EPA/635/R-09/004F
www.epa.gov/iris
&EPA
TOXICOLOGICAL REVIEW
OF
PENTACHLOROPHENOL
(CAS No. 87-86-5)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
September 2010
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
TOXICOLOGICAL REVIEW OF PENTACHLOROPHENOL
(CAS No. 87-86-5)
LIST OF TABLES vii
LIST OF FIGURES ix
LIST OF ABBREVIATIONS AND ACRONYMS x
FOREWORD xiii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xiv
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKINETICS 6
3.1. PCP LEVELS IN GENERAL AND OCCUPATIONALLY EXPOSED
POPULATIONS 6
3.2. ABSORPTION, DISTRIBUTION, METABOLISM, AND EXCRETION 7
3.2.1. Oral Studies 7
3.2.1.1. Absorption 7
3.2.1.2. Distribution 9
3.2.1.3. Metabolism 9
3.2.1.4. Excretion 13
3.2.2. Inhalation Studies 16
3.2.3. Dermal Studies 17
3.2.4. Other Studies 17
3.3. PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS 18
4. HAZARD IDENTIFICATION 19
4.1. STUDIES IN HUMANS 19
4.1.1. Studies of Cancer Risk 19
4.1.1.1. Case Reports and Identification of Studies for Evaluation of Cancer Risk 19
4.1.1.2. Cohort Studies 20
4.1.1.3. Case-Control Studies of Specific Cancers and PCP 28
4.1.1.4. General Issues—Interpretation of the Epidemiologic Studies 36
4.1.1.5. Specific Cancers 38
4.1.2. Studies ofNoncancerRisk 40
4.1.2.1. Case Reports of Acute, High-Dose Exposures 40
4.1.2.2. Studies of Clinical Chemistries, Clinical Examinations, and Symptoms 42
4.1.2.3. Studies of Neurological Outcomes 46
4.1.2.4. Studies of Reproductive Outcomes 51
4.1.2.5. Summary of Studies ofNoncancerRisk 53
4.2. SHORT-TERM, SUBCHRONIC, AND CHRONIC STUDIES AND CANCER
BIOASSAYS IN ANIMALS—ORAL AND INHALATION 53
4.2.1. Oral Studies 54
4.2.1.1. Short-term Studies 54
4.2.1.2. Subchronic Studies 59
4.2.1.3. Chronic Studies—Noncancer 68
in
-------
4.2.2. Inhalation Studies 74
4.2.2.1. Subchronic Studies 74
4.2.2.2. Chronic Studies 75
4.2.3. Other Routes of Exposure 75
4.2.4. Cancer Studies 75
4.2.4.1. Oral Studies 75
4.2.4.2. Inhalation Studies 84
4.3. REPRODUCTIVE, ENDOCRINE, AND DEVELOPMENTAL STUDIES 84
4.3.1. Reproductive and Endocrine Studies 84
4.3.2. Developmental Studies 88
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES 94
4.4.1. Oral 94
4.4.1.1 Acute Studies 94
4.4.1.2. Immunotoxicity Studies 94
4.4.1.3. Thyroid Hormone Studies 99
4 A.I A. Endocrine Disruption Studies 101
4.4.1.5. Neurotoxicity Studies 102
4.4.2. Inhalation 104
4.4.2.1. Acute Studies 104
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE
OF ACTION 104
4.5.1. Genetic Toxicity Studies 104
4.5.1.1. In Vitro Studies 105
4.5.1.2. In Vivo Studies 109
4.5.2. DNA Adduct Formation Ill
4.5.2.1. In Vitro Studies Ill
4.5.2.2. In Vivo Studies 112
4.5.3. Protein Adduct Formation 112
4.5.4. Oxidative DNA Damage and 8-Hydroxy-2'-Deoxyguanosine Formation 114
4.5.4.1. In Vitro Studies 114
4.5.4.2. In Vivo Studies 117
4.5.5. Uncoupling of Oxidative Phosphorylation 118
4.5.6. Cytotoxicity 120
4.5.7. Lipid Peroxidation 121
4.5.8. Inhibition of Gap Junction Intercellular Communication 121
4.6. SYNTHESIS AND EVALUATION OF MAJORNONCANCER EFFECTS 123
4.6.1. Oral 123
4.6.2. Inhalation 129
4.6.3. Mode-of-Action Information 130
4.6.4. Comparison of Toxic Effects of aPCP with tPCP or Commercial-Grade PCP 131
4.6.4.1. Short-term and Subchronic Studies 131
4.6.4.2. Chronic Studies 134
4.6.4.3. Developmental Studies 135
4.7. EVALUATION OF CARCINOGENICITY 136
4.7.1. Summary of Overall Weight of Evidence 136
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence 136
4.7.2.1. Human Epidemiologic Evidence 136
4.7.2.2. Animal Cancer Evidence from Oral Exposure 138
4.7.2.3. Animal Cancer Evidence from Inhalation Exposure 140
IV
-------
4.7.3. Mode-of-Action Information 141
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 145
4.8.1. Possible Childhood Susceptibility 145
4.8.1.1. Evidence in Humans 145
4.8.1.2. Evidence of Reproductive/Developmental Toxicity and Teratogenicity
in Animals 146
4.8.1.3. Evidence of Thyroid Hormone Perturbation in Animals 146
4.8.1.4. Other Considerations 147
4.8.1.5. Conclusions Concerning Childhood Susceptibility 148
4.8.2. Possible Gender Differences 148
4.8.3. Other Susceptible Populations 149
5. DOSE-RESPONSE ASSESSMENTS 151
5.1. ORAL REFERENCE DOSE (RfD) 151
5.1.1. Choice of Principal Study and Critical Effect—Rationale and Justification 151
5.1.2. Methods of Analysis 155
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs) 155
5.1.4. RfD Comparison Information 156
5.1.5. Previous RfD Assessment 161
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 161
5.3. UNCERTAINTIES IN THE RfD AND RfC 162
5.4. CANCER ASSESSMENT 163
5.4.1. Choice of Study/Data—Rationale and Justification 163
5.4.2. Dose-Response Data 166
5.4.3. Dose Adjustments and Extrapolation Method(s) 167
5.4.4. Oral Slope Factor and Inhalation Unit Risk 170
5.4.5. Uncertainties in Cancer Risk Values 174
5.4.6. Previous Cancer Assessment 179
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
DOSE RESPONSE 181
6.1. HUMAN HAZARD POTENTIAL 181
6.1.1. Noncancer 181
6.1.2. Cancer 183
6.2. DOSE RESPONSE 184
6.2.1. Noncancer—Oral Exposure 184
6.2.2. Cancer 185
7. REFERENCES 188
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION A-l
APPENDIX B: PHYSIOCHEMICAL DATA FOR PCP AND THE IDENTIFIED
TECHNICAL- AND COMMERCIAL-GRADE CONTAMINANTS B-1
APPENDIX C: PCP LEVELS IN OCCUPATIONALLY EXPOSED HUMANS C-l
v
-------
APPENDIX D: DOSE-RESPONSE MODELING OF CARCINOGENICITY DATA FOR
PENTACHLOROPHENOL D-l
D.I. METHODS D-l
D.2. RESULTS D-3
APPENDIX E: COMBINED ESTIMATES OF CARCINOGENIC RISK E-l
VI
-------
LIST OF TABLES
2-1. Chemical and physical properties of PCP 3
2-2. Impurities and contaminants in different grades of PCP 5
3-1. Summary of some toxicokinetic parameters in rats, monkeys, and humans for orally
administered PCP 16
4-1. Summary of cohort studies of cancer risk and PCP exposure, by specificity of exposure
assessment 21
4-2. Cancer mortality and incidence risk in relation to estimated PCP exposure in sawmill
workers, British Columbia, Canada 26
4-3. Summary of case-control studies of lymphomarisk and PCP exposure 29
4-4. Summary of case-control studies of soft tissue sarcoma risk and PCP exposure 30
4-5. Summary of case-control studies of chlorophenol and soft tissue cancer risk included
in Hardell et al. (1995) meta-analysis 33
4-6. Prevalence of medical conditions and physical symptoms, and associations with PCP
exposure, in 293 timber workers in New Zealand 45
4-7. Prevalence of neuropsychological symptoms, and associations with PCP exposure, in
293 timber workers inNew Zealand 49
4-8. Prevalence of abnormalities seen in neurological examination, and associations with
PCP exposure, in 293 timber workers in New Zealand 50
4-9. Comparison of the effects of three grades of PCP administered continuously in feed
to male (M) and female (F) B6C3Fi mice for 30 days 56
4-10. Summary of effects and NOAELs/LOAELs for short-term studies on PCP 58
4-11. Comparison of the effects of four grades of PCP administered continuously in feed
to male (M) and female (F) B6C3Fi mice for 6 months 60
4-12. Summary of NOAELs/LOAELs for oral subchronic studies for PCP 67
4-13. Liver histopathology, incidence, and severity in dogs exposed to tPCP 70
4-14. Summary of NOAELs/LOAELs for oral chronic studies for PCP 74
4-15. Treatment-related tumors in maleB6C3Fi mice fed tPCP or EC-7 for 2 years 77
4-16. Treatment-related tumors in female B6C3Fi mice fed tPCP or EC-7 for 2 years 78
vn
-------
4-17. Incidences of treatment-related tumors in male F344 rats fed purified PCP for up
to 2 years 81
4-18. Hepatocellular tumors inB6C3Fi mice in initiation/promotion studies 82
4-19. Summary of NOAELs/LOAELs for developmental and reproductive studies
for PCP 93
4-20. Summary of selected in vitro genotoxicity studies of PCP 108
4-21. Summary of selected in vitro genotoxicity studies of metabolites of PCP 109
4-22. Summary of selected in vivo genotoxicity studies of PCP Ill
4-23. Subchronic, chronic, developmental, and reproductive oral toxicity studies for PCP 124
5-1. Candidate PODs for hepatotoxicity with applied UFs and potential reference values 158
5-2. Candidate PODs for reproductive and developmental toxicity in rats with applied
UFs, and potential reference values 161
5-3. Incidence of tumors in B6C3Fi mice exposed to tPCP and EC-7 in the diet for
2 years 167
5-4. Summary of BMD modeling for PCP cancer data in male and female B6C3Fi mice 169
5-5. Summary of BMDLio/HED and cancer slope factors derived from PCP cancer data in
male and female B6C3Fi mice 170
5-6. Human-equivalent combined risk estimates for liver, adrenal, and circulatory tumors
inB6C3Fimice 172
5-7. Summary of uncertainties in the PCP cancer risk assessment 175
B-l. Physicochemical data for dioxin contaminants of PCP B-l
B-2. Physicochemical data for furan contaminants of PCP B-l
B-3. Average daily dose of PCP (mg/kg) and contaminants (ug/kg) to B6C3Fi mice
in the 2-year feeding study B-2
C-l. PCP levels in worker and residential populations (with >15 individuals per group) C-l
C-2. PCP levels in occupationally exposed populations (with >15 individuals per group) C-2
D-l. Incidence of tumors in B6C3Fi mice exposed to tPCP and EC-7 in the diet for
2 years D-l
Vlll
-------
D-2. Summary ofBMD modeling results based onNTP (1989) D-3
E-l. Results of simulation analyses characterizing combined cancer risk estimates for
male and female mice E-3
LIST OF FIGURES
3-1. Proposed PCP metabolism to quinols, benzosemiquinones, and benzoquinones 13
5-1. Array of candidate PODs with applied UFs and reference values for a subset of
hepatotoxic effects of studies in Table 5-1 157
5-2. Array of candidate PODs with applied UFs and reference values for a subset of
reproductive and developmental effects of studies in Table 5-2 160
IX
-------
LIST OF ABBREVIATIONS AND ACRONYMS
y-GTP y-glutamyl transpeptidase
SMC 3-methylcholanthrene
8-OH-dG 8-hydroxy-2'-deoxyguanosine
ADME absorption, distribution, metabolism, and excretion
AHH arylhydrocarbon hydroxylase
ALP alkaline phosphatase
ALT alanine aminotransferase
AML alpha mouse liver
AP apurinic
aPCP analytical-grade PCP
AST aspartate aminotransferase
AUC area under the curve
BMD benchmark dose
BMDL 95% lower bound of the BMD
BMDS Benchmark Dose Software
BMR benchmark response
BrdU bromodeoxyuridine
BRI biological reactive intermediates
BRL Bionetics Research Laboratory, Inc.
BSA bovine serum albumin
BUN blood urea nitrogen
BW3/4 body mass raised to the 3/4 power
CA chromosomal aberration
CASRN Chemical Abstracts Service Registry Number
CHO Chinese hamster ovary
CI confidence interval
Cmax maximum concentration
DEN diethylnitrosamine
DETAPAC diethylenetriamine pentaacetic acid
DMBA dimethylbenzanthracene
DMSO dimethylsulfoxide
DNA deoxyribonucleic acid
DNP-Ficoll 2,4-dinitrophenyl-amincethylcarbamylmethyl-Ficoll
DP-2 Dow PCP DP-2 Antimicrobial
EC-7 Dowicide EC-7
EMCV encephalomyocarditis virus
EMS ethyl methanesulfonate
FSH follicle stimulating hormone
GD gestation day
GJIC gap junction intercellular communication
GLP Good Laboratory Practice
HCB hexachlorobenzene
HED human equivalent dose
HPRT hypoxanthine phosphoribosyltransferase
HRP horseradish peroxidase
HxCDD hexachlorodibenzo-p-dioxin
x
-------
IQ
IRIS
ISF
i.p. interperitoneal(ly)
i.v. intravenous
IARC International Agency for Research on Cancer
ICD International Classification of Disease
median inhibitory dose
intelligence quotient
Integrated Risk Information System
isosafrole
median lethal dose
LDH lactate dehydrogenase
LF lipofuscin
LH luteinizing hormone
LOAEL lowest-observed-adverse-effect level
LPS lipopolysaccharide
MCS multiple chemical sensitivity
MSB MSV-transformed tumor cell
MSV Moloney sarcoma virus
MTD maximum tolerated dose
NHANES National Health and Nutrition Examination Survey
NIOSH National Institute for Occupational Safety and Health
NLM National Library of Medicine
NOAEL no-observed-adverse-effect level
NRC National Research Council
NTP National Toxicology Program
OCDD octachlorodibenzo-p-dioxin
OPP U. S. EPA' s Office of Pesticide Programs
OPPTS Office of Pollution, Prevention and Toxic Substances
OR odds ratio
OuaR ouabain resistance
PB phenobarbital
PBPK physiologically based pharmacokinetic
PCP pentachlorophenol
PFC plaque-forming cell
PNMT phenylethanolamine-N-methyltransferase
POD point of departure
RAL relative adduct level
RBC red blood cell
RED Reregi strati on Eligibility Decision
RfC reference concentration
RfD reference dose
ROS reactive oxygen species
RR relative risk
SAB Science Advisory Board
SCE sister chromatid exchange
SIR standardized incidence ratio
SMR standardized mortality ratio
SOD superoxide dismutase
SRBC sheep red blood cell
SSB single strand break
XI
-------
TS triiodothyronine
T4 thyroxine
TCDD tetrachlorodibenzo-p-dioxin
TCHQ tetrachlorohydroquinone
TCoBQ tetrachloro-o-benzoquinone
TCoHQ tetrachloro-1,2-hydroquinone
TCoSQ tetrachloro-1,2-benzosemiquinone
TCP tetrachlorophenol
TCpBQ tetrachloro-p-benzoquinone
TCpCAT tetrachlorocatechol
TCpSQ tetrachloro-1,4-benzosemiquinone
TGr 6-thioguanine resistance
Tmax time to peak plasma concentration
tPCP technical-grade PCP
TRH thyrotropin-releasing hormone
TSH thyroid-stimulating hormone
UCL upper confidence limit
UDS unscheduled DNA synthesis
UF uncertainty factor
UFA interspecies uncertainty factor
database deficiency uncertainty factor
intraspecies uncertainty factor
UFL LOAEL-to-NOAEL uncertainty factor
UFS subchronic-to-chronic uncertainty factor
U.S. EPA U.S. Environmental Protection Agency
WBC white blood cell
xn
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to
pentachlorophenol. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of pentachlorophenol.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose,
reference concentration and cancer assessment, where applicable, and to characterize the overall
confidence in the quantitative and qualitative aspects of hazard and dose response by addressing
the quality of data and related uncertainties. The discussion is intended to convey the limitations
of the assessment and to aid and guide the risk assessor in the ensuing steps of the risk
assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
xin
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGERS/AUTHORS
Catherine Gibbons, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Samantha J. Jones, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
AUTHORS
Glinda Cooper, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Geoff Patton, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Timothy F. McMahon, Ph.D.
Senior Toxicologist
Antimicrobials Division
Office of Pesticide Programs
U.S. Environmental Protection Agency
Washington, DC
Lynn Flowers, Ph.D., D.A.B.T.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
xiv
-------
Kowetha Davidson, Ph.D., D.A.B.T.
Chemical Hazard Evaluation Group
Toxicology and Risk Analysis Section
Life Sciences Division
Oak Ridge National Laboratory
Oak Ridge, TN
CONTRIBUTING AUTHORS
Karen Hogan, M.S.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Leonid Kopylev, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
REVIEWERS
This document has been provided for review to EPA scientists, interagency reviewers
from other federal agencies and White House offices, and the public, and peer reviewed by
independent scientists external to EPA. A summary and EPA's disposition of the comments
received from the independent external peer reviewers and from the public is included in
Appendix A.
INTERNAL EPA REVIEWERS
TedBerner, M.S.
National Center for Environmental Assessment
Office of Research and Development
Allan Marcus, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
Channa Keshava, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
James N. Rowe, Ph.D.
Office of Science Policy
Office of Research and Development
xv
-------
Linda Birnbaum, Ph.D., D.A.B.T.
Director, NIEHS
EXTERNAL PEER REVIEWERS
Scott M. Bartell, Ph.D.
Program in Public Health
University of California, Irvine
Nasrin Begum, Ph.D.
Tetrahedron, Inc.
Leslie T. Stayner, Ph.D.
Division of Epidemiology and Biostatistics
UIC School of Public Health
Paul B. Tchounwou, Sc.D., FABI, IOM
College of Science, Engineering & Technology
Jackson State University
Gary M. Williams, M.D.
Department of Pathology
New York Medical College
xvi
-------
1. INTRODUCTION
This document presents background information and justification for the Integrated Risk
Information System (IRIS) Summary of the hazard and dose-response assessment of
pentachlorophenol (PCP). IRIS Summaries may include oral reference dose (RfD) and
inhalation reference concentration (RfC) values for chronic and other exposure durations, and a
carcinogenicity assessment.
The RfD and RfC, if derived, provide quantitative information for use in risk assessments
for health effects known or assumed to be produced through a nonlinear (presumed threshold)
mode of action. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
inhalation RfC considers toxic effects for both the respiratory system (portal-of-entry) and for
effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
acute (<24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
lifetime) exposure durations, all of which are derived based on an assumption of continuous
exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
derived for chronic exposure duration.
The carcinogenicity assessment provides information on the carcinogenic hazard
potential of the substance in question and quantitative estimates of risk from oral and inhalation
exposure may be derived. The information includes a weight-of-evidence judgment of the
likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
effects may be expressed. Quantitative risk estimates may be derived from the application of a
low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk is a
plausible upper bound on the estimate of risk per |ig/m3 air breathed.
Development of these hazard identification and dose-response assessments for PCP has
followed the general guidelines for risk assessment as set forth by the National Research Council
(NRC, 1983). U.S. Environmental Protection Agency (U.S. EPA) Guidelines and Risk
Assessment Forum technical panel reports that may have been used in the development of this
assessment include the following: Guidelines for the Health Risk Assessment of Chemical
Mixtures (U.S. EPA, 1986a), Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b),
Recommendations for and Documentation of Biological Values for Use in Risk Assessment (U.S.
EPA, 1988), Guidelines for Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Interim
-------
Policy for Particle Size and Limit Concentration Issues in Inhalation Toxicity (U.S. EPA,
1994a), Methods for Derivation of Inhalation Reference Concentrations and Application of
Inhalation Dosimetry (U.S. EPA, 1994b), Use of the Benchmark Dose Approach in Health Risk
Assessment (U.S. EPA, 1995), Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA,
1996), Guidelines for Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council
Handbook: Risk Characterization (U.S. EPA, 2000a), Benchmark Dose Technical Guidance
Document (U.S. EPA, 2000b), Supplementary Guidance for Conducting Health Risk Assessment
of Chemical Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference
Concentration Processes (U.S. EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S.
EPA, 2005a), Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to
Carcinogens (U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U.S. EPA,
2006a), and A Framework for Assessing Health Risks of Environmental Exposures to Children
(U.S. EPA, 2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Number (CASRN) and at least one common name. Any pertinent
scientific information submitted by the public to the IRIS Submission Desk was also considered
in the development of this document. The relevant literature was reviewed through August 2010.
It should be noted that references have been added to the Toxicological Review after the external
peer review in response to peer reviewers' comments and for the sake of completeness. These
references have not changed the overall qualitative and quantitative conclusions. See Section 7
for a list of references added after peer review.
-------
2. CHEMICAL AND PHYSICAL INFORMATION
PCP (CASRN 87-86-5) is a chlorinated aromatic compound that appears in a solid
crystalline state and ranges in color from colorless to white, tan, or brown. The chemical, also
referred to as penta, pentachlorofenol, 2,3,4,5,6-PCP, and chlorophen, has a phenolic odor that is
pungent when heated. PCP is nonflammable and noncorrosive, and, although solubility is
limited in water, it is readily soluble in alcohol (Budavari et al., 1996; NTP, 1989). The
physical/chemical properties of PCP are summarized in Table 2-1.
Table 2-1. Chemical and physical properties of PCP
Chemical formula
Molecular weight
Melting point
Boiling point
Density
Vapor density
Vapor pressure
Log Kow
Log Koc
Water solubility
Henry's law constant
Conversion factors
C6HOC15
266.34
190-191°C
-309-3 10°C
1.978g/mL(at22°C/4°C)
9.20 (air = 1)
0.00011 (at 20°C)
5.01
4.5
80 mg/L (at 20°C), 14 mg/L (at 26.7°C)
2.45 x 10"8 (atm x m3)/mole
1 ppm = 10.9 mg/m3; 1 mg/m3 = 0.09 ppm; 1 ppm = 0.01088 mg/L;
1 mg/L = 99.1 ppm (at 25°C)
Sources: NLM (1999a, b); Budavari et al. (1996); Allan (1994); Royal Society of Chemistry (1991).
PCP has been used as a wood preservative to prevent decay from fungal organisms and
insect damage since 1936. The first pesticidal product containing PCP as an active ingredient
was registered in the United States in 1950 (U.S. EPA, 2008; Ahlborg and Thunberg, 1980).
Historically, PCP was widely used as a biocide and could also be found in ropes, paints,
adhesives, canvas, leather, insulation, and brick walls (U.S. EPA, 2008; Proudfoot, 2003;
ATSDR, 2001). Indoor applications of PCP were prohibited in 1984; PCP application was
limited to industrial areas (e.g., utility poles, cross arms, railroad cross ties, wooden pilings,
fence posts, and lumber/timbers for construction). Currently, products containing PCP remain
registered for heavy duty wood preservation, predominantly to treat utility poles and cross arms.
Pentachlorophenol is a restricted use pesticide available to certified applicators only (U.S. EPA,
2008).
PCP is produced via two pathways, either "by stepwise chlorination of phenols in the
presence of catalysts (anhydrous aluminum chloride or ferric chloride) or alkaline hydrolysis of
[hexachlorobenzene] HCB" (Proudfoot, 2003). In addition to industrial production of PCP, the
-------
degradation or metabolism of HCB (Rizzardini and Smith, 1982), pentachlorobenzene (Kohli et
al., 1976), or pentachloronitrobenzene (Renner and Hopfer, 1990) also yields PCP. Impurities
found in PCP are created during the production of the chemical. Technical-grade PCP (tPCP),
frequently found under the trade names Dowicide 7, Dowicide EC-7 (EC-7), Dow PCP DP-2
Antimicrobial (DP-2), Duratox, Fungol, Penta-Kil, and Permacide, is composed of
approximately 90% PCP and 10% contaminants. The impurities consist of several chlorophenol
congeners, chlorinated dibenzo-p-dioxins, and chlorinated dibenzofurans. Of the chlorinated
dibenzo-p-dioxin and dibenzofuran contaminants, the higher chlorinated congeners are
predominantly found as impurities within tPCP. In addition to the chlorinated dibenzo-p-dioxin
and dibenzofuran contaminants, HCB and chlorophenoxy constituents may also be present in
tPCP. Use of the analytical grade of PCP (aPCP) first requires a purification process to remove
the contaminants that were created during the manufacturing of PCP. The physicochemical
properties of these contaminants are listed in Appendix B in Tables B-l and B-2.
Grades described as analytical or pure are generally >98% PCP and the levels of dioxins
and furans are low to nondetectable. Purities of technical- and commercial-grade PCP
formulations are reported to be somewhat less than the analytical formulations, ranging from
85 to 91%. Hughes et al. (1985) reported that tPCP contains 85-90% PCP, 10-15%
trichlorophenol and tetrachlorophenol (TCP), and <1% chlorinated dibenzo-p-dioxins,
chlorinated dibenzofurans, and chlorinated diphenyl ethers. The compositions of different
grades of PCP, as reported by the National Toxicology Program (NTP) (and similar to values
reported in the general literature), are listed in Table 2-2.
-------
Table 2-2. Impurities and contaminants in different grades of PCP
Contaminant/impurity
PCP
Pure/analytical
98.6%
tPCP
90.4%
DP-2a
91.6%
EC-7a
91%
Chlorophenols
Dichlorophenol
Trichlorophenol
TCP
HCB
-
<0.01%
1.4%
lOppm
-
0.01%
3.8%
50 ppm
0.13%
0.044%
7.0%
15 ppm
-
0.007%
9.4%
65 ppm
Dioxins
Tetrachlorodibenzo-p-dioxin (TCDD)
Pentachlorodibenzo-p-dioxin
Hexachlorodibenzo-p-dioxin (HxCDD)
Heptachlorodibenzo-p-dioxin
Octachlorodibenzo-p-dioxin (OCDD)
<0.08 ppm
-
<1 ppm
-
<1 ppm
-
-
10.1 ppm
296 ppm
1,386 ppm
-
-
0.59 ppm
28 ppm
173 ppm
O.04 ppm
-
0.19 ppm
0.53 ppm
0.69 ppm
Ethers
Pentachlorodibenzofuran
Hexachlorodibenzofuran
Heptachlorodibenzofuran
Octachlorodibenzofuran
Hexachlorohydroxydibenzofuran
Heptachlorohydroxydibenzofuran
Chlorohydroxydiphenyl ethers
-
-
-
-
0.11%
0.22%
0.31%
1.4 ppm
9.9 ppm
88 ppm
43 ppm
0.16%
0.47%
5.58%
-
12.95 ppm
172 ppm
320 ppm
0.07%
0.31%
3.67%
-
0.13 ppm
0.15 ppm
-
-
-
-
"The DP-2 and EC-7 commercial formulations are no longer manufactured and are listed for informational purposes
only.
Source: NTP(1989).
-------
3. TOXICOKINETICS
The toxicokinetics of PCP have been studied in both humans and animals. These studies
show that PCP is rapidly and efficiently absorbed from the gastrointestinal and respiratory tracts
(Reigner et al., 1992a, b, c). Once absorbed, PCP exhibits a small volume of distribution.
Metabolism occurs primarily in the liver, to a limited extent, via oxidative dechlorination and
conjugation. Tetrachlorohydroquinone (TCHQ) and the conjugation product, PCP-glucuronide,
have been confirmed as the two major degradation products. PCP is predominantly excreted
unchanged and found in the urine in the form of the parent compound. The low degree of
metabolism is frequently attributed to extensive plasma protein binding.
3.1. PCP LEVELS IN GENERAL AND OCCUPATIONALLY EXPOSED
POPULATIONS
Several reports have provided data on levels of PCP in blood or urine samples in humans
(general population samples or groups with known exposures to PCP) indicating that PCP is
absorbed in humans. The correlation between blood and urinary values is relatively high when
the urinary data are corrected for creatinine clearance (0.92 in Cline et al. [1989] and 0.76 in
Jones et al. [1986]). Studies from Hawaii (Klemmer, 1972; Bevenue et al., 1967) and the United
Kingdom (Jones et al., 1986) have demonstrated blood (plasma or serum) and urine values of
PCP in workers with high PCP exposures (e.g., pesticide operators, wood treaters, and other
wood workers) that are approximately an order of magnitude higher than in nonexposed groups
within the same study.
People who lived or worked in buildings in which PCP-treated wood was used have been
found to have mean serum levels up to 10 times higher than groups that were not exposed
(Gerhard et al., 1999; Peper et al., 1999; Cline et al., 1989). Similar patterns were seen in the
urinary data. Sex differences were not noted for the PCP serum levels in log home residents, but
age differences were observed. Children (2-15 years old) had serum PCP levels 1.7-2.0 times
higher than those of their parents. Cline et al. (1989) attributed the higher PCP levels in children
to differences in the ventilation rate to body weight ratio, although Treble and Thompson (1996)
reported no age-related differences in urinary PCP concentrations in 69 participants ages 6-
87 years (mean 54.6 years) living in rural and urban regions of Saskatchewan, Canada. See
tables in Appendix C for further details on occupationally exposed humans.
Renner and Miicke (1986), in reviewing the metabolism of PCP, noted that establishing a
direct relationship between PCP exposure levels and PCP in body fluids may be difficult because
PCP is a metabolite of other environmental contaminants (e.g., HCB, pentachlorobenzene,
pentachloronitrobenzene) and is itself metabolized.
-------
Casarett et al. (1969) reported mean 10-day urine concentrations of 5.6 and 3.2 ppm in
two groups of workers handling PCP under different conditions. The mean decrease in urine
concentration in workers following different periods of absence from their jobs was 39% within
the first 24 hours and 60-82% over the next 17 days. Continued excretion of PCP was noted
after 18 days of absence from the job. A semilog plot shows a linear relationship between
plasma and urine concentrations at plasma concentrations of 0.1 ppm and a plateau for plasma
concentrations >10 ppm.
In another experiment by Casarett et al. (1969), air concentrations, blood levels, and
urinary excretion of PCP were measured 2 days before and 5 days after a 45-minute exposure to
PCP. Mean air concentrations of 230 and 432 ng/L (calculated doses were 90.6 and 146.9 jig,
respectively) were associated with 88 and 76% excretion of PCP in the urine, respectively.
Excretion was slow during the first 24 hours (half-life = 40-50 hours) and more rapid after the
first day (half-life =10 hours). In one subject, urine concentrations returned to baseline after
48 hours, but remained elevated in the other subject.
Begley et al. (1977) reported blood and urine PCP levels in 18 PCP-exposed workers
before, during, and after a 20-day absence from their jobs. Except for a brief rise on
postexposure day 6, blood PCP levels during a 20-day absence showed a steady decline to 50%
of the level measured on the last day of work (i.e., exposure). There was a 6-day lag in the
decrease in urine level; after day 20, urine levels had decreased about 50%. Begley et al. (1977)
also noted that the high PCP levels were accompanied by impaired renal function measured by
creatinine and phosphorus clearance and phosphorus reabsorption.
Ahlborg et al. (1974) detected PCP, as well as the metabolites TCHQ and
tetrachloropyrocatechol, in the urine of workers occupationally exposed to PCP. They did not
quantify the levels of metabolites in urine.
3.2. ABSORPTION, DISTRIBUTION, METABOLISM, AND ELIMINATION
3.2.1. Oral Studies
3.2.1.1. Absorption
Braun et al. (1979) orally dosed four male human subjects with 0.1 mg/kg unlabeled PCP
(ingested in 25 mL of water). The absorption half-life for the volunteers was 1.3 hours, with a
maximum plasma concentration (Cmax) of 0.245 ug/mL and a time to peak plasma concentration
(Tmax) of 4 hours. In another study, Braun et al. (1977) reported that the absorption rate
constants for PCP administered in corn oil to Sprague-Dawley rats were 1.95 and 1.52 hour"1 for
males and females, respectively. The plasma Tmax was 4-6 hours.
Larsen et al. (1975) observed that PCP levels (measured as percentage of administered
dose of [14C]PCP [99.54% radiochemical purity] and/or its metabolites per g of tissue) peaked in
maternal blood serum 8 hours after dosing 14 Charles River CD (Sprague-Dawley derived) rat
dams with 60 mg/kg on gestation day (GD) 15 (administered in a solution of olive oil; 100 mg/
-------
6 mL). The serum levels, peaking at approximately 1.13% [14C]PCP per g of blood serum,
steadily dropped during the remaining part of the 32-hour monitoring period for a final
measurement of 0.45% [14C]PCP per g of blood serum. [14C]PCP in the placenta peaked at
0.28% of administered dose 12 hours after dosing. The level reaching the fetus peaked at 0.08%
of the administered dose of [14C]PCP and remained extremely low throughout the monitoring
period. The levels of [14C]PCP per g of tissue measured in the placenta and fetus were much
lower than those levels found in the maternal blood serum.
Reigner et al. (1991) studied toxicokinetic parameters in 10 male Sprague-Dawley rats
administrated 2.5 mg/kg of aPCP (99% purity) via intravenous (i.v.) or gavage (five animals/
route) routes. Absorption was rapid and complete, with 91% bioavailability after oral
administration. Plasma levels peaked at 7.3 ug/mL after 1.5-2 hours and declined with a half-
life of 7.5 hours. Reigner et al. (1992c) examined the pharmacokinetics of orally administered
PCP (15 mg/kg) in male B6C3Fi mice. The data were consistent with an open one-compartment
model. Absorption followed first-order kinetics. Peak plasma concentration (28 ug/mL) was
achieved at 1.5 hours. Absorption was complete; bioavailability was measured as 106%.
Yuan et al. (1994) studied the toxicokinetics of PCP (>99% purity) administered to F344
male rats by gavage (n = 18) at doses of 9.5 or 38 mg/kg, or dosed feed (n = 42) containing
302 or 1,010 ppm PCP (21 or 64 mg/kg-day, respectively) for 1 week. In addition, groups of
18 male and 18 female rats were administered PCP at a dose of 5 mg/kg by i.v. injection.
Following gavage administration, the absorption half-life of 1.3 hours and plasma concentrations
that peaked in approximately 2-4 hours indicated very rapid absorption from the gut. For the
dosed feed study, absorption was also rapid and followed first-order kinetics. Plasma
concentrations showed repeated cycles of peaks and troughs, coinciding with feeding cycles (i.e.,
highest concentrations at night and lowest during the day); however, plasma concentration did
not reach pretreatment levels during the day. Absorption from the gut was estimated as 52 and
30% for administered doses of 21 and 64 mg/kg-day (302 and 1,010 ppm), respectively. The
bioavailability was much lower than the values obtained from the gavage study. The
investigators noted that the lower bioavailability for the dosed feed study suggests that PCP
interacts with components in feed. The data from the i.v. study were fitted to a two-compartment
model. The investigators stated that absorption and elimination half-lives were not affected by
the change from gavage to dosed feed administration.
Braun and Sauerhoff (1976) orally administered a single 10 mg/kg dose of [14C]PCP to
Rhesus monkeys in 10 mL of corn oil solution. The absorption kinetics of [14C]PCP were first
order with the absorption half-life ranging from 1.8 to 3.7 hours. Deichmann et al. (1942)
reported that absorption was immediate and rapid in rabbits given a single 18 mg/kg oral dose of
PCP (in feed), and peak blood levels were achieved 7 hours after dosing rabbits with 37 mg/kg
PCP (in feed). Deichmann et al. (1942) administered 90 successive (except Sundays) oral doses
of 0.1% PCP sodium salt (equivalent to 3 mg/kg) to 23 rabbits (sex not reported) in feed.
-------
Average peak blood concentrations of 0.6 mg PCP/100 mL blood were measured within 4 days
and did not change much for the remaining duration of the study. The investigators noted that
the blood concentrations of PCP were similar to those attained after 100 daily skin applications
of 100 mg each (0.45 mg PCP/100 mL of blood).
3.2.1.2. Distribution
Binding of PCP to specific components of liver cells or differential distribution of PCP to
different cellular organelles may affect its metabolic fate. Arrhenius et al. (1977a) administered
a 40 mg/kg dose of aPCP by gavage to rats; the animals were sacrificed 16 hours later. The
relative concentration of PCP in microsomes was 6 times greater than in mitochondria. PCP acts
as an inhibitor of mitochondrial oxidative phosphorylation (Weinbach, 1954) and has been
shown to inhibit the transport of electrons between a flavin and cytochrome P450, thereby
interrupting the detoxification enzyme system (Arrhenius et al., 1977a, b). Arrhenius et al.
(1977a) suggested that inhibition of microsomal detoxification and inhibition of mitochondrial
oxidative phosphorylation might be equally important.
Binding to plasma proteins plays a significant role in the distribution of PCP that likely
affects the amount available for metabolism and clearance. Uhl et al. (1986) found that >96% of
PCP was bound to plasma proteins in blood samples of three human males receiving an oral dose
of 0.016 mg/kg PCP (dissolved in 40% ethanol). Gomez-Catalan et al. (1991) found 97 ± 2% of
the administered dose of PCP (10-20 mg/kg in water and corn oil via gavage) bound to plasma
proteins in rats. Braun et al. (1977) examined tissues of rats orally administered PCP (in corn
oil) and showed the greatest accumulation of PCP in the liver and kidneys, with minimal levels
in the brain and fat. The study demonstrated that plasma protein binding accounted for
approximately 99% of the PCP. The authors noted that tissue/plasma ratios and renal clearance
rates following oral administration of PCP were much lower than would be predicted based on
the octanol/water coefficient and the glomerular filtration rate and suggested that the plasma
protein binding resulted in low renal clearance and tissue accumulation.
3.2.1.3. Metabolism
Studies in animals and humans indicate that PCP is metabolized primarily in the liver.
However, PCP is not extensively metabolized; a large portion of the administered dose is
excreted unchanged in the urine. The major metabolic pathways are oxidative dechlorination to
form TCHQ and conjugation with glucuronide. Extensive plasma protein binding occurs that
may account, at least in part, for the low degree of metabolism.
Braun et al. (1979) measured 86% of the administered dose of PCP (0.1 mg/kg; ingested
in 25 mL of water) in the urine and 4% in feces of four human males 8 days after ingestion of
PCP. The study reported that human male subjects excreted 74 and 2% of the administered dose
in urine and feces, respectively, as unmetabolized PCP. PCP, as the conjugated glucuronide, was
-------
measured as 12 and 2% of the administered dose in urine and feces, respectively. TCHQ was not
identified.
Ahlborg et al. (1974) detected PCP, as well as the metabolites TCHQ and
tetrachloropyrocatechol, in the urine of workers occupationally exposed to PCP. They did not
quantify the levels of metabolites in urine. Uhl et al. (1986) found that PCP-glucuronide
conjugate accounted for about 28% of the PCP in the urine of human males on day 1 and about
60% from days 15 to 38 after dosing with 0.31 mg/kg PCP (dissolved in 40% ethanol). The
percentage of PCP-glucuronide conjugate measured in this study is similar to reported levels in
urine of nonoccupationally exposed people. Although previous studies found urinary
metabolites TCHQ and TCP in humans, and TCHQ in animals (Kalman, 1984; Edgerton et al.,
1979; Ahlborg et al., 1974), the authors noted that the data showed no traces of these metabolites
of PCP.
Mehmood et al. (1996) studied the metabolism of PCP (purity not reported) in
microsomal fractions and whole cells of Saccharomyces cerevisiae expressing human CYP3A4.
PCP was transformed to TCHQ, although, in contrast to expected results, further hydroxylations
were not observed. In transformed animals in which CYP3A4 was lacking, metabolism of PCP
was not detected. In humans, this enzyme has low activity in the first month of life, but
approaches adult levels by 6-12 months of age. Adult activity may be exceeded between 1 and
4 years of age, although activity usually declines to adult levels at the end of puberty. Functional
activity of CYP3A7 in the fetus is approximately 30-75% of adult levels (Leeder and Kearns,
1997). aPCP (>99%) was identified as an inducer of CYP3A7 in studies in cultured rat
hepatocytes, quail hepatocytes, and human hepatoma (Hep G2) cells (Dubois et al., 1996).
Juhl et al. (1985) studied the metabolism of PCP in human S9 liver fractions from biopsy
patients and compared the results with those obtained from S9 liver preparations from
noninduced and Aroclor 1254-induced male Wistar rats. Human S9 fractions converted PCP to
TCHQ. Maximum conversion occurred after incubation for 3 hours, after which the level of
TCHQ steadily declined to nondetectable levels at 24 hours. The authors attributed the decline
to the oxidation capacity of the liver preparation or the further oxidation of TCHQ to
semiquinone radicals. The patterns of conversion of PCP to TCHQ in human and rat liver S9
preparations showed very little difference. Juhl et al. (1985) and the more recent study by
Mehmood et al. (1996) report the formation of the TCHQ metabolite of PCP in human liver
tissue and are supportive of the earlier findings of Kalman (1984), Edgerton et al. (1979), and
Ahlborg etal. (1974).
Braun et al. (1977) administered 10 or 100 mg/kg [14C]PCP (in corn oil) to rats. After
administration of a 10 mg/kg dose, approximately 80% of the dose was excreted in urine and
about 19% was excreted in feces of both male and female rats. After administration of
100 mg/kg, males excreted 72% of the administered dose in urine and 24% in feces (which is
similar to the excretion measured in male and female rats administered 10 mg/kg), whereas
10
-------
100 mg/kg females excreted 54% in urine and 43% in feces. The reason for the difference in
excretion in the females administered the higher dose of PCP is unknown; however, the decrease
in the amount of PCP excreted in urine is likely reflected in the increase in the amount of PCP
excreted in the feces, relative to that observed in the males at 100 mg/kg and male and female
rats at 10 mg/kg. Expired air accounted for a small amount of the administered dose.
Unmetabolized PCP accounted for 48% of the administered dose in urine; TCHQ and
PCP-glucuronide conjugate accounted for 10 and 6%, respectively.
PCP metabolites were measured in urine and feces from male Wistar rats administered
8 mg/kg-day PCP by gavage for 19 days (Engst et al., 1976). Under these conditions, most of
the PCP in urine was unmetabolized; small amounts of 2,3,4,5-TCP, 2,3,4,6-TCP and/or
2,3,5,6-TCP, and 2,3,4-trichlorophenol were found. No metabolites and only a small amount of
unmetabolized PCP were identified in feces.
Van Ommen et al. (1986a) studied the in vitro metabolism of PCP (100 uM) utilizing rat
liver microsomal preparations from untreated male and female Wistar rats and from rats treated
with HCB, phenobarbital (PB), 3-methylcholanthrene (3MC), or isosafrole (ISF). Rat liver
microsomes converted PCP only to TCHQ and tetrachloro-l,2-hydroquinone (TCoHQ) via
cytochrome P450 enzymes. The conversion rate (pmol total soluble metabolite formed per mg
protein per minute) increased sevenfold in rat microsomes induced with ISF and three- to
fourfold in HCB-induced rats. PB and 3MC increased the conversion rate two- to threefold over
the controls. The ratios of TCHQ/TCoHQ production were 4.9:1 for male rats and 1.6:1 for
female rats receiving no inducer. The ratio decreased in rats treated with the enzyme inducers in
the following order: HCB > PB > 3MC ~ ISF. The sex difference observed in untreated rats
was not observed in rats treated with the inducers, although there was no change in the
conversion rate in female rats (as opposed to male rats) treated with PB.
Van Ommen et al. (1986b) found that PCP binds to microsomal proteins. Protein binding
was dependent on metabolism, and the amount bound did not vary considerably with the
microsomal preparations (63-75 pmol/mg protein-minute) except for that obtained from
PB-induced female rats (104 pmol/mg protein-minute). Van Ommen et al. (1986b) indicated
that the "benzoquinone or the semiquinone form" of TCHQ and TCoHQ "is responsible for the
covalent binding properties." Protein binding was inhibited by glutathione through conjugation
with benzoquinone. When the covalent binding was inhibited through reduction of
benzoquinones and semiquinones to the hydroquinone form by ascorbic acid, the formation of
TCHQ and TCoHQ increased. Deoxyribonucleic acid (DNA) binding also occurred, but to a
lesser degree than protein binding. Covalent binding to DNA was 12 ± 3 pmol/mg DNA-minute,
while the average microsomal protein binding was 63 pmol/mg protein-minute. The Km value
for covalent binding to protein and conversion to hydroquinone was 13 uM, and the authors
suggested that these activities resulted from the same reaction (Van Ommen et al., 1986a).
11
-------
Tsai et al. (2001) attempted to analyze two proposed pathways of PCP (purity not
reported) metabolism. Additionally, the authors were interested in illustrating any differences in
metabolism between rats and mice that may explain the varied tumor patterns observed in the
two species of rodents (NTP, 1999, 1989). One potential metabolism pathway involves
cytochrome P450-mediated dechlorination of PCP to the quinols TCHQ and tetrachlorocatechol
(TCpCAT), which are oxidized to the respective benzoquinones and semiquinones in both
Sprague-Dawley rats and B6C3Fi mice. Alternatively, PCP is oxidized via peroxidase to
tetrachloro-p-benzoquinone (TCpBQ) by a direct P450/peroxidase-mediated oxidative pathway.
The formation of tetrachloro-o-benzoquinone (TCoBQ) via the latter pathway has not been
verified.
Tsai et al. (2001) found that liver cytosol and cumene hydroperoxide in either the
presence or absence of microsomes activated PCP and resulted in a greater production of
PCP-derived adducts (quinones or semiquinones) than when PCP was activated with microsomes
and NADPH. The investigators demonstrated that induction of microsomes, via SMC or PB, led
to PCP metabolism resulting in the formation of TCpBQ in both rats and mice. Increased
metabolism to the adduct-forming benzoquinones following induction by SMC and PB was
observed in both rats and mice, although the mice exhibited an increase in BQ adduct formation
that was significantly greater than that in rats. Other adducts measured, such as TCpBQ, did not
exhibit an induction greater than the controls. Results of this study as well as others (Mehmood
et al., 1996; Van Ommen et al., 1986a) indicate that various isozymes of P450 are responsible for
metabolism of PCP. The authors "speculate that the increased SMC-related induction of specific
P450 isozymes in mice (eightfold increase versus control) compared with rats (2.4-fold increase
versus control), may have played a role in the formation of liver tumors in mice (but not rats)
dosed with PCP."
Lin et al. (2002) proposed a metabolism pathway for PCP (Figure 3-1) that, similar to
Tsai et al. (2001) and Van Ommen et al. (1986a, b), involved oxidative dechlorination of PCP to
benzoquinones via the corresponding semiquinones (also referred to as benzosemiquinones).
The authors reported metabolites of PCP as TCHQ and TCpCAT. Both of these metabolites are
thought to undergo oxidation to tetrachloro-l,4-benzosemiquinone (TCpSQ) and tetrachloro-
1,2-benzosemiquinone (TCoSQ). The semiquinones subsequently undergo further oxidation to
form the corresponding TCpBQ and TCoBQ.
12
-------
OH
O2
H202
•Cl
OH
Cl
Oxidative DNA
Damage
OH
NAD(P)
NAD(P)H
H2O2
NAD(P)
NAD(P)H
Direct DNA
Adducts
•ci
TCpBQ
Source: Recreated from Lin et al. (2002).
Figure 3-1. Proposed PCP metabolism to quinols, benzosemiquinones, and
benzoquinones.
3.2.1.4. Excretion
Uhl et al. (1986) measured elimination half-lives of 18-20 days in urine and 16 days in
blood in human males orally administered 0.055, 0.061, 0.15, or 0.31 mg/kg PCP (dissolved in
13
-------
40% ethanol). Urinary clearance was 1.25 mL/minute for free (unconjugated) PCP, while
clearance for total PCP (free PCP and conjugated PCP-glucuronide) was shown to be very slow,
only 0.07 mL/minute. Considering that >96% of the administered PCP was bound to plasma
proteins in blood measurements, the authors suggested that bound PCP resulted in a relatively
long elimination half-life and slow clearance.
Braun et al. (1979) reported elimination half-lives of 30 and 33 hours for plasma
elimination and urinary excretion, respectively, in four human male subjects orally administered
0.1 mg/kg PCP (in 25 mL of water). Elimination was consistent with a first-order, one-
compartment pharmacokinetic model. While plasma concentration peaked at 4 hours, peak
urinary excretion occurred 42 hours after dosing; the delay in time was attributed to entero-
hepatic recirculation of PCP.
Braun et al. (1977) described a two-compartment open system model in rats administered
PCP in corn oil, where the PCP elimination half-life of the rapid phase was 13-17 hours for both
doses, while the slower phase was 33-40 hours at the 10 mg/kg dose and 121 hours for the
100 mg/kg dose in males. Females, however, did not show biphasic elimination at the 100 mg/
kg dose; the rapid phase accounted for >90% elimination of the dose.
Larsen et al. (1972) reported that <0.04% of a 59 mg/kg oral dose of [14C]PCP (99.5%
purity; dissolved in olive oil) administered to male and female rats (strain not reported) was
eliminated in expired air as 14CC>2 within 24 hours. After administration of 37-41 mg/kg,
females excreted 41% of the radioactivity in urine within 16 hours, 50% within 24 hours, 65%
within 72 hours, and 68% within 10 days. Fecal excretion accounted for 9.2-13.2% of the
administered dose. Excretion showed a biphasic pattern: a rapid excretion phase during the first
24 hours and a slower phase thereafter.
Ahlborg et al. (1974) reported that NMRI mice and Sprague-Dawley rats excreted <50%
of radioactivity in urine during the first 96 hours after oral administration of 25 mg/kg [14C]PCP
(dissolved in olive oil), with about twice as much appearing in the urine of rats compared with
mice. About 70% of the radioactivity appeared in the urine after interperitoneal (i.p.) injection of
25 mg/kg. The radioactivity in the urine of mice and rats was 41 and 43% PCP and 24 and 5%
TCHQ, respectively. Another metabolite, TCpCAT, made up 35% of the radioactivity in urine
in the mouse and 52% in the rat. Because TCHQ inhibited p-glucuronidase activity, the degree
of glucuronide conjugation could not be determined. However, boiling the urine with
hydrochloric acid to release free metabolites from conjugates converted the entire radioactivity to
PCP and TCHQ, with a nearly identical distribution of radioactivity between these metabolites
(54 and 57% PCP and 46 and 43% TCHQ, respectively, in mice and rats).
Reigner et al. (1991) investigated PCP elimination in male Sprague-Dawley rats given
2.5 mg/kg PCP by either i.v. or gavage administration. The study authors reported biphasic
plasma elimination with half-lives of 0.7 and 7.1 hours with i.v. administration. The data were
fitted with an open two-compartment model. The areas under the curve (AUCs) were similar for
14
-------
i.v. and oral administration (96 and 94 ug-hours/mL, respectively). Total excretion was 68 and
62% and total urinary excretion was 58 and 52% of the PCP doses for i.v. and gavage
administration, respectively. Total urinary TCHQ excretion was 31 and 27% of the PCP dose for
i.v. and gavage administration, respectively. These data are similar in recovery to other studies
in male rats (Braun et al., 1977), and in rats and mice (Ahlborg et al., 1974). However, the
plasma elimination after oral administration (in corn oil) observed in male rats by Braun et al.
(1977), while also following a biphasic pattern, showed much longer half-lives than those
obtained by gavage administration in Reigner et al. (1991). Reigner et al. (1992c) reported that
the elimination half-life in male B6C3Fi mice was 5.8 hours. An analysis of metabolites
revealed that only 8% of the administered PCP was excreted as parent compound. Yuan et al.
(1994) noted sex differences in F344 rats with regard to elimination half-life (5.6 hours for males
and 9.5 hours for females) and volume of distribution (0.13 L/kg for males and 0.19 L/kg for
females). Bioavailability estimated from the AUC for i.v. injection and gavage administration
was 100% at 9.5 mg/kg and 86% at 38 mg/kg PCP.
Rozman et al. (1982) demonstrated a significant effect of biliary excretion on disposition
of orally administered PCP. Three male Rhesus monkeys equipped with a bile duct bypass were
administered 50 mg/kg of [14C]PCP by stomach intubation. During the first 24 hours, 21% of the
administered dose was excreted into urine, 0.3% into feces, and 19% into bile. From day 2 to 7
after dosing, 35% of the administered dose was excreted into urine, 3% into feces, and 70% into
bile. The monkeys received a second dose of 50 mg/kg [14C]PCP, followed 24 hours later by 4%
cholestyramine (binds phenols) in the diet for 6 days. Cumulative excretion of PCP into urine
and bile was reduced to 5 and 52%, respectively, of the administered dose, whereas cumulative
excretion into feces was increased to 54% of the dose. The data suggest that enterohepatic
recirculation of PCP plays a major role in urinary excretion of the compound. In Rhesus
monkeys administered a single 10 mg/kg dose of [14C]PCP, the plasma elimination half-lives
ranged from 72 to 84 hours, and the urinary excretion half-lives were 41 hours for males and
92 hours for females (Braun and Sauerhoff, 1976). Urinary and fecal excretion accounted for
69-78 and 12-24% of the administered dose, respectively. Unlike humans and rats, all of the
PCP eliminated in the urine of monkeys was unchanged parent compound (Braun and Sauerhoff,
1976). The Rozman et al. (1982) data are not directly comparable with those obtained by Braun
and Sauerhoff (1976) because of the bile duct bypass; however, a relative correlation with the
excretion pattern is indicated.
Deichmann et al. (1942) administered 0.1% PCP sodium salt (equivalent to 3 mg/kg; in
feed) to rabbits repeatedly for 90 successive (except Sundays) doses and about 92% of the dose
was recovered in urine, feces, and tissues combined (-71% in urine and feces) within the first
24 hours, and elimination from the blood was almost complete within 4 days after dosing. The
largest fractional tissue dose was recovered from muscle, bone, and skin; however, 0.7-2% of
the dose was recovered in the liver. Deichmann et al. (1942) also showed that rabbits orally
15
-------
administered 25 and 50 mg/kg PCP sodium salt (in feed) excreted 64-70 and 49-56% of the dose
in urine and feces, respectively, within 7 and 12 days.
The absorption and elimination half-lives and the maximum plasma concentrations for
orally administered PCP in rats, mice, and monkeys are summarized in Table 3-1. Human data
from Braun et al. (1979) are also included for comparison. The kinetics of orally administered
PCP, for all of the species studied, are consistent with a one- or two-compartment open model
exhibiting first-order kinetics. Based on the available data, the toxicokinetics of PCP in humans
may be more similar to those of rats and mice than Rhesus monkeys.
Table 3-1. Summary of some toxicokinetic parameters in rats, monkeys, and
humans for orally administered PCP
Species
Human
Rhesus
monkey
Rat
Rat
Mouse
Absorption
half-life (hrs)
1.3
3.6 (males)
1.8 (females)
1.3
0.6
Plasma
Tmax (hrs)
4
12-24
4-6
2-4
1.5
Elimination
half-life (hrs)
30-33 (males)
72 (males)
84 (females)
Fast: 13 (females)-17 (males)
Slow: 33 (females)-40 (males)
5.6 (males)
9.5 (females)
5.8
Process description
First-order, one-
compartment
One-compartment,
open
Two-compartment,
open
First-order, one-
compartment
First-order, one-
compartment, open
Reference
Braun etal. (1979)
Braun and Sauerhoff
(1976)
Braun etal. (1977)
Yuan etal. (1994)
Reigneretal. (1992c)
3.2.2. Inhalation Studies
PCP inhaled by rats showed rapid uptake from the respiratory tract and excretion from
the body. Hoben et al. (1976a) exposed Sprague-Dawley rats to PCP aerosols at a dose of
5.7 mg/kg for 20 minutes and measured PCP at 0, 6, 12, 24, 48, and 72 hours after exposure.
Between 70 and 75% of the PCP could be accounted for as unmetabolized PCP within the first
24 hours; the highest levels were found in urine > liver = plasma > lungs. PCP in lung and liver
showed a steady decrease throughout the study; plasma levels showed a steady decrease after a
peak at 6 hours; and urine showed a steady decrease after 24 hours. The estimated half-life was
24 hours, and there was no evidence of accumulation or tissue binding.
Rats exposed to PCP aerosols repeatedly for 20 minutes/day for 5 days showed only a
slight net increase in lung and plasma levels immediately after the second exposure with no net
increase in liver levels (Hoben et al., 1976a). Twenty-four hours after each exposure, lung, liver,
and plasma levels were lower but urine levels increased, suggesting that increased urinary
excretion may explain the lack of accumulation of body burden upon repeated exposures.
However, the study authors noted that increased urinary excretion did not account entirely for the
lack of accumulation; they also concluded that metabolism was likely involved.
16
-------
3.2.3. Dermal Studies
Bevenue et al. (1967) reported a case in which a man immersed his hands for 10 minutes
in a solution containing PCP (0.4%). The initial urinary concentration measured 2 days after the
incident was 236 ppb. The urinary level declined to 34% of the initial concentration by day 3,
20% after 2 weeks, 27% after 3 weeks, 10% after 1 month, and 7% after 2 months. This report
shows that PCP is rapidly absorbed through the skin. Elimination was rapid during the first
4 days and proceeded more slowly thereafter. Because elimination is initially rapid, the
concentration of PCP in urine was likely much higher during the first 24 hours after exposure
than after 2 days.
Wester et al. (1993) reported on the absorption of PCP through the skin of female Rhesus
monkeys. PCP-contaminated soil (17 ppm [14C]PCP) or [14C]PCP in acetone) was applied
9
topically at a concentration of 0.7 or 0.8 ug/cm of skin, respectively, for 24 hours. The
percutaneous absorption levels were determined by comparing the urinary excretion levels of
[14C]PCP following either topical or i.v. administration. The measured percent dose peaked on
day 1 for topical and on day 2 for i.v. application, and exhibited a steady decline for
approximately 7 days followed by relatively level daily excretion rates. Over the 14-day
collection period, 45, 11, and 13% of the applied dose was excreted in the urine following i.v.,
topical-soil, and topical-acetone applications, respectively. Percutaneous absorption was similar
for both vehicles with 24 and 29% of the applied dose recovered for soil and acetone,
respectively. The [14C] half-life for excretion was 4.5 days after i.v. administration. Similarly,
the topical administration of PCP, either in soil or acetone, also indicated [14C] half-lives of
4.5 days. The efficient absorption of PCP from skin is indicative of high bioavailability. Similar
to that observed in humans by Bevenue et al. (1967), the relatively long half-life of PCP
observed in the dermal application increases the potential for biological interaction.
3.2.4. Other Studies
Jakobson and Yllner (1971) exposed mice to 1 or 0.5 mg [14C]PCP via i.p. injection. The
investigators reported the greatest amount of PCP distributed in the mice was found in the liver,
intestines, and stomach. Lesser amounts of the dose were found in the heart, kidney, and brain.
Within 96 hours after injection, 72-83% of the dose was excreted in urine and 3.8-7.8% was
excreted in feces; the remainder of the dose was found in specific organs and the carcass. Rapid
absorption and excretion of PCP was exhibited by the appearance of 45-60% of the dose in urine
within the first 24 hours. The authors found that approximately 30% of the PCP measured in the
urine of mice administered 1 or 0.5 mg [14C]PCP was unmetabolized, 7-9% was bound but
released by acid treatment, and 15-26% was the metabolite TCHQ.
17
-------
3.3. PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS
No physiologically based pharmacokinetic (PBPK) models for the oral or inhalation
routes of exposure in humans or animals are available.
18
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS
This section reviews the available evidence of health effects in humans resulting from
exposure to PCP, focusing on carcinogenicity, acute toxicity, and neurological, developmental,
and reproductive effects of chronic exposures.
4.1.1. Studies of Cancer Risk
4.1.1.1. Case Reports and Identification of Studies for Evaluation of Cancer Risk
Significant production of PCP began in the 1930s. The earliest report of cancer was
about 40 years later when Jirasek et al. (1976 [in German]) examined the condition of 80 factory
workers. In addition to porphyria and other serious conditions, two workers died of
bronchogenic carcinoma, which the authors attributed to contamination from 2,3,7,8-tetra-
chlorodibenzo-p-dioxin (TCDD). Other case reports published around this time described non-
Hodgkin's lymphoma among PCP manufacturing workers (Bishop and Jones, 1981) and
Hodgkin's disease in employees of a fence installation company who experienced high exposure
to PCP through the application of the wood preserving solution (Greene et al., 1978).
Several epidemiologic studies conducted in the 1970s and 1980s examined cancer risk in
relation to broad occupational groups (e.g., wood workers, agricultural and forestry workers)
(Pearce et al., 1985; Greene et al., 1978; Brinton et al., 1977). Some subsequent studies focused
on specific workplaces and jobs with known exposures to PCP (e.g., PCP manufacturing plants,
sawmills in which industrial hygiene assessments had been made). Other studies were conducted
in general population samples and used exposure assessments that attempted to distinguish
specific exposures, which sometimes included PCP, within broad occupational groups (e.g.,
specific farming-related activities or exposures).
Studies with PCP-specific data are described in the subsequent section. Some studies
provide data pertaining to exposure to chlorophenols. These studies were included in this
summary when specific information was presented in the report pertaining to PCP (for example,
results for specific jobs that would be likely to have used PCP, rather than other chlorophenols).
Studies that presented data only for a combined exposure (e.g., chlorophenols, or chlorophenols
and phenoxy herbicides) are not included (Richardson et al., 2008; 't Mannetje et al., 2005;
Mirabelli et al., 2000; Garabedian et al., 1999; Hooiveld et al., 1998; Hoppin et al., 1998;
Kogevinas et al., 1997; Ott et al., 1987; Mikoczy et al., 1996; Johnson et al., 1990). A cohort
study of sawmill workers in Finland and a study of cancer incidence in the area surrounding a
mill were identified but not included (Lampi et al., 1992; Jappinen et al., 1989) because the
chlorophenol exposure was primarily to TCP, with PCP representing <10% of the chlorophenol
exposure. Two papers describing studies of surveys of exposed workers contained some
19
-------
information pertaining to cancer mortality. Cheng et al. (1993) examined a small, relatively
young cohort (n = 144) from a PCP manufacturing plant in China, where a total of three deaths
occurred during follow-up, and Gilbert et al. (1990) examined mortality rates in 125 wood
workers in Hawaii, where a total of six deaths occurred. The mortality data in these studies were
very limited (cohort size <200; lack of information pertaining to follow-up and other
methodologic details; limited comparison data, particularly with respect to cancer-specific
mortality); therefore, these studies are not included in this section.
The studies summarized in this review include 3 cohort studies of workers occupationally
exposed to PCP (plywood mill workers, PCP manufacturing workers, and sawmill workers); 12
case-control studies (4 of which were summarized in a meta-analysis) of lymphoma, soft tissue
sarcoma, or multiple myeloma. When two papers on the same cohort were available, the results
from the longer period of follow-up are presented in the summary. Information from earlier
reports is used when these reports contained more details regarding working conditions, study
design, and exposure assessment. The study setting, methods (including exposure assessment
techniques), results pertaining to incidence or mortality from specific cancers, and a brief
summary of primary strengths and limitations are provided for each selected study. The limited
data pertaining to liver cancer are presented because the liver is a primary site seen in the mouse
studies (NTP, 1989). Other data emphasized in this summary relate to lymphatic and
hematopoietic cancers and soft tissue sarcoma, because of the quantity of data and interest in this
area. The description of individual studies is followed by a summary of the evidence available
from all studies reviewed relating to specific types of cancer.
4.1.1.2. Cohort Studies
Three cohort studies of workers exposed to PCP have been conducted, and in two of
these, a PCP-specific exposure measure was developed and used in the analysis (Table 4-1).
20
-------
Table 4-1. Summary of cohort studies of cancer risk and PCP exposure, by
specificity of exposure assessment
Reference, cohort,
location
Sample size,
duration of
work, and
follow-up
Inclusion
criteria
Exposure
assessment
Outcome
assessment
Results— PCP risk"
PCP, specific exposure
Ramlow et al.
(1996), Dow
manufacturing
plant, United States
(Michigan)
Demers et al.
(2006)
Hertzman et al.
(1997)
Heacock et al.
(2000), sawmill
workers, Canada
(British Columbia)
n = 770 men
mean duration:
not reported
mean follow-
up: 26.1 yrs
n =
23,829 men
mean duration:
9.8 yrs
mean follow-
up: 24.5 yrs
Worked
sometime
between 1937
and 1980 in a
relevant
department
Worked at least
1 yr (or 260 d
total) between
1950 and 1985
Work history (job
records) and
industrial hygiene
assessment;
developed exposure
intensity and
cumulative exposure
scores for PCP and
dioxinsb
Work history (job
records) and
industrial hygiene
assessment;
developed
cumulative exposure
scores for PCP and
TCP
Death
certificate
(underlying
cause)
Death
certificate
(underlying
cause);
cancer registry
(incidence)
Elevated risk of
lymphatic cancer
mortality, particularly
at higher intensity
exposures (relative risk
[RR] 2.58, 95%
confidence interval [CI]
0.98-6.8)c; similar
associations seen with
measures of other
dioxins
Elevated risk of non-
Hodgkin's lymphoma
and multiple myeloma
incidence and
mortality; evidence of
exposure -effect
response; weaker or no
risk seen with TCP (see
Table 4-2); no
increased risk of
childhood cancer in
offspring of workers
PCP, nonspecific exposure
Robinson et al.
(1987), plywood
mill workers,
United States
(Pacific Northwest)
n = 2,283 men
mean duration:
not reported
mean follow-
up: 25.2 yrs
Worked at least
1 yr between
1945 and 1955
Work history (job
records); subgroup
analysis of
818 workers known
to have worked in
areas with PCP or
formaldehyde
exposure
Death
certificate
(underlying
cause)
Elevated risk of
lymphatic and
hematopoietic cancer
mortality (standardized
mortality ratio [SMR] =
1.56, 95% CI 0.90-
2.52); stronger when
considering latency and
duration, and when
limited to subgroup
with PCP or
formaldehyde exposure
"Results are described as "elevated" if the SMR was >1.5, regardless of the precision of the estimate or power of the
statistical test; more detailed information on the results is presented in the text.
b2,3,7,8-TCDD and the hexachlorinated to octachlorinated dioxin ratio.
Tor the category of "other and unspecified lymphopoietic cancers" (now classified as multiple myeloma and non-
Hodgkin's lymphoma).
Ramlow et al. (1996) examined the mortality risk in a cohort of 770 male workers at a
large U.S. chemical manufacturing plant (Dow Chemical Company, Michigan Division) that
manufactured PCP from the late 1930s to 1980. This cohort was a subset of a larger cohort of
21
-------
workers in departments with potential for exposure to tPCP. Exposure to dioxins, primarily
hexa-, hepta-, and octa-chlorinated dibenzodioxins and dibenzofurans also occurred within this
cohort (Ott et al., 1987). Men who were employed at the Michigan plant between 1937 and 1980
were included in the study. Follow-up time was calculated through 1989. The mean durations of
work or exposure were not reported, although the mean duration of follow-up was 26.1 years.
The potential for exposure to PCP was assessed by evaluating available industrial
hygiene data, including some quantitative environmental and personal breathing zone PCP
measurements in conjunction with detailed employment records with information on job title and
location. Potential exposures for each job held by cohort members were assigned an estimated
exposure intensity score on a scale of 1 (low) to 3 (high). An estimated cumulative exposure
index was calculated for each subject by multiplying duration for each job by the estimated
exposure intensity for the job and summing across jobs. The cumulative exposure scores were
<1 for 338 (44%), 1-2.9 for 169 (22%), 3-4.9 for 74 (10%), 5-9.9 for 83 (11%), and >10 for
106 (14%) of the workers. A similar process was used to estimate cumulative exposure to
2,3,7,8-TCDD and the hexachlorinated to octachlorinated dioxin ratio. Standardized mortality
ratios (SMRs) were calculated comparing age- and period-specific mortality rates in the cohort
and the U.S. white male population. The cumulative exposure metric was used with an internal
reference group, allowing for examination of exposure-response in analyses estimating relative
risk (RR) controlling for age, period of employment, and general employment status (hourly
versus salaried).
Mortality risk for all causes of cancer was not elevated (SMR 0.95, 95% confidence
interval [CI] 0.71-1.25), and there were no reported cases of mortality due to liver cancer, soft
tissue sarcoma, or Hodgkin's disease. The SMR was 2.31 (95% CI 0.48-6.7) for kidney cancer
(International Classification of Disease [ICDJ-S* revision codes 189; three cases), with the
highest risk seen in the high-exposure group (defined as cumulative exposure >10; RR 4.16
[95% CI 1.43-12.09; trends-value 0.03]). An elevated kidney cancer mortality risk was also
seen with increased dioxin measures in this cohort (for TCDD, trends-value = 0.04; for
hexachlorinated to octachlorinated dioxin ratio, trends-value = 0.02). The SMR for all
lymphopoietic cancers (ICD-8* revision codes 200-209; seven cases) was 1.4 (95% CI 0.56-
2.88). This latter observation was driven by the results for the "other and unspecified
lymphopoietic cancers" (ICD-8* revision codes 200, 202-203, 209; five cases), with an SMR of
2.0 (95% CI 0.65-4.7). Two of these cases were multiple myeloma, and three would now be
classified as non-Hodgkin's lymphoma. Similar results were seen in analyses using a 15-year
latency period. In the exposure-response analysis, the RR in the high-exposure group (defined as
cumulative exposure >1) compared with the no-exposure group was 1.91 (95% CI 0.86-4.24,
trends-value 0.23) for all lymphopoietic cancers and 2.58 (95% CI 0.98-6.8, trends-value 0.08)
for other and unspecified lymphopoietic cancers. There was some indication of an increased risk
22
-------
of lymphopoietic cancer with the other dioxin measures, primarily seen in the "very low" or
"low" exposure groups.
The exposure assessment methodology, allowing for the analysis of PCP and various
forms of dioxins exposure, is the primary strength of this study. The cumulative exposure metric
used in the analysis was based on work duration data in conjunction with a semi quantitative
intensity score for specific jobs; the semiquantitative nature of this measure presents challenges
to its use in dose-response modeling for risk assessment. It is a relatively small cohort, however,
resulting in limited power to assess associations with relatively rare cancers, including the
various forms of lymphomas, soft tissue sarcoma, and liver cancer. Other limitations of this
study are its use of mortality data, rather than incidence data, and the difficulty in separating the
effects of exposures to different dioxins that occurred as part of the production process.
Hertzman et al. (1997) conducted a large cohort study of male sawmill workers from
14 mills in Canada (British Columbia), and this study was updated by Demers et al. (2006).
Sodium salts of PCP and TCP were used as fungicides in 11 of these mills from 1950 to 1990.
Workers from the mills that did not use the fungicides (n = 2,658 in Hertzman et al., 1997;
sample size not specified in Demers et al., 2006) were included in the unexposed group in the
exposure-response analyses. The updated study includes 26,487 men who had worked at least
1 year (or 260 days total) between 1950 and 1995. Record linkage through the provincial and
national death files and cancer incidence registries were used to assess mortality (from first
employment through 1995) and cancer incidence (from 1969, when the provincial cancer registry
began, through 1995) (Demers et al., 2006). The mean duration of work in the mills was not
given in the 2006 update by Demers et al. (2006), but in the earlier report of outcomes through
1989 (Hertzman et al., 1997), the mean duration of employment was 9.8 years and the mean
duration of follow-up was 24.5 years. Approximately 4% of the cohort was lost to follow-up,
and these individuals were censored at date of last employment.
Plant records were available to determine work histories for study cohort members,
including duration of work within different job titles. Representative exposures were determined
for three or four time periods for each mill. Historical exposure measurements had not been
made, so a retrospective exposure assessment was developed based on interviews with senior
workers (>5 years of experience) at each mill (9-20 workers for each time period; mean of
15 years of experience). This process was compared, for current exposures, to urinary
measurements, with correlation coefficients of 0.76 and 0.72 in two different sampling periods
(summer and fall) (Hertzman et al., 1988). Because only one sample was collected in each
period, day-to-day variation in job activities, and thus exposures, would not be captured by the
urine measure; the authors indicate that additional samples would likely result in increased
correlation coefficients. The validity of this method was also demonstrated in comparison with a
method based on an industrial hygienist assessment (Teschke et al., 1996, 1989).
23
-------
Information from the senior workers was used to develop a cumulative dermal
chlorophenol exposure score, calculated for each worker by summing, across all jobs, the
product of the job title specific exposure score and the length of employment in that job. One
exposure year was defined as 2,000 hours of dermal contact. Records from each mill were used
to determine the specific chlorophenol content of the fungicides used at specific time periods. In
general, TCP was increasingly used in place of PCP after 1965. This information was used to
develop PCP- and TCP-specific exposures scores. The correlation between the estimated PCP
and TCP exposures was 0.45 (Demers et al., 2006).
Soft tissue sarcoma is difficult to ascertain accurately without review of the available
histological information. Demers et al. (2006) did not include an analysis of soft tissue cancer
mortality risk (which would have had to rely only on death certificate classification data). The
authors based the analysis of incident soft tissue sarcoma on cancer registry data pertaining to
site (connective tissue) and histology.
SMR and standardized incidence ratios (SIRs) were calculated using reference rates
based on data for the province of British Columbia. Analyses using the quantitative exposure
measure used workers in the cohort with <1 exposure-year as the internal referent group. All
analyses were adjusted for age, calendar period, and race.
In the analyses using the external referent group (i.e., population rates), there was no
increased risk with respect to cancer-related mortality (SMR 1.00, 95% CI 0.95-1.05) or
incidences of all cancers (SIR 0.99, 95% CI 0.95-1.04) in the cohort of sawmill workers. For
liver cancer, the SMR was 0.98 (95% CI 0.62-1.49) and the SIR was 0.79 (95% CI 0.49-1.21);
corresponding estimates for kidney cancer were 1.31 (95% CI 0.98-1.73) for the SMR and 1.10
(95% CI 0.88-1.38) for the SIR. The SMR for non-Hodgkin's lymphoma was 1.02 (95% CI
0.75-1.34) and the SIR was 0.99 (95% CI 0.81-1.21). For multiple myeloma, the SMR was 0.94
(95% CI 0.60-1.41) and the SIR was 0.80 (95% CI 0.52-1.18). The SMR for brain cancer was
0.99 (95% CI 0.73-1.31) and the SIR was 1.08 (95% CI 0.80-1.43).
In the analyses of PCP exposure, there was evidence of an exposure effect for non-
Hodgkin's lymphoma and multiple myeloma in the mortality and in the incidence analyses
(Table 4-2). The risk of non-Hodgkin's lymphoma in relation to TCP was similar to or
somewhat smaller than for PCP, and no association was seen between TCP exposure and
multiple myeloma. The number of incident cases of soft tissue sarcoma was small (n = 23), and
lower risks of this cancer were seen in the higher exposure groups for PCP and for TCP. There
was some evidence of an increased risk of kidney cancer incidence or mortality for PCP and TCP
exposures (Table 4-2). Liver cancer, a relatively rare cancer, was associated with PCP exposure,
but the sparseness of data did not allow assessment at the highest exposure level (>5 exposure
years). Consideration of a 10- or 20-year exposure lag period had little effect on the risks seen
with respect to PCP exposure and risk of non-Hodgkin's lymphoma, multiple myeloma, and
kidney cancer incidence. The 20-year lag resulted in a reduction in the number of liver cancer
24
-------
cases in the exposed categories from 18 to 2; and thus, the pattern of increased risk was no
longer seen. Friesen et al. (2007) examined these data using different models and exposure
metrics, and using the best-fitting lagging period as seen in the Demers et al. (2006) analysis.
The results of Friesen et al. (2007) study indicate that for non-Hodgkin's lymphoma and kidney
cancer, the PCP risk was stronger than that seen for TCP or total chlorophenols.
25
-------
Table 4-2. Cancer mortality and incidence risk in relation to estimated PCP exposure in sawmill workers, British
Columbia, Canada"
Cancer
Non-
Hodgkin's
lymphoma
Multiple
myeloma
Soft tissue
sarcoma0
Kidney
Liver
Exposure-
yrs
<1
1-2
2-5
5+
(trendb)
<1
1-2
2-5
5+
(trendb)
<1
1-2
2-5
5+
(trendb)
<1
1-2
2-5
5+
(trendb)
<1
1-2
2-5
5+
(trendb)
PCP exposure
Mortality
Obs
15
6
18
10
4
5
4
10
15
6
17
12
4
5
8
5
RR
1.0
1.21
2.44
1.77
1.0
3.30
1.58
4.80
1.0
1.33
2.59
2.30
1.0
3.46
3.72
2.53
95% CI
(referent)
0.46-3.2
1.2-5.1
0.75^.2
(0.03)
(referent)
0.87-12.5
0.38-6.6
1.4-16.5
(0.03)
(referent)
0.51-3.5
1.22-5.5
1.00-5.3
(0.02)
(referent)
0.91-13.2
1.04-13.3
0.61-10.4
(0.10)
Incidence
Obs
38
13
24
17
6
4
4
11
18
3
2
0
32
9
22
16
3
4
12
2
RR
1.0
1.33
1.88
1.71
1.0
2.09
1.30
4.18
1.0
0.64
0.18
1.0
1.03
1.79
1.66
1.0
4.09
8.47
1.41
95% CI
(referent)
0.70-2.5
1.1-3.3
0.91-3.2
(0.06)
(referent)
0.57-7.6
0.34-5.0
1.4-12.9
(0.02)
(referent)
0.18-2.2
0.04-0.85
(0.11)
(referent)
0.49-2.2
0.99-3.2
0.85-3.2
(0.07)
(referent)
0.89-18.8
2.2-32.4
0.21-9.2
(0.18)
TCP exposure
Mortality
Obs
29
5
13
2
15
0
4
4
25
5
14
6
4
8
RR
1.0
0.93
1.96
0.63
1.0
0.00
0.94
1.84
1.0
0.94
2.09
1.87
1.0
0.95
95% CI
(referent)
0.36-2.43
0.99-3.89
0.15-2.69
(0.44)
(referent)
0.31-2.91
0.59-5.78
(0.55)
(referent)
0.36-2.46
1.07-4.08
0.75-4.67
(0.04)
(referent)
0.38-2.4
Incidence
Obs
50
11
20
11
15
1
5
4
16
3
4
0
47
6
14
12
11
7
3
0
RR
1.0
0.91
1.34
1.54
1.0
0.27
1.06
1.80
1.0
0.77
0.66
1.0
0.55
1.01
1.80
1.0
2.65
0.52
95% CI
(referent)
0.47-1.75
0.80-2.26
0.79-2.99
(0.14)
(referent)
0.04-2.04
0.38-2.94
0.58-5.60
(0.48)
(referent)
0.23-2.66
0.22-1.99
(0.43)
(referent)
0.23-1.28
0.56-1.84
0.94-3.43
(0.31)
(referent)
1.03-6.85
0.14-1.88
(0.58)
aObs = number of observed cases. Analyses based on Poisson regression using the lowest exposure group as the referent group, adjusting for age and time period.
bTrend/>-value.
°The authors used histology data for the classification of soft tissue sarcoma; therefore, mortality data (from death certificates, without detailed histology information)
were not analyzed for this disease.
Source: Demers et al. (2006).
26
-------
Heacock et al. (2000) examined risk of childhood cancer among the offspring of the male
workers in the British Columbia sawmill workers cohort. (An additional study by Dimich-Ward
et al. [1996], based on this cohort, of pregnancy outcomes, including prematurity, stillbirths, and
congenital anomalies, is discussed in Section 4.1.2.4, Studies of Reproductive Outcomes.)
Marriage and birth records were linked to identify 19,675 children born to these fathers between
1952 and 1988. Forty incident childhood cancers were identified within these children (with
follow-up through age 19 years) through the linking of these birth records to the provincial
cancer registry. Eleven of the cancers were leukemias, nine were brain cancers, and four were
lymphomas. The incidence rates were similar to those expected based on sex, age, and calendar
year standardized rates, with SIRs of 1.0 (95% CI 0.7-1.4) for all cancers, 1.0 (95% CI 0.5-1.8)
for leukemia, and 1.3 (95% CI 0.6-2.5) for brain cancer.
The large size and long follow-up period are important strengths of the British Columbia
sawmill cohort studies (Demers et al., 2006; Heacock et al., 2000; Hertzman et al., 1997), but
even with this size, there is limited statistical power to estimate precise associations with
relatively rare cancers such as liver cancer and soft tissue sarcoma. Other strengths of the study
include the detailed exposure assessment (for PCP and TCP), completeness of follow-up, and
analysis of cancer incidence (through the coverage of the population-based cancer registry) in
addition to mortality. The observed associations are not likely to be explained by confounding:
common behaviors, such as smoking and use of alcohol, have not been associated with the types
of cancers that were associated with PCP exposure in this study (non-Hodgkin's lymphoma,
multiple myeloma); the use of an internal comparison group for the analyses using the exposure
measures reduces the likelihood of potential confounders affecting the results; and the difference
in the patterns with respect to cancer risks seen between PCP and TCP and between PCP and
dioxins also argues against a role of other occupational exposures or contaminants of PCP as an
explanation for the observed associations. (See Section 4.1.1.4, General Issues—Interpretation
of the Epidemiologic Studies, for additional discussion of this issue.) No information is
provided, however, about the effect of adjustment for TCP exposure on the PCP results. Since
the correlation between the two measures is relatively low (r = 0.45), and for many of the cancers
of interest, the PCP associations are stronger than those seen with TCP, it is unlikely that this
adjustment would greatly attenuate the observed associations with PCP. Additional analyses by
the study authors could address this issue, although the relatively small number of observed
cases for specific cancers of interest is likely to be a limitation of this kind of analysis.
Robinson et al. (1987) examined mortality in a cohort of 2,283 male plywood mill
workers employed at four softwood plywood mills in Washington and Oregon (Table 4-1).
Protein glues were used to join the veneer plies, and PCP was often added to the glues as a mold
preventative. PCP was also added to oils used as mold release agents during finishing of the
plywood panels. Other exposures in the various jobs at the mills included wood dust, wood
volatiles, formaldehyde, and carbon disulfide. One subgroup analysis was conducted of workers
27
-------
(n = 818) who had worked in areas with PCP or formaldehyde exposures. There was no
increased risk of mortality for all sites of cancer (SMR 0.70). Data pertaining to cancer of the
liver were not reported. The SMR was 1.56 (95% CI 0.90-2.52) for lymphatic and
hematopoietic cancers (ICD-7th edition codes, ICD, 200-203, 205; based on 12 cases) and
0.86 (95% CI not reported) for leukemia (ICD code 204, based on 5 cases). For lymphatic and
hematopoietic cancers, this increased risk was stronger when using a latency period of 20 years
(SMR 1.95) and when the analysis was limited to duration of employment of >20 years
(SMR 2.50). The risk of lymphopoietic cancer was also stronger in the subgroup of workers
designated as exposed to PCP or formaldehyde (SMR 2.50 [95% CI 0.61-6.46] for lymphatic
cancer and 3.33 [95% CI 0.59-10.5] for Hodgkin's lymphoma). A major limitation of this study
is that there is no analysis specifically focused on PCP exposure. The co-exposure with
formaldehyde could result in a biased effect of PCP exposure on risk of lymphopoietic cancers
due to confounding if formaldehyde was related both to the disease and to PCP exposure in this
work setting.
4.1.1.3. Case-Control Studies of Specific Cancers and PCP
Six case-control studies have reported data pertaining to PCP exposure in relation to risk
of lymphoma (Table 4-3). Three of these studies also included analyses of risk of soft tissue
sarcoma, and five additional case-control studies of soft tissue sarcoma (four of which were
summarized in the meta-analysis by Hardell et al. [1995]) are also available (Table 4-4). Case-
control studies of multiple myeloma (Pearce et al., 1986a) and of childhood and young adult
cancers (Ali et al., 2004) are also included in this summary.
28
-------
Table 4-3. Summary of case-control studies of lymphoma" risk and PCP exposure
Reference, location,
demographic data, diagnosis yrs
Cases (n, source),
controls (n, source)
Source of exposure
data
Results
Detailed PCP assessment
Kogevinas et al. (1995), Europeb
Hardell etal. (1994, 1981), Sweden,
men, age 25-85 yrs, 1974-1978
Hardell and Eriksson (1999), Sweden,
men, age >25 yrs, 1987-1990
32 cases (death certificates for all countries;
cancer registries for seven countries),
158 controls (nested case-control study within
cohort study of exposed workers'3)
105 cases (hospital records) (62% deceased);
355 population controls (matched by vital
status)
442 cases (cancer registry) (43% deceased)
741 population controls (matched by vital
status)
Company records and
industrial hygienist
review
Serf-administered
questionnaire with
follow-up phone
interview if needed0
Serf-administered
questionnaire with
follow-up phone
interview if needed0
PCPs: OR 2.75 (95% CI 0.45-17.0)
High PCPs: OR 4. 19 (95% CI 0.59-29.6)
High (>1 wk continuously or 1 mo total) exposure to
PCPs: OR 8.8 (95% CI 3.4-24)
PCPs: OR 1.2 (95% CI 0.7-1.8)
Limited PCP assessment
Pearce et al. (1986b), New Zealand,
men, age <70 yrs, 1977-1981
Woods et al. (1987), United States
(Washington), men, age 20-79 yrs,
1983-1985
Smith and Christophers (1992),
Australia, men, age >30 yrs, 1976-
1980
83 cases (cancer registry) (% deceased not
specified)
168 cancer controls (% deceased not specified)
and 228 population controls
576 cases (cancer registry) (30% deceased)
694 population controls (32% deceased)
52 cases (cancer registry),
52 cancer controls and 52 population controls
Deceased cases and controls excluded
Structured interview0
Structured interview0
Structured interview
Chlorophenols: OR 1.3 (95% CI 0.6-2.7)
Fencing work: OR 2.0 (95% CI 1.3-3.01)
Chlorophenols: OR 0.99 (95% CI 0.8-1.2)
Increased risk (OR >1.5) for wood preservers and
Chlorophenols manufacturers but not for lumber grader
(OR 0.94)
Chlorophenols: OR 1.4 (95% CI 0.3-6.1)
Four cases and four controls (one population and three
cancer controls) had definite PCP exposure
aNon-Hodgkin's lymphoma except for Smith and Christophers (1992), which includes non-Hodgkin's and Hodgkin's.
bTwenty cohorts from 10 countries workers; total n = 13,898; workers exposed to phenoxy herbicides or Chlorophenols. The follow-up period varied among the cohorts:
follow-up began between 1942 and 1973 and ended between 1987 and 1992 (Kogevinas et al., 1997).
°Proxies included for deceased cases and controls.
OR = odds ratio
29
-------
Table 4-4. Summary of case-control studies of soft tissue sarcoma risk and PCP exposure
Reference, location,
demographics, diagnosis yrs
Cases (n, source),
controls (n, source)
Source of exposure
data
Results
Detailed PCP assessment
Kogevinas et al. (1995), Europe3
Hardell et al. (1995) meta-analysis of
four studies'3, Sweden, men, ages 25-
80 yrs, 1970-1983
1 1 cases (death certificates for all countries;
cancer registries for seven countries), 55
controls (nested case-control study within
cohort study of exposed workers)
434 cases (hospital records; cancer registry),
948 population controls
Company records and
industrial hygienist
review
Serf-administered
questionnaire with
follow-up phone
interview if needed0
PCPs: no exposed cases or controls
High (>1 wk continuously or 1 mo total) exposure to
PCPs: OR 2.8 (95% CI 1.5-5.4)
Limited PCP assessment
Smith et al. (1984), New Zealand,
males, age 20-80 yrs, 1976-1980
Woods et al. (1987), United States
(Washington), men, age 20-79 yrs,
1983-1985
Smith and Christophers (1992),
Australia, men, age >30 yrs, 1976-
1980
82 cases (cancer registry) (% deceased not
specified)
92 cancer controls (% deceased not specified)
128 cases (cancer registry) (24% deceased)
694 population controls (32% deceased)
30 cases (cancer registry),
30 cancer controls and 30 population controls
Excludes deceased cases and controls
Structured interview0
Structured interview0
Structured interview
Chlorophenols: OR 1.5 (95% CI 0.5-4.5)
Variable results (ORs 0.7-1.9) for fencing and
sawmill/timber merchant jobs
Chlorophenols: OR 0.99 (95% CI 0.7-1.5)
Lumber grader: OR 2.7 (95% CI 1. 1-6.4)
Variable results (ORs 0.79-4.8) for other "high,"
"medium," or "low" exposure jobs
Chlorophenols >1 d: 0 cases with this exposure
0 cases and 2 controls (1 population and 1 cancer control)
had definite PCP exposure
aTwenty cohorts from 10 countries workers; total n = 13,898; workers exposed to phenoxy herbicides or Chlorophenols. The follow-up period varied among the cohorts:
follow-up began between 1942 and 1973 and ended between 1987 and 1992 (Kogevinas et al., 1997).
bThe four case-control studies are described in Eriksson et al., 1990; Hardell and Eriksson, 1988; Eriksson et al., 1981; and Hardell and Sandstrom, 1979. More detailed
information the individual studies is shown in Table 4-5.
°Proxies included for deceased cases and controls.
30
-------
Case-control studies oflymphoma. Three case-control studies provided data pertaining to
risk of non-Hodgkin's lymphoma in relation to PCP using relatively detailed exposure data
(Table 4-3). Kogevinas et al. (1995) conducted a nested case-control study of non-Hodgkin's
lymphoma in the large, international cohort of 13,989 workers exposed to phenoxy herbicides or
chlorophenols assembled from 20 cohorts in 10 countries. Job records and company records
pertaining to chemicals used during specific processes were used by three industrial hygienists to
evaluate exposure to 21 specific chemicals (phenoxy herbicides, chlorophenols, polychlorinated
dibenzodioxins, furans, and process chemicals and raw materials). Cases of non-Hodgkin's
lymphoma (n = 32) were identified by review of death certificates (underlying and contributing
causes of death) for all countries, and review of cancer registries for the seven countries that had
national registries. Five controls were selected per case from within the cohort, matched by age,
sex, and country, for a total of 158 controls. The estimated associations in this study are
relatively imprecise, given the small size, but there is evidence of an association with any PCP
exposure (odds ratio [OR] 2.75, 95% CI 0.45-17.0) and specifically with the high exposure,
cumulative exposure category (OR 4.19, 95% CI 0.59-29.6). Increased risks were not observed
(i.e., ORs between 0.65 and 1.03) with the other specific chlorophenols examined
(2,4-dichlorophenol, 2,4,5-trichlorophenol, 2,4,6-trichlorophenol, and 2,3,4,6-TCP), and the
associations seen with phenoxy herbicides and dioxins were also weaker than those seen with
PCP (OR 1.84 for any dioxin or furan, 1.93 for 2,3,7,8-TCDD). Although this is a small study, it
is based within a large cohort for which detailed exposure assessments for a variety of
compounds are available.
Hardell et al. (1994, 1981) conducted a population-based, case-control study of non-
Hodgkin's lymphoma in men ages 25-85 years in Umea, Sweden. Cases (n = 105) with a
diagnosis occurring from 1974 to 1978 were identified through hospital records; 355 population
controls were identified through a population registry (for matching to living cases) and the
national death registry (for matching to deceased cases); individual-matching, rather than
frequency matching, was used. A self-administered questionnaire with follow-up phone
interview if needed was used to obtained detailed information pertaining to work history,
including information on specific jobs, and exposures. Next-of-kin proxy respondents were used
for deceased cases and controls. The questionnaire information was used to create an exposure
measure for specific chemicals, including chlorophenols and PCPs. The follow-up interview and
the evaluation of the questionnaire information was conducted without knowledge of case or
control status of the respondent. Exposures in the 5 years immediately preceding diagnosis (or a
corresponding reference year for controls) were excluded to account for a minimum latency
period. High exposure was defined as either >1 week of continuous exposure, or exposure for
>1 month in total; exposures less than this were considered low-grade. A strong association (OR
8.8, 95% CI, 3.4-24) was observed between high exposure to PCP (the predominant
chlorophenol used in this area) and risk of non-Hodgkin's lymphoma.
31
-------
A subsequent case-control study of non-Hodgkin's lymphoma covering a larger study
area (seven counties in northern and in mid-Sweden) was conducted by Hardell and Eriksson
(1999). This study was limited to men ages >25 years diagnosed between 1987 and 1990.
Procedures for case identification and recruitment of controls from the National Population
Registry or, for matching to deceased cases the National Registry for Causes of Death, were
similar to those used by Hardell et al. (1994, 1981). The study included 404 cases (43%
deceased) and 741 controls. Exposure information was collected with a self-administered mailed
questionnaire with follow-up phone interview if needed for clarification. Next-of-kin proxy
respondents were used for the deceased cases and controls. The work history included questions
on specific jobs, pesticides, and organic solvents. Exposures up to the year prior to diagnosis or
corresponding reference year for controls were included in the analysis, which was conducted
using conditional logistic regression. Increased risks were not seen with either chlorophenol (OR
1.1, 95% CI 0.7-1.8) orPCP (OR 1.2, 95% CI 0.7-1.8) exposure. A higher risk was seen with
PCP exposures that occurred between >20 and 30 years before diagnosis (OR 2.0, 95% CI 0.7-
5.3) compared with >10-20 years (OR 1.0, 95% CI 0.3-2.9) or >30 years (OR 1.1, 95% CI 0.7-
1.8). The authors noted that chlorophenols had been banned from use in Sweden in 1977 (or, as
noted in Hardell and Eriksson [2003] in January 1978), resulting in a different exposure period
relative to diagnosis for cases included in this study compared to their earlier study conducted
among cases diagnosed between 1974 and 1978 (Hardell et al., 1994, 1981).
Hardell and Eriksson (2003) discuss the trends in use of phenoxyacetic acids and
chlorophenols in relation to trends in the incidence of non-Hodgkin's lymphoma. Exposures to
these compounds peaked in the 1970s; incidence rates increased from 1960 to the late 1980s and
then were relatively steady through 2000. The authors note that these two trends are consistent
with a relatively short latency period between first exposure and disease onset.
Two other case-control studies of non-Hodgkin's lymphoma assessed occupational
exposure to chlorophenols with limited data specifically relating to potential exposure to jobs or
activities with likely exposure to PCP (Woods et al., 1987; Pearce et al., 1986b) (Table 4-3).
These studies reported no or weak (ORs <1.5) associations with chlorophenols, but somewhat
stronger risks with some specific jobs involving wood preservation or fencing work. Smith and
Christophers (1992) included Hodgkin's and non-Hodgkin's lymphoma in a small (52 cases)
study conducted in Australia using the area cancer registry. One cancer control and one
population-based control (from electoral rolls) were matched to each case based on age and place
of residence. The measure of association (point estimate or statistical significance), based on the
conditional logistic regression analysis of the matched triad data for PCP was not presented, but
this type of exposure was noted in four cases, one population control and three of the cancer
controls.
Case-control studies of soft tissue sarcoma. As with the studies of lymphoma, the case-
control studies of soft tissue sarcoma can be categorized based on the level of detail of the PCP
32
-------
assessment (Table 4-4). In the international nested case-control study by Kogevinas et al. (1995)
described above, 11 cases of soft tissue sarcoma and 55 matched controls were identified among
the 13,989 workers exposed to phenoxy herbicides or chlorophenols. None of these cases or
controls had been exposed to PCP. A meta-analysis of four separate but related (in terms of
exposure assessment methodology and other design features) case-control studies conducted in
different areas of Sweden (Eriksson et al., 1990, 1981; Hardell and Eriksson, 1988; Hardell and
Sandstrom, 1979) (Table 4-5) was published in 1995 (Hardell et al., 1995). The methodology
was based on the process described above for a study of lymphoma by Hardell et al. (1994,
1981).
Table 4-5. Summary of case-control studies of chlorophenol and soft tissue
cancer risk included in Hardell et al. (1995) meta-analysis
Reference
Hardell and
Sandstrom (1979)
Eriksson et al.
(1981)
Hardell and
Eriksson (1988)
Eriksson et al.
(1990)
Region of Sweden
Umea (northern)
Five counties,
(southern)
Three counties
(northern)
Upsala (middle)
Case accrual
1970-1977,
hospital records
1974-1978,
cancer registry
1978-1983,
cancer registry
1978-1986,
cancer registry
Age and sex criteria
Males, ages 26-80,
controls matched by
vital status, sex, age,
and area
Age and sex not
specified, controls
matched by vital status,
age, and area
Males, ages 25-80,
controls matched by
vital status, age, and
area
Males, ages 25-80,
controls matched by
vital status, age, and
area
n cases (percent deceased),
n controls"
52 cases (60% deceased),
208 controls
1 10 cases (35% deceased),
220 controls
54 cases (67% deceased),
311 population controls (33%
deceased), 179 cancer controls
(59% deceased)
218 cases (64% deceased),
2 12 controls
aThe matching design used in all of the studies except Hardell and Eriksson (1988) resulted in an equal proportion
of deceased cases and controls within each study.
Population controls were identified through a population registry or the national death
registry, and were individually matched to the cases by age and area of residence. A total of
434 cases and 948 controls are included in the meta-analysis. Work history data were obtained
through a self-administered questionnaire (completed by next-of-kin for deceased cases and
controls) with a follow-up phone interview (if needed to clarify responses). The work history
data were used to create an exposure measure for specific chemicals, including various forms of
phenoxyacetic acids and chlorophenols. Exposures in the 5 years immediately preceding
diagnosis (or a corresponding reference year for controls) were excluded to account for a
minimum latency period, and only "high" exposures (defined as >1 week continuously or
>1 month in total) were included in the meta-analysis. A strong association was observed
33
-------
between high exposure to PCP and soft tissue sarcoma risk (OR 2.8, 95% CI 1.5-5.4). The
primary strength of this meta-analysis is the relatively large number of cases obtained, which is
difficult to achieve in single-site studies of this rare disease.
The studies used in the meta-analysis were conducted by the same group of investigators
using a relatively common protocol across studies, which makes them suitable for this kind of
combined analysis. The exposure assessment was relatively detailed. There was a relatively
high proportion of deceased cases (and controls) in these studies (reflecting the high mortality
rate in this disease). The completeness and level of detail of the work history and exposure data
are likely to be lower in proxy- compared with self-respondents, resulting in a loss of precision
and possibly attenuation to the null.
The other three case-control studies of soft tissue sarcoma risk with more limited data
pertaining to PCP (Smith and Christophers, 1992; Woods et al., 1987; Smith et al., 1984) are
summarized in Table 4-4. These studies present variable results pertaining to various jobs with
potential exposure to PCP.
Case-control study of multiple myeloma. Pearce et al. (1986a) conducted a case-control
study of farming-related exposures and multiple myeloma risk in New Zealand. Men <70 years
old who had been hospitalized with a diagnosis of multiple myeloma (ICDs code 203) from 1977
to 1981 were recruited as cases. Controls, drawn from the Cancer Registry, were matched by
age and sex (all men) to the cases. A structured interview, completed by 76 (82%) of the
93 eligible cases and 315 (81%) of the 389 eligible controls, was used to collect data pertaining
to work history, with a particular focus on farming-related activities. There was little evidence
of an association with the general category of chlorophenol exposure (OR 1.1, 95% CI 0.4-2.7)
and work in a sawmill or timber merchant (OR 1.1, 95% CI 0.5-2.3). Stronger associations were
seen with a history of doing fencing work (OR 1.6, 95% CI 0.9-2.7) and jobs that involved
potential exposure to chlorophenols in a sawmill or as a timber merchant (OR 1.4, 95% CI 0.5-
3.9).
Case-control study of leukemia and brain cancer. Ali et al. (2004) reported results from
a case-control study of leukemia (ICDs-9th revision codes 204-208) and brain cancer (benign and
malignant, ICDs-9th revision codes 191, 192, 194.3, 194.4, and 225) in patients <30 years old at
diagnosis in Kaoshiung, Taiwan. Incident cases were drawn from a cancer registry and reviewed
by a pathologist to confirm diagnoses. Population-based controls were drawn using a
randomization scheme based on personal identification numbers and were matched to cases
based on age and sex, with three controls matched to each case. The mean ages of the brain
cancer and leukemia cases were 18 and 11 years, respectively. Participation rates for controls
were 61% for the brain cancer controls and 56% of the leukemia controls. Occupational history
(name of company, location, industry, duties, hours/week, and start and end dates) for jobs held
longer than 6 months since age 16 was obtained using a structured interview with each of the
parents. Additional interviews were conducted with any patient (or control) who was at least 16
34
-------
years old. The Taiwanese occupational and industrial coding system was used to assign 4-digit
job codes based on this information. The specific time periods of exposure examined in the
study were preconception (any job ending earlier than 1 year before the child's birth), prenatal
(any job held between 1 year prior to the child's birth and the child's birth), and postnatal (any
job held after the child's birth). Analyses were conducted using conditional logistic regression
among the discordant pairs, adjusting for smoking history (of the participant and the parents) and
exposure to medical radiation. Strong, but imprecise given the sample size, associations were
seen between paternal work as a wood-treater and risk of leukemia (for any exposure period, five
exposed cases, two exposed controls, OR 16.0, 95% CI 1.8-145.4; for preconception period, four
exposed cases, one exposed control, OR 12.2, 95% CI 1.4-109.2; for perinatal period, four
exposed cases, one exposed control, OR 13.0, 95% CI 1.4-125.5). No other information is
available pertaining to the specific material used by these workers (email dated September 19,
2006, from Dr. David Christiani, Harvard School of Public Health, Boston, Massachusetts, to Dr.
Glinda Cooper, U.S. EPA).
Ruder et al. (2009) examined risk factors for brain cancer (glioma) in a non-metropolitan
area, focusing on farming-related exposures including pesticides. This population-based case-
control study was based in non-metropolitan counties in Iowa, Michigan, Minnesota, and
Wisconsin. Patients with histologically-confirmed primary intracranial glioma diagnosed
between January 1, 1995 and January 31, 1997 were identified through medical facilities and
neurosurgeons' offices. Case ascertainment, compared with state tumor registries, was estimated
to be 78%. Controls were identified through driver's license records (ages 18-64) and Medicare
records (ages 65-80), based on residence in the study area on January 1, 1995. Controls were
selected based on age and sex strata defined by the distribution of these factors among the cases.
The number of cases identified was 872 and the participation rate was 91% (798). The number
of controls identified was 1,559 and the participation rate was 70.4% (n = 1,175). In-person
interviews were conducted with 438 cases and 360 case proxies (next-of-kin) and with the
controls. The structured questionnaire included detailed questions about types of crops and
livestock, use of pesticides, and other farm activities. These questions were asked of the
participants who reported they had lived or worked on a farm after age 17. Questions about use
of wood preservatives included ever used, time spent (hours per year), type of wood preservative
used (ever used creosote, ever used PCP). There was no association between wood preservatives
ever used in the full sample (OR 0.96, 95% CI 0.63-1.47) or in the sample excluding proxy
respondents (OR 0.91, 95% CI 0.53-1.55). Similar results were seen with time spent using wood
preservatives. In the analysis of type of wood preservative used, however, the OR was 1.93
(95% CI 0.61-6.10) for the full sample and was stronger in the analysis excluding proxy
respondents (OR 4.55, 95% CI 1.14-18.1). Elevated odds ratios were not seen with respect to
use of creosote as wood preservative (OR 0.83, 95% CI 0.33-2.11 for the full sample and OR
0.42, 95% CI 0.13-1.32 for the sample excluding proxy respondents).
35
-------
4.1.1.4. General Issues—Interpretation of the Epidemiologic Studies
The strongest of the cohort studies, in terms of design, is the large sawmill cohort study
conducted in British Columbia, Canada and updated by Demers et al. (2006). As noted
previously, important design features that add to the strengths of this study include its size
(n = 23,829 workers), the exposure assessment procedure developed specifically to address the
exposure situations and settings of the study, use of an internal referent group, analysis of PCP
and TCP exposures, the low loss to follow-up, and the use of a population-based cancer registry
that allowed for the analysis of cancer incidence. In contrast, the other cohort studies in a
manufacturing plant (Ramlow et al., 1996) and a plywood mill (Robinson et al., 1987) were
much smaller (n = 770 and 2,283, respectively), and did not present analyses that allow for
differentiation of risk between potential co-exposures (e.g., dioxins and furans in the
manufacturing plant, and formaldehyde in the plywood worker cohort). Even with the large size
of the British Columbia sawmill cohort, however, there is limited statistical power to estimate
precise associations with relatively rare cancers.
Case-control studies offer the potential for increased statistical power for assessing
associations with rare cancers such as liver cancer and various forms of lymphomas; however,
there is a considerable range in the detail and quality of the exposure assessment used in case-
control studies. Population-based, case-control studies rarely include specific exposure
measurements taken at specific worksites of individual study participants. Although it is more
difficult to determine absolute exposure levels without these individual measurements, the
exposure assessment methodology does allow ranking of exposure levels and useful between-
group comparisons of risk. Among the case-control studies with data pertaining to cancer risk
and PCP exposure, the studies with the strongest designs in terms of exposure assessment are the
nested case-control study by Kogevinas et al. (1995), conducted within a large, multinational
cohort of workers, and the collection of studies from Sweden (Hardell et al., 1995, 1994). These
studies used population-based cancer registries for case ascertainment. The nested case-control
study included detailed information pertaining to exposures for specific jobs, periods, and
locations. The Swedish studies obtained detailed information about work histories (rather than
just the usual or most recent job). The inclusion of work history from interviews with next-of-
kin (for cases and controls) in the Swedish studies, however, is most likely to result in
nondifferential misclassification of exposure, and thus attenuation in the observed associations.
Although there are demographic risk factors (e.g., age, sex, race) for non-Hodgkin's
lymphoma, multiple myeloma, and soft tissue sarcoma, "lifestyle" behaviors (e.g., smoking
history, alcohol use) have not been associated with these diseases. The large cohort study of
sawmill workers by Demers et al. (2006), the smaller cohort study by Ramlow et al. (1996), and
the nested case-control study by Kogevinas et al. (1995) all used internal comparison groups,
which would also reduce the potential influence of confounders.
36
-------
Contamination of PCP with dioxins and related byproducts is known to occur as part of
the production process. Several studies have examined the level of various dioxins and furans
among workers in the PCP and trichlorophenol production workers at the Michigan Division of
the Dow Chemical Company (Collins et al., 2007, 2006; Ott et al., 1993). The primary
contaminants are hexa-, hepta-, and octa-chlorinated dibenzodioxins and higher-chlorinated
dibenzofurans, rather than 2,3,7,8-TCDD.
There are several reasons that it is unlikely that the associations observed in the
epidemiologic studies described above are due to these contaminants. Although 2,3,7,8-TCDD
is associated with an increased risk of cancer, the available epidemiologic studies most
consistently demonstrate this association with all cancers, rather than with individual cancers
(NAS, 2006; Steenland et al., 2004). In contrast, none of the epidemiologic studies of PCP
exposure have demonstrated an increased risk for all cancers, but there is evidence of
associations (ORs, some of which are relatively strong) with various forms of lymphopoietic
cancers (non-Hodgkin's lymphoma, multiple myeloma) and soft tissue sarcoma. Thus, the
patterns observed differ substantially for PCP and dioxins.
Another argument against the influence of contaminants as the explanation for the
observations pertaining to PCP is based on the comparisons within a study of effects of different
chemicals. In the nested case-control study conducted within the large international cohort of
workers exposed to phenoxy herbicides or chlorophenols (Kogevinas et al., 1995), the observed
association between PCP exposure and non-Hodgkin's lymphoma (OR 2.75, 95% CI 0.45-17.0)
was stronger than the associations observed with the other dioxin and furan exposures, and there
was little evidence of an association with other types of chlorophenols. Also, in the large cohort
study of sawmill workers by Demers et al. (2006), the associations with multiple myeloma were
considerably stronger (based on RR), and the association with non-Hodgkin's lymphoma were
similar or somewhat stronger, for PCP than for TCP, but there is little difference in the
contaminants. The levels of contaminants are similar between the two chemicals, except that in
PCP, the levels of octachlorodibenzo-p-dioxin (OCDD) and octachlorodibenzofuran are greater
compared with those found in TCP (Schwetz et al., 1974a, b).
De Roos et al. (2005) reported results from a case-control study of non-Hodgkin's
lymphoma that examined plasma levels of various polychlorinated biphenyls, dioxins, furans,
and pesticides (PCP was not included in their analyses). There was no association between
OCDD levels and lymphoma risk. The strongest association was seen with 1,2,3,4,7,8-hexa-
chlorodibenzofuran, with an OR of 2.64 (95% CI 1.14-6.12) per 10 pg/g lipid. However, in a
study of the Dow Chemical Company chlorophenol production workers in Michigan (Collins et
al., 2007), the biggest difference in serum concentration of dioxin and furan congeners among
PCP exposed workers compared with various referent groups was in OCDD levels (mean 2,594,
509, and 439 pg/g lipid in the PCP workers, a worker non-exposed comparison group, and a
community comparison group, respectively; p< 0.05 for comparisons between PCP and each of
37
-------
the referent groups). Much smaller elevations (i.e., mean values of approximately 10 pg/g lipid
compared with 8 pg/g lipid) were seen for some of the hexa- or heptachlorodibenzofurans, but
the authors noted there was little evidence of increased furan levels in the PCP exposed workers.
Collins et al. (2007, 2006) also noted that although furan contaminants have been detected in
commercial PCP, they have rarely been found in blood samples from PCP workers. Thus, it is
unlikely that the observations pertaining to non-Hodgkin's lymphoma risk and PCP exposure can
be attributed to heptachlorodibenzofuran. McLean et al. (2009a) also reported increased levels
of OCDD in serum samples collected from PCP exposed sawmill workers 20 years after the last
exposure to PCP, with mean levels of 309.25 and 157.83 pg/g lipid in exposed and non-exposed
workers, respectively; data on furan levels were not provided.
The classifications used for the various subtypes of lymphomas, leukemias, and sarcomas
can be confusing and may not be applied similarly in different studies, particularly when
conducted over different time periods, or in different locations by different investigators. This
potential inconsistency may contribute to differences in results for these subtypes seen across
different studies, but any differences in disease definitions should not produce a biased result
within a study since the disease classification methods in the available studies (e.g., Demers et
al., 2006; Hardell et al., 1995) were independent of the exposure classification system.
4.1.1.5. Specific Cancers
Considering the issues described above with respect to the strengths and limitations of the
available epidemiologic studies, the following summary of the evidence relating to PCP
exposure and specific types of cancer can be made.
Liver cancer. An increased risk of liver cancer in relation to PCP exposure, but not TCP
exposure, was seen in the large cohort study of 23,829 sawmill workers in British Columbia
(Demers et al., 2006). There was little evidence of an increased risk when considering a 10- or
20-year exposure lag period. The difference between the results in the no-lagged and lagged
analyses may reflect the effect of PCP as a promoter, rather than an initiator, of liver cancer;
alternatively, it may reflect the influence of chance given the relatively low statistical power, and
thus lack of precision, inherent in a study of this relatively rare cancer even in this large-sized
cohort. No case-control studies of liver cancer risk in relation to PCP exposure were identified;
the plywood mill workers cohort study (Robinson et al., 1987) focused on lymphatic and
hematopoietic cancers and did not present liver cancer data; no cases of liver cancer were
observed in the small cohort study of 770 men in a PCP manufacturing plant (Ramlow et al.,
1996). The available epidemiologic studies, in combination with the observation of liver tumors
in mice (NTP, 1989), suggest a relationship between PCP and carcinogenic effects, although it
should be noted that this determination is based on limited human data.
Lymphomas (non-Hodgkin 's lymphoma, multiple myeloma). There was substantial
evidence of an association between PCP exposure and the incidence of non-Hodgkin's
38
-------
lymphoma and multiple myeloma, including an exposure-response trend across categories
reflecting higher exposures, in the large cohort study of sawmill workers (Demers et al., 2006).
For multiple myeloma, the risk ratios in the highest category of exposure were strong (>4.0), and
there was no evidence of similar patterns in the analyses of TCP exposure. For non-Hodgkin's
lymphoma, Demers et al. (2006) observed approximately a twofold increased risk in the highest
two categories of exposure, with a slight attenuation seen in the mortality analysis. An
attenuation of the exposure-response in the highest exposure category is commonly seen in
epidemiologic studies of occupational cohorts (Stayner et al., 2003).
The nested case-control study by Kogevinas et al. (1995), conducted within the combined
international cohorts of exposed phenoxy herbicide workers, also provides support for an
association between PCP (but not other chlorophenols) and non-Hodgkin's lymphoma risk. One
case-control study in Sweden with a relatively specific exposure measure of PCP also reported
very strong associations (OR 8.8) with non-Hodgkin's lymphoma. A subsequent study by the
same investigators did not observe this type of association (OR 1.2). There are no case-control
studies of multiple myeloma with a similarly focused type of exposure estimate. The available
epidemiologic studies strongly suggest that PCP exposure is associated with non-Hodgkin's
lymphoma and multiple myeloma risk. For the reasons described above, it is unlikely that this
association can be explained by co-exposures or contamination with other chlorophenols,
dioxins, or furans.
Soft tissue sarcoma. There was no association between PCP exposure and increased risk
of soft tissue sarcoma in the large sawmill worker cohort study by Demers et al. (2006). The
trend, based on small numbers, was for a decreased risk with higher exposures. None of the
11 cases or 55 controls in the nested case-control study by Kogevinas et al. (1995) were exposed
to PCP. However, the number of cases was insufficient to conclude that there is no association
between exposure to PCP and soft tissue sarcoma. These observations, within both of these
studies, reflect the difficulty in studying such a rare disease, even in large cohorts. In the
collection of case-control studies conducted in Sweden, summarized by Hardell et al. (1995), a
strong association (OR 2.8) was seen with their measure of PCP exposure (>1 week continuously
or 1 month total), based on structured interviews. A limitation of these studies is the relatively
large proportion of proxy respondents used (cases and matched controls), which is likely to result
in a loss of precision and possible attenuation of the observed association. In almost all cases,
the proportion of proxy respondents (i.e., because the case or control was deceased) was similar
for cases and controls. The available epidemiologic studies provide some evidence of an
association between PCP exposure and soft tissue sarcoma risk. The low incidence rate,
combined with a need to consider histology to accurately make a classification, and a fairly high
case fatality rate make it difficult to conduct definitive epidemiologic studies of this disease.
Brain cancer. The available data pertaining to brain cancer (glioma) risk among adults in
relation to PCP exposure is limited, and is based primarily on the case-control study of farming-
39
-------
related exposures by Ruder et al. (2009). In this study, an increased risk was seen in relation to
use of PCP as a wood preservative, particularly when proxy respondents were excluded from the
analysis (OR 4.55). In the large cohort study of PCP-exposed saw mill workers by Demers et al.
(2006), the SMR for brain cancer was 0.99 (95% CI 0.73-1.31) and the SIR was 1.08 (95% CI
0.80-1.43). This study did not present exposure-response modeling for brain cancer using the
internal comparison group.
Childhood cancers. There was little evidence of an association between paternal
exposure to PCP and the incidence of childhood cancers in the large sawmill worker cohort study
(Heacock et al., 2000), although with only 40 incident cancers, even this large cohort is of
limited statistical power for the analysis of these cancers. A small case-control study in Taiwan
reported strong associations with childhood leukemia in relation to paternal exposure
(particularly in the preconception and perinatal periods) (Ali et al., 2004). The available
epidemiologic data are too limited to assess with confidence whether parental, prenatal, or early
childhood exposure to PCP affects risk of childhood cancers.
4.1.2. Studies of Noncancer Risk
4.1.2.1. Case Reports of Acute, High-Dose Exposures
One of the earliest reports recognizing the toxic effects of PCP in humans was published
by Truhaut et al. (1952). The authors described the then-current procedures for treatment of
lumber to prevent rotting. Workers known as "treaters" soaked freshly sawn lumber in tubs
containing a 3% solution of a mixture of 80% pentachlorophenate of sodium and 20%
tetrachlorophenate of sodium. After soaking, the lumber was then carried to other workers called
"stackers" to be put in stacks. Based on examinations of >100 lumber treaters, symptoms of
PCP exposure included skin irritation with blisters, congestion of mucous membranes of eyes
and nose, loss of appetite, loss of weight, constriction of throat, respiratory stress, and fainting.
Urine levels of PCP in 16 workers who had worked for 2 months as treaters were between 3 and
10 mg/L. Truhaut et al. (1952) also describe the deaths of two workers following exposure to
PCP. Autopsy findings included liver poisoning, degenerative lesions in kidney, considerable
edema in the lungs, the presence of PCP in liver, kidney, blood, stomach, intestine, heart, lung,
and urine in one case, and considerable congestion and edema of the lungs and albumin in the
urine in the other case.
An incident of accidental PCP poisoning occurred in a nursery for newborn infants in St.
Louis in 1967 (Smith et al., 1996; Armstrong et al., 1969). Sodium pentachlorophenate had been
used as an antimildew agent by the hospital laundry. Nine cases of illness were seen with fever
and profuse sweating. As the disease progressed, respiratory rates increased and breathing
became labored. Other common findings included rapid heart rate, enlarged liver, and irritability
followed by lethargy. Laboratory tests showed progressive metabolic acidosis, proteinuria,
increased levels of blood urea nitrogen (BUN), and x-rays suggestive of pneumonia or
40
-------
bronchiolitis. Two of the cases were fatal. The only source of exposure for the infants was skin
absorption of the residues of sodium pentachlorophenate on the diapers, undershirts, and
bedding. The product label warned against use in laundering diapers and the amount used was
3-4 times the amount recommended for regular laundry. Analysis of freshly laundered diapers
showed a quantity of PCP ranging from 1.4 to 5.7 mg per diaper. One infant had 11.8 mg of
PCP/100 mL of serum before a transfusion was performed. A fatal case was found to have 2.1-
3.4 mg/100 g in various body tissues. The average duration of the hospital stay in the nursery
(when contaminated diapers were used) until the appearance of the first symptoms was 9 days.
Acute poisonings, including two fatalities, were reported in a study of workers in wood
preservative manufacturing plants (Wood et al., 1983). A general air sample taken from the
work area of one of the deceased workers found PCP levels of 4.6 mg/m3, which is 9 times the
Occupational Safety and Health Administration standard. Another case report described the
occurrence of pancreatitis in a wood worker (joiner) who had been applying a wood preservative
that contained PCP and zinc napththanate (Cooper and Macaulay, 1982). Gray et al. (1985)
reported the case of a 3 3-year-old man who used a jackhammer to break up large blocks of PCP,
which were ground into powder. He developed lethargy, rapid respiration, and sweating, which
led to his hospitalization, coma, pulmonary edema, and death.
From 1993 through 1996, 122 unintentional exposures were reported to the Toxic
Exposure Surveillance System of the American Association of Poison Control Centers. Children
<6 years of age were involved in 32 of the exposures, and half of these were followed to
determine outcome. Only five of the children were reported to have developed symptoms, all of
which were minor. Six of the children were seen in a health care facility and one was
hospitalized. There were 90 exposures in adults and older children, 30 of which had a minor
outcome and 9 had a moderate outcome. One case was considered life-threatening. Thirty-four
cases were seen in a health care facility; two were hospitalized and one was admitted for critical
care.
Detailed descriptions of 71 cases of PCP exposure and health effects submitted to the
California Pesticide Illness Surveillance Program (1982-1996) were evaluated. Irritative effects
to the eye and skin were observed in 58% of the total reports of illness in California, while the
remaining 42% exhibited effects systemic in nature, including symptoms of headache, nausea,
and difficulty breathing. Only cases with a definite, probable, or possible relationship were
reviewed. PCP was judged to be responsible for the health effects in 48 of these cases. Only
half of the systemic cases were classified as having a probable or definite relationship between
the exposure and the health effects. One individual was hospitalized in 1982 for skin grafts due
to second- and third-degree burns after carrying PCP-treated lumber for 4 weeks. The burns
were reported to the shoulder, neck, chin, back, and thigh, and were characterized as an allergic
reaction by one investigator.
41
-------
Dust and mist concentrations >1.0 mg/m3 can result in painful irritation of the upper
respiratory tract resulting in violent sneezing and coughing in persons not previously exposed to
PCP (U.S. EPA, 1980). Some nose irritation has been reported at levels as low as 0.3 mg/m3.
4.1.2.2. Studies of Clinical Chemistries, Clinical Examinations, and Symptoms
Chloracne has been often reported in studies of workers involved in the production of
chlorophenols. Contamination with chlorinated dioxins and dibenzofurans is a likely cause of
this association. Cole et al. (1986) described a case of chloracne in a carpenter with substantial,
prolonged dermal contact to PCP-treated lumber. Several studies have reported a high
prevalence of chloracne among workers involved in the manufacture of PCP. Bond et al. (1989)
examined 2,072 workers at the Dow Chemical Company manufacturing plant in Michigan.
O'Malley et al. (1990) examined 648 workers in Illinois. Cheng et al. (1993) examined
109 workers at a production plant in China. The prevalence of chloracne was 15% in Michigan,
7% in Illinois, and 73% in China.
PCP was used extensively in Hawaii as a wood preservative for protection against
termites and fungi endemic to the tropical climate. Studies of the health effects in workers
occupationally exposed, and in the general population exposed through residential contact and
diet, were begun in the 1960s (Bevenue, 1967). In a study of 18 exposed workers examined with
serial blood and urine measures before and after a 21-day vacation, creatine clearance and
phosphorus reabsorption were significantly decreased during the work period compared with the
vacation period (Begley et al., 1977). Klemmer et al. (1980) reported data from a study of
47 Hawaiian workers involved with treatment of wood products with PCP, 333 workers with
mixed exposures to various pesticides while working as farmers or pest control operators, and
42 controls with no history of occupational pesticide exposure (total n = 422). Blood and urinary
measures of PCP were elevated in the exposed workers, particularly among those who had
worked with an open-vat process (e.g., mean serum concentrations of 3.78, 1.72, 0.25, and
0.32 ppm in the open-vat wood treaters, pressure-tank wood treaters, farmers and pest control
operators, and controls). Results of clinical laboratory analyses showed that PCP exposure was
highly associated with increased numbers of immature leucocytes (band cells), increased levels
of blood plasma cholinesterase, alkaline phosphatase (ALP), gamma-globulin, basophils, and
uric acid, and reduced serum calcium. These analyses were limited to individuals with no
missing data for any of the parameters, and included only 7 open-vat wood treaters, 10 pressure-
tank wood treaters, 155 farmers and pest control operators, and 17 controls. Age-standardized
prevalence rates for conjunctivitis, chronic sinusitis, and chronic upper respiratory conditions
were approximately 3 times higher among the workers exposed to PCP than among the controls.
Prevalence rates of infections of the skin and subcutaneous tissue and of gout were
approximately 1.7 times higher in the PCP-exposed individuals. The authors noted that the
conjunctivitis cases only occurred among workers involved in pressure treatment and, therefore,
42
-------
had mixed exposure to PCP and other chemicals, and that the increased prevalence of gout may
have been due to a greater proportion of Filipinos in the PCP-exposed group, since the
prevalence of this condition is increased in this ethnic group.
Gilbert et al. (1990) examined clinical and laboratory parameters in another study of male
wood treaters in Hawaii. The 88 study participants were drawn from a total of 182 workers who
had worked for long periods and had chronic, low-level exposure to wood-treating chemicals
including PCP. Exposed workers had to be currently employed in a Hawaiian wood treatment
company for at least 3 months at the time of recruitment for the study or have been previously
employed for at least 12 months in a Hawaiian wood treatment company since 1960, including at
least one 3-month period of continuous employment as a wood treater. A comparison group of
58 men was selected from various unions (e.g., carpenters, masons) and from friends and
relatives of the exposed group. The comparison group was similar to the age, race, level of
physical activity, and weight distribution of the exposed group. The level of urinary PCP was
higher among the exposed (means of 174 and 35 ppb in the exposed and comparison groups,
respectively). The clinical examination of study participants included a complete review of
systems, lipid profile, and liver and kidney function tests. The authors reported no statistically
significant differences between the groups in the elements of the clinical examination or
symptoms (e.g., fever, skin rash, eye irritation, wheezing, cough). Although a few of the
laboratory results (e.g., heart rate, systolic blood pressure) differed between cases and controls,
additional analyses of trends across PCP exposure groups (based on urinary values) did not
provide evidence of differences that could be attributed to this exposure.
Walls et al. (1998) examined medical history and current symptoms in 127 sawmill
workers in New Zealand, many of whom were self-identified as having health concerns related
to PCP exposure. Study participants were primarily recruited through the Wood Industries
Union of Aoteoros and timber companies. Many also had exposures to other chemicals typically
used in the timber industry (e.g., arsenic) and to organopesticides. Data on occupational and
lifestyle histories (e.g., tobacco and alcohol use), exposure to PCP, medical history, and current
symptoms were collected using a structured questionnaire. An exposure metric incorporating
length of PCP exposure and a cumulative score for types of PCP work, type of vehicle, use of
personal protection, and intensity of exposure was calculated for each participant. Based on this
exposure metric, participants were categorized into three groups: low (n = 45), medium (n = 39),
and high (n = 43) exposure. There was no control group. An increased prevalence of weight
loss, fevers, excess fatigue, upper respiratory tract symptoms, history of emphysema or
bronchitis, and current or history of nausea was seen in the high-exposure group, and for many of
these symptoms, an exposure-effect gradient was seen across the three exposure groups (trend
p < 0.05). The authors describe these results as consistent with their clinical impressions, and as
hypothesis-generating observations that warrant additional research of a representative sample of
workers exposed to PCP.
43
-------
McLean et al. (2009b) followed up the findings of Walls et al. (1998) with an expanded
study of former New Zealand sawmill and timber workers. Employment records from three
employers (two sawmills and the New Zealand Forest Service) covering the period from 1970 to
1990 (McLean et al., 2007) were obtained and used to identify a cohort of workers. This follow-
up study included workers who had worked for >12 months and who were alive after 2003 and
living in New Zealand. After excluding 249 individuals who declined to participate (n = 146) or
who were not able to be contacted by post (n = 103), a pool of 776 potential participants
remained. From this pool, 338 were recruited and consented to participate, and 293 completed
the study. The study involved an in-person interview, clinical examination (including a
standardized neurological exam), and a blood sample (used for a nonfasting blood glucose test).
These activities were conducted either at a medical center or in a participant's home; the exam
was conducted separately from the interview, by a different member of the study team. The
interview included more detailed information about work history and tasks and a health history,
which included questions about diagnosis with respiratory and other conditions and 10 physical
symptoms (Table 4-6).
44
-------
Table 4-6. Prevalence of medical conditions and physical symptoms, and associations with PCP exposure, in
293 timber workers in New Zealand
History
Non-exposed
(n = 177)
n
(%)
Exposed
(n = 116)
n
(%)
OR
(95% CI)
Low exposure"
(n = 58)
n
(%)
OR
(95% CI)
High exposure"
(n = 58)
n
(%)
OR
(95% CI)
Conditions
Asthma
Nasal allergies
Eczema
Acne
Chronic bronchitis
Tuberculosis, pleurisy, or pneumonia
Diabetes
Thyroid disorder
Impaired kidney function
Impaired liver function
30
75
49
61
22
13
8
7
21
15
(17.1)
(42.6)
(27.8)
(34.7)
(12.5)
(7.4)
(4.6)
(4.0)
(11.9)
(8.5)
27
37
44
35
15
24
10
6
14
18
(23.3)
(31.9)
(37.9)
(30.4)
(13.0)
(20.7)
(8.6)
(5.2)
(12.2)
(15.5)
1.46
0.62
1.50
0.87
1.01
3.04
1.95
1.50
0.94
1.94
(0.79-2.68)
(0.37-1.03)
(0.90-2.50)
(0.52-1.47)
(0.48-2.13)
(1.46-6.33)
(0.73-5.23)
(0.47-4.85)
(0.43-2.02)
(0.92-4.10)
11
18
19
18
5
13
7
2
6
11
(19.0)
(31.0)
(23.8)
(31.0)
(8.6)
(22.4)
(12.1)
(3.5)
(10.5)
(19.0)
1.56
0.61
1.20
0.88
0.65
3.41
2.93
1.00
0.83
2.42
(0.53-2.53)
(0.32-1.16)
(0.62-2.29)
(0.46-1.69)
(0.23-1.85)
(1.46-7.94)
(0.99-8.72)
(0.19-5.11)
(0.30-2.26)
(1.03-5.72)
16
19
25
17
10
11
3
4
8
7
(27.6)
(32.8)
(43.1)
(29.8)
(17.5)
(19.0)
(5.2)
(6.9)
(13.8)
(12.1)
1.79
0.59
1.87
0.86
1.43
2.68
1.07
2.03
1.05
1.46
(0.87-3.70)
(0.31-1.11)
(0.99-3.51)
(0.44-1.68)
(0.60-3.38)
(1.11-6.48)
(0.27-4.31)
(0.54-7.64)
(0.41-2.68)
(0.55-3.88)
Physical symptoms
Unintentional weight loss
Unexplained persistent fevers
Persistent fatigue
Eye discomfort (reddened and dry)
Pins and needles, hands or feet
Numbness, hands or feet
Loss of muscle power, hands or feet
Recurrent nausea
Recurrent diarrhea
Recurrent bowel upsets
12
7
37
49
82
58
25
6
8
18
(6.8)
(4.0)
(21.0)
(27.8)
(46.6)
(33.0)
(14.2)
(3.4)
(4.6)
(10.2)
14
10
31
28
52
38
21
12
14
15
(12.1)
(8.6)
(26.7)
(24.1)
(44.8)
(32.8)
(18.0)
(10.3)
(12.1)
(12.9)
1.57
2.08
1.26
0.88
0.80
0.95
1.64
2.42
2.68
1.28
(0.68-3.62)
(0.76-5.73)
(0.72-2.22)
(0.51-1.53)
(0.48-1.31)
(0.57-1.60)
(0.69-2.58)
(0.85-6.87)
(1.07-6.71)
(0.61-2.72)
9
8
16
14
23
14
10
3
10
10
(15.5)
(10.3)
(27.6)
(24.1)
(39.7)
(24.1)
(17.2)
(5.2)
(17.2)
(17.2)
2.13
2.58
1.31
0.87
0.68
0.65
1.31
1.18
4.08
1.81
(0.83-5.47)
(0.82-8.09)
(0.66-2.63)
(0.43-1.74)
(0.36-1.28)
(0.32-1.29)
(0.58-2.98)
(0.28-5.08)
(1.51-11.0)
(0.78-4.28)
5
4
15
14
29
24
11
9
4
5
(8.6)
(6.9)
(25.9)
(24.1)
(50.0)
(41.4)
(19.0)
(15.5)
(6.9)
(15.2)
1.05
1.60
1.21
0.89
0.93
1.34
1.37
3.71
1.42
0.80
(0.34-3.22)
(0.44-5.79)
(0.59-2.46)
(0.43-1.81)
(0.50-1.74)
(0.71-2.51)
(0.61-3.05)
(1.21-11.4)
(0.40-4.98)
(0.28-2.29)
""Cumulative exposure metric, based on product of intensity and duration; low score = 0-120; high score = >120. Similar patterns seen with the intensity score.
Sources: McLean et al. (2009b, 2007).
45
-------
Exposure status was based on review of job history records, with a semi-quantitative
intensity score based on job title (taking into account degree of direct contact with PCP) and
specific high-exposure tasks (mixing PCP solutions, cleaning sludge from PCP dip tanks, and
backpack spraying) (McLean et al., 2009b). A cumulative exposure measure was based on the
product of the intensity and work duration data. Exposure classification was conducted separate
from the clinical examination and interview. Among the 293 study participants, 177 and
116 were classified as non-exposed and exposed to PCP, respectively. Categories used for
analysis of total intensity score were 2.0-4.9 (n = 86) and >5.0 (n = 30); categories for the
cumulative exposure analysis were <120 (n = 58) and>120 (n = 58).
Analyses were adjusted for age (as a continuous variable, gender, and smoking status
(never, former, and current). Comparing the non-exposed and exposed categories, an association
was seen between exposure and chronic respiratory disease (OR 3.04, 95% CI 1.46-6.33) and
recurrent diarrhea (OR 2.68, 95% CI 1.07-6.71). Other outcomes that were elevated (OR>1.5),
with higher risks (OR approximately >2.0) seen in the higher exposure groups, were eczema,
thyroid disorder, unexplained persistent fevers, and recurrent nausea (Table 4-6).
Two reports have described health effects of nonoccupational exposure to PCP (Lambert,
1986; CDC, 1980). The U.S. EPA conducted a survey of PCP-treated log homes and their
occupants at the request of the Kentucky Department of Health Services (CDC, 1980).
Environmental and medical data were collected for 32 individuals in 21 homes. No significant
associations were reported between serum or urinary levels of PCP and health complaints,
laboratory parameters of liver function, microsomal enzyme induction, renal function,
neurological examination, or presence of lymphadenopathy. However, there was an association
between a finding of skin abnormalities and serum and urinary levels of PCP. The types of skin
abnormalities were not described. The author noted that skin abnormalities might lead to
increased absorption of PCP resulting in higher biologic PCP concentrations in blood and urine,
rather than PCP being a cause of skin abnormalities. In another report of nonoccupational PCP
exposure, Lambert et al. (1986) described the development of pemphigus vulgaris, a serious
autoimmune disease involving successive blisters (bullae) in a 41-year-old man who had
purchased a PCP-treated bookcase and in a 28-year-old woman who had several rafters in the
living room treated with PCP. A third case involving urticaria (hives) occurred in a 35-year-old
male who worked with PCP-treated wooden framework. The authors noted a "striking
parallelism" in all three cases between the disease course and PCP serum levels and stated that
these cases suggest "possible new hazardous effects of PCP."
4.1.2.3. Studies of Neurological Outcomes
Two of the studies of general health effects described in this section also contain data
pertaining to neurobehavioral function (Walls et al., 1998; Cheng et al., 1993). In the study of
127 sawmill workers in New Zealand by Walls et al. (1998), a questionnaire developed to screen
46
-------
for neuropsychological impairment within the context of solvent exposures was used. This
measure of neuropsychological dysfunction was associated with PCP exposure level, with 62%
of the low-exposure group, 74% of the medium-exposure group, and 81% of the high-exposure
group characterized as positive on this screening test (trend/? < 0.05). Cheng et al. (1993)
included a nerve conduction test in a study of workers at a PCP production plant and a
comparison group of desalination plant workers. A slower conduction time was seen among
workers (n = 10) in the trichlorobenzene building (in which non-y-hexachlorocyclohexane was
heated and decomposed into trichlorobenzene and hydrogen chloride) compared with the
controls. However, there was no reduction in conduction time among workers in the other
production areas.
Triebig et al. (1987) conducted a longitudinal study of nerve conduction velocity on
10 individuals who had worked with PCP or PCP-containing substances including TCP,
y-hexachlorocyclohexane (lindane), and aldrin for an average of 16 years (range = 4-24 years).
Nerve conduction velocity measurements were available for comparison for years 1980 and 1984
for the 10 subjects. In addition, serum and urine concentrations of PCP were measured. Limited
industrial hygiene data showed that PCP concentrations in the air during the subjects'
employment were below the maximum allowable concentration of 500 ug/m3. Results of
biological monitoring showed serum concentrations of PCP between 38 and 1,270 ug/L (upper
normal limit =150 ug/L) and urine concentrations between 8 and 1,224 ug/L (upper normal
limit = 60 ug/L) showing definite internal exposure. However, no significant changes in nerve
conduction velocity during the period 1980-1984 were demonstrated in any of the subjects, and
there was no observed correlation between nerve velocity and level of PCP exposure.
Peper et al. (1999) examined neurobehavioral measures in 15 women exposed to wood
preserving chemicals in their residence and a comparison group of 15 unexposed women. Both
groups were drawn from a larger study of women seen at a university hospital in Heidelberg,
Germany, for reproductive and menopausal-related (but not neurological) complaints. Wood
preserving chemicals, usually containing PCP and/or lindane, had been used on interior wood in
this region. Exposure status was based on answers to a questionnaire pertaining to
environmental risk factors (e.g., treatment of wood in the home) and serum levels of PCP and
lindane. The exposed group consisted of women who indicated exposure to wood preserving
chemicals for >5 years who had a blood level >25 ug/L PCP and 0.1 ug/L lindane. The mean
(standard deviation) blood levels in the exposed and control groups, respectively, were
43.6 (31.2) ug/L and 11.8 (4.5) ug/L for PCP (p = 0.001), 0.085 (0.086) ug/L and 0.043 (0.025)
for lindane (p = 0.007), and 0.497 (0.964) ug/L and 0.268 (0.164) ug/L for p-hexachloro-
cyclohexane (p > 0.05). Neurobehavioral assessment included a 27-item questionnaire used to
derive scores for three factors relating to attention (distractibility and slowing of mental
processes, fatigue and slowing of practical activities, and motivation and drive), an emotional
mood scale, the Beck Depression Inventory, and the Freiburg Personality Inventory to assess
47
-------
primary personality traits. Study participants also underwent a neuropsychological examination
focusing on tests sensitive to cortico-striatal dysfunction, an intelligence quotient (IQ) test, tests
of attention and of psychomotor speed, visual and verbal span subtests of the Wechsler Memory
Scale-Revised, and the "Tower of Hanoi task" test of motor skills. A close relative of each study
participant also completed a rating scale of behavior. Several differences between the exposed
and control groups in these neurological tests were seen, including higher (i.e., worse
functioning) scores on the Beck Depression Inventory, three of the four measures of mood
(depression, fatigue, irritability), and some of the memory and attention tests. These differences
were all statistically significant (p < 0.05 with Bonferoni correction), although group means did
not fall within a range that would be classified as "impaired". This set of analyses did not
distinguish between the effects of PCP, y-hexachlorocyclohexane, or other compounds, but
serological measures of these exposures (PCP, y-hexachlorocyclohexane, and p-hexachloro-
cyclohexane) were used in analyses of the correlation between specific exposures and the
neurological measures. Serum PCP level was inversely correlated (r ~ -0.65) with reading speed
and naming speed, and positively associated (r ~ 0.60) with error rates in the paired-association
test and the Benton visual retention test. These correlations were statistically significant
adjusting for age, and were stronger than those seen with y-hexachlorocyclohexane. In contrast,
the correlations seen with y-hexachlorocyclohexane were with measures of memory
performance. Exposure to p-hexachlorocyclohexane was not correlated with any of the effect
measures, and none of the exposures were correlated with the self-reported symptom data. This
small study provides data suggesting the types of neurobehavioral effects that may be seen in
chronic exposure to PCP.
The study of 293 sawmill and timber workers in New Zealand by McLean et al. (2009b),
described in Section 4.1.2.2, also included 17 neuropsychological symptoms in the interview
(Table 4-7), and a standardized neurological examination (Table 4-8). Adjusting for age (as a
continuous variable), gender, and smoking status (never, former, and current), heart palpitations
(OR 1.92, 95% CI 1.06-3.50) and unexplained sweating (OR 2.10, 95% CI 1.14-3.87) were
associated with PCP exposure; for heart palpitations, a stronger risk was seen in the higher
exposure group (Table 4-7). An association was also seen between exposure and straight leg
raising (OR 2.10, 95% CI 1.16-3.81), with a weaker association seen in the cranial nerve exam
(OR 1.64, 95% CI 0.94-2.88) (Table 4-8).
48
-------
Table 4-7. Prevalence of neuropsychological symptoms, and associations with PCP exposure, in 293 timber workers in
New Zealand
Symptoms
Short memory
Often need to take notes
Often go back to check things
Hard to get meaning from reading
Problem concentrating
Feel depressed
Abnormally tired
Less interested in sex
Heart palpitations
Feel an oppression in chest
Sweat with no reason
Headache at least once a wk
Painful tingling
Problem buttoning or unbuttoning
Trouble sleeping
Frequent mood changes
Bothered by noise more than in past
Non-exposed
(n = 177)
n
72
98
82
40
55
32
45
28
29
36
26
39
39
16
54
37
72
(%)
(40.9)
(55.7)
(46.6)
(22.7)
(31.3)
(18.2)
(25.6)
(15.9)
(16.5)
(20.5)
(14.8)
(22.2)
(22.2)
(9.1)
(30.7)
(21.0)
(40.9)
Exposed
(n = 116)
n
47
49
57
24
38
30
34
24
31
29
31
24
34
11
42
35
52
(%)
(40.5)
(42.2)
(49.1)
(20.7)
(32.8)
(25.9)
(29.3)
(20.7)
(26.7)
(25.0)
(26.7)
(20.7)
(29.3)
(9.5)
(36.2)
(30.2)
(44.8)
OR
1.02
0.60
1.16
0.73
0.97
1.57
1.24
1.40
1.92
1.26
2.10
0.86
1.31
1.05
1.28
1.52
1.11
(95% CI)
(0.62-1.68)
(0.37-0.97)
(0.72-1.89)
(0.40-1.32)
(0.58-1.64)
(0.88-2.82)
(0.72-2.13)
(0.75-2.63)
(1.06-3.50)
(0.71-2.25)
(1.14-3.87)
(0.47-1.56)
(0.75-2.28)
(045-2.43)
(0.77-2.14)
(0.86-2.69)
(0.68-1.81)
Low exposure"
(n = 58)
n
33
35
41
18
27
22
26
16
20
18
23
18
25
8
31
26
41
(%)
(38.4)
(40.7)
(47.7)
(20.9)
(31.4)
(25.6)
(30.2)
(18.6)
(23.3)
(20.9)
(26.7)
(20.9)
(29.1)
(9.3)
(36.1)
(30.2)
(47.7)
OR
0.97
0.58
1.15
0.76
0.94
1.58
1.30
1.26
1.65
1.02
2.15
0.90
1.37
1.07
1.30
1.64
1.29
(95% CI)
(0.56-1.69)
(0.34-0.99)
(0.67- 1.95)
(0.40- 1.45)
(0.53-1.67)
(0.84-2.97)
(0.73-2.33)
(0.63-2.53)
(0.85-3.19)
(0.53-1.97)
(1.12-4.16)
(0.47-1.72)
(0.75-2.50)
(0.43-2.69)
(0.74-2.26)
(0.88-3.04)
(0.76-2.20)
High exposure"
(n = 58)
n
14
14
16
6
11
8
8
8
11
11
8
6
9
3
11
9
11
(%)
(46.7)
(46.7)
(53.3)
(20.0)
(36.7)
(26.7)
(26.7)
(26.7)
(36.7)
(36.7)
(26.7)
(20.0)
(30.0)
(10.0)
(36.7)
(30.0)
(36.7)
OR
1.17
0.65
1.22
0.65
1.06
1.55
1.07
1.85
2.84
2.12
1.94
0.75
1.15
0.99
1.24
1.25
0.71
(95% CI)
(0.52-2.64)
(0.29-1.45)
(0.55-2.71)
(0.24-1.76)
(0.46-2.44)
(1.62-3.90)
(0.44-2.63)
(0.72-4.74)
(1.18-6.80)
(0.90-4.99)
(0.75-4.99)
(0.28-2.03)
(0.48-2.7)
(0.26-3.79)
(0.54-2.85)
(0.50-3.08)
(0.31,1.62)
""Cumulative exposure metric, based on product of intensity and duration; low score = 0-120; high score = >120. Similar patterns seen with the intensity score.
Sources: McLean et al. (2009b, 2007).
49
-------
Table 4-8. Prevalence of abnormalities seen in neurological examination, and associations with PCP exposure, in
293 timber workers in New Zealand
Test
Cranial nerves
Sensory exam by cotton wool
Sensory exam by pin prick
Vibration sense
Joint position
Two point discrimination
Wasting
Power upper limb
Power lower limb
Reflexes
Straight leg raising
Gait
Tests of coordination
Non-exposed
(n = 177)
n
46
19
20
13
4
50
4
5
7
35
28
4
8
(%)
(26.4)
(10.9)
(11.6)
(7.5)
(2.3)
(28.7)
(2.3)
(2.9)
(4.1)
(20.1)
(17.3)
(2.5)
(4.6)
Exposed
(n = 116)
n
39
11
11
6
o
3
36
2
1
1
16
32
2
3
(%)
(33.9)
(9.5)
(9.7)
(5.2)
(2.6)
(31.3)
(1.8)
(0.9)
(1.0)
(13.9)
(31.7)
(1.8)
(2.6)
OR
1.64
0.79
0.75
0.68
1.21
1.11
0.63
0.33
0.22
0.60
2.10
1.04
0.64
(95% CI)
(0.94-2.88)
(0.35-1.79)
(0.33-1.68)
(0.24-1.94)
(0.25-5.78)
(0.65-1.91)
(0.10-3.83)
(0.04-3.05)
(0.03-1.91)
(0.31-1.17)
(1.16-3.81)
(0.17-6.57)
(0.15-2.59)
Low exposure"
(n = 58)
n
27
7
7
4
2
29
0
1
0
13
21
1
1
(%)
(31.8)
(8.1)
(8.4)
(4.7)
(2.4)
(34.1)
(0.0)
(1.2)
(0.0)
(15.3)
(28.4)
(1.2)
(1.2)
OR
1.45
0.71
0.68
0.61
1.10
1.26
-
0.50
-
0.69
1.78
0.52
0.29
(95% CI)
(0.78-2.68)
(0.28-1.83)
(0.27-1.72)
(0.18-2.00)
(0.19-6.26)
(0.71-2.25)
-
(0.05-4.61)
-
(0.34-1.41)
(0.92-3.44)
(0.06-4.81)
(0.03-2.51)
High exposure"
(n = 58)
n
12
4
4
2
1
7
2
0
1
3
11
1
2
(%)
(40.0)
(13.3)
(13.3)
(6.7)
(3.3)
(23.3)
(6.9)
(0.0)
(3.3)
(10.0)
(40.7)
(3.3)
(6.7)
OR
2.35
0.97
0.92
0.93
1.78
0.74
2.74
-
0.82
0.38
3.28
1.64
1.64
(95% CI)
(0.97-5.68)
(0.29-3.22)
(0.28-3.04)
(0.18-4.78)
(0.18-17.3)
(0.28-1.90)
(0.38-19.7)
-
(0.09-7.68)
(0.11-1.36)
(1.32-8.11)
(0.17-15.8)
(0.30-8.99)
""Cumulative exposure metric, based on product of intensity and duration; low score = 0-120; high score = >120. Similar patterns seen with the intensity score.
Sources: McLean et al. (2009b, 2007).
50
-------
4.1.2.4. Studies of Reproductive Outcomes
Two studies examined reproductive outcomes in relation to exposure to PCP and/or
lindane in residences or places of work in Germany (Gerhard et al., 1999; Karmaus and Wolf,
1995). Karmaus and Wolf (1995) studied reproductive outcomes among daycare center workers
who were exposed at their place of work to wood preservatives. Because of concerns about
indoor air exposure to these chemicals, measurements of PCP concentrations in all daycare
centers in Hamburg were conducted by the government in 1986. In 24 centers, PCP
concentrations in the wood of > 100 ppm were found. Indoor air concentrations of PCP, lindane,
pentachlorodibenzo-dioxin, and pentachlorodibenzofuran were conducted in these centers. The
median concentrations in these samples were 0.25 ug/m3 for PCP, 0.2 ug/m3 for lindane, and
0.5 pg/m3 toxic equivalent factors for polychlorinated dibenzo-p-dioxins/dibenzofurans. Women
who worked in any of these daycare centers during a pregnancy and a comparison group of
women who had worked in other daycare centers were recruited through the employer's
insurance program. The study included 214 exposed women and 184 control women, with
49 pregnancies (32 live births) during an exposure period and 506 nonexposed pregnancies
(386 live births). The nonexposed pregnancies included pregnancies among exposed women that
did not occur while working at the place of exposure, and pregnancies among the controls.
Study participants completed an interview focusing on occupational, lifestyle, and reproductive
histories. Information on pregnancy outcomes, birth weight, and birth length was validated by
review of medical cards for a subgroup of 220 (59%) participants. In analyses excluding twins
and adjusting for age at conception and gestational age, employment at the high-exposure
daycare centers during pregnancy was associated with an approximately 220 g decrease in birth
weight and a 1.1 cm decrease in birth length.
Gerhard et al. (1999) conducted a study of 171 women who were referred to a
gynecological clinic in Germany because of infertility or other gynecological and/or endocrine-
related conditions to investigate possible effects of PCP exposure on the endocrine system.
Exposure status was based on serum levels of PCP, with the "exposed" defined as >20 ug/L
(n = 65). The other 106 women who served as controls (PCP levels <20 ug/L) were matched to
the exposed women on age, underlying condition, and geographical region. Gonadotropin and
estradiol analyses were based on blood samples taken on days 2-5 of the menstrual cycle, and
progesterone was based on two samples taken during the luteal phase of the cycle. Thyroid-
stimulating hormone (TSH) was measured in an unstimulated (baseline) sample and 30 minutes
after administration of 200 ug of thyrotropin-releasing hormone (TRH). Cortisol and various
androgen hormones were also measured with a baseline sample and after administration of
0.25 ug of adrenocorticotrophic hormone.
The median PCP level in the PCP group was 35.9 ug/L compared to 9.5 ug/L for the
controls. Small differences in follicle stimulating hormone (FSH) levels (median 5.9 and
6.9 niE/mL in exposed and controls, respectively, p = 0.0053) and triiodothyronine (Ts) (median
51
-------
0.98 and 1.02 ng/mL in exposed and controls, respectively,/? = 0.046) were observed. Euthyroid
goiters were found more frequently in the PCP group than the controls (50 versus 30%). There
was no difference in the baseline cortisol levels between the PCP and control groups, but a larger
increase was seen in the PCP group after adrenocorticotrophic hormone stimulation. Baseline
levels of testosterone and other androgens, and 17-hydroxypregnenolone and 17-hydroxy-
progesterone were lower in the PCP group, but there was no difference between the PCP and
control group in these hormone levels seen in response to the adrenocorticotrophic hormone
stimulation. This study showed that relatively high serum PCP levels in women are associated
with a number of endocrine effects, particularly related to androgen responsiveness, among
patients seen for infertility and endocrine disorders.
Dimich-Ward et al. (1996) conducted a nested case-control study of reproductive
outcomes among offspring of 9,512 male production and maintenance workers in the British
Columbia sawmill workers cohort described in Section 4.1.1, Studies of Cancer Risk).
Chlorophenates (primarily PCP and TCP) were used at the 11 sawmills in this study from 1950
to 1989, with TCP use increasing around the mid 1960s. These workers were the basis for the
large cohort study reported by Demers et al. (2006) of cancer risks described in Section 4.1.1.2
(Studies of Cancer Risk—Cohort Studies). Marriage and birth records were linked to identify
19,675 children born to these fathers between 1952 and 1988, and born after their father began
employment at the study sawmills. Cases of congenital anomalies were identified within these
children through the linking of these birth records to the British Columbia Health Surveillance
Registry. These outcomes were coded based on 3-digit ICD-9* revision categories. Other
reproductive outcomes selected for study were prematurity (born at <37 weeks gestation), low
birth weight (<2,500 g), small for gestational age (less than the 10* percentile of gestation-
specific weight based on British Columbia births), neonatal deaths (death of a liveborn infant
before the age of 1 year), and stillbirths (pregnancy of at least 28 weeks of gestation). For each
case of any of these outcomes, five controls were chosen matching to the year of birth of the
cases. Gender was an additional matching criterion for the congenital anomalies, and was used
as an adjustment variable for the other outcomes. Exposure assessments for each job title were
made by experienced workers for each mill for time periods characterized as having relatively
constant exposure. Each worker's exposure estimate was calculated by multiplying this
exposure constant by duration of employment in each job for each time period. The exposure
measures used in the analyses included a cumulative exposure estimate for each of three time
windows relative to time of conception (up to 3 months prior to conception, in the 3 months prior
to conception, through the period of pregnancy), and a measure of the maximum exposure
(hours/year) for any sawmill job up to 3 months prior to conception.
There was no association between any of the exposure measures and the risk of
premature birth, low birth weight, small for gestational age, neonatal death, or stillbirth.
Congenital anomalies of the eye (ICD-9* revision code 743, 22 cases) were associated with the
52
-------
cumulative exposure measure for each of the three time periods (but most strongly for the
measures limited to the 3 months prior to conception and to the pregnancy period). This was
seen when analyzed as a continuous variable/100 hours of estimated exposure (ORs 2.01 and
1.21 for the 3 months prior to conception and to the pregnancy period measures, respectively,
p < 0.005) and in analyses comparing the 75 percentile with the 25* percentile of exposure
(ORs 2.87 and 2.59 for the 3 months prior to conception and to the pregnancy period measures,
respectively). Further analyses indicated that strong associations were seen with congenital
cataracts (ICD-9 revision code 743.3, 11 cases). In the comparison of the 75* percentile with
the 25th percentile of exposure, the ORs for this outcome were 5.68 and 4.34 for the 3 months
prior to conception and to the pregnancy period measures, respectively. Weaker associations
(ORs around 1.3 in the analyses by percentile) were seen for spina bifida (ICD-9th revision code
741, 18 cases) and for anomalies of genital organs (ICD-9th revision code 752, 105 cases). The
strengths of this study include its large size and the specificity of the measured outcomes.
4.1.2.5. Summary of Studies of Noncancer Risk
Instances of PCP poisoning have been documented, indicating the potentially severe
consequences of acute, high-dose exposures. Few studies have examined the effects of the lower
exposures that occurred in occupational settings or through residential or environmental sources.
Many of the available studies are relatively small (<50 participants) (Peper et al., 1999; Triebig
et al., 1987; Klemmer et al., 1980; Begley et al., 1977) or may not be representative of the
exposed population (Gerhard et al., 1999; Walls et al., 1998). Despite these limitations, there are
indications of specific types of neurobehavioral effects seen with chronic exposure to PCP in
nonoccupational settings (Peper et al., 1999). A larger study of 293 former sawmill workers in
New Zealand also suggests neuropsychological effects and respiratory diseases (McLean et al.,
2009b). In addition, the large nested cohort study of reproductive outcomes in offspring of
sawmill workers (Dimich-Ward et al., 1996) indicates that specific types of birth defects warrant
additional research.
4.2. SHORT-TERM, SUBCHRONIC, AND CHRONIC STUDIES AND CANCER
BIOASSAYS IN ANIMALS—ORAL AND INHALATION
This section presents the available PCP toxicity studies that characterize the effects
associated with PCP exposure to animals via the oral and inhalation routes. Although studies
have been summarized and presented according to their route and duration of exposure, some of
the toxicity studies within the database have utilized various forms of PCP. During manufacture
of PCP, the chemical becomes contaminated with impurities. These impurities are other
chlorophenols, such as TCP, chlorinated dibenzo-p-dioxins, and chlorinated dibenzofurans.
Studies investigating the toxicity of PCP generally employ the technical grade, which is
composed of approximately 90% PCP and 10% of the various contaminants. The tPCP is
53
-------
frequently found under the trade names Dowicide 7, EC-7, DP-2, Duratox, Fungol, Penta-Kil,
and Permacide. Use of EC-7 and DP-2 are identified where possible; all other forms of the
technical grade of PCP will be referred to in the document as tPCP. To achieve an analytical
grade of PCP, an additional purification step to remove the contaminants that were
simultaneously created during the manufacturing of PCP is required. Although the use of the
aPCP is limited, there are several studies within the database that employ the relatively pure form
of the chemical (99% purity). Where possible, the type of PCP utilized within the studies has
been identified.
4.2.1. Oral Studies
4.2.1.1. Short-term Studies
Kerkvliet et al. (1982a) found that B6 mice treated with 1,000 ppm aPCP (average dose
estimated as 195 mg/kg-day) for 4 days exhibited no changes in body weight compared with
controls. Relative liver and spleen weights were significantly elevated 76 and 26%, respectively,
compared with controls.
NTP (1999) reported a 28-day toxicity study in groups of 10 male and 10 female F344N
rats administered aPCP (99% purity) in the diet at concentrations of 200, 400, 800, 1,600, or
3,200 ppm (average doses estimated as 20, 40, 75, 150, and 270 mg/kg-day, respectively). One
male and two females receiving 270 mg/kg-day died before the end of the study. Statistically
significant decreases in the final mean body weights of males and female rats were observed at
the two highest doses. Male body weights were reduced 14 and 47% at 150 and 270 mg/kg-day,
respectively. Females exhibited 19 and 43% reductions in mean final body weights at the
150 and 270 mg/kg-day concentrations, respectively. Decreased food consumption was
measured in male and females in the 150 and 270 mg/kg-day dose groups on day 1 and in males
in the 270 mg/kg-day dose group on day 28. It is possible that the reduction in food
consumption contributed to the decreased body weight at the two highest doses for both sexes.
Microscopic effects of aPCP administration were confined to the liver (hepatocyte degeneration
and centrilobular hypertrophy) and testes (degeneration of the germinal epithelium). The
incidence and severity of hepatocyte degeneration were statistically, significantly increased in
males receiving >40 mg/kg-day and in females receiving >75 mg/kg-day. The incidence of
centrilobular hypertrophy was significantly increased only at 270 mg/kg-day in both sexes.
Degeneration of the testicular germinal epithelium occurred in all males receiving 270 mg/kg-
day but in none of the control or lower dose group males. Mild to chronic active inflammation
was observed in the nasal sections of all control males and in some males of each dose group.
NTP (1999) did not determine no-observed-adverse-effect level (NOAEL) or lowest-observed-
adverse-effect level (LOAEL) values. The EPA determined that, for male rats, the NOAEL was
20 mg/kg-day and the LOAEL was 40 mg/kg-day, based on significant hepatocyte degeneration.
54
-------
In females, the NOAEL was 40 mg/kg-day and the LOAEL was 75 mg/kg-day, based on
significant hepatocyte degeneration.
In an NTP (1989) study, groups of male and female B6C3Fi mice were fed tPCP (90.4%
purity), EC-7 (91% purity), or aPCP (98.6% purity) for 30 days. There were 19 and 11 controls
for the male and female groups, respectively; 15 mice/group treated with tPCP and 5 mice/group
treated with EC-7 or aPCP. The administered doses corresponding to the dietary concentrations
of 20, 100, 500, 2,500, or 12,500 ppm PCP are estimated as 4, 19, 95, 593, or 5,367 mg/kg-day
for males and 5, 25, 126, 645, or 3,852 for females, respectively. Treatment-related effects
included clinical signs, increased mortality, decreased body weight gain, leukopenia, liver
toxicity, and induction of hepatic microsomal enzymes (Table 4-9). The data show that effects
occurred primarily at concentrations >95 mg/kg-day for males and 126 mg/kg-day for females;
however, liver lesions observed in one female mouse receiving 25 mg/kg-day aPCP are likely
treatment related. Effects other than those listed in Table 4-9 are discussed below. Statistical
analysis data were not reported for these effects. Rectal temperature was decreased by at least
1 degree in most groups of mice receiving all grades of PCP at 593 or 5,367 mg/kg-day in males
and 645 or 3,852 in females. Urine color ranged from yellow to dark brown in males and
females fed the mid and high doses of all PCP grades. Total liver porphyrins were increased in
males receiving all three grades and in females receiving tPCP and aPCP. Uncoupling of
mitochondrial oxidative phosphorylation (decreased phosphate:oxygen ratio) was observed at the
high dose of aPCP, at the low dose of tPCP, and at the lower doses of EC-7 (<593 mg/kg-day for
males or 645 mg/kg-day for females). The phosphate:oxygen ratio was increased at 593 mg/kg-
day for males and at 645 mg/kg-day for females. The study authors did not determine
NOAELs/LOAELs for the 30-day study. The EPA determined that the LOAELs were 95 mg/kg-
day for males with all three grades of PCP, based on dose-related increases in liver lesions
including hepatocyte degeneration and necrosis, centrilobular cytomegaly, karyomegaly, and
nuclear atypia. For females, the LOAELs were 126 mg/kg-day for tPCP based on dose-related
increases in liver lesions, 645 mg/kg-day for EC-7 based on liver lesions and decreased body
weight gain, and 25 mg/kg-day for aPCP based on liver lesions. The NOAELs were 19 mg/kg-
day in males for all grades and 25, 126, and 5 mg/kg-day in females for tPCP, EC-7, and aPCP,
respectively.
55
-------
Table 4-9. Comparison of the effects of three grades of PCP administered
continuously in feed to male (M) and female (F) B6C3Fi mice for 30 days
Effect3
tPCP (90.4% purity)
EC-7 (91.0% purity)
aPCP (98.6% purity)
Concentrations: 20, 100, 500, 2,500, 12,500 ppm; average doses
4, 19, 95, 593, 5,367 mg/kg-d for males and 5, 25, 126, 645, 3,852 mg/kg-d for females
Mortality
Clinical signs
Body weight
Liver weights
Serum enzymes
Serum y-glutamyl
transpeptidase
(y-GTP)
Hematology
Hepatic microsomal
enzymes
Liver lesionsb
LOAEL
NOAEL
14/19 (M), 7/15 (F) at
12,500 ppm
19/19 (M), 5/5 (F) at
12,500 ppm
9/19 (M), 1/5 (F) at 2,500 ppm
19/19 (M), 5/5 (F) at
12,500 ppm
2/19 (M) at 2,500 ppm
Weakness, lethargy, shallow breathing, severe weight loss, convulsions, and death at
12,500 ppm
Weight loss in both sexes,
12,500 ppm
Decreased weight gain (M),
2,500 ppm
Decreased weight gain (M) at
2,500 ppm
Decreased weight gain in both
sexes at 2,500 ppm
Absolute and relative weights statistically significantly increased at higher concentrations,
both sexes
ALP, cholesterol, ALT increased in all animals, both sexes
Greatly increased in both
sexes at 2,500 and
12,500 ppm
No treatment-related increase
Clinically significantly marked reduction in leukocyte count, primarily affecting lymphocytes
(M) and monocytosis (statistically significant in EC-7 females) in both sexes
Platelet count increased, both
sexes
No increase in platelet count
AHH activity increased for both sexes, dose-related for tPCP; P450 levels increased in both
sexes, dose-related for tPCP and aPCP
>500ppm, 100% of animals
of both sexes, more diffuse
and severe than with other
grades
500 ppm for both sexes
95 mg/kg-d (M), 126 mg/kg-
d(F)
100 ppm for both sexes
19 mg/kg-d (M), 25 mg/kg-d
(F)
>500 ppm (M, 40%),
>2,500 ppm (F, 100%)
500 ppm, 95 mg/kg-d (M),
2,500 ppm, 645 mg/kg-d (F)
100 ppm, 19 mg/kg-d (M),
500 ppm, 126 mg/kg-d (F)
>500 ppm (M, 100%),
>100ppm(F, 100%)
500 ppm, 95 mg/kg-d (M),
100 ppm, 25 mg/kg-d (F)
100 ppm, 95 mg/kg-d (M),
20 ppm, 5 mg/kg-d (F)
"Statistical analyses were not reported for all effects.
bCentrilobular cytomegaly, karyomegaly, nuclear atypia, degeneration, or necrosis.
AHH = Arylhydrocarbon hydroxylase; ALT = alanine aminotransferase
Source: NTP(1989).
Renner et al. (1987) reported on the toxicity of aPCP (99% purity) administered by
gavage to rats for 4 weeks followed by 2 weeks of recovery. Groups of 24 female Sprague-
Dawley rats (3 months old) were given 0.2 mmol/kg/day (53 mg/kg-day), 1 mL/day corn oil
(vehicle), or no treatment for the entire study duration. The results showed that body weights
were not significantly affected by treatment with aPCP. No clinical signs were observed, but
56
-------
three aPCP-treated animals died on day 28 or 32 of the study. Relative liver weight was elevated
during treatment, but returned to normal after treatment. Red blood cell (RBC), hematocrit, and
hemoglobin were decreased throughout treatment and showed no evidence of reversal during
recovery. The erythrocytes were polychromatic and anisocytotic in appearance. Microscopic
effects in the liver consisted of enlarged pleomorphic hepatocytes with degeneration of liver cells
and acidophilic bodies in the sinusoids. Statistical analysis was not reported. The EPA
determined that the LOAEL was 53 mg/kg-day (the only dose used), based on decreased RBCs,
hematocrit, and hemoglobin, and increased liver effects. The NOAEL could not be established
as effects were noted at the only dose administered.
In a study on 6-week-old pigs, tPCP (purity not reported; contained 4.7% TCP and
3.2 ppm total OCDDs and octachlorodibenzofurans) was administered, in capsules at doses of 5,
10, or 15 mg/kg-day, to groups of six pigs (sex not reported) for 30 days (Greichus et al., 1979).
No overt clinical signs or weight changes were noted in the tPCP-treated pigs compared with the
controls. RBC parameters evaluated at 15 and 30 days showed no significant changes from
controls. The white blood cell (WBC) count was significantly lower than control values for the
10 mg/kg-day dose group at 30 days and for the 15 mg/kg-day dose group at 15 and 30 days;
values were near the lower limits of the normal range. The only serum chemistry change
observed was significantly elevated BUN in the 10 and 15 mg/kg-day dose groups after 15 days
of treatment. The elevated BUN value, measured at study termination, for the 15 mg/kg-day
dose group did not achieve statistical significance. The relative liver weights were significantly
increased by 18 and 17% at 10 and 15 mg/kg-day, respectively. Histopathological findings in
the liver of tPCP-treated pigs consisted of nonspecific cloudy swelling of hepatocytes
accompanied by cellular enlargement, finely vacuolated cytoplasm, and decreased sinusoids.
The investigators did not include incidence or severity of liver lesions for individual dose groups.
Blood tPCP levels for all doses ranged from 63 to 71.5 ppm and from 67.6 to 78.1 ppm at 15 and
30 days of treatment, respectively, and no clear dose effect was observed. The highest tissue
levels were measured in the liver and kidney followed by the muscle. The study authors did not
determine NOAEL/LOAELs. The EPA determined that the LOAEL for pigs treated with tPCP
for 30 days was 10 mg/kg-day, based on significantly increased relative liver weight
accompanied by histopathological effects, significantly decreased WBC, and significantly
increased BUN. The NOAEL was 5 mg/kg-day. The short-term oral studies for PCP are
summarized in Table 4-10.
57
-------
Table 4-10. Summary of effects and NOAELs/LOAELs for short-term studies on PCP
Species, strain
Rat, F344
(10/sex/dose)
Rat, Sprague-Dawley
(24 females)
Mouse, B6C3FJ
(15/sex/dosefortPCP;
5/sex/dose for EC -7 and
aPCP)
Pig
(6/dose; sex not reported)
Dose (mg/kg-d), duration
20,40,75, 150, or 270
(feed)
28 d
53
(feed)
28 d
4, 19, 95, 593, or 5,367 (M)
(feed)
30 d
5,25, 126, 645, or 3,852 (F)
(feed)
30 d
5, 10, or 15
(capsule)
30 d
Grade/type
of PCP
aPCP
aPCP
tPCP
EC-7
aPCP
tPCP
EC-7
aPCP
tPCP
NOAEL (mg/kg-
d)a
20 (M)
40 (F)
Not established
19
25
126
5
5
LOAEL
(mg/kg-d)a
40 (M)
75 (F)
53
95
126
645
25
10
Effect
Hepatocellular degeneration.
Decreased RBC, hematocrit, and
hemoglobin. Polychromatic and
anisocytotic erythrocytes. Hepatocellular
degeneration, enlarged pleomorphic
hepatocytes, and acidophilic bodies in the
sinusoids.
Liver lesions including hepatocellular
degeneration and necrosis, centrilobular
cytomegaly and karyomegaly, and nuclear
atypia.
Increased relative liver weight, cloudy
swelling of hepatocytes, finely vacuolated
cytoplasm, decreased sinusoids,
significantly elevated BUN, and decreased
WBCs.
Reference
NTP (1999)
Renner et al.
(1987)
NTP (1989)
Greichus et al.
(1979)
aNOAELs and LOAELs determined by the EPA for these studies; values are for both genders unless otherwise specified.
58
-------
4.2.1.2. Subchronic Studies
In a 6-month study conducted by NTP (1989), groups of 25 male and 10 female B6C3Fi
mice received either tPCP (90.4% purity) at 200, 600, or 1,800 ppm; EC-7 (91% purity) at 200,
600, or 1,200 ppm; DP-2 (91.6% purity) at 200, 600, or 1,200 ppm; or aPCP (98.6% purity) at
200, 500, or 1,500 ppm for 26-27 weeks. The average administered doses are estimated to be
38 and 301 mg/kg-day for males and 52 and 163 mg/kg-day for females fed 200 and 600 ppm
tPCP, respectively. There was 100% mortality in the 1,800 ppm dose group and average doses
could not be estimated. In animals fed 200, 600, or 1,200 ppm EC-7, the average doses are
estimated for males as 36, 124, or 282 mg/kg-day and for females as 54, 165, or 374 mg/kg-day,
respectively. The estimated average doses for 200, 600, or 1,200 ppm DP-2 are 40, 109, or
390 mg/kg-day for males and 49, 161, or 323 mg/kg-day for females, respectively. Males and
females fed aPCP at dietary concentrations of 200, 500, or 1,500 ppm received estimated average
doses of 102, 197, or 310 mg/kg-day for males and 51, 140, or 458 mg/kg-day for females,
respectively. The estimated average dose administered to the low-dose group is approximately
twofold greater for those males fed aPCP than the other grades of PCP. The average doses were
estimated by the EPA, using the feed intake values reported by NTP (1989). Statistical analyses
were not reported for all effects.
Effects of administration of the four grades of PCP to mice for 6 months are summarized
in Table 4-11. All groups of female mice receiving each grade of PCP had significantly
increased absolute and relative liver weights. Groups of male mice receiving>38 mg/kg-day
tPCP, >102 mg/kg-day aPCP, >109 mg/kg-day DP-2, and 282 mg/kg-day EC-7 also had
significantly increased liver weights. Spleen weights were increased for all groups of male mice
except the low dose of each grade, while spleen weights were significantly decreased in females
at 163 mg/kg-day tPCP, 374 mg/kg-day EC-7, and 323 mg/kg-day DP-2. Thymus weights were
not significantly affected. Liver lesions consisting of karyomegaly, cytomegaly, hepatocellular
degeneration, and necrosis occurred in all males and females at all doses and grades of PCP.
Liver pigmentation was observed in at least 6-10 males and females administered all doses of
tPCP, the mid and high dose of DP-2 or EC-7, and the high dose of aPCP. Liver inflammation
was observed in 8-10 high-dose male mice receiving tPCP, DP-2, and aPCP and in the females
receiving tPCP. Bile duct hyperplasia occurred in all high-dose mice receiving tPCP. In
addition, degenerative changes in the spleen, bone marrow, thymus, and testes occurred in
animals that died before study termination. Effects observed with tPCP were generally more
severe than those observed with other grades; however, nasal lesions were seen only with aPCP
and EC-7. Other effects included dark urine color and elevated urine creatinine levels in high-
dose males administered each grade and dark urine color in high-dose females administered
EC-7 and aPCP. In contrast to the 30-day study, rectal temperature was not elevated and
leukocyte counts were not affected.
59
-------
Table 4-11. Comparison of the effects of four grades of PCP administered
continuously in feed to male (M) and female (F) B6C3Fi mice for 6 months
Effect3
Estimated
average dose
Mortality
Clinical signs
Final body
weights
Body weight
gain
tPCP (90.4% purity)
200, 600, l,800b ppm
M: 38, 301 mg/kg-d
F: 52, 163 mg/kg-d
100% (M, F) at
1,800 ppm; 0% at
lower doses
Piloerection, hunched
posture, enophthalmos,
thinness, weakness, and
inactivity at 1,800 ppm
No effect on survivors
No effect on survivors
EC-7 (91.0% purity)
200, 600, 1,200 ppm
M: 36, 124, 282 mg/kg-d
F: 54, 165, 374 mg/kg-d
1/10 (M) at 200 ppm; no
other mortality observed
None
11-13% decrease
| at 1,200 ppm (M, F)
DP-2 (91.6% purity)
200, 600, 1,200 ppm
M: 40, 109, 390 mg/kg-d
F: 49, 161, 323 mg/kg-d
2/10 (M) at 1,200 ppm;
no other mortality
observed
Piloerection, hunched
posture, enophthalmos,
thinness, weakness, and
inactivity at 1,200 ppm
No effect
| at 1,200 ppm (M)
aPCP (98.6%
purity)
200, 500, 1,500 ppm
M: 102, 197,
3 10 mg/kg-d
F: 51, 140,
458 mg/kg-d
2/20 (M) at 200 ppm;
no other mortality
observed
None
No effect
| at 1,500 ppm (M,
F)
Liver effects
ALT
AST
y-GTP
Hepatic
microsomal
AHH
induction
Hepatic P450
induction
Liver weight
Hepatocellular
lesions0
Liver pigment
Bile duct
hyperplasia
Dose-related, statistically significant t all animals, except EC -7 and DP-2 at 200 ppm
Significant t at
600 ppm (M, F)
No effects (not reported
for females)
200 and 600 ppm (M)
200 and 600 ppm
Significant t at 200 and
600 ppm (M, F)
No treatment-related t
No effects (not reported
for females)
1,200 ppm
1,200 ppm
Significant t at 1,200 ppm
(M); >200 ppm (F)
Significant t at
1,200 ppm (M)
Significant t at
>600 ppm (M)
All doses, maximum at
600 ppm
All doses
Significant t at 600 and
1,200 ppm (M);
>200 ppm (F)
Significant t at
1,500 ppm (F)
Significant t at
1,500 ppm (M)
1,500 ppm
1,500 ppm
Significant t all
doses (M, F)
All doses, less severe in females than in males
All doses (M, F)
All animals at
1,800 ppm
600 and 1,200 ppm (M, F)
No effect
600 and 1,200 ppm
(M,F)
No effect
1,500 ppm (M, F)
No effect
Other effects
Urinary
bladder
pigmentation
Nasal lesionsd
Spleen weight
Minimal severity at all doses, less severe in females than in males receiving EC-7 or aPCP
No effect
t all males except low-
dose group; significant
I in females at
600 ppm
>600 ppm (M); all doses
(F)
t all males except low-
dose group; significant J,
in females at 1,200 ppm
No effect
t all males except low-
dose group; significant J,
in females at 1,200 ppm
1,500 ppm (M); all
doses (F)
t all males except
low-dose group; no
effect in females
60
-------
Table 4-11. Comparison of the effects of four grades of PCP administered
continuously in feed to male (M) and female (F) B6C3Fi mice for 6 months
Effect3
LOAEL
NOAEL
tPCP (90.4% purity)
200, 600, l,800b ppm
EC-7 (91.0% purity)
200, 600, 1,200 ppm
DP-2 (91.6% purity)
200, 600, 1,200 ppm
aPCP (98.6%
purity)
200, 500, 1,500 ppm
200 ppm for all grades of PCP (approximately 38 mg/kg-d for tPCP, DP -2, and EC-7 and
102 mg/kg-d for aPCP males, respectively; approximately 52 mg/kg-d for all grades of PCP in
females, based on liver lesions observed in all groups of mice tested
None established; effects at all concentrations
"Statistical analyses not reported for all effects.
bAll animals in this group died and the estimated average doses could not be calculated.
°Cytomegaly, karyomegaly, degeneration, and necrosis.
dNasal mucosal metaplasia and goblet cell hyperplasia.
t = increase; J, = decrease; AST = aspartate aminotransferase
Source: NTP(1989).
The study authors did not determine the NOAELs/LOAELs for this subchronic study.
The EPA determined that the LOAELs were 49-54 mg/kg-day for females for all four grades of
PCP and at the low dose for males for all grades (36-40 mg/kg-day for tPCP, DP-2, and EC-7;
102 mg/kg-day for aPCP), based on dose-related increases in incidence and severity of liver
lesions including hepatocellular degeneration and necrosis, karyomegaly, and cytomegaly.
NOAELs were not established for males and females for any grade of PCP because liver toxicity
was observed at all doses for all grades.
Kerkvliet et al. (1982a) administered 50, 250, or 500 ppm tPCP (average doses estimated
as 10, 51, or 102 mg/kg-day) to groups of six Swiss-Webster female mice in the diet for 8 weeks,
followed by an 8-week recovery. Animals were sacrificed at 2-week intervals throughout
treatment and recovery. Additionally, groups of 15-16 B6 female mice were administered 50,
100, or 250 ppm aPCP (average doses estimated as 10, 20, or 49 mg/kg-day, respectively) for
8 weeks. No treatment-related effects were observed on body weights of either strain.
In the serial sacrifice study, relative liver weight, liver toxicity (hepatocyte swelling,
nuclear swelling and vacuolization with eosinophilic inclusions in nuclear vacuoles, and mild-to-
moderate multifocal necrosis), serum alanine aminotransferase (ALT), and lactate
dehydrogenase (LDH) levels in Swiss-Webster mice were elevated as early as 2 weeks after
treatment with 51 mg/kg-day tPCP. Complete recovery occurred by 4-6 weeks after treatment
was stopped. B6 mice exhibited significant increases in relative liver weight, liver toxicity, and
decreases in thymus weight at doses of >20 mg/kg-day. Liver weights were significantly
increased at the mid (13-18%) and high (34-57%) doses for both strains. Thymus weights were
reduced at the high dose for both strains, significantly for B6 mice at 49 mg/kg-day. The results
of this aPCP study showed that effects on the liver can be caused by PCP alone in the absence of
contaminants. The study authors did not determine the NOAELs/LOAELs. The EPA
61
-------
determined that the LOAEL was 51 mg/kg-day for the tPCP-treated Swiss-Webster mice and
20 mg/kg-day for aPCP-treated B6 mice, based on dose-related increases in incidence and
severity of multifocal necrosis, hepatocellular and nuclear swelling, hepatocellular vacuolization,
and eosinophilic inclusion bodies in nuclear vacuoles. The NOAEL was 10 mg/kg-day for both
tPCP- and aPCP-treated mice strains.
Kerkvliet et al. (1982b) reported that 20 male B6 mice/dose administered 50 or 500 ppm
(average doses estimated as 10 or 98 mg/kg-day) tPCP (86% purity) or aPCP (>99% purity) for
12 weeks showed no effects on growth rate, overt signs of toxicity, or microscopic changes in the
kidney, spleen, or adrenal gland. However, dose-related mild to marked hepatocyte swelling was
observed in the livers of animals exposed to both grades of PCP. Hepatocyte swelling, nuclear
swelling, and vacuolization with eosinophilic inclusions in nuclear vacuoles were observed at
10 and 98 mg/kg-day. Mild-to-moderate multifocal necrosis was observed at 98 mg/kg-day.
The EPA determined that the LOAEL was 10 mg/kg-day, based on dose-related increases in
hepatic effects. The NOAEL could not be determined as effects were noted at the lowest dose
tested.
In a study conducted by Knudsen et al. (1974), 10 Wistar rat weanlings/dose/sex were fed
diets containing 25, 50, or 200 ppm tPCP (average doses estimated as 2, 5, or 18 mg/kg-day for
males and 3, 5, or 21 mg/kg-day for females, respectively) for 12 weeks. The only biologically
significant effects were a dose-related increase in aniline hydroxylase in liver microsomes and
centrilobular vacuolation. Aniline hydroxylase activity was consistently increased at the low
dose in males and females at 6 and 12 weeks, and significantly elevated in the 18 mg/kg-day
male rats at 6 or 12 weeks and 21 mg/kg-day female rats at 6 weeks. The incidence of
centrilobular vacuolation was increased in male rats at 5 (4/10) and 18 mg/kg-day (5/10)
compared with 2/10 for the control and 0/10 for the 2 mg/kg-day group. The study authors
determined that the LOAEL for this study was 5 mg/kg-day based on statistically significant
increased incidence of liver effects; the NOAEL was 2 mg/kg-day for males and 3 mg/kg-day for
females.
Johnson et al. (1973) described a study in which Sprague-Dawley rats (number of rats not
reported) were fed diets containing three grades of PCP (described in general terms as
commercial, improved, or chemically pure) for 90 days. None of these grades contained TCDD.
The commercial PCP was 85-90% pure and contained 19 ppm hexachlorodibenzo-p-dioxin
(HxCDD) and 1,980 ppm OCDD, the improved PCP was 88-93% pure and contained 1 ppm
HxCDD and 26 ppm OCDD, and the chemically pure PCP (>99%) contained no detectable
levels of chlorinated dioxins. The specific contaminant congeners were not identified. Treated
rats received PCP at doses of 3, 10, or 30 mg/kg-day. There were no effects on body weight with
any of the three grades of PCP. Treatment with commercial PCP caused elevated serum ALP
levels and liver and kidney weights at all concentrations. Serum albumin was increased at
10 and 30 mg/kg-day while erythrocyte count, hemoglobin concentration, and hematocrit were
62
-------
depressed at 30 mg/kg-day. Microscopic liver lesions (minimal focal hepatocellular
degeneration and necrosis) were seen only at 30 mg/kg-day. The only effects observed after
administering improved PCP and chemically pure PCP were elevated liver weight at 10 and
30 mg/kg-day and elevated kidney weight at 30 mg/kg-day. Quantitative changes and statistical
analyses were not reported. The study authors did not determine NOAELs and LOAELs. The
EPA determined that the LOAELs were 3 mg/kg-day (lowest dose tested) for commercial PCP
based on dose-related elevated serum ALP and increased liver and kidney weight and 10 mg/kg-
day for improved and pure PCP based on increased liver weight. The NOAEL was 3 mg/kg-day
for improved and pure PCP, and could not be determined for commercial PCP.
Kimbrough and Linder (1975) reported light microscopic and ultrastructural effects in the
liver of male rats (strain not specified) administered 1,000 ppm tPCP or aPCP (average dose
estimated as 87 mg/kg-day) for 90 days. PCP treatment and control groups each consisted of
10 male rats. Statistical analysis was not reported. The liver was enlarged in all animals treated
with PCP. Light microscopy revealed foamy cytoplasm or pronounced vacuolation of
hepatocytes, single hepatocellular necrosis, cytoplasmic inclusions, slight interstitial fibrosis,
prominent brown pigment in macrophages, and Kupffer cells in the livers of rats fed tPCP.
Ultrastructurally, the smooth endoplasmic reticulum was increased, many lipid vacuoles were
present, and the mitochondria had an atypical appearance. In rats fed aPCP, the hepatocytes
were enlarged and many cells contained cytoplasmic inclusions; ultrastructurally, a slight
increase in smooth endoplasmic reticulum, some lipid vacuoles, and atypical mitochondria were
observed. This study showed that tPCP and aPCP cause similar ultrastructural effects in the
liver. The study authors did not establish a LOAEL or NOAEL. The EPA determined that the
LOAEL was 87 mg/kg-day for tPCP and aPCP, based on hepatocellular vacuolation, cytoplasmic
inclusion, slight interstitial fibrosis, brown pigment in macrophages and Kupffer cells, and
atypical mitochondria. A NOAEL could not be determined.
Deichmann et al. (1942) administered tPCP in the diet to groups of 10 rats at a dose of
5 mg/day in 8.5 g of food for 26 weeks or 3.9 mg/day in 13 g of food for 28 weeks. The
comparison group was not described. No growth occurred in rats administered 5 mg/day, and
the rats receiving 3.9 mg/day had body weights below normal. No gross findings were noted for
either group, and microscopic findings were considered insignificant.
Villena et al. (1992) examined the microscopic lesions in liver, kidney, and sciatic nerve
of rats receiving PCP (grade not specified) for varied treatment times. Groups (number not
reported) of male Wistar rats were given drinking water containing PCP at concentrations of
0.3 mM (80 mg/L) for 60 days, 1.0 mM (266 mg/L) for 60 or 90 days, 3.0 mM (800 mg/L) for
120 days, or drinking water without added PCP. The investigators did not describe effects in rats
given 80 or 266 mg/L PCP for 60 days. Microscopic effects in the liver at 266 mg/L for 90 days
or 800 mg/L for 120 days consisted of increased granular endoplasmic reticulum, hydropic
vacuolar degeneration, and total cell degeneration (necrosis), congested portal veins, enlarged
63
-------
and congested sinusoids, and bile duct hyperplasia. The nephritis in the kidneys occurred
primarily in the cortex and was characterized by glomerular congestion with thickening of the
capillary wall, glomerular hyalinization, and hyaline casts in the lumen of the proximal
convoluted tubules. The investigators noted that the kidney was more affected than the liver, and
the effects imply that destruction could progress to loss of function in the kidney. The
investigators did not state whether the animals were treated with free tPCP, aPCP, or sodium
salts. This specific information is important considering that PCP has low solubility in water
(80 mg/L) (Budavari et al., 1996), while the sodium salt is freely soluble in water. Additionally,
effects on body weight, food and water consumption, or clinical signs were not described. The
authors did not establish a NOAEL or LOAEL. Based on the data presented in the report, the
EPA determined that the NOAEL was 80 mg/L and the LOAEL was 266 mg/L, based on dose-
related increases in severity of liver and kidney toxicity.
Deichmann et al. (1942) reported no deaths or signs of toxicity in a group of 23 rabbits
given 3 mg/kg of tPCP as a 1% aqueous solution (dosing method not reported) for 90 successive
doses except on Sundays. In another study by Deichmann et al. (1942), five rabbits were
administered tPCP orally at a dose of 35 mg/kg-day as a 0.5% solution for 15 days followed by a
5% solution to gradually increase the dose to 600 mg/kg-day (twice the lethal dose) during the
next 19 days. All animals died, one after ingesting a total dose of 1.9 g, two after ingesting 2.9 g,
and two after ingesting 3.9 g. Effects attributed to tPCP administration included weight loss and
anemia.
McConnell et al. (1980) administered either 100% aPCP, 10% tPCP/aPCP mix, 35%
tPCP/aPCP mix, or 100% tPCP to groups of three yearling (10-14 months old) Holstein cattle to
determine the effect of contaminants on PCP toxicity. The purity of PCP was not reported. Each
treatment group was given 647 ppm PCP in feed (20 mg/kg) for 42 days, which was then
decreased to 491 ppm (15 mg/kg) for the remaining 118 days of the study (total treatment time =
160 days). A group of three yearlings served as controls. The diet containing 100% tPCP
produced more untoward effects than that of the 100% aPCP diet. Growth and feed efficiency
were depressed by all PCP treatments but more severely by tPCP. The general appearance of
tPCP-treated yearlings was unthrifty toward the end of the study. Yearlings receiving tPCP had
a number of clinical and pathological abnormalities including anemia, increased hepatic mixed
function oxidase and y-glutamyl transpeptidase (y-GTP) activities, increased relative liver and
lung weights, thymus atrophy, and marked villous hyperplasia of the urinary bladder mucosa,
which extended into the renal pelvis, renal papillae, and terminal portions of the collecting ducts
(most striking lesion). Additionally, the yearlings exhibited signs of hyperplasia of the gall
bladder and bile duct mucosa, hyperkeratosis of ductal lining and dilated ducts containing
keratinaceous material in the Meibomian glands in the eyelid, and hyperkeratosis of the skin.
Many of these effects can be associated with exposure to dioxin and/or furan contaminants in
PCP and were dose-related with respect to tPCP (i.e., the effects were more severe in cattle given
64
-------
100% tPCP). In the 100% aPCP group, effects were limited to decreased concentrations of
serum T3 and thyroxine (14) and increased arylhydrocarbon hydroxylase (AHH) activity.
Kinzell et al. (1981) reported on the treatment of four lactating Holstein dairy cattle
(6 weeks postpartum) with dietary tPCP (85-90% purity). Cattle were given a dose of
0.2 mg/kg-day for 75-84 days followed by 2 mg/kg-day for an additional 56-60 days (total
treatment time, 131-144 days). tPCP administration had no effect on body weight, food
consumption, hematology, clinical chemistry, or urinalysis tests. Relative organ weights for
liver, lung, kidney, and adrenals were increased by 23-27% compared with control (n = 4)
weights; gross and microscopic lesions were observed in the kidney (chronic diffuse interstitial
nephritis) and urinary bladder (thickening of bladder wall). In vitro tests revealed impairment of
kidney function (decreased PAH, tetraethyl ammonium, and a-aminoisobutyrate uptake). These
kidney effects were also observed in younger Holstein calves and attributed to PCP and not the
contaminants (Hughes et al., 1985). No histopathologic effects attributable to tPCP were
observed in the liver.
Hughes et al. (1985) fed tPCP (85-90% purity) or aPCP (99.02% purity) to 15 Holstein
bull calves (7 days old) twice daily at doses of 0, 2, or 20 mg/kg-day. One calf in each of the
high-dose groups fed aPCP or tPCP died after acute toxicity (elevated temperature, rapid
respiration, severe diarrhea, acute purulent pneumonia). After 5 days, the doses of 2 and
20 mg/kg-day were lowered to 1 and 10 mg/kg-day, respectively, and treatment was continued
for total treatment duration of 42 or 43 days. Severe toxic effects occurred following PCP
administration, primarily in calves receiving tPCP. One calf treated with 10 mg/kg-day was
moribund at the time of necropsy. Body weight gain, measured up to day 35 of treatment, was
decreased in the 10 mg/kg-day dose groups when compared to that of controls. Body weight
gain was decreased by 80 and 41% in calves receiving 10 mg/kg-day tPCP and aPCP,
respectively. The overall marked decrease in weight was due primarily to a 93% decrease in
weight gain for tPCP-treated calves relative to controls between days 20 and 35; the decrease for
aPCP-treated calves was only 17%. Calves receiving 1 mg/kg-day of tPCP or aPCP gained
slightly less weight than controls. During the last 3 weeks of treatment, tPCP-treated calves
consumed only 15% as much grain as controls.
Thyroid hormone levels in serum were measured during the first 35 days of treatment.
Serum T3 levels were statistically significantly reduced by 58-69% after treatment with
10 mg/kg-day tPCP and 49-55% with 10 mg/kg-day aPCP. Treatment with 1 mg/kg-day
reduced serum T3 levels 44-56% with tPCP and 22-27% with aPCP. Reductions of 37-58 and
25% were observed in the calves' serum T4 levels following treatment with 1 mg/kg-day tPCP
and aPCP, respectively. T3 and T4 responsiveness to the TRH challenge were not affected by
treatment with either grade. Organ weights most notably affected by PCP treatment were
thymus and spleen in calves treated with 10 mg/kg-day tPCP or aPCP. The thymus weight was
reduced by 83% with tPCP and 54% with aPCP. Microscopic lesions consistent with thymus
65
-------
atrophy (cortical atrophy) were observed in tPCP-treated calves. Spleen weights were reduced
by 52% with 10 mg/kg-day tPCP and by 32% with 10 mg/kg-day aPCP. Squamous metaplasia
was observed in the Meibomian gland of the eyelid of the three calves treated with 10 mg/kg-day
tPCP, but in none of the calves treated with aPCP. The investigators attributed the eye effects to
contaminants in PCP and not PCP itself. Statistically significantly elevated serum gamma-
glutamyl transferase was observed with tPCP at 10 mg/kg-day. A decrease in serum protein
concentration was noted at 10 mg/kg-day for both tPCP and aPCP.
In vitro tests to examine kidney function by observing p-aminohippurate and tetraethyl
ammonium uptake indicated that 10 mg/kg-day PCP and not the contaminants impaired these
energy-dependent functions. During treatment, Hughes et al. (1985) measured plasma PCP
levels in calves. PCP levels rapidly increased then plateaued between 5 and 10 days. No
difference was observed between the maximum plasma levels attained with tPCP and aPCP,
although there were dose-related differences. The plasma PCP concentrations leveled off at
approximately 100 ppm in calves given 10 mg/kg-day and at approximately 13-14 ppm in calves
given 1 mg/kg-day. The PCP level in the plasma of control calves did not exceed 1 ppm. The
authors did not establish NOAEL/LOAEL values. The EPA determined a NOAEL of 1 mg/kg-
day and a LOAEL of 10 mg/kg-day, based on decreased body weight gain, significantly elevated
serum gamma glutamyl transferase, decreased serum protein concentration, significantly
decreased TS and T4 levels, and decreased kidney function. The subchronic studies for PCP are
summarized in Table 4-12.
66
-------
Table 4-12. Summary of NOAELs/LOAELs for oral subchronic studies for PCP
Species, strain
Mouse, Swiss-
Webster
(6 females/dose)
Mouse, B6
(15-16 female
mice/dose)
Mouse, B6
(20 males/dose)
Rat, Wistar
weanlings
(10/sex/dose)
Rat, Sprague-Dawley
(number not
reported)
Rat (10 males/dose)
Rat, male Wistar
(number not
reported)
Mouse, B6C3FJ
(25 males/dose;
10 females/dose)
Dose (mg/kg-d),
duration
10, 51, or 102
(feed)
8wks
10, 20, or 49
(feed)
8wks
10 or 98
(feed)
12wks
2, 5, or 18 (M)
3,5,or21(F)
(feed)
12wks
3, 10, or 30
(feed)
90 d
87
(feed)
90 d
80, 266, or 800 mg/L
(drinking water)
60-120 d
38 or 301 (M)
(feed)
26-27 wks
52 or 163 (F)
(feed)
26-27 wks
36, 124, or 282 (M)
(feed)
26-27 wks
54, 165, or 374 (F)
(feed)
26-27 wks
40, 109, or 390 (M)
(feed)
26-27 wks
49, 161, or 323 (F)
(feed)
26-27 wks
102, 197, or 3 10 (M)
(feed)
26-27 wks
51, 140, or 458 (F)
(feed)
26-27 wks
Grade/type
of PCP
tPCP
aPCP
tPCP
aPCP
tPCP
Commercial
Improved
Pure
tPCP
aPCP
Not reported
tPCP
EC-7
DP-2
aPCP
NOAEL
(mg/kg-d)
10
10
Not established
2
3
Not established
3
3
Not established
80
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
LOAEL
(mg/kg-d)
51
20
10
5
5
3
10
10
87
266
38 (M)
52 (F)
36 (M)
54 (F)
40 (M)
49 (F)
102 (M)
51 (F)
Reference
Kerkvliet et al.
(1982a)a
Kerkvliet et al.
(1982b)a
Knudsen et al.
(1974)
Johnson et al.
(1973)a
Kimbrough and
Linder (1975)a
Villena et al.
(1992)a
NTP (1989)a
aNOAELs and LOAELs determined by the EPA for these studies; values are for both genders unless otherwise
specified.
67
-------
4.2.1.3. Chronic Studies—Noncancer
In a chronic toxicity study in dogs (Mecler, 19961), tPCP (90.9% purity) was fed by
gelatin capsules to four beagle dogs/sex/dose at 0, 1.5, 3.5, or 6.5 mg/kg-day for 52 weeks. At
6.5 mg/kg-day, one male and one female dog were sacrificed in extremis on days 247 and 305,
respectively, due to significant clinical toxicity (significant weight loss, lethargy, marked
dehydration, vomiting, icterus). The morbidity was presumed due to hepatic insufficiency based
on profuse toxicity in the liver that consisted of histologic lesions; multifocal, moderate
hepatocellular swelling and degeneration of hepatocytes; fibrosis; bile duct hyperplasia; foci of
hepatocellular hypertrophy; and hyperplasia consistent with cirrhosis. The mean body weight in
surviving males in the 6.5 mg/kg-day dose group was decreased 18% when compared with
controls. The decrease in body weight was not considered statistically significant as calculated
by the study authors. Absolute body weight was only slightly decreased at the lower doses
(4 and 6% at 1.5 and 3.5 mg/kg-day, respectively). Female dogs in the 6.5 mg/kg-day dose
group exhibited a 20% decrease in absolute body weight that was statistically significantly less
than controls at week 13 and for the remainder of the study. At the lower doses of 1.5 and
3.5 mg/kg-day, the absolute body weights of females were decreased 9 and 13%, respectively.
In contrast to males, the decrease in absolute body weight in treated females was dose-related.
Only group means were reported; individual animal data and standard deviations were not
included.
There were dose-related, mild-to-moderate decreases in three hematological parameters
measured in male dogs for all dose groups, although not all changes were considered statistically
significant (in calculations performed by study authors). Statistically significant decreases (15%)
in red cell counts were observed in males at the 3.5 mg/kg-day dose, while the 1.5 mg/kg-day
group showed only a 3% decrease. In males at the 6.5 mg/kg-day dose, RBC counts and
hemoglobin levels were statistically significantly reduced by 21 and 16%, respectively,
compared with controls. In females, statistically significant decreases of 10-17% in these
hematological parameters were observed at 6.5 mg/kg-day from week 26 until study termination.
In contrast to males, the hematological effects in females were not dose-related.
Activities of ALP, aspartate aminotransferase (AST), and ALT were elevated for both
sexes throughout the study. At study termination, ALP activity was increased, compared with
controls, in the serum of males (1.9-, 2.3-, and 4.9-fold) and females (1.9-, 2.6-, and 6.8-fold) at
all three doses (1.5, 3.5, and 6.5 mg/kg-day, respectively). AST activity increased slightly at
doses >3.5 mg/kg-day, although never >1.7-fold greater than in controls. The serum activity of
ALT was similar to the control at 1.5 mg/kg-day, although ALT activity was observed at levels
2.8- and 3.1-fold greater than the controls for males and females, respectively, in the 3.5 mg/kg-
lrThis study was submitted to the Agency as part of the process for the development of the Reregistration Eligibility
Decision (RED) document by the U.S. EPA's Office of Pesticide Programs (OPP). Mecler (1996) satisfied the
guideline requirements (OPPTS 870.4100) for a chronic toxicity study in non-rodents and is classified as an
"acceptable" Good Laboratory Practice (GLP) study.
68
-------
day dose group. Exposure to 6.5 mg/kg-day of PCP resulted in ALT levels 3.9- and 8.8-fold
greater than in controls for males and females, respectively.
Male dogs exhibited increases of 10, 31, and 32% over controls in measurements of
absolute liver weight at the 1.5, 3.5, and 6.5 mg/kg-day dose levels, respectively; these were not
considered statistically significant by the study authors. However, increases of 14, 39, and 66%
in relative liver weights of males were significantly greater than in controls in the 1.5, 3.5, and
6.5 mg/kg-day dose groups, respectively. Absolute and relative liver weights were significantly
elevated at 1.5, 3.5, and 6.5 mg/kg-day doses in females by 24, 22, and 49% (absolute liver
weights) and 37, 40, and 94% (relative liver weights), respectively. Thyroid weight
measurements in males were increased when compared with controls, but did not show a linear
dose-response relationship. Absolute and relative thyroid weights were statistically significantly
increased in females at the 6.5 mg/kg-day dose by 78 and 138%, respectively. Relative thyroid
weight was also increased at the 1.5 (72%) and 3.5 mg/kg-day (64%) doses.
An increased incidence of gross stomach lesions consisting of multiple, raised mucosal
foci were observed in all treated groups (1.5, 3.5, and 6.5 mg/kg-day) of male (2/4, 3/4, and 2/3,
respectively, versus 0/4 in controls) and female (2/4, 4/4, and 2/3, respectively, versus 1/4 in
controls) dogs. Male dogs exhibited dark, discolored livers in 1/4, 1/4, and 3/3 dogs, while 3/4,
3/4, and 2/3 females exhibited the discolored livers in the 1.5, 3.5, and 6.5 mg/kg-day treatment
groups, respectively. Microscopically, liver lesions associated with tPCP treatment consisted of
pigmentation, cytoplasmic vacuolization, minimal necrosis, and chronic inflammation; incidence
and severity generally increased with dose. The incidence and severity of the liver lesions in
male and female dogs are shown in Table 4-13. The authors noted that the pigmentation was
approximately 2-4 microns in diameter and segregated near the cytoplasmic membrane. The
pigment was sometimes observed in the regions of canaliculi of adjacent hepatocytes, and less
frequently in the cytoplasm of Kupffer cells and histiocytes within periportal regions. The
authors considered the pigment consistent with lipofuscin (LF), noting that biotransformation of
chlorinated phenolic compounds occurs via CYP450 enzymes, during which time lysosome-
related peroxidation of intracellular lipids produces LF pigment. The study authors determined
that the LOAEL was 6.5 mg/kg-day tPCP, based on morphologic effects in the liver. The
NOAEL was 3.5 mg/kg-day. However, considering the progression of lesions observed with
increasing dose and the morbidity observed in both sexes at the 6.5 mg/kg-day dose, the EPA
determined that the LOAEL was 1.5 mg/kg-day (lowest dose tested), based on liver pathology
consisting of dose-related increases in incidence and severity of hepatocellular pigmentation,
cytoplasmic vacuolation, and chronic inflammation, and significant increases in relative liver
weight and increases in absolute liver weight (significant in females), and increased serum
enzyme activity. The NOAEL could not be established.
69
-------
Table 4-13. Liver histopathology, incidence, and severity in dogs exposed to
tPCP
Dose
(mg/kg-d)
Number examined
Females
0
4
1.5
4
3.5
4
6.5
3
Males
0
4
1.5
4
3.5
4
6.5
3
Lesion3
Pigment
Cytoplasmic vacuolization
Minimum necrosis
Chronic inflammation
0
3(1)
0
2(1)
4 (2.3)
3(2)
0
2(1.5)
4 (2.8)
4 (2.3)
0
4(1.8)
3 (3.3)
3 (3.3)
2(1)
3 (1.7)
0
1(3)
0
0
4(3)
1(2)
0
4(1)
4(3)
4 (2.8)
0
4(1.3)
3 (3.3)
3 (3.3)
1(1)
3 (1.3)
aThe values in parentheses are grades of severity for the lesion: 1 = minimal; 2 = mild; 3 = moderate; 4 = marked.
Source: Mecler (1996).
In a study conducted by NTP (1989), groups of 50 B6C3Fi mice/sex/dose were
administered feed containing 100 or 200 ppm tPCP (90.4% purity) or 100, 200, or 600 ppm EC-7
(91% purity) continuously for 2 years. Two groups of mice (35 animals/sex) were maintained on
untreated feed to serve as controls. The average administered dose in the treated feed was
calculated as 18 or 35 mg/kg-day for males and 17 or 35 mg/kg-day for females for the 100 or
200 ppm dose groups, respectively, for tPCP or 18, 37, or 118 mg/kg-day for males, and 17, 34,
or 114 for females for the 100, 200, or 600 ppm dose groups, respectively, for EC-7. Both tPCP
and EC-7 contain approximately 90% PCP, but different levels of contaminants. The average
daily PCP and contaminant doses associated with each dietary concentration are summarized in
Table B-3 in Appendix B. Mean body weights of male and female mice receiving either tPCP or
EC-7 were similar to control weights throughout the study with one exception. Female mice
receiving 114 mg/kg-day EC-7 weighed 78-91% of the control weights during the second year
of the study. No statistically significant effects were observed on survival in either male or
female mice receiving tPCP or EC-7, although the survival rate of tPCP male controls was
abnormally low (34%) at the end of the study.
This study showed that the liver was the primary target for systemic toxicity for both
grades of PCP and in both sexes. The following liver lesions occurred at statistically significant
higher incidences in PCP-treated males at all doses of tPCP and EC-7 than in the control: clear
cell focus, acute diffuse necrosis, diffuse cytomegaly, diffuse chronic active inflammation,
multifocal accumulation of brown pigmentation (LF and cellular debris) in Kupffer cells, and
proliferation of hematopoietic cells (extramedullary hematopoiesis). Males also had a
significantly higher incidence of bile duct hyperplasia at both doses of tPCP, but only at the
114 mg/kg-day dose of EC-7. Females receiving all doses of tPCP and EC-7 exhibited
incidences of the following liver lesions that were significantly higher than controls:
cytomegaly, necrosis, inflammation, and pigment accumulation. In addition, the incidence of
70
-------
clear cell focus was significantly increased compared with controls in females treated with
17 mg/kg-day tPCP and 34 and 114 mg/kg-day EC-7. The incidence of extramedullary
hematopoiesis was higher in females exposed to 35 mg/kg-day tPCP and all doses of EC-7 when
compared with that in controls. In contrast to males, the female mice did not exhibit a significant
increase in bile duct hyperplasia with tPCP, although the hyperplasia was significantly higher in
females treated with 114 mg/kg-day EC-7. This was the only lesion that the investigators related
solely to the impurities within PCP.
Other treatment-related nonneoplastic findings were observed in the spleen and nose of
male and female mice and in the mammary glands of females. The incidence of extramedullary
hematopoiesis in the spleen was significantly higher in tPCP males at 18 and 35 mg/kg-day and
in females at 35 mg/kg-day. Acute focal inflammation of the mucosal gland and focal
metaplasia of the olfactory epithelium were increased in male (118 mg/kg-day) and female mice
(114 mg/kg-day) receiving EC-7; these lesions did not occur in any mouse receiving tPCP. In
tPCP females, the incidence of cystic hyperplasia of the mammary gland was significantly higher
at 35 mg/kg-day (59%) than in tPCP controls (23%) but not when compared with the EC-7
control (58%). Therefore, this lesion was not considered related to treatment by investigators.
Under the conditions of these studies, tPCP and EC-7 were equally effective in male mice except
for induction of bile duct hyperplasia. In female mice, tPCP was generally more effective than
EC-7 except for induction of bile duct hyperplasia and nasal lesions. The study authors did not
determine LOAELs/NOAELs. The EPA determined that the LOAELs were 18 mg/kg-day for
males and 17 mg/kg-day for females for both tPCP and EC-7, based on statistically significant
increases in liver lesions. NOAELs could not be established for either tPCP or EC-7, because
effects in the liver occurred at the lowest doses tested in male and female mice. Some findings
occurred at incidences approaching 100% at 100 ppm (17-18 mg/kg-day), indicating that a lower
dose could have been tested and the potential for low-dose toxicity exists.
In a chronic toxicity study, Schwetz et al. (1978) administered EC-7, a commercial-grade
PCP (91% purity), in the diet of male and female Sprague-Dawley rats at doses of 0, 1,3, 10, or
30 mg/kg-day. Treated or control diets were fed to males for 22 months and females for
24 months. Each group consisted of 25 rats of each sex. Statistical analysis was not reported.
No treatment-related effects were observed for clinical signs, food consumption, survival,
hematological parameters, or organ weights. The investigators stated that mean body weights of
high-dose females were significantly less than those of controls during most of the study. Serum
ALT activity was slightly increased (<1.7-fold) in both sexes at the highest dose when measured
at study termination. Histopathological examination showed pigment accumulation in the
centrilobular hepatocytes of the liver in 30% of females given 10 mg/kg-day and in 59% of
females given 30 mg/kg-day. Similarly, 26 and 70% of females receiving 10 and 30 mg/kg-day
EC-7 exhibited pigment accumulation in the epithelial cells of the proximal convoluted tubules
in the kidney. This effect was not detected in the females of the lower dose or control groups.
71
-------
Only 1 of the 27 male rats given EC-7 (30 mg/kg-day) exhibited the brown pigment in
hepatocytes. The study authors determined that the LOAEL was 30 mg/kg-day for males and
10 mg/kg-day for females, based on dose-related increased pigment accumulation in the liver and
kidney. The NOAELs were 10 mg/kg-day for males and 3 mg/kg-day for females.
Kimbrough and Linder (1978) compared the effect of tPCP (84.6%) and aPCP (>99%)
fed to male and female Sherman rats for 8 months, observing that effects following
administration of tPCP were more severe than those of aPCP. PCP was administered at
concentrations of 20, 100, or 500 ppm (average doses estimated as 2, 9, or 44 mg/kg-day for
males and 2, 10, or 48 mg/kg-day for females, respectively). No signs of mortality were
observed with either tPCP or aPCP. Final body weights were significantly reduced 15-16% for
both male and females fed the high dose of tPCP and 5 and 10% for females and males,
respectively, fed the high dose of aPCP. Dose-related effects were observed in the liver,
particularly in rats fed tPCP (effects were described qualitatively; the quantitative changes were
not reported). Liver weights were elevated in both sexes (statistically significant in the males) at
the high dose of tPCP. Animals treated with 44 (males) or 48 mg/kg-day (females) tPCP
exhibited liver toxicity (statistical analyses not reported), manifested by periportal fibrosis,
hepatocyte hypertrophy, vacuolation, pleomorphism, bile duct proliferation, adenofibrosis
(cholangiofibrosis), cytoplasmic hyaline inclusions, and abundant brown pigment in
macrophages and Kupffer cells (porphyria) in one or both sexes. At 9 (males) or 10 mg/kg-day
(females) tPCP, similar but less severe effects than those observed at the high doses were
observed, although adenofibrosis and bile duct proliferation did not occur at this dose. A small
neoplastic nodule was observed in the liver of one mid-dose female rat. At the lowest dose of
2 mg/kg-day tPCP, slight hepatocyte hypertrophy and vacuolation were observed in all males
and one female. In rats administered aPCP at doses of 44 (males) and 48 mg/kg-day (females),
effects in the liver included slight hepatocyte hypertrophy, eosinophilic cytoplasmic inclusions,
and brown pigment in macrophages in animals of one or both sexes. There were no effects
observed in rats treated with the two lower doses of aPCP. The EPA determined that the
LOAELs were 2 mg/kg-day (lowest dose tested) for tPCP and 44 mg/kg-day in males and
48 mg/kg-day in females for aPCP, based on dose-related increases in incidence and severity of
liver effects and statistically significant decreases in body weight. The NOAEL could not be
determined for tPCP. The NOAELs were 9 and 10 mg/kg-day for males and females,
respectively, for aPCP.
NTP (1999) examined groups of 50 F344 rats/sex/dose administered aPCP (99% purity,
with no detectable levels of chlorinated dibenzo-p-dioxin, dibenzofuran, diphenyl ether, or
hydroxydiphenylether) in feed at concentrations of 0, 200, 400, or 600 ppm (average doses of 0,
10, 20, or 30 mg/kg-day, respectively) for 105 weeks. In an additional stop-exposure study,
groups of 60 rats/sex were maintained on feed containing 1,000 ppm aPCP (average dose of
60 mg/kg-day) for 52 weeks followed by untreated feed until study termination at 2 years. This
72
-------
study was also reported by Chhabra et al. (1999). Survival rates of male rats receiving
30 mg/kg-day for 2 years or 60 mg/kg-day for 52 weeks significantly exceeded those of controls
(62 or 64%, respectively, versus 24% for controls), while survival of the other groups was
similar to that of controls. Mean body weights were decreased in both male and female rats at
various times during the study. Mean body weights were 94, 91, 89, and 82% of the control
weights in males and 94, 91, 84, and 78% of the control weights in females receiving 10, 20, 30,
and 60 mg/kg-day aPCP, respectively. In the stop-exposure study, body weights recovered to
within 4% of the control weight after treatment stopped at 52 weeks.
The liver was the primary target for nonneoplastic toxicity, particularly in male rats. The
incidence of cystic degeneration was significantly increased at 20 (56%) and 30 (78%) mg/kg-
day. In addition, the incidence of hepatodiaphragmatic nodules was significantly increased in all
groups of males receiving aPCP (10-16 versus 0% for controls), although no clear dose-response
was observed. Hepatodiaphragmatic nodules were described as developmental anomalies
commonly observed in F344 rats; therefore, the increased incidence observed in this study was
not considered related to exposure to aPCP. The incidences of liver lesions in female rats in the
2-year study were similar to or significantly lower than those of controls (cytoplasmic hepatocyte
vacuolation in 2 versus 14% for controls).
Interim evaluation (7 months) of the stop-exposure group exhibited significantly elevated
(20-90%) serum ALP levels in males and sorbitol dehydrogenase levels in males and females
compared with control levels. The ALT level in males was elevated by 46%, but this was not
statistically significant as calculated by the investigators. Microscopic examination of 60 mg/kg-
day rats, sacrificed at 7 months, showed significantly higher incidences of centrilobular
hepatocyte hypertrophy in both male and female rats (60%) and cytoplasmic hepatocyte
vacuolization in male rats (80%) compared with the controls (0%). These microscopic lesions
were also observed in male and female rats of the 2-year study; however, incidences were not
significantly increased. The 60 mg/kg-day males exhibited a significantly greater incidence,
compared with controls, of liver lesions consisting of chronic inflammation (64 versus 44% for
controls), basophilic focus (62 versus 34% for controls), and cystic degeneration of hepatocytes
(56 versus 32% for controls). The study authors did not determine LOAELs and NOAELs. This
study showed that male rats were more susceptible to aPCP exposure than female rats with one
exception; males and females were equally responsive to aPCP in the stop-exposure study. The
EPA determined that the LOAEL was 20 mg/kg-day for male rats based on statistically
significant increases in cystic degeneration; the NOAEL was 10 mg/kg-day. The LOAEL was
30 mg/kg-day for female rats based on a biologically significant decrease in body weight; the
NOAEL was 20 mg/kg-day. The chronic studies for PCP are summarized in Table 4-14.
73
-------
Table 4-14. Summary of NOAELs/LOAELs for oral chronic studies for
PCP
Species, strain
Rat, Sherman
(10/sex/dose)
Dog, beagle
(4/sex/dose)
Rat, F344
(50/sex/dose)
Rat, Sprague-Dawley
(25/sex/dose)
Mouse, B6C3FJ
(50/sex/dose)
Dose (mg/kg-d),
duration
2, 9, or 44 (M)
2, 10, or 48 (F)
8 mo (feed)
2, 9, or 44 (M)
2, 10, or 48 (F)
8 mo (feed)
1.5, 3.5, or6.5
1 yr (gelatin capsule)
10, 20, or 30
2 yrs (feed)
1,3, 10, or 30
2 yrs (feed)
18 or 35 (M)
17 or 35 (F)
2 yrs (feed)
18, 37, or 118 (M)
17, 34, or 114 (F)
2 yrs (feed)
Grade/type
of PCP
aPCP
tPCP
tPCP
aPCP
EC-7
tPCP
EC-7
NOAEL
(mg/kg-d)
9(M)
10 (F)
Not
established
Not
established
10 (M)
20 (F)
10 (M)
3(F)
Not
established
Not
established
LOAEL
(mg/kg-d)
44 (M)
48 (F)
2
1.5
20 (M)
30 (F)
30 (M)
10 (F)
18 (M)
17 (F)
18 (M)
17 (F)
Reference
Kimbrough and
Linder (1978)a
Mecler(1996)a
NTP (1999)a
Schwetz et al.
(1978)
NTP (1989)a
aNOAELs and LOAELs determined by the EPA for these studies; values are for both genders unless otherwise
specified.
4.2.2. Inhalation Studies
4.2.2.1. Subchronic Studies
No subchronic inhalation studies that examined the effects of PCP in humans are
available. A Chinese study (Ning et al., 1984; translation) exposed weanling male rats to 3.1 or
21.4 mg/m3 PCP (reagent grade, Sodium-PCP) 4 hours/day, 6 days/week, for 4 months. Rats in
the 21.4 mg/m3 group exhibited significant increases, compared with control, in lung, kidney,
liver, and adrenal gland weight. Additionally, the levels of blood-glucose were elevated in rats
exposed to the high concentration of PCP. Ning et al. (1984) also observed statistically
significantly increased serum y-globulin (although not a-globulin, p-globulin, or serum albumin)
and lung and liver weights in six rabbits (pooled males and females) exposed, in a similar
manner, to 21.4 mg/m3. Demidenko (1969) reported results in which anemia, leukocytosis,
eosinophilia, hyperglycemia, and dystrophic processes in the liver were observed in rats and
rabbits exposed to 28.9 mg/m3 PCP (high concentration; purity not reported) for 4 hours/day for
4 months. Animals exposed to the low concentration (2.97 mg/m3) exhibited effects on liver
function, cholinesterase activity, and blood sugar that were considered minor and were not
observed 1 month following exposure completion. Kunde and Bohme (1978), calculated an
estimated dose of 0.3 mg/kg-day PCP based on the 2.97 mg/m3 concentration reported by
Demidenko (1969). This calculation assumed 100% pulmonary uptake and absorption.
74
-------
4.2.2.2. Chronic Studies
No chronic inhalation studies that examined the effects of PCP in humans or animals are
available.
4.2.3. Other Routes of Exposure
A 13-week dermal toxicity study was conducted in groups of 10 male and 10 female
Sprague-Dawley rats/dose receiving 0, 100, 500, or 1,000 mg/kg-day doses of tPCP (88.9%
purity) applied to clipped dorsal skin for 6 hours/day for 91 days (Osheroff et al., 1994). tPCP,
applied without a vehicle, was held in place by a gauze patch. Some degree of skin irritation
(acanthosis and chronic inflammation) was observed in both sexes at all doses of tPCP. Chronic
inflammation was observed in 10, 80, and 100% of males and 0, 100, and 100% of females
treated with 100, 500, and 1,000 mg/kg-day tPCP, respectively. Hepatocellular degeneration
was observed in 90 and 100% of males at the mid and high doses, respectively, and in 20, 100,
and 100% of females in the low, mid, and high doses, respectively. ALT was statistically
significantly increased 4.3- and 7.6-fold in males and 2.5- and 5.4-fold in females in the 500 and
1,000 mg/kg-day dose groups, respectively, and AST was statistically significantly increased
2.3- and 3.3-fold in males and 1.8- and 3.1-fold in females in the 500 and 1,000 mg/kg-day dose
groups, respectively. Relative liver weights were statistically significantly increased over
controls in the 100 (11%), 500 (18%), and 1,000 (30%) mg/kg-day dose groups for male rats. In
females, the relative liver weights in animals of the 500 (18%) and 1,000 (36%) mg/kg-day dose
groups were significantly greater than controls. Additionally, relative kidney weights were
increased 20% in 1,000 mg/kg-day males and 56 and 16% in 500 and 1,000 mg/kg-day females,
respectively. This study showed that PCP is absorbed from the skin at levels that caused liver
toxicity. The study authors determined that the LOAEL for this study was 500 mg/kg-day based
on dose-related increases in liver toxicity (hepatocellular degeneration, chronic inflammation,
and statistically significant increases in hepatic enzyme induction). The NOAEL was
100 mg/kg-day.
4.2.4. Cancer Studies
4.2.4.1. Oral Studies
NTP (1989) administered feed containing 100 or 200 ppm tPCP (90.4% purity) or 100,
200, or 600 ppm EC-7 (91% purity) to B6C3Fi mice (50/sex/group) continuously for 2 years
(NTP, 1989). Two groups of 35 mice of each sex maintained on untreated feed served as
controls for each grade of PCP. The average daily doses were estimated as 18 and 35 mg/kg-day
for 100 and 200 ppm tPCP males, respectively, and 17 and 34 mg/kg-day for 100 and 200 ppm
tPCP females, respectively. The doses of EC-7 administered to male and female mice were
estimated as 18, 37, or 118 mg/kg-day for males, and 17, 34, or 114 for females, respectively.
75
-------
The average daily PCP and contaminant doses associated with each dietary concentration are
summarized in Table B-3 of Appendix B. Statistical analyses included the life table test that
considered tumors as fatal in animals dying before study termination, the logistic regression test
that regarded all lesions as nonfatal, and the Fisher's exact and Cochran-Armitage trend test that
compared the overall incidence rates of treated groups with controls. Nonneoplastic findings are
discussed in Section 4.2.1.
The incidences of treatment-related tumors and results of the statistical analyses are
presented in Tables 4-15 (males) and 4-16 (females). In male mice, the incidence of
hepatocellular adenoma and carcinoma were statistically significantly elevated by both grades of
PCP compared with controls. The incidence of hepatocellular adenoma was statistically
significantly elevated in males receiving 18 mg/kg-day tPCP diet (43 versus 16% for controls),
but not in males receiving the 18 mg/kg-day EC-7 diet (27 versus 14% for controls). The
incidence of hepatocellular carcinoma in males was only marginally statistically increased
(p = 0.06 or 0.07) by both grades at 18 mg/kg-day (21% in tPCP and 15% in EC-7), although the
incidence was statistically significantly increased at 35 mg/kg-day for tPCP (25%) and at
37 mg/kg-day for EC-7 (15%) when compared with individual control groups. However, the
incidence of hepatocellular carcinoma in the 18 mg/kg-day dose groups was statistically
significantly (p = 0.006) elevated when compared with the combined control groups. The
incidence of hepatocellular adenoma/carcinoma was statistically significantly increased with all
doses of tPCP and EC-7. The incidences were greater in male mice receiving tPCP (55 and 77%
at 18 and 35 mg/kg-day, respectively) than in males receiving EC-7 (40, 44, and 69% at 18, 37,
and 118 mg/kg-day, respectively). In female mice, the incidence of hepatocellular adenoma
(63%) was statistically significantly elevated only at the 114 mg/kg-day dose of EC-7 when
compared with the control group, and the incidence of hepatocellular carcinoma (range of 2-4%)
was not significantly elevated in females treated with either grade of PCP. If incidence of
hepatocellular adenoma in female groups treated with tPCP is compared with the combined
control groups, then statistical significance is achieved at 17 mg/kg-day (p = 0.05; 16%) with
marginal significance at 34 mg/kg-day (p = 0.06; 16%).
76
-------
Table 4-15. Treatment-related tumors in male B6C3Fi mice fed tPCP or EC-7 for 2 years
Organ/lesions"
tPCP
Control
18 mg/kg-d
35 mg/kg-d
EC-7
Control
18 mg/kg-d
37 mg/kg-d
118 mg/kg-d
Liver — hepatocellular
Adenoma
Carcinoma
Adenoma/carcinoma
5/32
2/32
7/32
20/47c'd
10/47
26/47c'd
33/48b'c'd
12/48c'd
37/48b,c,d
5/35
1/35
6/35
13/48
7/48
19/48b'c'd
17/48b,c,d
7/48b'c
21/48b,c,d
32/49b,c,d
9/49b'°'d
34/49b,o,d
Adrenal gland/medulla
Pheochromocytoma
Malignant pheochromocytoma
Pheochromocyto ma/malignant
0/31
10/45b'c'd
23/45b'c'd
0/34
1/34
1/34
4/48
0/48
4/48
21/48b,c,d
0/48
21/48b,c,d
44/49b.°.d
3/49
45/49b,o,d
aData reported as number of animals with tumors/number of animals examined at the site.
bStatistically significant as calculated by life table analysis.
Statistically significant as calculated by the logistic regression test.
dStatistically significant as calculated by the Cochran-Armitage trend or Fisher's exact test.
eNo statistical analyses reported.
Source: NTP(1989).
77
-------
Table 4-16. Treatment-related tumors in female B6C3Fi mice fed tPCP or EC-7 for 2 years
Organ/lesions"
tPCP
Control
17 mg/kg-d
35 mg/kg-d
EC-7
Control
17 mg/kg-d
34 mg/kg-d
114 mg/kg-d
Liver — hepatocellular
Adenoma
Carcinoma
Adenoma/carcinoma
3/33
0/33
3/33
8/49
1/49
9/49
8/50
1/50
9/50
1/34
0/34
1/34
3/50
1/50
4/50
6/49
0/49
6/49
30/48b'c'd
2/48
31/48b'c'd
Adrenal gland/medulla
Pheochromocytoma
Malignant pheochromocytoma6
Pheochromocytoma/malignant
2/33e
2/48e
l/49e
0/35
0/35
0/35
1/49
1/49
2/49
2/46
0/46
2/46
38/49b'c'd
1/49
38/49b'c'd
Circulatory system
Hemangioma"
Hemangiosarcoma
Hemangioma/hemangiosarcoma
0/35
3/50
6/50b'c'd
0/35
0/35
0/35
0/50
1/50
1/50
0/50
3/50
3/50
1/49
9/49b'c'd
9/49b'c'd
aData reported as number of animals with tumors/number of animals examined at the site.
bStatistically significant as calculated by life table analysis.
Statistically significant as calculated by the logistic regression test.
dStatistically significant as calculated by the Cochran-Armitage trend or Fisher's exact test.
eNo statistical analyses reported.
Source: NTP(1989).
78
-------
Adrenal gland medullary pheochromocytomas occurred in 22 and 51% of male mice
receiving 18 and 35 mg/kg-day tPCP, respectively, and in 44 and 90% of male mice receiving
37 and 118 mg/kg-day EC-7, respectively, but in none of the controls. Pheochromocytomas also
developed in 78% of females receiving 114 mg/kg-day compared with only one or two female
mice in the control groups or 17 and 34 mg/kg-day dose groups. Hemangiosarcomas, which
developed primarily in the liver and spleen, were observed in 6 and 12% of females receiving
17 and 34 mg/kg-day tPCP, and 2, 6, and 18% receiving 17, 34, and 114 mg/kg-day EC-7, and
none in the 70 controls examined. Hemangiosarcomas were also observed in male mice
administered both grades of PCP, although the incidences were low (4-6% in tPCP-exposed
mice and 6-10% in EC-7-exposed mice versus 3% in control) and were not statistically
significantly different from the control.
The results of this study show that tumors were induced in mice exposed to tPCP and
EC-7. The latter contains relatively low levels of dioxin and furan impurities compared to tPCP.
Based on tumor response, tPCP was slightly more potent. NTP (1989) and McConnell et al.
(1991) compared the concentrations of HxCDD, a known contaminant of PCP, in tPCP and EC-7
with that known to induce liver tumors in mice and concluded that the carcinogenic response in
mice can be attributed primarily to PCP and that the impurities provided a minor contribution.
NTP (1989) concluded that PCP is primarily responsible for the carcinogenicity observed in
mice and that impurities played only a small part in the neoplastic process, at least in the liver of
male mice. NTP further concluded that there was clear evidence of carcinogenic activity for
male mice receiving tPCP and male and female mice receiving EC-7 and some evidence of
carcinogenic activity for female mice receiving tPCP.
Bionetics Research Laboratory, Inc. (BRL, 1968) carried out two long-term (18-month)
studies of EC-7 (90% purity) in B6C3Fi and B6AKFi mice, one using continuous oral
administration and the other a single subcutaneous injection. In the first study, mice
(18 mice/sex/strain) were exposed to EC-7 by gavage (in 0.5% gelatin) at a dose of 46.4 mg/kg-
day starting on day 7 of age through weanling (day 28 of age). Thereafter, mice received EC-7
in the diet at a dose initially corresponding to 46.4 mg/kg-day; dosing continued for up to
18 months of total exposure. No adjustments to the dietary concentration were made for body
weight gain during the study. In the second experiment, 28-day-old mice of the same strains
(18 mice/sex/strain) received a single, subcutaneous injection of 46.4 mg/kg EC-7 in the neck
and were examined at 18 months. Male and female mice exposed to EC-7 in this study did not
develop tumors that were considered statistically significantly greater in incidence than tumors
observed in control animals.
In the NTP (1999) study, groups of 50 male and 50 female F344 rats were administered
aPCP (99% purity) in feed at concentrations of 0, 200, 400, or 600 ppm continuously for
105 weeks; additional groups of 60 male and 60 female rats were maintained on feed containing
1,000 ppm aPCP for 52 weeks followed by untreated feed until study termination at 2 years in a
79
-------
stop-exposure study. The average doses of PCP were reported as 10, 20, 30, and 60 mg/kg-day
for male and female rats fed the 200, 400, 600, and 1,000 ppm diets, respectively.
Histopathologic examination showed a statistically significantly higher incidence (18%) of
malignant mesothelioma in 60 mg/kg-day males compared with controls; the incidence exceeded
the range of historical controls. The mesotheliomas originated from the tunica vaginalis. The
incidence of nasal squamous cell carcinomas was also elevated (10%) in 60 mg/kg-day males.
At study termination (2 years), the nasal tumors spread to the oral cavity in one of the male rats
in this dose group. When compared with concurrent controls, the tumor incidence in male rats
did not achieve statistical significance but did exceed the range of historical controls. Nasal
squamous cell carcinoma at 10 mg/kg-day was the only neoplastic finding in male rats treated for
the entire 2 years that occurred with a higher incidence (6%) than that of historical controls.
However, NTP (1999) did not consider the finding at 10 mg/kg-day to be treatment related
because the incidence at 20 (2%) and 30 mg/kg-day (0%) was less than or no greater than that of
concurrent controls (2%). Therefore, the only treatment-related tumors that occurred in male rats
were in those animals exposed to 60 mg/kg-day PCP in the stop-exposure study. The tumors
observed in the stop-exposure study were observed earlier than tumors at other doses (45 days
earlier for nasal tumors and 91 days earlier for mesotheliomas) and did not regress during the
observation year in which animals were administered untreated feed. There were no treatment-
related increases in tumor incidence at any site in females receiving aPCP. These data and
results of the statistical analyses are presented in Table 4-17. NTP concluded that this study
showed some evidence of carcinogenic activity of PCP in male F344 rats, based on increased
incidences of mesothelioma and nasal squamous cell carcinoma in the stop-exposure study.
Additionally, the tumors observed in the 1-year stop-exposure study did not regress when
animals were examined 1 year after exposure stopped.
80
-------
Table 4-17. Incidences of treatment-related tumors in male F344 rats fed
purified PCP for up to 2 years
Tumors and statistical analysis
Malignant mesothelioma
Overall rateb
Adjusted rate0
Statistical analysis
Poly-3 testd
Fisher's exact test
Historical control incidence
(mean ± standard deviation)
Nasal squamous cell carcinoma
Overall rateb
Adjusted rate0
Statistical analysis
Poly-3 testd
Fisher's exact test
Historical control incidence
(mean ± standard deviation)
Dose (mg/kg-d)
0
1/50 (2%)
2.6%
p = 0.447N
10
0/50 (0%)
0%
p = 0.509N
p = 0.500N
20
2/50 (4%)
5.1%
^ = 0.511
;? = 0.500
30
0/50 (0%)
(0%)
p = 0.472N
p = 0.500N
60a
9/50 (18%)
20.6%
;? = 0.014
p = 0.008
40/1,354 (3.0 ± 2.3%), range = 0-8%
1/50 (2%)
2.7%
^ = 0.171N
3/50 (6%)
8.1%
;? = 0.299
;? = 0.309
1/50 (2%)
2.6%
p = 0.756N
0/50 (0%)
(0%)
^ = 0.471N
p = SOON
5/50 (10%)
11.7%
;? = 0.128
;? = 0.102
5/1,314 (0.5 ± 1.0%); range = 0-4%
"Stop-exposure study; rats received treated feed for 52 wks and untreated feed until study termination at 2 yrs.
bNumber of animals with tumors/number of animals examined.
°Poly-3 estimated incidence after adjustment for intercurrent mortality.
dTrend-test under control column (60 mg/kg-d group excluded); pair-wise comparison test under treatment group
column. Poly-3 test accounts for intercurrent mortality; N refers to negative trend.
Source: NTP(1999).
Schwetz et al. (1978) conducted a 2-year study in 25 male and 25 female Sprague-
Dawley rats maintained on diets containing EC-7 (90.4% purity) at concentrations delivering
doses of 3, 10, or 30 mg/kg-day; males were fed the diets for 22 months and females for
24 months. Tumors typical of this strain of rat (i.e., pituitary, adrenal and thyroid glands, testes,
and pancreas tumors in males and pituitary, thyroid, mammary glands, and uterus tumors in
females) were noted in 41% of the male controls and 100% of the female controls. The treated
animals exhibited tumors that were also observed in the control animals. There were no
statistically significant increases in incidence of tumors noted in the treated animals when
compared with the controls. Information concerning individual tumors was not included in the
report.
Initiation/promotion studies. Umemura et al. (1999) examined the initiating and
promoting activity of aPCP (98.6% purity) administered in the diet to 20 male B6C3Fi mice/
group. Diethylnitrosamine (DEN) was given as the initiator when the promoting activity of
aPCP was assessed, and PB was administered as the promoter when the initiating activity of
aPCP was assessed. Table 4-18 summarizes the treatment protocol and response of each group
to treatment. Three groups of mice received no treatment during the 13-week initiating phase but
81
-------
were administered a basal diet, 600 ppm aPCP, or 500 ppm PB during the 25-week promoting
phase. DEN was administered in drinking water to four groups for 13 weeks at a concentration
of 20 ppm followed by a 4-week rest period. Following the rest period, animals were treated
with a basal diet, 500 ppm PB in drinking water, or 300 or 600 ppm aPCP in the diet for
25 weeks to assess promoting activity of aPCP. aPCP was administered at 1,200 ppm during the
initiating phase followed by no treatment for 29 weeks. Two groups of mice received aPCP at
concentrations of 600 or 1,200 ppm in the diet for 13 weeks, followed by 500 ppm of PB for
29 weeks (no rest period). The doses corresponding to dietary concentrations of 300, 600, and
1,200 ppm aPCP were estimated to be 54, 108, and 216 mg/kg-day, respectively.
Table 4-18. Hepatocellular tumors in B6C3Fi mice in initiation/promotion
studies
Treatment"
Initiation
(13 wks)
Untreated
Untreated
Untreated
DEN (20 ppm)
DEN (20 ppm)
DEN (20 ppm)
DEN (20 ppm)
aPCP (216 mg/kg-d)
aPCP (216 mg/kg-d)
aPCP (216 mg/kg-d)
Promotion
(25 wks)
Basal diet
aPCP
(108 mg/kg-d)
PB (500 ppm)b
Basal diet
PB (500 ppm)
aPCP
(54 mg/kg-d)
aPCP
(108 mg/kg-d)
PB (500 ppm)b
PB (500 ppm)b
Untreated
Incidences
Altered foci
0/20
1/19
(5%)
8/20
(40%)
7/15
(47%)
6/19
(32%)
8/15
(53%)
13/18
(72%)
5/20
(25%)
2/20
(10%)
2/17
(12%)
Adenomas
0/20
0/19
0/20
4/15
(27%)
10/19
(53%)
10/15C
(67%)
13/18
(72%)d
0/20
0/20
0/17
Carcinomas
0/20
0/19
0/20
0/15
1/19
(5%)
2/15
(13%)
4/18
(22%)
0/20
0/20
0/17
Adenoma/
carcinoma
0/20
0/19
0/20
4/15
(27%)
10/19
(53%)
10/15
(67%)c
13/18
(72%)d
0/20
0/20
0/17
Tumor
multi-
plicity
0
0
0
0.33
1.42C
1.27C
2.22C
0
0/20
0/17
"Vehicle: aPCP in feed; DEN and PB in drinking water; a 4-wk rest period followed the initiation phase.
''No rest period, PB given for 29 wks.
°p < 0.05.
dp < 0.01 (compared with DEN + PB).
Source: Umemuraetal. (1999).
Survival of mice was reduced in animals administered 108 (19/20) and 216 mg/kg-day
(17/20) of aPCP alone. DEN-treated animals also exhibited a decrease in survival with basal diet
(15/20), PB (19/20), and 54 (15/20) and 108 (18/20) mg/kg-day aPCP. Body weight
82
-------
measurements recorded at the end of the 42-week study showed significant reductions of 20, 22,
24, and 29% in mice receiving DEN followed by basal diet, PB, and 54, and 108 mg/kg-day
aPCP, respectively, compared with mice receiving only the basal diet. Hepatomegaly was
observed with aPCP or PB following DEN treatment. Liver weights were increased in mice
receiving 108 mg/kg-day aPCP with (1.9-fold) or without (1.3-fold) prior DEN treatment. Liver
weights in animals treated with PB alone (1.3-fold) or after aPCP treatment (1.4- and 1.3-fold
with 108 and 216 mg/kg-day, respectively) were also increased. Liver weights were not
increased after administering 216 mg/kg-day aPCP for 13 weeks, followed by no treatment for
29 weeks.
There was an increase in incidence of hepatocellular altered foci for all mice in the
treated groups, although the only statistically significant increase (5.7-fold) in multiplicity was
observed with DEN initiation and 108 mg/kg-day aPCP promotion. All groups initiated with
DEN exhibited hepatocellular adenomas and carcinomas with the exception of the DEN control
group, which only developed adenomas. The incidence of liver tumors was statistically
significantly higher in mice initiated with DEN and promoted with 54 (67%) or 108 mg/kg-day
PCP (72%) than in control mice receiving DEN only (27%). Tumor multiplicity was statistically
significantly increased in 54 and 108 mg/kg-day aPCP-promoted mice (1.27 and 2.22 tumors/
mouse, respectively) and 500 ppm PB (1.42 tumors/mouse) compared with DEN controls
(0.33 tumors/mouse). No liver tumors developed in mice initiated with aPCP with or without
subsequent promotion with PB. In this study, aPCP, at approximate doses of 54 and 108 mg/kg-
day, showed promoting, but not initiating, activity in mice that were initiated with DEN.
Umemura et al. (1999) concluded that aPCP exerts a promoting effect on liver carcinogenesis.
In another promotion study, Chang et al. (2003) administered an initiator, 100 jig
dimethylbenzanthracene (DMBA) in acetone (100 uL), in a single application to the back of
10 CD-I female mice/dose followed 1 week later by promotion treatment with 2.5, 50, or
1,000 jig PCP or TCHQ (purities not reported) in acetone twice weekly painted onto the skin of
the mice for total treatment times of 20 or 25 weeks. DMBA treatment followed by PCP or
TCHQ promotion resulted in a dose-related increase (>1.6-fold) in epidermal hyperplasia and
elevated proliferating cell nuclear antigen expression (>2.2-fold), with TCHQ being slightly
more effective than PCP. One or two skin tumors were observed in week 6 (30%) and week 11
(20%) in mice treated with PCP (0.2-0.4 tumors/mice average) and TCHQ (0.1-0.7 tumors/
mouse average), respectively. Systemic effects include dose-related decreases in body weight in
which TCHQ induced a greater loss in body weight than PCP (16 versus 7%, respectively). The
kidneys were significantly enlarged for all treated mice. Liver and spleen weights were
increased with PCP and decreased with TCHQ following treatment. However, PCP (not TCHQ)
promotion also caused lymphomas. Initiating ability of PCP or TCHQ was not tested in this
study.
83
-------
4.2.4.2. Inhalation Studies
No chronic cancer bioassays by the inhalation route of exposure are available.
4.3. REPRODUCTIVE, ENDOCRINE, AND DEVELOPMENTAL STUDIES
4.3.1. Reproductive and Endocrine Studies
Schwetz et al. (1978) conducted a one-generation reproductive toxicity study in which
groups of 10 male and 20 female Sprague-Dawley rats were administered EC-7 (90% purity) in
the diet. Dietary concentrations were adjusted monthly to deliver doses of 3 or 30 mg/kg-day.
The test material was administered continuously for 62 days prior to mating and during mating,
gestation, and lactation. All animals including pups were sacrificed after the litters were weaned
on lactation day 21 (169 days for males; -110 days for females). Toxic effects were noted in the
animals and pups of the high dose only. There were no significant effects on survival, body
weight, or litters at the low dose. Decreased body weight was noted in high-dose rats, with an
8% decrease in males and a 10% decrease (statistically significant) in females. At 30 mg/kg-day,
fewer pups were born alive and the survival of pups decreased throughout lactation, leading to
significantly decreased litter sizes measured on days 7, 14, and 21 of lactation. In addition, mean
pup weights were significantly decreased by 14-27% at birth and throughout lactation at
30 mg/kg-day compared with the controls. Decreases in pup weight gain (28%) and survival
(79%) during the first 14 days of lactation in the 30 mg/kg-day dose group are suggestive of a
lactational effect of EC-7. The study authors noted that an increased incidence of litters with
skeletal variations (lumbar spurs and vertebra with unfused centra) occurred at 30 mg/kg-day
compared with controls. The study authors determined that the LOAEL for this study was
30 mg/kg-day for statistically significant changes in reproductive and developmental effects
(decreased survival and growth, and skeletal variations); the NOAEL was 3 mg/kg-day.
In a two-generation reproductive toxicity study (Bernard et al., 2002), tPCP (88.9%
purity) in corn oil was administered by gavage 7 days/week to groups of 30 male and 30 female
Sprague-Dawley rats at doses of 10, 30, or 60 mg/kg-day. FO male and female rats were given
PCP for at least 70 days prior to mating and during mating, gestation, and lactation until weaning
of litters, after which all FO animals were sacrificed. Fl male and female rats were similarly
exposed, starting at weaning and continuing to the day before sacrifice. In addition to indices of
reproductive performance, parameters of reproductive function (vaginal patency, preputial
separation, estrous cycle, and sperm morphology) were also evaluated.
Absolute body weight of the 30 and 60 mg/kg-day groups of FO and Fl parental male rats
were statistically significantly decreased by 5.3 and 15%, respectively, compared with controls
from day 36 throughout the remainder of the study. Significantly decreased absolute body
weight was observed in 60 mg/kg-day females during the premating, gestation, and lactational
periods. No treatment-related effect was observed on body weight in females receiving
30 mg/kg-day, except for lactation days 10 and 15-17 in which body weight was statistically
84
-------
significantly lower (-8%) than controls. Systemic effects in parental animals (FO and Fl male
rats) were observed at 30 and 60 mg/kg-day dose levels and included increased liver weight,
enlarged liver (FO males only), and microscopic liver lesions ranging from centrilobular
hypertrophy and vacuolation, multifocal inflammation, and single cell necrosis to a centrilobular
pigment identified as LF. Centrilobular hypertrophy, vacuolation, and multifocal inflammation
were also observed at the lowest dose of 10 mg/kg-day in FO and Fl males. The liver weight in
FO females was significantly greater than controls in the 30 and 60 mg/kg-day dose groups.
Parental females exhibited histopathological effects similar to males, including centrilobular
hypertrophy and vacuolation, multifocal inflammation, single-cell necrosis (except for
Fl females), and LF pigment at tPCP doses of 10, 30, and 60 mg/kg-day. Additionally, bile duct
proliferation was observed at 60 mg/kg-day tPCP.
The fertility index and the number of litters produced were decreased at 60 mg/kg-day in
Fl females. Days to vaginal patency and preputial separation were statistically significantly
increased in Fl females (at doses >10 mg/kg-day) and males (at doses >30 mg/kg-day),
respectively. The length of the estrous cycle was not significantly affected in either FO or
Fl females. Sperm morphology and count were not affected in FO males, although testicular
spermatid count and testes weight were decreased at 30 and 60 mg/kg-day in Fl males.
Offspring evaluations showed significant reduction in mean litter size, number of live pups,
viability index, and lactation index for Fl and/or F2 pups at 60 mg/kg-day tPCP compared with
the controls. Body weight of pups was statistically significantly decreased by 6-9% at
10 mg/kg-day (lactation days 1-4), by 10-15% at 30 mg/kg-day (lactation days 1-28), and by
11-39% at 60 mg/kg-day (lactation days 1-28). In addition, decreased weights of the liver,
brain, spleen, and thymus were observed in F2 pups at 60 mg/kg-day. The study authors
determined that the parental LOAEL was 30 mg/kg-day for male and female rats based on
significantly decreased body weight and weight gain in Fl generation parental rats, and testicular
effects in Fl male rats (decreased testis weight, decreased spermatid count). The investigators
noted that reproductive and developmental toxicity in the rats of this study were only observed at
doses that also induced systemic toxicity. The EPA determined that the parental LOAEL was 10
mg/kg-day (lowest dose tested) for male and female parental rats, based on effects in the liver
characterized by single cell necrosis, LF, centrilobular hypertrophy, cytoplasmic vacuolation,
and multifocal inflammation. The parental NOAEL could not be determined. The reproductive
LOAEL was 10 mg/kg-day (lowest dose tested) based on statistically significantly decreased
group mean pup weight and statistically significantly increased vaginal patency in females. The
reproductive NOAEL could not be determined.
Beard et al. (1997) conducted a study using mink to assess the effect of PCP in a one-
generation study. Groups of 10 female mink (9 months old) received 1 mg/kg-day PCP (purity
not stated; confirmed as aPCP [CalEPA, 2006]) in the diet continuously for 3 weeks before and
during mating, and throughout gestation and lactation of one litter of kits. Each female was
85
-------
mated twice with an untreated male mink, with an interval of 7-8 days between matings.
Treatment with 1 mg/kg-day aPCP had no effect on clinical signs, body weight gain, or food
consumption. No effect was observed on females accepting males during the first mating, but
statistically significantly fewer aPCP-treated females accepted males during the second mating,
resulting in significantly fewer pregnant females. Implantations were not affected by aPCP
treatment, but only 70% of the treated mink with implantation sites eventually whelped
compared with 88% of controls. In aPCP-treated mink, 46.7% of embryos were lost compared
with 40.5% of control embryos, which resulted in smaller litter sizes (3.40 versus 4.45 for
controls). The decreased implantation rate and reduced embryo survival after implantation were
not statistically significantly different from the controls; however, the combined effect of these
decreases contributed to the lower whelping rate. Uterine cysts were present in both control and
treated mink, although the severity was greatest in the treated animals (severity grade 1.33 in
treated versus 0.19 in controls). The study authors suggested that aPCP may have contributed to
the increased loss of embryos. Beard et al. (1997) noted that the uterine cysts may have been
associated with uterine infection and could indicate an immunosuppressive activity on the uterus
by aPCP. Additionally, aPCP treatment resulted in a longer duration of pregnancy (4-5 days
longer) compared with controls. aPCP treatment had no effect on serum levels of progesterone,
estradiol, cortisol, or T4 in adult female mink at weaning of their litters. Mink are seasonal
breeding animals (in that ovulation is induced by copulation and implantation is delayed) which,
according to the investigators, may result in these animals being particularly sensitive to aPCP
(mild effects on reproduction were noted at a dose that was an order of magnitude lower than the
NOAEL for a two-generation study in rats [Bernard et al., 2002]). A decrease (not considered
statistically significantly different from controls) in the whelping rate was observed in mink at
1 mg/kg-day aPCP; however, it is unknown if this is a result of the embryo loss or the reduction
in mating response. The study authors did not determine a NOAEL or LOAEL for this study.
The EPA established a free-standing NOAEL of 1 mg/kg-day (only dose used), based on the
absence of treatment-related lexicologically significant effects.
Beard and Rawlings (1998) examined reproduction in a two-generation study in mink
exposed to 1 mg/kg-day PCP (purity not reported); 10 controls/generation were included. Dams
(number of animals not reported) were administered PCP, in feed, 3 weeks prior to mating and
continued through gestation until weaning of offspring (8 weeks postpartum). Eight
Fl generation females (from treated dams) were administered PCP in their feed starting at
weaning; these animals were maintained on the treated diet as they grew and were mated with
untreated males. Treatment continued throughout gestation and lactation, and was terminated
with sacrifice of Fl females 3 months after the end of the lactation period. Six Fl generation
males were administered PCP in their feed starting at weaning until maximal development of the
testis (approximately 42 weeks of age), at which time the Fl males were sacrificed. Ten
F2 generation females were administered PCP-treated feed from weaning until mink reached full
86
-------
body size (approximately 30 weeks of age). Eight F2 generation males were administered
PCP-treated feed from weaning until the mink reached sexual maturity in their first breeding
season. The study authors noted that all of the animals received PCP-treated feed continuously
from conception to maturity. The only change observed in the body weights of PCP-treated
mink was a 17% increase over controls in the body weight of Fl males. There were no changes
in the proportion of Fl generation accepting the first and second mating. Additionally, no
temporal changes were noted during the matings. PCP treatment did not affect whelping date or
duration of gestation in the mink. Mean testis length was greater in PCP-treated Fl male mink
compared with controls, although this difference was not apparent in examination (length and
mass measurements) of testes after removal. Interstitial cell hyperplasia of greater severity was
noted in the testes of Fl generation males compared with controls (severity scores for left and
right testes were 1.0 and 0.6 for controls versus 2.3 and 2.5 for treated animals, respectively).
The severity of cystic hyperplasia in the prostate gland of Fl males was statistically significant
(0.9) compared with controls (0). A higher serum testosterone concentration was associated with
the mild multifocal cystic hyperplasia, noted in 50% of the PCP-treated mink.
Serum T4 secretion was statistically significantly decreased in the Fl (-21%) and
F2 (-18%) males and in the F2 females (-17%). T4 secretion was presented graphically in Beard
and Rawlings (1998); therefore, percent changes are reported as approximate values estimated
from the graphs. Thyroid mass was decreased in both Fl and F2 generation animals, although
the reduction was statistically significant only in F2 females (-27%). There was a significant
increase in size (42%) of the adrenal gland in the Fl females, but no change in the F2 females.
Interestingly, decreased mating and whelping rates were observed in mink treated with 1 mg/kg-
day PCP in the one-generation study by Beard et al. (1997) compared with no changes in mating
or whelping rates of 1 mg/kg-day PCP-treated mink in the two-generation reproductive study by
Beard and Rawlings (1998). The authors noted that the treatment-related cystic hyperplasia of
the prostate and interstitial hyperplasic testes may be associated with PCP-induced
hypothyroidism. The study did not report a NOAEL or LOAEL. The EPA determined a
LOAEL of 1 mg/kg-day based on significant decreases in T4 secretion.
In a one-generation study, groups of 13 ewes (1-3 years old) received an untreated diet or
a diet treated with PCP (purity not reported) at a concentration delivering a dose of 1 mg/kg-day
(Beard et al., 1999a). The ewes were treated for 5 weeks prior to mating (with untreated rams),
during gestation, and until 2 weeks after weaning their lambs. The ewes were sacrificed at the
end of treatment. Clinical signs, blood hormone levels, ovarian function, embryonic growth,
reproductive function, and histopathologic lesions were assessed during the study. No clinical
signs or treatment-related decreases in body weight were observed. One ewe died of a cause
unrelated to treatment with PCP. No effects on reproductive function (i.e., ovulation rate,
fertility rate, lambing rate, mean number of lambs born per ewe, and mean gestation rate) were
observed. The male:female ratio showed an excess of ewe lambs born (5:13). There was a slight
87
-------
but statistically significant decrease in the weight of ewe lambs at weaning (86% of control
weight). Ovarian function (follicle number and corpora lutea size), fetal growth (measured by
head diameter), and postweaning serum levels of luteinizing hormone (LH), FSH, and cortisol
were not affected by treatment with PCP. However, maximum serum 14 levels in PCP-treated
ewes were statistically significantly lower (approximately 25%) than in control ewes with or
without prior administration of TSH. The increases in serum T4 levels compared with the
pretreatment level were 190% for PCP-treated ewes and 169% for controls.
Beard et al. (1999b) described a study in sheep in which the ram lambs born of ewes
maintained on untreated or PCP-treated diets were examined. A dose of 1 mg/kg-day PCP
(purity not reported) was administered starting at week 5 prior to mating and continuing through
weaning of lambs. The lambs were maintained on the same diets as the ewes from weaning until
puberty at 28 weeks of age. The lambs exhibited no overt signs of toxicity or treatment-related
decreases in body weight. Testes diameter was unaffected at 10 and 14 weeks of age, but scrotal
circumference measured at intervals between 16 and 26 weeks was statistically significantly
increased in PCP-treated rams. There was no effect of PCP on age at puberty, sperm count, or
sperm motility at 27 weeks of age. Scores for different measures of sexual behavior were
consistently lower in PCP-treated rams than in controls at 26 weeks of age, but the differences
were not statistically significant. T4 levels were statistically significantly lower at 6-16 weeks,
similar at 18-26 weeks, and lower at 28 weeks of age, compared with control levels. The
response to TSH stimulation was unaffected by treatment with PCP. The serum levels of other
endocrine hormones were unaffected by treatment with PCP. Microscopic examination of the
testes and epididymides showed seminiferous tubular atrophy, reduced production of
spermatocytes in the seminiferous tubules, and reduced density of sperm in the body of the
epididymides but not in the head and tail of the epididymides. The investigators attributed the
spermatogenic findings to the reduced thyroid hormone levels.
4.3.2. Developmental Studies
Larsen et al. (1975) reported on groups of 10 pregnant CD Sprague-Dawley rats
administered 60 mg/kg aPCP (>99% purity) in olive oil by gavage on GDs 8, 9, 10, 11, 12, or 13
and maintained until GD 20. Controls received olive oil only. The percentages of resorptions
ranged from 2.0 to 11.6% for controls and from 1.6 to 13.5% for treated dams. Additionally, the
temperature of the treated animals increased significantly (increases ranged from 0.5 to 1.14°C)
in animals treated on GDs 8, 9, or 10. The fetuses from dams receiving aPCP on GDs 8, 9, 10,
or 12 weighed 12-20% less than those from controls; the weight of fetuses from dams treated on
GDs 11 or 13 were similar to those of controls. There was a small increase in the percentage of
fetuses with malformations: 2% after treatment on GD 8 and 5.8% after treatment on GD 9. No
malformations were observed in control fetuses. The investigators attributed the fetal effects to
maternal toxicity because a placental transfer experiment, performed concurrently with this
-------
study, indicated that only very small amounts (<0.1% of the administered dose/g of tissue) of
aPCP cross the placental barrier.
In a study conducted by Welsh et al. (1987), 20 Sprague-Dawley rats/sex/dose were
administered diets containing aPCP (>99% purity) at dose levels of 60, 200, or 600 ppm (4, 13,
or 43 mg/kg-day, respectively) for 181 days prior to mating. At the end of the 181-day dosing
phase, male and female rats were mated for teratological evaluation. After mating, PCP
administration in the diet continued through gestation until GD 20 when dams were sacrificed.
Body weight gain in maternal rats exposed to aPCP was statistically significantly decreased at
the high dose (76% of control). Food consumption was increased for all dose groups in the early
part of gestation. Ringed eye (50%) and vaginal hemorrhaging (25%) were observed in dams of
the 43 mg/kg-day dose group. The investigators suggested that the hemorrhaging was most
likely related to the pregnancies. Pregnancy rates were low in all dose groups (77.5, 55, 84.2,
and 85% for the 0, 4, 13, and 43 mg/kg-day dose groups, respectively); however, there was no
effect on fertility. There were no dose-related effects on corpora lutea, implantation efficiency,
or average number of implants/female. Decreased numbers of viable fetuses (due to early death)
were observed at 43 mg/kg-day. Statistically significant increases in the percentage of females
with two or more resorptions were observed at 13 and 43 mg/kg-day.
Dose-related decreases in fetal body weight were observed in males (10%) and females
(8%) in the 13 mg/kg-day dose group and in males (36%) in the 43 mg/kg-day dose group.
Analysis at the 43 mg/kg-day dose level was not complete due to an alteration in the sex ratio at
this dose (100% male sex ratio at this dose was reported). Crown-rump lengths were decreased
in a dose-related manner for males and females at doses >13 mg/kg-day. No significant
alterations in external or sternebral observations were reported at any dose of aPCP in this study.
An increased incidence of misshapen centra and an increase in fetal litters with at least two
skeletal variations were observed at 13 mg/kg-day aPCP. The results of this study demonstrate
toxicity of aPCP at 13 mg/kg-day in the form of increased percentage of female rats with two or
more resorptions. However, this study is confounded by a lack of fetal data at the high dose and
inconsistent and low percentages of pregnancy at each dose level of aPCP tested. The
researchers suggest that PCP is embryotoxic and embryolethal rather than teratogenic. The EPA
determined that the maternal LOAEL was 13 mg/kg-day, based on significantly increased
resorptions, and the maternal NOAEL was 4 mg/kg-day. The developmental LOAEL was
13 mg/kg-day, based on dose-related increases in the incidence of skeletal variations and
decreases in fetal body weight and crown-rump lengths. The developmental NOAEL was
4 mg/kg-day.
In a study conducted by Schwetz et al. (1974a), doses of 5.8, 15, 34.7, or 50 mg/kg-day
tPCP (88.4% purity) or 5, 15, 30, or 50 mg/kg-day aPCP (>98% purity) prepared in corn oil were
administered by gavage to groups of pregnant Sprague-Dawley rats on GDs 6-15 (inclusive).
The control group consisted of 33 rats. The numbers of animals were 18, 17, 19, and 15 in the
89
-------
5.8, 15, 34.7, and 50 mg/kg-day tPCP dose groups, respectively, and 15, 18, 20, and 19 in the 5,
15, 30, and 50 mg/kg-day aPCP dose groups, respectively. Additional groups of rats were
administered 30 mg/kg-day aPCP and tPCP on GDs 8-11 or 12-15. Maternal toxicity from
aPCP was evidenced by decreased maternal weight gain in the 34.7 and 50 mg/kg-day tPCP dose
groups and 30 and 50 mg/kg-day aPCP dose groups for GDs 6-21 (74% compared with control).
For tPCP, weight gain was decreased 22 and 43% at the 34.7 and 50 mg/kg-day doses,
respectively, when compared with controls. The dams were more affected by aPCP than tPCP.
No other significant signs of maternal toxicity were observed.
The incidence of resorptions was increased at the three highest dose groups for both
aPCP (statistically significant in the 30 and 50 mg/kg-day dose groups) and tPCP (statistically
significant in all three dose groups). At the aPCP 50 mg/kg-day dose level, there were 100%
resorptions; thus, no measurements were recorded for aPCP-treated animals at values
>30 mg/kg-day. Resorptions were measured in 7, 9, 27, and 58% of fetuses and 56, 65, 95, and
93% of litters treated with 5.8, 15, 34.7, and 50 mg/kg-day tPCP, respectively. In animals
treated with 5, 15, 30, and 50 mg/kg-day of aPCP, resorptions were found in 4, 6, 97, and 100%
of fetuses and 5, 4, 100, and 100% of litters, respectively. Fetal body weight was statistically
significantly decreased for aPCP at 30 mg/kg-day and for tPCP at 34.7 and 50 mg/kg-day, but
actual values were not reported. The sex ratio showed a significant change from the controls
with a predominance of male survivors in the 30 and 50 mg/kg-day doses of aPCP and 34.7 and
50 mg/kg-day doses of tPCP. Crown-rump length was decreased at 30 mg/kg-day aPCP
(statistically significant) and 34.7 and 50 mg/kg-day tPCP. The litter incidence of soft tissue
anomalies (subcutaneous edema) and skeletal anomalies (lumbar spurs and supernumerary
lumbar, or fused ribs) was statistically significantly increased at 15, 34.7, and 50 mg/kg-day
tPCP, but the data did not indicate a clear dose-response (i.e., the number of litters affected were
greater at 34.7 than at 50 mg/kg-day). The litter incidence for similar soft tissue and skeletal
anomalies was also statistically significantly increased at 15 and 30 mg/kg-day aPCP. The
skeletal anomalies of the vertebrae and sternebrae occurred in a dose-related manner that was
statistically significant at doses >30 mg/kg-day for both tPCP and aPCP. At the 5 mg/kg-day
aPCP dose, the only significant effect observed was an increased number of fetal rats with
delayed ossification of the skull (threefold increase over controls).
Rats were treated on GDs 8-11 or 12-15 with 34.7 mg/kg-day tPCP or 30 mg/kg-day
aPCP to examine the effects on early or late organogenesis. Maternal body weight was
significantly decreased following treatment with aPCP (67%) and tPCP (27%) on GDs 8-11.
There were no dose-related decreases in maternal body weight in animals treated on GDs 12-15.
Resorptions in the GDs 8-11 treatment group were significantly increased in the aPCP- and
tPCP-treated rats. Fetal body weight and crown-rump length were significantly decreased in
animals treated with aPCP and tPCP on GDs 8-11. For the resorptions and changes in fetal body
weight and crown-rump length, aPCP-treated animals exhibited more severe effects than those
90
-------
treated with tPCP, but both formulations showed significantly elevated levels of fetal resorptions
when treated on GDs 8-11. On GDs 8-11, both aPCP and tPCP caused significant decreases in
fetal body weight and crown-rump length at the 30 and 34.7 mg/kg-day doses, respectively, but
only aPCP also significantly reduced these endpoints when administered on GDs 12-15.
Incidence of subcutaneous edema was statistically significant in fetuses treated with aPCP
(100%) and tPCP (82%) during GDs 8-11 and with aPCP (95%) during GDs 12-15. Skeletal
anomalies of the ribs, vertebrae, and sternebrae were found in approximately 100% of the fetuses
treated with aPCP or tPCP during GDs 8-11. The only skeletal effects observed during GDs 12-
15 were significant increases in the incidence of delayed skull ossification (aPCP, 70%) and
sternebrae anomalies (aPCP, 85%; tPCP, 82%). The authors postulated that this study was
limited due to the increased resorptions and correspondingly reduced litter sizes at higher dose
levels, but the results at lower doses were sufficient to indicate that the developing embryo is
more susceptible to PCP during early organogenesis. The study authors identified the
developmental NOAEL for tPCP as 5 mg/kg-day, which is equivalent to the adjusted dose of
5.8 mg/kg-day that the authors used for this grade of PCP to account for impurities.
Based on the results of this study, aPCP was more toxic than tPCP in maternal and fetal
rats. The EPA determined that the maternal LOAELs were 34.7 mg/kg-day for tPCP and
30 mg/kg-day for aPCP, based on significantly increased incidence of resorptions and decreased
body weight; the maternal NOAEL was 15 mg/kg-day. The developmental endpoints differed
according to the formulation of PCP used. The developmental LOAEL for aPCP was 5 mg/kg-
day based on dose-related, significantly delayed ossification of the skull. The developmental
NOAEL could not be established. The developmental LOAEL for tPCP was 15 mg/kg-day,
based on dose-related, statistically significant increases in soft tissue and skeletal anomalies.
The developmental NOAEL was 5.8 mg/kg-day.
Bernard and Hoberman (2001) observed effects in Crl:CD BR VAF/plus (Sprague-
Dawley) rats administered tPCP (88.9% purity; >97.5% chlorinated phenols) that were similar to
but less severe than those reported by Schwetz et al. (1974a). Groups of 25 pregnant rats were
administered tPCP in corn oil via gavage at doses of 0, 10, 30, or 80 mg/kg-day on GDs 6-15
(inclusive). Animals were sacrificed for maternal and fetal examinations on GD 21. The mean
maternal body weight gain was reduced by 15% at 80 mg/kg-day. Significant decreases in
maternal food consumption at 80 mg/kg-day were 15 and 11% less than controls on GDs 6-9 and
9-12, respectively. Additionally, increased numbers of dams with resorptions (83 versus 41%
for controls) were reported at 80 mg/kg-day.
Developmental toxicity was also observed at 80 mg/kg-day. Effects following tPCP
administration included decreased litter size (86% of controls) and reduced fetal body weight
(79% of controls). Litters from dams treated with 80 mg/kg-day had significantly increased
incidences of visceral (27 versus 5% for controls) and skeletal malformations/variations
(96 versus 27% for controls). The visceral malformations included hydrocephaly, diaphragmatic
91
-------
hernia, and dilation of renal pelvis, while skeletal malformations were of the vertebral and
sternebral type of anomalies. This study showed similar effects to those reported by Welsh et al.
(1987) in Sprague-Dawley rats, but this particular strain may not be as sensitive to tPCP, or tPCP
is not as toxic to the fetus as aPCP. The study authors determined that the maternal NOAEL for
this study was 30 mg/kg-day and the maternal LOAEL was 80 mg/kg-day, based on increased
incidence of resorptions and decreased maternal body weight gain. The developmental NOAEL
was 30 mg/kg-day and the developmental LOAEL was 80 mg/kg-day, based on significantly
increased visceral malformations and skeletal variations and decreased live litter size and fetal
body weight.
Bernard et al. (2001) examined inseminated New Zealand white rabbits (20 rabbits/dose)
administered tPCP (88.9% purity) by gavage at doses of 0, 7.5, 15, and 30 mg/kg-day on GDs 6-
18 (inclusive). The dams were sacrificed for maternal and fetal examinations on GD 29. There
was no dose-related maternal mortality or overt toxicity at any dose level. Decreases in maternal
mean body weight were statistically significant for GDs 6-12 and 9-12 at 30 mg/kg-day. At this
dose, body weight gain and food consumption showed overall decreases of 29 and 10%,
respectively, when compared with controls. The decreases were too small to be considered
statistically significant. The 15 mg/kg-day dose group showed a significant decrease in body
weight gain for GDs 9-12 only.
The fetuses did not exhibit signs of mortality, and developmental parameters were
unaffected by the treatment. The researchers noted a dose-related reduction in implantations per
doe that was consistent with a decrease in litter size, although these changes were not statistically
significant. With one exception, there were no significant external, visceral, or skeletal
malformations observed in the fetuses of treated does. In this study, treatment with tPCP up to
30 mg/kg-day did not result in developmental effects in rabbits. Since rabbits did not receive the
80 mg/kg-day dose that the rats in the Bernard and Hoberman (2001) study, it is not possible to
compare the sensitivity of rabbits with that of the CD rat. The study authors determined that the
maternal LOAEL was 15 mg/kg-day, based on significantly reduced body weight gain; the
NOAEL was 7.5 mg/kg-day. The developmental LOAEL could not be established; the NOAEL
was 30 mg/kg-day (the highest dose tested). The developmental and reproductive studies for
PCP are summarized in Table 4-19.
92
-------
Table 4-19. Summary of NOAELs/LOAELs for developmental and
reproductive studies for PCP
Species, strain
Mink
(10 F/dose)
Mink
(8 F/dose)
Sheep
(13 F/dose)
Sheep
Rat, Sprague-Dawley
(10 M and 20 F/dose)
Rat, Sprague-Dawley
(30/sex/dose)
Rat, Sprague-Dawley
(20/sex/dose)
Rat, Sprague-Dawley
(15-20 pregnant
dams/dose)
Rat, Sprague-Dawley
(10 pregnant
dams/dose)
Rat, Sprague-Dawley
(15-20 pregnant
dams/dose)
Rabbit, New Zealand
(20 pregnant
dams/dose)
Dose (mg/kg-d),
route/duration
Oorl
(feed)
One generation
Oorl
(feed)
Two generations
Oorl
(feed)
One generation
Oorl
(feed)
Two generations
0,3, or 30
(feed)
110 d, one generation
0, 10, 30, or 60
(gavage)
110 d, two generations
0,4, 13, or 43
(feed)
181 d
0,5, 15, 30, or 50
(gavage)
CDs 6-15
0,5.8, 15, 34.7, or 50
(gavage)
CDs 6-15
Oor60
(gavage)
CDs 8, 9, 10, 11, 12, or
13-20
0, 10, 30, or 80
(gavage)
CDs 6-15
0,7.5, 15, or 30
(gavage)
CDs 6-18
Grade/type of
PCP
aPCP
PCPb
PCPb
PCPb
EC-7
tPCP
aPCP
aPCP
tPCP
aPCP
tPCP
tPCP
NOAEL
(mg/kg-d)
1
Not
established
1
Not
established
3
Not
established
4
Not
established
5.8
Not
established
30
30
LOAEL
(mg/kg-d)
Not
established
1
Not
established
1
30
10
13
5
15
60
80
Not
established
Reference
Beard et al.
(1997)a
Beard and
Rawlings (1998)a
Beard et al.
(1999a)a
Beard et al.
(1999b)a
Schwetz et al.
(1978)
Bernard et al.
(2002)a
Welsh et al.
(1987)a
Schwetz et al.
(1974a)a
Larsen et al.
(1975)
Bernard and
Hoberman
(2001)
Bernard et al.
(2001)
aNOAELs and LOAELs determined by the EPA for these studies; values are for both genders unless otherwise
specified.
bPurity not reported.
93
-------
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES
4.4.1. Oral
4.4.1.1 Acute Studies
The oral median lethal dose (LD50) for male and female rats receiving tPCP (90.4%) by
gavage was reported as 155 mg/kg for males and 137 mg/kg for females by Norris (1972).
Deichmann et al. (1942) reported oral LDso values of 27.3 mg/kg for rats administered PCP in
0.5% Stanolex fuel oil, 77.9 mg/kg for PCP administered in 1% olive oil, and 210.6 mg/kg for
sodium pentachlorophenate administered in a 2% aqueous solution. Oral LDso values for mice,
rats, and hamsters ranged from 27 to 175 mg/kg as reported by the International Agency for
Research on Cancer (IARC, 1999). Clinical signs observed in dogs, rabbits, rats, and guinea pigs
consisted of increased blood pressure, hyperpyrexia, hyperglycemia, glucosuria, and
hyperperistalsis; increased urinary output followed by decreased urinary output; and rapidly
developing motor weakness. Dying animals showed signs of complete collapse, asphyxial
convulsive movements, and rapid onset of rigor mortis upon death. Necropsy examinations
showed vascular damage with heart failure and involvement of parenchymous organs
(Deichmann et al., 1942).
4.4.1.2. Immunotoxicity Studies
McConnachie and Zahalsky (1991) reported that 38 individuals exposed to PCP (in
PCP-treated log homes) for various times ranging from 0 to 13 years had activated T-cells,
autoimmunity, functional immunosuppression, and B-cell dysregulation. In addition, females,
but not males, exhibited statistically significantly increased natural killer cell function. The
exposed individuals consisted of 17 females 9-60 years of age (mean: 30.1 years) and 21 males
8-60 years of age (mean: 31.8 years). The exposed group was compared with a control group
consisting of 120 individuals; 81 females and 39 males ranging in age from 11 to 50 years and
from 24 to 67 years, respectively. Blood serum PCP concentrations ranged from 0.01 to
3.40 ppm (blood serum of 17 individuals was not analyzed for PCP content).
Daniel et al. (1995) studied immune response using peripheral lymphocytes from
188 patients exposed to PCP-containing pesticides for >6 months. Of those tested, the mitogenic
response was impaired in 65% of patients. The likelihood of an impaired response was greatest
in patients with blood PCP levels >10 ug/L (68%) and particularly for those with levels >20 ug/L
(71%). Only 50% of patients with blood levels <10 ug/L had impaired immune response. The
impaired response persisted for up to 36 months in some patients. Patients with impaired
mitogenic response were also likely to have significantly elevated (3.2-fold) interleukin-8 levels
and increased proportion of peripheral monocytes (18%) compared with patients with normal
responses. The study authors concluded that PCP-exposed patients had moderate to severe
immune dysregulation involving T and B lymphocytes. They further noted that immune
94
-------
dysfunction may explain chronic infection, chronic fatigue, and hormonal dysregulation seen in
PCP-exposed patients.
Exon and Roller (1983) conducted a study in rats to examine the effects of aPCP (97%
purity) on cell-mediated immunity, humoral immunity, and macrophage function. Groups of
male and female Sprague-Dawley rats were administered 5, 50, or 500 ppm aPCP (estimated
average dose of 0.4, 4, or 43 mg/kg-day for males and 0.5, 5, or 49 mg/kg-day for females)
continuously in the diet from weaning until 3 weeks after parturition. Offspring were treated
similarly to the parents and treatment continued until 13 weeks of age. Immune response of
offspring showed significant depression at all doses for cell-mediated immunity measured by
delayed-type hypersensitivity reaction and humoral immunity measured by antibody production
to bovine serum albumin (BSA). However, a clear dose-response relationship was not seen for
either endpoint. In contrast to the lack of effect of aPCP in adult rats, exposure of rat offspring
from the time of conception to 13 weeks of age produced effects on both humoral and cell-
mediated immunity. Macrophage function measured by the rats' ability to phagocytize sheep red
blood cells (SRBCs) increased in a dose-related manner that was statistically significant at 4 and
43 mg/kg-day for males and 5 and 49 mg/kg-day for females. In addition, there was an increase
in the number of macrophages harvested from the peritoneal exudate.
An NTP study (1989) conducted in B6C3Fi mice assessed the immunotoxic effect of
aPCP at 200, 500, or 1,500 ppm; DP-2 and EC-7 at 200, 600, or 1,200 ppm; and tPCP at 200,
600, or 1,800 ppm in the diet for 6 months. Immunotoxicity was determined by measuring
hemagglutination liters and plaque-forming cells (PFCs) in response to SRBC immunization.
Mice showed marked decreases of 89 and 57% in PFCs in spleen cells in animals treated with
200 and 600 ppm tPCP (38 and 301 mg/kg-day for males; 52 and 163 mg/kg-day for females)
respectively, and 45, 56, and 85% with 200, 600, and 1,200 ppm DP-2 (40, 109, and 390 mg/kg-
day for males; 49, 161, and 323 mg/kg-day for females), respectively. EC-7 and aPCP
measurements of PFCs increased and decreased, respectively, relative to controls, although
results were not dose related. The hemagglutination liters were decreased in mice exposed to
tPCP and DP-2, similar to the PFC response but with less consistency. The investigators
suggested that this may have been due to the lack of sensitivity of the test. No dose-related
effects were observed in measurements of hemagglutination with EC-7 or aPCP exposure.
Kerkvliet et al. (1982a) assessed the humoral immune response in groups of random-bred
Swiss-Webster female mice fed tPCP (86% purity) at concentrations of 50, 250, or 500 ppm
(estimated doses of 10, 51, or 102 mg/kg-day, respectively) and in B6 female mice fed 50, 100,
or 250 ppm (estimated doses of 10, 20, or 49 mg/kg-day, respectively) for 8 weeks. In a separate
experiment, groups of Swiss-Webster female mice were fed 250 ppm (51 mg/kg-day) tPCP with
serial sacrifice at 2-week intervals during an 8-week feeding period and an 8-week recovery
period to determine the time of onset and recovery from PCP-induced toxicity. In addition,
groups of B6 female mice were fed 1,000 ppm (195 mg/kg-day) aPCP (>99% purity) for 8 weeks
95
-------
to assess the effect on immune function of a dose of aPCP fourfold higher than the tPCP dose.
The effect of tPCP on the primary and secondary splenic antibody response to T-dependent
SRBCs in Swiss-Webster mice was measured using the hemolytic antibody isotope release
assay. The direct effect of tPCP on B-cells in B6 mice was measured using the splenic hemolytic
plaque assay and the serum antibody response to the T-independent antigen, 2,4-dinitrophenyl-
amincethylcarbamylmethyl-Ficoll (DNP-Ficoll).
tPCP caused a dose-dependent suppression of the primary and secondary T-dependent
immune responses in Swiss-Webster mice and the T-independent immune response in B6 mice.
The kinetics of the response, peak of the response, and/or the magnitude of the pre- and postpeak
antibody response to SRBCs were affected by tPCP at all doses. The IgM response was more
sensitive to tPCP exposure than the IgG response. The serial sacrifice study in Swiss-Webster
mice showed that significant immunosuppression was evident after only 2 weeks of tPCP
treatment and persisted for the 8-week treatment and recovery periods. In contrast to tPCP,
aPCP at a fourfold higher dose had no effect on humoral immune response in mice.
Kerkvliet et al. (1982b) studied the effect of tPCP and aPCP on susceptibility of mice to
tumor growth and viral infection by assessing the function of cytotoxic T-cells and phagocytic
macrophages. Male B6 mice were administered aPCP (>99% purity) or tPCP (86% purity) in
the diet at concentrations of 50 or 500 ppm (average estimated doses of 10 or 102 mg/kg-day) for
12 weeks before testing for immune competence. In vivo immunotoxicity tests included:
(1) growth of transplanted syngeneic 3MC-induced sarcoma cells, (2) susceptibility to Moloney
sarcoma virus (MSV) inoculation followed by challenge with MSV-transformed tumor cells
(MSB), and (3) susceptibility to encephalomyocarditis virus (EMCV) infection.
Progressive tumor growth was not affected by aPCP; the incidence was 35% for controls
and 31 and 40% for the 10 or 102 mg/kg-day dose groups, respectively. The incidence of
progressive tumor growth in tPCP-treated animals was significantly increased to 67 and 82% at
10 and 102 mg/kg-day, respectively. After MSV inoculation, all animals developed primary
tumors that regressed, although at a slower rate in mice treated with 102 mg/kg-day tPCP. The
tumor reappeared in 55% of the 102 mg/kg-day tPCP mice and two additional mice developed
secondary tumors after challenge with MSBs for a total incidence of 73%. Secondary tumors
developed in only 19% of controls and 18% of aPCP-treated mice, while 45% of tPCP-treated
mice (10 mg/kg-day) developed secondary tumors. Splenic tumors were observed in
MSB-challenged animals administered 10 (22%) and 102 mg/kg-day (44%) aPCP and 10 mg/kg-
day (50%) tPCP, but not in the remaining 102 mg/kg-day tPCP-treated animals. In contrast to
increased tumor susceptibility, susceptibility to EMCV-induced mortality was not significantly
affected by either aPCP or tPCP. Of particular interest is the observation that treated mice
showed significant depression of T-lymphocyte cytolytic activity and enhancement of
macrophage phagocytosis after tPCP treatment, but not after aPCP treatment. It is possible that
these immune effects could be the result of exposure to the dioxin-like contaminants present in
96
-------
tPCP (and not present in aPCP). However, Exon and Roller (1983) reported significant increases
in macrophage phagocytosis in aPCP-treated rats.
Kerkvliet et al. (1985a) conducted a study to examine the effect of tPCP on the humoral
immune response. B6C3Fi mice were administered 15, 30, 60, or 120 mg/kg tPCP (86% purity)
by gavage 2 days before challenge with SRBCs. The peak splenic IgM antibody response was
measured 5 days after the challenge. The 120 mg/kg dose was given in two 60 mg/kg fractions
on 2 consecutive days because a single 120 mg/kg dose was lethal to about one-half of the group
of 32 animals. A dose-related immunosuppressive effect was observed with a 50% response
(ID50 = median inhibitory dose) relative to controls at 83 mg/kg. aPCP (99% purity) at the same
doses had no effect on the IgM antibody response. The investigators tested three contaminant
fractions from tPCP at doses equivalent to that of the tPCP IDso dose and found that the
chlorinated dioxin/furan fraction had a significant immunosuppressive effect, whereas
chlorinated phenoxyphenol and the chlorinated diphenyl ether fractions were ineffective.
Additionally, a comparison was made regarding the immunosuppressive effect of dietary
tPCP administered for 6 weeks to two strains of mice (B6C3Fi and DBA/2) at 10 or 250 ppm
(average doses estimated as 2 and 49 mg/kg-day, respectively). Following tPCP administration,
B6C3Fi mice exhibited a greater immunotoxic effect than DBA/2 mice. The antibody response
was suppressed 28 and 75% at 2 and 49 mg/kg-day tPCP, respectively, in B6C3Fi mice
compared with no significant suppression and 45% in DBA/2 mice, respectively. The
investigators attributed the difference in the two strains to Ah-receptor responsiveness in B6C3Fi
mice and Ah-receptor-nonresponsiveness in DBA/2 mice (Kerkvliet et al., 1985a).
In another study, Kerkvliet et al. (1985b) examined the sensitivity of T-cells,
macrophages, and natural killer cells in naive and interferon-induced female C57BL/6J (B6)
mice to tPCP (86% purity) administered in the diet at concentrations of 100, 250, or 500 ppm
(estimated average doses are 20, 49, or 98 mg/kg-day, respectively) for 8 weeks. Immune
function tests included T-cell (concanavalin A and phytohemagglutinin induced) and B-cell
mitogenesis (lipopolysaccharide [LPS] induced), mixed lymphocyte response (proliferation and
cytotoxicity), spontaneous and boosted natural killer cytotoxicity, and phagocytic activity of
resident peritoneal macrophages (thioglycollate-induced and tumor activated). Body weight was
not affected, but the relative liver weights were significantly increased at all doses. The only
effect observed was the mixed lymphocyte proliferative response to allogeneic stimulation.
However, there was no effect on the generation of cytotoxic effector cells (measured by response
to P815 mastocytoma cells); the peak proliferative response of mixed lymphocyte cultures did
not show a clear dose-response. The T- and B-cell mitogenic response, natural killer cell
activity, macrophage phagocytic activity, and bone marrow cellularity were not affected by
exposure to tPCP. The investigators attributed the differences (i.e., humoral immunity was
affected by tPCP, but cellular immunity was not) in response of humoral and cell-mediated
immunity to inhibitory effects of tPCP.
97
-------
Holsapple et al. (1987) administered PCP by gavage to groups of eight female B6C3Fi
mice at doses of 10, 30, or 100 mg/kg-day tPCP (purity not reported) or 100 mg/kg-day EC-7
(purity not reported) for 14 consecutive days. Spleen cells were harvested, cultured, and exposed
to three antigens (LPS, DNP-Ficoll, and SRBCs) on day 15. Neither tPCP nor EC-7 affected the
antibody response in the splenic cells immunized in vitro to LPS, DNP-Ficoll, or SRBCs. In
another experiment, animals were treated as described above, but on day 10 or 11, the mice were
immunized with SRBCs and sacrificed on day 15. The response of IgM-producing spleen cells
was decreased in a dose-related manner with tPCP; the lowest dose of 10 mg/kg-day resulted in
statistically significant reductions of 44 and 31% on day 4 (peak response) and day 5,
respectively, compared with the controls. The study authors did not determine LOAEL/NOAEL
levels.
White and Anderson (1985) demonstrated that tPCP (90.4% purity) administered to
B6C3Fi mice by gavage for 14 days inhibited the functional activity of complement measured by
the microtiter hemolytic assay. The classical complement, spontaneous autoactivation, and
alternative pathways were inhibited at the high dose (100 mg/kg). At 10 and 30 mg/kg, tPCP
resulted in inhibitory effects that were less pronounced than high-dose effects. Animals that
returned to the control diet after the 14-day treatment period showed only a partial recovery by
30 days postexposure. Animals treated with 100 mg/kg EC-7 (91.0% purity, which contains
relatively fewer dibenzo-p-dioxin/dibenzofuran contaminants compared with tPCP) exhibited no
effects on complement levels. The investigators concluded that a contaminant or contaminants
were responsible for the effect on the complement system.
In a study on cattle, McConnell et al. (1980) administered groups of three yearling (10-
14 months old) Holstein cattle 100% aPCP, 10% tPCP/aPCP mix, 35% tPCP/aPCP mix, or tPCP
to determine the effect of the level of contaminants in PCP. Each treatment group was given
647 ppm as PCP in feed (20 mg/kg-day body weight) for 42 days and then 491 ppm (15 mg/kg-
day body weight) for 118 days of the study (total treatment time =160 days). A group of three
yearlings served as controls. McConnell et al. (1980) reported that IgG2 levels decreased as the
proportion of tPCP increased. The decrease in IgM levels did not show a dose-related trend.
Lymphocyte proliferation was increased in calves treated with tPCP following Concanavalin A
and pokeweed mitogen activation. The increase was both time- and dose-related. Proliferation
was not enhanced with the administration of aPCP, possibly suggesting that the dioxin/furan
contaminants within tPCP were responsible for the proliferation.
Two groups of four female Holstein-Friesian cattle received either a control diet or
tPCP-treated (purity 85-90%) diet corresponding to a dose of 0.2 mg/kg-day for 75-84 days
followed by 2.0 mg/kg-day for 56-62 days (Forsell et al., 1981). Immunologic parameters
measured included peripheral T- and B-cell populations, serum IgG, IgA, and IgM levels,
mitogen-induced lymphocyte blastogenesis, and antibody response to SRBCs. The investigators
observed no treatment-related effect on immune function in lactating cattle fed tPCP for up to
98
-------
146 days. These results are in contrast to those reported by McConnell et al. (1980), although
the doses used by McConnell et al. (1980) were 7-10 times greater than the highest dose used by
Forselletal. (1981).
4.4.1.3. Thyroid Hormone Studies
Jekat et al. (1994) conducted a study to examine the effect of aPCP and tPCP (purity not
reported) on thyroid hormones in female Wistar rats maintained on a normal iodine diet or a low
iodine diet and pretreated with propylthiouracil to exacerbate the thyroid deficiency. Each group
of eight female rats was administered 3 mg/kg-day tPCP, 3 or 30 mg/kg-day aPCP, or the vehicle
only (0.5% tylose solution). The test materials were administered by gavage, twice a day at
12-hour intervals, 7 days/week for 28 days. Iodine deficiency caused a 182% increase in thyroid
weight and decreased levels of total and free serum T4 and T3 and thyroid gland T4, and T3, and a
decrease in the T4:T3 ratio in the serum and thyroid gland.
Treatment with 3 mg/kg-day aPCP caused decreases in total and free serum T4, T4:T3 ratio
in serum, and serum TSH. Treatment with 3 mg/kg-day tPCP caused decreases in serum T4,
serum T3, T4, and T3 in the thyroid, T4:T3 ratio in serum, and serum TSH. Except for serum TSH,
aPCP caused greater decreases in thyroid measurements for iodine-deficient rats than in normal
rats. Because TSH levels were not elevated in response to the reduced thyroid hormone levels,
the investigators concluded that PCP interfered with thyroid hormone regulation at the
hypothalamic and pituitary levels. They also stated that peripheral interference with thyroid
hormone metabolism was suggested by the greater reduction in T4 compared with T3. The study
authors concluded that the NOAEL for this study was 3 mg/kg-day.
In a study by Rawlings et al. (1998), mature ewes in age groups of 1, 1-2, and >3-4 years
were given capsules directly into the rumen twice weekly for approximately 6 weeks. The
capsules contained 2 mg/kg aPCP (99.9% purity) or were empty (control). Blood was collected
for serum analysis of T4, LH, FSH, estradiol, progesterone, cortisol, and insulin on day 36 of
treatment. A marked decrease in serum T4 levels was observed in mature ewes at 36 days. In
addition to statistically significant decreased serum T4 levels, aPCP-treated ewes had
significantly increased serum insulin levels. However, no treatment-related changes were
observed in cortisol, LH, FSH, estradiol, or progesterone levels. No clinical signs or treatment-
related weight changes were observed during treatment. The only microscopic change observed
was increased severity of intraepithelial cysts in both oviducts.
In a study on cattle, McConnell et al. (1980) administered groups of three yearling (10-
14 months old) Holstein cattle 100% aPCP, 10% tPCP/aPCP mix, 35% tPCP/aPCP mix, or tPCP
to determine the effect of the level of contaminants in PCP. Each treatment group was given
647 ppm as PCP in feed (20 mg/kg-day body weight) for 42 days and then 491 ppm (15 mg/kg-
day) for 118 days of the study (total treatment time =160 days). A group of three yearlings
served as controls. Treatment with aPCP caused statistically significant decreases in serum T4
99
-------
(60-71% of control level) and T3 levels (56-65% of control level). The effect on thyroid
hormones is attributable to PCP and not the contaminants, because hormone levels were similar
among all treated groups of various grades of PCP. The investigators noted that thyroid follicles
were smaller and more numerous in animals receiving 100% tPCP; they did not describe the
thyroid of animals receiving aPCP.
Hughes et al. (1985) fed tPCP (85-90% purity) or aPCP (99.02% purity) to 15 Holstein
bull calves (7 days old) twice daily at doses of 0, 2, or 20 mg/kg-day. One calf in each of the
high-dose groups fed aPCP or tPCP died after acute toxicity (elevated temperature, rapid
respiration, severe diarrhea, acute purulent pneumonia). After 5 days, the doses of 2 and
20 mg/kg-day were lowered to 1 and 10 mg/kg-day, respectively, and treatment was continued
for a total duration of 42 or 43 days. Thyroid hormone levels in serum were measured during the
first 35 days of treatment. Serum T3 levels were reduced by 58-69% after treatment with
10 mg/kg-day tPCP and 49-55% with 10 mg/kg-day of aPCP. Treatment with 1 mg/kg-day
reduced serum T3 levels 44-56% with tPCP and 22-27% with aPCP. Reductions of 37-58 and
25% were observed in the calves' serum T4 levels following treatment with 1 mg/kg-day tPCP
and aPCP, respectively. T3 and T4 responsiveness to the TRH challenge were not affected by
treatment with either grade. Organ weights most notably affected by PCP treatment were
thymus and spleen in calves treated with 10 mg/kg-day tPCP or aPCP. The thymus weight was
reduced by 83% with tPCP and 54% with aPCP. Microscopic lesions consistent with thymus
atrophy were observed in tPCP-treated calves. Spleen weights were reduced by 52% with
10 mg/kg-day tPCP and by 32% with 10 mg/kg-day aPCP. Squamous metaplasia was observed
in the Meibomian gland of the eyelid of the three calves treated with 10 mg/kg-day tPCP, but in
none of the calves treated with aPCP. The investigators attributed the above eye effects to
contaminants in PCP and not to PCP itself.
Beard and Rawlings (1998) examined reproduction in a two-generation study in mink
exposed to 1 mg/kg-day PCP (purity not reported); 10 controls/generation were included. Dams
(number of animals not reported) were administered PCP in feed 3 weeks prior to mating and
continued through gestation until weaning of offspring (8 weeks postpartum). Eight
Fl generation females (from treated dams) were administered PCP in their feed starting at
weaning and maintained on the treated diet as animals grew and were mated with untreated
males. Treatment continued throughout gestation and lactation, and was terminated with
sacrifice of Fl females 3 months after the end of the lactation period. Six Fl generation males
were administered PCP in their feed starting at weaning until maximal development of the testis
(approximately 42 weeks of age), at which time the Fl males were sacrificed. Ten F2 generation
females were administered PCP-treated feed from weaning until mink reached full body size
(approximately 30 weeks of age). Eight F2 generation males were administered PCP-treated
feed from weaning until the mink reached sexual maturity in their first breeding season. The
study authors noted that all of the animals received PCP-treated feed continuously from
100
-------
conception to maturity. T4 secretion was presented graphically in Beard and Rawlings (1998);
therefore, percent changes are reported as approximate values estimated from the graphs.
Observed treatment-related effects included a statistically significant decrease in serum T4
secretion in the Fl (21%) and F2 (18%) males and in the F2 females (17%). Thyroid mass was
decreased in both Fl and F2 animals, although the reduction was statistically significant only in
F2 females (27%).
In a one-generation study, groups of 13 ewes (1-3 years old) received an untreated diet or
a diet treated with PCP (purity not reported) at a concentration delivering a dose of 1 mg/kg-day
(Beard et al., 1999a). The ewes were treated for 5 weeks prior to mating (with untreated rams),
during gestation, and until 2 weeks after weaning their lambs. The ewes were sacrificed at the
end of treatment. Maximum serum T4 levels in PCP-treated ewes were statistically significantly
lower (approximately 25%) than in control ewes with or without prior administration of TSH.
The decrease in serum T4 levels was observed over time, decreasing as night progressed.
Beard et al. (1999b) described a study in sheep in which the ram lambs born of five ewes
maintained on untreated or PCP-treated diets were examined. A dose of 1 mg/kg-day PCP
(purity not reported) was administered starting at week 5 prior to mating and continuing through
weaning of lambs. The lambs were maintained on the same diets as the ewes from weaning until
puberty at 28 weeks of age. T4 levels were statistically significantly lower than control levels
from 6 to 16 weeks, similar from 18 to 26 weeks, and lower again at 28 weeks of age. The
response to TSH stimulation was unaffected by treatment with PCP. The serum levels of other
endocrine hormones were unaffected by treatment with PCP. Microscopic examination of the
testes and epididymides showed seminiferous tubular atrophy, reduced production of
spermatocytes in the seminiferous tubules, and reduced density of sperm in the body of the
epididymides, but not in the head and tail of the epididymides. The investigators attributed the
spermatogenic findings to the reduced thyroid hormone levels.
4.4.1.4. Endocrine Disruption Studies
Orton et al. (2009) analyzed several pesticides, including PCP, for their ability to act as
agonists or antagonists in estrogenic and androgenic receptor-mediated activity in vitro and in
vivo. In yeast estrogen and androgen screen assays, PCP showed no agonistic activity, but was
the most potent compound tested in antiestrogenic and antiandrogenic effects, which were
statistically significant at concentrations of 0.015-7.8 uM (p < 0.004) and 0.015-3.9 uM (p <
0.02), respectively. In an ovulation assay, the ovaries were removed from female Xenopus laevis
and monitored for dissociation of ovulated oocytes and hormone levels by radioimmunoassay
following in vitro exposure to PCP at concentrations of 0.00625, 0.0625, 0.625, 6.25, and
62.5 ug/L. At the two highest concentrations, PCP statistically significantly depressed estradiol
(62.5 ug/L,/? < 0.001; 6.25 ug/L,/? < 0.01) and testosterone (p < 0.001); 62.5 ug/L also
depressed progesterone levels (p < 0.001). These effects were concurrent with statistically
101
-------
significantly decreased ovulation at the highest three concentrations (62.5 and 6.25 ug/L,/> <
0.001; 0.625,/?< 0.01).
In the same study, adult female Xenopus laevis were consistently exposed to low,
environmentally relevant concentrations of 0.1 and 1 ug/L PCP for 6 days and monitored for
hormone fluctuations and alterations in ovarian morphology and function (Orton et al., 2009).
Measured plasma progesterone levels were slightly elevated in both dose groups compared to the
controls, although these were not significant unless both dose groups were pooled and compared
to the controls (ANOVA/? = 0.036). Alternately, the progesterone and testosterone levels from
cultured ovarian tissue were lower than controls, with the low-dose group more affected than the
high-dose group, but these data was not reported. In addition, degenerative ovarian features and
abnormal oocytes were observed at higher levels in the low-dose (6 and 22%, respectively) and
high-dose (11 and 22%, respectively) groups compared to controls (0 and 10%, respectively),
although these levels did not reach statistical significance.
4.4.1.5. Neurotoxicity Studies
4.4.1.5.1. In vitro studies. Igisu et al. (1993) demonstrated that acetylcholinesterase activity in
human erythrocytes is inhibited by PCP at temperatures ranging from 13 to 37°C. Using isolated
sciatic nerve-sartorius muscle preparations from toads, Montoya and Quevedo (1990)
demonstrated a dose-dependent irreversible reduction of end plate potential at the neuromuscular
junction using PCP (purity not reported) concentrations between 0.01 and 0.1 mM. Axonal
conduction, using an in vitro preparation of toad sciatic nerve, was shown to be blocked
(concentration- and time-dependent) irreversibly by PCP (Sigma chemical; purity not reported
but likely aPCP in the ionized form) at concentrations ranging from 0.3 to 10 mM (Montoya et
al., 1988). PCP may not have reached the site of action as effectively in the ionized form as it
would have been expected to if it were in the nonionized form. PCP was more potent
(approximately twofold) in causing axonal conduction block than procaine. The median
effective dose for PCP was 1 mM. PCP was also able to cause a dose- and time-dependent
irreversible ganglionic synaptic transmission block at concentrations ranging from 0.003 to
0.03 mM. PCP is believed to have an effect during depolarization due to interference with
Ca++ influx (Montoya and Quevedo, 1990).
Folch et al. (2009) exposed primary rat cerebellar granule neurons to 0.1-1,000 jiM PCP
in vitro for 16-hour incubations and measured cell viability, apoptosis, reactive oxygen species
(ROS) generation, and transcriptional activity of selected genes relevant to PCP-induced toxicity.
In cells exposed to 100-1,000 jiM PCP, a statistically significant and dose-dependent loss of cell
viability and increases in apoptosis, nuclear condensation, and ROS production were observed.
These effects were concomitant with a significant upregulation of genes related to oxidative
stress (catalase, glutathione-S-transferase A5, glutathione peroxidase-1, and superoxide
dismutase [SOD]-1), apoptosis (caspases 3 and 8, p53, and Bcl-2 associated death promoter), cell
102
-------
cycle control (cyclins Dl, A, and E; cyclin-dependent kinases 2 and 4; cyclin-dependent kinase
inhibitor 2B), and DNA damage (phosphorylated p53).
4.4.1.5.2. In vivo studies. Savolainen and Pekari (1979) studied the neurochemical effects of
tPCP (86.1% purity, sodium salt and 2.4% TCP) and the body burden of chlorophenols on
groups of five male Wistar rats administered tPCP in drinking water at a concentration of
20 mg/L for 3-14 weeks. One group was allowed to recover for 4 weeks (total study duration
18 weeks). tPCP and TCP levels in the liver and brain (PCP only) remained stable between
3 and 14 weeks, whereas the levels in perirenal fat continued to increase during the treatment
time. tPCP and TCP levels in liver, brain (PCP only), and fat decreased during the 4-week
recovery period. Neurochemical studies showed that acid proteinase or SOD activities in the
right cerebral hemisphere were statistically significantly increased at 8 or 14 weeks, respectively.
NADPH-diaphorase activity was statistically significantly decreased in the right hemisphere at
3 and 18 weeks. Glutathione peroxidase activity in the right hemisphere was not significantly
affected. Glutathione levels and SOD activity were decreased (statistically significant) in glial
cells at 7 and 12 weeks. Glutathione levels were not affected in neuronal cells and glutathione
peroxidase activity was not affected in glial cells. The study authors concluded that treatment
with tPCP caused transient biochemical effects in the rat brain and that the effects were
associated with body burden of chlorophenols and possibly dibenzo-p-dioxin and dibenzofuran
contaminants.
Villena et al. (1992) examined the microscopic lesions in nerves of rats receiving PCP
(purity not reported) under different experimental conditions. This study also included an
examination of lesions in kidney and liver. Groups (number not reported) of male Wistar rats
were given drinking water containing PCP at concentrations of 0.3 mM for 60 days, 1.0 mM for
60 or 90 days, 3.0 mM for 120 days, or drinking water without added PCP. Sciatic nerves were
examined by electron and light microscopy. No effects were seen in rats given 0.3 or 1.0 mM
for 60 days. Exposure to 1.0 mM PCP for 90 days or 3.0 mM PCP for 120 days caused changes
in approximately 10% of type A and B nerve fibers in the myelin sheath. The effect was more
severe in animals receiving the highest dose. Visible damage to the sciatic nerve fibers was
characterized by variable degrees of dissociation of the myelin sheath, including complete
dissociation, profound invagination of the myelin, advanced degeneration of the neuroglial coat,
and variable losses of neurotubule neurofilaments, and other axoplasmic components. The
investigators did not state whether the animals were treated with free tPCP, aPCP, or sodium
salts. This specific information is important, considering that PCP has relatively low solubility
in water (80 mg/L) (Budavari et al., 1996), while the sodium salt is freely soluble in water. It
was noted that interference with food intake (malnutrition) can impair myelin development in
maturing animals, but the study did not investigate whether PCP caused effects on body weights,
food or water consumption, or clinical signs in this study.
103
-------
As part of its investigation into the carcinogenicity of PCP in mice, NTP (1989) also
conducted studies in groups of 10 B6C3Fi mice/sex/dose to assess the neurobehavioral effect of
PCP. Estimated doses of tPCP (38 and 301 mg/kg-day for males and 52 and 163 mg/kg-day for
females), DP-2 (40, 109, or 390 mg/kg-day for males and 49, 161, or 323 mg/kg-day for
females), EC-7 (36, 124, or 282 mg/kg-day for males and 54, 165, or 374 mg/kg-day for
females), or aPCP (102, 197, or 310 mg/kg-day for males and 51, 140, or 458 mg/kg-day for
females) were administered in the diet for 6 months. Neurobehavioral effects were assessed at
weeks 5 and 26. The battery of tests included the presence or absence of autonomic signs;
pinnal, corneal, and righting reflexes; spontaneous motor activity; acoustical startle response;
visual placement response; grip strength; and rotarod tests.
At week 5, the only neurobehavioral effects observed were dose-related decreases in
motor activity and rotarod performance in mice administered tPCP. At week 26, dose-related
increases in motor activity and startle response were observed in female mice administered all
four grades of PCP, while this effect in males was only observed in those receiving tPCP. Actual
incidence data were not published in the NTP report; therefore, the effect level is not known with
certainty.
4.4.2. Inhalation
4.4.2.1. Acute Studies
Hoben et al. (1976b) conducted a study in which groups of 12 male Sprague-Dawley rats
were exposed to PCP (purity not reported) aerosols by inhalation exposure. Assuming an
inhalation rate of 80 mL/minute, rats received calculated PCP doses of 10.1 and 14.5 mg/kg
following exposure durations of 28 and 44 minutes, respectively. The dose-response curve was
very steep; 33% of animals receiving 10.1 mg/kg died and 83.3% receiving 14.5 mg/kg died.
The LD50 was 11.7 mg/kg.
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION
4.5.1. Genetic Toxicity Studies
Genotoxicity studies following PCP exposure have shown that, while mutations have not
been detected in prokaryotic systems, there is evidence both in subcellular systems and in human
cells in vitro that PCP can induce damage to DNA and proteins via oxidative mechanisms. In
addition, gene mutation and recombination in fungi, clastogenic effects in mammalian systems in
vitro, and a weakly positive indication of transplacental mutation in mice have been have been
observed in assays with PCP. TCHQ, a metabolite of PCP, has also been shown to induce DNA
damage in in vitro studies and oxidative damage in both in vitro and in vivo studies.
104
-------
4.5.1.1. In Vitro Studies
Exposure to tPCP (90.6 purity) at concentrations of 0.3, 1, 3, 10, or 30 ug/plate for
20 minutes did not induce mutations in Salmonella typhimurium tester strains TA98, TA100,
TA1535, or TA1537 with or without the microsomal fraction (S9) from Aroclor 1254-induced
rat or hamster liver (Haworth et al., 1983). Waters et al. (1982) reported PCP, at concentrations
up to 10 ug/plate, was negative for mutations in S. typhimurium (tester strains TA98, TA100,
TA1535, TA1537, and TA1538) in the presence and absence of S9. Donnelly et al. (1998)
reported no increases in mutations in S. typhimurium (tester strains TA97a, TA98, and TA100)
incubated with aPCP (>98% purity) at concentrations of 2, 20, 50, 100, or 200 ug/plate.
Buselmaier et al. (1973) reported that PCP was negative for mutations in S. typhimurium in the
presence of S9. Gopalaswamy and Nair (1992) incubated 50 or 100 ug/plate PCP with
S. typhimurium tester strain TA98, with and without S9. The changes relative to control could
not be calculated; however, the authors reported a positive response in the number of revertants
per plate (albeit a weak response) with both doses of PCP in the presence of S9 only.
Fahrig (1974) incubated 0.19 mM PCP with S. cerevisiae for 6 hours to measure the
mitotic gene conversion at the ade2 and trp5 loci. The number of convertants/105 survivors was
measured as a 15- and 12-fold increase over control at the ade2 and trp5 loci, respectively. The
survival was reported as 30%.
Jansson and Jansson (1986) reported that forward mutations (6-thioguanine resistance
[TGr]) were not induced in V79 Chinese hamster cells incubated for 24 hours with 6.25-
50 ug/mL PCP (>99.5% purity). Cell survival was reduced (100, 90, 73, 53, and 27% cell
survival) with increasing doses (0, 6.5, 12.5, 25, and 50 ug/mL, respectively). The authors
concluded that the dose-dependent decrease in survival was possibly a result of PCP-induced
inhibition of oxidative phosphorylation.
Jansson and Jansson (1991) examined the effects of two PCP metabolites, TCHQ (doses
of 4, 20, 40, and 60 uM) and TCpCAT (doses of 15, 30, 60, and 120 uM), on TGr at the
hypoxanthine phosphoribosyltransferase (HPRT) locus and ouabain resistance (OuaR) at the
Na/K-ATPase locus in V79 Chinese hamster cells in the absence of exogenous activation. The
study demonstrated that the metabolite, TCHQ, induced TGr at concentrations >20 uM.
However, TCpCAT did not induce TGr at any of the administered doses. Neither TCHQ nor
TCpCAT affected the frequency of OuaR mutants. The authors suggested that autoxidation of
TCHQ to form the semiquinone radical or ROS would result in DNA damage (Jansson and
Jansson, 1991).
Jansson and Jansson (1992) investigated the induction of micronuclei in V79 Chinese
hamster cells treated with 5, 10, 15, or 20 uM TCHQ (>99% purity) for 3 hours. The survival of
the V79 cells was significantly reduced following administration of TCHQ, and an LD50 of
12 uM was identified. Cells with micronuclei (per 2,000 cells scored) were significantly
105
-------
increased at doses >10 uM (increased threefold or more over controls) and was dose-dependent.
The 5 uM dose induced micronuclei, but the increase was not considered statistically significant.
Galloway et al. (1987) assayed chromosomal aberrations (CAs) in Chinese hamster ovary
(CHO) cells treated with 3, 10, 30, or 100 ug/mL with S9 and 10, 30, or 100 ug/mL without S9.
tPCP produced a weakly positive response with added S9 at concentrations of 80 and
100 ug/mL; the response was negative without S9. Fahrig (1974) reported a weakly positive CA
response with PCP in human lymphocytes in the absence of S9.
Galloway et al. (1987) investigated the effects of 1, 3, 10, or 30 ug/mL tPCP (91.6%
purity) in the presence and absence of S9 in CHO cells. Weakly positive results were observed
in the induction of sister chromatid exchanges (SCEs) in the absence of S9. The relative changes
in SCEs per chromosome in treated versus control cells were 98.8, 120.5, 108.4, and 113.3% for
1,3, 10, and 30 ug/mL, respectively. All but the lowest dose exhibited changes that were
statistically significant. A negative response was observed in the CHO cells treated with tPCP in
the presence of the S9 fraction.
Ehrlich (1990) showed that PCP (purity not reported) at 5, 10, or 20 ug/mL was not
effective in inducing single strand breaks (SSBs) in CHO cells, whereas its metabolite, TCHQ,
was very effective. At a concentration of 10 ug/mL, PCP failed to induce SSBs after incubating
with CHO cells for 2 hours; this concentration was only slightly toxic to cells after 3 days. After
incubation for 2 days at a concentration of 20 ug/mL, PCP stopped growth of CHO cells. At
concentrations of 2, 5, and 10 ug/mL, TCHQ caused a dose-related increase in SSBs. Toxicity
tests showed that 5 ug/mL of TCHQ inhibited growth of CHO cells, 10 ug/mL stopped growth,
and 20 ug/mL was toxic and killed the cells. Carstens et al. (1990) also found SSBs with TCHQ
exposure when they administered 50 uM TCHQ to PM2 DNA. Within 1 hour of incubation,
0.58 SSBs per PM2 DNA molecule were observed.
Dahlhaus et al. (1995) combined Chinese hamster V79 lung fibroblasts with 6.25, 12.5,
25, or 50 jiM TCHQ for 1 hour. There was no change in SSBs at doses <12.5 jiM; however,
SSB increases were statistically significant at the 25 and 50 jiM doses, compared with control.
As cytotoxicity can induce SSBs, Dahlhaus et al. (1995) also examined the cytotoxic effects of
TCHQ. The cytotoxicity at 25 jiM was statistically significant, but low, and did not parallel the
SSBs. At 50 jiM, the cytotoxicity was much greater and corresponded with an increase in SSBs.
The authors suggested that the toxic effects to the cells may also result in SSBs in DNA. In
another study, Dahlhaus et al. (1996) found that 25 |iM TCHQ or TCpBQ incubated with
Chinese hamster V79 cells significantly induced DNA fragmentation while TCoHQ, TCoBQ,
and PCP did not.
Lin et al. (2000) examined the effects of DNA fragmentation using TCHQ and TCpBQ in
the presence of the reducing agent NADPH and Cu(II), which have been shown to induce redox
cycling in quinones. Calf thymus DNA treated with either TCHQ (100 uM and 1 mM) and
100 uM Cu(II) or TCpBQ (1 and 10 uM) and 100 uM Cu(II) and NADPH caused an increase in
106
-------
SSBs that was dose-dependent. TCpBQ alone (TCHQ was not analyzed alone) did not induce
SSBs.
Epithelial cells were isolated by Tisch et al. (2005) from human nasal tissue removed in
the surgical treatment of chronic sinusitis and nasal concha hyperplasia. Cultures were exposed
to aPCP (0.3, 0.75, and 1.2 mmol) for 1 hour and then examined for single and double strand
breaks. DNA migration length was measured in treated cells and migration exceeding 35 um
was considered indicative of cell damage. There was an increase in the damaged cells observed
in the middle nasal concha with 0.3 (1.4-fold), 0.75 (2.2-fold), and 1.2 mmol/mL (2.8-fold) PCP
compared with the control. Similarly, the inferior nasal concha exhibited 1.2-, 1.7-, and 2.3-fold
increases in damaged cells compared to the control following administration of 0.3, 0.75, and
1.2 mmol/mL PCP, respectively. Cells from both the inferior and middle (location of most of the
wood dust-induced adenocarcinomas of the nose) nasal conchae were found to have severely
fragmented DNA, observed with clear dose dependence. DNA damage in the middle nasal
concha was observed in >50, 70, and 92% of PCP-treated cells. The inferior nasal concha
exhibited less sensitive effects, with only 64% of treated cells showing DNA damage at the high
dose (1.2 mmol/mL). While supportive of other in vitro testing, it should be noted that this ex
vivo work used cells lacking the protective mucosal barrier present in vivo.
Purschke et al. (2002) used normal human fibroblasts to assess DNA damage via comet
assay and DNA repair via unscheduled DNA synthesis (UDS) resulting from exposure to TCHQ
or TCBQ at concentrations up to 60 jiM. These experiments were designed to establish whether
TCHQ or its metabolic byproduct, H2O2, caused DNA damage. There were dose-dependent
increases in DNA breakage with concentrations >20 jiM H2O2 and >5 jiM TCHQ, indicating that
TCHQ caused DNA damage similar to H2O2, although at lower concentrations. TCHQ was far
more potent than H2O2 in inducing DNA damage at concentrations between 0.5 and 10 jiM,
while TCBQ was less potent than H2O2. DNA damage produced by TCHQ, as measured by the
relative tail moment, was still measurable at 24 hours after exposure, while damage produced by
H2O2 had disappeared after 6 hours. In the UDS test, TCHQ-induced [3H]thymidine
incorporation peaked at 10 jiM but fell to near-control levels at 25 jiM, while H2O2-induced
UDS continued to rise linearly up to at least 60 jiM, indicating that TCHQ inhibited repair of the
DNA damage it induced, while H2O2 did not. The fact that TCBQ, the autoxidation product of
TCHQ, did not display the same genotoxic potency as TCHQ, was seen as evidence that redox
cycling was not involved in the observed effects. The authors suggested that the
tetrachlorosemiquinone radical may be responsible for any genotoxic activity of TCHQ.
Additionally, Purschke et al. (2002) exposed human fibroblasts to TCHQ to discern
whether the semiquinone or the hydroxyl radical formed during redox cycling was responsible
for the DNA damage by comparing TCHQ with H2O2. Based on kinetics of [3H]thymidine
incorporation, the authors suggested that DNA repair may be different following TCHQ
exposure, as compared to H2O2 exposure. Mutagenicity of TCHQ, shown previously by Jansson
107
-------
and Jansson (1991) at cytotoxic concentrations, was confirmed here at nontoxic concentrations;
H2O2 did not induce mutants at concentrations 5 times higher than those needed for DNA
damage (up to 50 jiM). However, TCHQ mutation frequency (as measured in V79 cells with the
HPRT assay) was significantly increased at 5 and 7 jiM. These results confirmed the ability of
TCHQ to induce mutations and that the effect was not caused by the metabolic byproduct, H2O2.
The study indicates that in blocking DNA repair, TCHQ exposure permits sustained DNA
damage that could lead to mutations.
Synopses of findings from genotoxicity studies with PCP are provided in Table 4-20, and
results of genotoxicity studies with PCP metabolites are provided in Table 4-21.
Table 4-20. Summary of selected in vitro genotoxicity studies of PCP
Test system
Reverse mutation in S. typhimurium
Reverse mutation in S. typhimurium
Forward mutation (TGr) in V79 Chinese hamster cells
at the HPRT locus
DNA damage in Bacillus subtilis
DNA damage in S. cerevisiae D3
DNA damage in S. cerevisiae MP-1
DNA damage in polA~ Escherichia coli
SSBs in V79 Chinese hamster cells
SSBs in CHO cells
SSBs in mouse embryonic fibroblasts
Single and double strand breaks in human mucosal
cells
CAs in CHO cells
CAs in human lymphocytes
SCE in CHO cells
Result (S9)
Negative (+/-)
Negative (+)
Negative (-)
Positive
Positive
Positive (-)
Negative
Negative (-)
Negative (-)
Weakly positive (+)
Positive (-)
Negative (-)
Weakly positive (+)
Weakly positive (-)
Negative (-)
Weakly positive (+)
Reference
Haworthetal. (1983)
Gopalaswamy and Nair (1992)
Jansson and Jansson (1986)
Waters etal. (1982)
Waters etal. (1982)
Fahrig (1978)
Waters etal. (1982)
Dahlhaus etal. (1996)
Ehrlich(1990)
Wang and Lin (1995)
Tisch etal. (2005)
Galloway etal. (1987)
Galloway etal. (1987)
Fahrig (1974)
Galloway etal. (1987)
Galloway etal. (1987)
108
-------
Table 4-21. Summary of selected in vitro genotoxicity studies of metabolites
ofPCP
Test system
Result (S9)
Reference
TCHQ
Forward mutation (TGr) in V79 Chinese hamster cells
at the HPRT locus
Forward mutation (OuaR) in V79 Chinese hamster cells
at the HPRT locus
Forward mutation in V79 Chinese hamster cells at the
HPRT locus
SSBs in V79 Chinese hamster cells
SSBs in CHO cells
SSBs in human fibroblasts
SSBs in calf thymus DNA
Strand breaks in human fibroblasts
Positive (-)
Negative (-)
Positive
Positive (-)
Positive (-)
Positive
Positive
Positive
Jansson and Jansson (1991)
Jansson and Jansson (1991)
Purschke et al. (2002)
Dahlhaus et al. (1996, 1995)
Ehrlich(1990)
Carstens et al. (1990)
Lin et al. (2000)
Purschke et al. (2002)
TCoHQ
SSBs in V79 Chinese hamster cells
Negative (-)
Dahlhaus etal. (1996)
TCpBQ
SSBs in V79 Chinese hamster cells
SSBs in calf thymus DNA
Positive (-)
Positive
Dahlhaus etal. (1996)
Lin et al. (2000)
TCpCAT
Forward mutation (TGr) in V79 Chinese hamster cells
at the HPRT locus
Forward mutation (OuaR) in V79 Chinese hamster cells
at the HPRT locus
Negative (-)
Negative (-)
Jansson and Jansson (1991)
Jansson and Jansson (1991)
4.5.1.2. In Vivo Studies
A bone marrow micronucleus test was conducted using male and female CD-I mice
dosed by gavage with 24, 60, or 120 mg/kg tPCP (88.9% purity) for males and 10, 50, or
100 mg/kg tPCP for females; tPCP produced no increases in the frequency of micronuclei in this
study conducted with male and female CD-I mice (Xu, 1996).
In a bone marrow micronucleus test, male F344/N rats (five animals/dose) were treated
i.p. with 25, 50, or 75 mg/kg PCP 3 times at intervals of 24 hours (NTP, 1999). Similarly, male
B6C3Fi mice were treated with 50, 100, or 150 mg/kg PCP. Neither the rats nor the mice
showed an increase in micronucleated polychromatic erythrocytes at any dose of PCP. The high
dose was lethal in the rats (75 mg/kg) and mice (150 mg/kg).
Daimon et al. (1997) conducted an in vivo/in vitro study that showed that PCP (purity not
reported) induced a significant increase in SCEs in hepatocytes isolated from male F344 rats
injected i.p. with 10 mg/kg PCP. This was not accompanied by an increase in replicative DNA
synthesis, indicating that cell proliferation was not a factor in SCE induction. CAs, however,
were not observed in these cells.
109
-------
Spalding et al. (2000) used nine chemicals, among them PCP (purity not stated), in two
different transgenic mouse models: the heterozygous p53 knockout (p53+/-) mouse that is able
to discriminate between genotoxic carcinogens and noncarcinogens and the v-Ha-ras gene
(Tg-AC) transgenic mouse that can differentiate between genotoxic and nongenotoxic
carcinogens and noncarcinogens. The findings were compared with results from standard 2-year
bioassays conducted by NTP. PCP was administered to p53+/- mice for 26 weeks at 100, 200,
or 400 ppm in the feed (estimated doses of 18, 35, or 70 mg/kg-day, respectively) and to TgAC
mice via skin painting 5 days/week for 20 weeks at 30, 60, or 120 mg/kg-day. All doses used in
this study were based on maximum tolerated doses (MTDs) from the corresponding 2-year
bioassays. The highest dose of PCP in the feed, 400 ppm, caused signs of liver toxicity in the
p53+/- mice, indicating that the MTD had been reached, but did not induce any tumors. In the
Tg-AC mice, however, PCP induced papillomas in a dose-dependent fashion, with time-to-tumor
decreasing with increasing dose, and tumor multiplicity increasing with dose. PCP induced
some mortality in this study, but it showed inverse dose dependence (i.e., the highest mortality
[38.5%] was observed at the lowest dose).
Yin et al. (2006) exposed 10 adult zebrafish/dose to 0.5, 5.0, or 50 ug/L aPCP (>98%
purity) for 10 days to examine point mutations in the p53 gene. The number of mutated
molecules measured in amplified liver cells of the zebrafish was significantly increased in the
5 and 50 ug/L dose groups compared with the control plasmid. The mutation rates were 7.33 x
10"4 and 10.73 x 10"4 at 5 and 50 ug/L aPCP, respectively. These mutation rates were more than
threefold greater than those in control. The authors suggested that the induction of point
mutations in p53 at concentrations as low as 5 ug/L aPCP may play a role in the carcinogenesis
of PCP.
Peripheral lymphocytes of 22 male workers engaged in the manufacture of PCP
(8 workers) or sodium-PCP (14 workers) were analyzed for chromosome aberrations; all
22 workers were smokers (Bauchinger et al., 1982; Schmid et al., 1982). Airborne PCP
concentrations during the 3 years before the analysis showed 18/67 measurements <0.01 mg/m3
and 10/67 measurements >0.5 mg/m3 for the PCP workplace and 7/55 measurements <0.1 mg/m3
and 8/55 measurements >0.5 mg/m3 for the sodium-PCP workplace. The results for the workers
exposed to PCP were compared with a group of 22 controls matched for age and social
environment; 9 were smokers and 13 nonsmokers. The frequency of chromosome type
aberrations (dicentrics and acentrics) was increased in PCP-exposed workers compared with the
controls. The frequency of chromatid type aberrations (breaks and exchanges) was not
statistically significantly increased compared with controls. A comparison of the SCE frequency
in PCP workers who were all smokers with that of control smokers and control nonsmoker
subgroups showed that the SCE frequency could be attributed to smoking and not to PCP
exposure.
110
-------
Ziemsen et al. (1987) studied the frequency of SCEs and CAs in the lymphocytes of
20 adult workers occupationally exposed to airborne PCP at concentrations ranging from 1.2 to
180 ug/m3 for 3-34 years. Fourteen workers were smokers and six were nonsmokers. Some
workers were exposed via inhalation to dry PCP (96% pure) dust, technical water-soluble
sodium-PCP (85% pure), or finished PCP solutions. Blood PCP concentrations ranged from
23 to 775 ug/L serum. No exposure-related effect was observed on the frequency of SCEs or
chromosome aberrations in these 20 workers.
Table 4-22 presents a synopsis of the result from selected in vivo genotoxicity studies
with PCP.
Table 4-22. Summary of selected in vivo genotoxicity studies of PCP
Test system
Micronucleus formation in mice
Micronucleus formation in rats
Sex-linked recessive lethal mutation in Drosophila melanogaster
Point mutations inp53 gene in hepatocytes of zebrafish
Tumor multiplicity in v-Ha-ras transgenic mice TG- AC)
CAs in human lymphocytes
CAs in human lymphocytes
CAs in male rat hepatocytes
SCE in human lymphocytes
SCE in human lymphocytes
SCE in male rat hepatocytes
Result
Negative
Negative
Negative
Positive
Positive
Weakly positive
Negative
Negative
Negative
Negative
Weakly positive
Reference
NTP (1999); Xu (1996)
NTP (1999)
Vogel and Chandler (1974)
Yin et al. (2006)
Spalding et al. (2000)
Bauchinger et al. (1982)
Ziemsen etal. (1987)
Daimonetal. (1997)
Bauchinger et al. (1982)
Ziemsen etal. (1987)
Daimonetal. (1997)
4.5.2. DNA Adduct Formation
4.5.2.1. In Vitro Studies
Lin et al. (2000) incubated two PCP metabolites, TCHQ and TCpBQ, at concentrations of
1 or 5 mM with 500 jig calf thymus DNA for 2 hours. TCpBQ induced the formation of four
major adducts in a dose-dependent fashion. Estimated relative adduct levels (RALs) were 3.5 ±
0.93/105 total nucleotides at the high dose (5 mM). There were no adducts visible with controls.
The authors reported (data not provided) that 1 mM TCHQ (with and without Cu(II)) induced a
pattern of DNA adducts similar to those induced by TCpBQ with an estimated RAL of 5.3 ±
0.1.8/107 total nucleotides.
Additionally, Lin et al. (2000) attempted to induce depurination of these DNA adducts
using thermal hydrolysis. The stability of the four major adducts following thermal hydrolysis
indicated that apurinic (AP)/apyrimidinic sites observed with TCpBQ were not formed from
depurination/depyrimidination of the adducts.
Dai et al. (2003) incubated deoxynucleosides (2 mM) in the presence of PCP (100 |iM),
H2C>2 (100 jiM and 1 mM), and myeloperoxidase and horseradish peroxidase (HRP). They
111
-------
found formation of an adduct between the oxygen of PCP and C8 of deoxyguanosine, but not
with the three other deoxynucleosides. The reaction was specific for HRP, which is known to
oxidize PCP to the phenoxy radical. However, when these researchers used rat liver microsome
preparations with an NADPH-regenerating system and the same concentrations of PCP and
nucleoside as above, a different adduct was formed, derived from TCpBQ. The results suggest
that under in vivo conditions, PCP is likely to undergo two dechlorination steps before a DNA
adduct can be formed. In a subsequent paper, Dai et al. (2005) presented evidence that
p-benzoquinone derivatives can react with the amino and imino groups in the pyrimidine portion
of the guanosine molecule to form a tricyclic benzetheno adduct.
4.5.2.2. In Vivo Studies
Lin et al. (2002) administered PCP (purity not reported, although likely aPCP as authors
compared results to NTP [1999], which used aPCP, and earlier studies by Lin et al. [1999, 1997]
used aPCP) to groups of three or four male F344 rats at concentrations of 30, 60, or 120 mg/kg-
day for 1 day and concentrations of 30 or 60 mg/kg-day for 5 days and also obtained tissues from
the livers of 10 F344 rats fed 60 mg/kg-day aPCP for 27 weeks in a 2-year bioassay conducted
by NTP (1999). While no adducts were observed in the 1- or 5-day experiments, two adducts
were identified in liver DNA in rats treated for 27 weeks. RALs were estimated as 0.78 ±
0.04 adducts/10"7 total nucleotides. Based on the chromatographic behavior of one of the
identified adducts, the authors suggested that it was derived from TCpBQ.
The study noted that PCP-induced DNA adducts have been found at much higher
occurrences (adduct levels of 8 x 10'7, 3.2 x 10'7, and 1.7 x 10'6 for PCP, TCHQ with HRP and
H2O2, and TCBQ, respectively) in mouse liver (Bodell and Pathak, 1998), possibly as a
consequence of higher amounts of PCP quinone metabolites found in mouse liver as compared
with rat liver (Lin et al., 1997). PCP formed direct DNA adducts in vitro with HRP and H2O2,
but formed DNA adducts in vivo only after dehalogenation and quinone formation (Lin et al.,
2002).
4.5.3. Protein Adduct Formation
NTP (1999) reported protein adducts of chlorinated quinones and semiquinones in tissue
samples from F344 rats after 7 months of dosing with 1,000 ppm (60 mg/kg-day) dietary aPCP
(99% purity). The level of hemoglobin adducts was elevated in male and female rats.
Lin et al. (1999) investigated the production of chlorinated quinone and semiquinone
adducts in the livers of Sprague-Dawley rats and B6C3Fi mice following a single oral dose of 0-
40 mg/kg PCP and in male F344 rats following chronic ingestion of 60 mg PCP/kg for 6 months.
At low PCP doses (<4-10 mg/kg), TCoSQ-protein adduct formation in liver cytosol and nuclei
was higher in rats than in mice. At high PCP doses (>60-230 mg/kg), however, TCpBQ adducts
were higher in mice than in rats. Moreover, there was a fourfold difference in the nuclear total of
112
-------
quinone metabolites in the mouse compared with that in the rat (Lin et al., 1997). Lin et al.
(1999) speculated that such differences in the metabolism of PCP to semiquinones and quinones
might be responsible for the production of liver tumors in mice but not rats.
Waidyanatha et al. (1996) examined adducts to blood proteins, albumin, and hemoglobin,
in three male Sprague-Dawley rats/dose treated with a single dose (gastric intubation) of 5, 10,
20, or 40 mg/kg aPCP (99% purity). Rats were sacrificed 24 hours following administration of
PCP. Protein adducts involving reactive metabolites of PCP, TCpBQ (specifically mono-, di-,
and tri-substituted forms of chlorinated benzoquinones), TCpSQ, and TCoSQ were identified for
both albumin and hemoglobin following administration of PCP. TCoBQ adducts were not
identified in the blood of the rats in this study. The authors performed a linear regression for
each of the hemoglobin and albumin adducts in vivo as pM adducts per mg PCP/kg rat body
weight and reported the resulting slopes.
The benzoquinone adducts were detected at greater concentrations in albumin compared
with hemoglobin, while the semiquinones were present in greater amounts in hemoglobin. The
greatest concentration of adducts was observed with the tri-substituted benzoquinone, C^BQ-Y
(where Y represents the protein). For the adducts, C13BQ-Y, 2,3-Cl2BQ-Y2, 2,5-and 2,6-Cl2BQ-
Y2, C1BQ-Y3, TCoSQ-Y, and TCpSQ-Y, the slopes were reported as 79 ± 8.84, 11.4 ± 1.3,
8.28 ±1.18, ND, 47.9 ± 3.44, and 20.2 ± 4.04 for formation in hemoglobin, respectively, and
200 ± 13.3, 14.2 ± 1.65, 8.75 ±0.33, 1.06 ±0.065, 13.9 ± 1.47, and 13.7 ± 0.98 for formation in
albumin, respectively. Based on the observed proportional relationship between the adduct
levels and TCpBQ, the authors concluded that the adducts were produced in a dose-dependent
manner following administration of PCP. These results provided further evidence that PCP
administered to rodents results in the formation of adducts via the oxidative dechlorination of
PCP to reactive quinones and semiquinones.
In a second experiment, Waidyanatha et al. (1996) administered a single dose via gastric
intubation of 20 mg/kg aPCP to three male Sprague-Dawley rats/group to investigate the stability
of PCP-induced protein adducts. The eight groups of rats were characterized by the duration of
time between treatment and sacrifice of 0, 2, 4, 8, 24, 48, 168, or 336 hours. Following 8 and
24 hours, the adduct levels achieved a maximum concentration and declined at times exceeding
24 hours. Two adducts were presented to serve as a representative measurement for the
remaining identified adducts. The di- and tri-substituted benzoquinones, 2,3-Cl2BQ-Y2 and
C^BQ-Y, reached maximum levels of 8 and 60 pmol/g for hemoglobin and 150 and 800 pmol/g
for albumin, respectively (values were estimated and extracted from a graph). Elimination half-
lives for these adducts were calculated as 155 and 41 hours for the hemoglobin and albumin
adducts, respectively. Both of these durations are shorter than the normal rate of turnover for
both erythrocytes and serum albumin. The authors suggested that the adducts identified in vivo
were somewhat unstable and attributed this to continuing sulfhydryl group reactions.
113
-------
The available DNA and protein adduct studies provide further evidence that PCP, or
more specifically the quinone (hydro- or benzo-) and semiquinone metabolites of PCP, can
interact with DNA in rodents. Furthermore, the liver, considered to be the target organ of both
noncancer toxicity and carcinogenicity, is susceptible to DNA alteration via PCP exposure and
the subsequent formation of DNA and/or protein adducts.
4.5.4. Oxidative DNA Damage and 8-Hydroxy-2'-Deoxyguanosine Formation
4.5.4.1. In Vitro Studies
ROS generated by metabolic processes may have a role in PCP-induced oxidative DNA
damage. Research initiatives have focused on the question of whether ROS and/or biological
reactive intermediates (BRIs) were the ultimate causative agents in DNA damage and cancer.
Carstens et al. (1990) reported an increase in SSBs in DNA of cultured human fibroblasts
following administration of 50 jiM TCHQ. They observed highly effective suppression in
TCHQ-induced SSBs in the presence of the hydroxyl radical scavengers, dimethylsulfoxide
(DMSO), ethanol, or mannitol; the metal chelator, deferoxamine; and the enzyme catalase. The
metal chelator, diethylenetriamine pentaacetic acid (DETAPAC), and the enzyme, SOD, had
little effect on the TCHQ-induced SSBs. DMSO was similarly effective in preventing DNA
breakage induced by 10 or 30 jiM TCHQ in cultured human fibroblasts. The researchers used
electron spin resonance to show that the tetrachlorosemiquinone radical, an autoxidation product
of TCHQ, was present in the reaction mixtures at up to 60% of the original TCHQ
concentrations. Formation of this radical entails the production of superoxide radicals that
produce hydroxyl radicals. The low efficiency of SOD and DETAPAC, which block the iron-
catalyzed Haber-Weiss reaction of the superoxide radical, was seen as an indication that the
superoxide radical plays a minor role in TCHQ-induced DNA damage. However, the
suppressive effect that deferoxamine, which blocks the semiquinone radical-driven Fenton
reaction, had on the SSBs indicated that the semiquinone radical was the major DNA-damaging
agent. The high efficiency of the hydroxyl radical scavengers, however, suggested also an
important function for the hydroxyl radical. Thus, both ROS and BRI were involved in TCHQ-
induced DNA damage.
Lin et al. (2000) found a dose-dependent increase in the number of AP sites following
incubation of calf thymus DNA with 1, 2.5, or 5 mM TCpBQ. The increase over control was
roughly threefold at 5 mM TCpBQ. In another experiment, 1 or 10 uM TCpBQ was incubated
with calf thymus DNA in the presence of 100 uM NADPH and 100 uM Cu(II) to determine if
ROS formed from the redox cycling of TCpBQ induced by the reducing agent, NADPH, and
copper resulted in the AP sites previously observed with TCpBQ. At the uM concentrations,
much lower than previous concentrations (e.g., 1, 2.5, or 5 mM), TCpBQ with NADPH and
Cu(II) induced statistically significant increases in the AP sites when compared with control.
Approximately 5- and 10-fold increases in AP sites were observed with 1 and 5 uM TCpBQ,
114
-------
respectively, in the presence of NADPH and Cu(II). The authors suggested that this effect could
be attributed to redox cycling of TCpBQ.
Similar experiments with 300 uM TCHQ showed no increase in AP sites, although the
addition of 100 uM Cu(II) resulted in a sixfold increase (10.8 ± 0.5 AP sites/105 nucleotides)
over control (1.6 ± 0.2 AP sites/105 nucleotides). The increase in AP sites observed with TCHQ
and Cu(II) was dose-dependent for concentrations of TCHQ from 0.5 to 300 uM. Additionally,
the number of AP sites was reduced with the addition of 5U catalase, suggesting that hydrogen
peroxide was involved in the formation of the AP sites (Lin et al., 2001).
Jansson and Jansson (1992) showed a significant induction of micronuclei in V79
Chinese hamster cells treated with 10, 15, and 20 uM TCHQ (>99% purity). Combined
administrations of TCHQ with DMSO (a hydroxyl radical scavenger) and ethyl methane-
sulfonate (EMS; an alkylating agent) and DMSO were performed to determine if hydroxyl
radicals were involved in the TCHQ-induced chromosomal damage. A 5% solution of DMSO
combined with 15 uM TCHQ partially inhibited the micronucleus formation observed with
TCHQ alone. Because DMSO did not similarly inhibit the formation of micronuclei following
EMS treatment, the authors concluded that these results provide support for a role of hydroxyl
radicals in the chromosomal damage associated with TCHQ.
Lin et al. (2001) assayed calf thymus DNA treated with TCpBQ to determine if the
benzoquinone induced changes in the levels of oxidative DNA damage indicator 8-hydroxy-
2'-deoxyguanosine (8-OH-dG) and whether these changes were related to TCpBQ-induced AP
sites. While the control measurement of 8-OH-dG was high (the authors treated this as "an
artifact of commercial isolation"), the levels of 8-OH-dG increased in a statistically significant,
dose-dependent fashion. Approximately 2-, 2.5-, and 3-fold increases in 8-OH-dG/105 dG were
observed with 1, 2.5, and 5 mM of TCpBQ. This change in 8-OH-dG occurred parallel to
formation of AP sites, leading the authors to suggest that the AP sites formed as a result of
oxidative stress-induced DNA damage. Additionally, parallel increases in SSBs were dose-
dependent, with amplified DNA fragmentation at 1 and 10 uM TCpBQ in the presence of Cu(II)
and NADPH, but not with 5 mM TCpBQ alone.
TCHQ, at concentrations ranging from 0.5 uM to 1 mM, incubated with calf thymus
DNA failed to induce 8-OH-dG compared with controls. However, the addition of 100 uM
Cu(II) to TCHQ resulted in a statistically significant, dose-dependent increase in 8-OH-dG.
TCHQ (with 100 uM Cu(II)) at a concentration of 300 uM produced a threefold increase in
8-OH-dG/105 dG compared with controls. The authors suggested that the metal facilitated
TCHQ autooxidation, generating ROS and subsequently oxidative DNA damage. Additionally,
dose-dependent increases in DNA SSBs were observed parallel to increased 8-OH-dG levels
(Lin etal., 2001).
Naito et al. (1994) investigated the mechanism of PCP metabolite-induced DNA damage
in vitro. They incubated TCHQ with calf thymus DNA in the presence or absence of cations
115
-------
(Cu2+, Mn2+, or Fe3+) that are known to be involved in redox cycling, and found that Cu2+
facilitated 8-OH-dG formation in the presence of TCHQ. This effect was not suppressed by
typical hydroxyl scavengers but was abolished by bathocuproine (a Cu+ chelator) or catalase,
from which the authors concluded that Cu+ and H2O2 were involved in the production of reactive
species causing DNA damage. The authors concluded that it was not the semiquinone, but rather
redox cycling with superoxide and H2O2 formation and the subsequent metal-catalyzed
decomposition into hydroxyl radicals that played the crucial role in oxidative DNA damage.
Dahlhaus et al. (1995) treated Chinese hamster V79 lung fibroblasts with 0, 6.25, 12.5,
25, or 50 jiM TCHQ for 1 hour and measured 8-OH-dG formation immediately or up to 2 hours
after treatment. After normalizing for variable background levels of 8-OH-dG in control V79
cells, they found that 25 and 50 jiM (but not 6.25 and 12.5 jiM) caused approximately twofold
increases in 8-OH-dG. The 25 jiM concentration was associated with low cytotoxicity, while the
50 jiM concentration exhibited appreciable cytotoxicity. The increase in 8-OH-dG correlated
with the cytotoxicity at 25 jiM, although 50 jiM presented similar levels of 8-OH-dG as observed
with the lower dose. The increase in 8-OH-dG formation was optimal after 1 hour of TCHQ
exposure, but was much reduced after 2 hours of exposure. The authors suggested that this was a
sign of activation of a repair system in the V79 cells.
Dahlhaus et al. (1996) investigated PCP, TCHQ, TCpBQ, TCoHQ, and TCoBQ for the
ability to produce oxidative DNA damage in Chinese hamster V79 cells. Changes in 8-OH-dG
in the DNA of the V79 cells were examined after exposure for 1 hour to 25 jiM PCP or one of its
metabolites. TCHQ, TCpBQ, and TCoBQ produced 8-OH-dG at levels approximately 2- to
2.5-fold greater than those observed with either PCP or the control. TCoHQ and PCP did not
show an increase in 8-OH-dG. The authors discussed their findings in terms of redox cycling
leading to ROS (i.e., direct attack of hydroxyl radicals, excision repair of hydroxylated DNA
bases, or cytotoxic effects) as the possible causes of this DNA damage.
As a means of further investigating the mechanism of redox cycling by PCP metabolites,
electron spin resonance spin trapping was used by Zhu and Shan (2009) to identify the
metabolite involved in producing hydroxyl radicals. Previous work (Zhu et al., 2000) using the
salicylate hydroxylation method had shown that in the presence of hydrogen peroxide, both
TCHQ and TCpBQ were able to produce hydroxyl radicals in a metal-independent reaction,
implicating an alternate pathway to the Fenton reaction in producing these radicals. Based on the
reaction products, the authors determined that a novel mechanism involving a nucleophilic
reaction between TCpBQ and hydrogen peroxide leads to the formation of an unstable
hydroperoxyl-l,4-benzoquinone intermediate, which decomposes homolytically to produce
hydroxyl radicals and trichloro-hydroxy-l,4-benzoquinone radicals. However, it is not been
determined if this reaction has relevance in vivo.
116
-------
4.5.4.2. In Vivo Studies
Lin et al. (2002) administered PCP (purity not reported, although likely aPCP as authors
compared results to NTP [1999] which used aPCP, and earlier studies by Lin et al. [1999, 1997]
used aPCP) to groups of three or four male F344 rats at concentrations of 30, 60, or 120 mg/kg-
day for 1 day and concentrations of 30 or 60 mg/kg-day for 5 days. Additionally, Lin et al.
(2002) obtained tissues from the livers of 10 F344 rats fed 60 mg/kg-day aPCP for 27 weeks in a
2-year bioassay conducted by NTP (1999). The induction of the 8-OH-dG lesion in rat liver
DNA was evaluated for the rats exposed to aPCP. There was no induction in 8-OH-dG at the 30,
60, or 120 mg/kg-day dose groups treated with PCP for 1 or 5 days when compared with
controls. However, there was a statistically significant increase (1.8 ± 0.65 x 10"6) in the level of
8-OH-dG/106 dG that was twofold greater in rats fed 60 mg/kg-day aPCP for 27 weeks
compared to controls (0.91 ± 0.42 x 10"6). Lin et al. (2002) noted that the liver adducts observed
in another assay were present at levels well below (10-fold lower) the 8-OH-dG concentration.
However, it was observed that two distinct types of DNA adducts formed in parallel to the
8-OH-dG lesions in the liver of rats chronically administered PCP. DNA adducts were also
detected in rat kidney, but at levels 10-fold lower than the adducts and 8-OH-dG lesions in the
liver.
Sai-Kato et al. (1995) studied the influence of PCP on the formation of 8-OH-dG in the
liver of B6C3Fi mice administered PCP by gavage at 30, 60, or 80 mg/kg as a single dose or five
consecutive doses to groups of five male mice. A clear dose-response relationship was also
observed with both treatments (no specific trend analysis was described). The 8-OH-dG
formation after a single dose (1.4- and 1.7-fold at 60 and 80 mg/kg, respectively) and repeated
exposures (1.5-, 1.9-, and 1.9-fold at 30, 60, or 80 mg/kg-day, respectively) was statistically
significantly increased compared with controls. Formation of 8-OH-dG was specific for the
target organ, liver; no significant increase in 8-OH-dG levels was observed in the kidney or
spleen. Based on evidence of the presence of a repair enzyme for 8-OH-dG in mammalian cells
(Yamamoto et al., 1992), the finding that elevation of 8-OH-dG levels was not observed at
24 hours after a single i.p. injection of an 80 mg/kg dose of PCP suggests that repair of this
oxidative DNA damage had occurred by that time point. However, single administration via
gavage and repeat administration of PCP caused elevated levels of 8-OH-dG at low doses (30 or
60 mg/kg-day). The authors concluded that long-term exposure of PCP may induce gradual
accumulation of oxidative DNA damage in the liver by overwhelming the repair potential and
that this cumulative oxidative DNA damage could cause critical mutations leading to
carcinogenesis (Sai-Kato et al., 1995).
Umemura et al. (1996) demonstrated that feeding aPCP (98.6% purity) to male B6C3Fi
mice for 2 or 4 weeks at concentrations of 41, 86, and 200 mg/kg-day resulted in dose-
dependent, statistically significant two- to threefold increases of 8-OH-dG formation in the liver.
In addition to the dose- and time-dependent elevation of 8-OH-dG, significantly elevated
117
-------
bromodeoxyuridine (BrdU) labeling index and hepatic DNA content (indicative of
hyperproliferation) led the authors to suggest that oxidative DNA damage in combination with
hyperproliferation might cause PCP-related cancer.
Umemura et al. (1999) fed mice 600 or 1,200 ppm PCP (98.6% purity; doses estimated as
108 and 216 mg/kg-day, respectively) for 8 weeks and noted that the oxidative lesion 8-OH-dG
in liver DNA was statistically increased to 2.5- and 3.8-fold at 108 and 216 mg/kg-day,
respectively, compared with the control levels. La et al. (1998a) reported that F344 rats fed PCP
for 27 weeks showed a twofold increase in the 8-OH-dG DNA lesion in liver. Another lesion
was noted and compared with in vitro PCP metabolite adducts. This lesion co-migrated with the
TCpBQ adduct but at an absolute level threefold lower than that of the oxidative lesion.
Dahlhaus et al. (1994) showed that the PCP metabolite, TCHQ, elicited an approximately
50% increase in 8-OH-dG formation in hepatic DNA of B6C3Fi mice fed 300 mg/kg TCHQ for
2 or 4 weeks. Single i.p. injections of 20 or 50 mg/kg TCHQ had no such effect.
4.5.5. Uncoupling of Oxidative Phosphorylation
The ability of PCP to uncouple mitochondrial oxidative phosphorylation was first
described by Weinbach (1954). This study measured the net uptake of phosphate and oxygen in
rat liver mitochondria during the oxidation of a-ketoglutarate to succinate to indicate the extent
of oxidative phosphorylation uncoupling. PCP induced the uncoupling of oxidative
phosphorylation in a dose-dependent manner. At the lowest concentration tested, 10"6 M, PCP
showed signs of suppressing phosphate uptake, but this was accompanied with a stimulation of
oxidation. However, at concentrations of 10"5 and 10"4 M, PCP suppressed phosphate uptake
while having little effect on oxidation, indicative of uncoupling as respiration was being
stimulated without concomitant phosphorylation. At concentrations of >10"3 M, PCP completely
inhibited both phosphorylation and oxidation. PCP also accelerated the breakdown of
mitochondrial ATP, which the author theorized was a consequence of altered membrane
permeability, as PCP had a suppressive effect on ATPase (Weinbach, 1954).
Arrhenius et al. (1977a) observed that PCP, not a metabolite, exerted a strong inhibition
of electron transport between a flavin coenzyme and CYP450. In the second part of that study,
Arrhenius et al. (1977b) looked at the effects of PCP on cellular detoxification mechanisms.
Their main focus was to examine whether PCP acts only as an inhibitor of oxidative
phosphorylation in mitochondria or if it exerts an additional effect on the microsomal electron
transport. The experiments were conducted in vitro with the subcellular fraction from liver of
male Wistar rats, using oxygen consumption as the measure of respiration. PCP was about twice
as potent in mitochondria as the commonly used uncoupler, dinitrophenol. The authors
concluded that the parent compound, not a metabolite, was the active toxicant and that it
inhibited the electron transport from flavin to CYP450. The authors discussed their findings in
terms of a possible effect of lipophilic chlorophenols on membrane function.
118
-------
Varnbo et al. (1985) used a murine neuroblastoma-derived cell line to investigate the
influence of a variety of toxicants on respiratory activity as measured by oxygen consumption.
aPCP was used at concentrations between 100 jiM and 1 mM and caused a brief spike in oxygen
consumption followed by a dose-dependent decrease that reached approximately 70% inhibition
within 30 minutes at 1 mM aPCP.
A series of experiments was conducted with female Wistar rats that were fed 0.2% HCB
in the diet for up to 60 days (Trenti et al., 1986a, b; Masini et al., 1985, 1984a, b). PCP is
chemically similar to HCB, which is a benzene ring with a chlorine bound to each of the six
carbons. In the PCP molecule, one chlorine atom present in HCB is replaced with a hydroxyl
(OH) group, rendering the molecule somewhat electrophilic. One of the pathways for HCB
metabolism produces PCP. Animals were sacrificed at 20, 40, and 60 days of feeding, and
mitochondria were prepared from their livers. Masini et al. (1984a) observed that the porphyrin
content of liver mitochondria increased with time, but porphyrins were not detectable in urine or
feces. Using oligomycin, the authors found that the change in ratio of state 3 to state 4
respiration (i.e., respiratory control index) was due to uncoupling of oxidative phosphorylation.
The effect was reversible by addition of BSA, a scavenger for uncoupling agents. The authors
speculated that phenolic metabolites of HCB, specifically PCP, caused the uncoupling of
oxidative phosphorylation.
Masini et al. (1984b) recorded the transmembrane potentials of mitochondria from
HCB-treated animals and control mitochondria with added micromolar concentrations of PCP
and found that they were highly similar. Subsequently, the same investigators (Masini et al.,
1985) reported a time-dependent increase, up to 600-fold, of porphyrins in the urine, liver, and
mitochondria of female Wistar rats. PCP levels in livers and liver mitochondria of HCB-treated
animals rose with time in parallel with HCB levels, amounting to about 10% of the HCB load per
g of liver tissue, and per mg protein (liver mitochondria). To strengthen their hypothesis that the
HCB metabolite PCP might be responsible for the observed effects, these researchers added PCP
to a mitochondrial suspension at 0.25-2.5 jiM, which caused a dose-dependent inhibition of
oxidative phosphorylation that was reversible by the addition of BSA.
Trenti et al. (1986a) found that oxygen usage per mg mitochondrial protein was almost
doubled by treatment with either 0.2% HCB or 1 jiM PCP. The effect was fully reversible by the
addition of 0.1% BSA to the medium. The authors concluded that the increased oxygen usage
observed after HCB feeding was entirely caused by the HCB metabolite, PCP. In a parallel
experiment, Trenti et al. (1986b) fed female Wistar rats 0.2% HCB in the diet for up to 60 days
and prepared mitochondria from their livers after 20, 40, and 60 days of feeding. There was a
constant decline in the respiratory control index (ratio of state 3 to state 4 respiratory rate), the
ADP:oxygen ratio, and the transmembrane potential with time. The investigators also observed
that PCP concentrations in liver and mitochondria increased with time, paralleled by an increase
119
-------
in porphyrins. However, they concluded that porphyrin formation was unrelated to uncoupling
of oxidative phosphorylation.
4.5.6. Cytotoxicity
Freire et al. (2005) evaluated the potential cytotoxic effects of PCP on Vero monkey cells
(from the kidney of the African green monkey) by incubating cultures with PCP concentrations
of 1, 5, 10, 50, or 100 uM (0.26-26.63 ug/mL) for 24, 48, or 72 hours. There was a statistically
significant increase in cytotoxicity at the 5 uM concentration of PCP with cell viabilities of 72,
70, and 45% of the control for the 24-, 48-, and 72-hour incubation periods, respectively. The
cytotoxicity increased in a dose- and time-dependent manner. The viabilities of the Vero cells
measured at the higher concentrations of PCP were <40% of the control for all three incubation
periods.
Additionally, Freire et al. (2005) looked at effects on lysosomes and mitochondria in cells
incubated with 10, 40, or 80 uM PCP for 3 or 24 hours. Damaged lysosomes or a reduced
number of intact lysosomes increased in a dose- and time-dependent manner. Large vacuoles,
potentially indicative of lysosomal fusion or swelling, were observed at all doses after 24 hours.
A disturbance in the transmembrane potential of the mitochondria in the Vero cells was observed
after 3 hours of incubation with the 40 and 80 uM dose groups of PCP. After 24 hours, the cells
exhibited severely compromised mitochondria (with 80 uM) and statistically significant
morphological changes (chromatin condensation and nuclear fragmentation) that were indicative
of apoptosis (with all doses).
Dorsey et al. (2004) incubated alpha mouse liver 12 (AML 12) hepatocytes with PCP at
concentrations of 1.95, 3.95, 7.8, 15.6, or 31.2 ug/mL (98% purity) for 48 hours to examine the
cytotoxic effects of PCP. The viability of the cells treated with the lower doses (<7.8 ug/mL)
was greater than that measured with the control; however, at the two highest doses, 15.6 and
31.2 ug/mL, cell viability was statistically significantly reduced by >50% compared with
controls. Additionally, the authors examined morphology of the AML 12 hepatocytes following
incubation with PCP. Morphologic effects were observed as changes in cell shape and in the
monolayer after 48 hours of incubation with 15.6 ug/mL PCP.
In the same study, Dorsey et al. (2004) looked at the mitogenic effects of 0.975, 1.95,
3.95, or 7.8 ug/mL PCP on AML 12 hepatocytes after 12 and 24 hours of incubation.
Stimulatory patterns of cell proliferation in treated heptocytes were compared with untreated
cells. Cell proliferation was statistically significantly increased one- to threefold at all doses and
both durations of incubation with PCP. The authors noted that PCP was mitogenic at low doses
in the AML 12 mouse hepatocytes.
This group also observed, in previous studies, dose-dependent cytotoxic effects in HepG2
cells (LD50 = 23.0 ± 5.6 ug/mL) with decreased viabilities that were 95, 90, 40, 30, and 10% of
the control following incubation with 6.25, 12.5, 25, 50, or 100 ug/mL PCP, respectively, for
120
-------
48 hours (Dorsey and Tchounwou, 2003). The decreased cell viability was statistically
significant at all doses except the lowest dose of 6.25 ug/mL. PCP exerted mitogenic effects on
HepG2 cells with one- to fivefold increases in cell proliferation at doses ranging from 0.20 to
3.25 ug/mL (Dorsey and Tchounwou, 2003). Suzuki et al. (2001) observed cytotoxicity,
measured by release of LDH from Wistar rat hepatocytes. Cytotoxicity was significantly
increased (20-35% release of LDH) following incubation with 1 mM PCP for 1 hour compared
with controls.
4.5.7. Lipid Peroxidation
Suzuki et al. (2001) isolated Wistar rat hepatocytes and incubated them for 1 hour with
1 mM PCP (purity not reported) to examine the lipid peroxidative and cytotoxic effects. PCP
induced a slight but statistically significant increase in cellular phospholipoperoxides.
Additionally, glutathione was nearly depleted with administration of PCP. The authors
suggested that this depletion may have induced the lipid peroxidation.
4.5.8. Inhibition of Gap Junction Intercellular Communication
Sai et al. (1998) investigated the possible role that inhibition of gap junction intercellular
communication (GJIC), a nongenotoxic mechanism, may play in contributing to tumor
promotion. They used WB-F344 rat epithelial cell lines with concentrations ranging from 25 to
200 jiM PCP (<24 hours) and TCHQ (1 hour). Incubations with PCP at concentrations >40 and
>75 jiM for TCHQ were found to induce cytotoxicity. Subsequent GJIC experiments were
conducted under conditions that did not elicit cytotoxicity. A time course of GJIC inhibition by
PCP revealed a 40% inhibition by 4 hours, a return to normal levels by 6-8 hours, and a second
phase of inhibition up to 50%, lasting from 16 to 24 hours. The effect displayed dose-
dependence from 10 to 40 jiM PCP. When cells were incubated with 20 or 40 jiM PCP for 4 or
24 hours and then reincubated in the absence of PCP, normal GJIC was restored within 4-
6 hours. Four hours of exposure to 40 jiM PCP significantly reduced the levels of connexin
(CX43), a GJIC-specific protein, in WBCs but did not affect its localization on the cell surface.
Removal of PCP restored CX43 levels within 6 hours. Phosphorylation of CX43 was not
affected by 40 jiM PCP, while strong phosphorylation was achieved by the potent tumor
promoter, tetradecanoylphorbol acetate (concentration not stated). The authors concluded that
the PCP-induced GJIC inhibition was not based on changes in CX43 phosphorylation, but more
likely represented a posttranslational event. TCHQ did not affect GJIC in WBCs, but it is
possible that the time of exposure (1 hour) was too short to elicit measurable changes.
In a subsequent study, Sai et al. (2000) administered green tea (in place of drinking
water) for 3 weeks to male B6C3Fi mice. For the latter 2 weeks of treatment, the animals were
exposed to 300 or 600 ppm PCP (doses estimated as 54 and 108 mg/kg-day, respectively) via
feed (these doses were chosen because they had demonstrated tumor-promoting activity in an
121
-------
initiation-promotion assay [Umemura et al., 1999]). PCP alone inhibited GJIC up to 60% in a
dose-dependent manner; a similar, albeit reduced inhibition (maximally 10%) was observed in
the animals co-treated with green tea. Expression of CX32, another GJIC-specific marker, on
the cytoplasmic membrane was attenuated by PCP treatment. This effect was prevented by
green tea treatment.
Exposure to 54 and 108 mg/kg-day PCP in feed for 2 weeks increased cell proliferation
(as evidenced by the BrdU labeling index) 6- and 15-fold, respectively, compared with controls.
Co-treatment with green tea lessened this proliferative effect by 60-70%. Because green tea
contains highly effective antioxidants, the authors suggested that PCP caused GJIC inhibition by
means of oxidative stress. They did not elaborate as to whether the formation of oxygen radicals
and oxidative stress required metabolism of PCP (Sai et al., 2000).
Sai et al. (2001) conducted another study of the effects of aPCP on GJIC in which they
evaluated possible mechanistic links to apoptosis, using a WB-F344-derived rat epithelial cell
line. An aPCP concentration of 2 jiM was chosen for the tests based on the observation that
1 |iM was minimally effective, while 3 jiM marked the beginning of cytotoxicity. Apoptosis
was induced by serum deprivation of the cultured cells, which takes 3-6 hours to first become
evident in the form of cell detachment from the dish and is at a maximum by 12 hours after
serum removal. Three different methods were used: apoptosis staining using Hoechst 33342;
the terminal deoxynucleotidyltransferase mediated deoxyuridine 5'-triphosphate-biotin nick-end
labeling test; and DNA ladder formation. By all three measures, aPCP inhibited serum
deprivation-induced apoptosis at 2 jiM in a time-dependent manner. While serum deprivation
alone did not affect GJIC until 12 hours after removal, aPCP caused a significant inhibition of
GJIC within 1 hour. Additionally, aPCP caused up to a 60% drop in the protein level of p53, an
apoptosis-inducing protein, in the serum-deprived cells over a period of 12 hours. Subsequent
decreases in mRNA levels of p53 were observed, as well as a similar decrease in the level of
GJIC-specific CX43. The authors considered these findings evidence that aPCP inhibits GJIC
formation that would be required for propagation of the "death signal," thus preventing apoptosis
and the elimination of transformed cells. The aPCP-induced effects on p53 and CX43 may
explain the decrease in apoptosis and GJIC. It was suggested that the suppression of apoptosis
and GJIC could lead to tumor promotion.
122
-------
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
4.6.1. Oral
The liver is the primary target for noncancer effects of oral exposure to PCP. Numerous
short- and long-term oral studies show that PCP is toxic to the liver of rats, mice, and dogs (see
Table 4-23). Liver toxicity is generally manifested by increased absolute and relative weights
and a wide spectrum of microscopic lesions. Liver toxicity in long-term studies in rats was
primarily characterized by pigment accumulation (Schwetz et al., 1978), chronic inflammation at
high doses, and cystic degeneration at lower doses in males (NTP, 1999); female rats were not as
sensitive as males in the NTP study. Liver toxicity in mice exposed orally to PCP was
manifested primarily by necrosis, cytomegaly, chronic active inflammation, and bile duct lesions
(NTP, 1989). Liver toxicity was more severe in mice than rats at similar doses, which could be
partially attributable to differences in biotransformation of PCP. Additionally, rats in one of the
chronic studies (NTP, 1999) were treated with aPCP, whereas mice in the chronic NTP (1989)
study received either tPCP or EC-7 grades of PCP, which are higher in chlorinated dibenzo-p-
dioxins and dibenzofuran contaminants and may contribute to the severity of the response in
mice compared with rats. The NTP (1989) studies showed very little difference between the
toxicity of tPCP and EC-7 in mice, except for bile duct hyperplasia, which may be associated
with the impurities in tPCP. Liver lesions in the dog (Mecler, 1996) were similar to those
observed in the mouse (NTP, 1989), but the doses inducing the lesions in the dog were lower
than those that induced these lesions in the mouse (1.5 mg/kg-day compared with 17-18 mg/kg-
day for the mouse). Studies in domestic animals showed that pigs, but not cattle, exhibited liver
lesions similar to those observed in mice. The pig exhibited liver toxicity at a lower dose
(10 versus 17-18 mg/kg-day for the mouse) and for a shorter duration (30 days versus 2 years)
than the mouse. Other noncancer targets identified in long-term studies include the kidney
(pigment deposition in the proximal convoluted tubules) of rats (Schwetz et al., 1978) and the
spleen (decrease in organ weight) of mice (NTP, 1989), rats (Bernard et al., 2002), and calves
(Hughes et al., 1985).
123
-------
Table 4-23. Subchronic, chronic, developmental, and reproductive oral toxicity studies for PCP
Species, strain
Dose (mg/kg-d),
duration
Grade/type of
PCP
NOAEL
(mg/kg-d)
LOAEL
(mg/kg-d)
Effect(s) at the LOAEL
Reference
Subchronic
Mouse, Swiss-Webster
(6 females/dose)
Mouse, B6
(15-16 females/dose)
Mouse, B6
(20 males/dose)
Rat, Wistar weanlings
(10/sex/dose)
Rat, Sprague-Dawley
(number not reported)
Rat
(10 males/dose)
Rat, male Wistar
(number not reported)
0, 10, 51, or 102
(feed)
8wks
0, 10, 20, or 49
(feed)
8wks
0, 10, or 98
(feed)
12wks
0, 2, 5, or 18 (M)
(feed)
12wks
0,3,5,or21(F)
(feed)
12wks
0, 3, 10, or 30
(feed)
90 d
Oor87
(feed)
90 d
0, 80, 266, or
800 mg/L
(drinking water)
60-120 d
tPCP
aPCP
tPCP
aPCP
tPCP
Commercial
Improved
Pure
tPCP
aPCP
Not reported
10
10
Not established
2
o
J
Not established
3
o
J
Not established
80
51
20
10
5
5
3
10
10
87
266
Dose-related increases in hepatocellular
multifocal necrosis, hepatocellular and nuclear
swelling, hepatocellular vacuolation, and
eosinophilic inclusion bodies in nuclear vacuoles.
Dose-related increases in hepatocellular swelling,
nuclear swelling and vacuolation with
eosinophilic inclusion bodies.
Centrilobular vacuolationb, increased aniline
hydroxylase activity in liver microsomes.
Dose-related elevated serum ALP and increases
in liver and kidney weight.
Increased liver weight.
Enlarged liver, single hepatocellular necrosis,
hepatocellular vacuolation, cytoplasmic
inclusion, slight interstitial fibrosis, brown
pigment in macrophages and Kupffer cells, and
atypical mitochondria.
Enlarged liver, hepatocellular vacuolation,
cytoplasmic inclusion, and atypical mitochondria.
Dose-related increases in hepatocellular
degeneration and necrosis, increased granular
endoplasmic reticurum, congested portal veins,
enlarged and congested sinusoids, and bile duct
hyperplasia. Nephritis in kidney including
glomerular congestion and hyalinization.
Kerkvliet et al.
(1982a)a
Kerkvliet et al.
(1982b)a
Knudsenetal. (1974)
Johnson etal. (1973)a
Kimbrough and
Linder (1975)a
Villenaetal. (1992)a
124
-------
Table 4-23. Subchronic, chronic, developmental, and reproductive oral toxicity studies for PCP
Species, strain
Mice, B6C3FJ
(25 males/dose;
10 females/dose)
Dose (mg/kg-d),
duration
0, 38, or 301 (M)
(feed)
26-27 wks
0, 52, or 163 (F)
(feed)
26-27 wks
0, 36, 124, or 282 (M)
(feed)
26-27 wks
0, 54, 165, or 374 (F)
(feed)
26-27 wks
0,40, 109, or 390 (M)
(feed)
26-27 wks
0,49, 161, or 323 (F)
(feed)
26-27 wks
0, 102, 197, or 3 10
(M)
(feed)
26-27 wks
0,51, 140, or 458 (F)
(feed)
26-27 wks
Grade/type of
PCP
tPCP
EC-7
DP-2
aPCP
NOAEL
(mg/kg-d)
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
Not established
(M)
Not established
(F)
LOAEL
(mg/kg-d)
38 (M)
52 (F)
36 (M)
54 (F)
40 (M)
49 (F)
102 (M)
51 (F)
Effect(s) at the LOAEL
Dose-related increases in incidence and severity
of liver lesions including hepatocellular
degeneration and necrosis, karyomegaly, and
cytomegaly.
Reference
NTP (1989)a
125
-------
Table 4-23. Subchronic, chronic, developmental, and reproductive oral toxicity studies for PCP
Species, strain
Dose (mg/kg-d),
duration
Grade/type of
PCP
NOAEL
(mg/kg-d)
LOAEL
(mg/kg-d)
Effect(s) at the LOAEL
Reference
Chronic
Rat, Sherman
(10/sex/dose)
Dog, beagle
(4/sex/dose)
Rat, F344 (50/sex/dose)
Rat, Sprague-Dawley
(25/sex/dose)
Mouse, B6C3FJ
(50/sex/dose)
0, 2, 9, or 44 (M)
0, 2, 10, or 48 (F)
(feed)
8 mo
0, 1.5, 3.5, or 6.5
(gelatin capsule)
lyr
0, 10, 20, or 30
(feed)
2yrs
0, 1,3, 10, or 30
(feed)
2yrs
tPCP: 0, 18, or
35 (M); 0, 17, or
35 (F)
EC-7: 0, 18, 37, or
118(M);0, 17, 34, or
114(F)
(feed)
2yrs
tPCP
aPCP
tPCP
aPCP
EC-7
tPCP/EC-7
Not established
9(M)
10 (F)
Not established
10 (M)
20 (F)
10 (M)
3(F)
Not established
2
44 (M)
48 (F)
1.5
20 (M)
30 (F)
30 (M)
10 (F)
18 (M)
17 (F)
Dose-related increases in centrilobular
hepatocyte hypertrophy and vacuolation; at
higher doses, pleomorphism, bile duct
proliferation, adenofibrosis, cytoplasmic hyaline
inclusions, abundant brown pigment in
macrophages and Kupffer cells, and statistically
significantly increased liver weight.
Statistically significant decrease in body weight,
slight hepatocyte hypertrophy, eosinophilic
cytoplasmic inclusions, and brown pigment in
liver.
Dose-related increases in incidence and severity
of hepatocellular pigmentation, cytoplasmic
vacuolation, chronic inflammation; significantly
increased serum ALT and AST; significantly
increased relative liver weight; and increased
absolute liver wt (significant in females).
Increased cystic degeneration13 and decreased
body weight.
Decreased body weight.
Dose-related increases in pigmentation in liver.
Dose-related increases in pigmentation in liver
and kidney.
blncreased clear cell focus, acute diffuse necrosis,
diffuse cytomegaly, diffuse chronic active
inflammation, multifocal accumulation of brown
pigmentation (LF and cellular debris) in Kupffer
cells in the liver, and proliferation of
hematopoietic cells (extramedullary
hematopoiesis).
Kimbrough and
Linder (1978)a
Mecler (1996)a
NTP (1999)a
Schwetzetal. (1978)
NTP (1989)a
126
-------
Table 4-23. Subchronic, chronic, developmental, and reproductive oral toxicity studies for PCP
Species, strain
Dose (mg/kg-d),
duration
Grade/type of
PCP
NOAEL
(mg/kg-d)
LOAEL
(mg/kg-d)
Effect(s) at the LOAEL
Reference
Developmental/reproductive
Rat, Sprague-Dawley
(15-20 pregnant
dams/dose)
Rat, Sprague-Dawley
(15-20 pregnant
dams/dose)
Rat, Sprague-Dawley
(10 M and 20 F/dose)
Rat, Sprague-Dawley
(30/sex/dose)
Rat, Sprague-Dawley
(20/sex/dose)
tPCP: 0, 5.8, 15, 34.7,
or 50
aPCP: 0, 5, 15, 30, or
50
(gavage)
CDs 6-15
0, 10, 30, or 80
(gavage)
CDs 6-15
0,3, or 30
(feed)
110 d, one-generation
0, 10, 30, or 60
(gavage)
110 d, two-generations
0,4, 13, or 43
(feed)
18 Id plus CDs 1-20
tPCP
aPCP
tPCP
EC-7
tPCP
aPCP
5.8
Not established
30
3
Not established
4
15
5
80
30
10
13
Increased incidence of soft tissue and skeletal
anomalies'3.
Delayed ossification of the skullb.
Increased incidence of malformations'3 and
skeletal variations'3, and decreased live litter size
and fetal body weight.
Decreased pup survival and growth, and
increased skeletal variations.
Delay in vaginal patency13.
Increased skeletal variations'3, and dose-related
decreases in fetal body weight and crown-rump
length.
Schwetz et al.
(1974a)a
Bernard and
Hoberman(2001)
Schwetz etal. (1978)
Bernard et al. (2002)a
Welsh etal. (1987)a
aNOAELs and LOAELs determined by the EPA for these studies; values are for both genders unless otherwise specified.
bDenotes statistical significance.
F = female; M = male
127
-------
A two-generation reproductive toxicity study in rats showed that exposure to tPCP is
associated with decreased fertility, delayed puberty, testicular effects, decreased litter size,
decreased viability, and decreased pup weights at a dose of 30 mg/kg-day (Bernard et al., 2002).
These effects occurred at the same doses causing systemic toxicity in parental animals. A one-
generation reproductive study in mink (1 mg/kg-day aPCP) showed evidence of reproductive
effects in which many of the dams refused to accept the males for a second mating.
Additionally, the whelping rate was reduced (Beard et al., 1997). However, a two-generation
reproductive study of similar design reported no reproductive effects in mink administered
1 mg/kg-day PCP (Beard and Rawlings, 1998). Additionally, no effects on reproduction were
noted in sheep (both ewes and rams) at a PCP dose of 1 mg/kg-day (Beard et al., 1999a, b).
The majority of developmental toxicity studies on PCP provided no evidence of
teratogenic effects, but some older studies showed toxic effects of PCP in offspring that occurred
at dose levels below those producing maternal toxicity. In Welsh et al. (1987), effects were
observed in rat fetuses at 13 mg/kg-day compared with 43 mg/kg-day in the dams. Schwetz et
al. (1974a) similarly reported sensitivity in fetuses at 5 mg/kg-day aPCP and 15 mg/kg-day tPCP
compared with 30 mg/kg-day in the dams treated with either grade of PCP.
Studies show that treatment with PCP affected the levels of circulating thyroid hormones,
TS and T4. Serum TS and T4 levels were significantly decreased by both aPCP and tPCP in rats
(at a dose of 3 mg/kg-day [Jekat et al., 1994]) and cattle (at doses of 1 mg/kg-day [Hughes et al.,
1985] and 15 mg/kg-day [McConnell et al., 1980]). Serum T4 levels were significantly
decreased by PCP (purity not reported) in ram and ewe lambs, and mink (at a of dose 1 mg/kg-
day [Beard et al., 1999a, b; Beard and Rawlings, 1998]), and by aPCP in mature ewes (at a dose
of 2 mg/kg-day [Rawlings et al., 1998]). PCP treatment did not affect the degree to which TSH
stimulated thyroid hormone levels (Beard et al., 1999a, b). Only Jekat et al. (1994) reported
changes in TSH levels following administration of PCP to rats for 28 days. Along with a
decrease in T4, there was a noted decrease in TSH. Because TSH levels were not elevated in
response to the reduced thyroid hormone levels, the investigators concluded that PCP interfered
with thyroid hormone regulation at the hypothalamic and pituitary levels. Additionally, the
peripheral interference with thyroid hormone metabolism was suggested by the greater reduction
in T4 compared with Tj, (Jekat et al., 1994).
The mechanism by which PCP affects thyroid hormones has not been identified. Van
den Berg (1990) reported that PCP competitively binds T4 sites (i.e., for transthyretin, albumin,
and thyroid binding globulin) and consequently induces inhibitory effects. Additionally, Den
Besten et al. (1991) observed that PCP showed greater affinity for binding the T4-binding site on
thyretin (major T4 transport protein) than T4. The authors speculated that the binding to thyretin
most likely resulted in the effects on thyroid homeostasis (Den Besten et al., 1991). Considering
that similar effects were observed in rats and cattle with both tPCP and aPCP, the effect on
serum thyroid hormone levels is attributed to PCP and not its impurities.
128
-------
Studies examining the immunotoxic effects of PCP showed that the humoral response
and complement activity in mice were impaired by tPCP, but not by aPCP, when administered to
adult animals (at doses as low as 38 mg/kg-day [NTP, 1989]; 10 mg/kg-day [Holsapple et al.,
1987; Kerkvliet et al., 1982a, b]; and 2 mg/kg-day [Kerkvliet et al., 1985a, b]). Treatment of
mice with doses as low as 4 mg/kg-day from the time of conception to 13 weeks of age resulted
in impaired humoral- and cell-mediated immunity (Exon and Roller, 1983). Blood
measurements in humans with known exposure to PCP showed that immune response was
impaired in patients who had blood PCP levels >10 ug/L and in particular in those whose levels
were >20 ug/L (Daniel et al., 1995; McConnachie and Zahalsky, 1991).
In vitro neurotoxicity studies showed that 0.003-0.03 mM PCP caused a dose-dependent
irreversible reduction in endplate potential at the neuromuscular junction and interference with
axonal conduction in the sciatic nerve from the toad (Montoya and Quevedo, 1990; Montoya et
al., 1988). An NTP (1989) study in mice showed decreased motor activity in rotarod
performance in male rats treated with tPCP for 5 weeks and increases in motor activity and
startle response in females receiving aPCP and tPCP for 26 weeks. Another in vivo study
showed that treatment of rats with 20 mg/L PCP for up to 14 weeks caused biochemical effects
in the rat brain (Savolainen and Pekari, 1979), although the authors considered these transient
effects. The most definitive study showed that rats receiving 3 mM PCP in drinking water for at
least 90 days had marked morphological changes in sciatic nerves (Villena et al., 1992). It is
possible that some of the neurotoxic effects are related to PCP contaminants. Most of the
neurotoxicity studies were performed using tPCP or PCP of unknown purity. NTP (1989)
utilized four grades (aPCP, tPCP, DP-2, and EC-7) of PCP, at doses ranging from 36 to
458 mg/kg-day, and found that the majority of the neurotoxic effects were observed in male mice
with tPCP; however, similar effects were also observed in female mice treated with all four
grades of PCP. Effects were observed at the lower doses (36-102 mg/kg-day) and exhibited
dose-related increases.
4.6.2. Inhalation
There are no human or animal data available to evaluate the consequences of long-term
inhalation exposure to PCP. Toxicokinetic studies show that PCP is efficiently absorbed from
the respiratory tract after single or repeated exposures and that a large portion of PCP is excreted
in the urine as the unmetabolized parent compound with little evidence of binding in the tissues
or plasma (Hoben et al., 1976a). In subchronic studies in rats and rabbits (Demidenko, 1969),
minor liver function, cholinesterase activity, and blood sugar effects were reported in animals
exposed to 2.97 mg/m3 (calculated as 0.3 mg/kg-day PCP by Kunde and Bohme [1978]), a dose
that is lower than the lowest NOAELs (1 mg/kg-day) observed in animals orally exposed to PCP.
Demidenko (1969) reported anemia, leukocytosis, eosinophilia, hyperglycemia, and dystrophic
processes in the liver of rats and rabbits exposed to 28.9 mg/m3 PCP. Ning et al. (1984) reported
129
-------
significant increases in organ weights (lung, liver, kidney, and adrenal glands), serum y-globulin,
and blood-glucose levels at 21.4 mg/m3.
4.6.3. Mode-of-Action Information
Liver necrosis, chronic inflammation, hepatocellular vacuolation, pigmentation, and
hepatic hypertrophy following chronic oral exposure to relatively low doses (1.5-30 mg/kg-day)
of PCP demonstrate that the liver is the target organ involved in PCP-induced toxicity. Liver
necrosis was observed in subchronic-duration (NTP, 1989; Kerkvliet et al., 1982b) and chronic-
duration studies in mice (NTP, 1989), in subchronic-duration studies in rats (Villena et al., 1992;
Johnson et al., 1973), and in a two-generation reproductive study in rats (Bernard et al., 2002).
Chronic exposure to PCP induced inflammation in the liver of mice (NTP, 1989), rats (Bernard
et al., 2002; NTP, 1999; Kimbrough and Linder, 1978; Schwetz et al., 1978), and dogs (Mecler,
1996), and in olfactory epithelium of rats (NTP, 1999). Additional evidence of lethal
hepatocellular damage was reported by the majority of the studies in the database.
Oxidation/reduction processes have repeatedly been shown to be involved in PCP
toxicity at doses of 60 mg/kg-day (NTP, 1999) and 25 pM (Dahlhaus et al., 1996, 1994).
Dahlhaus et al. (1994) also observed oxidative stress at 300 mg/kg TCHQ (metabolite of PCP)
after 2 or 4 weeks of exposure. Damaged lipid membranes and induction of apoptosis (Wang et
al., 2001) are some of the effects observed following exposure to 15 and 40 mg/kg PCP. The
uncoupling of oxidative phosphorylation has long been associated with exposures ranging from
0.25 pM to 1 mM PCP (Gravance et al., 2003; Wang et al., 2001; Trenti et al., 1986a, b; Masini
et al., 1985, 1984a, b; Varnbo et al., 1985). The earliest detectable intracellular indication of an
adverse redox shift is the appearance of lamellar aggregations of damaged lipid membranes (at
the electron microscopy level), followed by uncoupling of oxidative phosphorylation and
induction of apoptosis (Wang et al., 2001). PCP, as low as 0.1 mM, accelerated the breakdown
of mitochondrial ATP, a likely consequence of changed membrane permeability (Weinbach,
1954). PCP was noted as inhibiting the electron transport between flavin coenzyme and CYP450
(which may explain the limited metabolism associated with PCP). Thus, PCP was recognized as
capable of interacting with, and interfering with, multiple molecular intracellular target
molecules and cellular processes. The inhibition of oxidative phosphorylation, at 40 mg/kg, has
been suggested to precede heptatocellular necrosis (Arrhenius et al., 1977a). Increased cellular
phospholipoperoxides and greatly decreased glutathione have been observed following
incubation with 1 mM PCP (Suzuki et al., 2001). Antioxidant protective systems can become
overwhelmed in the presence of intracellular redox disruption. Depletion of glutathione
combined with the potential for oxidative damage suggests that PCP can induce nonneoplastic
effects in multiple animal species.
130
-------
4.6.4. Comparison of Toxic Effects of aPCP with tPCP or Commercial-Grade PCP
PCP is manufactured in a multistage chlorination process that results in contamination
with dioxins, furans, and other chlorophenols. Consequently, the formulation that is employed
and that people are exposed to is a chemical grade that has a purity of approximately 90%, and is
commonly referred to as the technical or commercial grade of PCP. Depending on the specific
synthesis process, the level of these impurities may vary with differing grades of manufactured
PCP. aPCP is only achieved after the impurities are removed. Therefore, the information
available on toxic effects from PCP alone is limited. There are studies in the database that have
examined the toxicity of aPCP, either alone or concurrently with the technical/commercial
grades (tPCP, EC-7, and/or DP-2). The toxicity database for PCP contains many studies that did
not characterize the type and/or level of the contaminants. The uncertainty surrounding the
presence of these contaminants confounds the characterization of PCP itself. However, a
comparison of toxicity studies conducted with the analytical grade (>99% purity) with studies
using commercial preparations is useful.
4.6.4.1. Short-term and Subchronic Studies
In a subchronic study, rats exhibited increased liver weight at doses of 10 and 30 mg/kg-
day and increased kidney weight at 30 mg/kg-day (Johnson et al., 1973, 90-day feed study) with
both aPCP and an "improved" grade (88-93% purity) of PCP. tPCP administration elicited
elevated liver and kidney weight at 3, 10, and 30 mg/kg-day. Additionally, at a dose level of
30 mg/kg-day tPCP, serum albumin and hepatic microscopic lesions (minimal focal
hepatocellular degeneration and necrosis) were elevated and erythrocyte count, hemoglobin
concentration, and hematocrit were reduced. For aPCP, Renner et al. (1987) reported decreased
erythrocyte parameters (RBC, hemoglobin, and hematocrit) throughout 4 weeks of treatment
(53 mg/kg-day) via gavage. Liver effects, including enlarged pleomorphic hepatocytes,
degeneration of liver cells, and acidophilic bodies in sinusoids, were observed in addition to the
hematological effects. The hepatic and hematological effects observed with 30 mg/kg-day tPCP
and not aPCP in Johnson et al. (1973) were seen with aPCP at a concentration of 53 mg/kg-day
in Renner et al. (1987). In an NTP (1999) study, hepatocyte degeneration increased in incidence
and severity at aPCP doses of 40 and 75 mg/kg-day in male and female rats, respectively.
Degeneration of germinal epithelium in testes in males and centrilobular hypertrophy in males
and females were observed at 270 mg/kg-day aPCP (highest dose) (NTP, 1999, 28-day study).
Kimbrough and Linder (1975) reported cytoplasmic inclusions and ultrastructural effects
(increased smooth endoplasmic reticulum, presence of lipid vacuoles, and atypical appearance of
mitochondria) at 1,000 ppm (approximately 87 mg/kg-day) of either tPCP or aPCP for 90 days.
In addition, tPCP-treated animals exhibited hepatic effects consisting of foamy cytoplasm,
pronounced vacuolation of hepatocytes, single hepatocellular necrosis, slight interstitial fibrosis,
and prominent brown pigment in macrophages and Kupffer cells in liver. In Kimbrough and
131
-------
Linder (1978), rats administered tPCP and aPCP for 8 months showed signs of liver toxicity at
500 ppm (approximately 46 mg/kg-day), including cytoplasmic hyaline inclusions,
hepatocellular hypertrophy, and abundant brown pigment in macrophages and Kupffer cells. As
in the 1975 study, additional liver effects were observed in those animals treated with tPCP
(periportal fibrosis, adenofibrosis, vacuolation, pleomorphism, and bile duct proliferation).
Hepatic effects were also observed at 10 mg/kg-day, although these effects were limited to
animals treated with tPCP.
NTP (1989) noted liver lesions consisting of centrilobular cytomegaly, karyomegaly,
nuclear atypia, and degeneration, and necrosis in male mice treated for 30 days with 500 ppm
(95 mg/kg-day for males and 126 mg/kg-day for females) tPCP, EC-7, and aPCP. Female mice
showed signs of liver toxicity with EC-7 and aPCP at doses of 645 and 25 mg/kg-day,
respectively. The report stated that hepatic lesions in animals treated with EC-7 and aPCP were
less diffuse and less severe than with tPCP. However, the incidences of the lesions were similar
for tPCP and aPCP for all doses. All grades of PCP exhibited increases in absolute and relative
liver weights, liver porphyrins, P450 levels, and serum enzymes (ALP, cholesterol, and ALT),
and a decrease in leukocyte count (males only).
In a 27-week study (NTP, 1989), mice treated with tPCP, EC-7, DP-2, and aPCP showed
results similar to the 30-day study. Hepatic cytomegaly, karyomegaly, degeneration, and
necrosis were observed in males and females at all doses (estimated average doses of 36-
458 mg/kg-day) and grades of PCP. While all four grades elicited effects at the high dose,
including liver pigmentation, liver inflammation, dark urine, and urine creatinine, only tPCP
showed signs of bile duct hyperplasia. Liver pigments were seen at the low and mid dose for
tPCP and at the mid dose for DP-2 and EC-7. aPCP-treated animals did not show signs of liver
pigmentation, inflammation, or urinary effects at doses other than the high dose. Similar
hepatotoxic effects were shown for aPCP and tPCP, including mild to marked hepatocyte
swelling, and increases in relative liver weight, nuclear swelling, vacuolization with eosinophilic
inclusions in nuclear vacuoles, and mild-to-moderate multifocal necrosis in the liver (Kerkvliet,
1982a, b).
tPCP was observed to have significantly higher levels of chlorinated dibenzo-p-dioxins
and dibenzofurans than either DP-2 or EC-7. Specifically, the concentration of
heptachlorodibenzo-p-dioxin was observed to be approximately 10 and 500 times higher for
tPCP than for DP-2 and EC-7, respectively. Higher concentrations were also observed for
OCDD and HxCDD. Thus, mice were exposed to higher levels of these contaminants from
tPCP-treated feed than from DP-2- or EC-7-treated feed (NTP, 1989). Despite this, there were
no differences in liver toxicity caused by tPCP and EC-7, suggesting that PCP, itself, causes liver
toxicity in the mice. Only tPCP resulted in significant increases in the incidences of lesions in
the spleen of male mice and mammary gland of female mice, suggesting that these lesions were
caused by impurities. Lesions in the nose were prominent in mice receiving EC-7 but not in
132
-------
mice receiving tPCP, suggesting that a specific EC-7 impurity (possibly TCP, which is present in
greater amounts in EC-7 compared with tPCP) caused these lesions.
Dose-dependent decreases in motor activity and rotarod performance were found in mice
treated with tPCP only. Immunosuppression in the form of inhibition of plaque-forming
response following immunization with SRBCs was seen at all doses of tPCP and at the highest
dose of DP-2 and was not observed with EC-7 or aPCP. NTP (1989) stated that the degree of
immunosuppression is consistent with exposure to dioxin and furan contamination. Studies in
Swiss Webster, C57BL/6J, and DBA/2J mice showed immunosuppressive effects in animals
treated with tPCP but not with aPCP (Kerkvliet 1985a, b; 1982a, b). In an experiment looking at
tPCP only, mice exhibited a significant increase in relative liver weight as well as effects on
humoral but not cellular immunity (Kerkvliet, 1985b). The remaining studies observed
differences in effects from treatment with aPCP and tPCP. Significant depression of
T-lymphocyte cytolytic activity and enhancement of macrophage phagocytosis (Kerkvliet,
1982b) as well as early immunosuppressive effects on humoral response (Kerkvliet, 1982a) were
observed with tPCP treatment and no effects were seen with aPCP, even at doses fourfold greater
than tPCP doses. Additionally, contaminant fractions from tPCP, at equivalent doses to tPCP,
were examined for immunotoxic effects. The chlorinated dioxin/furan fraction had a significant
immunosuppressive effect, whereas the chlorinated phenoxyphenol and the chlorinated diphenyl
ether fractions were ineffective in affecting the immune response (Kerkvliet, 1985a). These
studies show that the chlorinated dioxin and furan contaminants present in tPCP and not PCP are
likely responsible for the immunotoxic effects observed in mice. However, Exon and Roller
(1983) reported a significant depression in immune response (humoral and cell-mediated
immunity) in offspring of male and female Sprague-Dawley rats administered 4 or 43 mg/kg-day
and 5 or 49 mg/kg-day aPCP, respectively, continuously in the diet from weaning until 3 weeks
after parturition. Offspring were treated similarly to the parents and treatment continued until
13 weeks of age. Macrophage function measured by the rats' ability to phagocytize SRBCs
increased in a dose-related manner that was statistically significant at 4 and 43 mg/kg-day for
males and 5 and 49 mg/kg-day for females. In addition, there was an increase in the number of
macrophages harvested from the peritoneal exudate.
In cattle, aPCP caused significant decreases in serum Tj, and T4 levels at 10 (Hughes et
al., 1985) and 15 mg/kg-day (McConnell et al., 1980). However, tPCP-treated animals also
exhibited microscopic lesions consistent with thymus atrophy, squamous metaplasia in the
Meibomian gland of the eyelid (Hughes et al., 1985; McConnell et al., 1980), and smaller and
more numerous thyroid-follicles (McConnell et al., 1980). McConnell et al. (1980) attributed the
dose-related effects that were observed with tPCP and not aPCP to the dioxin and furan
contaminants in tPCP. Jekat et al. (1994) reported decreases in total and free serum T4,
T4:T3 ratio in serum, and serum TSH in female Wistar rats administered 3 mg/kg-day aPCP or
tPCP by gavage for 28 days. In a two-generation study in mink exposed to 1 mg/kg-day PCP,
133
-------
Beard and Rawlings (1998) reported statistically significant decreases in serum T4 secretion in
the Fl (21%) and F2 (18%) males and F2 females (17%). Thyroid mass was decreased in both
Fl and F2 generation animals, although the reduction was statistically significant only in
F2 females (27%). Rawlings et al. (1998) administered 2 mg/kg aPCP to mature ewes for
approximately 6 weeks. A marked decrease in serum T4 levels was observed in mature ewes at
36 days. In addition to statistically significant decreased serum T4 levels, aPCP-treated ewes had
significantly increased serum insulin levels. However, no treatment-related changes were
observed in cortisol, LH, FSH, estradiol, or progesterone levels. Beard et al. (1999a) noted that
maximum serum T4 levels in ewes treated with 1 mg/kg-day PCP were statistically significantly
lower (approximately 25%) than controls with or without prior administration of TSH.
4.6.4.2. Chronic Studies
Within the PCP database, only one study examined the effects of chronic exposure to
aPCP. NTP (1999) reported significantly increased cystic degeneration of hepatocytes in male
rats at 20 and 30 mg/kg-day in a 2-year bioassay. However, in an additional stop-exposure
portion of this study, rats administered 60 mg/kg-day for 1 year exhibited significantly elevated
serum ALP and cytoplasmic hepatocyte vacuolization in males, increased sorbitol
dehydrogenase, and incidences of centrilobular hypertrophy in both males and females. ALT
levels were elevated in male rats, although this increase was not considered statistically
significant. In another chronic study in rats, Schwetz et al. (1978) reported slightly increased
(<1.7-fold) serum ALT activity in both sexes at 30 mg/kg-day EC-7.
Additionally, rats treated with 60 mg/kg-day aPCP (NTP, 1999) exhibited liver lesions
including chronic inflammation, basophilic focus, and cystic degeneration of hepatocytes. Renal
tubule pigmentation was observed in all rats of this study at doses ranging from 10 to 60 mg/kg-
day (2-year bioassay and 1-year stop-exposure). Analyses of the pigment were inconclusive as a
result of contrasting staining results. Histopathological examination in Schwetz et al. (1978)
showed pigment accumulation in the centrilobular hepatocytes of the liver in 30% of females
given 10 mg/kg-day and in 59% of females given 30 mg/kg-day. Similarly, 26 and 70% of
females receiving 10 and 30 mg/kg-day EC-7 exhibited pigment accumulation in the epithelial
cells of the proximal convoluted tubules in the kidney. This effect was not detected in the lower
dose or control groups of the female rats. Only 1 of the 27 male rats given EC-7 (30 mg/kg-day)
exhibited the brown pigment in hepatocytes. NTP (1989) reported hepatotoxic effects in mice at
doses as low as 17 mg/kg-day EC-7 or tPCP that are similar to those reported in rats ranging
from 10 to 60 mg/kg-day reported by NTP (1999) for aPCP and Schwetz et al. (1978) for EC-7.
134
-------
4.6.4.3. Developmental Studies
Schwetz et al. (1974a) examined the maternal and fetal effects of rats administered tPCP
or aPCP on GDs 6-15. Similar effects were observed for both grades of PCP, including
significant decreases in maternal and fetal weight gain at 30 and 50 mg/kg-day. A statistically
significant increased incidence of resorptions was noted at 15 mg/kg-day for tPCP and 30 mg/kg-
day for aPCP. While tPCP did not seem to affect fetal crown-rump length, aPCP-treated rats
exhibited significantly decreased crown-rump length at 30 mg/kg-day. Soft-tissue and skeletal
anomalies were induced with doses >15 mg/kg-day tPCP and >5 mg/kg-day aPCP. In a timing
evaluation of PCP administration, significant decreases in fetal body weight and crown-rump
length and increased incidence of subcutaneous edema and rib, vertebral, and sternebral
anomalies were observed following administration of 30 mg/kg-day PCP on GDs 8-11 for tPCP
and aPCP and on GDs 12-15 for aPCP only. The authors stated that aPCP exhibited greater
toxicity than tPCP, especially in the latter stage of gestation. The effects observed in the
developing rat embryo and fetus were attributed to PCP and not the contaminants (Schwetz et al.,
1974a).
Developmental toxicity was noted at a dose level of 60 mg/kg-day in the Larsen et al.
(1975) study in which rats exposed to aPCP during gestation had fetuses with reduced body
weight and increased malformations. The authors concluded that the maternal toxicity resulted
in the observed fetal effects. This was based on other study findings indicating limited transfer
of PCP through the placental barrier. However, Larsen et al. (1975) did not report the maternal
toxicity data. Welsh et al. (1987) also observed fetal effects following administration of aPCP at
doses of 13 and 43 mg/kg-day. Significantly decreased body weight and crown-rump length and
increased skeletal variation (misshaped centra) were observed in fetuses at 13 and 43 mg/kg-day.
The dams exhibited signs of toxicity, such as decreased mean weight gain (GDs 7-20) and
decreased number of viable fetuses, because of significant resorption at the 43 mg/kg-day dose
level.
Summary of comparison of toxic effects ofaPCP with technical/commercial PCP.
Repeated dose toxicity studies with tPCP, EC-7, DP-2, and/or aPCP formulations all show the
liver to be a major target. Many of the studies comparing tPCP and aPCP showed similar toxic
effects following exposure to each formulation. Studies that compared toxicity of purified PCP
and tPCP show a broader spectrum of liver toxicity occurring at similar or slightly lower doses
with tPCP than aPCP (NTP, 1989; Hughes et al., 1985; McConnell et al., 1980; Kimbrough and
Linder, 1978; Johnson et al., 1973). Therefore, the EPA determined that studies using technical
or commercial grades of PCP are representative of PCP itself, and that an RfD based on these
studies should also apply to pure PCP.
135
-------
4.7. EVALUATION OF CARCINOGENICITY
4.7.1. Summary of Overall Weight of Evidence
Under the EPA's Guidelines for Carcinogen Risk Assessment (2005a), PCP is "likely to be
carcinogenic to humans." This cancer weight of evidence determination is based on:
(1) evidence of carcinogenicity from oral studies in male mice exhibiting hepatocellular adenomas
and carcinomas, and pheochromocytomas and malignant pheochromocytomas, and in female
mice exhibiting hepatocellular adenomas and carcinomas, pheochromocytomas and malignant
pheochromocytomas, and hemangiomas and hemangiosarcomas (NTP, 1989), (2) some evidence
of carcinogenicity from oral studies in male rats exhibiting malignant mesotheliomas and nasal
squamous cell carcinomas (Chhabra et al., 1999; NTP, 1999), (3) strong evidence from human
epidemiologic studies showing increased risks of non-Hodgkin's lymphoma and multiple
myeloma, some evidence of soft tissue sarcoma, and limited evidence of liver cancer associated
with PCP exposure (Demers et al., 2006; Hardell et al., 1995, 1994; Kogevinas et al., 1995), and
(4) positive evidence of hepatocellular tumor-promoting activity (Umemura et al., 2003a, b, 1999)
and lymphoma and skin-adenoma promoting activity in mice (Chang et al., 2003).
The U.S. EPA's Guidelines for Carcinogen Risk Assessment (2005a) indicate that for
tumors occurring at a site other than the initial point of contact, the cancer descriptor may apply
to all routes of exposure that have not been adequately tested at sufficient doses. An exception
occurs when there is convincing toxicokinetic data that absorption does not occur by other
routes. Oral studies of PCP carcinogenicity demonstrate that tumors occur in tissues remote
from the site of absorption, including the liver, adrenal gland, circulatory system, and nose.
Information on the carcinogenicity of PCP via the inhalation and dermal routes is unavailable.
Studies of the absorption of PCP indicate that the chemical is readily absorbed via all routes of
exposure, including oral, inhalation, and dermal. Therefore, based on the observance of systemic
tumors following oral exposure, and in the absence of information to indicate otherwise, it is
assumed that an internal dose will be achieved regardless of the route of exposure. Accordingly,
PCP is considered "likely to be carcinogenic to humans" by all routes of exposure.
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence
4.7.2.1. Human Epidemiologic Evidence
Epidemiological studies of various designs (cohort, population-based case-control, and
nested case-control within occupationally exposed workers) have examined the relationship
between occupational PCP exposure and cancer risk. The most comprehensive of the cohort
studies, in terms of design, is the sawmill cohort study conducted in British Columbia, Canada,
updated by Demers et al. (2006). In addition to the sample size, the design of this study
including the exposure assessment procedure; use of an internal referent group; analysis of PCP
and TCP exposures; low loss to follow-up; and use of a population-based cancer registry add to
the strengths of this study. Even with this size, however, there is limited statistical power to
136
-------
estimate precise associations with relatively rare cancers. The case-control studies of non-
Hodgkin's lymphoma and soft tissue sarcoma (Hardell and Eriksson, 1999; Kogevinas et al.,
1995; Hardell et al., 1995, 1994) specifically address this limitation by focusing on these
outcomes. Kogevinas et al. (1995) has the additional attribute of providing estimates for the
effects of other phenoxy herbicides or chlorophenols, which provides information regarding the
issue of co-exposures.
In these studies, moderately high associations (i.e., a two- to fourfold increased risk) were
generally seen between occupational exposure to PCP and non-Hodgkin's lymphoma (Demers et
al., 2006; Kogevinas et al., 1995; Hardell et al., 1994), multiple myeloma (Demers et al., 2006),
or soft tissue sarcoma (four studies summarized in a meta-analysis by Hardell et al., 1994).
However, there are some inconsistencies, most notably for soft tissue sarcoma. The relative
rarity of this cancer (e.g., only 12 cases were found in the nested case-control study of
13,898 workers exposed to phenoxy herbicides or chlorophenols by Kogevinas et al. [1995]), and
the difficulty in classifying the disease, even with a review of the histology, may be reasons for
this inconsistency. In contrast to the studies from the 1970s and 1980s, the most recent case-
control study of non-Hodgkin's lymphoma, conducted in cases diagnosed 9-13 years after PCP
had been banned from use in Sweden, did not observe an association (OR 1.2) with PCP
exposure (Hardell and Eriksson, 1999). The lack of association in this study could reflect a
relatively short latency period between exposure and disease, as has been seen with other
lymphoma-inducing agents (e.g., Krishnan and Morgan, 2007).
Demers et al. (2006) developed a cumulative dermal chlorophenol exposure score based
on a retrospective exposure assessment that was validated for current exposures in comparison
with urinary measurements and with industrial hygienist assessments. This detailed exposure
measure allowed for analysis of an exposure-response gradient, with evidence of a trend of
increasing mortality or incidence risk seen for non-Hodgkin's lymphoma and multiple myeloma.
The other studies with a relatively detailed exposure assessment (Hardell et al., 1995, 1994;
Kogevinas et al., 1995) also demonstrated stronger associations with the more refined (e.g.,
higher exposure probability or frequency) measures of exposure compared with the associations
seen with "any pentachlorophenols."
The possibility of the carcinogenic effects of PCP resulting solely from the presence of
contaminants of dioxins and furans was examined in this assessment (Demers et al., 2006). The
primary contaminants were hexa-, hepta-, and octa-chlorinated dibenzodioxins, and higher-
chlorinated dibenzofurans. There are several reasons, as noted in Section 4.1.1.4 (General
Issues—Interpretation of the Epidemiologic Studies) that this contamination is an unlikely
explanation for the observed effects. Specific furans are not generally seen at higher levels in
blood from PCP workers compared with the general population (Collins et al., 2007). The
cancer risks seen in the large cohorts of workers exposed to dioxins (consistent observations of
an exposure-response gradient with total cancer risk) (NAS, 2006; Steenland et al., 2004) differ
137
-------
from the observations seen in studies of PCP exposure. In addition, the associations seen with
specific cancers (e.g., non-Hodgkin's lymphoma) and PCP are generally stronger than the
associations seen between these cancers and dioxin or other chlorophenol exposures in studies
with both of these measures (Demers et al., 2006; Kogenivas et al., 1995).
An increased risk of liver cancer associated with exposure to PCP was seen in the large
cohort study of sawmill workers in British Columbia (Demers et al., 2006), and as noted in the
previous discussion of non-Hodgkin's lymphoma, an attenuation in the highest exposure group
was observed. This study identified strong associations between exposure to PCP and liver
cancer, with at least a doubling of the risk in almost all of the exposure categories.
Evidence for PCP-induced DNA damage has been presented in numerous animal or in
vitro studies and was equivocal in studies of PCP-exposed workers (Ziemsen et al., 1987;
Bauchinger et al., 1982; Schmid et al., 1982). Evidence for cytotoxicity or apoptosis, reparative
cell proliferation, and gap junction inhibition usually cannot be obtained in human studies.
PCP-induced effects on the immune system have been found in humans and animals.
Blakley et al. (1998) reported stimulation of mitogen effects in low-dose, gavage-treated male
rats. Daniel et al. (1995) observed exposure-dependent impairment of mitogen response in
lymphocytes of PCP-exposed humans, and McConnachie and Zahalsky (1991) reported
heightened immune response in PCP-exposed humans. Finally, symptoms of porphyria were
identified in PCP-exposed humans (Cheng et al., 1993) and animals (NTP, 1989; Kimbrough and
Linder, 1978). These findings make a strong point for the plausibility of PCP-related
carcinogenesis in humans. In summary, the weight of evidence for the carcinogenic action of
PCP (U.S. EPA, 2005a) suggests that this compound by itself (i.e., in the absence of
contaminants) is likely to be a human carcinogen.
4.7.2.2. Animal Cancer Evidence from Oral Exposure
Long-term animal studies employing the oral route of exposure are available that assess
the carcinogenicity of PCP in animals. An NTP feeding study in B6C3Fi mice demonstrated that
tPCP (17-18 or 35-36 mg/kg-day) and EC-7 (17-18, 35-36, or 117-118 mg/kg-day) caused
statistically significant increases in the incidence of hepatocellular adenomas/carcinomas and
adrenal gland pheochromocytomas in males and females, and an increased incidence of
hemangioma/hemangiosarcoma in female mice (NTP, 1989). tPCP was slightly more effective
than EC-7, suggesting that chlorinated dibenzo-p-dioxin and dibenzofuran impurities in tPCP
may have only exacerbated the carcinogenic effect of PCP in mice.
Another NTP (1999) feeding study conducted in F344/N rats provided some evidence of
carcinogenic activity, demonstrated by increased incidence of mesotheliomas and nasal
squamous cell carcinomas in males exposed to aPCP (10-60 mg/kg-day). NTP (1999)
concluded that there was no evidence of carcinogenic activity for female rats fed aPCP.
138
-------
Umemura et al. (1999) examined the initiating and promoting activity of aPCP (98.6%
purity) administered in the diet to 20 male B6C3Fi mice/group. DEN was given as the initiator
when the promoting activity of aPCP was assessed, and PB was administered as the promoter
when the initiating activity of aPCP was assessed. The incidence of liver tumors was statistically
significantly higher in mice initiated with DEN and promoted with PCP than in control mice
receiving DEN only. Tumor multiplicity was statistically significantly increased in mice
promoted with aPCP and PB compared with DEN controls. No liver tumors developed in mice
initiated with aPCP with or without subsequent promotion with PB. In this study, aPCP showed
promoting, but not initiating, activity in mice that were initiated with DEN. Umemura et al.
(1999) concluded that aPCP exerts a promoting effect on liver carcinogenesis.
A study by BRL (1968) showed no carcinogenic response in male and female B6C3Fi
and B6AKFi mice administered EC-7 at a dose of 46.4 mg/kg-day for up to 18 months. This
exposure may not have been long enough to reveal carcinogenic effects. BRL (1968) also
reported that mice administered 46.4 mg/kg-day EC-7 as a single, subcutaneous injection did not
develop tumors that were considered statistically significantly greater than tumors observed in
control animals. Schwetz et al. (1978) reported no carcinogenic response in male and female
Sprague-Dawley rats administered EC-7 in the diet at doses up to 30 mg/kg-day for 22-
24 months. A lack of body or organ weight changes even at the highest dose raise the possibility
that an MTD was not reached in this study.
Potential toxicity of contaminants. The potential carcinogen! city of the contaminants
associated with PCP was considered when assessing the carcinogenicity associated with
exposure to PCP. NTP (1989) listed an estimate of the total contaminant exposure associated
with tPCP and EC-7 in the mouse 2-year bioassay. Most importantly, the most potent
carcinogenic promoter ever studied (Pitot et al., 1980), TCDD, has not been detected in the PCP
preparations. Contaminant levels increased with the degree of chlorination; the highest levels
were detected for OCDD (400 and 800 |ig from tPCP, or 0.2, 0.4, and 1.2 |ig from EC-7). Total
exposure to pentachlorodibenzofuran was estimated at approximately 0.01-0.03 jig/kg-day for
tPCP at the 17-18 and 35-36 mg/kg-day doses over the full 2-year period. This compound was
not detected in EC-7. Additional contaminants identified at comparatively high levels in tPCP
were octachlorohydroxydiphenyl ether (0.2-0.4 mg/kg-day), nonachlorohydroxydiphenyl ether
(0.4-0.8 mg/kg-day), hexachlorohydroxydibenzofuran (0.02-0.04 mg/kg-day), and
heptachlorohydroxydibenzofuran (0.05-0.1 mg/kg-day). These ether contaminants were not
detected in EC-7. A complete list of the contaminants can be found in Table 2-2 and estimated
daily doses can be found in Table B-3.
NTP (1989) and McConnell et al. (1991) compared the concentrations of HxCDD in
tPCP and EC-7 with that known to induce liver tumors in mice and concluded that the
carcinogenic response in mice can be attributed primarily to PCP. Hepta- and
octachlorodibenzo-p-dioxins and dibenzofurans, because of their very poor bioavailability and
139
-------
metabolism, have comparatively low toxicity. Toxicity data for the higher chlorinated
hydroxydibenzofurans or hydroxydiphenyl ethers are not available.
The major contaminant measured in both formulations of PCP utilized by NTP (1989)
was TCP, present at levels yielding doses of 0.4-0.9 mg/kg-day in tPCP at the 17-36 mg/kg-day
doses and 1.0-6.0 mg/kg-day in EC-7 at the 17-118 mg/kg-day doses, respectively. In the
absence of a slope factor for any of the TCP congeners, the possible contribution of this
contaminant to the carcinogen!city of tPCP or EC-7 cannot be determined. However,
considering the difference in the amount of TCP that was found in tPCP versus EC-7 compared
to the similar tumor responses observed for the two formulations, a reasonable assumption would
be that, at the given doses, the contribution of TCP to the carcinogenicity of tPCP or EC-7 is
likely to be minimal.
4.7.2.3. Animal Cancer Evidence from Inhalation Exposure
There are no known chronic duration inhalation exposure studies in humans or laboratory
animals. Limited evidence concerning the potential effects induced by PCP inhalation is based
on evidence of respiratory tract effects in three animal studies. In the NTP (1999) stop-exposure
oral study of F344/N rats showing nasal squamous cell carcinomas in males, Chhabra et al.
(1999) suggested that the cancers were chemical related, either via systemic exposure, via direct
nasal contact with PCP vapors during feeding, or via PCP-containing feed dust. In an earlier
NTP (1989) study, increased incidences of acute focal inflammation of the nasal mucosa (males:
4/35, 1/13, 3/16, 47/49; females: 0/35, 0/14, 2/5, 46/48) and focal metaplasia of the olfactory
epithelium (males: 2/35, 1/13, 2/16, 46/49; females: 1/35, 0/14, 2/5, 45/48) were observed in
mice that received EC-7 in feed (at doses of 0, 17-18, 34-37, and 114-118 mg/kg-day,
respectively) but not in mice exposed to tPCP (NTP, 1989).
NTP (1989) conducted a 6-month range-finding study in B6C3Fi mice fed four different
preparations of PCP (tPCP, DP-2, EC-7, and aPCP). Increased incidences of nasal mucosal
metaplasia/goblet cell hyperplasia were seen in female mice that received doses of 54 or
51 mg/kg-day EC-7 or aPCP, respectively, or 323 mg/kg-day DP-2 and in male mice that
received doses of 124 mg/kg-day EC-7 or 102 mg/kg-day aPCP. Mice, both male and female,
administered tPCP (38-301 mg/kg-day) did not show any of the nasal effects. Females were
more sensitive to the nasal effects than male mice.
Tisch et al. (2005) obtained evidence for single and double strand breaks in ex vivo
cultures of human mucosal cells of the inferior and middle nasal conchae treated with 0.3, 0.75,
and 1.2 mmol/mL aPCP. According to the authors of the study, as much as 1.5 mmol PCP has
been measured in nasal mucosa in the presence of dust contaminated with PCP in occupational
inhalation studies. These results indicate that humans may be exposed to concentrations of PCP
that have induced DNA damage in human mucosal cells, although Tisch et al. (2005) observed
the damage in cells that lacked a protective mucosal barrier normally present in humans in vivo.
140
-------
While many of the human epidemiological studies (Kogevinas et al., 1992; Saracci et al., 1991;
Brinton et al., 1977) suggest an inhalation cancer risk the lack of useable exposure levels,
possible presence of contaminants and other study limitations prevent clear associations between
PCP exposure and cancer in these reports.
4.7.3. Mode-of-Action Information
PCP can interact directly via parent compound or indirectly via metabolites with cellular
biomolecules, including lipids, proteins, and nucleotides. PCP has not shown strong mutagenic
activity in standard genotoxicity tests such as the Ames assay (Seller, 1991). Positive results
have been observed for PCP in tests that respond to molecular action other than direct mutation,
such as SCE induction; however, PCP-induced SCEs could not be confirmed in exposed humans
(Ziemsen et al., 1987; Bauchinger et al., 1982; Schmid et al., 1982). SSBs and CAs were
observed in animals and exposed humans in assays using PCP or TCHQ. The metabolites of
PCP, specifically TCHQ, TCoHQ, TCpBQ, and TCpCAT, have shown some evidence of SSBs
in in vitro assays. TCHQ was positive for forward mutations in V79 Chinese hamster cells at the
HPRT locus (Jansson and Jansson, 1991). Carstens et al. (1990) suggested that superoxide
formation with TCHQ and reduction of H2O2 by TCSQ (in the Fenton reaction) may result in
cellular toxicity and genotoxicity. However, PCP is rather poorly metabolized in animals (see
Section 3.1) and to what extent the metabolites are formed is unknown. Without more
information on the formation of the metabolites, it is difficult to determine the influence that the
parent compound or the metabolites have on mutagenic activity.
While standard mutagenicity assays have produced weak or equivocal evidence for PCP,
there is some in vitro and in vivo evidence for the ability of PCP to cause oxidative DNA
damage. Several studies presented evidence that long-term administration of PCP results in
measurable 8-OH-dG formation in hepatic nuclear DNA of mice (Umemura et al., 1996; Sai-
Kato et al., 1995) and rats (Lin et al., 2002). Naito et al. (1994) demonstrated that PCP induced
DNA damage via 8-OH-dG formation through its metabolite, TCHQ, in calf thymus DNA in
vitro. Dahlhaus et al. (1994) showed that TCHQ elicited increased 8-OH-dG formation in
hepatic DNA of B6C3Fi mice fed this PCP metabolite for 2 or 4 weeks, while single i.p.
injections had no such effect. Dahlhaus et al. (1996, 1995) found that TCHQ, TCpBQ, and
TCoBQ produced 8-OH-dG, while TCoHQ and PCP did not. Formation of 8-OH-dG was
specific for the liver, the target organ. Significant decreases in the levels of glutathione, a
protective antioxidant, were observed following exposure to PCP (Suzuki et al., 2001, 1997;
Savolainen and Pekari, 1979) and TCHQ (Wang et al., 1997).
In addition to oxidative stress-induced DNA damage, the formation of DNA adducts by
metabolites of PCP has been observed in both in vitro and in vivo studies. TCpBQ was
frequently identified as the major metabolite responsible for the formation of the DNA and
protein adducts associated with PCP exposure. Studies have shown that dechlorination of PCP
141
-------
to the 1,4-chlorinated benzoquinone resulted in increases of DNA adducts in vitro at 100 uM
(Dai et al., 2005, 2003) and at 1 or 5 mM (Lin et al., 2001) and in vivo (Lin et al., 2002; Bodell
and Pathak, 1998). Rats exhibited DNA adducts following administration of PCP, TCHQ, and
TCpBQ. Typically, PCP and TCHQ are oxidized to facilitate the formation of the benzoquinone
radical, which is believed to be the reactive intermediate in the adduct formation (Lin et al.,
2002). Additionally, protein adducts in albumin and hemoglobin were observed in rats exposed
to TCpBQ, TCpSQ, and TCoSQ, but not TCoBQ (Waidyanatha et al., 1996), providing further
evidence of oxidative stress induced DNA damage. Oxidative stress-induced DNA damage that
occurs in concert with the formation of chemical-specific DNA adducts may enhance the
genotoxic effects of PCP.
Lin et al. (1999) suggested that species differences in the metabolism of PCP to
semiquinone and quinone metabolites may be responsible for the observed species differences in
liver carcinogenicity (i.e., PCP induced liver tumors in mice but not rats). At low PCP doses
(<4-10 mg/kg), TCoSQ-protein adduct formation in liver cytosol and nuclei was higher in rats
than in mice. At high PCP doses (>60-230 mg/kg), however, TCpBQ adducts were higher in
mice than in rats. Moreover, there was a fourfold difference in the nuclear total of quinone
metabolites in the mouse compared with that in the rat (Lin et al., 1997). Lin et al. (1999)
speculated that such differences in the metabolism of PCP to semiquinones and quinones might
be responsible for the production of liver tumors in mice but not rats. This is supported by the
results in Dahlhaus et al. (1996, 1995) in which TCHQ and TCpBQ, but not TCoHQ, induced the
formation of 8-OH-dG.
Various isozymes of P450 are responsible for metabolism of PCP and these may differ
between the two rodent species. Specific enzyme induction in mice (eightfold increase versus
control) versus the rat (2.4-fold increase versus control) may also be involved in the different
tumor patterns for these animals (Mehmood et al., 1996; Van Ommen et al., 1986a). PCP-DNA
adducts have been found at much higher amounts in mouse liver (Bodell and Pathak, 1998),
possibly a consequence of higher amounts of PCP quinone metabolites found in mouse liver as
compared with rat liver (Lin et al., 1997). Evidence of varied oxidative stress-generated
quinone-DNA adducts in rats and mice administered PCP (La et al., 1998b) combined with the
production of superoxide anion radical by mice, more so than other species (Parke and loannides,
1990), suggests species differences in the PCP-induced effects. These differences may explain
the distinctive tumor patterns in mice and rats. Additionally, the findings concerning species
differences in liver carcinogenicity of PCP were corroborated in other studies in which PCP
induced hepatocellular karyomegaly, cytomegaly, and degeneration in mice but only mild
hepatotoxicity in exposed rats (NTP, 1989; Kimbrough and Linder, 1978).
A number of studies have shown that PCP causes not only oxidative DNA damage, but
also oxidative damage to other subcellular systems, specifically cellular membranes (Suzuki et
al., 1997; Wang et al., 1997; NTP, 1989). It is well known that these events disrupt electron
142
-------
transport and metabolic energy synthesis (Freire et al., 2005; Masini et al., 1985; Arrhenius et al.,
1977b; Weinbach, 1954), thereby contributing to cell death. Suzuki et al. (1997) reported a
fivefold increase in cellular phospholipid hydroperoxide levels that were induced by PCP, while
cellular glutathione was virtually eliminated by PCP treatment. The latter effect is a potentially
critical event for PCP, allowing for oxidative stress to damage membranes, proteins, and
nucleotides. Wang et al. (1997) reported depletion of glutathione by TCHQ. These results
suggest that oxidative damage to cellular membrane phospholipids may be responsible for the
cytotoxicity induced by PCP.
Several responses to PCP exposure—including necrosis and chronic inflammation
leading to reparative cell proliferation/regeneration, and interference with GJIC—are consistent
with a promoting effect of PCP. Liver cell necrosis, the prerequisite for reparative cell
proliferation, has been observed in many experimental settings involving PCP exposure. Liver
necrosis was observed in subchronic (NTP, 1989; Kerkvliet et al., 1982b) and chronic (NTP,
1989) duration studies in mice, in subchronic (Villena et al., 1992; Kimbrough and Linder, 1975;
Johnson et al., 1973) duration studies in rats, and in two-generation reproductive studies in rats
(Bernard et al., 2002). Many studies have shown that PCP causes liver necrosis in experimental
animals, but no systematic studies to elucidate whether necrosis is followed by DNA resynthesis
have been conducted.
Chronic inflammation is another stimulus that can lead to cell regeneration. Several
studies have shown chronic inflammation to occur in liver, olfactory epithelium, and skin of
PCP-exposed laboratory animals, but, again, no studies were identified that demonstrate for PCP
that this event was a precursor of cell proliferation. However, Umemura et al. (1996)
demonstrated that 2-4 weeks of PCP administration to mice resulted in increased DNA content
and BrdU labeling of liver cells. Dose- and time-dependent elevation of 8-OH-dG combined
with an increase of DNA in the liver, indicating hyperproliferation, suggests that oxidative DNA
damage following PCP administration may lead to cellular proliferation that, if sustained, could
lead to tumorigenesis in the livers of mice.
Sai et al. (2001, 2000, 1998) demonstrated that aPCP, via decreased levels of the p53
tumor suppressor, inhibited GJIC. Gap junctions form between cells with the help of specialized
proteins, connexins. These junctions allow many molecules to pass from one cell to another,
enabling one cell to supply the other with metabolites required for survival, or, in the case of
apoptosis, to transfer what has been called the death signal, triggering programmed death in cells
that are attacked or damaged by certain toxicants. If a chemical prevents gap junctions from
forming, then programmed cell death may not occur in a transformed cell that will eventually
undergo clonal expansion and develop into a tumor. Many tumor promoters, such as the phorbol
esters or PB, have been shown to inhibit GJIC, while other substances that inhibit tumor
development, such as corticosteroids or retinoids, have been shown to strengthen GJIC.
Specifically, Sai et al. (2001) found that PCP inhibited apoptosis and that this coincided with a
143
-------
60% drop in the cellular level of p53. The 8-OH-dG moiety in DNA can lead to base-pair
exchanges that result in p53 gene mutations. PCP- or PCP metabolite-induced DNA damage,
inhibition of GJIC, and increased cellular proliferation have all been shown to be reduced by
antioxidants. Considering that PCP can reduce glutathione levels, the results reported by Sai et
al. (2001, 2000, 1998) provide support for another mechanism by which PCP potentially
promotes DNA damage.
A promoting effect of PCP has also been demonstrated in in vivo studies. In a study
designed to look at initiation and promotion activity, Umemura et al. (1999) found that PCP
exerted a promoting, but not initiating, effect on mouse liver carcinogenesis. Chang et al. (2003)
found that PCP or TCHQ applied repeatedly to mouse skin promoted skin tumor development.
Conclusions about the hypothesized mode of action. PCP induces tumors in rodents and
there is some evidence of carcinogenicity in humans; however, available experimental
information does not support the identification of key events in the mode of action of PCP
carcinogenicity. The potential for PCP to induce oxidative DNA damage is mostly supported by
a few animal and in vitro studies. The available evidence suggests that PCP's para- and possibly
orthohydroquinone and benzoquinone metabolites are the principal biologically reactive
intermediates. These intermediates can form direct DNA adducts; however, because there is
weak evidence for PCP-induced direct mutations in traditional tests, the intermediates are likely
unstable. The hydroquinone/benzoquinone metabolites undergo redox cycling resulting in the
formation of ROS and 8-OH-dG that, in turn, can result in chromosomal damage. SCEs, CAs,
and SSBs have been demonstrated in animals in vivo and in cell culture, but similar evidence in
PCP-exposed humans has been less than conclusive. The influence of oxidative stress on the
DNA-damaging action by PCP is supported by reduction of these effects with the application of
ROS scavengers and other antioxidants (Lin et al., 2001; Jansson and Jansson, 1992).
The available data suggest that PCP enters the cell and interacts with multiple targets,
with oxidative stress involved in both metabolism and proliferative signals. Damaged DNA can
lead to apoptosis, necrosis, inappropriate replication, CAs, SCEs, gene mutations, and DNA
strand breaks. It is possible that tumors could arise from cells that progressed through mitosis
with damaged DNA and failed cell cycle arrest.
Indicators of oxidative stress that were observed in studies with PCP have also been
identified in human cancers. The presence of 8-OH-dG and ROS (via oxidative phosphorylation,
P450 metabolism, redox cycling, etc.) as well as the formation of DNA adducts have been noted
in human carcinogenesis (Klaunig et al., 1998). Other mechanisms such as decreased GJIC have
been measured in the cancer process and observed in human carcinogenesis (Trosko and Ruch,
1998; Krutovskikh and Yamasaki, 1997). Oxidative stress is believed to play a role in human
carcinogenicity (Loft and M011er, 2006; Klaunig and Kamendulis, 2004; Klaunig et al., 1998;
Trush and Kensler, 1991), although the mechanisms involved and the extent to which oxidative
stress contributes are not fully understood. The available evidence in animals suggests that the
144
-------
metabolites TCHQ and TCBQ, as well as ROS formed in the course of redox cycling of these
metabolites, are involved in PCP-induced carcinogenicity in mammalian cells. However,
information on the metabolism of PCP to the quinone metabolites is limited and the level of
metabolite(s) associated with a dose of PCP cannot be quantified. It is plausible that long-term
exposure to PCP may induce gradual accumulation of oxidative DNA damage in the liver by
overwhelming the repair potential and this cumulative oxidative DNA damage could cause
critical mutations leading to carcinogenesis; however, the key events are unknown. While data
are limited and the mode of action by which PCP exerts its carcinogenic effect cannot be
characterized, the available evidence in both animals and humans suggests that induction of both
indirect and direct DNA damage and subsequent carcinogenicity via oxidative stress is possible.
The available data indicate that multiple modes of action for carcinogenicity are possible, but
none have been defined sufficiently (e.g., key events for carcinogenicity, temporal relationships)
to inform the human relevance or low-dose extrapolation for the estimate of the carcinogenicity
of PCP.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
4.8.1.1. Evidence in Humans
There are a number of cases from poison control centers, as outlined in Section 4.1,
where children have been exposed to PCP. In the cases involving small children, no serious
outcomes were reported, and in the cases with older children, only one case required critical care.
However, an incident where newborns in a nursery were accidentally exposed to PCP via their
diapers resulted in severe illness with two fatalities. Blood and tissue measurements of PCP in
affected or deceased children showed extreme PCP levels: almost 12 mg/100 mL serum in one
child who survived, and tissue levels in excess of 3 mg/100 g tissue in one of the fatalities.
Biomonitoring studies have shown higher levels of PCP in children compared with
similarly exposed adults, although differences in toxicological response based on these higher
levels are unknown. Kutz et al. (1992) reported higher urinary levels of PCP in adolescents
compared to adults, using data from the National Health and Nutrition Examination Survey
(NHANES), a representative sample of the U.S. population. A study on residents of PCP-treated
log homes (Cline, 1989) also found higher serum PCP levels in children compared with their
parents. The contribution of biological differences and of differences in exposure to this
observed age difference is unknown. One other study of 69 participants, ages 6-87 years (mean
54.6 years), in Saskatchewan, Canada, did not observe any age-related difference in urinary PCP
concentrations (Treble and Thompson, 1996).
There are some data from epidemiologic studies suggesting a susceptibility to adverse
health effects (birth defects or childhood cancers) from paternal-mediated exposure during the
preconception or perinatal periods. A case-control study in Taiwan reported strong associations
145
-------
(adjusted ORs >12.0) with childhood leukemia (103 cases) in relation to paternal work as a wood
treater in the preconception and perinatal periods (Ali et al., 2004), but there was no association
(RR = 1.0) between paternal exposure to PCP and the incidence of childhood leukemia
(11 cases) in the large sawmill worker cohort study (Demers et al., 2006; Heacock et al., 2000).
Another study of the pregnancy outcomes within this sawmill cohort reported associations
between paternal exposure (3 months prior to conception and during the pregnancy) and
congenital anomalies of the eye (Dimich-Ward et al., 1996).
4.8.1.2. Evidence of Reproductive/Developmental Toxicity and Teratogenicity in Animals
Early studies of reproductive or developmental toxicity suggested that PCP is fetotoxic
and teratogenic (Williams, 1982), but these findings were attributed to the chlorinated dibenzo-p-
dioxin and dibenzofuran contaminants. However, a considerable number of studies exist where
laboratory animals or livestock were exposed to both contaminated and pure PCP during
pregnancy, indicating that the contaminants are not solely responsible for the observed fetotoxic
effects. A one-generation study in rats (Schwetz et al., 1978) produced evidence of fetotoxicity
at maternally toxic doses, but also produced evidence of skeletal variations, and of neonatal
toxicity when exposure of the offspring was extended through lactation. A two-generation study
in rats (Bernard et al., 2002) showed evidence of hepatotoxicity from PCP in the offspring.
Fertility was decreased at high doses, some maturational landmarks were delayed in male and
female offspring, and there was evidence for interference with testicular development. Increased
maternal body temperature and resorptions and decreased fetal weights were observed in rats
exposed on various days of pregnancy to aPCP or tPCP (Larsen et al., 1975). Dosing on GDs 9
or 10 induced the highest level of fetotoxicity. No fetal malformations were observed, and the
authors attributed the fetal effects to maternal toxicity.
Two studies of the reproductive toxicity of PCP were performed in mink (Beard and
Rawlings, 1998; Beard et al., 1997). Sex hormone levels in females of the FO generation were
measured, but no changes were observed. However, short-term exposure to PCP (Beard et al.,
1997) reduced reproductive efficiency of the dams at a dose that was 10 times lower than the
dose that caused developmental toxicity in rats (Bernard et al., 2002). Reproductive efficiency
of mink was not affected with long-term exposure to PCP (Beard and Rawlings, 1998).
However, testicular toxicity consisting of interstial cell hyperplasia and testes length was noted
in Fl generation male mink, but they were not as severe in the F2 generation (Beard and
Rawlings, 1998).
4.8.1.3. Evidence of Thyroid Hormone Perturbation in Animals
McConnell et al. (1980) showed that exposure of 10-14-month-old Holstein cattle to PCP
for 160 days resulted in significantly lowered levels of the thyroid hormones 14 and 13. Beard et
al. (1999b) exposed pregnant rams to PCP and found effects on genital development in the male
146
-------
offspring. T4 levels were temporarily decreased during the postnatal period, but other hormone
levels were not affected. The authors suggested that the lowered T4 levels were to blame for the
impaired sexual development of the males. Beard et al. (1999a) conducted a one-generation
reproductive study in sheep exposed to PCP. Reproductive function of the ewes (the rams were
not exposed) was not affected by PCP, although T4 levels were significantly reduced. The
significant thyroid hormone-lowering effect of both aPCP and tPCP has also been demonstrated
in nonpregnant female rats (Jekat et al., 1994). Beard and Rawlings (1998) reported significant
decreases in serum T4 in mink fed 1 mg/kg-day PCP.
Changes in thyroid hormones have been associated with effects (i.e., delayed
myelination, neuronal proliferation, and synapse formation) on neurons. Considering that
thyroid hormones may play a role in neurodevelopmental processes, the disruption of thyroid
homeostasis that has been observed with PCP indicates a potential concern for the critical period
of development of the nervous system (CalEPA, 2006). However, the downstream effects
associated with PCP and decreased T4 levels have not been explored.
A study on pregnant women in Germany has correlated gynecological hormonal
effects—specifically, lower TS levels—with PCP exposure (Gerhard et al., 1999). No conclusive
data exist in support of an estrogenic action of PCP that would be of special concern to humans.
Findings in various animal species exposed to PCP point in the same direction, but no evidence
has been presented in human or animal carcinogenicity evaluations to suggest that PCP-induced
low thyroid hormone levels would be associated with thyroid cancers.
4.8.1.4. Other Considerations
One interesting aspect emerges from one of the CYP450 isozymes, CYP3A4, which is
thought to metabolize PCP in humans (Mehmood et al., 1996). This enzyme is not expressed in
humans before birth; instead, humans express a fetal form, CYP3 A7, which exists for a limited
time after birth. By 1 year, only CYP3 A4 can be found (Williams et al., 2002). Considering that
the metabolites of PCP may be the active form of the compound, if CYP3A4 is not present to
metabolize PCP (this information is unavailable), it is possible that PCP would be less toxic in
humans before they begin to express CYP3A4. An evaluation of published drug clearance data
indicates that clearance of drugs metabolized by CYP3A4 is 3 times lower in neonates compared
with adults, while in children 1-16 years of age, it is about 1.4 times that of adults (Dome et al.,
2005; Dome, 2004). If the metabolites are responsible for the toxic effects, the latter age group
would have an increased risk for PCP-induced toxicity.
The U. S. EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (2005b) refers to stop-exposure studies as possible sources of
information concerning childhood susceptibility. The NTP (1999) rat bioassay included one
dosing regimen where male and female rats were exposed to the same cumulative dose, either
60 mg/kg-day for 1 year or 30 mg/kg-day for 2 years (all animals were sacrificed at 105 weeks).
147
-------
In contrast to the mouse bioassay (NTP, 1989), where the animals were first dosed at 9 weeks of
age, the rats were first dosed at 6 weeks, an age that is considered juvenile. In this study, an
elevated incidence of tumors, mesotheliomas, and nasal squamous cell carcinomas was observed
exclusively in males subjected to the stop-exposure regimen. The findings of the stop-exposure
study (NTP, 1999) suggest that young rats may be more susceptible to the toxicity of PCP
delivered at a high-dose rate.
Data suggest that PCP exposure may result in oxidative DNA damage leading to the
formation of cancers. Few data are available that describe young animals' or children's ability to
repair oxidative stress-induced DNA damage compared with adults. Thus, young animals or
children may be more susceptible to the carcinogenicity of PCP. However, a mitigating factor is
that cell replication and mitotic indices are higher in young organisms than in adults; however,
because these processes tend to promote the propagation of cells with DNA damage or
mutations, it may be assumed that suitable repair mechanisms are in place to prevent that from
happening.
4.8.1.5. Conclusions Concerning Childhood Susceptibility
Evidence in laboratory animals exists to support some reproductive or developmental
toxicity of PCP in laboratory animals. PCP is a weak teratogen, if at all. Many of the effects
reported in fetuses may be linked to maternal toxicity and/or the uncoupling of oxidative
phosphorylation by PCP. However, the thyroid hormone-lowering effect of PCP seen in
animals, and corroborated in one study in human females, is a matter of concern, as low thyroid
levels during pregnancy are known to adversely affect child development (cretinism as the
extreme outcome).
It is unknown if the thyroid hormone-lowering and porphyrogenic effects of PCP have
any potential impact on cancer development in children. One of the possible modes of action for
PCP-induced cancer, oxidative DNA damage, may have a more profound impact in children
compared with adults considering the greater activity (1.4 times higher) of the CYP3A4 pathway
in humans 1-16 years of age compared with adults. In humans, however, CYP3A4 activity can
vary at least 20-fold (Kadlubar et al., 2003; von Ahsen et al., 2001). In the absence of any
knowledge concerning the metabolism of PCP at early life stages, pre- and postnatal
development of DNA repair systems, control of cell proliferation, and plasticity of the immune
system in humans, it is not known whether children are at an increased risk of PCP-induced
cancer.
4.8.2. Possible Gender Differences
There is some indication that PCP is a testicular toxicant in rats (NTP, 1999) and mink
(Beard and Rawlings, 1998). Few published studies have directly compared the effects of PCP
exposure in males and females. Most studies in which PCP was administered to both sexes of a
148
-------
species did not provide substantial or consistent evidence for a difference in gender susceptibility
toward the toxicity of PCP. However, both of the NTP bioassays in mice (NTP, 1989) and rats
(NTP, 1999) found that males were more susceptible to PCP than females for many of the
examined endpoints.
The Hazardous Substances Data Bank, an online database of the National Library of
Medicine (NLM), lists a 20% higher LDso for female rats (175 mg/kg) as compared with male
rats (146 mg/kg) (NLM, 2006). Braun et al. (1977) reported that the toxicokinetics of PCP
differed between male and female rats, with elimination rate constants in females being 20-30%
higher than in males. This finding could explain the slightly lower toxicity of PCP in female
rats.
The NTP stop-exposure study (NTP, 1999) found some sex-related differences in tumor
susceptibility. Increased incidences of nasal squamous cell carcinomas and mesotheliomas were
observed in male rats but not female rats. Given that females were less susceptible to PCP
toxicity than males, this may indicate that a sufficiently high dose was not achieved in females.
The NTP mouse feed study (NTP, 1989) produced similar types of liver cancer in both genders,
although only females had elevated incidences of hemangiomas or hemangiosarcomas in the
liver and spleen. Mode-of-action information to explain gender differences is not available.
Two epidemiologic studies conducted on PCP-exposed women in Germany (Gerhard et
al., 1999; Karmaus and Wolf, 1995) suggest that PCP may affect pregnancy and pregnancy
outcome. Significantly lowered FSH and T3 levels in pregnant, PCP-exposed women compared
with levels in unexposed pregnant women were reported in one study (Gerhard et al., 1999).
Both studies evaluated women exposed to tPCP used as a wood preservative that contained other
toxic agents as contaminants. Because men were not examined in these studies, it cannot be
determined whether the observed hormone disturbances are specific to women. Dimich-Ward et
al. (1996) present epidemiologic evidence for an uncommon paternally transmitted
developmental toxicity in PCP-exposed male workers, suggesting that PCP could be a male
reproductive toxicant.
4.8.3. Other Susceptible Populations
No published experimental animal or human epidemiological studies are available to
evaluate the effects of PCP in a geriatric population or in individuals with a compromised health
status, such as asthmatics, or those with respiratory impairments. A German language
retrospective study (Lohmann et al., 1996; English abstract only) examined possible correlations
among exposures to certain environmental contaminants, neurotoxicity, and multiple chemical
sensitivity (MCS). In almost two-thirds of the cases, exposure to PCP or lindane was associated
with symptoms of neurotoxicity and MCS. The authors emphasized that their study was not
based on a full-fledged epidemiologic evaluation and was therefore purely descriptive.
149
-------
However, it may be suggested that the condition of MCS heightens the sensitivity to neurotoxic
effects in humans exposed to wood preservatives.
Many animal studies provide evidence that it is not the parent compound itself but
hydroquinone and benzoquinone metabolites of PCP that are the biologically reactive
intermediates. This implies that metabolism is required for toxicity to occur. Mehmood et al.
(1996), using yeast cells expressing human CYP450 isozymes, identified CYP3A4 as one
isozyme that can metabolize PCP. Metabolism studies in animals using inducers for specific
CYP450 isozymes, however, indicated that more than one isozyme is responsible for PCP
metabolism (Tsai et al., 2001; Van Ommen et al., 1986a, b). In humans, CYP3A4 activity varies
at least 20-fold and displays gene polymorphism, with numerous known variants (He et al.,
2005; Kadlubar et al., 2003; Hsieh et al., 2001; von Ahsen et al., 2001). Some of the variants
whose catalytic activities have been investigated differ by factors of about two (He et al., 2005;
Amirimani et al., 2000). However, there are also a number of mutant alleles with no catalytic
activity at all (Hsieh et al., 2001). Because these alleles occur very rarely, it may be concluded
that, for CYP3 A4 at least, gene polymorphism does not contribute greatly toward a specific
susceptibility of humans to PCP-induced toxicity. Other enzymes involved in the metabolism of
PCP, such as sulfotransferases or glucuronidases, have not been characterized in detail to warrant
an extensive examination of possible gene polymorphisms.
150
-------
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect—Rationale and Justification
In the absence of human studies on the noncancer effects of PCP, toxicity studies in
experimental animals were considered as the basis for the derivation of the oral RfD for PCP.
The numerous acute, subchronic, and chronic studies characterizing the systemic toxicity of oral
exposure to PCP have been performed in rats, mice, dogs, pigs, rabbits, cattle, mink, and sheep.
The primary target for PCP toxicity with both analytical- and commercial-grade formulations
was consistently identified by the available animal studies as the liver. Hepatotoxicity has been
observed in various animal species after both short- and longer-term exposure to PCP. Other
effects have been reported, including reproductive and developmental toxicity, kidney toxicity,
neurotoxicity, immunotoxicity, and endocrine effects at doses equal to or greater than those
doses eliciting hepatotoxicity.
Many studies in the PCP database were considered to be of limited suitability for
derivation of the oral RfD based on incomplete examination of the animals; failure to report
grade, purity, and effects of PCP; and/or the use of only one experimental dose of PCP. The
remaining studies consist of five chronic studies: three in rats (NTP, 1999; Kimbrough and
Linder, 1978; Schwetz et al., 1978), one in mice (NTP, 1989), and one in dogs (Mecler, 1996).
Additionally, there are five developmental and reproductive studies in rats (Bernard et al., 2002;
Bernard and Hoberman, 2001; Welsh et al., 1987; Schwetz et al., 1978, 1974a).
The Mecler (1996) study examined the toxic effects of tPCP in dogs fed 1.5, 3.5, or
6.5 mg/kg-day tPCP. Decreased absolute body weight (9%) in females was noted at 1.5 mg/kg-
day, and mean body weight and body weight gain continued to decline in both male (decreased
4, 6, and 18% at 1.5, 3.5, and 6.5 mg/kg-day, respectively) and female dogs (decreased 13 and
20% at 3.5 and 6.5 mg/kg-day, respectively) as the dose increased. Hepatotoxic effects were
noted at 1.5 mg/kg-day with increased incidence of liver pigmentation (in 100% of males and
females) consistent with LF, cytoplasmic vacuolation (25% of males, 75% of females), chronic
inflammation (100% of males, 50% of females), and severely dark, discolored livers (25% of
males, 75% of females) accompanied by significantly increased serum ALP activity (twofold
increase over controls for both sexes), and significantly increased relative liver weight in males
(14%) and females (37%), and absolute liver weight in females (24%). Absolute liver weight
was increased in males (10%), but was not considered statistically significantly greater than
controls. As the dose of tPCP increased, the effects observed in the animals of the 1.5 mg/kg-
day dose group increased in incidence and severity. Additional effects observed at the 3.5 and
6.5 mg/kg-day doses include increases in serum activity of ALP (2.3- and 4.9-fold in males and
2.6- and 6.8-fold in females at 3.5 and 6.5 mg/kg-day, respectively), ALT (2.8- and 3.9-fold in
151
-------
males and 3.1- and 8.8-fold in females at 3.5 and 6.5 mg/kg-day, respectively), and AST
(1.2- and 1.25-fold in males and 1.1- and 1.7-fold in females, respectively), and minimum
hepatocellular necrosis (25% of males, 50% of females). Additionally, foci of hepatocellular
hypertrophy, hyperplasia consistent with cirrhosis, fibrosis and decreased hematological
parameters (including RBC count, hemoglobin, and hematocrit) were noted in the treated
animals. The two animals that were sacrificed in extremis due to morbidity following exposure
to tPCP at 6.5 mg/kg-day were characterized as moribund from hepatic insufficiency (Mecler,
1996). The LOAEL was 1.5 mg/kg-day (lowest dose tested), based on dose-related increases in
incidence of hepatocellular pigmentation, cytoplasmic vacuolation, chronic inflammation, and
severely discolored livers accompanied by statistically significantly increased relative liver
weights and serum enzymes, and increased absolute liver weights (significant in females). A
NOAEL was not established.
Kimbrough and Linder (1978) fed tPCP and aPCP to male and female rats for 8 months
in the diet. A decrease in final body weight (15-16% in tPCP-treated animals; 5 and 10% in
aPCP females and males, respectively) and dose-related increases in incidence of liver lesions,
including hepatocyte hypertrophy, vacuolation, pleomorphism, periportal fibrosis, abundant
brown pigment in macrophages and Kupffer cells, bile duct proliferation, adenofibrosis, and
cytoplasmic hyaline inclusions, were observed in rats exposed to doses starting at 2 mg/kg-day
for tPCP and at 44 or 48 mg/kg-day (males and females, respectively) for aPCP; however, no
incidence data for these effects were reported. Effects were more severe in rats treated with
tPCP. The LOAELs, based on hepatotoxicity, were 2 mg/kg-day for males and females exposed
to tPCP and 44 and 48 mg/kg-day for males and females, respectively, exposed to aPCP. The
NOAEL could not be determined for tPCP. The NOAELs were 9 and 10 mg/kg-day for male
and females, respectively, exposed to aPCP.
NTP (1999) reported significantly increased cystic degeneration of hepatocytes in 56 and
78% of males following administration of 20 and 30 mg/kg-day aPCP and eosinophilic focus in
18% of males at 30 mg/kg-day aPCP. Increased centrilobular hepatocyte hypertrophy was noted
in 60% of males and females and cytoplasmic hepatocyte vacuolization was observed in 80% of
males examined in an interim evaluation after 7 months of administration of 60 mg/kg-day.
Increases in serum activity of ALT (1.5-fold for males, 1.1-fold for females), ALP (1.2-fold for
males, 1.1-fold for females), and sorbitol dehydrogenase (1.9-fold for males, 1.4-fold for
females) were measured in rats administered 60 mg/kg-day aPCP for 7 months. After 2 years
(only 1 year of exposure), male rats exhibited increased incidences of liver lesions including
basophilic focus (62%), chronic inflammation (68%), cytoplasmic vacuolization (26%), and
cystic degeneration of hepatocytes (56%) at 60 mg/kg-day aPCP. In females, clear cell focus
(32%) and cytoplasmic vacuolization (18%) were slightly increased after 1 year of treatment
with 60 mg/kg-day followed by 1 year of nontreatment. The EPA determined that the LOAEL
was 20 mg/kg-day for male rats based on liver toxicity; the NOAEL was 10 mg/kg-day. The
152
-------
LOAEL was 30 mg/kg-day for female rats based on a biologically significant decrease in body
weight; the NOAEL was 20 mg/kg-day.
Rats treated with 1, 3, 10, or 30 mg/kg-day EC-7 (Schwetz et al., 1978) for approximately
2 years exhibited slight increases (~1.7-fold) in serum ALT activity at 30 mg/kg-day. Pigment
accumulation in the centrilobular hepatocytes of the liver occurred in 30 and 59% of females
given 10 and 30 mg/kg-day. Similarly, 26 and 70% of females receiving 10 and 30 mg/kg-day
EC-7 exhibited pigment accumulation in the epithelial cells of the proximal convoluted tubules
in the kidney. The study authors reported that the LOAEL was 30 mg/kg-day for males and
10 mg/kg-day for females, based on pigment accumulation in the liver and kidney. The NOAEL
was 10 mg/kg-day for males and 3 mg/kg-day for females.
NTP (1989) reported an increased incidence of liver lesions, including clear cell focus
(23 and 40%), acute diffuse necrosis (87 and 98%), diffuse cytomegaly (100% for both
formulations), diffuse chronic active inflammation (89 and 75%), and multifocal accumulation of
brown pigmentation (LF and cellular debris) in Kupffer cells (96 and 83%) in male mice
administered 18 mg/kg-day tPCP and EC-7, respectively. Incidence of lesions generally
increased with increasing dose. Female mice exhibited clear cell focus (6 and 4%), acute diffuse
necrosis (90 and 42%), diffuse cytomegaly (98 and 74%), diffuse chronic active inflammation
(69 and 8%), and multifocal accumulation of brown pigmentation (76 and 65%) at doses of
17 mg/kg-day for tPCP and EC-7, respectively. Similar to male mice, the incidence of hepatic
lesions in females increased with increasing dose. The EPA determined that the LOAELs were
18 mg/kg-day for males and 17 mg/kg-day for females for both tPCP and EC-7. NOAELs could
not be established for either tPCP or EC-7 because effects in the liver occurred at the lowest
doses tested in male and female mice.
Results of studies that examined the effects of PCP on the liver indicate that rats, mice,
and rabbits (NTP, 1999, 1989; Kimbrough and Linder, 1978; Schwetz et al., 1978) are less
sensitive to the hepatotoxicity of PCP than the beagle dog (Mecler, 1996). Hepatotoxic effects
were observed in rodent and rabbit studies at doses that exceeded those that caused effects in
dogs. Specifically, Mecler (1996) reported that a 1-year exposure to tPCP at a concentration of
1.5 mg/kg-day induced hepatotoxicity characterized by increases in hepatic lesions (including
liver pigmentation, cytoplasmic vacuolation, chronic inflammation, and the appearance of dark,
discolored livers) accompanied by increases in absolute and relative liver weight and serum
activity of ALT and ALP in male and female dogs.
Reproductive evaluation of PCP (EC-7) toxicity revealed treatment-related effects in rats
at doses of 30 mg/kg-day (Bernard et al., 2002; Schwetz et al., 1978). Decreased parental (8 and
10% in males and females, respectively) and fetal body weight (14-27%), reduced number of
pups born alive (6%), pup survival (79%), and increased fetal skeletal variations (quantitative
data not reported) were observed at 30 mg/kg-day in a study of rats exposed to 0, 3, or 30 mg/kg-
day of PCP (Schwetz et al., 1978). Bernard et al. (2002) reported reductions of 5.3 and 15% for
153
-------
body weight in 30 and 60 mg/kg-day tPCP treated parental males, respectively. Parental female
body weights were reduced 8.3% in the 60 mg/kg-day tPCP dose group. Body weights of the
Fl generation rats were reduced 10 and 30% in males and 6 and 23% in females at 30 and
60 mg/kg-day, respectively. Increased liver weight, enlarged liver, centrilobular hypertrophy/
vacuolation (100% of males and females), multifocal inflammation (20 and 57% of males;
62 and 63% of females), single-cell necrosis (13 and 70% of males; 38 and 80% of females), and
pigmentation (LF; 13 and 37% of males; 45 and 87% of females) were observed in parental rats
treated with 30 and 60 mg/kg-day, respectively. Centrilobular hypertrophy (76% of males; 43%
of females), pigmentation (10% of females), and multifocal inflammation (7% of males; 13% of
females) were observed at the 10 mg/kg-day dose of tPCP. Preputial separation was delayed
(~2 days) and spermatid count decreased (10%) in Fl males in the 30 mg/kg-day dose group,
while vaginal patency was delayed 1 day in females of the 10 mg/kg-day dose group.
Reproductive effects associated with the Fl generation included decreases in live litter size
(22%) and viability index (94.4 versus 98.8% in controls) at 60 mg/kg-day, a dose that exceeded
that of parental toxicity. The F2 generation presented similar reproductive effects at 60 mg/kg-
day (Bernard et al., 2002).
Bernard and Hoberman (2001) reported reductions in maternal (15%) and fetal body
weight (79% of controls) and litter size (86% of controls) and increased resorptions (83% of
dams versus 41% of controls), and visceral (27%) and skeletal malformations/variations (96%)
in rats developmentally exposed to 80 mg/kg-day of tPCP. Decreased maternal body weight
gain (22 and 74% for tPCP and aPCP, respectively) and fetal effects, including decreased body
weight and crown-rump length (13 and 22% for tPCP and aPCP, respectively), and increased
resorptions (27% of fetuses and 95% of litters for tPCP; 97% of fetuses and 100% of litters for
aPCP) were observed in rats administered 30 mg/kg-day (Schwetz et al., 1974a). The incidence
of delayed ossification of the skull (threefold increase over controls) was noted at a lower dose
(5 mg/kg-day) by Schwetz et al. (1974a). Similar to the other developmental studies, Welsh et
al. (1987) reported a decrease in maternal body weight gain (76% of control) and the number of
viable fetuses (99% decrease) at 43 mg/kg-day of aPCP. Rats exposed to 13 mg/kg-day PCP
exhibited an increase in percentage of females with one or more (87.5% of treated versus 67.74%
of controls) or two or more resorptions (81.25% of treated versus 41.94% of controls), and
fetuses showed an increase in incidence of misshapen centra (36%), and at least two skeletal
variations (2.4-fold increase over controls) (Welsh et al., 1987). A developmental study in
rabbits showed slight, but significant, decreases in maternal body weight gain of 12 and 29% at
15 and 30 mg/kg-day tPCP, respectively (Bernard et al., 2001).
Reproductive and developmental effects in rodents and rabbits as well as additional
effects (kidney, immunological, and neurological; see Section 4.6.1 for more detailed discussion)
occurred at doses of PCP that exceeded the doses that elicited hepatotoxicity in dogs (as reported
by Mecler, 1996). Therefore, the chronic study by Mecler (1996) in male and female beagle
154
-------
dogs was selected as the principal study for RfD derivation as it identified effects
(hepatotoxicity) at the lowest dose of any of the available studies. The EPA established a
LOAEL of 1.5 mg/kg-day based on hepatotoxicity in dogs (Mecler, 1996) characterized by dose-
related increases in incidence and severity of pigmentation, cytoplasmic vacuolation, chronic
inflammation, and severely discolored livers accompanied by increased relative liver weight and
serum enzymes, and increased absolute liver weight (statistically significant in females).
5.1.2. Methods of Analysis
Hepatotoxicity of PCP was evident in the histopathological results of tPCP administration
in dogs of the Mecler (1996) study. The observed hepatotoxicity was present in many of the
treated dogs (both male and female) at the lowest dose tested, 1.5 mg/kg-day. These effects were
minimally present, if at all, in the control animals. For the 3.5 and 6.5 mg/kg-day doses, the
hepatotoxicity was present in all animals that survived and the severity of the effects increased
with dose. A NOAEL/LOAEL approach is used to derive the RfD for PCP based on the LOAEL
of 1.5 mg/kg-day for hepatotoxicity identified by Mecler (1996) in dogs.
In general, the benchmark dose (BMD) approach is preferred over the NOAEL/LOAEL
approach for identifying a point of departure (POD). In this particular case, however, the
incidence of two of the key liver effects (i.e., hepatocellular pigmentation in males and females
and chronic inflammation in males) increased from 0% in the controls to 100% in the low-dose
group, and then remained at 100% in both the mid- and high-dose groups. Because of the 100%
response at all doses tested, these data are not amenable to BMD modeling, as none of the dose-
response models in the EPA's BMD Software (BMDS) can adequately accommodate this steep
increase. Thus, the NOAEL/LOAEL approach was employed to identify the POD.
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)
The derivation of the RfD for liver effects from the 1-year toxicity study in beagle dogs
(Mecler, 1996) is calculated from the LOAEL by application of a composite UF as follows:
RfD = LOAEL -T- UF
RfD = 1.5 mg/kg-day + 300 = 0.005 mg/kg-day = 5 x 10"3 mg/kg-day
The composite UF of 300 consists of individual UFs of 10 for intraspecies variation,
10 for interspecies variation, and 3 for the use of a LOAEL instead of a NOAEL. The UFs were
applied to the POD as described below:
• A default intraspecies uncertainty factor (UFn) of 10 was applied to account for
variability in susceptibility among members of the human population in the absence of
quantitative information on the variability of human response to PCP. Current
information is unavailable to assess human-to-human variability in PCP toxicokinetics
155
-------
and toxicodynamics; therefore, to account for these uncertainties, a factor of 10 was
applied for individual variability.
A default interspecies uncertainty factor (UFA) of 10 was applied to account for the
potential pharmacokinetic and pharmacodynamic differences between dogs and humans.
Although toxicokinetic data are available in some animals, a description of toxicokinetics
in either dogs or humans is limited or not available. In the absence of data to quantify
specific interspecies differences, a factor of 10 was applied.
• A LOAEL to NOAEL uncertainty factor (UFi,) of 3 was applied to account for the
extrapolation from a LOAEL to a NOAEL. The 1.5 mg/kg-day dose level was selected
as the LOAEL based on histopathological changes in the liver, consisting of increased
incidence of pigmentation in both males and females; minimal chronic inflammation in
males; and increased relative liver weights in males and absolute and relative liver weight
in females. These effects were accompanied by small changes (less than twofold) in
serum enzymes (ALT in males and ALP in males and females), indicating an effect of
minimal toxicological significance. Therefore, a factor 3 was applied to account for the
use of a LOAEL that is characterized by effects that can be considered mild.
• A UF of 1 was applied to extrapolate from a subchronic to a chronic (UFs) exposure
duration because the RfD was derived from a study using a chronic exposure protocol.
• A UF of 1 was applied to account for database deficiencies (UFo). The database for PCP
contains human studies; chronic studies in rats, mice, and dogs; subchronic studies in
various animal species; neurological, reproductive, endocrine, and developmental and
reproductive toxicity studies; and a two-generation reproductive toxicity study.
5.1.4. RfD Comparison Information
The predominant noncancer effect of subchronic and chronic oral exposure to PCP is
hepatic toxicity. Figure 5-1 provides a graphical display of dose-response information from six
studies (summarized in Table 5-1) that reported liver toxicity in experimental animals following
chronic oral exposure to PCP, focusing on candidate PODs that could be considered in deriving
the oral RfD. As discussed in Sections 5.1.1 and 5.1.2, among those studies that demonstrated
liver toxicity, the study by Mecler (1996) provided the most sensitive data set for deriving the
RfD. Potential reference values that might be derived from each of the other studies are also
presented.
156
-------
Dog
Rat
Rat
Rat
Mouse
100
6-
Q o.i
o.oi
o.ooi
0 Point of Departure
• UFL, LOAEL to
NOAEL
«V] UFA, Interspecies
^-^
T\l UFn, Intraspecies
o
Reference Dose
Increased liver Lesions Pigment NOAEL Decreased body Lesions NOAEL Necrosis, lesions
weight, serum LOAEL (tPCP) (EC-7) weight, lesions (aPCP) LOAEL (tPCP, EC-7)
enzymes, lesions Kimbrough and Schwetz et al., NOAEL (aPCP) NTP, 1999; 2-yr NTP, 1989; 2-yr
LOAEL (tPCP) Linder, 1978; 8-m 1978; 2-yr Kimbrough and
Mecler, 1996; 1-yr Linder, 1978; 8-m
Figure 5-1. Array of candidate PODs with applied UFs and reference values for a subset of hepatotoxic effects
of studies in Table 5-1.
157
-------
Table 5-1. Candidate PODs for hepatotoxicity with applied UFs and
potential reference values
Endpoint
Increased liver weight
and serum enzymes;
hepatocellular lesions;
LOAEL
Dog, 1 yr
Hepatocellular lesions;
LOAEL
Rat, 8 mo
Pigment; NOAEL
Rat, 2 yr
Decreased body
weight, hepatocellular
lesions; NOAEL
Rat, 8 mo
Lesions; NOAEL
Rat, 2 yr
Necrosis,
hepatocellular lesions;
LOAEL
Mouse, 2 yr
Candidate
POD
(mg/kg-d)
1.5
2
3
9(M)
10 (F)
10
18
UFs
Composite
UF
300
300
100
100
100
1,000
UFL
3
3
1
1
1
10
UFA
10
10
10
10
10
10
UFH
10
10
10
10
10
10
Potential
reference
value
(mg/kg-d)
0.005
0.007
0.03
0.09 (M)
0.1 (F)
0.1
0.018
Reference
Mecler(1996)
(tPCP)
Kimbrough and
Linder (1978)
(tPCP)
Schwetz et al.
(1978) (EC-7)
Kimbrough and
Linder (1978)
(aPCP)
NTP (1999)
(aPCP)
NTP (1989)
(tPCP, EC-7)
158
-------
Reproductive and developmental studies in experimental animals have found that PCP
can produce prenatal loss, skeletal and soft-tissue variations, delays in puberty, and decreased
fetal weight; these doses also produced toxic effects in the dams. These studies show that the
developing embryo and fetus may be a target of PCP toxicity; however, study results indicate
that PCP is more likely to be embryo- and fetotoxic rather than teratogenic. Figure 5-2 provides
a graphical display of dose-response information from two reproductive and four developmental
studies (summarized in Table 5-2). For the reasons discussed above and in Section 5.1.1, liver
effects in the dog observed in the study by Mecler (1996) are considered the most sensitive
effects to serve as the basis for the derivation of the RfD for PCP. The potential reference value
associated with delayed ossification of the skull in fetuses of rats administered 5 mg/kg-day
aPCP from GDs 6 to 15 (Schwetz et al., 1974a) is identical to the RfD based on hepatotoxicity in
dogs administered 1.5 mg/kg-day tPCP (Mecler, 1996). The POD for hepatotoxicity is the same
as or lower than that for reproductive and developmental toxicity, and the resulting RfD should
protect against reproductive and developmental effects of PCP.
159
-------
Repro
Dev
Repro
Dev/Repro
0 Point of Departure
^ffl L!FL, LOAEL to
HHi NOAEL
fj/VJ UFA, Interspecies
I; ; I UFa Intraspecies
o
Reference Dose
Decreased survival and
growth, increased
skeletal variations
NOAEL (EC-7)
Schwetz et al., 1978
Increased resorptions Delayed Increased Delayed vaginal Increased resorptions,
and skeletal ossification of the resorptions, soft patency decreased litter size and
variations, decreased skull tissue and skeletal LOAEL (tPCP) fetal weight, increased
fetal weight and
crown-rump length
NOAEL (aPCP)
Welsh etal., 1987
LOAEL (aPCP) anomalies Bernard et al., 2002 visceral malformations
Schwetz et al., NOAEL (tPCP) and skeletal anomalies
1974a Schwetz et al., 1974a NOAEL (tPCP)
Bernard and Hoberman,
2001
Figure 5-2. Array of candidate PODs with applied UFs and reference values for a subset of reproductive and
developmental effects of studies in Table 5-2.
160
-------
Table 5-2. Candidate PODs for reproductive and developmental toxicity in
rats with applied UFs, and potential reference values
Endpoint
Decreased survival and
growth, increased
skeletal variations;
NOAEL
Increased resorptions
and skeletal variations,
decreased fetal weight
and crown-rump length;
NOAEL
Delayed ossification of
the skull; LOAEL
Increased resorptions,
soft tissue and skeletal
anomalies;
NOAEL
Delayed vaginal
patency; LOAEL
Increased resorptions,
decreased litter size and
fetal weight, increased
visceral malformations,
and skeletal anomalies;
NOAEL
Candidate
POD
(mg/kg-d)
3
4
5
5.8
10
30
UFs
Composite
UF
100
100
1,000
100
1,000
100
UFL
1
1
10
1
10
1
UFA
10
10
10
10
10
10
UFH
10
10
10
10
10
10
Potential
reference values
(mg/kg-d)
0.03
0.04
0.005
0.06
0.01
0.3
Reference
Schwetz et al.
(1978) (EC-7)
Welsh etal. (1987)
(aPCP)
Schwetz et al.
(1974a) (aPCP)
Schwetz et al.
(1974a) (tPCP)
Bernard et al.
(2002) (tPCP)
Bernard and
Hoberman(2001)
(tPCP)
5.1.5. Previous RfD Assessment
The previous RfD, posted to the IRIS database in January 1987, was based on a chronic
oral rat study by Schwetz et al. (1978). Investigators administered 0, 3, 10, or 30 mg/kg-day
PCP in feed ad libitum to 25 rats/sex/dose for 22 (males) or 24 months (females). Derivation of
the RfD of 3 x 10"2 mg/kg-day was based on a NOAEL of 3 mg/kg-day for liver and kidney
pathology, evidenced by pigmentation of the liver and kidneys in female rats at 10 mg/kg-day
(LOAEL). A composite UF of 100 (UFn of 10 for intraspecies variability and UFA of 10 for
interspecies variability) was applied to the NOAEL.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
Adequate data are not available to derive an inhalation RfC. No chronic or subchronic
animal studies for inhalation exposure are available. The previous IRIS assessment did not
derive an RfC.
161
-------
5.3. UNCERTAINTIES IN THE RfD AND RfC
Uncertainties associated with the RfD in the assessment for PCP are identified in the
following discussion. As presented earlier in Section 5.1, UFs were applied to the POD, a
LOAEL, for deriving the RfD. Factors accounting for uncertainties associated with a number of
steps in the analyses were adopted to account for extrapolating from an animal bioassay to
human exposure and for a diverse population of varying susceptibilities. These extrapolations
are carried out with default approaches given the limitations of experimental PCP data for the
interspecies and intraspecies differences.
A range of animal toxicology data is available for the hazard assessment of PCP, as
described in Section 4. Included in these studies are short-term and long-term studies in dogs,
rats, and mice and developmental and reproductive toxicity studies in rats, as well as numerous
supporting studies. Toxicity associated with oral exposure to PCP is observed as hepatic and
reproductive and developmental endpoints. Critical data gaps have been identified in Section 4
and uncertainties associated with data deficiencies are more fully discussed below.
Consideration of the available dose-response data to determine an estimate of oral
exposure that is likely to be without an appreciable risk of adverse health effects over a lifetime
led to the selection of the 1-year oral study in beagle dogs (Mecler, 1996) as the principal study
and hepatotoxicity (characterized by increased incidence and severity of liver pigmentation,
cytoplasmic vacuolation, chronic inflammation, and severely discolored livers, significantly
increased absolute [females only] and relative liver weights, and increased serum enzyme
activity) as the critical effect for deriving the RfD for PCP. The dose-response relationships for
oral exposure to PCP and hepatotoxicity in rats and mice are also available for deriving an RfD,
but are associated with higher NOAELs/LOAELs that would be protected by the selected critical
effect and corresponding POD.
The Mecler (1996) study used a PCP formulation that was 90.0% pure. As discussed in
Section 5.4.4, impurities in the formulation could influence the toxicity of the test compound.
Whether these impurities would reduce or increase the toxicity relative to aPCP is unknown.
The derived RfD was quantified using a LOAEL for the POD. A POD based on a
NOAEL or LOAEL is, in part, a reflection of the particular exposure concentration or dose at
which a study was conducted. It lacks characterization of the dose-response curve and for this
reason it is less informative than a POD obtained from BMD modeling. In this particular case,
however, BMD modeling was not utilized for the determination of the POD for hepatotoxicity in
Mecler (1996) because the incidence of two of the key liver effects (i.e., hepatocellular
pigmentation in males and females and chronic inflammation in males) increased from 0% in the
controls to 100% in the low-dose group, and then remained at 100% in both the mid- and high-
dose groups. Because none of the dose-response models in BMDS can adequately accommodate
this steep increase in response at the lowest dose, it was determined that the critical data set was
162
-------
not amenable to BMD modeling, and the NOAEL/LOAEL approach was used to identify the
POD.
The oral reproductive and developmental toxicity studies indicate that the developing
embryo and/or fetus may be a target of PCP toxicity. However, observed toxic effects were not
teratogenic in nature, but rather embryo- or fetotoxic. Systemic effects were frequently observed
in the dams at similar doses. In the two-generation reproductive study, hepatotoxic effects were
noted in the dams at doses that elicited delayed vaginal patency in the Fl offspring females. The
potential reference value associated with delayed ossification of the skull in fetuses of rats
administered 5 mg/kg-day aPCP from GDs 6 to 15 (Schwetz et al., 1974a) is identical to the RfD
based on hepatotoxicity in dogs administered 1.5 mg/kg-day tPCP (Mecler, 1996). The POD for
hepatotoxicity is lower than that for reproductive and developmental toxicity, and the resulting
RfD should protect against reproductive and developmental effects of PCP.
A LOAEL was identified based on hepatotoxicity in dogs administered tPCP in Mecler
(1996). The hepatotoxicity was observed at all doses, including the lowest dose tested; therefore,
a NOAEL was not established. In the absence of an established NOAEL, the LOAEL was used
as the POD to derive the RfD. A threefold UF was applied to account for the use of a POD
characterized by effects that can be considered mild at the dose established as the LOAEL.
Extrapolating from animals to humans embodies further issues and uncertainties. The
effect and its magnitude associated with the concentration at the POD in dogs are extrapolated to
human response. Pharmacokinetic models are useful for examining species differences in
pharmacokinetic processing; however, dosimetric adjustment using pharmacokinetic modeling
was not available for oral exposure to PCP. Information was unavailable to quantitatively assess
toxicokinetic or toxicodynamic differences between animals and humans, so the 10-fold UF was
used to account for uncertainty in extrapolating from laboratory animals to humans in the
derivation of the RfD.
Heterogeneity among humans is another uncertainty associated with extrapolating doses
from animals to humans. In the absence of PCP-specific data on human variation, a factor of
10 was used to account for uncertainty associated with human variation in the derivation of the
RfD.
5.4. CANCER ASSESSMENT
5.4.1. Choice of Study/Data—Rationale and Justification
The available epidemiologic studies support an association between PCP exposure and
development of specific cancers, i.e., non-Hodgkin's lymphoma, multiple myeloma, soft tissue
sarcoma, and liver cancer (Section 4.1.1). However, the lack of an exposure estimate that allows
for an absolute rather than a relative level of exposure renders these studies unsuitable for
deriving cancer risk estimates for PCP via the oral or inhalation routes. The most detailed
exposure assessment was in the large cohort study of over 26,000 sawmill workers in British
163
-------
Columbia (Demers et al., 2006). This study used a metric based on a cumulative dermal
chlorophenol exposure score, with 1 exposure year defined as 2,000 hours of dermal contact.
Two well-conducted studies provide data for the carcinogenicity of PCP via the oral route
in laboratory animals: one study utilizing B6C3Fi mice (NTP, 1989) and another study in F344
rats (NTP, 1999). Two types of PCP (tPCP and EC-7) were carcinogenic in the mouse.
Hepatocellular adenomas/carcinomas and adrenal medullary pheochromocytomas developed in
male mice treated with tPCP or EC-7, and hepatocellular adenomas/carcinomas and
hemangiosarcomas developed in female mice treated with tPCP or EC-7 and adrenal medullary
pheochromocytomas developed in female mice treated with EC-7.
In the mouse study, the carcinogenicity of tPCP, which contains appreciable amounts of
chlorinated dibenzo-p-dioxins and dibenzofurans, was compared with the carcinogenicity of
EC-7, which contains relatively low levels of the dioxins and furans. Mice were administered
tPCP (90.4% purity; 18 or 35 mg/kg-day for males and 17 or 35 mg/kg-day for females) or EC-7
(91.9% purity; 18, 37, or 118 mg/kg-day for males and 17, 34, or 114 mg/kg-day for females) in
feed for 2 years. In male mice, the incidence of hepatocellular adenomas and carcinomas
combined showed a statistically significantly elevated trend with increasing levels of tPCP and
EC-7. In female mice, the incidence of hepatocellular adenomas and carcinomas combined
showed a statistically significantly elevated trend with increasing levels of EC-7. The incidence
of hepatocellular adenomas and carcinomas combined was statistically significantly elevated
only at 114-118 mg/kg-day EC-7 when compared with the control group. The remaining
exposures exhibited an increase in hepatocellular adenomas and carcinomas; however, these
were not considered statistically significant when compared with control values.
Adrenal gland medullary pheochromocytomas and malignant pheochromocytomas were
observed in all dose groups of both tPCP and EC-7 grades of PCP. There was a statistically
significant increase in the incidence of combined pheochromocytomas and malignant
pheochromocytomas in male mice at all doses of tPCP and all doses of EC-7, except 18 mg/kg-
day. Pheochromocytomas were also observed in female mice administered tPCP and EC-7,
although the appearance of tumors in tPCP mice did not exhibit a dose-related increase and the
only statistically significant increase in incidence was observed in the 114-118 mg/kg-day EC-7
dose group. A significant positive trend was observed for pheochromocytomas in male mice
treated with tPCP and male and female mice treated with EC-7.
Hemangiosarcomas were observed in male mice administered both grades of PCP,
although the incidences were slight and not considered statistically significant. Female mice
administered tPCP showed an increase in hemangiosarcomas at both doses, but the increase was
only significant at the high dose (35 mg/kg-day for tPCP). Increased incidences of combined
hemangiomas and hemangiosarcomas were observed in EC-7 females, and the incidence in the
high-dose (118 mg/kg-day) group was significantly elevated compared with controls.
164
-------
The rat bioassay (NTP, 1999) examined the effects of aPCP in male and female F344
rats. There was some evidence of carcinogenicity in the male rats that exhibited a significantly
higher incidence of malignant mesothelioma at 60 mg/kg-day (dose used in the 1-year stop-
exposure study) compared with that of controls. The incidence exceeded the range of historical
controls. The incidence of nasal squamous cell carcinomas was also elevated in 60 mg/kg-day
males, and while the incidence did not achieve statistical significance compared with that of
concurrent controls, it did exceed the range of historical controls. Nasal squamous cell
carcinomas were observed in male rats administered 10 mg/kg-day and were the only neoplastic
finding in male rats treated for the full 2 years of the bioassay that occurred with a higher
incidence than that of historical controls. However, nasal tumors were not considered treatment-
related because the incidence at 20 and 30 mg/kg-day was less than or equal to the control
incidence. There were no treatment-related increases in the incidences of tumors in female rats
receiving aPCP. This study showed some evidence of carcinogenicity of aPCP in male F344 rats
exposed to 60 mg/kg-day aPCP, based on increased incidences of mesothelioma and nasal
squamous cell carcinoma in the stop-exposure study.
The mouse study was selected for dose-response assessment based on statistically
significant increased incidences of hepatocellular adenomas and carcinomas, adrenal
pheochromocytomas and malignant pheochromocytomas, and hemangiomas and
hemangiosarcomas (in liver and spleen) at multiple exposure levels in males and females. The
study by NTP (1989) was used for development of an oral slope factor. This was a well-
designed study, conducted in both sexes of B6C3Fi mice with two grades of PCP (tPCP and
EC-7) and with 50 male and 50 female mice per dose group (typical for NTP-type bioassays).
The test animals were allocated among two dose levels for tPCP and three dose levels for EC-7
with untreated control groups for each PCP formulation. Animals were observed twice daily and
examined weekly (for 12-13 weeks) and then monthly for body weight and monthly for feed
consumption. Animals were necropsied and all organs and tissues were examined grossly and
microscopically for histopathological lesions. Tumor incidences were elevated with increasing
exposure level at multiple sites in both sexes, including the liver, adrenal gland, and circulatory
system.
The male F344 rat tumor incidence data (NTP, 1999), while demonstrating some
evidence of carcinogenicity, were not used for deriving low-dose quantitative risk estimates.
The responses of increased incidence of mesothelioma and nasal squamous cell carcinoma in
male rats were lower than those of the mice (NTP, 1989) at a greater exposure level, suggesting
greater sensitivity of the mice. The toxicological database for PCP studies in rodents has shown
the mouse model, rather than the rat, to be a more sensitive model of PCP hepatotoxicity.
Additionally, the differences in the presence of metabolites, TCpBQ in mice versus TCoBQ in
rats and subsequent formation of DNA adducts via TCpBQ that is believed to be associated with
the oxidative stress-related toxicity and the proposed mode of action, also suggest that the mice
165
-------
are more sensitive than the rats. Although the NTP (1999) bioassay in rats administered aPCP
reported mesotheliomas and nasal squamous cell carcinomas, the tumor incidence was
statistically significantly elevated only at the high dose (1-year exposure). The lack of a
significant dose-response trend in the rat data and the observation of consistently greater
sensitivity to PCP in mice, rather than rats, led to the use of the mouse data for the derivation of
the slope factor. Consequently, dose-response modeling was not carried out with the rat tumor
data.
5.4.2. Dose-Response Data
Oral cancer risk estimates were calculated based on the incidences of hepatocellular and
adrenal medullary tumors in male mice, and hepatocellular tumors, adrenal medullary tumors,
and hemangiomas/hemangiosarcomas in female mice treated with tPCP or EC-7 (NTP, 1989).
Data are not available to indicate whether malignant tumors developed specifically from
progression of benign tumors; however, etiologically similar tumor types (i.e., benign and
malignant tumors of the same cell type) were combined for these analyses because of the
possibility that the benign tumors could progress to the malignant form. Thus, adenomas and
carcinomas of the liver were considered together because adenomas develop from the same cell
lines and can progress to carcinomas. The adrenal medullary tumors, distinguished as either
pheochromocytomas or malignant pheochromocytomas, were also considered together. The
classification of malignant pheochromocytoma was assigned if the pheochromocytoma
progressed and was observed as obliterating the cortex (outer layer of the adrenal gland) or
penetrating the capsule of the adrenal gland. Hemangiosarcomas differed from the hemangiomas
in that the hemangiosarcomas consisted of a greater amount of pleomorphic and anaplastic
endothelial cells (NTP, 1989); these tumors were also considered together.
The male and female mice were exposed to tPCP and EC-7, two formulations of PCP that
are approximately 90% pure; however, the composition of the impurities identified in these two
formulations differs both qualitatively and quantitatively (e.g., see Table B-3 in Appendix B). In
addition, although the tumors in male and female mice exposed to tPCP were qualitatively the
same as the tumors in EC-7-exposed mice, the tumor incidences were quantitatively different
(see Table 5-3). Therefore, given the differences in the contaminant profiles for the two PCP
formulations and the quantitative differences in tumor incidences across the formulations, the
tumor data sets for tPCP and EC-7 were modeled separately. Animals dying before the first
appearance of tumors during the first year of exposure in any group of that sex were censored
from the group totals when figuring the denominators. This adjustment was made so that the
denominators included only those animals at risk for developing tumors. The incidences of
tumors in mice treated with tPCP and EC-7 are presented in Table 5-3.
166
-------
Table 5-3. Incidence of tumors in B6C3Fi mice exposed to tPCP and EC-7
in the diet for 2 years
Tumor type
Males
Hepatocellular
adenoma/carcinoma
Adrenal benign/malignant
pheochromocytoma
Females
Hepatocellular
adenoma/carcinoma
Adrenal benign/malignant
pheochromocytoma
Hemangioma/hemangio-
sarcoma
tPCP,
ppm in diet
0
100
200
mg/kg-da
0
7/32b
(7/28)d
0/3 lb
(0/26)
0
3/33
(3/31)
2/33
(2/31)
0/3 5b
(0/33)
18
26/47c
(26/46)
10/45C
(10/41)
17
9/49
(9/49)
2/48
(2/48)
3/50
(3/50)
35
37/48c
(37/46)
23/45c
(23/44)
35
9/50
(9/48)
1/49
(1/47)
6/50c
(6/48)
EC-7,
ppm in diet
0
100
200
600
mg/kg-da
0
6/3 5b
(6/33)
l/34b
(1/32)
0
l/34b
(1/34)
0/3 5b
(0/35)
0/3 5b
(0/35)
18
19/48C
(19/45)
4/48
(4/45)
17
4/50
(4/49)
2/49
(2/48)
1/50
(1/49)
37
21/48C
(21/38)
21/48C
(21/39)
34
6/49
(6/49)
2/46
(2/46)
3/50
(3/50)
118
34/49c
(34/47)
45/49c
(45/47)
114
31/48C
(31/48)
38/49c
(38/49)
9/49c
(9/49)
aAverage daily doses estimated by the researchers.
bStatistically significant trend (p < 0.05) by the Cochran-Armitage trend test.
Statistically significant difference from controls (p < 0.05) by Fisher's exact test.
dCensored data used for modeling are shown in parentheses; see text for description of censoring procedure.
Source: NTP(1989).
r\
Following statistical analysis (Fisher's exact and x tests), the responses in male mice
control groups between the tPCP and EC-7 groups were judged to be similar for both
hepatocellular and adrenal tumors. Additionally, the responses in female control mice for
hepatocellular, adrenal, and circulatory tumors were similar for the tPCP and EC-7 experiments.
Therefore, all dose-response analyses were conducted using combined controls.
5.4.3. Dose Adjustments and Extrapolation Method(s)
The U.S. EPA's Guidelines for Carcinogen Risk Assessment (2005a) recommend that the
method used to characterize and quantify cancer risk from a chemical is determined by what is
known about the mode of action of the carcinogen and the shape of the cancer dose-response
curve. The dose response is assumed to be linear in the lowest dose range when evidence
supports a genotoxic mode of action because of DNA reactivity, or if another mode of action is
applicable that is anticipated to be linear. A nonlinear approach is appropriate when there are
sufficient data to ascertain the mode of action and conclude that it is nonlinear (e.g., when the
carcinogenic action is secondary to another toxic effect that itself has a threshold). The linear
approach to low-dose extrapolation is taken for agents where the mode of action is uncertain
(U.S. EPA, 2005a).
167
-------
As discussed in Section 4.7.3, the available data indicate that multiple modes of
carcinogenic action are possible, but none have been defined sufficiently (e.g., key events for
carcinogenicity, temporal relationships) to inform the human relevance or low-dose extrapolation
for the carcinogenicity of PCP. Therefore, as recommended in the U.S. EPA's Guidelines for
Carcinogen Risk Assessment (2005a), "when the weight of evidence evaluation of all available
data are insufficient to establish the mode of action for a tumor site and when scientifically
plausible based on the available data, linear extrapolation is used as a default approach."
Accordingly, for the derivation of a quantitative estimate of cancer risk for ingested PCP, a linear
extrapolation was performed to determine the cancer slope factor.
The multistage model has been used by EPA in the vast majority of quantitative cancer
assessments because it is thought to reflect the multistage carcinogenic process and it fits a broad
array of dose-response patterns. Occasionally, the multistage model does not fit the available
data, in which case, alternatives should be considered. Alternatives include dropping higher
exposure groups if, for example, the responses plateau at the higher exposures and the potential
POD is in the range covered by the remaining exposure levels. Alternate models may be used if
dropping groups is not feasible. Use of this decision scheme has contributed to greater
consistency among cancer risk assessments. Consequently, the multistage model was the
primary tool considered for fitting the dose-response data and is given by:
P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)],
where:
P(d) = lifetime risk (probability) of cancer at dose d
q; = parameters estimated in fitting the model, i = 1, ..., k
The multistage model in BMDS (version 1.3.2) (U.S. EPA, 2004) was used for all model
fits, and complete results are shown in Appendix D. Adequate fits were obtained for each of the
data sets as assessed by the %2 goodness-of-fit statistic (p > 0.1). In one case, adrenal
pheochromocytomas for male mice exposed to EC-7, an adequate fit was achieved after dropping
the highest exposure group. The BMD modeling results and their 95% lower bounds (BMDLs)
derived from each endpoint for the individual data sets are summarized in Table 5-4.
168
-------
Table 5-4. Summary of BMD modeling for PCP cancer data in male and
female B6C3Fi mice
Test
material
tPCP
EC-7
Sex
M
M
F
F
M
M
F
F
F
Endpoint
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/
malignant pheochromocytoma
Hepatocellular adenoma/carcinoma
Hemangioma/hemangiosarcoma
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/
malignant pheochromocytoma
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/
malignant pheochromocytoma
Hemangioma/hemangiosarcoma
Model degree
One stage
One stage
One stage
One stage
One stage
Two stage
Two stage
Two stage
One stage
BMD10a
(mg/kg-d)
3.12
6.45
21.3
27.8
11.0
12.6
36.9
45.5
61.7
BMDL10b
(mg/kg-d)
2.27
4.47
11.7
16.3
7.59
5.75
16.4
29.6
37.9
aBMDs, calculated using polynomial multistage model of BMDS version 1.3.2, associated with a 10% extra risk.
bBMDL = 95% lower confidence limit on the BMD.
Source: NTP(1989).
A BW3/4 (body mass raised to the 3/4 power) scaling factor was used to convert the PODs
in the mouse study to human equivalent doses (HEDs), in accordance with the Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 2005a). This procedure presumes that equal doses in
these units (i.e., in mg/kg3/4-day), when administered daily over a lifetime, will result in equal
lifetime risks of the critical effect across mammalian species (U.S. EPA, 1992). The HED may
be calculated as follows (U.S. EPA, 2005a, 1992):
HED (mg/kg-day) = dose in animals (mg/kg-day) x (BWa/BWh)
where:
Dose = average daily dose in animal study
BWa = animal body weight (kg)
= reference human body weight (70 kg)
0.25
The time-weighted average body weights in the combined controls were used to represent
animal body weights in the above equation (0.037 kg for males and 0.038 kg for females). The
cross-species scaling factor of 0.15 was used to calculate the HEDs shown in Table 5-5.
169
-------
Table 5-5. Summary of BMDLio/HED and cancer slope factors derived from
PCP cancer data in male and female B6C3Fi mice
Test
material
tPCP
EC-7
Sex
M
M
F
F
M
M
F
F
F
Endpoint
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/malignant
pheochromocytoma
Hepatocellular adenoma/carcinoma
Hemangioma/hemangiosarcoma
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/malignant
pheochromocytoma
Hepatocellular adenoma/carcinoma
Adrenal pheochromocytoma/malignant
pheochromocytoma
Hemangioma/hemangiosarcoma
BMD10/HEDa
(mg/kg-d)
0.475
0.981
3.24
4.23
1.68
1.92
5.61
6.93
9.24
BMDL10/HEDa
(mg/kg-d)
0.35
0.68
1.79
2.48
1.15
0.88
2.50
4.51
5.76
Slope factorb
(mg/kg-d)1
2.9 x 10 -1
1.5 x 1Q-1
5.6 x 1Q-2
4.0 x 10'2
8.7 x 10'2
1.1 x 10'1
4.0 x 10'2
2.2 x 10'2
1.7 x 10'2
aBMD(L)HED = BMD(L) x BW3/4 scaling factor.
bCancer slope factor calculated by dividing the risk at the POD by the BMDLnED at the POD (0.1/BMDL10/HED).
Source: NTP(1989).
Alternatively, the cross-species scaling factor could have been applied to the individual
exposure levels for each dose-response analysis, prior to modeling. When the cross-species
factor is the same across groups, because of no appreciable difference in body weights in a data
set, it is numerically equivalent to apply the factor after modeling to the BMDs only, as in this
assessment.
5.4.4. Oral Slope Factor and Inhalation Unit Risk
A low-dose linear extrapolation approach results in calculation of an oral slope factor that
describes the cancer risk per unit dose of the chemical at low doses. The oral slope factors for
each data set considered were calculated by dividing the risk at the POD by the corresponding
BMDL (0.1/BMDLio/HEo). The site-specific oral slope factors are summarized in Table 5-5.
The slope factors ranged from 1.7 x 10"2 to 8.7 x 10"2 (mg/kg-day)"1 for EC-7 and from
4 x 10"2 to 2.9 x 10"1 (mg/kg-day)"1 for tPCP. The highest PCP cancer slope factor (2.9 x
10"1 (mg/kg-day)"1) resulted from the analysis of combined incidences for hepatocellular
adenomas and carcinomas in tPCP male mice. Considering the multiple tumor types and sites
observed in the mice exposed to PCP, the estimation of risk based on only one tumor type/site
may underestimate the overall carcinogenic potential of PCP.
The U.S. EPA's cancer guidelines (2005a, b) identify two ways to approach this issue—
analyzing the incidences of tumor-bearing animals or combining the potencies associated with
significantly elevated tumors at each site. The NRC (1994) concluded that an approach based on
170
-------
counts of animals with one or more tumors would tend to underestimate overall risk when tumor
types occur independently, and that an approach based on combining the risk estimates from
each separate tumor type should be used. The NRC (1994) recommended an approach based on
simulations. Therefore, a bootstrap analysis (Efron and Tibshirani, 1993) was used to derive the
distribution of the BMD for the combined risk of liver, adrenal gland, and circulatory system
tumors observed in male and female mice with oral exposure to PCP. A simulated incidence
level was generated for each exposure group using a binomial distribution with probability of
success estimated by a Bayesian estimate of probability. Each simulated data set was modeled
using the multistage model in the same manner as was done for the individual risks associated
with the liver, adrenal gland, and circulatory system tumors. The 5* percentile from the
distribution of combined BMDs was used to estimate the BMDL corresponding to an extra risk
of 1% for any of the three tumor sites. This analysis is described in greater detail in Appendix E
(see Table E-l).
The results of combining risks across sites within data sets are shown in Table 5-6. The
highest combined risk observed, similar to the individual cancer risk estimates, was in
tPCP-exposed male mice. The male mice were consistently more sensitive than female mice to
PCP tumor-induction. The 95% upper confidence limit (UCL) on the combined risk for male
mice that developed liver and/or adrenal gland tumors was 4.0 x 10"1 (mg/kg-day)"1, which is
about 38% higher than the 2.9 x 10"1 (mg/kg-day)"1 cancer slope factor estimated from liver
tumors only in tPCP-exposed male mice. The risk estimates for the tPCP-exposed males and
females tend to be higher than those for the EC-7-exposed animals, by approximately twofold for
the central tendency estimates and for the upper bound estimates. These differences suggest a
slightly greater potency for the technical grade.
171
-------
Table 5-6. Human-equivalent combined risk estimates for liver, adrenal,
and circulatory tumors in B6C3Fi mice
Sex
Endpoints
Human-equivalent combined risk
(mg/kg-d)3
Central tendency
Upper bound
tPCP
Male
Female
Hepatocellular adenoma/carcinoma or adrenal
pheochromocytoma/malignant
pheochromocytoma
Hepatocellular adenoma/carcinoma, adrenal
pheochromocytoma/malignant
pheochromocytoma, or hemangioma/
hemangiosarcoma
2.9 x 10'1
5.2 x ID'2
4.0 x 10'1
8.3 x ID'2
EC-7
Male
Female
Hepatocellular adenoma/carcinoma or adrenal
pheochromocytoma/malignant
pheochromocytoma
Hepatocellular adenoma/carcinoma, adrenal
pheochromocytoma/malignant
pheochromocytoma, or
hemangioma/hemangiosarcoma
1.1 x ID'1
2.8 x ID'2
1.7 x ID'1
4.8 x ID'2
"See the text and Appendix E for details of the derivation of combined risk estimates.
For oral exposure to tPCP and aPCP (pure PCP), the recommended slope factor is 4 x
10"1 (mg/kg-day)"1. This slope factor should not be used with exposures >0.3 mg/kg-day (the
POD for the site with the greatest response for tPCP-exposed male mice), because above this
point, the slope factor may not approximate the observed dose-response relationship adequately.
For oral exposure to EC-7, the recommended slope factor is 2 x 10"1 (mg/kg-day)"1. This
slope factor should not be used with exposures >1 mg/kg-day (the POD for the site with the
greatest response for EC-7-exposed male mice), because above this point, the slope factor may
not approximate the observed dose-response relationship adequately.
Several issues bear consideration before recommending a slope factor for oral exposure
to PCP only. A key issue concerns the impurities in the test materials and whether they
contribute to the carcinogenicity associated with PCP. Limited quantitative information is
available on the carcinogenic potential of the impurities in the formulations of PCP (tPCP and
EC-7) tested by NTP (1989). Based on the NTP (1989) calculations, the tPCP formulation is
comprised of approximately 90% PCP, 4% TCP, 6% chlorohydroxydiphenyl ethers, and trace
amounts of chlorinated dibenzodioxins and dibenzofurans. The EC-7 formulation is comprised
of approximately 91% PCP, 9% TCP, and trace amounts of chlorinated dibenzodioxins and
dibenzofurans. The oral slope factor of 4 x 10"1 (mg/kg-day)"1 for tPCP may be associated with
cancer risk from both PCP and its impurities. Available information addressing carcinogenicity
of the impurities varies widely, from a slope factor for hexachlorodibenzodioxins (U.S. EPA,
172
-------
1988) to no information regarding the carcinogenicity for most of the impurities.
Hexachlorodibenzodioxins comprise 0.001% of tPCP and 0.00002% of EC-7, about a 50-fold
difference. The most common impurity in both formulations, TCP, at 3.8% in tPCP and 9.4% in
EC-7, shows some evidence of carcinogenicity (see Section 4.1). Although the available data do
not support a quantitative risk estimate for TCP, the difference in potencies between the two
formulations (if there truly is one) does not suggest a role for TCP, since the difference in
potencies is in the opposite direction to the relative amounts of TCP in each formulation.
Estimation of bounding conditions may help in considering the possible impact of the
impurities. First, if any carcinogenic risk associated with each set of impurities is negligible
relative to that from PCP alone, then in order to use the estimated slope factor for a PCP-only
exposure, the slope factor should be adjusted to reflect that the exposure levels in the bioassay
were not completely PCP. That is, the slope factor would be multiplied by I/purity, or 1/0.9 =
1.1, an increase of 10%, because both formulations were approximately 90% PCP.
On the other hand, if the carcinogenic activity of the impurities is not negligible, then the
estimated risk attributable to PCP should be reduced. Starting with hexachlorodibenzodioxins,
T I r\
the slope factor was estimated at 6 x 10 (mg/kg-day)" (U.S. EPA, 1988 ). For an exposure
level of 1 mg/kg-day of tPCP, there would be 0.00001 mg/kg-day of hexachlorodibenzodioxins,
r\
for an estimated lifetime upper bound extra risk of 6 x 10" , about sevenfold lower than the
estimated lifetime risk using the slope factor for tPCP (4 x 10"1). Note that about seven
impurities are present in tPCP at higher levels than hexachlorodibenzodioxins. Similarly, at
1 mg/kg-day of EC-7, there would be 2 x 10"7 mg/kg-day of hexachlorodibenzodioxins, for an
estimated lifetime upper bound extra risk of 1.2 x 10"3, about 160-fold lower than the estimated
lifetime risk using the slope factor for EC-7 (2 x 10"1). Also note that about five other
chlorinated phenols, dioxins, and furans are present in EC-7 at higher levels than the
hexachlorodibenzodioxins. These risk comparisons are only approximate, but in view of the
other related chemicals present in these formulations without cancer assessments, they suggest
that the slope factors estimated from tPCP and EC-7 data are more relevant for exposures to
those formulations and less relevant for PCP alone or in mixtures other than tPCP and EC-7.
However, based on either low toxicity or the presence of minute quantities, the chlorinated
dibenzodioxins and dibenzofurans may contribute only slightly to the cancer risk associated with
tPCP.
Comparison of the two formulations identifies a common contaminant, TCP. It is
unlikely, based on the quantities present in both formulations of PCP, that TCP is largely
2The reported slope factor for hexachlorodibenzodioxins was a geometric mean of the slope factors for male mice
and female rats: female rat = 3.5 x 103 per mg/kg-day, male mouse = 1.1 x 104 per mg/kg-day. Using the more
sensitive response, and adjusting for the current interspecies scaling factor based on B W3/4 rather than B W2/3 (by
multiplying by (BWa/BWh)033/(BWa/BWh)025 = 0.083/0.152 = 0.54), an approximate slope factor for comparison
with the PCP slope factors is given by 1.1 x 104 per mg/kg-day x 0.54-6 x 103 per mg/kg-day, essentially the same
as the reported slope factor for hexachlorodibenzodioxins.
173
-------
responsible for the difference in the oral slope factors for tPCP and EC-7. The assumption that
TCP minimally contributes to the estimated cancer risk for EC-7 indicates that the oral slope
factor of 2 x 10"1 (mg/kg-day)"1 underestimates the risk associated with aPCP. It is possible that
the hydroxydiphenyl ether contaminants are responsible for the difference in cancer potency
between tPCP and EC-7; however, given the lack of information on these ethers contaminants,
their potential contribution to the carcinogenicity of tPCP cannot be characterized.
In summary, the presence of contaminants in the formulations of PCP tested by NTP
(1989), tPCP and EC-7, could have contributed to the carcinogenicity of the formulations.
However, the degree of influence of the contaminants on the cancer potency of either
formulation (e.g., whether these contaminants resulted in an over- or underestimation of the
potency of PCP alone) cannot be determined. Therefore, in the absence of information to
indicate the formulation which best represents PCP carcinogenicity, EPA selected the most
sensitive cancer risk estimate, the slope factor of 4 x 10"1 (mg/kg-day)"1 derived for tPCP, which
is the higher cancer potency of the two formulations, to represent the cancer risk estimate for
PCP.
An inhalation unit risk was not derived in this assessment. Data on the carcinogenicity of
the compound via the inhalation route is unavailable, and route-to-route extrapolation was not
possible due to the lack of a PBPK model.
5.4.5. Uncertainties in Cancer Risk Values
As in most risk assessments, extrapolation of the available experimental data for PCP to
estimate potential cancer risk in human populations introduces uncertainty in the risk estimation.
Several types of uncertainty may be considered quantitatively, whereas others can only be
addressed qualitatively. Thus, an overall integrated quantitative uncertainty analysis cannot be
developed. Major sources of uncertainty in the cancer assessment for PCP are summarized
below and in Table 5-7.
174
-------
Table 5-7. Summary of uncertainties in the PCP cancer risk assessment
Consideration/
approach
Overall carcinogenic
potential
Human relevance of
male mouse tumor
data
Bioassay
Dose metric
Low-dose
extrapolation
procedure
Cross-species scaling
Statistical
uncertainty at POD
Human population
variability in
metabolism and
response/sensitive
subpopulations
Impact on oral slope
factor
Slope factor could J,
by ~1. 4-fold if based
on most sensitive site
only
Human risk could J,
or t, depending on
relative sensitivity
Alternatives could t
or J, slope factor by
an unknown extent
Alternatives could t
or J, slope factor by
an unknown extent
Departure from the
U.S. EPA's
Guidelines for
Carcinogen Risk
Assessment POD
paradigm, if justified,
could | or t slope
factor an unknown
extent
Alternatives could J,
or t slope factor (e.g.,
3. 5 -fold J, [scaling by
BW] or t twofold
[scaling by BW2/3])
J, slope factor 1.4-fold
if a central tendency
estimate (i.e., BMD)
is used rather than
lower bound on POD
Low-dose risk t or J,
to an unknown extent
Decision
Combined risk,
across sites thought
to be independent
Liver and adrenal
gland tumors in
male mice are
relevant to human
exposure
NTP study
Used administered
exposure
Multistage model to
determine POD,
linear low-dose
extrapolation from
POD (default
approach)
BW3/4 (default
approach)
BMDL (default
approach for
calculating
reasonable upper
bound slope factor)
Considered
qualitatively
Justification
Basing risk on one site underestimates overall
risk when multiple tumor types occur.
There are no mode-of-action data to guide
extrapolation approach for any choice. It was
assumed that humans are as sensitive as the most
sensitive rodent gender/species tested; true
correspondence is unknown. The carcinogenic
response occurs across species. PCP is a
multisite carcinogen, although direct site
concordance is generally not assumed (U.S. EPA,
2005a); consistent with this view, some human
tumor types are not found in rodents.
Alternative bioassays were unavailable.
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites are not clearly identified.
Available mode-of-action data do not inform
selection of dose-response model; the linear
approach is applied in the absence of support for
an alternative.
There are no data to support alternatives.
Because the dose metric was not an AUC, BW3/4
scaling was used to calculate equivalent
cumulative exposures for estimating equivalent
human risks.
Limited size of bioassay results in sampling
variability; lower bound is 95% CI on
administered exposure.
No data to support range of human
variability/sensitivity, including whether children
are more sensitive.
175
-------
Overall carcinogenic potential. Considering the multiple tumor types and sites observed
in the mice exposed to PCP, the estimation of risk based on only one tumor type/site, even if the
most sensitive, may underestimate the overall carcinogenic potential of PCP. An approach based
on counts of animals with one or more tumors is expected to underestimate overall risk when
tumor types occur independently (NRC, 1994). The modes of action of the liver, adrenal gland,
and circulatory system tumors are unknown, so it cannot be verified whether or not these tumors
develop independently with PCP exposure. (Note that within sites, adenomas and carcinomas
were not assumed to be independent.) The NRC (1994) recommended a simulation approach for
combining the risk estimates from each separate tumor type in order to derive the distribution of
the BMD for the combined risk of liver, adrenal gland, or circulatory system tumors observed in
male and female mice with oral exposure to PCP. A bootstrap analysis (Efron and Tibshirani,
1993) was implemented for these data. For male mice, the overall unit risk was approximately
1.4-fold higher than that from liver tumors alone. If there is some dependency between the sites
considered, then the overall carcinogenic potential would be somewhat reduced.
Relevance to humans. Tumors in animals are considered relevant to humans unless data
are available to indicate otherwise. Furthermore, site concordance between animals and humans
is not necessary to consider a tumor observed in animals to be relevant to humans. As noted in
U.S. EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), "... agents observed
to produce tumors in both humans and animals have produced tumors either at the same site
(e.g., vinyl chloride) or different sites (e.g., benzene) (NRC, 1994). Hence, site concordance is
not always assumed between animals and humans." These assumptions related to human
relevance of animal tumors introduce some degree of uncertainty into the assessment; the
magnitude of that uncertainty is unknown.
The tumor types used to derive the slope factor for PCP were male mouse liver tumors
and adrenal gland tumors (pheochromocytomas). Data are limited and preclude the
characterization of the mode of action by which PCP exerts its carcinogenic effect in the liver.
Oxidative stress may play a role in the carcinogenicity of PCP observed in mice. Indicators of
oxidative stress that were observed in animal studies with PCP have also been identified in
human cancers. There is some evidence in humans (sawmill workers) for hepatic cancer
associated with PCP exposure (Demers et al., 2006). The relevance of the mode of action of
liver tumor induction is discussed in Section 4.7.3.
The relevance of rodent pheochromocytomas as a model for human cancer risk has been
the subject of discussion in the scientific literature (e.g., Greim et al., 2009; Powers et al., 2008).
In humans, pheochromocytomas are rare catecholamine-producing neuroendocrine tumors that
are usually benign, but may also present as or develop into a malignancy (Eisenhofer et al., 2004;
Lehnert et al., 2004; Edstrom-Elder et al., 2003; Goldstein et al., 1999). Hereditary factors in
humans have been identified as important in the development of pheochromocytomas
(Eisenhofer et al., 2004). Parallels between human and mouse pheochromocytomas have led
176
-------
investigators to conclude that the mouse might be an appropriate model for human adrenal
medullary tumors (Tischler et al., 1996). Like the human, pheochromocytomas in the mouse are
relatively rare, as are metastases. The lifetime incidence in wild-type laboratory mice has been
reported as <3% by Tischler et al. (2004, 1996). Both the morphological variability of the mouse
pheochromocytomas and the morphology of the predominant cells are comparable to those of
human pheochromocytomas. An important characteristic of mouse pheochromocytomas is
expression of immunoreactive phenylethanolamine-N-methyltransferase (PNMT); human
pheochromocytomas are also usually PNMT-positive (Tischler et al., 1996). The mode of action
for the induction of adrenal gland tumors (pheochromocytomas and malignant
pheochromocytomas) in mice by PCP is unknown. The available experimental evidence for
pheochromocytomas more generally, however, supports a conclusion that mouse
pheochromocytomas are relevant to humans.
Bioassay selection. The study by NTP (1989) was used for development of an oral slope
factor. This was a well-designed study, conducted in both sexes of B6C3Fi mice with
50 animals/sex/dose group, which is typical for carcinogenicity studies. Test animals were
allocated among two dose levels of tPCP and three dose levels of EC-7 and an untreated control
group for each formulation. Animals were observed twice daily and examined weekly (for 12-
13 weeks) for body weight and monthly for feed consumption. Animals were necropsied and all
organs and tissues were examined grossly and microscopically for histopathological lesions for a
full set of toxicological endpoints in both sexes. Alternative bioassays for quantitative analysis
were unavailable. Overall responses across the sexes of the two grades of PCP were similarly
robust, although the responses tended to be greater in those animals treated with tPCP than those
treated with EC-7.
Choice of species/gender. The oral slope factor for PCP was quantified using the tumor
incidence data for male mice, which were judged to be more sensitive than female mice to the
carcinogenicity of PCP. The male rat tumor incidence data, while demonstrating some evidence
of carcinogenicity, were not utilized for deriving low-dose quantitative risk estimates. The
responses of increased incidence of mesothelioma and nasal squamous cell carcinoma in male
rats were lower than those of the mice (NTP, 1989) at a greater exposure level, suggesting
greater sensitivity of the mice. Moreover, the toxicological database for PCP studies in rodents
has shown the mouse model, rather than the rat, to be a more sensitive model of PCP
hepatotoxicity. Although the NTP (1999) bioassay in rats administered aPCP reported
mesotheliomas and nasal squamous cell carcinomas, the tumors occurred in male rats of multiple
dose groups, but only in the high dose (1-year exposure) was the tumor incidence statistically
significant. The lack of a significant dose-response trend in the rat data and the observation of
consistently greater sensitivity to PCP in mice, rather than rats, led to the use of the mouse data,
specifically the male mouse data (relatively most sensitive), for the derivation of the slope factor.
Consequently, dose-response modeling was not carried out with the rat tumor data.
177
-------
Dose metric. PCP is metabolized to hydroquinone and benzoquinone metabolites;
however, it is unknown whether a metabolite or some combination of parent compound and
metabolites is responsible for the observed toxicity of PCP. If the actual carcinogenic moiety is
proportional to administered exposure, then use of administered exposure as the dose metric
provides an unbiased estimate of carcinogenicity. On the other hand, if this is not the correct
dose metric, then the impact on the slope factor is unknown.
Choice of low-dose extrapolation approach. The mode of action is a key consideration in
clarifying how risks should be estimated for low-dose exposure. A linear low-dose extrapolation
approach was used as a default to estimate human carcinogenic risk associated with PCP
exposure due to the limited availability of data to determine the mode of carcinogenic action of
PCP. The extent to which the overall uncertainty in low-dose risk estimation could be reduced if
the mode of action for PCP were known is of interest, but the mode of action is not known.
Etiologically different tumor types were not combined across sites prior to modeling, in
order to allow for the possibility that different tumor types can have different dose-response
relationships because of varying time courses or other underlying mechanisms or factors. The
human equivalent oral slope factors estimated from the tumor sites with statistically significant
increases ranged from 0.017 to 0.29 per mg/kg-day, a range less than two orders of magnitude,
with the greater risk coming from the male mice tPCP data.
However, given the multiplicity of tumor sites, basing the oral slope factor on one tumor
site may underestimate the carcinogenic potential of PCP. Following the recommendations of
the NRC (1994) and the U.S. EPA's Guidelines for Carcinogen Risk Assessment (2005a), an
approach based on combining the risk estimates from each separate tumor type was used. Total
carcinogenic risk was estimated using a bootstrap analysis (Efron and Tibshirani, 1993; see
Section 5.3) to derive the distribution of the BMD for the combined risk of liver and adrenal
gland tumors observed in male mice and the combined risk of liver, adrenal gland, and
circulatory system tumors observed in female mice with oral exposure to PCP. Note that this
estimate of overall risk describes the risk of developing any combination of the tumor types
considered, not just the risk of developing all three simultaneously. The highest combined risk
observed, similar to the individual cancer risk estimates, was in tPCP-exposed male mice. The
95% UCL on the combined risk for male mice that developed liver and/or adrenal gland tumors
was 4.0 x 10"1 (mg/kg-day)"1, which is about 38% higher than the 2.9 x 10"1 (mg/kg-day)"1 cancer
slope factor estimated from liver tumors only in tPCP-exposed male mice.
Choice of model. In the absence of mechanistic data for biologically based low-dose
modeling or mechanistic evidence to inform the low-dose extrapolation (see the discussion at the
beginning of Section 5.4.3), a linear low-dose extrapolation was carried out from the BMDLio.
It is expected that this approach provides an upper bound on low-dose cancer risk for humans.
The true low-dose risks cannot be known without additional data.
178
-------
With respect to uncertainties in the dose-response modeling, the two-step approach of
modeling only in the observable range (U.S. EPA, 2005a) and extrapolating from a POD in the
observable range is designed in part to minimize model dependence. Furthermore, the
multistage model used provided an adequate fit to all the data sets. The ratio of the BMDio
values to the BMDLio values give some indication of the uncertainties in the dose-response
modeling. The ratio between BMDs and BMDLs is typically <2 when modeling cancer data
(i.e., NTP or other bioassay data with about 50 animals/group). This ratio characterizes the
experimental variability inherent in the data. For the tumor sites evaluated for PCP, this ratio
was <1.8, indicating that the estimated risk is not influenced by any unusual variability relative
to other assessments. Additional uncertainty introduced by estimating combined risks reflecting
multiple tumor sites is related to the assumption of independence of tumors. As NRC (1994)
stated: "...a general assumption of statistical independence of tumor-type occurrences within
animals is not likely to introduce substantial error in assessing carcinogenic potency," so that the
uncertainty related to combining tumors is considered small.
Cross-species scaling. An adjustment for cross-species scaling (BW3 4) was applied to
address toxicological equivalence of internal doses between mice and humans, consistent with
the 2005 Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). It is assumed that
equal risks result from equivalent constant lifetime exposures.
Human population variability. Neither the extent of interindividual variability in PCP
metabolism nor human variability in response to PCP has been characterized. Factors that could
contribute to a range of human response to PCP include variations in CYP450 levels because of
age-related differences or other factors (e.g., exposure to other chemicals that induce or inhibit
microsomal enzymes), nutritional status, alcohol consumption, or the presence of underlying
disease that could alter metabolism of PCP or antioxidant protection systems. Incomplete
understanding of the potential differences in metabolism and susceptibility across exposed
human populations represents a major source of uncertainty.
5.4.6. Previous Cancer Assessment
The previous cancer assessment, posted to the IRIS database in March 1991, included an
oral slope factor of 1.2 x 10"1 (mg/kg-day)"1. While also based on the NTP (1989) study that
currently serves as the basis for the quantitative cancer assessment, the previous oral slope factor
was derived using the pooled incidence of tumors in female mice (now thought to underestimate
total risk), the linearized multistage procedure, a cross-species scaling factor based on BW2/3
(resulting in a twofold higher risk than current methods), and a geometric mean of the slope
factors associated with each formulation of PCP, tPCP, and EC-7 (tending toward the lower
slope factor of those estimated). The incidence of tumors in the female mice, rather than the
males, was used to derive an oral slope factor because hemangiomas and hemangiosarcomas
were observed in females. The male mice did not exhibit a significant increase in incidence of
179
-------
hemangiomas and hemagiosarcomas. The hemangiosarcomas were judged to be the tumor of
greatest concern because they are morphologically related to known fatal human cancers that are
induced by xenobiotics. Based on a preference for the data on hemangiosarcomas and because
some male groups experienced significant early loss (observed in male controls in the tPCP study
and in male mice in the mid-dose group in the EC-7 study, although the current analysis has
shown a lack of significant effect resulting from the early loss in these groups), only the female
mice were used in the quantitative risk assessment.
180
-------
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF
HAZARD AND DOSE RESPONSE
6.1. HUMAN HAZARD POTENTIAL
6.1.1. Noncancer
PCP is a nonflammable, noncorrosive chemical that was first registered in the United
States in 1936 as a wood preservative to prevent decay from fungal organisms and insect
damage. It was widely used as a biocide and could also be found in ropes, paints, adhesives,
canvas, insulation, and brick walls. After use was restricted in 1984, PCP applications were
limited to utilization in industrial areas, including utility poles, cross arms, railroad cross-ties,
wooden pilings, fence posts, and lumber/timbers for construction. Currently, products
containing PCP remain registered for wood preservation, and utility poles and cross arms
represent approximately 92% of all uses for PCP-treated lumber.
During manufacture of PCP, the chemical is contaminated with impurities that consist of
several congeners of the chlorophenols, chlorinated dibenzo-p-dioxins, and chlorinated
dibenzofurans. Of the chlorinated dibenzo-p-dioxin and dibenzofuran contaminants, the higher
chlorinated congeners are predominantly found as impurities within tPCP (approximately 90%
purity). Use of the aPCP first requires a purification process to remove the contaminants that are
simultaneously created during the manufacturing of PCP.
Instances of PCP poisoning have been documented, indicating the potentially severe
consequences of acute, high-dose exposures. Few studies have examined the effects of the lower
exposures that occurred in occupational settings or through residential or environmental sources.
Many of the available studies are relatively small (<50 participants) (Peper et al., 1999; Triebig
et al., 1987; Klemmer et al., 1980; Begley et al., 1977) or may not be representative of the
exposed population (Gerhard et al., 1999; Walls et al., 1998). Despite these limitations, there are
indications of specific types of neurobehavioral effects seen with chronic exposure to PCP in
nonoccupational settings (Peper et al., 1999). A larger study of 293 former sawmill workers in
New Zealand also suggests neuropsychological effects and respiratory diseases (McLean et al.,
2009b). In addition, the results from a large nested cohort study of reproductive outcomes in
offspring of sawmill workers (Dimich-Ward et al., 1996) indicate that specific types of birth
defects warrant additional research.
The toxicity of PCP in orally exposed animals was investigated in numerous studies in
experimental animals. These studies indicate that PCP is toxic to the liver. In chronic studies in
rats and dogs, liver toxicity was characterized primarily by increased incidence of chronic
inflammation, cytoplasmic vacuolization, pigmentation, and hepatocellular necrosis as well as
changes in liver weight (NTP, 1999; Mecler, 1996; Schwetz et al., 1978). Liver toxicity in mice
was exhibited as necrosis, cytomegaly, chronic active inflammation, pigmentation, and bile duct
181
-------
lesions (NTP, 1989). The increased severity of liver toxicity observed in mice versus rats could
be based, in part, on differences in biotransformation of PCP (Lin et al., 1997), but it is also
noted that in the mouse studies, the PCP test material contained higher concentrations of
chlorinated dibenzo-p-dioxin or dibenzofuran contaminants, which could contribute to the
severity of the liver response. Liver toxicity in the dog (Mecler, 1996) was similar to that of the
mouse, but the doses inducing toxicity were lower than those in the mouse (i.e., 1.5 mg/kg-day in
the dog versus 17-18 mg/kg-day in the mouse). Studies using domestic or farm animals showed
that pigs, but not cattle, exhibited similar liver toxicity as that observed in mice. Pigment
deposition was also observed in the proximal convoluted tubules in the kidneys of rats (NTP,
1999). Developmental toxicity studies (Welsh et al., 1987; Schwetz et al., 1974a) indicated toxic
effects in offspring at dose levels below those producing maternal toxicity. Studies in mink
indicate some reproductive effects following exposure to PCP (Cook et al., 1997). The spleen
weights of mice (NTP, 1989), rats (Bernard et al., 2002), and cattle (Hughes et al., 1985) were
decreased following exposure to PCP.
Disruption of thyroid homeostasis has been observed following the administration of
PCP. Several studies have reported decreased serum T4 and TS levels in rats (Jekat et al., 1994)
and cattle (Hughes et al., 1985; McConnell et al., 1980). Decreases in serum T4 have been
observed in ram and ewe lambs (Beard et al., 1999a, b), mature ewes (Rawlings et al., 1998), and
mink (Beard and Rawlings, 1998) after administration of PCP. TSH was unaffected by treatment
with 1 mg/kg-day PCP in calves (Hughes et al., 1985) and sheep (Beard et al., 1999b). However,
Jekat et al. (1994) reported a decrease in TSH accompanying the decrease in T4 levels in rats
administered 3 mg/kg-day tPCP and aPCP. Considering that TSH acts on the thyroid to control
production of T4, the concurrent decrease in TSH is in contrast to the expected TSH response to a
decrease in T4 (TSH is generally expected to increase in response to a decrease in T4), which led
Jekat et al. (1994) to suggest that this was due to interference with thyroid hormone regulation at
the hypothalamic/pituitary level and possibly increased peripheral thyroid hormone metabolism.
However, the available data do not allow for determination of the mechanism involved in the
effects on TS, T4, and TSH following exposure to PCP. The effect of PCP on thyroid hormone
homeostasis has been attributed to PCP and not to contaminants. Changes in thyroid hormones
have been associated with effects (i.e., delayed myelination, neuronal proliferation, and synapse
formation) on neurons. Considering that thyroid hormones may play a role in
neurodevelopmental processes, the disruption of thyroid homeostasis that has been observed with
PCP indicates a potential concern for critical period of development of the nervous system
(CalEPA, 2006). However, the downstream effects associated with PCP and decreased T4 levels
have not been explored.
Studies examining the immunotoxic effects of PCP showed that the humoral response
and complement activity in mice were impaired by tPCP, but not by aPCP, when administered to
adult animals (NTP, 1989; Holsapple et al., 1987; Kerkvliet et al., 1985a, b; 1982a). However,
182
-------
treatment of mice with aPCP from the time of conception to 13 weeks of age resulted in impaired
humoral and cell-mediated immunity (Exon and Roller, 1983), suggesting that PCP, and not just
the contaminants, induce immunotoxicity. Human studies showed that immune response was
impaired in patients who had blood PCP levels >10 ug/L and in particular in those whose levels
were >20 ug/L (Daniel et al., 1995; McConnachie and Zahalsky, 1991). Based on the limited
available information, immunotoxic effects of PCP may be elicited, in part, through the presence
of the dioxin/furan contaminants within PCP.
In vitro neurotoxicity studies showed that PCP causes a dose-dependent irreversible
reduction in endplate potential at the neuromuscular junction and interferes with axonal
conduction in the sciatic nerve from the toad (Montoya and Quevedo, 1990; Montoya et al.,
1988). An NTP (1989) study in mice showed only decreased motor activity in rotarod
performance in male rats treated with tPCP for 5 weeks and increases in motor activity and
startle response in females receiving purified and tPCP for 26 weeks. Another in vivo study
showed that treatment of rats with PCP for up to 14 weeks caused biochemical changes in the rat
brain (Savolainen and Pekari, 1979). The most definitive study showed that rats receiving PCP
in drinking water for at least 90 days had marked morphological changes in sciatic nerves
(Villena et al., 1992).
Elevated blood sugar levels (considered minor by Demidenko, 1969) and increases in
organ weights were observed in rats and rabbits exposed to 21-29 mg/m3 PCP by inhalation for
4 months (Ning et al., 1984; Demidenko, 1969). Additional effects included anemia,
leukocytosis, eosinophilia, hyperglycemia, and dystrophic processes in the liver. Minor effects
were noted on the liver, cholinesterase activity, and blood sugar effects of animals exposed to
2.97 mg/m3 (calculated as 0.3 mg/kg-day PCP by Kunde and Bohme [1978]), a dose that is lower
than the lowest NOAELs (1 mg/kg-day) observed in animals orally exposed to 28.9 mg/m3 PCP
(Demidenko, 1969). Ning et al. (1984) reported significant increases in organ weights (lung,
liver, kidney, and adrenal glands), serum y-globulin, and blood-glucose levels at 21.4 mg/m3.
Studies examining the mutagenicity of PCP have shown that in a variety of test systems,
PCP is nonmutagenic, with the exception of one study (Gopalaswamy and Nair, 1992) in which
PCP exhibited a positive response for mutagenicity in the Ames Salmonella assay. In contrast to
data on PCP, data for the TCHQ metabolite of PCP show positive mutagenic effects in CHO
cells (Jansson and Jansson, 1991; Carstens et al., 1990; Ehrlich, 1990), an increase in
micronuclei using V79 cells (Jansson and Jansson, 1992), covalent binding to DNA (Witte et al.,
2000, 1985), and induction of DNA SSBs (Witte et al., 1985).
6.1.2. Cancer
The available epidemiologic studies support an association between PCP exposure and
development of specific cancers: non-Hodgkin's lymphoma, multiple myeloma, soft tissue
sarcoma, and liver cancer (limited evidence). These studies used PCP-specific exposure
183
-------
assessment and in some cases, additional assessment of other chlorophenols and potential
contaminants. PCP preparations are produced with methods that allow for the formation of
contaminants, and degradation products occur naturally in most formulations. However, these
contaminants are unlikely to spuriously produce the observed associations seen in the
epidemiologic studies, given the difference in the patterns of cancer risk seen in studies of
dioxins compared with the studies of PCP, and the relative strengths of the effects of different
chemicals (PCP, other chlorophenols, dioxins, and furans) in the studies that examined more than
one of these chemicals. It should be noted that in the epidemiological studies examining the
cancer risk associated with exposure to PCP, exposures occurred predominantly via the
inhalation and dermal routes.
Animal studies with PCP show evidence of adrenal medullary and hepatocellular tumors
in male and female mice, hemangiosarcomas and hemangiomas in female mice, and nasal
squamous cell carcinomas and mesotheliomas in male rats. Two well-conducted studies provide
data for the carcinogenicity of PCP via the oral route in laboratory animals: one study in
B6C3Fi mice (NTP, 1989) and another study in F344 rats (NTP, 1999). Two formulations of
PCP (tPCP and EC-7) were carcinogenic in the mouse. Hepatocellular adenomas/carcinomas
and adrenal medullary pheochromocytomas developed in male mice treated with tPCP or EC-7,
and hepatocellular adenomas/carcinomas and hemangiosarcomas developed in female mice
treated with tPCP or EC-7 and adrenal medullary pheochromocytomas developed in female mice
treated with EC-7.
Under the U.S. EPA's Guidelines for Carcinogen Risk Assessment (2005a), PCP is
characterized as "likely to be carcinogenic to humans" by all routes of exposure.
6.2. DOSE RESPONSE
6.2.1. Noncancer—Oral Exposure
The most sensitive endpoints identified for effects of PCP by oral exposure relate to liver
toxicity in the chronic gelatin capsule study Mecler (1996) in beagle dogs. Mecler (1996) was
selected for the derivation of the oral RfD. This study was conducted in accordance with good
laboratory practice guidelines valid at that time and included both sexes of beagle dogs, four
animals per sex and dose group, and three dose groups plus controls (0, 1.5, 3.5, and 6.5 mg/kg-
day). The study reported multiple toxic endpoints, including changes in absolute and relative
organ weights, changes in hematological parameters, and histopathologic outcomes.
Hepatotoxicity characterized by dose-related increases in incidence and severity of hepatic
lesions (including liver pigmentation, cytoplasmic vacuolation, chronic inflammation, and the
appearance of dark, discolored livers) accompanied by significant increases in absolute (in
females only) and relative liver weight, and serum activity of ALT and ALP in dogs was
considered the critical effect. Another target of PCP toxicity following oral exposure considered
in the selection of the critical effect was the developing organism. Studies in experimental
184
-------
animals found that PCP exposure during gestation can produce prenatal loss, skeletal variations,
visceral malformations, decreased fetal weight, and delayed puberty; these doses also produced
toxic effects in the dams. However, PCP doses associated with liver toxicity were lower than
those associated with developmental toxicity.
Dose-response data of Mecler (1996) were evaluated by using the NOAEL/LOAEL
approach with an increase in the incidence of hepatic effects identified as the critical effect. The
POD was 1.5 mg/kg-day, the LOAEL. A composite UF of 300 was applied to derive the oral
RfD of 5 x 10"3 mg/kg-day. The composite UF of 300 consists of an interspecies UF of 10 for
extrapolation from animals to humans, an intraspecies UF of 10 to adjust for sensitive human
subpopulations, and a UF of 3 to account for the use of a LOAEL instead of a NOAEL.
Confidence in the principal study, Mecler (1996), is medium. The 52-week study in
beagle dogs is an unpublished, Office of Pollution, Prevention and Toxic Substances (OPPTS)
guideline study that used three dose groups plus a control and collected interim data at 13, 26,
and 39 weeks. The study is limited by the use of relatively small group sizes (4 dogs/sex/dose).
Because the incidence of two of the key liver effects (i.e., hepatocellular pigmentation in males
and females and chronic inflammation in males) increased from 0% in the controls to 100% in
the lowest dose tested, and remained at 100% in both the mid- and high-dose groups, the study
provided limited resolution of the dose-response curve at low doses. However, liver effects
observed in this study (i.e., the critical effect for the RfD) are well-supported by other oral
subchronic and chronic studies. PCP also induced toxicity in reproductive and immunological
studies, but at doses higher than those used in the principal study. Confidence in the database is
high because the database includes acute, short-term, subchronic, and chronic toxicity studies
and developmental and multigenerational reproductive toxicity studies in multiple species, and
carcinogenicity studies in two species. Overall confidence in the RfD is medium.
6.2.2. Cancer
The NTP (1989) mouse study was selected for dose-response assessment based on
statistically significant increased incidence of hepatocellular adenomas and carcinomas and
adrenal pheochromocytomas and malignant pheochromocytomas in male and female mice and
hemangiomas and hemangiosarcomas (in liver and spleen) in female mice. The study was used
for development of an oral slope factor. This was a well-designed study, conducted in both sexes
of B6C3Fi mice with two formulations of PCP (tPCP and EC-7) and with 50 mice/sex/dose.
Test animals were allocated among two dose levels for tPCP and three dose levels for EC-7 and
untreated control groups for each formulation. Animals were observed twice daily and examined
weekly (for 12-13 weeks) and then monthly for body weight and monthly for feed consumption.
Animals were necropsied and all organs and tissues were examined grossly and microscopically
for histopathological lesions for a full set of toxicological endpoints in both sexes. Tumor
185
-------
incidences were elevated with increasing exposure level at multiple sites in both sexes, including
the liver, adrenal gland, and circulatory system.
The male F344 rat tumor incidence data (NTP, 1999), while demonstrating some
evidence of carcinogenicity, were not utilized for deriving low-dose quantitative risk estimates,
based on evidence of greater sensitivity of the mice to PCP.
A linear approach was applied in the dose-response assessment for PCP, in which the
mode of action is uncertain, consistent with U.S. EPA's Guidelines for Carcinogen Risk
Assessment (2005a). The guidelines recommend the use of a linear extrapolation as a default
approach when the available data are insufficient to establish a mode of action for a tumor site.
As discussed in Section 4.7.3, the mechanism leading to the formation of liver, adrenal, and
circulatory tumors in mice following PCP ingestion is unknown. There is some evidence of
oxidative damage to cells and DNA adducts from prominent reactive metabolites, and some
evidence of cytotoxicity observed in animal and in vitro studies; however, these data do not
allow for the identification of key events or support a mode of carcinogenic action. Therefore, a
linear extrapolation was used to derive the cancer slope factor for ingested PCP.
Increased incidence of hepatocellular adenomas and carcinomas, benign and malignant
adrenal medullary tumors, and hemangiomas and hemangiosarcomas in a 2-year mice bioassay
(NTP, 1989) served as the basis for the oral cancer dose-response analysis. A multistage model
using linear extrapolation from the POD (combined risk estimates based on increased incidence
of both hepatocellular and adrenal gland tumors in male mice) was performed to derive an oral
slope factor of 4 x 10"1 (mg/kg-day)"1 for PCP. The recommended slope factor should not be
used with exposures >0.3 mg/kg-day (POD for the site with the greatest response for
tPCP-exposed male mice), because above this point, the slope factor may not approximate the
observed dose-response relationship adequately.
Extrapolation of the experimental data to estimate potential cancer risk in human
populations introduces uncertainty in the risk estimation for PCP. Uncertainty can be considered
quantitatively; however, some uncertainty can only be addressed qualitatively. For this reason,
an overall integrated quantitative uncertainty analysis cannot be developed. However, a major
uncertainty considered was the observation of multiple tumor types and sites in the mice exposed
to PCP. Risk estimated using only one tumor type/site, even if the most sensitive, may
underestimate the overall carcinogenic potential of PCP. Therefore, an upper bound on
combined risk was derived in order to gain some understanding of the overall risk resulting from
tumors occurring at multiple sites. A bootstrap analysis (Efron and Tibshirani, 1993) was used
to derive the distribution of the BMD for the combined risk of liver and adrenal gland tumors
observed in male rats with oral exposure to PCP. A simulated incidence level was generated for
each exposure group using a binomial distribution with probability of success estimated by a
Bayesian estimate of probability. Each simulated data set was modeled using the multistage
model in the same manner as was done for the individual risks associated with the liver, adrenal
186
-------
gland, and circulatory system tumors. The 5th percentile from the distribution of combined
BMDs was used to estimate the BMDL corresponding to an extra risk of 1% for any of the three
tumor sites. The results of combining risks across sites within data sets are shown in Table 5-6.
The highest combined risk observed, similar to the individual cancer risk estimates, was in tPCP-
exposed male mice. The 95% UCL on the combined risk for animals that developed liver and/or
adrenal gland tumors was 4.0 x 10"1 (mg/kg-day)"1, which is about 38% higher than the 2.9 x
10"1 (mg/kg-day)"1 cancer slope factor estimated from liver tumors only in tPCP-exposed male
mice. The risk estimates for the tPCP-exposed males and females tend to be higher than those
for the EC-7-exposed animals, by approximately twofold for both the central tendency and upper
bound estimates.
A biologically-based model was not supported by the available data; therefore, a
multistage model was the preferred model. The multistage model can accommodate a wide
variety of dose-response shapes and provides consistency with previous quantitative dose-
response assessments for cancer. Linear low-dose extrapolation from a POD determined by an
empirical fit of tumor data has been judged to lead to plausible upper bound risk estimates at low
doses for several reasons. However, it is unknown how well this model or the linear low-dose
extrapolation predicts low-dose risks for PCP. An adjustment for cross-species scaling (BW3/4)
was applied to address toxicological equivalence of internal doses between mice and humans
based on the assumption that equal risks result from equivalent constant lifetime exposures.
An inhalation unit risk was not derived in this assessment. Data on the carcinogenicity of
the compound via the inhalation route is unavailable, and route-to-route extrapolation was not
possible due to the lack of a PBPK model. However, it is proposed that PCP is likely to be
carcinogenic to humans by the inhalation route since the compound is well-absorbed, and in oral
studies induces tumors at sites other than the portal of entry.
187
-------
7. REFERENCES
Ahlborg, UG; Thunberg, TM. (1980) Chlorinated phenols: occurrence, toxicity, metabolism, and environmental
impact. CRC Crit Rev Toxicol 7(1): 1-35.
Ahlborg, UG; Lindgren, JE; Mercier, M. (1974) Metabolism of pentachlorophenol. Arch Toxicol 32:271-281.
Ali, R; Yu, CL; Wu, MT; et al. (2004) A case-control study of parental occupation, leukemia, and brain tumors in
an industrial city in Taiwan. J Occup Environ Med 46:985-992.
Allan, R. (1994) Phenols and phenolic compounds. In: Clayton, G; Clayton, F; eds. Patty's industrial hygiene and
toxicology. 4th edition. Vol. 2, pt. B. New York, NY: John Wiley and Sons, pp. 1567-1630.
Amirimani, B; Weber, B; Rebbeck, T. (2000) Regulation of reporter gene expression by a CYP3A4 promoter
variant in primary human hepatocytes. Proc Am Assoc Cancer Res 41:18. (abstract)
Armstrong, RW; Eichner, ER; Klein, DE; et al. (1969) Pentachlorophenol poisoning in a nursery for newborn
infants. II. Epidemiologic and toxicologic studies. J Pediatr 75:317-325.
Arrhenius, E; Renberg, L; Johansson, L. (1977a) Subcellular distribution, a factor in risk evaluation of
pentachlorophenol. Chem Biol Interact 18:23-34.
Arrhenius, E; Renberg, L; Johansson, L; et al. (1977b) Disturbance of microsomal detoxication mechanisms in liver
by chlorophenol pesticides. Chem Biol Interact 18:35-46.
ATSDR (Agency for Toxic Substances and Disease Registry). (2001) Toxicological profile for pentachlorophenol
(CAS # 87-86-5). Public Health Service, U.S. Department of Health and Human Services, Atlanta, GA. Available
online at http://www.atsdr.cdc.gov/toxprofiles/index.asp.
Bauchinger, M; Dresp, J; Schmid, E; et al. (1982) Chromosome changes in lymphocytes after occupational exposure
to pentachlorophenol (PCP). MutatRes 102:83-88.
Beard, AP; Rawlings, NC. (1998) Reproductive effects in mink (Mustela vison) exposed to the pesticides lindane,
carbofuran and pentachlorophenol in a multigeneration study. J Reprod Fertil 113:95-104.
Beard, AP; McRae, AC; Rawlings, NC. (1997) Reproductive efficiency in mink (Mustela vison) treated with the
pesticides lindane, carbofuran and pentachlorophenol. J Reprod Fertil 111:21-28.
Beard, AP; Bartlewski, PM; Rawlings, NC. (1999a) Endocrine and reproductive function in ewes exposed to the
organochlorine pesticides lindane or pentachlorophenol. J Toxicol Environ Health A 56:23-46.
Beard, AP; Bartlewski, PM; Chandolia, RK; et al. (1999b) Reproductive and endocrine function in rams exposed to
the organochlorine pesticides lindane and pentachlorophenol from conception. J Reprod Fertil 115:303-314.
Begley, J; Reichert, EL; Rashad, MN; et al. (1977) Association between renal function tests and pentachlorophenol
exposure. Clin Toxicol 11:97-106.
Bernard, BK; Hoberman, AM. (2001) A study of the developmental toxicity potential of pentachlorophenol in the
rat. IntJ Toxicol 20:353-362.
Bernard, BK; Ranpuria, AK; Hoberman, AM. (2001) Developmental toxicity study of pentachlorophenol in the
rabbit. IntJ Toxicol 20:345-352.
Bernard, BK; Hoberman, AM; Brown, WR; et al. (2002) Oral (gavage) two-generation (one litter per generation)
reproduction study of pentachlorophenol (penta) in rats. Int J Toxicol 21:301-318.
188
-------
Bevenue, A; Wilson, J; Casarett, L; et al. (1967) A survey of pentachlorophenol content in human urine. Bull
Environ Contam Toxicol 2:319-332.
Bishop, CM; Jones, AH. (1981) Non-Hodgkin's lymphoma of the scalp in workers exposed to dioxins. Lancet
2:369.
Blakley, BR; Yole, MJ; Brousseau, P; et al. (1998) Effect of pentachlorophenol on immune function. Toxicology
125:141-148.
Bodell, WJ; Pathak, DN. (1998) Detection of DNA adducts in B6C3F1 mice treated with pentachlorophenol. Proc
Am Assoc Cancer Res 39:332. (abstract)
Bond, GG; McLaren, EA; Brenner, FE; et al. (1989) Incidence of chloracne among chemical workers potentially
exposed to chlorinated dioxins. J Occup Med 31(9):771-774.
Braun, WH; Sauerhoff, MW. (1976) The pharmacokinetic profile of pentachlorophenol in monkeys. Toxicol Appl
Pharmacol 38:525-533.
Braun, WH; Young, JD; Blau, GE; et al. (1977) The pharmacokinetics and metabolism of pentachlorophenol in rats.
Toxicol Appl Pharmacol 41:395-406.
Braun, WH; Blau, GE; Chenoweth, M. (1979) The metabolism/pharmacokinetics of pentachlorophenol in man, and
a comparison with the rat and monkey. In: Deichman, W; ed. Toxicology and occupational medicine: proceedings.
Developments in toxicology and environmental science. Vol. 4. New York, NY: Elsevier, pp. 289-296.
Brinton, LA; Blot, WJ; Stone, BJ; et al. (1977) A death certificate analysis of nasal cancer among furniture workers
in North Carolina. Cancer Res 37:3473-3474.
BRL (Bionetics Research Laboratories, Inc.). (1968) Evaluation of carcinogenic, teratogenic, and mutagenic
activities of selected pesticides and industrial chemicals. Vol. 1. Carcinogenic study. Study performed by Bionetics
Research Labs, Inc., Bethesda, MD, for the National Cancer Institute, Bethesda, MD. Available from the National
Technical Information Service, Springfield, VA; PB223159.
Budavari, S; O'Neil, M; Smith, A; et al. (1996) The Merck index: an encyclopedia of chemicals, drugs, and
biological. 12th edition. Whitehouse Station, NJ: Merck & Co., Inc., p. 1222.
Buselmaier, W; Rohrborn, G; Propping, P. (1973) Comparative investigations on the mutagenicity of pesticides in
mammalian test systems. Mutat Res 21:25-26.
CalEPA (California Environmental Protection Agency). (2006) Development of health criteria for school site risk
assessment pursuant to health and safety code section 901(g): proposed child-specific reference dose (chRfD) for
school site risk assessment. Manganese and pentachlorophenol. Office of Environmental Health Hazard
Assessment, California Environmental Protection Agency. Available online at
www. oehha. ca. gov/public_info/public/kids/index.html.
Carstens, CP; Blum, JK; Witte, I. (1990) The role of hydroxyl radicals in tetrachlorohydroquinone induced DNA
strand break formation in PM2 DNA and human fibroblasts. Chem Biol Interact 74:305-314.
Casarett, LJ; Bevenue, A; Yauger, WL, Jr; et al. (1969) Observations on pentachlorophenol in human blood and
urine. Am Ind Hyg Assoc J 30:360-366.
CDC (Centers for Disease Control and Prevention). (1980) Epidemiologic notes and reports: pentachlorophenol in
log homes-Kentucky. MMWR 29(36):431-437.
Chang, WC; Jeng, JH; Shieh, CC; et al. (2003) Skin tumor-promoting potential and systemic effects of
pentachlorophenol and its major metabolite tetrachlorohydroquinone in CD-I mice. Mol Carcinog 36:161-170.
Cheng, WN; Coenraads, PJ; Hao, ZH; et al. (1993) A health survey of workers in the pentachlorophenol section of a
chemical manufacturing plant. Am J Ind Med 24:81-92.
189
-------
Chhabra, RS; Maronpot, RM; Bucher, JR; et al. (1999) Toxicology and carcinogenesis studies of pentachlorophenol
in rats. Toxicol Sci 48:14-20.
Cline, RE; Hill, RH, Jr; Phillips, DL; et al. (1989) Pentachlorophenol measurements in body fluids of people in log
homes and workplaces. Arch Environ Contam Toxicol 18:475-481.
Cole, GW; Stone, O; Gates, D; et al. (1986) Chloracne from pentachlorophenol-preserved wood. Contact Dermatitis
15:164-168.
Collins, JJ; Bodner, KM; Wilken, M; et al. (2007) Serum concentrations of chlorinated dibenzo-p-dioxins and
dibenzofurans among former Michigan trichlorophenol and pentachlorophenol workers. J Expo Sci Environ
Epidemiol 17:541-548.
Cook, SJ; Beard, AP; McRae, AC; et al. (1997) Fertility in mink (Mustela vison) exposed to pesticides from
conception. Biol Reprod 56(Suppl 1):200.
Cooper, RG; Macauley, MB. (1982) Pentachlorophenol pancreatitis. Lancet 1:517.
Dahlhaus, M; Almstadt, E; Appel, KE. (1994) The pentachlorophenol metabolite tetrachloro-p-hydroquinone
induces the formation of 8-hydroxy-2-deoxyguanosine in liver DNA of male B6C3F1 mice. Toxicol Lett 74: 265-
274.
Dahlhaus, M; Almstadt, E; Henschke, P; et al. (1995) Induction of 8-hydroxy-2-deoxyguanosine and single-strand
breaks in DNA of V79 cells by tetrachloro-p-hydroquinone. Mutat Res 329:29-36.
Dahlhaus, M; Almstadt, E; Henschke, P; et al. (1996) Oxidative DNA lesions in V79 cells mediated by
pentachlorophenol metabolites. Arch Toxicol 70:457^4-60.
Dai, J; Wright, MW; Manderville, RA. (2003) An oxygen-bonded c8-deoxyguanosine nucleoside adduct of
pentachlorophenol by peroxidase activation: evidence for ambident c8 reactivity by phenoxyl radicals. Chem Res
Toxicol 16:817-821.
Dai, J; Sloat, AL; Wright, MW; et al. (2005) Role of phenoxyl radicals in DNA adduction by chlorophenol
xenobiotics following peroxidase activation. Chem Res Toxicol 18:771-779.
Daimon, H; Sawada, S; Asakura, S; et al. (1997) Inhibition of sulfotransferase affecting in vivo genotoxicity and
DNA adducts induced by safrole in rat liver. Teratog Carcinog Mutagen 17:327-337.
Daniel, V; Huber, W; Bauer, K; et al. (1995) Impaired in vitro lymphocyte responses in patients with elevated
pentachlorophenol (PCP) blood levels. Arch Environ Health 50:287-292.
Deichman, W; Machle, W; Kitzmiller, K; et al. (1942) Acute and chronic effects of pentachlorophenol and sodium
pentachlorophenate upon experimental animals. J Pharmacol Exp Ther 76:104-117.
Demers, PA; Davies, HW; Friesen, MC; et al. (2006) Cancer and occupational exposure to pentachlorophenol and
tetrachlorophenol. Cancer Causes Control 17(6):749-758.
Demidenko, MM. (1969) Materials for establishing the maximum permissible concentration in air. Gig Tr Prof
Zabol 7:58-60. (Russian)
Den Besten, C; Vet, JJRM; Besselink, HT; et al. (1991) The liver, kidney, and thyroid toxicity of chlorinated
benzenes. Toxicol Appl Pharmacol 111 :69-81.
De Roos, AJ; Hartge, P; Lubin, JH; et al. (2005) Persistent organochlorine chemicals in plasma and risk of non-
Hodgkin's lymphoma. Cancer Res 65(23):11214-11226.
Dimich-Ward, H; Hertzman, C; Teschke, K; et al. (1996) Reproductive effects of paternal exposure to chlorophenate
wood preservatives in the sawmill industry. Scand J Work Environ Health 22:267-273.
190
-------
Donnelly, KC; Claxton, LD; Huebner, HJ; et al. (1998) Mutagenic interactions of model chemical mixtures.
Chemosphere 37:1253-1261.
Dome, JL. (2004) Impact of inter-individual differences in drug metabolism and pharmacokinetics on safety
evaluation. Fundam Clin Pharmacol 18:609-620.
Dome, JL; Walton, K; Renwick, AG. (2005) Human variability in xenobiotic metabolism and pathway-related
uncertainty factors for chemical risk assessment: a review. Food Chem Toxicol 43:203-216.
Dorsey, WC; Tchounwou, PB. (2003) CYP1A1, HSP70, p53, and c-fos expression in human liver carcinoma cells
(HepG2) exposed to pentachlorophenol. Biomed Sci Instrum 39:389-396.
Dorsey, WC; Tchounwou, PB; Sutton, D. (2004) Mitogenic and cytotoxic effects of pentachlorophenol to AML 12
mouse hepatocytes. Int J Environ Res Pub Health 1(2): 100-105.
Dubois, M; Plaisance, H; Thome, JP; et al. (1996) Hierarchical cluster analysis of environmental pollutants through
P450 induction in cultured hepatic cells. Ecotoxicol Environ Saf 34:205-215.
Edgerton, TR; Moseman, RF; Linder, RE; et al. (1979) Multiresidue method for the determination of chlorinated
phenol metabolites in urine. J Chromatogr 170:331-342.
Edstrom Elder, E; Hjelm Skog, AL; Hoog, A; et al. (2003) The management of benign and malignant
pheochromocytoma and abdominal paraganglioma. Eur J Surg Oncol 29:278-283.
Efron, B; Tibshirani, RJ. (1993) An introduction to the bootstrap. San Francisco, CA: Chapman and Hall.
Ehrlich, W. (1990) The effect of pentachlorophenol and its metabolite tetrachlorohydroquinone on cell growth and
the induction of DNA damage in Chinese hamster ovary cells. Mutat Res 244:299-302.
Eisenhofer, G; Bornstein, SR; Brouwers, FM; et al. (2004) Malignant pheochromocytoma: current status and
initiatives for future progress. Endocr-Relat Cancer 11:423^36.
Engst, R; Macholz, RM; Kujawa, M; et al. (1976) The metabolism of lindane and its metabolites
gamma-2,3,4,5,6-pentachlorocyclohexene, pentachlorobenzene, and pentachlorophenol in rats and the pathways of
lindane metabolism. J Environ Sci Health B 11:95-l 17.
Eriksson, M; Hardell, L; Berg, NO; et al. (1981) Soft-tissue sarcomas and exposure to chemical substances: a case-
referent study. Br J Ind Med 3 8:27-3 3.
Eriksson, M; Hardell, L; Adami, HO. (1990) Exposure to dioxins as a risk factor for soft tissue sarcoma: a
population-based case-control study. J Natl Cancer Inst 82:486-490.
Exon, JH; Koller, LD. (1983) Effects of chlorinated phenols on immunity in rats. Int J Immunopharmacol 5:131-
136.
Fahrig, R. (1974) Comparative mutagencity studies with pesticides. In: Montesano, R; Tomatis, L; Davis, W; eds.
Chemical carcinogenesis essays. IARC scientific publications. No. 10. Lyon, France: International Agency for
Research on Cancer, pp. 161-178.
Fahrig, R; Nilsson, C; Rappe, C. (1978) Genetic activity of chlorophenols and chlorophenol impurities. In: Rao, K;
ed. Pentachlorophenol: chemistry, pharmacology, and environmental toxicology. New York and London: Plenum,
pp. 325-338.
Forsell, JH; Shull, LR; Kateley, JR. (1981) Subchronic administration of technical pentachlorophenol to lactating
dairy cattle: immunotoxicologic evaluation. J Toxicol Environ Health 8:543-558.
Freire, P.F; Labrador, V; Perez Martin, JM; et al. (2005) Cytotoxic effects in mammalian Vero cells exposed to
pentachlorophenol. Toxicology 210:37-44.
191
-------
Friesen, MC; Davies, HW; Teschke, K; et al. (2007) Impact of the specificity of the exposure metric on exposure-
response relationships. Epidemiology 18(l):88-94.
Galloway, SM; Armstrong, MJ; Reuben, C; et al. (1987) Chromosome aberrations and sister chromatid exchanges in
Chinese hamster ovary cells: evaluations of 108 chemicals. Environ Mol Mutagen 10(Suppl 10): 1-175.
Garabedian, MJ; Hoppin, JA; Tolbert, PE; et al. (1999) Occupational chlorophenol exposure and non-Hodgkin's
lymphoma. J Occup Environ Med 41 (4): 267-272.
Gerhard, I; Frick, A; Monga, B; et al. (1999) Pentachlorophenol exposure in women with gynecological and
endocrine dysfunction. Environ Res 80:383-388.
Gilbert, FI, Jr; Minn, CE; Duncan, RC; et al. (1990) Effects of pentachlorophenol and other chemical preservatives
on the health of wood-treating workers in Hawaii. Arch Environ Contam Toxicol 19:603-609.
Goldstein, RE; O'Neill, JA, Jr; Holcomb, GW, III; et al. (1999) Clinical experience over 48 years with
pheochromocytoma. Ann Surg 229:755-766.
Gomez-Catalan, J; To-Figueras, J; Rodamilans, M; etal. (1991) Transport of organochlorine residues in the rat and
human blood. Arch Environ Contam Toxicol 20:61-66.
Gopalaswamy, UV; Nair, CK. (1992) DNA binding and mutagenicity of lindane and its metabolites. Bull Environ
Contam Toxicol 49:300-305.
Gravance, CG; Garner, DL; Miller, MG; et al. (2003) Flow cytometric assessment of changes in rat sperm
mitochondrial function after treatment with pentachlorophenol. Toxicol In Vitro 17:253-257.
Gray, RE; Gilliland, RD; Smith, EE; et al. (1985) Pentachlorophenol intoxication: report of a fatal case, with
comments on the clinical course and pathologic anatomy. Arch Environ Health 40:161-164.
Greene, MH; Brinton, LA; Fraumeni, JF; et al. (1978) Familial and sporadic Hodgkin's disease associated with
occupational wood exposure. Lancet 2:626-627.
Greichus, YA; Libal, GW; Johnson, DD. (1979) Diagnosis and physiologic effects of pentachlorophenols on young
pigs. Part I. Effects of purified pentachlorophenol. Bull Environ Contam Toxicol 23:418^-22.
Hardell, L; Eriksson, M. (1988) The association between soft tissue sarcomas and exposure to phenoxyacetic acids.
A new case-referent study. Cancer 62(3):652-656.
Hardell, L; Sandstrom, A. (1979) Case-control study: soft-tissue sarcomas and exposure to phenoxyacetic acids or
chlorophenols. Br J Cancer 39(6):711-717.
Hardell, L; Eriksson, M; Lenner, P; et al. (1981) Malignant lymphoma and exposure to chemicals, especially organic
solvents, chlorophenols and phenoxy acids: a case-control study. Br J Cancer 43:169-176.
Hardell, L; Eriksson, M; Degerman, A. (1994) Exposure to phenoxyacetic acids, chlorophenols, or organic solvents
in relation to histopathology, stage, and anatomical localization of non-Hodgkin's lymphoma. Cancer Res 54:2386-
2389.
Hardell, L; Eriksson, M; Degerman, A. (1995) Meta-analysis of four Swedish case-control studies on exposure to
pesticides as risk-factor for soft-tissue sarcoma including the relation to tumour localization and histopathological
type. IntJOncol 6:847-851.
Haworth, S; Lawlor, T; Mortelmans, K; et al. (1983) Salmonella mutagenicity test results for 250 chemicals.
Environ Mutagen 5(Suppl 1): 1-142.
He, P; Court MH; Greenblatt, DJ; et al. (2005) Genotype-phenotype associations of cytochrome P450 3A4 and 3A5
polymorphism with midazolam clearance in vivo. Clin Pharmacol Ther 77:373-387.
192
-------
Heacock, H; Hertzman, C; Demers, PA; et al. (2000) Childhood cancer in the offspring of male sawmill workers
occupationally exposed to chlorophenate fungicides. Environ Health Perspect 108(6):499-503.
Hertzman, C; Teschke, K; Dimich-Ward, H; et al. (1988) Validity and reliability of a method for retrospective
evaluation of chlorophenate exposure in the lumber industry. Am JIndMed 14(6):703-713.
Hertzman, C; Teschke, K; Ostry, A; et al. (1997) Mortality and cancer incidence among sawmill workers exposed to
chlorophenate wood preservatives. Am J Public Health 87:71-79.
Hill, RH, Jr; Head, SL; Baker, S; et al. (1995) Pesticide residues in urine of adults living in the United States:
reference range concentrations. Environ Res 71:99-108.
Hoben, HJ; Ching, SA; Casarett, LJ. (1976a) A study of inhalation of pentachlorophenol by rats. IV. Distribution
and excretion of inhaled pentachlorophenol. Bull Environ Contam Toxicol 15:466-474.
Hoben, HJ; Ching, SA; Casarett, LJ. (1976b) A study of inhalation of pentachlorophenol by rats. III. Inhalation
toxicity study. Bull Environ Contam Toxicol 15:463-465.
Holsapple, MP; McNerney, PJ; McCay, JA. (1987) Effects of pentachlorophenol on the in vitro and in vivo antibody
response. J Toxicol Environ Health 20:229-239.
Hooiveld, M; Heederik, DJJ; Kogevinas, M; et al. (1998) Second follow-up of a Dutch cohort occupationally
exposed to phenoxy herbicides, chlorophenols, and contaminants. Am J Epidemiol 147(9):891-899.
Hoppin, JA; Tolbert, PE; Herrick, RF; et al. (1998) Occupational chlorophenol exposure and soft tissue sarcoma risk
among men aged 30-60 years. Am J Epidemiol 148(7):693-703.
Hsieh, KP; Lin, YY; Cheng, CL; et al. (2001) Novel mutations of CYP3A4 in Chinese. Drug Metab Dispos
29:268-273.
Hughes, BJ; Forsell, JH; Sleight, SD; et al. (1985) Assessment of pentachlorophenol toxicity in newborn calves:
clinicopathology and tissue residues. J Anim Sci 61:1587-1603.
IARC (International Agency for Research on Cancer). (1999) IARC monographs on the evaluation of carcinogenic
risk of chemicals to humans. Vol. 71. Polychlorophenols and their sodium salts (Group 2B). Lyon, France:
International Agency for Research on Cancer, p. 769. Available online at
http://www.inchem.org/documents/iarc/vol71/028-polychloroph.html.
Igisu, H; Hamasaki, N; Ikeda, M. (1993) Highly cooperative inhibition of acetylcholinesterase by pentachlorophenol
in human erythrocytes. Biochem Pharmacol 46:175-177.
Jakobson, I; Yllner, S. (1971) Metabolism of 14C-pentachlorophenol in the mouse. Acta Pharmacol Toxicol
29:513-524.
Jansson, K; Jansson, V. (1986) Inability of chlorophenols to induce 6-thioguanine-resistant mutants in V79 Chinese
hamster cells. MutatRes 171:165-168.
Jansson, K; Jansson, V. (1991) Induction of mutation in V79 Chinese hamster cells by tetrachlorohydroquinone, a
metabolite of pentachlorophenol. Mutat Res 260:83-87.
Jansson, K; Jansson, V. (1992) Induction of micronuclei in V79 Chinese hamster cells by tetrachlorohydroquinone,
a metabolite of pentachlorophenol. Mutat Res 279:205-208.
Jappinen, P; Pukkala, E; Tola, S. (1989) Cancer incidence of workers in a Finnish sawmill. Scand J Work Environ
Health 15:18-23.
Jekat, FW; Meisel, ML; Eckard, R; et al. (1994) Effects of pentachlorophenol (PCP) on the pituitary and thyroidal
hormone regulation in the rat. Toxicol Lett 71:9-25.
193
-------
Jirasek, L; Kalensky, J; Kubec, K; et al. (1976) Chloracne, porphyria cutanea tarda, and other poisonings due to the
herbicides. Hautarzt 27:328-333.
Johnson, RL; Gehring, PJ; Kociba, RJ; et al. (1973) Chlorinated dibenzodioxins and pentachlorophenol. Environ
Health Perspect 5:171-175.
Johnson, CC; Feingold, M; Tilley, B. (1990) A meta-analysis of exposure to phenoxy acid herbicides and
chlorophenols in relation to risk of soft tissue sarcoma. Int Arch Occup Environ Health 62:513-520.
Jones, RD; Winter, DP; Cooper, AJ. (1986) Absorption study of pentachlorophenol in persons working with wood
preservatives. HumToxicol 5:189-194.
Juhl, U; Witte, I; Butte, W. (1985) Metabolism of pentachlorophenol to tetrachlorohydroquinone by human liver
homogenate. Bull Environ Contam Toxicol 35:596-601.
Kadlubar, FF; Berkowitz, GS; Delongchamp, RR; et al. (2003) The CYP3A41B variant is related to the onset of
puberty, a known risk factor for the development of breast cancer. Cancer Epidemiol Biomarkers Prev 12:327-331.
Kalman, DA. (1984) Determination of pentachlorophenol and 2,3,4,6-tetrachlorophenol in human urine by high
resolution gas chromatography. J Chromotogr Sci 22:452^55.
Karmaus, W; Wolf, N. (1995) Reduced birthweight and length in the offspring of females exposed to PCDFs, PCP,
and lindane. Environ Health Perspect 103:1120-1125.
Kerkvliet, NI; Baecher-Steppan, L; Claycomb, AT; et al. (1982a) Immunotoxicity of technical pentachlorophenol
(PCP-T): depressed humoral immune responses to T-dependent and T-independent antigen stimulation in PCP-T
exposed mice. Fundam Appl Toxicol 2:90-99.
Kerkvliet, NI; Baecher-Steppan, L; Schmitz, JA. (1982b) Immunotoxicity of pentachlorophenol (PCP): increased
susceptibility to tumor growth in adult mice fed technical PCP-contaminated diets. Toxicol Appl Pharmacol 62:55-
64.
Kerkvliet, NI; Brauner, JA; Matlock, JP. (1985a) Humoral immunotoxicity of polychlorinated diphenyl ethers,
phenoxyphenols, dioxins and furans present as contaminants of technical grade pentachlorophenol. Toxicology
36:307-324.
Kerkvliet, NI; Brauner, JA; Baecher-Steppan, L. (1985b) Effects of dietary technical pentachlorophenol exposure on
T cell, macrophage and natural killer cell activity in C57B1/6 mice. Int J Immunopharm 7:239-247.
Kimbrough, RD; Linder, RE. (1975) The effect of technical and 99% pure pentachlorophenol on the rat liver. Light
microscopy and ultrastructure. Toxicol Appl Pharmacol 33:131-132.
Kimbrough, RD; Linder, RE. (1978) The effect of technical and purified pentachlorophenol on the rat liver. Toxicol
Appl Pharmacol 46:151-162.
Kinzell, JH; Ames, NK; Sleight, SD; et al. (1981) Subchronic administration of technical pentachlorophenol to
lactating dairy cattle: performance, general health, and pathologic changes. J Dairy Sci 64:42-51.
Klaunig, JE; Kamendulis, LM. (2004) The role of oxidative stress in carcinogenesis. Annu Rev Pharmacol Toxicol
44:239-267.
Klaunig, JE; Yong, X; Isenberg, JS; et al. (1998) The role of oxidative stress in chemical carcinogenesis. Environ
Health Perspect 106(l):289-295.
Klemmer, HW. (1972) Human health and pesticides—community pesticide studies. Residue Rev 41:55-63.
Klemmer, HW; Wong, L; Sato, MM; et al. (1980) Clinical findings in workers exposed to pentachlorophenol. Arch
Environ Contam Toxicol 9:715-725.
194
-------
Knudsen, I; Verschuuren, HG; den Tonkelaar, EM; et al. (1974) Short-term toxicity of pentachlorophenol in rats.
Toxicology 2:141-152.
Kogevinas, M; Saracci, R; Bertazzi, P; et al. (1992) Cancer mortality study from soft-tissue sarcoma and malignant
lymphomas in an international cohort of workers exposed to chlorophenoxy herbicides and chlorophenols.
Chemosphere 25(7-10):1071-1076.
Kogevinas, M; Kauppinen, T; Winkelmann, R; et al. (1995) Soft tissue sarcoma and non-Hodgkin's lymphoma in
workers exposed to phenoxy herbicides, chlorophenols, and dioxins: two nested case-control studies. Epidemiology
6(4):396^K)2.
Kogevinas, M; Becher, H; Benn, T; et al. (1997) Cancer mortality in workers exposed to phenoxy herbicides,
chlorophenols, and dioxins: an expanded and updated international cohort study. Am J Epidemiol 145(12):1061-
1075.
Kohli, J; Jones, D; Safe, S. (1976) The metabolism of higher chlorinated benzene isomers. Can J Biochem 54:203-
208.
Krutovskikh, V; Yamasaki, H. (1997) The role of gap junctional intercellular communication (GJIC) disorders in
experimental and human carcinogenesis. Histol Histopathol 12:761-768.
Kunde, M; Bohme, G. (1978) The toxicology of pentachlorophenol: a review (Zur toxikologie des
pentachlorphenols: eine ubersicht). Bundesgesundhbl 21(19-20):302-310. (German)
Kutz, FW; Cook, BT; Carter-Pokras, OD; et al. (1992) Selected pesticide residues and metabolites in urine from a
survey of the U.S. general population. J Toxicol Environ Health 37:277-291.
La, D; Lin, P; Swenberg, J. (1998a) Analysis of DNA adducts in rats chronically exposed to pentachlorophenol.
Proc Am Assoc Cancer Res 39:330. (abstract)
La, D; Lin, P; Swenberg, J. (1998b) Analysis of DNA adducts in rodent livers after chronic administration of
pentachlorophenol. Proc Am Assoc Cancer Res 39:2253. (abstract)
Lambert, J; Schepens, P; Janssens, J; et al. (1986) Skin lesions as a sign of subacute pentachlorophenol intoxication.
Acta Derm Venereol 66:170-172.
Lampi, P; Hakulinen, T; Luostarinen, T; et al. (1992) Cancer incidence following chlorophenol exposure in a
community in southern Finland (Jarvela). Arch Environ Health 47(3): 167-175.
Larsen, RV; Kirsch, LE; Shaw, SM; et al. (1972) Excretion and tissue distribution of uniformly labeled
14C-pentachlorophenol in rats. J Pharm Sci 61:2004-2006.
Larsen, RV; Born, GS; Kessler, WV; et al. (1975) Placenta! transfer and teratology of pentachlorophenol in rats.
EnvironLett 10:121-128.
Leeder, JS; Kearns, GL. (1997) Pharmacogenetics in pediatrics. Implications for practice. Pediatr Clin North Am
44:55-77.
Lehnert, H; Mundschenk, J; Halm, K. (2004) Malignant pheochromocytoma. Front Horm Res 31:155-162.
Lin, PH; Waidyanatha, S; Pollack, GM; et al. (1997) Dosimetry of chlorinated quinone metabolites of
pentachlorophenol in the livers of rats and mice based upon measurement of protein adducts. Toxicol Appl
Pharmacol 145:399-408.
Lin, PH; Waidyanatha, S; Pollack, GM; et al. (1999) Dose-specific production of chlorinated quinone and
semiquinone adducts in rodent livers following administration of pentachlorophenol. Toxicol Sci 47:126-133.
195
-------
Lin, PH; Nakamura, J; Yamaguchi, S; et al. (2001) Oxidative damage and direct adducts in calf thymus DNA
induced by the pentachlorophenol metabolites, tetrachlorohydroquinone and tetrachloro-l,4-benzoquinone.
Carcinogenesis 22:627-634.
Lin, PH; La, DK; Upton, PB; et al. (2002) Analysis of DNA adducts in rats exposed to pentachlorophenol.
Carcinogenesis 23:365-369.
Loft, S; Meller, P. (2006) Oxidative DNA damage and human cancer: need for cohort studies. Antioxid Redox
Signal 8(5-6):1021-1031.
Lohmann, K; Prohl, A; Schwarz, E. (1996) Multiple chemical sensitivity disorder in patients with neurotoxic
illnesses. Gesundheitswesen 58:322-331.
Masini, A; Ceccarelli-Stanzani, D; Trenti, T; et al. (1984a) Transmembrane potential of liver mitochondria from
hexachlorobenzene- and iron-treated rats. Biochim Biophys Acta 802:253-258.
Masini, A; Ceccarelli-Stanzani, D; Trenti, T; et al. (1984b) Structural and functional properties of rat liver
mitochondria in hexachlorobenzene induced experimental porphyria. Biochem Biophys Res Commun 118:356-363.
Masini, A; Ceccarelli-Stanzani, D; Tomasi, A; et al. (1985) The role of pentachlorophenol in causing mitochondria!
derangement in hexachlorobenzene induced experimental porphyria. Biochem Pharmacol 34:1171-1174.
McConnachie, PR; Zahalsky, AC. (1991) Immunological consequences of exposure to pentachlorophenol. Arch
Environ Health 46:249-253.
McConnell, EE; Moore, JA; Gupta, BN; et al. (1980) The chronic toxicity of technical and analytical
pentachlorophenol in cattle. I. Clinicopathology. Toxicol Appl Pharmacol 52:468-490.
McConnell, EE; Huff, JE; Hejtmancik, M; et al. (1991) Toxicology and Carcinogenesis studies of two grades of
pentachlorophenol in B6C3Fi mice. Fundam Appl Toxicol 17:519-532.
Mecler, F. (1996) Fifty-two week repeated dose chronic oral study of pentachlorophenol administered via capsule to
dogs. Study conducted by TSI Mason Laboratories, Worcester, MA; TSI Report #ML-PTF-J31-95-94. Submitted to
the Pentachlorophenol Task Force, c/o SRA International, Inc., Washington, DC. U.S. Environmental Protection
Agency, Washington, DC; MRID 439827-01. Unpublished report3.
Mehmood, Z; Williamson, MP; Kelly, DE; et al. (1996) Metabolism of organochlorine pesticides: the role of human
cytochrome P450 3A4. Chemosphere 33:759-769.
Mikoczy, Z; Schutz, A; Stromberg, U; et al. (1996) Cancer incidence and specific occupational exposures in the
Swedish leather tanning industry: a cohort based case-control study. Occup Environ Med 53(7):463-467.
Montoya, GA; Quevedo, L. (1990) The effects of pentachlorophenol (PCP) at the toad neuromuscular junction.
Comp Biochem Physiol C 96:193-197.
Montoya, GA; Roa, J; Cruz, F; et al. (1988) The actions of phenol and pentachlorophenol (PCP) on axonal
conduction, ganglionic synaptic transmission, and the effect of pH changes. Comp Biochem Physiol C 89:377-382.
Naito, S; Ono, Y; Somiya, I; et al. (1994) Role of active oxygen species in DNA damage by pentachlorophenol
metabolites. Mutat Res 310:79-88.
NAS (National Academy of Science). (2006) Health risks from dioxin and related compounds: evaluation of the
EPA reassessment. Washington, DC: National Academy Press.
3This study was submitted to the Agency as part of the process for the development of the RED document by the
U.S. EPA's OPP. Mecler (1996) satisfied the guideline requirements (OPPTS 870.4100) for a chronic toxicity study
in non-rodents and is classified as an "acceptable" GLP study.
196
-------
Ning, HS; Zehn, HQ; Li, LS; et al. (1984) Study of the toxicity of pentachlorophenol and recommendations of the
maximum allowable concentration in air. J Commun Ind Hyg (Rail Transp Syst) 4:7-16. (Chinese)
NLM (National Library of Medicine). (1999a) Pentachlorophenol. HSDB (Hazardous Substances Data Bank).
National Institutes of Health, U.S. Department of Health and Human Services, Bethesda, MD. Available online at
http://toxnet.nlm.nih.gov (accessed 1999).
NLM. (1999b) Pentachlorophenol. RTECS (Registry of Toxic Effects of Hazardous Chemicals). National Institutes
of Health, U.S. Department of Health and Human Services, Bethesda, MD. Available from National Institute for
Safety and Health, Cincinnati, OH.
NLM. (2006) Pentachlorophenol. HSDB (Hazardous Substances Data Bank). National Institutes of Health, U.S.
Department of Health and Human Services, Bethesda, MD. Available online at http://toxnet.nlm.nih.gov (accessed
2006).
Norris, J. (1972) Acute lexicological properties of XD-8108.00L. Antimicrobial. Study conducted by Toxicology
Research Laboratory, Health and Environmental Research, Dow Chemical, Midland, MI. Submitted to the U.S.
Environmental Protection Agency, Washington, DC; MRID 00101715. Unpublished report.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC. (1994) Science and judgment in risk assessment. Washington, DC: National Academy Press.
NTP (National Toxicology Program). (1989) NTP toxicology and carcinogenesis studies of two pentachlorophenol
technical-grade mixtures (CAS No. 87-86-5) inB6C3F! mice (feed studies). Public Health Service, U.S.
Department of Health and Human Services; NTP TR 349. Available from the National Institute of Environmental
Health Sciences, Research Triangle Park, NC.
NTP. (1999) NTP toxicology and carcinogenesis studies of pentachlorophenol (CAS NO. 87-86-5) inF344/N rats
(feed studies). Public Health Service, U.S. Department of Health and Human Services; NTP TR 483. Available
from the National Institute of Environmental Health Sciences, Research Triangle Park, NC.
O'Malley, MA; Carpenter, AV; Sweeney, MH; et al. (1990) Chloracne associated with employment in the
production of pentachlorophenol. Am J Ind Med 17:411-421.
Osheroff, M; Horvath, C; Abrams, K; et al. (1994) Ninety-one day repeated dose dermal toxicity study of
pentachlorophenol in Sprague-Dawley rats. Study conducted by TSI Mason Labs for the Pentachlorophenol Task
Force c/o SRA International, Inc., Washington, DC; Study No 2-J37. Submitted to the U.S. Environmental
Protection Agency, Washington, DC; MRID 43182301. Unpublished report.
Ott, MG; Olsen, RA; Cook, RR; et al. (1987) Cohort mortality study of chemical workers with potential exposure to
the higher chlorinated dioxins. J Occup Med 29(5):422-429.
Ott, MG; Messerer, P; Zober, A. (1993) Assessment of past occupational exposure to 2,3,7,8-tetrachlorodibenzo-p-
dioxin using blood lipid analyses. Int Arch Occup Environ Health 65(1): 1-8.
Parke, DV; loannides, C. (1990) Role of cytochromes P-450 in mouse liver tumor production. Prog Clin Biol Res
331:215-230.
Pearce, NE; Smith, AH; Fisher, DO. (1985) Malignant lymphoma and multiple myeloma linked with agricultural
occupations in a New Zealand cancer registry-based study. Am J Epidemiol 121:225-237.
Pearce, NE; Smith, AH; Howard, JK; et al. (1986a) Case-control study of multiple myeloma and farming. Br J
Cancer 54(3):493-500.
Pearce, NE; Smith, AH; Howard, JK; et al. (1986b) Non-Hodgkin's lymphoma and exposure to phenoxyherbicides,
chlorophenols, fencing work, and meat works employment: a case-control study. Br J Ind Med 43(2):75-83.
197
-------
Peper, M; Ertl, M; Gerhard, I. (1999) Long-term exposure to wood-preserving chemicals containing
pentachlorophenol and lindane is related to neurobehavioral performance in women. Am J Ind Med 35:632-641.
Pitot, HC; Goldsworthy, T; Campbell, HA; et al. (1980) Quantitative evaluation of the promotion by 2,3,7,8-
tetrachlorodibenzo-p-dioxin of hepatocarcinogenesis from diethylnitrosamine. Cancer Res 40:3616-3620.
Proudfoot, AT. (2003) Pentachlorophenol poisoning. Toxicol Rev 22:3-11.
Purschke, M; Jacobi, H; Witte, I. (2002) Differences in genotoxicity of H2O2 and tetrachlorohydroquinone in human
fibroblasts. Mutat Res 513:159-167.
Ramlow, JM; Spadacene, NW; Hoag, SR; et al. (1996) Mortality in a cohort of pentachlorophenol manufacturing
workers, 1940-1989. Am J Ind Med 30:180-194.
Rawlings, NC; Cook, SJ; Waldbillig, D. (1998) Effects of the pesticides carbofuran, chlorpyrifos, dimethoate,
lindane, triallate, trifluralin, 2,4-D, and pentachlorophenol on the metabolic endocrine and reproductive endocrine
system in ewes. J Toxicol Environ Health A 54:21-36.
Reigner, BG; Gungon, RA; Hoag, MK; et al. (1991) Pentachlorophenol toxicokinetics after intravenous and oral
administration to rat. Xenobiotica 21:1547-1558.
Reigner, BG; Bois, FY; Tozer, TN. (1992a) Assessment of pentachlorophenol exposure in humans using the
clearance concept. HumExp Toxicol 11:17-26.
Reigner, BG; Gungon, RA; Bois, FY; et al. (1992b) Pharmacokinetic concepts in assessing intake of
pentachlorophenol by rats after exposure through drinking water. J Pharm Sci 81:1113-1118.
Reigner, BG; Rigod, JF; Tozer, TN. (1992c) Disposition, bioavailability, and serum protein binding of
pentachlorophenol in the B6C3Fi mouse. Pharm Res 9:1053-1057.
Renner, G; Hopfer, C. (1990) Metabolic studies on pentachlorophenol (PCP) in rats. Xenobiotica 20:573-582.
Renner, G; Mticke, W. (1986) Transformations of pentachlorophenol Part 1: metabolism in animals and man.
Toxicol Environ Chem 11:9-29.
Renner, G; Hopfer, C; Gokel, J; et al. (1987) Subacute toxicity studies on pentachlorophenol (PCP), and isomeric
tetrachlorobenzenediols tetrachlorohydroquinone (TCH), tetrachlorocatecol (TCC), and tetrachlororesorcinol (TCR).
Toxicol Environ Chem 15:301-312.
Rizzardini, M; Smith, AG. (1982) Sex differences in the metabolism of hexachlorobenzene by rats and the
development of porphyria in females. Biochem Pharmacol 31:3 543-3 548.
Robinson, C; Fowler, D; Brown, D; et al. (1987) Plywood mill worker's mortality patterns 1945-1977 (revised
March 1987). Study performed by Stanford Research Institute, Menlo Park, CA, for the National Institute for
Occupational Health and Safety, Cincinnati, OH. Available from the National Technical Information Service,
Springfield, VA; PB90-147075.
Royal Society of Chemistry. (1991) Pentachlorophenol. In: Kidd, H; James, DR; eds. The agrochemicals handbook.
3rd edition. Cambridge, England: Royal Society of Chemistry.
Rozman, T; Ballhorn, L; Rozman, K; etal. (1982) Effect of cholestyramine on the disposition of pentachlorophenol
in Rhesus monkeys. J Toxicol Environ Health 10:277-283.
Sai, K; Upham, BL; Kang, KS; et al. (1998) Inhibitory effect of pentachlorophenol on gap junctional intercellular
communication in rat liver epithelial cells in vitro. Cancer Lett 130:9-17.
Sai, K; Kanno, J; Hasegawa, R; et al. (2000) Prevention of the down-regulation of gap junctional intercellular
communication by green tea in the liver of mice fed pentachlorophenol. Carcinogenesis 21:1611-1676.
198
-------
Sai, K; Kang, KS; Hirose, A; et al. (2001) Inhibition of apoptosis by pentachlorophenol in v-myc-transfected rat
liver epithelial cells: relation to down-regulation of gap junctional intercellular communication. Cancer Lett
173:163-174.
Sai-Kato, K; Umemura, T; Takagi, A; et al. (1995) Pentachlorophenol-induced oxidative DNA damage in mouse
liver and protective effect of antioxidants. Food Chem Toxicol 33:877-882.
Saracci, R; Kogevinas, M; Bertazzi, PA; et al. (1991) Cancer mortality in workers exposed to chlorophenoxy
herbicides and chlorophenols. Lancet 338:1027-1032.
Savolainen, H; Pekari, K. (1979) Neurochemical effects of peroral administration of technical pentachlorophenol.
Res Commun ChemPathol Pharmacol 23:97-105.
Schmid, E; Bauchinger, M; Dresp, J. (1982) Chromosome analyses of workers from a pentachlorophenol plant.
Prog ClinBiol Res 109:471-477.
Schwetz, BA; Keeler, PA; Gehring, PJ. (1974a) The effect of purified and commercial grade pentachlorophenol on
rat embryonal and fetal development. Toxicol Appl Pharmacol 28:151-161.
Schwetz, BA; Keeler, PA; Gehring, PJ. (1974b) The effect of purified and commercial grade tetrachlorophenol on
rat embryonal and fetal development. Toxicol Appl Pharmacol 28:146-150.
Schwetz, BA; Quast, J; Keeler, P; et al. (1978) Results of two-year toxicity and reproduction studies on
pentachlorophenol in rats. In: Rao, K; ed. Pentachlorophenol: chemistry, pharmacology, and environmental
toxicology. New York and London: Plenum, pp. 301-309.
Seiler, JP. (1991) Pentachlorophenol. MutatRes 257:27-47.
Smith, JG; Christophers, AJ. (1992) Phenoxy herbicides and chlorophenols: a case control study on soft tissue
sarcoma and malignant lymphoma. Br J Cancer 65:442-448.
Smith, AH; Pearce, NE; Fisher, DO; et al. (1984) Soft tissue sarcoma and exposure to phenoxyherbicides and
chlorophenols in New Zealand. J Natl Cancer Inst 79(5): 1111-1117.
Smith, J; Loveless, L; Belden, E. (1996) Pentachlorophenol poisoning in newborn infants—St. Louis, Missouri,
April-August 1967. MMWR 45:545-549.
Spalding, JW; French, JE; Stasiewicz, S; et al. (2000) Responses of transgenic mouse lines p53(+/-) and Tg.AC to
agents tested in conventional carcinogenicity bioassays. Toxicol Sci 53:213-223.
Steenland, K; Bertazzi, P; Baccarelli, A; et al. (2004) Dioxin revisited: developments since the 1997 IARC
classification of dioxin as a human carcinogen. Environ Health Perspect 112(13): 1265-1268.
Suzuki, T; Komatsu, M; Isono, H. (1997) Cytotoxicity of organochlorine pesticides and lipid peroxidation in
isolated rat hepatocytes. Biol Pharm Bull 20:271-274.
Suzuki, T; Ide, K; Ishida, M. (2001) Response of MCF-7 human breast cancer cells to some binary mixtures of
oestrogenic compounds in vitro. J Pharm Pharmacol 53:1549-1554.
Teschke, K; Hertzman, C; Dimich-Ward, H; et al. (1989) A comparison of exposure estimates by worker raters and
industrial hygienists. Scand J Work Environ Health 15(6):424-429.
Teschke, K; Marion, SA; Ostry, A; et al. (1996) Reliability of retrospective chlorophenol exposure estimates over
five decades. Am J Ind Med 30(5):616-622.
Tisch, M; Faulde, MK; Maier, H. (2005) Genotoxic effects of pentachlorophenol, lindane, transfluthrin, cyfluthrin,
and natural pyrethrum on human mucosal cells of the inferior and middle nasal conchae. Am JRhinol 19:141-151.
199
-------
Treble, RG; Thompson, TS. (1996) Normal values for pentachlorophenol in urine samples collected from a general
population. J Anal Toxicol 20:313-317.
Trenti, T; Ventura, E; Ceccarelli, D; et al. (1986a) Porphyrins and pentachlorophenol in rat-liver mitochondria in
hexachlorobenzene-induced porphyria. IARC Sci Publ 77:457-459.
Trenti, T; Ventura, E; Ceccarelli, D; et al. (1986b) Functional derangement of liver mitochondria from
hexachlorobenzene-treated rats. IARC Sci Publ 77:329-331.
Triebig, G; Csuzda, I; Krekeler, HJ; et al. (1987) Pentachlorophenol and the peripheral nervous system: a
longitudinal study in exposed workers. Br J Ind Med 44:638-641.
Trosko, JE; Ruch, R. (1998) Cell-cell communication and carcinogenesis. Front Biosci 3:208-236.
Truhaut, R; L'Epee, P; Boussemart, E. (1952) Studies on toxicology of pentachlorophenol. II. Occupational
poisoning in the wood processing industry; reports of two fatal cases. Arch Mai Prof 13:567-569.
Trash, MA; Kensler, TW. (1991) An overview of the relationship between oxidative stress and chemical
carcinogenesis. Free Radic Biol Medicine 10:201-209.
Tsai, CH; Lin, PH; Waidyanatha, S; et al. (2001) Characterization of metabolic activation of pentachlorophenol to
quinones and semiquinones in rodent liver. Chem Biol Interact 134:55-71.
Uhl, S; Schmid, P; Schlatter, C. (1986) Pharmacokinetics of pentachlorophenol in man. Arch Toxicol 58:182-186.
Umemura, T; Sai-Kato, K; Takagi, A; et al. (1996) Oxidative DNA damage and cell proliferation in the livers of
B6C3Fi mice exposed to pentachlorophenol in their diet. Fundam Appl Toxicol 30:285-289.
Umemura, T; Kai, S; Hasegawa, R; et al. (1999) Pentachlorophenol (PCP) produces liver oxidative stress and
promotes but does not initiate hepatocarcinogenesis in B6C3F! mice. Carcinogenesis 20:1115-1120.
Umemura, T; Kai, S; Hasegawa, R; et al. (2003a) Prevention of dual promoting effects of pentachlorophenol, an
environmental pollutant, on diethylnitrosamine-induced hepato- and cholangiocarcinogenesis in mice by green tea
infusion. Carcinogenesis 24:1105-1109.
Umemura, T; Kodama, Y; Kanki, K; et al. (2003b) Pentachlorophenol (but not phenobarbital) promotes intrahepatic
biliary cysts induced by diethylnitrosamine to cholangio cystic neoplasms in B6C3Fi mice possibly due to oxidative
stress. Toxicol Pathol 31:10-13.
U.S. EPA (Environmental Protection Agency). (1980) Creosote, inorganic arsenicals, pentachlorophenol. Position
document no. 2/3. Office of Pesticide Programs, Washington, DC.
U.S. EPA. (1986a) Guidelines for the health risk assessment of chemical mixtures. Federal Register
51(185):34014-34025. Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012. Available
online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk assessment.
Prepared by the Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment,
Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, DC; EPA 600/6-87/008.
Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1992) Draft report: a cross-species scaling factor for carcinogen risk assessment based on equivalence of
mg/kg3/4/day. Federal Register 57(109):24152-24173.
200
-------
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation toxicity studies.
Federal Register 59(206):53799. Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-90/066F. Available online at
http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/R-94/007. Available online at
http://cfpub.epa. gov/ncea/raf/recordisplay.cfm?deid=42601.
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274-56322.
Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-26954. Available
online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA 100-B-00-002. Available online at
http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2000b) Benchmark dose technical guidance document. External review draft. Risk Assessment Forum,
Washington, DC; EPA/630/R-00/001. Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2000c) Supplementary guidance for conducting health risk assessment of chemical mixtures. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/002. Available online at
http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2002) A review of the reference dose and reference concentration processes. Risk Assessment Forum,
Washington, DC; EPA/630/P-02/0002F. Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2004) Benchmark dose software (BMDS) version 1.3.2 (last modified May 23, 2003). Available online
at http ://www. epa. gov/ncea/bmds/about. html.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum, Washington, DC;
EPA/630/P-03/001F. Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. Risk
Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available online at
http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA/100/B-06/002. Available online at
http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (2006b) A framework for assessing health risk of environmental exposures to children. National Center
for Environmental Assessment, Washington, DC, EPA/600/R-05/093F. Available online at
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=158363.
Van den Berg, KJ. (1990) Interaction of chlorinated phenols with the thyroxin binding sites of human transthyretin,
albumin and thyroid binding globulin. Chem Biol Interact 76:63-75.
Van Ommen, B; Adang, A; Muller, F; et al. (1986a) The microsomal metabolism of pentachlorophenol and its
covalent binding to protein and DNA. Chem Biol Interact 60:1-11.
Van Ommen, B; Adang, AE; Brader, L; et al. (1986b) The microsomal metabolism of hexachlorobenzene. Origin of
the covalent binding to protein. Biochem Pharmacol 35:3233-3238.
201
-------
Varnbo, I; Peterson, A; Walum, E. (1985) Effects of toxic chemicals on the respiratory activity of cultured mouse
neuroblastoma cells. Xenobiotica 15:727-733.
Villena, F; Montoya, G; Klaasen, R; et al. (1992) Morphological changes on nerves and histopathological effects on
liver and kidney of rats by pentachlorophenol (PCP). Comp Biochem Physiol C 101:353-363.
Vogel, E; Chandler, JL. (1974) Mutagenicity testing of cyclamate and some pesticides mDrosophila melanogaster.
Experientia 30:621-623.
von Ahsen, N; Richter, M; Grupp, C; et al. (2001) No influence of the MDR-1 C3435T polymorphism or a CYP3A4
promoter polymorphism (CYP3 A4-V allele) on dose-adjusted cyclosporin A trough concentrations or rejection
incidence in stable renal transplant recipients. Clin Chem 47:1048-1052.
Waidyanatha, S; Lin, PH; Rappaport, SM. (1996) Characterization of chlorinated adducts of hemoglobin and
albumin following administration of pentachlorophenol to rats. Chem Res Toxicol 9:647-653.
Walls, CB; Glass, WI; Pearce, NE. (1998) Health effects of occupational pentachlorophenol exposure in timber
sawmill employees: a preliminary study. NZ Med Jill :362-364.
Wang, YJ; Lin, JK. (1995) Estimation of selected phenols in drinking water with in situ acetylation and study on the
DNA damaging properties of polychlorinated phenols. Arch Environ Contam Toxicol 28:537-542.
Wang, YJ; Ho, YS; Chu, SW; et al. (1997) Induction of glutathione depletion, p53 protein accumulation and cellular
transformation by tetrachlorohydroquinone, a toxic metabolite of pentachlorophenol. Chem Biol Interact 105:1-16.
Wang, YJ; Lee, CC; Chang, WC; et al. (2001) Oxidative stress and liver toxicity in rats and human hepatoma cell
line induced by pentachlorophenol and its major metabolite tetrachlorohydroquinone. Toxicol Lett 122:157-169.
Waters, MD; Sandhu, SS; Simmon, VF; et al. (1982) Study of pesticide genotoxicity. Basic Life Sci 21:275-326.
Weinbach, E. (1954) The effect of pentachlorophenol on oxidative phosphorylation. J Biol Chem 210:545-550.
Welsh, JJ; Collins, TF; Black, TN; et al. (1987) Teratogenic potential of purified pentachlorophenol and
pentachloroanisole in subchronically exposed Sprague-Dawley rats. Food Chem Toxicol 25:163-172.
Wester, RC; Maibach, HI; Sedik, L; et al. (1993) Percutaneous absorption of pentachlorophenol from soil. Fundam
Appl Toxicol 20:68-71.
White, KL, Jr; Anderson, AC. (1985) Suppression of mouse complement activity by contaminants of technical grade
pentachlorophenol. Agents Actions 16:385-392.
Williams, PL. (1982) Pentachlorophenol, an assessment of the occupational hazard. Am Ind Hyg Assoc J 43:799-
810.
Williams, JA; Ring, B J; Cantrell, VE; et al. (2002) Comparative metabolic capabilities of CYP3 A4, CYP3 A5, and
CYP3A7. Drug Metab Dispos 30:883-891.
Witte, I; Juhl, U; Butte, W. (1985) DNA-damaging properties and cytotoxicity in human fibroblasts of
tetrachlorohydroquinone, a pentachlorophenol metabolite. MutatRes 145:71-75.
Witte, I; Zhu, BZ; Lueken, A; et al. (2000) Protection by desferrioxamine and other hydroxamic acids against
tetrachlorohydroquinone-induced cyto- and genotoxicity in human fibroblasts. Free Radic Biol Med 28:693-700.
Wood, S; Rom, WN; White, GL, Jr; et al. (1983) Pentachlorophenol poisoning. J Occup Med 25:527-530.
Woods, JS; Polissar, L; Severson, RK; et al. (1987) Soft tissue sarcoma and non-Hodgkin's lymphoma in relation to
phenoxyherbicide and chlorinated phenol exposure in western Washington. J Natl Cancer Inst 78(5):899-910.
202
-------
Xu, J. (1996) In vivo test for chemical induction of micro nucleated polychromatic erythrocytes in mouse bone
marrow cells. Study conducted by SITEK Research Laboratories, Rockville, MD for the Pentachlorophenol Task
Force c/o SRA International, Inc., Washington, DC. Submitted to the U.S. Environmental Protection Agency,
Washington, DC; MRID 43911301. Unpublished report.
Yamamoto, F; Kasai, H; Bessho, T; et al. (1992) Ubiquitous presence in mammalian cells of enzymatic activity
specifically cleaving 8-hydroxyguanine-containing DNA. Jpn J Cancer Res 83:351-357.
Yin, D; Gu, Y; Li, Y; et al. (2006) Pentachlorophenol treatment in vivo elevates point mutation rate in zebrafish p53
gene. Mutat Res 609:92-101.
Yuan, JH; Goehl, TJ; Murrill, E; et al. (1994) Toxicokinetics of pentachlorophenol in the F344 rat. Gavage and
dosed feed studies. Xenobiotica 24:553-560.
Ziemsen, B; Angerer, J; Lehnert, G. (1987) Sister chromatid exchange and chromosomal breakage in
pentachlorophenol (PCP) exposed workers. Int Arch Occup Environ Health 59:413-417.
REFERENCES ADDED AFTER EXTERNAL PEER REVIEW
Bogen, KT; Seilkop, S. (1993) Investigation of independence in inter-animal tumor type occurrence with the NTP
rodent-bioassay database. Report prepared for the National Research Council Committee on Risk Assessment for
Hazardous Air Pollutants, Washington, DC. Available online at
http://www.osti.gov/bridge/product.biblio.jsp?query_id=0&page=0&osti_id=10121101 (accessed August 17, 2010).
Collins, JJ; Budinsky, RA; Burns,CJ; et al. (2006). Serum dioxin levels in former chlorophenol workers. J Expo Sci
Environ Epidemic 116:76-84.
Elmore, S A; Nyska, A; Tischler, AS. (2009) The adrenal medulla as a target organ in toxicologic studies of rats and
mice. In: Harvey, PW; Everett, DJ; Springall, CJ; eds. Adrenal toxicology, target organ toxicology series, Vol. 26.
New York, NY: Informa Healthcare USA, Inc., pp. 111-138.
Folch, J; Yeste-Velasco, M; Alvira, D; et al. (2009) Evaluation of pathways involved in pentachlorophenol-induced
apoptosis in rat neurons. NeuroToxicology 30:451-458.
Greim, H; Hartwig, A; Reuter, U; et al. (2009) Chemically-induced pheochromocytomas in rats: mechanisms and
relevance for human risk assessment. Crit Rev Toxicol 39(8):695-718,
Hardell, L; Eriksson, M. (1999) A case-control study of non-Hodgkin lymphoma and exposure to pesticides.
Cancer 85:1353-1360.
Hardell, L; Eriksson, M. (2003) Is the decline of the increasing incidence of non-Hodgkin lymphoma in Sweden and
other countries a result of cancer preventive measures? Environ Health Perspect 111: 1704-1706.
Krishnan, B; Morgan, GJ. (2007) Non-Hodgkin lymphoma secondary to cancer chemotherapy. Cancer Epidemiol
Biomarkers Prev 16(3):377-380.
McLean, D; Eng, A; Mannetje, A; et al. (2007) Health outcomes in former New Zealand timber workers exposed to
pentachlorophenol (PCP). Technical report no. 20. Wellington: CPHR. Available online at
http://www.dol.govt.nz/publications/research/pcp-report-2008/final_pcp_reportjan_2008_l 1 .asp (accessed August
17, 2010).
McLean, D; Eng, A; Walls, C; et al. (2009a) Serum dioxin levels in former New Zealand sawmill workers twenty
years after exposure to pentachlorophenol (PCP) ceased. Chemosphere 74:962-967.
McLean, D; Eng, A; Dryson, E; et al. (2009b) Morbidity in former sawmill workers exposed to pentachlorophenol
(PCP): a cross-sectional study in New Zealand. Am J Ind Med 52:271-281.
203
-------
Mirabelli, MC; Hoppin, JA; Tolbert, PE; et al. (2000) Occupational exposure to chlorophenol and the risk of nasal
and nasopharyngeal cancers among US men aged 30 to 60. Am J Ind Med 37:532-541.
Ohta, S; Lai, EW; Taniguchi, S; et al. (2006) Animal models of pheochromocytoma including NIH initial
experience. Ann NY Acad Sci 1073:300-305.
Orton, F; Lutz, I; Kloas, W; et al. (2009) Endocrine disrupting effects of herbicides and pentachlorophenol: in vitro
and in vivo evidence. Environ Sci Technol 43(6):2144-2150.
Powers, JF; Picard, KL; Nyska, A; et al. (2008) Adrenergic differentiation and Ret expression in rat
pheochromocytomas. EndocrPathol 19(1):9-16.
Richardson, DB; Terschuren, C; Hoffmann, W. (2008) Occupational risk factors for non-Hodgkin's lymphoma: a
population-based case-control study in Northern Germany. Am J Ind Med 51:258-268.
Ruder, AM; Carreon, T; Butler, MA; et al. (2009) Exposure to farm crops, livestock, and farm tasks and risk of
glioma: the Upper Midwest Health Study. Am JEpidemiol 169:1479-1491.
Salmenkivi, K; Heikkila, P; Haglund, C; et al. (2004) Malignancy in pheochromocytomas. APMIS 112:551-559.
Stayner, L; Steenland, K; Dosemeci, M; et al. (2003) Attenuation of exposure-response curves in occupational
cohort studies at high exposure levels. Scand J Environ Health 29:317-324.
Sweeney, AT. (2009) Pheochromocytoma. eMedicine. Available online at
http://emedicine.medscape.com/article/124059-overview (accessed February 9, 2010).
Tischler, AS; Sheldon, W; Gray, R. (1996) Immunohistochemical and morphological characterization of
spontaneously occurring pheochromocytomas in the aging mouse. Vet Pathol 33:512-520.
Tischler, AS; Powers, JF; Alroy, J. (2004) Animal models of pheochromocytoma. Histol Histopathol 19:883-895.
't Mannetje, A; McLean, D; Cheng, S; et al. (2005) Mortality in New Zealand workers exposed to phenoxy
herbicides and dioxins. Occup Environ Med 62:34-40.
U.S. EPA. (2006a) Reregistration Eligibility Decision for Pentachlorophenol. Prevention, Pesticides Toxic
Substances, Washington, DC; EPA 739-R-08-008. Available online at
http ://www. epa. gov/oppsrrd l/REDs/pentachlorophenol_red.pdf.
Zhu, B-Z; Shan, G-Q. (2009) Potential mechanism for pentachlorophenol-induced carcinogenicity: a novel
mechanism for metal-independent production of hydroxyl radicals. Chem Res Toxicol 22:969-977.
Zhu, B-Z; Kitrossky, N; Chevion, M. (2000) Evidence for production of hydroxyl radicals by pentachlorophenol
metabolites and hydrogen peroxide: a metal independent organic Fenton reaction. Bicohem Biophys Res Commun
270:942-946.
204
-------
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
The Toxicological Review of Pentachlorophenol (dated April 2009) has undergone a
formal external peer review performed by scientists in accordance with the U.S. EPA guidance
on peer review (2006a, 2000a). An external peer-review workshop was held August 4, 2009.
The external peer reviewers were tasked with providing written answers to general questions on
the overall assessment and on chemical-specific questions in areas of scientific controversy or
uncertainty. A summary of significant comments made by the external reviewers and the EPA's
responses to these comments arranged by charge question follow. In many cases the comments
of the individual reviewers have been synthesized and paraphrased in development of
Appendix A. The EPA also received scientific comments from the public. These comments and
the EPA's responses are included in a separate section of this appendix.
EXTERNAL PEER REVIEWER COMMENTS
The reviewers made several editorial suggestions to clarify specific portions of the text.
These changes were incorporated in the document as appropriate and are not discussed further.
When the external peer reviewers commented on decisions and analyses in the Toxicological
Review under multiple charge questions, these comments were organized under the most
appropriate charge question.
A. General Charge Questions
1. Is the Toxicological Review logical, clear and concise? Has EPA accurately, clearly and
objectively represented and synthesized the scientific evidence for noncancer and cancer
hazards?
Comments: Two reviewers considered the document to be well written. One of the reviewers
commented that the document is logical and clear, but could be more concise (i.e., less
repetitive). One reviewer found the document to provide an accurate, clear, and objective
presentation of the studies. This reviewer did note some editorial and grammatical errors.
Another reviewer commented that the presentation of the toxicological and epidemiological data
was very logical, clear, and concise. One reviewer considered the review of the literature to be
thorough and comprehensive and presented in a logical manner. This reviewer stated that the
weight of evidence of PCP toxicity to be objectively analyzed. One reviewer considered the
organization of the Toxicological Review to be cumbersome. This reviewer did not find the
science regarding the mode of action to have been adequately evaluated, in particular the failure
A-l
-------
to have incorporated the initiation/promotion study by Umemura et al. (1999) in the analysis of
the mode of action for liver tumors.
Response: The content of the Toxicological Review is consistent with the current outline for
IRIS Toxicological Reviews. The document was reviewed and edited to improve clarity and
reduce repetition. The study by Umemura et al. (1999) is described in detail in Section 4.2.4.1,
Initiation/promotion studies; discussion of the findings of this study was added to
Section 4.7.2.2, Animal Cancer Evidence from Oral Exposure.
2. Please identify any additional studies that should be considered in the assessment of the
noncancer and cancer health effects of PCP.
Comments: Two reviewers were not aware of any additional studies that should be included in
the assessment. Two reviewers identified the following published studies for consideration:
Folch, J; Yeste-Velasco, M; Alvira, D; et al. (2009) Evaluation of pathways involved in pentachlorophenol-
induced apoptosis in rat neurons. Neurotoxicology 30:451^58.
McLean, D; Eng, A; Walls, C; et al. (2009a) Serum dioxin levels in former New Zealand sawmill workers
twenty years after exposure to pentachlorophenol (PCP) ceased. Chemosphere 74:962-967.
McLean, D; Eng, A; Dryson, E; et al. (2009b) Morbidity in former sawmill workers exposed to
pentachlorophenol (PCP): a cross-sectional study in New Zealand. Am J Ind Med 52:271-281.
Mirabelli, MC; Hoppin, JA; Tolbert, PE; et al. (2000) Occupational exposure to chlorophenol and the risk
of nasal and nasopharyngeal cancers among US men aged 30 to 60. Am J Ind Med 37:532-541.
Orton, F; Lutz, I; Kloas, W; and Routledge, EJ. (2009) Endocrine disrupting effects of herbicides and
pentachlorophenol: in vitro and in vivo evidence. Environ Sci Technol 43:2144-2150.
't Mannetje, A; McLean, D; Cheng, S; et al. (2005) Mortality in New Zealand workers exposed to phenoxy
herbicides and dioxins. Occup Environ Med 62:34-40.
Zhu, BZ and Shan, GQ. (2009) Potential mechanism for pentachlorophenol-inducedcarcinogenicity: a
novel mechanism for metal-independent production of hydroxyl radicals. Chem Res Toxicol 22:969-977.
One of these reviewer identified a new National Institute for Occupational Safety and
Health (NIOSH) epidemiological study that was said to provide evidence for an association
between exposure to PCP and a risk of non-Hodgkin's lymphoma. However, this reviewer noted
that the study is currently unpublished and that only a conference abstract is available.
One reviewer noted that the findings in Umemura et al. (1999) comparing rat and mouse
liver effects that were discussed in Section 4.2.4.1 should have also been incorporated into the
discussions on mode of action (Sections 4.5 and 4.7.3).
A-2
-------
Response: Summaries of the McLean et al. cohort studies of serum dioxin levels (2009a) and
morbidity (e.g., respiratory and neurological effects, 2009b) in former New Zealand sawmill
workers exposed to PCP were added to Sections 4.1.2.2 and 4.1.2.3. These studies are
considerably larger than any other studies examining these types of effects.
A reviewer also suggested adding the Mirabelli et al. (2000) study of case-control study
of nasal and nasopharyngeal cancers in relation to chlorophenol exposure. This study is a
parallel study to the Hoppin et al. (1998) case-control study of soft tissue sarcoma; that is, these
two studies were conducted using the same study design and exposure assessment. A full
description of the Hoppin et al. (1998) study was not included in the Toxicological Review
because the authors presented data only for a combined exposure (e.g., chlorophenols, or
chlorophenols and phenoxy herbicides). However, this study, along with several other studies
that presented data only for a combined exposure (e.g., chlorophenols, or chlorophenols and
phenoxy herbicides), are noted in Section 4.1.1.1. Mirabelli et al. (2000) and 't Mannetje et al.
(2005) have been added to the studies listed in Section 4.1.1.1. These studies present
information for a combined category of chlorophenols, for five occupational exposure categories
that were the basis for estimating chlorophenol exposure (cutting oils, leather work,
saw/pulp/planning mill, shoe/leather dust, and wood preserving chemicals), and for the
occupational exposure category described as plywood/fiberboard/particleboard and wood/saw
dust. Because the authors did not include a discussion of the relative contribution of PCP to each
of these categories, these studies were not considered directly useful for an assessment of PCP
hazard.
The unpublished NIOSH study was not included in the assessment because it is not
currently part of the peer-reviewed literature. A published abstract based on a presentation at the
2009 Society for Epidemiologic Research was found (Ruder and Sweeney, 2009) but because the
information provided in the abstract is limited and was not subject to peer review, the abstract
was not included in the Toxicological Review.
Relevant information from the other studies identified by the reviewers were added to the
Toxicological Review. As noted in response to a comment under General Charge Question #1,
further discussion of the initiation/promotion study by Umemura et al. (1999) was added to
Section 4.7.2.2, Animal Cancer Evidence from Oral Exposure.
3. Please discuss research that you think would be likely to increase confidence in the
database for future assessments of PCP.
Comments: Reviewers offered suggestions for additional research to address the data gaps for
PCP, most of which focused on the need for further elucidation of a cancer mode of action and
the development of an RfC. Specific research recommendations included the following:
A-2
-------
• Epidemiologic studies focusing on quantitative exposure assessment;
• Studies of PCP metabolism;
• Development of toxicokinetic models for route-to-route extrapolation to allow for the
development of an RfC;
• Inhalation studies to support development of an RfC;
• Studies in the low-dose range, focusing on endpoints pertinent to endocrine disruption
and neurological effects;
• A study of aPCP that could further define the dose response to allow for BMD
modeling;
• Comparison of the quinone metabolites of PCP in the liver nuclei of dogs, mice, and
rats;
• Further research on the cancer mode of action to reduce uncertainty in the cancer
assessment, with one reviewer suggesting molecular techniques such as microarray
analysis, and another suggesting that genotoxicity testing, specifically the comet assay
or nucleotide postlabeling, be performed in target organs for PCP-induced
carcinogenicity; and
• Dermal toxicity studies.
Response: The EPA agrees that additional research in the areas recommended by the peer
reviewers would increase the confidence in the PCP database for future toxicological
assessments of this chemical.
4. Please comment on the identification and characterization of sources of uncertainty in
Sections 5 and 6 of the assessment document. Please comment on whether the key sources
of uncertainty have been adequately discussed. Have the choices and assumptions made in
the discussion of uncertainty been transparently and objectively described? Has the
impact of the uncertainty on the assessment been transparently and objectively described?
Comments: Two reviewers commented that the choices and assumptions made in the discussion
of uncertainty were transparently and objectively described. One of these reviewers specifically
noted that the section on uncertainty was concise and thoughtful. The other reviewer indicated
that the impact of the uncertainties identified in the assessment were adequately presented.
Several reviewers offered suggestions to more completely characterize the sources of
uncertainty associated with the PCP database. One reviewer offered comments on a specific UF;
these comments are summarized and addressed in response to RfD Charge Question #5.
A-4
-------
Two reviewers suggested that PODs for the cancer assessment be estimated using other
models in BMDS to provide quantitative information regarding the degree of uncertainty/
sensitivity associated with the choice of the low-dose extrapolation procedure. One reviewer
also thought that uncertainty related to tumor site concordance (i.e., that the quantitative cancer
assessment is based on liver, adrenal and circulatory system cancers, whereas the most
commonly reported association in the epidemiologic literature is lymphomas) should be
addressed. One reviewer questioned the conclusion in the Toxicological Review that no
additional uncertainty is added to the assessment by estimating combined risks reflecting
multiple tumor sites and the assumption that the carcinogenesis process is completely
independent across tumor sites. This reviewer suggested discussion of this assumption and the
choice of prior distribution as a source of uncertainty.
One reviewer identified uncertainties in the principal study associated with a study
duration that was shorter than other chronic studies of PCP and small numbers of animals tested.
Another reviewer suggested that a discussion of the uncertainty inherent in comparing effects
from different compositions of PCP would be helpful.
Response: With respect to the comment regarding tumor site concordance, the U.S. EPA's
Guidelines for Carcinogen Risk Assessment (2005a) state that "... agents observed to produce
tumors in both humans and animals have produced tumors either at the same site (e.g., vinyl
chloride) or different sites (e.g., benzene) (NRC, 1994). Hence, site concordance is not always
assumed between animals and humans." A discussion of the uncertainties associated with this
assumption was added to Section 5.4.5.
Identification of the potential for impurities to influence the toxicity of the PCP formation
tested by Mecler (1996) (90.6% PCP) was added as an additional area of uncertainty in
Section 5.3.
Comments on the selection of the principal study, the EPA's procedure for estimating
combined risks from multiple tumor sites and the assumption of independence of the
carcinogenesis process across tumor sites, and use of other models in BMDS to better evaluate
the sensitivity of the selected analysis are addressed in response to comments under RfD Charge
Question #1 and Cancer Charge Questions #2 and 6.
A-5
-------
Chemical-Specific Charge Questions:
B. Oral Reference Dose (RfD) for Pentachlorophenol
1. A 1-year oral study in dogs by Mecler (1996) was selected as the basis for the RfD.
Please comment on whether the selection of this study as the principal study is scientifically
justified. Has this study been transparently and objectively described in the document?
Are the criteria and rationale for this selection transparently and objectively described in
the document? Please identify and provide the rationale for any other studies that should
be selected as the principal study.
Comments: Four of the reviewers noted that the study was transparently and objectively
described in the document. Two of these reviewers agreed that the selection of the Mecler
(1996) study was scientifically justified; the other two reviewers commented that the selection of
the Mecler (1996) study as the principal study was the appropriate study on which to base the
RfD since this study identified hepatotoxicity at the lowest dose tested in the available studies.
One of these reviewers commented that the nature of the liver pigmentation described in the
Mecler (1996) study needed clarification. This reviewer also noted that there was no discussion
of absorption, distribution, metabolism, and excretion (ADME) in dogs.
The fifth reviewer questioned whether the study by Kimbrough and Linder (1978), rather
than Mecler (1996), resulted in the lowest RfD, and noted a discrepancy between the figure and
table presented in Section 5.1.4. This reviewer surmised that the decision to use a principal
study that did not apparently result in the lowest RfD was not a scientific choice but rather one
based on EPA risk assessment policy.
Response: The Mecler (1996) study was selected as the principal study because it represents the
most sensitive chronic oral study for PCP. The decision to use this study was based on a critical
review of the scientific literature and not on EPA policy. Table 5-1 and Figure 5-1 contained
incorrect information that misrepresented the potential RfDs calculated from the LOAELs
observed in the studies by Mecler (1996) and Kimbrough and Linder (1978). The table and
figure were corrected to show that the study by Kimbrough and Linder (1978) yields a candidate
RfD that is higher that the RfD derived from Mecler (1996). Discussion of uncertainty
associated with the Kimbrough and Linder (1978) study in Section 5.1.1 was revised. Text was
added to the summary of the Mecler (1996) study in Section 4.2.1.3 to better characterize the
nature of the liver pigmentation as described by the study authors. Information on ADME of
PCP in dogs is not available.
A-6
-------
2. An increase in hepatic effects (characterized by a dose-related increase in the incidence
of hepatocellular pigmentation, cytoplasmic vacuolation, chronic inflammation, and
severely discolored livers; statistically significant increases in absolute (females only) and
relative liver weights, and serum enzyme activity) as reported by Mecler (1996) was
selected as the critical effect for the RfD because these effects are considered by EPA to be
indicative of hepatocellular injury. Please comment on whether the rationale for the
selection of this critical effect is scientifically justified. Are the criteria and rationale for
this selection transparently and objectively described in the document? Please provide a
detailed explanation. Please identify and provide the rationale for any other endpoints that
should be considered in the selection of the critical effect.
Comments: All of the reviewers agreed with the selection of the critical effect. One reviewer,
while stating that use of the Mecler (1996) study and subclinical hepatic effects as the principal
study and critical effect was acceptable, questioned whether necrosis as reported in the study by
Kimbrough and Linder (1978) would yield a lower RfD and thus serve as a more appropriate
critical effect for derivation of the RfD. One reviewer commented that they agreed that based on
the available mode-of-action data, PCP is capable of inducing hepatocellular injury. This
reviewer stated that they would have selected a newer study utilizing lower doses if they were
available. The reviewer also stated that if newer studies were completed, they expected that
endocrine and neurological endpoints would be the most sensitive.
Response: The description of the study by Kimbrough and Linder (1978) incorrectly indicated
that liver necrosis was observed in rats exposed to the highest dose tested of tPCP. In addition,
necrosis was used to describe the effects observed at the LOAEL in Table 5-1 and Figure 5-1. In
fact, necrosis was not observed in this study. The text has been corrected to reflect the actual
findings.
No studies on neurological or endocrine-related endpoints are available for PCP. The
study by Mecler (1996) represents the most sensitive chronic oral study for PCP.
3. The hepatotoxic data and a NOAEL/LOAEL approach were used to derive the point of
departure (POD) for the RfD. Please provide comments with regard to whether this is the
best approach for determining the POD. Has it been transparently and objectively
described? Please identify and provide rationales for any alternative approaches for the
determination of the POD and discuss whether such approaches are preferred to EPA's
approach.
Comments: Three reviewers specifically agreed with the application of the NOAEL/LOAEL
approach to derive the POD. Two reviewers recommended a clearer, more convincing rationale
A-7
-------
for not conducting BMD modeling. One of these reviewers noted that a 100% response in dose
groups, the absence of a NOAEL, and small group sizes should not preclude BMD modeling.
One reviewer suggests that it would be possible to conduct BMD modeling on the data from the
Kimbrough and Linder (1978) study. In addition, one reviewer mentioned that it would be of
interest to compare RfD values derived from studies that produced a NOAEL value at the lowest
dose tested to those derived from the LOAEL identified by Mecler (1996) in order to see whether
the values were similar, thus providing further support for the POD chosen for deriving the RfD.
Response: In Mecler (1996), the incidence of hepatocellular pigmentation in males and females
and chronic inflammation in males increased from 0% in the controls to 100% in the low-dose
group; both the mid- and high-dose groups also had 100% responses. These data were not
amenable to BMD modeling, as there is no information to inform the shape of the dose-response
curve. Thus, the NOAEL/LOAEL approach was employed to identify the POD. The text that
stated the absence of a NOAEL and small group sizes in the Mecler (1996) study precluded
BMD modeling was removed and text was added to Section 5.1.2, Methods of Analysis, to
clearly articulate the rationale (i.e., 100% response at the lowest dose) behind selecting the
LOAEL/NOAEL approach.
Figures 5-1 and 5-2 provide graphical comparisons of candidate PODs that represent
NOAELs and LOAELs from alternative studies and data sets that were considered as the basis
for the PCP RfD.
4. The RfD is based on toxic effects observed in dogs (Mecler, 1996) administered a
technical grade formulation of PCP (90.9% purity). Considering the toxicological database
for PCP is largely comprised of studies that utilized similar formulations, as well as
commercial and analytical (pure) formulations, please provide comments with regard to
whether the use of data based on animal exposure to a technical grade PCP formulation of
this purity is the best approach and can be considered representative of pure PCP. If not,
please identify and provide the rationale for any alternative data sets, and the sufficiency of
such data sets, to support derivation of the RfD.
Comments: All five peer reviewers agreed that tPCP can be considered representative of pure
PCP.
Response: No response necessary.
5. Please comment on the selection of the uncertainty factors applied to the POD for the
derivation of the RfD. For instance, are they scientifically justified and transparently and
objectively described in the document? If changes to the uncertainty factors are proposed,
A-8
-------
please identify and provide a rationale(s). Please comment specifically on the following
uncertainty factor:
. An uncertainty factor of 3 was applied in deriving the RfD to account for the use of
a LOAEL rather than a NOAEL as the POD.
Comments: Two reviewers thought that the selection of a UFL of 3 was appropriate. Three other
reviewers questioned the characterization of the hepatic effects observed in the Mecler (1996)
study as mild, and did not find the rationale for using a UFL of 3 rather than 10 to be adequately
justified.
Response: The discussion of the UFL in Section 5.1.3 was revised to provide better justification
for the selection of a UF of 3 for extrapolation from a LOAEL to a NOAEL.
C. Inhalation Reference Concentration (RfC) for Pentachlorophenol
1. An RfC was not derived due to the lack of available studies to characterize the health
effects associated with pentachlorophenol administered via the inhalation route. Are there
available data that might support development of an RfC for pentachlorophenol?
Comments: All of the peer reviewers agreed that available data do not support the development
of an RfC.
Response: No response necessary.
D. Carcinogenicity of Pentachlorophenol
1. Under EPA's 2005 Guidelines for Carcinogen Risk Assessment
(http://www.epa.gov/iris/backgrd.html), the Agency concluded that pentachlorophenol is
"likely to be carcinogenic" to humans. Please comment on the cancer weight of evidence
characterization. Has the scientific justification for the weight of evidence descriptor been
sufficiently, transparently and objectively described? Do the available data for liver,
adrenal gland, and circulatory system tumors in mice and nasal tumors and mesotheliomas
in rats support the conclusion that PCP is a likely human carcinogen?
Comments: Four of the five reviewers agreed that the classification of PCP as "likely to be
carcinogenic to humans" was appropriate based on tumor incidence in animal studies and
epidemiological data. One of these reviewers thought that the rationale for selecting "likely to be
carcinogenic to humans" over other descriptors should have been explicitly stated. One reviewer
A-9
-------
stated that only a descriptor of "possibly carcinogenic to humans" could be supported by animal
tumor findings in the absence of a mode of action of established relevance to humans. This
reviewer was unable to locate the weight-of-evidence descriptor.
Response: As discussed in the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
supporting data for the descriptor of "likely to be carcinogenic to humans" may include human
studies demonstrating a plausible (but not definitively causal) association between exposure and
cancer and positive studies in animals in more than one species, sex, strain, site, or exposure
route. Section 4.7.1 of the Toxicological Review presents the data supporting the cancer
descriptor for PCP. PCP is "likely to be carcinogenic to humans" based on positive studies in
more than one species (mice and rats), both males and females, and multiple sites (hepatocellular
adenomas and carcinomas, pheochromocytomas and malignant pheochromocytomas,
hemangiomas and hemangiosarcomas, and malignant mesotheliomas and nasal squamous cell
carcinomas), along with supporting data demonstrating a plausible (but not definitively causal)
association between human exposure and cancer. Because the PCP carcinogenicity data are
consistent with the descriptor of "likely to be carcinogenic to humans," the rationale for not
selecting other descriptors would not be a useful addition.
A characterization of PCP as "possibly carcinogenic to humans" as suggested by one peer
reviewer is not one of the descriptors offered in the 2005 Cancer Guidelines. As noted in the
guidelines, knowing the cancer mode of action for a chemical might change or strengthen the
rationale for choosing a descriptor; however, not knowing the cancer mode of action does not
influence the choice of the descriptor. Thus, a descriptor of "likely to be carcinogenic to
humans" does not require that a mode of action establishing relevance of animal tumors to
humans be elucidated.
2. A quantitative oral cancer assessment has been derived for PCP. Do the data support
an estimation of a cancer slope factor for PCP? Please comment on the scientific
justification for deriving a quantitative cancer assessment. Has the rationale and scientific
justification for quantitation been transparently and objectively described?
Comments: All five reviewers agreed that the available data are sufficient for deriving a cancer
slope factor. Four of the reviewers commented that the rationale and justification for the
quantitative assessment was adequately described in the document. One reviewer provided no
comment regarding the rationale and justification for the quantitative assessment. One reviewer
commented that a comparison of different modeling approaches to better evaluate the sensitivity
of the selected analysis would be useful.
A-10
-------
Response: EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) encourage
consideration of alternative models where feasible. However, in cases like PCP in which only
curve-fitting models are available, the guidelines indicate that such a comparison of model
results would not be so useful. From p. 3-15:
Many different curve-fitting models have been developed, and those that fit the observed
data reasonably well may lead to several-fold differences in estimated risk at the lower
end of the observed range. Another problem occurs when a multitude of alternatives are
presented without sufficient context to make a reasoned judgment about the alternatives.
This form of model uncertainty reflects primarily the availability of different computer
models and not biological information about the agent being assessed or about
carcinogenesis in general. In cases where curve-fitting models are used because the data
are not adequate to support a toxicodynamic model, there generally would be no
biological basis to choose among alternative curve-fitting models. However, in situations
where there are alternative models with significant biological support, the decisionmaker
can be informed by the presentation of these alternatives along with their strengths and
uncertainties.
These concerns underscore the EPA's rationale as expressed in the Cancer Guidelines for the use
of a consistent model-independent approach and the use of an upper bound on risk that captures
some uncertainty in the estimation of the POD. Accordingly, EPA retained the quantitative
cancer approach utilizing the multistage model for PCP.
3. A two-year oral cancer bioassay (NTP, 1989) in mice was selected as the principal study
for the development of an oral slope factor. Please comment on the appropriateness of the
selection of the principal study. Has the rationale for this choice been transparently and
objectively described?
Comments: Three of the reviewers agreed that the NTP bioassay (1989) was the appropriate
choice. One reviewer commented that the presentation of the development of the slope factor
was well presented with objectivity and good transparency. This reviewer noted that selection of
the principal study with multiple tumor sites allowed for estimation of a statistically appropriate
upper bound on total risk (combined slope factor), which described the risk of developing any
combination of tumor types considered. One reviewer recommended that the possibility of a
threshold (or nonlinear approach) for cancer be considered since the mode of action for PCP
involves promoting action. This reviewer also commented that mouse liver carcinogenicity is
species specific and that justification for the selection of mouse liver tumor data should be
provided.
Response: The mode of action for PCP-induced carcinogenicity is unknown. In vivo study
results provide evidence that PCP has promoting activity (as summarized in Section 4.2.4.1);
A-ll
-------
however, as discussed in Sections 4.7.3 and 5.4.3, in vivo and in vitro data suggest that multiple
modes of carcinogenic action are possible, and that none have been defined sufficiently (e.g., key
events for carcinogenicity, dose-response and temporal relationships) to inform the shape of the
dose-response curve at low doses. Therefore, data to establish significant biological support for
a nonlinear approach are unavailable.
Because mice are relatively susceptible to liver tumors and because mouse liver tumors
can occur with a relatively high background, use of mouse liver tumor data in risk assessment
has been a subject of controversy. In the absence of mode-of-action data or other information
that establishes lack of human relevance, mouse liver tumors (as well as other rodent tumors) are
considered relevant to humans. A discussion of the uncertainties associated with the selection of
these tumor data sets for derivation of the slope factor as well as justification for their selection
was added to Section 5.4.5.
4. Data on the mode of action (MOA) of carcinogenicity of PCP were considered. Several
hypothesized MO As were evaluated within the Toxicological Review and EPA reached the
conclusion that a MOA(s) could not be supported for any tumor types observed in animal
models. Please comment on whether the weight of the scientific evidence supports this
conclusion. Please comment on whether the rationale for this conclusion has been
transparently and objectively described. Please comment on data available for PCP that
may provide significant biological support for a MOA beyond what has been described in
the Toxicological Review.
Comments: Four of the five reviewers agreed that the scientific evidence supports the
conclusion that a mode of action could not be established. One reviewer commented that the
mode-of-action data were not adequately discussed. Specifically, the reviewer commented on
the lack of consideration of studies finding mouse liver tumor promotion but a lack of initiation
following PCP exposure. This reviewer questioned why rat liver was not discussed as a target of
PCP metabolites as shown by Lin et al. (1977) and why the data for oxidative damage were
considered too limited to consider it as a possible mode of action. This reviewer also suggested
that further justification be provided to support the conclusion that oxidative stress-induced DNA
damage is thought to be related to the formation of electrophilic metabolites of PCP that are
capable of binding to DNA.
Response: In Section 4.7.3, Mode-of-Action Information, the available studies on PCP-induced
liver tumor promotion, the species differences in liver tumors between mice and rats, and the
oxidative damage induced by PCP metabolites are all discussed and considered. However, there
are several other responses to PCP exposure that can have promoting effects and could be
involved in the mode of action, including necrosis and chronic inflammation leading to
A-12
-------
reparative cell proliferation/regeneration and interference with GJIC, as well as other types of
genotoxic damage, including DNA adduct formation. Therefore, the mode of action for the
carcinogenic effects of PCP could not be determined. Clarification regarding the conclusions
related to oxidative stress-induced DNA damage were made to the Toxicological Review.
5. Increased incidence of tumors in male and female B6C3Fi mice was observed following
administration of two formulations of PCP [technical grade PCP and EC-7 (a commercial
grade of PCP)] that contain various chlorophenol and chlorinated dibenzodioxin and
dibenzofuran contaminants. The carcinogenic contributions of PCP versus those of
contaminants have been described qualitatively and to a limited extent quantitatively
within the document. The cancer assessment is based on the data sets resulting from
exposure to two different formulations that are approximately 90% PCP, with the
assumption that carcinogenic contributions from the contaminants are minimal. Please
comment on the scientific justification and transparency of this analysis. Please comment
on whether these are the appropriate data sets on which to base the cancer risk estimate
and, if not, please identify and provide the rationale for any alternative data sets, and the
sufficiency of such data sets, to support estimation of cancer risk.
Comments: The reviewers generally agreed with the EPA's assumption that the contribution of
chemical contaminants to the carcinogenic response of tPCP is minimal. One reviewer indicated
that it was not clear why the data for tPCP was chosen over data for EC-7 for characterizing PCP
cancer risk. Two reviewers suggested that the EPA consider pooling the findings from the NTP
study for tPCP with the EC-7 findings on the basis that the two formulations have similar PCP
content and the carcinogenicity for both is attributable to PCP, and assuming that these studies
do not have significantly different tumor results. One of these reviewers noted that it is possible
that differences in slope factors for tPCP and EC-7 are due to random variability in the
experimental responses, rather than some difference in the underlying formulations. The
reviewer considered this view supported by the fairly similar BMD values for tPCP and EC-7.
One reviewer considered the approach for reseating the slope factors by I/purity to be justified,
but noted that this purity reseating was not applied in estimating the slope factors for aPCP. One
of these reviewers observed that the mouse liver promotion study of Umemura et al. (1999)
performed with aPCP supports the interpretation that the liver effects are due to aPCP.
Response: The tumor incidence in mice exposed to tPCP or EC-7 for 2 years in the NTP (1989)
bioassays differed quantitatively. For example, the incidence of hepatocellular adenoma was
almost twofold higher in tPCP-exposed male mice compared to EC-7-exposed mice; similarly,
the incidence of adenoma or carcinoma was 1.8-fold higher in tPCP-exposed male mice
compared to EC-7-exposed mice. Whether these differences were attributable to random
A-13
-------
variability, different compositions of impurities in the two PCP formulations, or other factors is
uncertain. Given this uncertainty, the two data sets were modeled separately. The rationale for
modeling the data sets separately was clarified in Section 5.4.2. The text in Section 5.4.4 was
revised to clarify the rationale for selecting tPCP as the basis for characterizing PCP cancer risk.
Section 5.4.4 presents a discussion of the potential impact of impurities on the value of
the oral slope factor for aPCP derived from data for tPCP. If the carcinogenic risk associated
with impurities is negligible relative to that from PCP alone, scaling by I/purity (or 1/0.9), which
would increase the slope factor by 10%, is appropriate. On the other hand, if the carcinogenic
activity of impurities is not negligible, then the PCP slope factor should be reduced. In the
absence of information to establish the impact of impurities on the oral cancer potency,
adjustments to reduce or increase the slope factor were not applied. If scaling by I/purity was
applied, an increase in the estimated slope factor by 10% would not change the value of the PCP
slope factor when rounded to one significant figure [i.e., (4.0 x 10"1) x 0.1 = 4.4 x 10"1 or,
rounded to one significant figure, 4 x 10"1]. Text was added to Section 5.4.4 to clarify why the
oral slope factor was not adjusted to account for impurities in the tPCP.
6. Data on tumors in the liver and adrenal gland in B6C3F] male mice administered
technical PCP were used to estimate the oral cancer slope factor. Please comment on the
estimation of a statistically appropriate upper bound on total risk (combined slope factor),
which described the risk of developing any combination of tumor types considered. Please
comment on the scientific justification and transparency of the analysis for combining these
data to derive the oral cancer slope factor. Please comment on the use of data in male mice
exposed to technical PCP for a cancer risk estimate for both technical and analytical PCP.
Comments: Three reviewers agreed that an approach for deriving a slope factor for tPCP
involving a combined risk across tumor types was justified or reasonable. One of these
reviewers questioned aspects of the combined risk analysis, suggesting that a parametric
bootstrap technique would be advantageous (i.e., preclude the use of a uniform Bayesian prior
and having confidence limits more comparable with BMDS). A fourth reviewer said it was
appropriate if one can assume the sites are independent; this reviewer asked whether the
assumption of independence across cancer sites could be tested by further analyzing historical
tumor data from the NTP database. The fifth reviewer did not offer a response to this charge
question.
Response: The Bayesian binomial estimate of the probability, rather than a full-blown Bayesian
analysis, was utilized to prevent the re-sampling probability from being 0 or 1. For a control
group incidence of 0, as is the case with both male and female tPCP data (see Table D-l), the
A-14
-------
parametric estimate of the control probability is 0. Hence the reviewer's suggestion would likely
somewhat underestimate variability.
Tumor-type associations among individual animals in 62 B6C3Fi mouse studies and
61 F344 rat studies from the NTP database were evaluated by Bogen and Seilkop (1993). The
NRC (1994) considered this evaluation and reported that tumor-type occurrences in NTP
bioassays were in most cases nearly independent, and that the few departures that were detected
were small.
PUBLIC COMMENTS
Comment: One commenter expressed their support for the LOAEL to NOAEL UF of 3, noting
that a UF of 3 was applied to the same study (Mecler, 1996) used to derive the chronic RfD in
the EPA's Office of Pesticide Programs' 2008 RED document for PCP. This commenter
observed that questions have been raised about the relevance of the liver effects at the low dose
(1.5 mg/kg-day) and that selection of this dose as the LOAEL already reflects a conservative
choice.
Response: The LOAEL to NOAEL UF of 3 was retained. The justification for this selection in
Section 5.1.3 was clarified.
Comment: A commenter disagreed with the EPA's conclusion that the mode(s) of action for
liver tumors has not been defined sufficiently to inform low-dose extrapolation for estimation of
PCP carcinogenicity. The commenter stated that chronic oxidative stress leading to generation
of ROS and liver toxicity (as demonstrated by dose-related increases in AP and apyrimidinic
sites, 8-OH-dG, and SSBs in DNA) are strongly supported as the mechanism for tumor
induction, and that oxidative stress occurred only at high PCP exposures. The commenter also
observed that species differences in susceptibility to PCP could be explained by differences in
the rate of glutathione depletion, with conjugation of ROS with glutathione serving as a
protective mechanism against ROS-induced toxicity. Because glutathione depletion was not
expected to occur in humans, the commenter considered humans not to be susceptible to
glutathione depletion at current exposures to PCP. The commenter concluded that neither
oxidative stress nor glutathione depletion were likely to occur in humans at levels to which
humans are exposed to PCP. As such, the commenter argued that the mode of action for PCP is
expected to be nonlinear and that the current cancer assessment for PCP should be replaced by a
nonlinear assessment.
Response: EPA agrees that evidence supports oxidative stress as playing a role in tumor
induction; however, the available data indicate that multiple modes of carcinogenic action are
A-15
-------
possible and that none have been defined sufficiently (e.g., key events, dose-response and
temporal relationships) to inform low-dose extrapolation. Therefore, as discussed in Section
4.7.3, the EPA determined that the mode of action for PCP induction of liver tumors is unknown.
The Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend that linear
extrapolation of cancer risk be used as a default approach where the weight of evidence
evaluation is insufficient to establish the mode of action for a tumor site. Consistent with this
guidance, the EPA does not consider application of a nonlinear extrapolation approach to be
supported.
Comment: A commenter identified a 2009 paper on hydroxyl radical formation (Zhu and Shan,
2009) that was not included in the Toxicological Review.
Response: A summary of this paper was added to the Toxicological Review.
Comment: A commenter stated that because of the higher incidence of liver tumors with tPCP
than EC-7, the dioxin and furan contaminants that are at higher concentrations in tPCP may
contribute greatly to the increased incidence of liver tumors with tPCP, and it is possible that the
liver tumors induced by EC-7 could also be related to the contaminants and not PCP alone.
Response: In Section 5.4.4, specific consideration was given to the possible contributions of
impurities in tPCP and EC-7 on tumor response. While uncertainties associated with impurities
are acknowledged, the EPA concluded that the oral slope factor of 4 x 10"1 (mg/kg-day)"1 for
tPCP can be considered representative of the cancer risk associated with PCP alone. The
rationale for selecting tumor data for tPCP as the basis for characterizing PCP cancer risk in
Section 5.4.4 was clarified.
Comment: A commenter stated that there is no evidence that mouse pheochromocytomas are
relevant to the assessment of human risk because: (1) the incidence of malignant
pheochromocytomas was not statistically significantly increased in males or females, (2) benign
tumors are not considered relevant to humans, and (3) there is no epidemiology evidence to
support that pheochromocytomas can be induced in humans under any conditions, citing Elmore
et al. (2009) as support for this position. The commenter stated that recommendations of the
1990 Science Advisory Board (SAB) supported this opinion.
Response: Pheochromocytomas are catecholamine-producing neuroendocrine tumors. The
relevance of rodent pheochromocytomas as a model for human cancer risk has been the subject
of discussion in the scientific literature (e.g., Greim et al., 2009; Powers et al., 2008). In humans,
pheochromocytomas are rare and usually benign, but may also present as or develop into a
A-16
-------
malignancy (Eisenhofer et al., 2004; Lehnert et al., 2004; Salmenkivi et al., 2004; Edstrom-Elder
et al., 2003; Goldstein et al., 1999; Tischler et al., 1996). Rates of malignant transformation of
10% (Sweeney, 2009; Salmenkivi et al., 2004) to approximately 36% have been reported (Ohta
et al., 2006). Hereditary factors in humans have been identified as important in the development
of pheochromocytomas (Eisenhofer et al., 2004). Pheochromocytomas are more common in
laboratory rats, although evidence suggests that certain rat pheochromocytomas may have
similarity to human pheochromocytomas (Powers et al., 2008). Furthermore, mechanisms of
action inducing pheochromocytomas in rats are expected to occur in humans as well (Greim et
al., 2009).
Parallels between pheochromocytomas in the mouse and humans have led investigators
to suggest that the mouse might be an appropriate model for human adrenal medullary tumors
(Tischler et al., 1996). Like humans, the spontaneous occurrence of pheochromocytomas in the
mouse are relatively rare (<3%; Tischler et al., 2004, 1996), as are metastases. The
morphological variability of mouse pheochromocytomas and the morphology of the predominant
cells are comparable to those of human pheochromocytomas. An important characteristic of
mouse pheochromocytomas is expression of immunoreactive phenylethanolamine-N-methyl-
transferase (PNMT), the enzyme that produces epinephrine from norepinephrine; human
pheochromocytomas are also usually PNMT-positive (Tischler et al., 1996).
Elmore et al. (2009) states that there is no epidemiologic evidence that adrenal medullary
proliferative lesions can be induced in humans under any circumstances, but that the rodent
tumors express many of the same genes as their human counterparts and are potentially valuable
for mechanistic studies of the roles of those genes in tumor biology. No case-control or other
studies in humans that evaluated possible associations between pheochromocytomas and
environmental agents are available in the published peer-reviewed literature. Thus, while the
epidemiological literature does not provide evidence of pheochromocytoma induction by various
agents, it appears that no such studies have been performed.
The SAB committee stated that the increased incidence of pheochromocytomas and dose-
response pattern was related to PCP exposure. The committee noted that there is disagreement in
the interpretation of the meaning of pheochromocytomas in rodents and the diagnoses of these
lesions. The SAB questioned the human relevance of these tumors based on the fact that only
benign tumors were observed. However, the committee did not state that pheochromocytomas
are not relevant to humans.
No studies were identified to determine a mode of action for PCP-induced tumors of the
adrenal gland. Thus, the mode of action for pheochromocytomas observed following oral
exposure to PCP is unknown. Therefore, in the absence of information indicating otherwise,
adrenal gland tumors in rodents are considered relevant to humans.
A-17
-------
Comment: A commenter stated that the mode of action for hemangiomas and
hemangiosarcomas in female mice is consistent with oxidative stress, and that such a mode of
action would have a nonlinear dose-response. The commenter specifically cited research on
vinyl chloride, a chemical that induces hemangiosarcoma and is metabolized to chloroethylene
oxide, a mutagen that induces four adducts that have been shown to be induced by oxidative
stress as support for this determination.
Response: In the absence of any mechanistic data specific to the induction of hemangiomas and
hemangiosarcomas by PCP, a mode of action for this tumor type is unknown. Consistent with
the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), linear extrapolation of cancer
risk was applied to data for hemangiomas and hemangiosarcomas as a default approach where
the weight of evidence evaluation was insufficient to establish the mode of action.
Comment: A commenter stated that the most likely operative mode of action for mesotheliomas
of the peritoneal cavity, originating from the tunica vaginalis of the testes in male rats, was
oxidative stress. The commenter observed that these tumors were increased in male rats in the
stop-exposure component of the NTP (1999) study, but not the 2-year study. The commenter
also suggested that another mode of action contributing to this tumor type is hormonal imbalance
brought about by perturbations of the endocrine system, which is associated with the formation
of Leydig tumors of the testes that occur spontaneously at a high incidence in F344/N rats.
Response: The possible role of oxidative stress in PCP carcinogenicity is discussed in
Section 4.7.3. The EPA did not identify any literature that addresses PCP induction of
mesothelioma via hormonal imbalance. Chhabra et al. (1999) concluded that further studies are
needed to fully explain the molecular events leading to mesothelioma formation by PCP.
Mesothelioma in the male rat as observed in the stop-exposure study was included in the
evaluation of the overall weight of evidence for PCP carcinogenicity, but was not used as the
basis for slope factor derivation in this assessment.
Comment: A commenter raised doubts about the nasal squamous cell carcinomas observed in
the "stop-exposure" study in the rat (NTP, 1999) because the incidence of nasal tumors was not
increased in the full 2-year study, and further argued that this tumor was not relevant to humans
at current PCP exposure levels. Possible explanations offered for the increased tumor incidence
were direct contact of the nasal mucosa membrane with PCP vapor during feeding or to
PCP-containing feed dust, and oxidative damage.
Response: As noted by Chhabra et al. (1999), the nasal effects in the treated rats in the NTP
stop-exposure study may have been due to systemic exposure to PCP, direct contact of the nasal
A-18
-------
mucous membrane with PCP vapor during ingestion of feed, or PCP-containing dust from feed.
Studies providing support for PCP-induction of nasal cell carcinomas via direct contact with PCP
vapor or dust are not available. Nasal squamous cell carcinomas in the male rat as observed in
the stop-exposure study were included in the evaluation of the overall weight of evidence for
PCP carcinogenicity, but were not used as the basis for slope factor derivation.
Comment: A commenter suggested that the rat may be a more appropriate animal model than
the mouse for assessing risks of PCP to humans for the following reasons: (1) the high
spontaneous rate of liver adenomas/carcinomas in male mice, (2) the unusual sensitivity of the
B6C3Fi mouse, (3) a "surprisingly" low incidence of liver tumors in concurrent controls in the
mouse study, possibly due to low survival of control animals and smaller groups sizes (35 per
sex in the control group versus 50 in the treated groups), and (4) contribution of dioxins and
furans (at higher levels in tPCP) to the higher liver tumor incidence in mice in the tPCP study.
Further, the commenter observed that the incidence of liver tumors was not increased in
PCP-exposed F344 rats, which has a lower spontaneous incidence of liver tumors, and that such
an increase in the rat would have provided more convincing evidence that PCP is
hepatocarcinogenic.
This commenter observed that the incidence of hemangiomas/hemangiocarcinomas was
increased only in the female mouse (not in the male mouse or in rats of either sex) and only at
the highest doses tested in the studies of tPCP and EC-7.
Finally, this commenter stated that data from the rat study may be more relevant because
of the greater tendency for glutathione (an important detoxification mechanism of ROS)
depletion to occur in the mouse than the rat and human and the purity of PCP used in the rat
study compared to the mouse studies (99 and 90.4-91%, respectively).
Response: As noted in response to peer review comments under Cancer Charge Question #3, the
mouse model has been shown to be more sensitive to PCP carcinogenicity than the rat model. In
the absence of mode-of-action data to establish that the mouse model is not relevant to humans,
tumor data from the male mouse (specifically the combined risk of liver tumors and adrenal
gland pheochromocytomas) were used to derive the PCP cancer slope factor. Uncertainties
associated with the selection of this tumor data set for derivation of the slope factor are discussed
in Section 5.4.5.
In response to the comment regarding the findings for hemangiosarcomas/
hemangiocarcinomas in female mice, the EPA notes that the oral slope factor for PCP is based
on male mouse data.
The comment related to the contribution of dioxins and furans to the cancer response
observed in mice exposed to tPCP and EC-7 is addressed in the fourth public comment.
A-19
-------
Comment: A commenter claimed that the EPA ignored the recommendations of the 1990 SAB
related to the relevance or lack of relevance of tumors in the male and female mouse and with
regard to calculation of the oral slope factor, and suggested that disregarding available expert
advice can affect the credibility of the EPA's risk assessment process.
Response: The EPA performed the cancer assessment for PCP consistent with the Agency
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). Scientific understandings of
cancer and risk assessment practices have evolved in the two decades since the SAB's review of
the PCP assessment. It is not unexpected that some of the views and recommendations offered
by the SAB in 1990 might differ from current risk assessment practices.
Comment: A commenter cited several issues related to the EPA's derivation of the cancer slope
factor for PCP, including the following:
(1) A mode of action of oxidative stress has been shown for PCP. The commenter stated that
oxidative DNA damage is the most common endogenous DNA damage in cells, and that
exposure to PCP results in the formation of additional identical adducts. For agents
acting through this mode of action, a nonlinear approach should be used. To that end, the
commenter stated that the draft assessment ignored the U.S. EPA's (2005a) Guidelines
for Carcinogen Risk Assessment, under which the preferred method for risk assessment is
use of a biologically-based model that incorporates mode-of-action considerations.
(2) A slope factor based derived by combining adrenal and liver tumors is inconsistent with
the recommendations of the 1990 SAB.
(3) The slope factor based on data for tPCP is recommended for use with pure PCP, in spite
of the fact that EC-7 was indicated to have lower levels of dioxins and furans.
(4) The EPA chose to apply defaults of the risk assessment methodology, including the
assumption that humans were more sensitive than the most sensitive species (mouse).
The commenter reiterated the position related to the lack of relevance of mouse liver
tumors, that the benign tumor response in the mouse liver and adrenal are more reflective
of an epigenetic or nongenotoxic mode of action (proposing as modes of action a
sustained increased cellular turnover and hormonal challenge), and that the incidence of
hemangiomas/hemangiosarcomas may be increased due to oxidative stress that occurs at
high exposures or related to by the contaminants.
Response: Reponses to comments related to the use of a nonlinear analysis, concerns about
dioxin and furan impurities in tPCP, and tumor relevance are provided in responses to Charge
Question D.3 and other public comments (e.g., public comments 2, 4, 5, 6, 9, and 10).
The derivation of an oral slope factor based on the combined risk of liver and adrenal
gland tumors is consistent with the recommendations in the U.S. EPA's Guidelines for
Carcinogen Risk Assessment (2005a). As discussed in Section 5.4.5.1, given the multiplicity of
A-20
-------
tumors sites associated with PCP exposure, an oral slope factor based on one tumor site may
underestimate the carcinogenic potential of PCP.
Comment: A commenter questioned whether the Guidelines for Carcinogen Risk Assessment
(U.S. EPA, 2005a) had been followed, raising specific questions about the choice of
epidemiologic studies that the EPA relied on for the determination of the weight-of-evidence
descriptor in the assessment. This commenter stated that "it is unknown (and unexplained) why
the IRIS conclusions only relied on four studies (out of the many available)."
Response: The EPA evaluated a large number of studies with data pertaining to chlorophenols.
Section 4.1 presents summaries of the studies that included specific information relevant to PCP.
Although all of these studies were considered in the evaluation, the studies cited in the weight-
of-evidence narrative (Section 4.7) were the sawmill worker cohort study of cancer incidence
and mortality that included specific evaluation of PCP as well as trichlorophenol (Demers et al.,
2006) and the case-control studies with detailed PCP assessment (see Tables 4-3 and 4-4).
Cohort studies with nonspecific exposure assessment, cohorts set in manufacturing plants, and
case-control studies with limited PCP assessment were given little weight in the overall cancer
evaluation. The relative strengths and limitations of these sets of studies are described in
Section 4.1.
Comment: A commenter discussed the role that tests of statistical significance should play in the
evaluation of the body of evidence, stating that "Clearly, only statistically significant
associations between exposure and outcomes are judged to be relevant."
Response: EPA disagrees with the commenter's interpretation of the role of statistical
significance in terms of determining the relevance of the results of a study. Statistical
significance reflects the magnitude of the observed effect and the precision of the estimate. The
statistical power of the study (i.e., the probability of correctly identifying a true effect of a
specified size) should also be considered in the evaluation of the data; statistical power is directly
related to the size of the study (e.g., number of cases in a case-control study) and prevalence of
exposure. A study with a low power or probability of detecting a statistically significant result
should not necessarily be interpreted as a null or "no effect" study. As noted in the U.S. EPA's
Guidelines for Carcinogenic Risk Assessment (2005a), other considerations, including the
magnitude of the point estimate, the precision of this estimate, the appropriateness of the
statistical test that was used, the design of the study, the likelihood of exposure misclassification
leading to an attenuation of observed effects, and potential confounders should also be
considered (see, for example, Guidelines for Carcinogen Risk Assessment [U.S. EPA, 2005a],
pp. 2-3, in which is it noted that "Null results from epidemiologic studies alone generally do not
A-21
-------
prove the absence of carcinogenic effects because such results can arise either from an agent being
truly not carcinogenic or from other factors such as: inadequate statistical power, inadequate study
design, imprecise estimates, or confounding factors." These ideas are elaborated further in
Sections 2.2.1, Human Data, of the 2005 Cancer Guidelines).
Comment: A commenter also noted that for at least one of the four key studies relied on by the
EPA in the cancer weight-of-evidence evaluation, statistical significance was either not
considered or minimized, based on the presentation in Table 4-1 of results that were of a defined
magnitude (SMR > 1.5).
Response: The intent of Table 4-1 is to present a summary of the data relating to cancer risk
from the available cohort studies. Each of the studies is discussed in more detail in the text, and
in the case of the specific study noted by the commenter, the Demers et al. (2005) cohort study of
sawmill workers, an additional table (Table 4-2) is provided because of the breadth of the data
available from that study (i.e., both incidence and mortality data analyzed using a 4-group
exposure categorization, with an internal comparison group). The summary provided in
Table 4-1, "Elevated risk of non-Hodgkin's lymphoma and multiple myeloma incidence and
mortality; evidence of exposure-effect response; weaker or no risk seen with TCP (see Table 4-
2)" summarizes the data and points the reader to the more extensive presentation of the Demers
et al. (2005) results.
Comment: Referring to the body of research on phenoxy herbicides and chlorophenols, a
commenter noted that in some manufacturing scenarios, trichlorophenol rather than PCP was
used in the production process. The commenter notes that "only when such studies demonstrate
that exposure to PCP was explicitly considered can results be afforded greater weight in a weight
of evidence (WOE) analysis."
Response: The EPA agrees that studies focusing specifically on PCP should be the basis of the
analysis. Accordingly, the criteria used in the selection of studies emphasized the availability of
PCP-specific data (see description in Section 4.1.1.1), and relatively little weight was given to
the cohort study by Ramlow et al. (1996) set in a PCP manufacturing plant that did not include
detailed exposure assessment specific to PCP (see discussion in Section 4.1.1.4).
Comment: With respect to the case-control study of non-Hodgkin's lymphoma by Hardell et al.
(1994), a commenter noted a strong association observed with high-grade exposure to PCP
(defined as one or more week continuous exposure or one or more month total exposure) with an
OR of 8.8 (95% CI 3.4-24), but stated that such a high value for an OR is "not plausible and the
wide confidence intervals also cast some doubt on the validity of the findings." The commenter
A-22
-------
also stated that it is not clear if any of the chlorophenol-exposed subjects were also exposed to
the phenoxyacetic acids 2,4-D and 2,4,5-T, and questioned why the authors stated that most
subjects had been exposed to 2,4-D and 2,4,5-T when the overall prevalence of this exposure was
47 out of 105 cases and 51 out of 335 controls.
Response: The EPA agrees that the imprecision of the estimated association results in
considerable uncertainty with respect to whether a three-, four-, eightfold or higher risk was seen
in the high-PCP exposure group, but the data nonetheless indicate an increased risk. With
respect to co-exposure with the phenoxyacetic acids, the data indicate some overlap in these
groups (among 105 cases, 35 exposed to chlorophenols, 25 exposed to phenoxyacetic acids, and
47 exposed to either chlorophenols or phenoxyacetic acids; among 355 controls, 35 exposed to
chlorophenols, 24 exposed to phenoxyacetic acids, and 51 exposed to either chlorophenols or
phenoxyacetic acids). The similarity in these patterns among cases and controls, however,
argues against confounding as an explanation for the observed association with PCP. Hardell et
al. (1994) state in the results section that "Mostly, a combination of 2,4-D and 2,4,5-T had been
used in both occupational and leisure time exposure" and the statement in the abstract that "Most
cases and controls were exposed to a commercial mixture of 2,4-dichlophenoxyacetic acid and
2,4,5-trichlorophenoxyacetic acid" appear to be based on the data from Table 2 of the paper
showing the breakdown in frequency of specific phenoxyacetic acids, in which only three cases
and one control were exposed only to 2,4-D.
Comment: With respect to the nested case-control study of soft tissue sarcoma (11 cases) and
non-Hodgkin's lymphoma (32 cases) within 24 cohort studies conducted in 11 countries by
Kogevinas et al. (1995), the commenter summarized the results for chlorophenols and for PCP.
The association between any PCP exposure and non-Hodgkin's lymphoma was OR 2.75 (95%
CI 0.45-17.0), and the association between high PCP and non-Hodgkin's lymphoma was OR
4.19 (95% CI 0.59-29.59). The commenter observed that "[t]o the extent that this study was
able to isolate exposure to only chlorophenols or PCP specifically, there was no significant
increase in either STS [soft tissue sarcoma] or NHL [non-Hodgkin's lymphoma] associated with
exposure. These results suggest that neither exposure to chlorophenols, or to PCP in particular,
is associated with increased risk of STS or NHL."
Response: Given the limited number of observed cases of either disease in Kogevinas et al.
(1995) and the magnitude of the point estimates for the association between PCP and non-
Hodgkin's lymphoma (i.e., a three- or fourfold increased risk), it is inappropriate to characterize
the observed results as evidence of no association. As noted in the Guidelines for Carcinogenic
Risk Assessment (U.S. EPA, 2005a), pp. 2-3, the inadequate statistical power is a key factor that
should be considered in the interpretation of these data. In addition, as discussed in Section
A-23
-------
4.1.1.3, the different pattern of results seen for the chlorophenols other than PCP and for
phenoxy herbicides also suggests relative specificity of effects for PCP.
Comment: A commenter summarized the results of the comparisons between cancer rates
among the workers in a cohort study by Demers et al. (2006) and reference rates in British
Columbia, Canada and a set of exposure-response analyses using an internal referent group
within the cohort. These analyses indicated an association with non-Hodgkin's lymphoma,
multiple myeloma, and kidney cancer, but not with soft tissue sarcoma.
Response: The EPA agrees with the study summary, but also notes the additional evidence
concerning liver cancer seen in this cohort study.
Comment: A commenter provided an evaluation of the findings of Hardell et al. (1994, 1995),
Kogevinas et al. (1995), and Demers et al. (2006) in terms of strength of association, consistency
of association, dose-response relationship, temporality, specificity of association, and biological
plausibility. The commenter stated that only statistically significant associations between
exposure and outcomes were judged to be relevant and that strength of association refers
primarily to size of the RR, which must reach statistical significance.
Response: Both statistical significance and statistical power should be considered when
interpreting results of a study. To dismiss from consideration all results that do not reach
statistical significance, particularly those from low-powered studies, such as the Kogevinas et al.
(1995) study (with an odds ratio for the association between high PCP and non-Hodgkin's
lymphoma of 4.19 [95% CI 0.59-29.59], based on 32 observed cases) would be to dismiss pieces
of evidence that should be considered in the weight-of-evidence evaluation. EPA evaluate a
study or collection of studies with consideration of the magnitude of effects, precision of effect
estimates, the probability of the observed results under the null (no effect) hypothesis, and the
statistical power of the study with respect to specific alternative hypotheses.
Comment: A commenter stated that the Demers et al. (2006) cohort study found no significant
association or no association between exposure to all chlorophenols (i.e., PCP and
trichlorophenol) and non-Hodgkin's lymphoma, multiple myeloma, soft tissue sarcoma, and liver
cancer.
Response: These statements apply when only the results using the external comparison group,
i.e., the SIR and SMR data using the British Columbia population as a referent group (Table 2 of
Demers et al., 2006) are evaluated. The data from the internal cohort comparisons are important
in the evaluation of the association between exposure to chlorophenols and non-Hodgkin's
A-24
-------
lymphoma, multiple myeloma, soft tissue sarcoma, and liver cancer. Tables 3-7 of Demers et al.
(2006) present the more extensive analyses developed specifically for PCP and trichlorophenol,
rather than combining the exposures into the single category (along with unexposed individuals
within the plant) that was used in the analysis with the external comparison group. The use of an
internal comparison group reduces the likelihood of potential confounders affecting the results.
The EPA used these internal exposure-response analyses in the evaluation of the study (see Table
4-2).
Comment: A commenter stated that the strong associations between non-Hodgkin's lymphoma
and PCP seen in the Hardell et al. (1994) study were too strong and imprecise to be believable.
The commenter also stated that the results were more likely due to exposure to 2,4-D and
2,4,5-T.
Response: As noted previously, the EPA agrees that the imprecision of the estimated association
results in uncertainty regarding the magnitude of the observed risk, but does not consider this
imprecision to negate the presence of an increased risk. In addition, there is no indication of a
disproportionately higher rate of co-exposure with phenoxyacetic acids among cases compared
with controls.
Comment: A commenter provided a reference to a case-control study of non-Hodgkin's
lymphoma (Hardell and Eriksson, 1999) that had not been included in the Toxicological Review.
Response: A summary of this study was added to Section 4.1.1; the findings of this study are
also included in the discussion in Section 4.7. This case-control study included 404 male cases
age >25 years diagnosed with non-Hodgkin's lymphoma between 1987 and 1990 in northern
Sweden. The association seen with PCP exposure was OR 1.2, 95% CI 0.7-1.8. PCP use had
been banned in Sweden in 1977, so the exposure time period in relation to timing of diagnosis
differs in this study compared with the earlier studies from Sweden.
Comment: With respect to the discussion of "consistency," a commenter stated that consistency
refers to the presence of a [statistically] significant association in studies of similarly exposed
populations.
Response: The EPA does not view study results as dichotomous (i.e., either the presence or
absence of statistical significance); evaluating the consistency of results based only on the
presence or absence of statistical significance would be of limited value, and given the numerous
factors that should be considered when evaluating the consistency of results (see response on
p. A-21), potentially misleading. As noted in the discussion of the evaluation of consistency
A-25
-------
among study results in the Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a),
p. 2-13, "If there are discordant results among investigations, possible reasons such as differences in
exposure, confounding factors, and the power of the study are considered." Thus EPA evaluates the
magnitude and precision of the estimates in conjunction with relevant issues regarding the study
size (power), exposure measures, potential confounding factors, and other study design issues
when evaluating the consistency of results among studies.
Comment: A commenter discussed the available epidemiological studies of PCP and non-
Hodgkin's lymphoma, multiple myeloma, soft tissue sarcoma, and liver cancer. The summaries
are presented in terms of the presence or absence of a statistically significant association, and a
specific disease that was not examined in a study is described as being "not mentioned" by that
study (e.g., the commenter states that a case-control study of non-Hodgkin's lymphoma did not
mention multiple myeloma). The non-Hodgkin's lymphoma literature is summarized as showing
a significant association in Hardell et al. (1994) and no association in Hardell and Eriksson
(1999). The commenter concluded that there is a lack of consistency in the association for this
endpoint.
Response: The Demers et al. (2006) study reported an increased incidence and increased
mortality of non-Hodgkin's lymphoma (trends-values = 0.03 for incidence and 0.06 for
mortality; an approximate twofold increased risk in the two highest exposure groups in both
analyses). There is a difference in time period between the two Swedish case-control studies.
The earlier study (Hardell et al., 1994), in which strong associations were seen, was conducted in
cases diagnosed between 1974 and 1978; the later study (Hardell and Eriksson, 1999) was
conducted in cases diagnosed between 1987 and 1990. PCP use had been banned in Sweden in
1977, which would be expected to result in a considerably different set of exposure conditions.
All of these factors were considered in the evaluation of the consistency of the results.
Comment: With respect to multiple myeloma, a commenter stated that the studies cited as the
basis for the weight-of-evidence conclusion did not show a significant association with multiple
myeloma.
Response: The Demers et al. (2006) cohort study of sawmill workers shows an exposure-
response trend for both incidence (trends-value = 0.03) and mortality (trends-value = 0.02).
The risk ratios in the highest category of exposure were strong (>4.0), and there was no evidence
of similar patterns in the analyses of TCP exposure. None of the other studies cited examined
multiple myeloma. The Toxicological Review also described the study by Pearce et al. (1986a)
of farming-related exposures and multiple myeloma risk in New Zealand (76 cases, 315 controls
drawn from a population cancer registry). This study demonstrated that there was little evidence
A-26
-------
of an association with the general category of chlorophenol exposure (OR 1.1, 95% CI 0.4-2.7)
and work in a sawmill or timber merchant (OR 1.1, 95% CI 0.5-2.3) and stronger associations
were seen with a history of doing fencing work (OR 1.6, 95% CI 0.9-2.7) and jobs that involved
potential exposure to chlorophenols at a sawmill or timber merchant (OR 1.4, 95% CI 0.5-3.9).
Because of the limited information pertaining specifically to PCP in this study, it was not cited in
the weight-of-evidence summary.
Comment: For soft tissue sarcoma, a commenter stated that the association between exposure to
chlorophenols and PCP reported by Hardell et al. (1995) was not replicated in a later study
(Hardell and Eriksson, 1999), which was not cited in the Toxicological Review. In addition, the
commenter also noted that the study by Kogenias et al. (1995) showed no increase and that soft
tissue sarcomas were not mentioned in Hardell et al. (1994). The commenter considered that
these findings demonstrated a lack of consistency.
Response: The 1999 Hardell and Eriksson study did not fail to replicate the findings of the meta-
analysis of four case-control studies of soft tissue sarcoma presented by Hardell et al. (1995), but
rather, did not examine this question. Hardell and Eriksson (1999) is a case-control study of
non-Hodgkin's lymphoma, and thus does not provide any data regarding soft tissue sarcoma.
While Kogenias et al. (1995) did not show an association between PCP exposure and soft tissue
sarcoma risk, it should be noted that this analysis was based on only 11 cases.
Comment: A commenter described the results of specific diseases in terms of the statistical
significance of trend tests. For example, for non-Hodgkin's lymphoma observed in the Demers
et al. (2006) study, the commenter indicated that the study authors showed a significant dose-
response trend with incidence, but not with mortality and significant dose-response trends in the
incidence and mortality analyses that were lagged by 10 or 20 years. The commenter also
summarized the Demers et al. (2006) data with respect to multiple myeloma by noting "a
significant dose-response trend" in the incidence and mortality analyses in the lagged and
unlagged data.
Response: The argument that the incidence data show a significant dose-response trend that is
not seen in the mortality data rests solely on the statistical significance of the trend test, which is
p = 0.03 for the incidence data and 0.06 for the mortality data (see Table 4-2 of the Toxicological
Review). The EPA considered that both the incidence and the mortality data, in the lagged and
unlagged analyses, provide evidence of exposure-response trends, with approximately a twofold
increased risk in the highest two categories of exposure. The attenuation of the exposure-
response seen in the highest exposure category is commonly seen in epidemiologic studies of
occupational cohorts (Stayner et al., 2003). Further, the characterization of the pattern of
A-27
-------
response across exposure groups should not be based solely on the presence or absence of a test
for linear trend that is statistically significant at a specified alpha level. The actual pattern of
response should be examined when characterizing the data. For example, in addition to the trend
/7-value, Demers et al. (2006) observed approximately a fourfold increased risk of multiple
myeloma in the highest exposure group (see Table 4-2).
Comment: A commenter stated that none of the studies reported a significant dose-response
trend between soft tissue sarcoma and PCP.
Response: Because of the very low incidence of soft tissue sarcoma, the data available to
evaluate dose-response patterns are quite limited. The Kogevinas et al. (1995) nested case-
control study is based on 11 observed cases, and the Demers et al. (2006) cohort study is based
on 23 observed cases, with only five cases in the highest three quartiles of exposure. In Demers
et al. (2005), data suggest an inverse association (lower risk with higher exposure), but there is
considerable uncertainty in this analysis given the limited number of observations. Case-control
studies provide a study design that addresses the inherent limitations of cohort studies for
examining rare outcomes. The Hardell et al. (1995) meta-analysis of four case-control studies
included 434 observed cases of soft tissue sarcoma and observed a strong association between
high exposure to PCP, defined as 1 week or more continuously or at least 1 month in total, and
soft tissue sarcoma risk (OR = 2.8, 95% CI 1.5-5.4).
Comment: A commenter noted that the Demers et al. (2006) study is the only study with data
relating to liver cancer and that the data show no increases in this cancer or dose-response or
latency trends with exposure to PCP.
Response: Table 4-2 presents the data for the Demers et al. (2006) study. The internal cohort
exposure-response analyses was the primary focus of the Demers et al. (2006) study. As stated
above in the previous discussion of non-Hodgkin's lymphoma, an attenuation in the highest
exposure group was observed. Specifically, relatively strong associations (i.e., at least a
doubling of the risk in almost all of the exposure categories) were observed. The EPA concluded
that these data do not support the conclusion that there is no evidence of an association with liver
cancer.
Comment: A commenter noted inconsistencies in the results reported by Demers et al. (2006),
Hardell et al. (1995, 1994), and Kogevinas et al. (1995), and concluded that the criterion for
specificity of association (i.e., a single effect being produced by a particular exposure) was not
met.
A-28
-------
Response: The EPA agrees that the difference in the results among these studies is an important
consideration and the summaries of these studies therefore describe these data as providing some
evidence of carcinogenicity. In addition, the differences between results among these studies are
described. It is also important to note the methodological differences in the studies, specifically,
that the meta-analysis (Hardell et al., 1995) included 434 cases compared with the 23 observed
cases in Demers et al. (2006) and 11 observed cases in Kogevinas et al. (1995).
Comment: A commenter suggested that the data show a lack of site concordance between
animal and human studies (i.e., tumors seen in experimental exposure studies in rats or mice
correlate with the type of tumors seen in humans). This commenter further noted that
hepatocellular carcinoma, adrenal medullary neoplasms, and hemangiosarcomas (a histologic
form of soft tissue sarcoma) were seen in the animal studies and, thus, would be the expected
types of cancers that would be seen in epidemiological studies of PCP-exposed populations. The
commenter stated that the observation of hemangiosarcomas in rats is in contrast to the overall
weight of evidence in human studies suggesting that exposure to PCP is not associated with
increased risk of soft tissue sarcoma.
Response: The EPA concluded that the human studies provide some evidence of soft tissue
sarcoma and limited evidence of liver cancer associated with PCP exposure. In addition, the lack
of site concordance between animals and humans does not necessarily support a lack of
biological plausibility.
Comment: A commenter maintained that site-specificity is always found in epidemiologic
studies of chemical carcinogens. Specifically, that human response to exposures to carcinogens
are consistent (i.e., of the same type of nature) and that chemical carcinogens display target
organ specificity.
Response: Variability in the type of tumors observed in both human and animal studies is
common in studies of carcinogens. Many factors can influence the effect of a carcinogen in
human populations, including genetic susceptibility, nutritional status, and co-exposures.
Comment: A commenter noted that the summary statement in Section 5 regarding the basis for
the cancer weight-of-evidence descriptor as "likely to be carcinogenic to humans" by all routes
of exposure based on inadequate evidence from human studies and adequate evidence from
animal studies is inconsistent with the discussion presented in Section 4.7.
Response: The summary statement in Section 5 was revised to better reflect the discussion in
Section 4.7.
A-29
-------
APPENDIX B: PHYSIOCHEMICAL DATA FOR PCP AND THE IDENTIFIED
TECHNICAL- AND COMMERCIAL-GRADE CONTAMINANTS
Table B-l. Physicochemical data for dioxin contaminants of PCP
General chemical
formula
CgHClsO
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HXCDD
1,2,3,6,7,8-HXCDD
1,2,3,7,8,9-HXCDD
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7,8,9-OCDD
Common name
PCP
Pentachlorodibenzo-p-dioxin
HxCDD
HxCDD
HxCDD
Heptachlorodibenzo-p-dioxin
OCDD
Vapor
pressure
(mm Hg)
0.00415
4.4 x IQ-10
3.8 x 1Q-11
3.6 x 1Q-11
4.9 x 1Q-11
5.6 x ID'12
8.25 x ID'13
Water solubility
at 25°C
(mg/L)
14
0.000118
4.42 x 1Q-6
4.42 x 1Q-6
4.42 x 1Q-6
2.4 x 1Q-6
7.4 x 1Q-8
Henry's law
constant
(atm x m3/mol)
0.079
2.6 x 1Q-6
1.7 x 1Q-5
1.7 x 1Q-5
1.7 x 1Q-5
1.26 x 1Q-5
6.75 x ID'6
LogK,,w
-
6.64
7.8
7.8
7.8
8.0
8.2
Table B-2. Physicochemical data for furan contaminants of PCP
General chemical
formula
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
2,3,4,7,8-PCDF
1,2,3,4,6,7,8,9-OCDF
Common name
Pentachlorodibenzofuran
Pentachlorodibenzofuran
Hexachlorodibenzofuran
Hexachlorodibenzofuran
Hexachlorodibenzofuran
Heptachlorodibenzofuran
Heptachlorodibenzofuran
Pentachlorodibenzofuran
Octachlorodibenzofuran
Vapor
pressure
(mm Hg)
1.7 x 10'9
2.6 x 10'9
2.4 x lO'10
2.2 x lO'10
2.0 x 10'10
3.5 x KT11
1.07 x 10'10
No data
3.75 x 10'12
Water solubility
at 25°C
(mg/L)
-
2.36 x 10'4
8.25 x 10'6
1.77 x 10'5
No data
1.35 x 10'6
No data
No data
1.16 x 10'6
Henry's law
constant
(atm x m3/mol)
-
4.98 x 10'6
1.43 x 10'5
7.31 x 10'6
No data
1.41 x 10'5
No data
No data
1.88 x 10'6
LogK,,w
6.79
6.5
7.0
7.0
7.0
7.4
No data
No data
8.0
B-l
-------
Table B-3. Average daily dose of PCP (mg/kg) and contaminants (ug/kg) to
B6C3Fi mice in the 2-year feeding study
PCP/contaminant
PCPa
Trichlorophenol
TCP
HCB
TCDD
HxCDD
Heptachlorodibenzo-p-dioxin
OCDD
Pentachlorodibenzofuran
Hexachlorodibenzofuran
Heptachlorodibenzofuran
Octachlorodibenzofuran
Heptachlorohydroxydiphenyl
ether
Octachlorohydroxydiphenyl ether
Nonachlorohydroxydiphenyl
ether
Hexachlorohydroxydibenzofuran
Heptachlorohydroxydibenzofuran
Males
100 ppm
tPCP
18
1.1
430
0.6
-
0.11
3.3
15.6
0.016
0.11
1.0
0.5
10
220
400
20
50
EC-7
18
0.8
1,100
0.7
-
0.002
0.006
0.008
-
0.001
0.002
-
-
-
-
-
-
200 ppm
tPCP
35
2.3
860
1.1
-
0.23
6.7
31
0.03
0.24
2.0
1.0
20
430
800
40
110
EC-7
37
1.6
2,100
1.5
-
0.004
0.01
0.02
-
0.003
0.003
-
-
-
-
-
-
600 ppm
EC-7
118
4.7
6,300
4.4
-
0.01
0.04
0.05
-
0.009
0.01
-
-
-
-
-
-
Females
100 ppm
tPCP
17
1.1
415
0.54
-
0.11
3.2
15.1
0.014
0.11
1.0
0.5
10
210
390
20
50
EC-7
17
0.8
1,000
0.7
-
0.002
0.006
0.008
-
0.001
0.002
-
-
-
-
-
-
200 ppm
tPCP
35
2.2
830
1.1
-
0.22
6.5
31
0.03
0.22
1.9
1.0
20
420
780
30
100
EC-7
34
1.5
2,000
1.4
-
0.004
0.01
0.02
-
0.003
0.003
-
-
-
-
-
-
600 ppm
EC-7
114
4.6
5,800
4.2
-
0.01
0.03
0.05
-
0.008
0.01
-
-
-
-
-
-
"Daily dose in mg/kg body weight.
Source: NTP(1989).
B-2
-------
APPENDIX C: PCP LEVELS IN OCCUPATIONALLY EXPOSED HUMANS
Table C-l. PCP levels in worker and residential populations (with >15 individuals per group)
Population, location
Serum or plasma
n
Mean
(Range or standard
deviation)
Unit
Urine
n
Mean
(Range or
standard
deviation)
Unit
Reference
Occupationally exposed workers
Hawaii
Worker sample
Exposed (pesticide operators)
Nonexposed (other workers)
Population sample
Occupational exposures
No occupational exposures
Hawaii
Exposed (open vat wood treaters)
Exposed (pressure tank wood treaters)
Farmers (mixed pesticides exposure)
Controls (no occupational exposure)
United Kingdom
Exposed (formulators)
Exposed (sprayers)
Exposed (timberyard operators)
Nonexposed (furniture makers)
22
24
280
32
29
108
68
61
3.78
1.72
0.25
0.32
1.3
6.0
4.8
0.2
(4.00)
(2.02)
(0.88)
(1.26)
(0.4^.8)
(0.2-29.0)
(0.3-45.0)
(0.1-0.6)
ppm
ppm
ppm
ppm
mmol/L
mmol/L
mmol/L
mmol/L
130
117
121
173
18
23
210
32
26
112
54
—
1,802
40
465
44
0.95
0.27
0.01
0.03
39.6
274
74.0
(3-35,700)
(ND-1,840)
(3-38,642)
(3-570)
(1.93)
(0.56)
(0.05)
(0.18)
(7.4-300)
(11-1,260)
(5-655)
Not measured
ppb
ppb
ppb
ppb
ppm
ppm
ppm
ppm
nmol/mmol
creatinine
nmol/mmol
creatinine
nmol/mmol
creatinine
nmol/mmol
creatinine
Bevenue et al.
(1967)
Klemmer
(1972)
Jones et al.
(1986)
C-l
-------
Table C-2. PCP levels in occupationally exposed populations (with >15 individuals per group)
Population, location
Serum or plasma
n
Unit
Mean
(Range or standard
deviation)
Urine
n
Unit
Mean
(Range or standard
deviation)
Reference
Residential or work site exposure3
United States
Exposed (residential)
Exposed (telephone line workers)
Nonexposed
Germany
Exposed
Nonexposed
Germany
Exposed
Nonexposed
123
13
34
65
106
15
15
ppb
ppb
ppb
ug/L
ug/L
ug/L
ug/L
420
110
40
35.9
9.5
43.6
11.8
(39-1.340)
(26-260)
(15-75)
(20.7-133)
(2.8-19.3)
(31.2)
(4.5)
118
143
117
ppb
ppb
ng/mg
creatinine
69
3.4
3.1
(1-340)
(1-17)
(1-12)
Cline et al.
(1989)
Gerhard et al.
(1999)
Peper et al.
(1999)
General population
United States
NHANES III
ug/L
MS/g
creatinine
2.5
1.8
(Not detected-55)
(Not detected-29)
Hilletal.
(1995)
""Residents of homes or workers in work places in which PCP was used as a wood preservative on logs or wood used in the construction of these sites.
C-2
-------
APPENDIX D: DOSE-RESPONSE MODELING OF CARCINOGENICITY DATA FOR
PENTACHLOROPHENOL
D.I. METHODS
The multistage model included in EPA's BMD software (version 1.3.2) was fit to the
censored incidence data for selected tumors in male and female B6C3Fi mice exposed to either
tPCP or commercial (EC-7) grade PCP in the diet for 2 years (NTP, 1989). The raw and
censored incidence data are shown in Table D-l. Models were run restricting the fitted
parameters to be positive, in order to fit a monotonically increasing dose-response relationship.
The highest degree polynomial modeled for any data set was one less than the number of dose
groups. For each data set, successive decreasing polynomial degrees (down to the one-degree)
were modeled as well. Fit of a model to the data was assessed by the %2 goodness-of-fit test. A
X2/'-value >0.1 was considered to be an adequate fit (U.S. EPA, 2000b). Following the U.S.
EPA (2000b) methodology for the multistage model, the lowest degree polynomial that provided
adequate fit was selected as the source of the risk estimate for that data set. As recommended in
the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), a benchmark response
(BMR) near the lower end of the observed data, generally a 10% increase in extra risk, was used.
Table D-l. Incidence of tumors in B6C3F! mice exposed to tPCP and EC-7
in the diet for 2 years
Tumor type
Males (mg/kg-d)a
Hepatocellular adenoma/
carcinoma
Adrenal benign/malignant
pheochromocytoma
Females (mg/kg-d)a
Hepatocellular adenoma/
carcinoma
Adrenal benign/malignant
pheochromocytoma
Hemangioma/hemangio-
sarcoma
tPCP
(ppm in diet)
0
0
7/32b
(7/28)d
0/3 lb
(0/26)
0
3/33
(3/31)
2/33
(2/31)
0/3 5b
(0/33)
100
18
26/47c
(26/46)
10/45C
(10/41)
17
9/49
(9/49)
2/48
(2/48)
3/50
(3/50)
200
35
37/48c
(37/46)
23/45c
(23/44)
35
9/50
(9/48)
1/49
(1/47)
6/50c
(6/48)
EC-7
(ppm in diet)
0
0
6/3 5b
(6/33)
l/34b
(1/32)
0
l/34b
(1/34)
0/3 5b
(0/35)
0/3 5b
(0/35)
100
18
19/48C
(19/45)
4/48
(4/45)
17
4/50
(4/49)
2/49
(2/48)
1/50
(1/49)
200
37
21/48C
(21/38)
21/48C
(21/39)
34
6/49
(6/49)
2/46
(2/46)
3/50
(3/50)
600
118
34/49c
(34/47)
45/49c
(45/47)
114
31/48C
(31/48)
38/49c
(38/49)
9/49°
(9/49)
aAverage daily doses estimated by the researchers.
bStatistically significant trend (p < 0.05) by the Cochran-Armitage trend test.
Statistically significant difference from controls (p < 0.05) by Fisher's exact test.
dCensored data used for modeling are shown in parentheses; see text for description of censoring procedure.
Source: NTP (1989).
D-l
-------
Although survival was considered by NTP (1989) to be adequate for evaluation of
carcinogenicity in all groups, there were two survival-related issues that were considered for
potential impact on the dose-response assessment. First, males in the control group for the tPCP
study had unusually low survival, starting early in the study (first death at 15 weeks) and
continuing to termination. Survival at termination was only 34%, compared with 71% in the
EC-7 control males. The first hepatocellular tumor in this control group was observed in an
animal that died at 48 weeks and the second in an animal that died at 60 weeks. Hepatocellular
tumors in the low- and high-dose male tPCP groups were first observed at 59 and 54 weeks,
respectively. These findings suggest that survival as short as 48 weeks was adequate for
evaluation of liver tumors in the male mice. Despite the overall low survival and early onset of
mortality in the male tPCP control group, there were still only five deaths that occurred in
animals younger than 48 weeks. This compares to two deaths each in the low- and high-dose
male tPCP groups in the same time frame. Therefore, survival issues in the control male tPCP
group are expected to have little or no impact on the dose-response assessment.
The second survival-related issue was an increase in deaths occurring between weeks 40
and 80 in male mice in the mid-dose group in the EC-7 study (11 deaths, compared with 5 in
controls, 7 in the low-dose group, and 4 in the high-dose group). Neither hepatocellular nor
adrenal tumors were seen in any of these deaths among the mid-dose males. The earliest
appearance of these tumors in the male EC-7 study was 77 weeks for hepatocellular tumors and
66 weeks for adrenal pheochromocytomas, both in the high-dose group. However, as discussed
above, hepatocellular tumors were seen as early as 48 weeks in untreated males in the tPCP
study. Therefore, animals that died between 40 and 80 weeks in the EC-7 study were likely at
risk of developing tumors, and the greater number of such animals in the mid-dose group versus
the other groups is considered to be of little or no consequence for dose-response assessment.
Because survival issues were not expected to impact the dose-response assessment
significantly, time-to-tumor modeling was not performed. However, as a standard adjustment to
prevent counting animals that were never at risk of developing tumors, the incidence data were
censored to remove animals that died before appearance in the experiment of the first tumor of
the type in question in animals of the same sex and species (or 1 year, whichever occurred
earlier).
Statistical analysis (Fisher's exact and %2 tests of 2 x 2 contingency tables) showed no
difference in proportion of responders between male controls in the tPCP and EC-7 experiments
for hepatocellular adenoma/carcinoma or adrenal benign/malignant pheochromocytoma, or
between female controls in the tPCP and EC-7 experiments for hepatocellular adenoma/
carcinoma, adrenal benign/malignant pheochromocytoma, or hemangioma/hemangiosarcoma.
Therefore, dose-response analyses for each chemical formulation were conducted using the
combined control groups.
D-2
-------
In the NTP (1989) study, tumors were increased by PCP exposure at multiple sites — the
liver and adrenal gland in both male and female mice. The females had increased circulatory
tumors as well. There is a concern that in this situation a risk estimate based solely on one tumor
type may underestimate the overall cancer risk associated with exposure to the chemical.
D.2. RESULTS
The BMD modeling results for the individual data sets are summarized in Table D-2.
This table shows the BMDs and BMDLs derived from each endpoint modeled. BMDs and
BMDLs presented in this table are those predicted by the multistage model selected according to
the U.S. EPA (2000b) BMD methods, at 10% extra risk. All data sets were run using combined
control groups. Note that all risk estimates presented here are for mice; they have not been
converted to human equivalent values.
Table D-2. Summary of BMD modeling results based on NTP (1989)
Endpoint
Test
material
Model
degree
Goodness of fit
^-value
BMR,
extra risk
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
Males
Hepatocellular
adenoma/carcinoma
Adrenal
pheochromocytoma/
malignant
pheochromocytoma
Hepatocellular
adenoma/carcinoma
Adrenal
pheochromocytoma/
malignant
pheochromocytoma
tPCP
tPCP
EC-7
EC-7
One stage
One stage
One stage
Two stage
0.597
0.382
0.330
0.159
10%
10%
10%
10%
2.84
5.72
10.6
14.9
2.15
4.29
7.62
10.8
Females
Hepatocellular
adenoma/carcinoma
Hemangioma/
hemangiosarcoma
Hepatocellular
adenoma/carcinoma
Adrenal
pheochromocytoma/
malignant
pheochromocytoma
Hemangioma/
hemangiosarcoma
tPCP
tPCP
EC-7
EC-7
EC-7
One stage
One stage
Two stage
Three stage
One stage
0.336
0.998
0.952
0.79
0.986
10%
10%
10%
10%
10%
21.3
28.1
37.7
47.7
61.0
11.8
17.0
22.9
34.6
39.9
The detailed modeling results for each endpoint are presented below. The lowest BMD
(2.84 mg/kg-day) and BMDL (2.15 mg/kg-day) were for hepatocellular adenomas/carcinomas in
-------
male mice treated with tPCP. BMDLs for other data sets ranged up to 20-fold higher. Dividing
the extra risk level of 0.10 by the BMDL of 2.15 mg/kg-day yields an estimated slope factor of
0.046 (mg/kg-day)"1 for PCP based on this endpoint (U.S. EPA, 2005a).
D-4
-------
MODELING RESULTS BY ENDPOINT
Part 1. Hepatocellular adenoma/carcinoma in male B6C3F± mice treated with tPCP
adequate fit (p>0.1) with one-degree model
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
0.00
0.28
df
0
1
p-value
for model
fit
perfect
fit
0.5970
AIC for
fitted
model
177.664
175.945
BMD
(mg/kg)
3.86
2.84
BMDL
(mg/kg)
2.18
2.15
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 17:47:47 2006
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
Total number of observations = 4
Total number of records with missing values = 1
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.57
Beta(l) -0.57 1
D-5
-------
Variable
Background
Beta(1)
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood) Deviance Test DF
-85.8322
-85.9727 0.280935 1
-106.048 40.4321 2
P-value
13
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
2.83814
2.15146
D-6
-------
0.9
0.8
0.7
| 0.6
£
<
c 0.5
o
=5
2
"- 0.4
0.3
0.2
0.1
Multistage Model with 0.95 Confidence Level
Multistage
BMDL BMD
10
15 20
dose
25
30
35
17:4708/21 2006
D-7
-------
Part 2. Adrenal pheochromocytoma/malignant pheochromocytoma in male B6C3F± mice
treated with tPCP
adequate fit (p>0.1) with one-degree model
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
0.00
0.77
df
0
1
p-value
for model
fit
perfect
fit
0.3817
AIC for
fitted
model
122.564
121.347
BMD
(mg/kg)
9.22
5.72
BMDL
(mg/kg)
4.48
4.29
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 17:50:33 2006
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
Dependent variable = tPCP m rp a cc
Independent variable = tPCP_m_dose
Total number of observations = 4
Total number of records with missing values = 1
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(l) = 0.020577
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.64
Beta(l) -0.64 1
D-8
-------
Variable
Background
Beta(1)
Model Log(likelihood) Deviance Test DF
Full model -58.2818
Fitted model -58.6733 0.782979 1
Reduced model -78.4336 40.3037 2
AIC:
P-value
Goodness of Fit
Est._Prob. Expected Observed
Risk Type
Confidence level =
BMD =
BMDL =
1
10
P-value =
D-9
-------
0.6
0.5
0.4
o 0.3
=5
0.2
0.1
Multistage Model with 0.95 Confidence Level
Multistage
BMDL
BMD
10
15 20
dose
25
30
35
17:5008/21 2006
D-10
-------
Part 3. Hepatocellular adenoma/carcinoma in male B6C3F± mice treated with EC7
three- and two-degree models defaulted to the one-degree
adeguate fit (p>0.1) with one-degree model
model fit details
1 degree polynomial (pos
betas)
x2
2.22
df
2
p-value
for model
fit
0.3298
AIC for
fitted
model
238.389
BMD
(mg/kg)
10.61
BMDL
(mg/kg)
7.62
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 17:52:55 2006
The form of the probability function is:
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Default Initial Parameter Values
Background = 0.305226
Beta(l) = 0.00821465
Beta(2) = 0
Beta(3) = 0
Background
Beta (1)
D-ll
-------
Parameter Estimates
Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood) Deviance Test DF
-116.091
-117.194 2.20623 2
-131.634 31.0845 3
P-value
Goodness of Fit
Est._Prob. Expected Observed
0.1
Extra risk
0. 95
10.6138
7.62123
D-12
-------
o
••§
0.8
0.7
0.6
§• 0.5
0.4
0.3
0.2
0.1
Multistage Model with 0.95 Confidence Level
Multistage
BMDL BMD
20
17:52 08/21 2006
40
60
dose
80
100
120
D-13
-------
Part 4. Adrenal pheochromocytoma/malignant pheochromocytoma in male B6C3F± mice
treated with EC7
no adequate fit (p>0.1) with any models
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
5.56
11.55
df
1
2
p-value
for model
fit
0.0184
0.0031
AIC for
fitted
model
119.263
125.816
BMD
(mg/kg)
12.50
5.75
BMDL
(mg/kg)
7.25
4.61
High dose group dropped:
adequate fit (p>0.1) with two-degree model
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
1.98
7.96
df
1
2
p-value
for model
fit
0.1594
0.0048
AIC for
fitted
model
97.126
103.899
BMD
(mg/kg)
14.95
7.81
BMDL
(mg/kg)
10.79
5.63
High dose group dropped
Combined controls
Two-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 19:14:15 2006
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
Total number of observations = 4
Total number of records with missing values = 1
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
D-14
-------
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.000576302
Asymptotic Correlation Matrix of Parameter Estimates
Background
Background 1
Beta(2) -0.53
Parameter Estimates
Variable
Background
Beta(1)
Beta(2)
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood) Deviance Test DF
-45.4672
-46.563 2.19157 1
-67.6005 44.2666 2
P-value
Chi-square =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
14.9463
10.7929
D-15
-------
Multistage Model with 0.95 Confidence Level
0.7
0.6
0.5
0.4
£
° 0.3
2
LJ_
02
0.1
Multistage
19:1408/21 2006
BMDL
BMD
10
15 20
dose
25
30
35
40
D-16
-------
Part 5. Hepatocellular adenoma/carcinoma in female B6C3F± mice treated with tPCP
two-degree model defaulted to the one-degree
adeguate fit (p>0.1) with one-degree model
model fit details
1 degree polynomial (pos
betas)
x2
0.92
df
1
p-value
for model
fit
0.3362
AIC for
fitted
model
128.013
BMD
(mg/kg)
21.27
BMDL
(mg/kg)
11.79
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 18:00:37 2006
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
Total number of observations = 4
Total number of records with missing values = 1
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta (2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(l)
Background 1 -0.74
Beta(l) -0.74 1
Parameter Estimates
D-17
-------
Variable
Background
Beta(1)
Beta(2)
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood) Deviance Test DF
-61.5594
-62.0064 0.893896 1
-64.3577 5.59665 2
"hi-square =
4 65
9 49
9 48
P-value = 0.3362
0.1
Extra risk
0. 95
21.2708
11.7885
D-18
-------
Multistage Model with 0.95 Confidence Level
0.3
0.25
•action Affected
O
1^ O
01 k>
LJ_
0.1
0.05
: l\
I I | i i i
/I ulti stage
-i— i i i i i p— i— i i— i
~-
~-
~-
-
<
^^^^
BMDL
^^,
^~~^^
— i i— i— i i— |— i i i i i i— i— i i— |— i i— i— i i— i— i i i | i i— i— i i— i— i r— i—
y
^^^
^ ^
'-
-
-
;
-_
-
BMD ;
0 5 10 15 20 25 30 35
dose
18:0008/21 2006
D-19
-------
Part 6. Hemangioma/hemangiosarcoma in female B6C3F± mice treated with tPCP
adequate fit (p>0.1) with one-degree model
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
0.00
0.00
df
1
2
p-value
for model
fit
1.0000
0.9978
AIC for
fitted
model
62.8667
60.8711
BMD
(mg/kg)
28.11
28.06
BMDL
(mg/kg)
16.98
16.97
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 18:10:12 2006
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
Dependent variable = tPCP f rp bl cc
Independent variable = tPCP_f_dose
Total number of observations = 4
Total number of records with missing values = 1
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(l) = 0.00381681
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(1)
Beta(1)
Parameter Estimates
D-20
-------
Variable
Background
Beta(1)
Estimate
0
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
11.102
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
28.0602
16.972
D-21
-------
Multistage Model with 0.95 Confidence Level
0.25
0.2
0,5
0.1
0.05
Multistage
BMDL
BMD
10
15 20
dose
25 30 35
18:1008/21 2006
D-22
-------
Part 7. Hepatocellular adenoma/carcinoma in female B6C3F± mice treated with EC7
adequate fit (p>0.1) with two-degree model
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
0.10
7.48
df
2
2
p-value
for model
fit
0.9526
0.0238
AIC for
fitted
model
160.694
168.686
BMD
(mg/kg)
37.72
16.51
BMDL
(mg/kg)
22.86
12.48
Combined controls
Two-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 18:14:37 2006
BMDS MODEL RUN
The form of the probability function is:
Independent variable = EC7_f_dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Asymptotic Correlation Matrix of Parameter Estimates
and do not appear in the correlation matrix )
Background Beta (2)
1 -0.4
-0.4 1
D-23
-------
Parameter Estimates
Variable
Background
Beta(1)
Beta(2)
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model Log(likelihood) Deviance Test DF
Full model -78.2973
Fitted model -78.347 0.0992897 2
Reduced model -109.352 62.1099 3
AIC:
Goodness of Fit
Est._Prob. Expected Observed
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 37.7232
BMDL = 22.8618
D-24
-------
0.8
0.7
0.6
0.5
< 0.4
c
o
=5
S 0.3
0.2
0.1
Multistage Model with 0.95 Confidence Level
Multistage
18:1408/21 2006
BMDL
20
BMD
40
60
dose
80
100
120
D-25
-------
Part 8. Adrenal pheochromocytoma/malignant pheochromocytoma in female B6C3F± mice
treated with EC7
Adequate fit (p>0.1) with > two-degree models, no adequate fit with one-deqree model.
Three-staqe model, with only the third staqe coefficient fit, had the lowest AIC.
model fit details
4 deqree polynomial (pos
betas)
3 degree polynomial (pos
betas)
2 deqree polynomial (pos
betas)
1 deqree polynomial (pos
betas)
x2
0.08
0.47
3.75
21.43
df
1
2
2
2
p-value
for model
fit
0.7711
0.7903
0.1537
0.0000
AIC for
fitted
model
109.277
107.703
111.771
133.837
BMD
(mq/kq)
58.05
47.69
32.44
13.99
BMDL
(mq/kq)
35.88
34.65
26.92
10.81
Combined controls
Three-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 18:35:44 2006
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3)]
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Default Initial Parameter Values
Background = 0.0245017
Beta(l) = 0
Beta (2) = 0
Beta(3) = 9.91296e-007
D-26
-------
*** The model parameter(s) -Beta(l) -Beta (2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(3)
1 -0.28
-0.28 1
Parameter Estimates
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) Deviance Test DF
-51.5972
0.508423 2
111.931 3
"hi-square =
2 66
2 48
2 46
38 49
P-value = 0.7903
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
47.6898
34.6479
D-27
-------
Multistage Model with 0.95 Confidence Level
0.8
0.6
T3
0)
O
I
o 0.4
0.2
Multistage
18:3508/21 2006
BMDL
BMD
20
40
60
dose
80
100
120
D-28
-------
Part 9. Hemangioma/hemangiosarcoma in female B6C3F± mice treated with EC7
adequate fit (p>0.1) with models of all degrees, so choose simplest (one-degree)
model fit details
2 degree polynomial (pos
betas)
1 degree polynomial (pos
betas)
x2
0.11
0.14
df
2
3
p-value
for model
fit
0.9449
0.9862
AIC for
fitted
model
83.3146
81.3551
BMD
(mg/kg)
63.01
61.02
BMDL
(mg/kg)
40.03
39.91
Combined controls
One-degree model
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: C:\BMDS\DATA\PCP-REV.(d)
Gnuplot Plotting File: C:\BMDS\DATA\PCP-REV.plt
Mon Aug 21 18:39:55 2006
BMDS MODEL RUN
The form of the probability function is:
Independent variable = EC7_f_dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(1)
Beta(l) 1
Variable
D-29
-------
Background
Beta(1)
Model Log(likelihood) Deviance Test DF P-value
Full model -39.5989
Fitted model -39.6775 0.157225 3
Reduced model -49.135 19.0721 3
AIC: 81.3551
Goodness of Fit
Est. Prob. Expected Observed Size
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 61.0211
BMDL = 39.9095
0 68
1 49
3 50
9 49
P-value = 0.9862
D-30
-------
I
c
o
••s
s
0.3
0.25
0.2
0.15
0.1
0.05
Multistage Model with 0.95 Confidence Level
Multistage
BMDL
BMD
20
40
60
dose
80
100
120
18:3908/21 2006
D-31
-------
APPENDIX E: COMBINED ESTIMATES OF CARCINOGENIC RISK
Considering the multiple tumor types and sites observed in the mice exposed to PCP, the
estimation of risk based on only one tumor type/site may underestimate the overall carcinogenic
potential of PCP. The Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) identify
two ways to approach this issue — analyzing the incidences of tumor-bearing animals or
combining the potencies associated with significantly elevated tumors at each site. The NRC
(1994) concluded that an approach based on counts of animals with one or more tumors would
tend to underestimate overall risk when tumor types occur independently, and that an approach
based on combining the risk estimates from each separate tumor type should be used. The latter
approach was used in the cancer assessment for pentachlorophenol. This approach assumes
independence of tumors. This is a reasonable assumption since NRC (1994) stated that ". . .a
general assumption of statistical independence of tumor-type occurrences within animals is not
likely to introduce substantial error in assessing carcinogenic potency."
Cancer potencies are typically upper bound estimates; combining such upper bound
estimates across tumor sites is likely to overstate the overall risk. Therefore, following the
recommendations of the NRC (1994) and the Guidelines for Carcinogen Risk Assessment (U.S.
EPA, 2005a), a statistically valid upper bound on combined risk was derived in order to gain
some understanding of the overall risk resulting from tumors occurring at multiple sites. It is
important to note that this estimate of overall potency describes the risk of developing tumors at
any combination of the sites considered, and is not just the risk of developing tumors at all three
sites simultaneously.
For individual tumor data modeled using the multistage model,
(1) P(d) = 1 - exp[-(q0 + qjd + q2d2 + ...
and assuming independence of tumors, the model for the combined tumor risk is still multistage,
with a functional form that has the sum of stage-specific multistage coefficients as the
corresponding multistage coefficient;
(2) Pc(d) = 1 - exp[-(Zq0l + dlqh + d2Iq2l + ... + dklqkl)], for i = 1, . . . , m (m = total
number of sites).
The resulting equation for fixed extra risk (BMR) is polynomial in dose (when logarithms
of both sides are taken) and can be straightforwardly solved for a combined BMD. Confidence
bounds for that combined BMD are not estimated by the available BMD software. Consistent
E-l
-------
with the NRC (1994) recommendations, a bootstrap analysis (Efron and Tibshirani, 1993) was
used to derive the distribution of the BMD for the combined risk of multiple tumors observed in
male and female B6C3Fi mice exposed to tPCP or EC-7 (i.e., liver and adrenal gland tumors in
male mice exposed to tPCP or EC-7; liver and circulatory tumors in female mice exposed to
tPCP; and liver, adrenal gland, and circulatory tumors in female mice exposed to EC-7).
Within each of the individual tumor data sets (see Table E-l), a simulated incidence level
was generated for each exposure group using a binomial distribution with probability of success
estimated by a Bayesian (assuming a flat prior) estimate of probability given by (observed
incidence+l)/(total number+2). This adjustment is necessary in order to avoid underestimation
of variability when the observed incidence is 0 in any group and there is no prior estimate of
incidence (e.g., from historical controls), and must be applied to all groups to preserve the
differences between them. Each simulated data set was then modeled using the multistage model
in the same manner as those reported in Appendix D. The multistage parameter estimates from
the individual tumors were substituted in equation (2) above, which was then solved for the
BMD at an overall BMR of 1% extra risk. This process was repeated until there were 10,000
simulated experiments for each individual tumor. Whenever the multistage model could not
provide an adequate fit for any of the simulated data sets, the simulated experiments were
excluded from the analysis. The 5* percentile from the distribution of combined BMDs was
used to estimate the lower 95% bound on the dose (BMDL) corresponding to an extra risk of 1%
for any of the two tumor sites in male mice exposed to tPCP or EC-7 or three tumor sites in
female mice exposed to tPCP or EC-7.
E-2
-------
Table E-l. Results of simulation analyses characterizing combined cancer risk estimates for male and female mice
Endpoint
In terms of administered bio assay
exposures
BMD10
mg/kg-d
BMDL10
mg/kg-d
Human equivalents"
BMD10/HED
mg/kg-d
0.1/BMD10/HED
(mg/kg-d)1
BMDL10HED
mg/kg-d
0.1/BMDL10/HED
(mg/kg-d)1
Male mice, tPCP
Hepatocellular adenoma/carcinoma
Adrenal benign/malignant pheochromocytoma
Combined tumors
3.12
6.45
2.23
2.27
4.47
1.63
0.475
0.981
0.340
0.211
0.102
0.294
0.35
0.68
0.25
0.290
0.147
0.402
Male mice, EC-7
Hepatocellular adenoma/carcinoma
Adrenal benign/malignant pheochromocytoma
Combined tumors
11.0
12.6
6.2
7.59
5.75
3.7
1.68
1.92
0.944
0.060
0.052
0.106
1.15
0.88
0.57
0.087
0.114
0.174
Female mice, tPCP
Hepatocellular adenoma/carcinoma
Hemangioma/hemangiosarcoma
Combined tumors
21.3
27.8
12.6
11.7
16.3
7.88
3.24
4.23
1.91
0.031
0.024
0.052
1.79
2.48
1.20
0.056
0.040
0.083
Female mice, EC-7
Hepatocellular adenoma/carcinoma
Adrenal benign/malignant pheochromocytoma
Hemangioma /hemangiosarcoma
Combined tumors
36.9
45.5
60.7
23.2
16.4
29.6
37.9
13.6
5.61
6.93
9.24
3.52
0.018
0.014
0.011
0.028
2.50
4.51
5.76
2.07
0.040
0.022
0.017
0.048
aHED (mg/kg-d) = dose in animals (mg/kg-d) x (BWa/BWh); at 0.037 kg for male mice and 0.038 kg for female mice and 70 kg for humans, the cross-species scaling
factor was 0.15.
Source: NTP(1989).
E-3
-------
The results of combining risks across sites within data sets are shown in Table 5-6. The highest
combined risk observed, similar to the analysis of cancer risk estimates for individual tumors, was in
tPCP-exposed male mice. The 95% UCL on the combined risk for animals that developed liver and/or
adrenal gland tumors was 4.0 x 10"1 (mg/kg-day)"1, which is about 30% higher than the 3.1* 10"1
(mg/kg-day)"1 oral slope factor estimated from liver tumors only in tPCP-exposed male mice. The risk
estimates for the tPCP-exposed males and females are higher than those for the EC-7-exposed animals
by approximately twofold for both the central tendency and upper bound estimates.
E-4
------- |