internet at http://www.epa.gov/docs/ORD
or for purchase from the National Technical Information Service, 5285 Port Royal Road,
Springfield, VA 22161; telephone 703-487-4650; facsimile 703-321-8547. The NTIS order
number is PB96-140603; the price is $19.50 for paper and $9.00 for microfiche. Copies will, be
available for inspection at the EPA libraries. The EPA Headquarters Library is located at 401 M
Street, S.W., Washington, DC; the library is open Monday through Friday between 10:00 a.m. .
-------
-------
PCBs: Cancer Dose-Response Assessment
2bf2
and 2:00 p.m., except for Federal holidays. Unfortunately, due to budget restrictions, printed
copies of the document are not available from the National Center for Environmental Assessment.
FOR FURTHER INFORMATION CONTACT: Dr. Tim Cogliano, National Center for
Environmental Assessment/Washington Office (8602), U.S. Environmental Protection Agency,
401M Street, S.W., Washington DC 20460. Telephone: 202-260-3830;
facsimile: 202-260-3803; E-mail: cogliano.jim @ epamail.epa.gov.
SUPPLEMENTARY INFORMATION: The draft report updates the cancer dose-response
assessment for PCBs and shows how information on toxicity, disposition, and environmental
processes can be considered together to evaluate health risks from PCB mixtures in the
environment. Guidance is given on applying the assessment to environmental mixtures, different
exposure routes, partial lifetime exposure, and mixtures containing dioxin-like compounds. In the
Spring, the Agency will convene an external peer-review panelfor a workshop that will be
announced in the Federal Register. After the peer review workshop, EPA will incorporate the
panel's comments and issue a final report. The expected date for the final report is September 1,
1996. At the same time, a summary of the final report will be loaded onto the Agency's on-line
database, the Integrated Risk Information System (IRIS).
2-
Dated
Deputy Assistant Administrator
For Research and Development
-------
------- |