vvEPA
          United States
          Environmental Protection
          Agency
           Office of Research and
           Development
           Washington DC 20460
EPA/600/6-91/002F
September 1991
Response to Issues and
Data Submissions on the
Carcinogenicity of
Tetrachloroethylene
(Perchloroethylene)

-------


-------
                                          EPA/600/6-91/002F
                                          September 1991
RESPONSE TO ISSUES AND DATA SUBMISSIONS
                                      j
       ON THE CARCINOGENICITY OF

         TETRACHLOROETHYLENE

          (PERCHLOROETHYLENE)
        Human Health Assessment Group    •
   Office of Health and Environmental Assessment
      U.S. Environmental Protection Agency
           Washington, D.C. 20460       ;
                                       Printed on Recycled Paper

-------
                                  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.
                                          11

-------
                                  CONTENTS


 Preface  	                     vj

 Authors  	;	  yii

 1.  INTRODUCTION  	                    l

 2.  BACKGROUND  	  3

    2.1. Prior EPA Analyses	  3

    2.2. Animal Studies of Perchloroethylene Carcinogenicity .     	  5

    2.3. Purpose of This Paper	  6

 3.  THE METABOLISM OF PERCHLOROETHYLENE	  8

 4.  MUTAGENICITY OF PERCHLOROETHYLENE AND ITS METABOLITES	14

    4.1. Data on Mutagenicity of Perchloroethylene per se	14

    4.2. Mutagenicity of Perchloroethylene Metabolites	17

 5.  MOUSE LIVER TUMORS	23
                                                         i
    5.1. Carcinogenicity Bioassay Data and EPA's Position	23

    5.2. Peroxisome Proliferation and Perchloroethylene	27

 6.  KIDNEY TUMORS IN MALE RATS	 .	33

    6.1. Alpha-2u-Globulin in Renal Carcinogenesis in Male Rats  	35

   6.2. Sustained Chronic Nephrotoxicity as a Possible Mechanism Independent of
        Alpha-2u-Globulin Accumulation	41

   6.3. A Mutagenic Mechanism of Perchloroethylene-Induced Carcinogenesis in Male
        Rats	                   42

7. MONONUCLEAR CELL LEUKEMIA IN RATS	i	46
                                      m

-------
                        CONTENTS (continued)






8.  SUMMARY AND CONCLUSIONS  	50




References 	61
                                 IV

-------
                                CONTENTS (continued)
                                  LIST OF TABLES


 1.  Summary of Genotoxicity Testing of Tetrachloroethylene .....  .........       15

 2.  Summary of Genotoxicity Testing of Tetrachloroethylene Metabolites  .........   i8



                                 LIST OF FIGURES


 1.  Oxidative metabolism of perchloroethylene

2.  Enzyme-catalyzed metabolism of perchloroethylene to its glutathione conjugate

    ^^S-co^ °f ** .glUtamyl and ^^ reSidU6S t0 yidd itS ^-Ponding

3.  Further metabolism in the kidney of the perchloroethylene intermediate 1 1 2-
    trichlorovmylcysteine, leading to mutagenic metabolites ............ '.'...       12

-------
                                      PREFACE

       This document, prepared by the Office of Health.and Environmental Assessment
(OHEA), responds to data and comments submitted to the U.S. Environmental Protection
Agency (EPA) and discusses how this information influences the overall weight-of-evidence
classification for a perchloroethylene human cancer hazard.  Relevant literature through early
1991 has been critically evaluated.
       The Agency's Science Advisory Board reviewed the February 1991 draft of this
report at a public meeting and found the report to be well written and of high scientific
quality, although the Board did not agree with the Agency's recommended weight-of-
evidence classification for perchloroethylene of B2, probable human carcinogen.  The
Science Advisory Board, expressing  its views in an August 1991 letter to EPA Administrator
William Reilly, offered advice on the weight-of-evidence classification in a spirit of
flexibility  encouraged by the 1986 Risk Assessment Guidelines for Carcinogenicity and
recommended placing perchloroethylene on a continuum between Group B2 and Group C
weight-of-evidence categories. The  Board stated that because the major issues arising from
the assessment of perchloroethylene  had not changed since  1987,  its previous response is
 appropriate and cautioned, as in its 1988 response, that "from a scientific point of view, it
 seems inappropriate for EPA and other agencies to regulate substances that are classified B2
 and not to consider regulations of compounds classified as  C, regardless  of the level of
 human exposure	A substance classified as  C for which exposure is  high may represent a
 much greater threat to human health."
                                             VI

-------
                                   AUTHORS


      Members of the Human Health Assessment Group of the Office of Health and

Environmental Assessment (OHEA) prepared this document.
                                                          i
                                                          i

PRIMARY AUTHOR

Jean C. Parker, Ph.D.                                       i
Carcinogen Assessment Toxicology Branch
U.S. Environmental Protection Agency                         j
Washington, DC


CONTRIBUTING AUTHORS

Vicki Vaughan-Dellarco, Ph.D.
Genetic Toxicology Assessment Branch
U.S. Environmental Protection Agency                         i
Washington, DC

David Reese, Ph.D.
Genetic Toxicology Assessment Branch
U.S. Environmental Protection Agency
Washington, DC
                                       Vll

-------

-------
                                 1. INTRODUCTION

       The scientific debate over the potential carcinogenicity of tetrachloroethylene
(perchloroethylene, perc, PCE) spans several years. The Office of Health and
Environmental Assessment (OHEA) within the U.S. Environmental Protection Agency's
(EPA's) Office of Research and Development has been considering the issues and current
thinking pertaining to weight of evidence for the human cancer hazard from exposure to
perchloroethylene.  Several issues were raised by the EPA's Science Advisory Board (SAB,
personal communication) during its review of an addendum (U.S. EPA, 1986a) to the Health
Assessment Document for Tetrachloroethylene (U.S. EPA, 1985a). New information that
has bearing on the issues also became available over the last two to three years.
       Recently generated laboratory data have led to the development of hypotheses about
the mechanisms of perchloroethylene tumorigenesis. Biological arguments suggested species
specificity for some of the proposed tumorigenic mechanisms.  Such arguments  imply that
certain experimental results are of questionable predictive validity with respect to human
health hazards and risks. While some evidence supports  these arguments, several critical
experimental  elements are missing to determine cause-and-effect relationships in the test
animals or to answer the human relevancy question with certainty. Because the data  are
equivocal regarding mechanisms, a conclusion about the carcinogenic potential of
perchloroethylene in humans must be one of judgment, considering the weight of the
pertinent evidence.
       The objectives of this paper are to review the current issues and hypotheses
surrounding perchloroethylene  carcinogenesis, to evaluate these hypotheses hi light of
recently published studies, and to develop the EPA's response to issues and data included hi
comments submitted to the Agency on the overall weight of evidence for the potential human
cancer hazard.  This paper reviews the issues considered by the EPA's Science Advisory
Board during its review  of the  draft Addendum to the Health Assessment Document for
Tetrachloroethylene (U.S. EPA, 1986a) and discusses relevant research data published since
1986.  The topics include three tumor end points observed hi rodents exposed to
perchloroethylene:                                             '

-------
       1.  Hepatocellular carcinoma in male and female mice,
       2.  Renal tubule neoplasia in male rats, and
       3.  Mononuclear cell leukemia in male and female rats,
and the recent data on metabolism, mutagenicity, peroxisome proliferation, and alpha-2u-
globulin.
       The EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986b) provide a
framework for assessing the likelihood of a substance being a human cancer hazard.  Under
these guidelines animal studies and human data are first analyzed separately.  The evidence
from laboratory animal studies, along with other relevant information, may be classified as
"sufficient," "limited," "inadequate,"  "no data," or "no evidence" hi animals. The human
data are classified as "sufficient," "limited," "inadequate," "no data," or "no evidence" in
humans. The two sets of information are merged with respect to assessing potential
carcinogenicity hi humans (see Guidelines  for Carcinogen Risk Assessment, U.S. EPA,
1986b). The classifications refer only to the weight of the experimental evidence that a
chemical is carcinogenic and not to its potency of carcinogenic action.  The overall challenge
is not only to determine how the currently available data influence the Agency's previous
categorization of the experimental animal evidence on the carcinogenicity of
perchloroethylene  as "sufficient," but also to determine whether the "sufficient" animal data
hi the case of perchloroethylene  signify a potential human hazard, as would be ordinarily
assumed (U.S. EPA,  1986b; OSTP, 1985; IARC, 1982, 1987).

-------
                                   2.  BACKGROUND

  2.1. PRIOR EPA ANALYSES
         The EPA published the Health Assessment Document (HAD) for Tetrachloroethylene
  (Perchloroethylene) in July 1985 (U.S. EPA,  1985a).  The Office of Health and
  Environmental Assessment, in consultation with an Agency workgroup, prepared the HAD as
  a source document for the entire EPA (U.S. EPA, 1985a, preface).  The document
  underwent extensive expert peer review and review by the Environmental Health Committee
  of the Agency's Science Advisory Board (SAB) before publication.  Based on the EPA's
  interpretation of the overall weight of evidence, the HAD placed perchloroethylene into
  Group C (possible human carcinogen).  This categorization was in accordance with the
  Agency's proposed Guidelines for Carcinogen Risk Assessment (published in final form in
  September  1986).  The classification was based primarily on the finding that "in a gavage
  bioassay, perchloroethylene induced a statistically significant increase of malignant liver
  tumors in both male and female B6C3F1 mice."  This decision reflected a "limited" number
  of studies showing a robust positive response that was a commonly observed tumor type as
 opposed to  a very rare tumor type, rather than "limited" evidence, such as a borderline
 response, derived from a number of adequately run studies.  In view of the pending release
 of National Toxicology Program (NTP) reports on long-term animal inhalation studies with
 perchloroethylene, the HAD stated that the carcinogenicity conclusions were interim and
 would be updated if necessary  when the NTP reports were evaluated (U.S. EPA, 1985a,
 preface).                                                             !
       A draft Addendum to the Health Assessment Document for Tetrachloroethylene
 (Perchloroethylene) (U.S. EPA, 1986a), prepared by OHEA,  analyzed the results of the
 inhalation bioassays performed by Battelle Pacific Northwest Laboratories for the NTP.  The
 results of these studies revealed perchloroethylene-associated increases  in the incidences of
 hepatocellular carcinomas-the same tumor type seen in the gavage study-in both sexes  of
 B6C3F1 mice, mononuclear cell leukemia in both sexes of F344/N rats, and uncommon renal
tubule neoplasms and some evidence for gliomas of the brain in male rats.   The authors of

-------
the addendum concluded that perchloroethylene is a B2 chemical (probable human
carcinogen) because:

       •  The NTP inhalation bioassay demonstrated that perchloroethylene can induce
          carcinogenic effects at multiple sites in both rats and mice (a replicate finding)
          through inhalation exposure (a second route), and
       •  The earlier National Cancer Institute (NCI) bioassay provided positive evidence of
          hepatocellular carcinomas in mice administered perchloroethylene by gavage.

       On May 15,  1986, the draft addendum underwent peer review by the SAB's
 Halogenated Organics Subcommittee in a public meeting held in Madison, Wisconsin. The
 SAB's initial comments appeared in a letter to EPA Administrator Lee Thomas dated January
 27, 1987 (personal communication). In this letter the SAB concluded "that perchloroethylene
 belongs in the overall weight-of-evidence category C (possible human carcinogen)."
        Further, the  SAB judged the evidence for carcinogenicity in animals to be "limited"
 because "the National Toxicology Program bioassay does not provide a scientific basis to
 associate either lesion (in rats) with inhalational exposure to perchloroethylene";  thus, "the
 evidence arises only from a single strain of mouse"  and "the kind of tumor associated with
 perchloroethylene exposure in this mouse strain makes it difficult to create an inference
 regarding human carcinogenicity."
        Because of the SAB's conclusions, Agency scientists and managers reexamined the
 assessment of perchloroethylene and again concluded that perchloroethylene should be
 classified as a B2 chemical.  Administrator Thomas responded to the SAB in an August 3,
 1987, letter that clarified the Agency's position on perchloroethylene and requested additional
 consultative advice from the SAB on specific issues relating to liver tumors in B6C3F1  mice
 and kidney  tumors  in male rats.  More detailed comments were presented in "EPA Staff
 Comments on Issues Regarding the Carcinogenicity of Perchloroethylene (Perc) Raised by
 the SAB," a paper that was enclosed with the Administrator's August 3 letter.
         The SAB responded to the second request for  advice in a letter dated March 9, 1988,
  to Administrator Thomas.  In this letter the SAB concluded that "the overall weight of

-------
  evidence lies on the continuum between the categories B2 and C of EPA's risk assessment
  guidelines  for cancer."  In an attempt to put this conclusion in perspective,  the SAB also
  remarked that:
        A substance classified as C (limited evidence in animals) for which human
        exposure is high may represent a much greater potential threat to human
        health.  EPA and other agencies (including those in state governments) may
        therefore, wish to take steps to reduce high exposures to substances in the C
        category whenever there appears to be a potentially significant threat to  human
        health (in the sense that the plausible upper bound estimate of potency times
        lifetime exposure is above the threshold where regulation may be judged
        appropriate).
        Since then, the EPA has received public comments on perchloroetitrylene in several
 regulatory actions that include consideration of the perchloroethylene weight-of-evidence
 classification.  These public comments pertain to Resource Conservation and Recovery Act
 listing (U.S. EPA, 1989a) and Comprehensive Environmental Response Compensation and
 Liability Act reportable quantity rules (U.S. EPA, 1989b), and to maximum contaminant
 level goal and maximum contaminant level proposals for drinking water (U.S.  EPA, 1989c).
       After its review of a 1991 draft of this report on perchloroethylene, the Agency's
 Science Advisory Board concluded that its previous advice remained appropriate and that the
 weight of evidence indicates that perchloroethylene lies on a continuum between categories
 B2 and C.  The Board also maintained its stance that, from a scientific perspective, exposure
 should be considered more important than classification category in determining potential
 threat to human health and whether or not a chemical substance should be regulated.

2.2.  ANIMAL STUDIES OF PERCHLOROETHYLENE CARCINOGrENICITY
       Perchloroethylene has shown cancer-causing activity in male and female mice and in
male and female rats in the NCI and NTP studies.  In both sexes of mice, perchloroethylene
administered  by oral gavage or by inhalation induced dose-related statistically significant
increases in hepatocellular carcinomas.  Statistically significant increased incidences of

-------
mononuclear cell leukemia and the presence of uncommon renal tubule neoplasms and some
evidence of gliomas in the brain were observed in male rats exposed to perchloroethylene by
inhalation.  Renal tubule tumors were also detected in male rats exposed by gavage.  Female
rats exhibited an increase of mononuclear cell leukemia when exposed to perchloroethylene
by inhalation.
       Although perchloroethylene increased the incidence of cancer at three different sites
and in two species, controversy surrounds each of the tumor end points.  Considerable
scientific debate has focused on the predictive validity of mouse liver and male rat kidney
tumors as well as rat mononuclear cell leukemia. In addition to the general controversies
surrounding these tumor end points, chemical-specific data that may be pertinent to
evaluating the effect of perchloroethylene on tumor incidence have generated concern.

2.3.  PURPOSE OF THIS PAPER
        The intent of this paper is to respond to data and comments submitted to the Agency.
 Studies related to perchloroethylene tumorigenesis have been published subsequent to EPA's
 1986 draft addendum and have been formally  submitted to the EPA. These studies were
 designed to elucidate the mechanism of action in animals and provide better understanding of
 the relevance of animal data to human hazard.  The studies report mechanistic data germane
 to a number of issues concerning the etiology of perchloroethylene-induced rodent tumors
 and their relevance to humans.  The EPA must  evaluate information from these studies and
 assess whether the data reflect an understanding of underlying tumorigenesis mechanisms that
 would lead to a conclusion that modes of perchloroethylene cancer-causing activity are not
 operative in humans. The purpose of this paper, then, is to provide a response to the data
 and comments submitted to the Agency that have bearing on the relevancy issues and to
 discuss how this information influences the overall weight-of-evidence classification for a
 perchloroethylene human cancer hazard.
         This paper summarizes OHEA's critical evaluation of three perchloroethylene-related
  tumor end points in rats and mice in light of new scientific findings and incorporates  the
  recent information into the weight of evidence  for human hazard.

-------
       The paper addresses recent literature on perchloroethylene and its biometabolites as
they relate to metabolism, mutagenicity, cytotoxicity, and proliferative changes in mouse
liver, nephrotoxicity, and renal tubule neoplasia in male rats and mononuclear cell leukemia
in male and female rats.                                              :

-------
                3.  THE METABOLISM OF PERCHLOROETHYLENE
       The cancer-causing activity of halogenated ethylenes is generally considered to reside
primarily in biometabolites rather than in the parent compounds themselves.  Studies in
animals and humans indicate that metabolism of perchloroethylene is relatively limited, as
evidenced by the high percentage of absorbed dose excreted hi the breath as the parent
molecule (U.S. EPA, 1985a, 1986a; Ohtsuki et al., 1983; Ikeda et al., 1980; Henschler,
1977a; Yllner, 1961). In human studies, however, only about half of the perchloroethylene
absorbed has been accounted for through excretion of parent compound or metabolites (U.S.
EPA, 1985a, 1986a).
       Estimates of the extent of metabolism in humans have been made from balance studies
by accounting for a retained dose after inhalation exposure by measuring trichloro-
compounds excreted hi the urine. Metabolites other than those measured may be excreted in
the urine or bile. Thus, the additional perchloroethylene in humans  may be metabolized to
compounds that were not measured.  Other as yet unrecognized pathways for
perchloroethylene that have not been considered may exist hi humans.
       Perchloroethylene is metabolized through at least two distinct pathways.  Oxidative
metabolism via the  cytochrome P450 system has been extensively reviewed in the HAD (U.S.
EPA, 1985a). Recent investigations have revealed a glutathione conjugative pathway that
appears to be a minor but important route that has been shown to generate a mutagenic
constituent. The oxidative and conjugative  pathways are summarized in figures 1, 2, and 3.
       Oxidative metabolism of perchloroethylene (dependent on cytochrome P45o) probably
occurs mostly hi the liver but may occur at other sites.  This pathway is operative in humans
as well as in rodents and leads to the production of several metabolites (figure 1).  There is
no basis  to believe that qualitative differences exist between species with respect to known
pathways of oxidative metabolism of perchloroethylene.  There are,  however, quantitative
differences among the metabolic rates of different species. The major metabolite of the
oxidative pathway is trichloroacetic acid (TCA), which is excreted hi the urine of all species
tested. Other identified urinary metabolites are designated in figure  1. Some of the
intermediates hi the oxidative pathway are known to possess cytotoxic or genotoxic activity
                                           8

-------
                                           .a
                                           =3 u
1
u

5 •*•
en
 CM

S

o
 I
S-f

5"!

v!


X
0
1
CM
?•*
LE
5N«
§
15
1
CM
z
0
1 *.
« s
|1

°*o'S|
                                           O
                                           -<
                                             2
                                          II

                                          £2
                                           x «
                                          O T3
                                            a



                                           el
                                           ? S

                                          S 2

                                       B
                                       p

-------
(e.g., perchloroethylene-epoxide, chloroacetaldehydes; see section 4), and several have been
shown to cause cancer (e.g., DCA, TCA, and chloroacetaldehydes).
       Recent studies in rats (reviewed by Anders et al., 1988 and Dekant et al., 1989) have
demonstrated the formation of cytotoxic and mutagenic metabolites of perchloroethylene that
arise from hepatic glutathione conjugative pathways (see figures 2 and 3). This secondary
metabolic pathway is initially catalyzed by hepatic cytosolic and microsomal glutathione S-
transferases to yield S-(l,2,2-trichlorovinyl) glutathione (TCVG).  After transport to the
kidney, TCVG is metabolized to S-(l,2,2-trichlorovinyl) cysteine (TCVC) by the enzymatic
removal of glutamyl and glycine residues.  TCVC is acetylated via a reversible reaction to
N-acetyl-S-(l,2,2-trichlorovinyl)-L-cysteine, which is excreted in the urine.  However,
TCVC is also a substrate for renal beta-lyases, which cleave TCVC to yield an unstable thiol
that may give rise to cytotoxic and mutagenic intermediates (Vamvakas et al., 1987, 1989a,
 1989b, 1989c).
        In  vivo and in vitro experiments in rodents provide evidence to support this
 conjugative metabolic scheme.  Dekant et al.  (1987) and Green et al. (1990) have
 demonstrated hepatic conjugation of perchloroethylene with glutathione by rat liver fractions
 in in vitro experiments.  Vamvakas et al. (1989b) found TCVG in the bile excreted by
 perchloroethylene-perfused rat livers, and Green et al.  (1990) reported the presence of the
 glutathione conjugate in the bile of rats administered perchloroethylene by gavage.
        Some limited evidence suggests that humans may not metabolize perchloroethylene by
 the conjugative pathway.  Human liver samples have been compared with rat and mouse  liver
 samples with respect to ability to conjugate perchloroethylene with glutathione (Green et  al.,
 1990).  These investigators found low levels  of conjugation by rat and mouse livers but were
 unable to demonstrate conjugation by human livers.  To confirm the viability of glutathione
 S-transferase in the rat and human liver samples studied by using a more powerful protocol,
 these workers compared the two species with respect to their ability to conjugate 1-chloro-
 2,4-dinitrobenzene with glutathione.  Both species carried out this conjugation rapidly and at
 essentially the same rate.  This finding indicates that the reduced ability of the human liver
  samples to conjugate perchloroethylene was not attributable to an inactive  glutathione
  S-transferase.  Because of the very low levels of enzyme activity being measured and the
                                             10

-------
Cl
                    Cl
  X - _ « X
  .A* -""^v
ci'      sci
Perchloroethylene
                         GSH (glutatMone)
                                   C'
                 GSH-S-transferas*
                                                        NH-CH2-COOH
                                      .C=<

                                   or
                                     i-CH9-CH
                                             AH
                                           1,1,2-TrichIorovlnylgIutathlone
O              COOH
II               I
C-CH2-CH2-CH

               NH,
                              a-glutamyltranspeptldase
                   civ

                   cix
                                      .Cl
                                                     gtutamate
                                                  NH-CH9-COOH
                                               "C
                                     NS-CH2-CH

                                                NH,
                             cystelnylglydnedlpeptldase
                                                     gtydne
                                             Ov   OH

                             V
-------
            Cl
            Cl'
       Cl
                C   N-acetyltransterase^

        !-CH9-CH  «           -
                                 NH,
            1 ,1 ^-Trichloro vlnylcystelne*
 Cl
                            Ci
                                   rearrangement
 Cl         SH


Trlchlorovlnylthlol
Cl


Cl
                                                               Cl
xs.          I
         S-CH2-CH   O
               2   I     II
                   NH-C-CH3

1,1,2-Trichlorovinyl-N-acetylcysteIne*
Cl         Cl

  V=<
Cl'       S
                                                 Dichlorothioacetyl Chloride
                          + Pyruvate

                          + Ammonia
                 Cl

                   vc=c=s

                 Cl'

                Dichlorothloketene
                                                H2o
                                                 OH
                                                      Cl
                                      Cl
                                     DIchloroacetIc Acid*
            •Identified urinary metabolites.



Figure 3.  Further metabolism in the kidney of the perchloroethylene intermediate
1,1,2-trichIorovinylcysteine, leading to mutagenic metabolites.  These pathways occur
in humans as well as in rodents.


Source:  Dekant et al., 1987; Green et ah, 1990; ECETOC, 1990.
                                           12

-------
                 of human liver sampies tested, it „ premature ^      d
          c step » untteiy to occur in humans.  Additional confirmatory studies ^clearly
   needed.  Human liver docs conjugate glutathione  with hexach,oro-l 3-LdienlT
   cMoroalkene rdated to percUoroethylene (Oesch and Wo,f  19m  'l^™' *



            Lvorry     ystetae metaboute- ta reiaaveiy ^ «™ «-» *- «-.
          s  have been measured in the urine of occupation^ exposed subjects  These data
         a h,gn ^iihood of perciUoroethyiene being n^ed via this athway ^N
                        .cys.in,  rhe precursors of these  urtaa, metaLKs
                     icagen^hcnbioacUvatedbybeta-.yase.

                       activdy
       He bewyase paftway has been shown to produce cytotoxic and mutagcnic
    .-. to .e fo^on of .e toxic metaboltes in anim           d «

~,es in bacteria, mode, is aiso present in human proxima, tubuiar cei^h et a,

1990; Green et al. ,  1990). Human proximal tubular cells have been shown r

* ~» of g-utathione andVor cysteine conjugates of a ^££.

fluoroalkenes that are activated via the beta-lyase pathway (Chen et a,., 1990).
                                      13

-------
  4. MUTAGENICITY OF PERCHLOROETHYLENE AND ITS METABOLITES

     Genetic alterations are critica, events in the carcinogenesis process. Thus, evidence
          of an agent to produce heritable genetic lesions (e.g., gene mutauons, stable

-r: ir. :===
observations of mutagenic noncarcinogens and nonmutagenic carcmogens as well as
           985 EPA Heal* Assessment Document for TetracMoroethylene provded a
 to determine if the earlier conclusions are still valid.

 4 !  DATA ON MUTAGENICTTY OF PERCHLOROETHYIJSNE PER SE
      Til in table 1, perchloroethvlene has not been clearly shown to be an mducer
  o, gene mutations in routinery used assays.  In bacteria, assays for reverse mutation
   1— /mammalian microsome test) in the presence or — «^T
  activation, perchloroemylene exposures produced largely negative results (Bartsch e, a.
  r979  Majard, 1978; Haworth e, a,., 1983; Warner et a,., 1988). In srud.es reportmg



   !Tpi tested. Jen highly purified perch.oroethy.ene was evaluated  in a des,ccator with
                                        14

-------
  Table 1. Summary of Genotoxicity Testing of Tetrachloroethylene
A. Gene Mutation Tests
Salmonella/Ames assay
Escherichia coll K12/343/113
Multipurpose test
Yeast reverse mutation test
Drosophila sex-linked recessive lethal test
L5178YTK+/TK- mouse lymphoma cell assay
B. Chromosomal Aberration Tests
Chinese hamster ovary cells (CHO)
Rat bone marrow assay
Peripheral lymphocytes from exposed humans
C. Other Tests Indicative of DNA-Damaging
Activity
Unscheduled DNA synthesis in WI-38
Hepatocyte primary cuIture/DNA repair test
Mitotic recombination tests in yeast
Sister chromatid exchange formation in CHO cells
DNA strand breaks (alkaline elution test assay) in
mouse kidney and liver cells
D. DNA Binding Studies
Mice
Mice and rats
Results3
mostly —
+6
-
-
-
-, ?c

-
_c
-

_c
+b,-
+,-
-
wk

-
9
References
Shimada et al., 1985; Margard, 1978;
Bartsch et al., 1979; Haworth et al., 1983;
Warner et al., 1988
Henschler, 1977b; Greim et al., 1975
Callen et al., 1980; Bronzetti et al.; 1983
Beliles et al., 1980; Valencia et al., 1985
Myhr et al., 1986; Galloway et al., 1987;
McGregor et al., 1988
I
Galloway et al., 1987
Rampy et al., 1978); Beliles et al., 1980
Dceda et al., 1980

Beliles et al., 1980
Shimada et al., 19S5; Costa and Ivanetich,
1984; Goldsworthy et al., 1988
Callen et al., 1980; Bronzetti et al., 1983
Koch et al., 1988; Galloway et al., 1987
Walles, 1986

Schumann et al., 1980
Mazzullo et al., 1987
a+  designates positive;  - negative; wk weak response; ? inconclusive test.  Dose-response relationships were not
 established for the reported + results or wk results.
Positive results are considered weak because large amounts of material were needed to elicit the responses.  Results may
 also be explained by  mutagenic stabilizers or contaminants.                                   :
Questionable evidence for weak or borderline activity in specific data sets.
                                                         15

-------
the Salmonella/mammalian microsome test, negative results were obtained (Shimada et al.,
1985).
       Perchloroethylene has also tested negative in yeast for reverse mutations (Callen et
al., 1980; Bronzetti et al.,  1983) and Drosophila for sex-linked recessive lethal mutations
(Valencia et al., 1985). Responses after perchloroethylene treatment in the L5178Y
Tk+/Tk~ mouse lymphoma cell assay for forward mutations have been either negative or
equivocal (Myhr et al., 1986; McGregor et al., 1988).
       Perchloroethylene has not been demonstrated to be a clastogen (chromosome-breaking
activity).  Negative results were found for the induction of chromosomal aberrations in
cultured Chinese hamster ovary cells (Galloway et al., 1987) and in the bone marrow assay
for rats and mice (Cerna and Kypenova, 1977; Rampy et al., 1978; Beliles et al., 1980). A
cytogenetic study of humans exposed to perchloroethylene did not show elevated frequencies
of chromosomal aberrations or sister chromatid exchanges in peripheral  lymphocytes (Ikeda
etal., 1980).
       Chemical adduct formation is a prerequisite step in certain types  of mutagenesis.
Schumann et al. (1980) reported no detectable DNA binding in livers of mice exposed to
inhaled 14C-labeled perchloroethylene.  It is important to  note that this was not a sensitive
test, however, because the specific activity of the label was too low to preclude the
possibility of DNA binding (i.e., this test could not detect slightly fewer than 10'5 alkylations
per nucleotide; recent protocols are able to detect 10'9 to 1042). In a more  recent study,
Mazzullo et al. (1987) reported low levels of DNA binding in the liver,  kidney, lung, and
 stomach of the mouse and rat after intraperitoneal injection of perchloroethylene.  These low
 levels of DNA binding cannot be distinguished from binding as a result of biosynthetic
 incorporation of the label  into DNA, and thus, it is questionable whether exposure to
 perchloroethylene results in the formation of DNA adducts.
        Perchloroethylene  exposures have produced negative, questionable, or weak results in
 tests that do not measure mutation per  se but are indicative of DNA-damaging activity. Tests
 for DNA repair synthesis in hepatocytes  (Shimada et al., 1985; Costa and Ivanetich, 1984;
 Goldsworthy et al., 1988), mitotic recombination in yeast (Bronzetti et  al., 1983;  Koch et
 al., 1988), and sister chromatid exchange formation in culture Chinese  hamster cells
                                             16

-------
   (Galloway et al., 1987) have been predominantly negative.  Perchloroethylene has been
   reported to be a weak inducer of DNA single-strand breaks in mouse liver and kidney
   (Walles, 1986).  Although DNA strand breaks may lead ,o mutagenieity, agents tot can be
   demonstrated to induee only DNA strand breakage shou!d not be viewed as possessing the
   same genetie hazard potential as agents mat have been shown to induce gene mutations or
   stable chromosomal aberrations.

  4.2.  MUTAGENICITY OF PERCHLOROETHYLENE METABOLITES
         When the 1985 HAD was being prepared, literature on the mutagenicity of
  Perchloroethylene metabolites was limited. However, several studies are now available
  (summarized in table 2), and their results warrant some consideration.   \
        Oxidative metabolism of perchloroethylene (dependent on cytochrome P450) occurs
  mostly in the liver.  This pathway is operative in both rodents and humans and leads to the
  production of several metabolites (U.S. EPA, 1985a;  Yllner, 1961) (figure 1).
  Perchloroethylene-epoxide, a hypothesized intermediate in perchloroethylene  oxidative
  metabolism (Henschler et al., 1977a,b), has been shown to be mutagenic in the
  Salmonella/mammalian microsome test (Kline et al., 1982)
        Chloral hydrate  (trichloroacetaldehyde), a known metabolite of trichloroethylene and
 likely a perchloroethylene metabolite, has been produced both in vitro and in vivo   Several
 studies are available on the ability of chloral hydrate to produce aneuploidy (i.e., loss or gain
 of whole chromosomes) in both mitotic and meiotic.cells, including yeast (Singh and Sinha
 1976, 1979; Kafer, 1985; Gualandi,  1987; Sora and Agostini-Carbone, 1987), cultured
 mammalian somatic cells (Degrassi and Tanzarella, 1988), and spermatocytes  of mice (Russo
 et al., 1984; Liang and  Pacchierotti,  1988).  It should be pointed out that this type of genetic
 effect is most likely due to interference of spindle function rather than a DNA-reactive
 mechanism.  Chloral hydrate has also been shown to block spindle elongation  in insect
 spermatocytes (Ris, 1949).  Chloral hydrate has been reported to be weakly mutagenic in the
 Salmonella/mammon microsome test (Haworth et al., 1983) but negative for sex-linked
recessive lethal mutations in Drosophila (Yoon et al., 1985).  It has  been  reported to induce
single-strand breaks in hepatic DNA of mice and rats (Nelson and Bull, 1988)  and mitotic
                                          17

-------
Table 2. Summary of Genotoxicity Testing of Tetrachloroethylene Metabolites
  Metabolite
  _...         —
  Perchloroethylene-epoxide
  —i             •—	     ""
  Chloral hydrate
    Trichloroacetic acid
     Dichloroacetaldehyde
     Monochloroacetaldehyde
     S-<1,2,2-trichloro-vinyl)
     glutathione
            (Result)3/Assay

(+) SalmonellatAiass assay

(+) Aneuploidy/yeast

	•	
 (+) Aneuploidy/mammalian cells in vitro

 (+) Aneuploidy/spermatocytes of mice

	
 (wk) Salmonella/Ames test

 (-) Drosophila sex-linked recessive
 lethal mutation
              ~
     DNA strand breaks in mice and rats

 (+) mitotic gene conversion in yeast
            ^~*^^^—^™^^^^™^~"

         .	•	
      DNA strand break in mice and rats

  (_) DNA strand break in mice and rats

  (?) in vivo cytogenetics

  (-) Salmonella!Ames assay

 	—	
  (+) Salmonellal'Ames assay

  (+) DNA strand breaks in human cells
  in vitro
                                       (+) DNA strand breaks in human cells
                                       in vitro
                                       (+) Salmonellal'Ames assay
                                       (+) unscheduled DNA synthesis in
                                       LLC-PK, cells
            Reference
           _	
 Kline et al., 1982
_.
 Singh and Sinha,  1976,  1979; Kafer,
 1985; Gualandi, 1987; Sora and
 Aeostini-Carbone, 1987

 Degrassi and Tanzarella, 1988

 Russo et al., 1984; Liang and
 Pacchierotti, 1988
 —_-^	•—	
 Haworth et al., 1983
  Nelson and Bull, 1988

  Bronzetti et al., 1984
 __——^———

 	.	
  Nelson and Bull,  1988
 ^
  Chang et al.,  1989

  Bhunya and Behera, 1987

  Waskell, 1978
               "

  —.            —
  Bignamiet al., 1980

   Chang etal., 1989

  	                 ~

                    •
   Chang et al., 1989
  	
   Green and Odum, 1985; Dekant et
   al., 1986
  _«^_i^—^—^—•
   Vamvakas et al.,  1989b

   Vamvakas et al.,  1989a
    »+ designates positive; - negative; wk weak response; ? inconclusive test.
                                                           18

-------
    gene conversion in yeas. (Bronzetti e, al., 1984). Chloral hydrate also has been observed ,o
    cause tumors in mice (Rijhsinghani e, a,., 1986). Other chtoroacetaMehydes are potently
    mutagenic.  Dichloroacetaldehyde (DCAA) is mutagenic in «he &«a/mamma,ian
    m.crosome test (Bignami et a,., 1980), and monochloroacetaMehye and, to a lesser extent
   DCAA appear to induce DNA single-strand breaks in cultured human ceils (Chang et al.,'

         Few genotoxicity studies are available on the carcinogenic perchloroethylene
   metabolites trichloroacetic acid (TCA) and dichloroaeetic acid (DCA).  TCA and DCA have
   been reported to produce stogie-strand breaks to hepatic DNA of mice and rats.  This action
   » mdependen. of peroxisome proliferation and of liver necrosis  (Nelson Snd Bull 1988-
   Nelson e, al., 1989).  The induction of DNA stogfe-strand breaks could not be confirmed by
  other laboratories, but a different methodology was used (Chang et al.,  1989)  TCA was
  reported as positive  for the induction of chromosomal aberrations and micronuclei to the
  bone marrow of mice (Bhunya and Behera, 1987). This finding is questionaWe,  however
  because of the low background frequencies report for chromosomal  aberrations and the'
  anomalous dose response seen for mieronuclei formation in nonnochromatic erythrocyfcs
        In rats, perchloroethytene has been shown to be metabolized ,o a cytotoxic, mutagenic
  fractton through a conjugate beta-lyase pathway (reviewed by Anders e, al.  1988 and
  Dekant e, al., 1989;  see also section 3 of this document).  This secondary metabolic
 pathway, although quantitatively minor, may be of major toxicologic importance  The
 pathway is toitiaily catalyzed by hepatic cy.oso.ic and microsoma. glutathione S-^ansferases
 to y,e.d S-a.^-trichlorovinyl) glutathkme ^^  After ^^ ^ ^        ^
 me.abo.ized to S-(l,2,2-tricblorovmyl) cysteme (TCVC) by the enzymatic remova! of
 glutamy, and giyeme  r^idues.  TCVC is ace«yla.ed by a reversible reaction to N-acetyl-S-
 (l,2,2-tnchlorovtayl)-L-cysteine, which is excreted in the urine.  However, TCVC is  also a
 substrate for renal beta-lyases that cleave TCVC to yie.d  an unstab.e thio, that may give rise
 to cytotoxic and mutagenic intermediates.
       In vivo and in  vitro experiments provide evidence  to support this metabolic  and
cytotoxic/mutagenic scheme. Dekant et al. (1987) and Green et al  (1990) have
demonstrated hepatic conjugation of perchloroethylene with glutathione by rodent liver
                                         19

-------
fractions in in vitro experiments.  Vamvakas et al. (1989b) found TCVG in the bile excreted
by perchloroethylene-perfused rat livers, and Green et al. (1990) reported the presence of the
glutathione conjugate in the bile of rats administered perchloroethylene by gavage.
       Human liver samples have been compared with rat and mouse liver samples with
respect to their ability to conjugate perchloroethylene with  glutathione (Green et al., 1990).
These investigators found low levels of conjugation by rat and mouse livers but were unable
to demonstrate conjugation by the human livers.  To confirm the viability of glutathione
S-transferase in the rat and human liver samples studied, these workers compared the two
species with respect to their ability to conjugate l-cWoro-2,4-dirntrobenzene with glutathione.
Both species carried out this conjugation rapidly and at essentially the same rate.  This
 finding indicates that failure of human liver samples to conjugate perchloroethylene was not
 attributable to an inactive glutathione S-transferase.
        An inability of human liver to conjugate perchloroethylene with glutathione indicates
 that toxicologic effects attributable to conjugative metabolites in animals would have little, if
 any, relevance to human health hazard.  However, because of the limited number of human
 livers tested,  it is impossible to conclude that humans are unable to carry out this metabolic
 step.  Additional confirmatory studies are clearly needed. Human liver does conjugate other
 chloroalkenes, including trichloroethylene, with glutathione (Oesch and Wolf, 1989).
        Vamvakas et al. (1989b) studied the mutagenicity of chemically synthesized TCVG in
  a modified Ames protocol employing Salmonella typhimurium TA100.  In the absence of an
  exogenous activating system, the conjugate produced a weak mutagenic response. In the
  presence of rat kidney microsomes, mitochondria, or cytosolic fractions (sources of gamma
  glutamyl transferase [GGT] and dipeptidase), TCVG caused marked, dose-related mutagenic
  responses.  These responses were reduced when the protein fraction was pretreated with
  either a beta-lyase or a GGT inhibitor.  A mutagenic jesponse was not observed when
  hepatic enzymes were used in place of kidney fractions.
         The results of these experiments show that TCVG requires metabolic activation to
  express  its marked mutagenic activity.  Further, the enzymes required to carry out this
  activation are found in the rat kidney, not in the liver.  This distribution of enzymes is
   consistent with  the production of renal tubule neoplasia in perchloroethylene-treated rats.
                                             20

-------
         Bile collected from rat livers perfused with perchloroethylene was found to contain
  TCVG and,  when tested in the Ames protocol using kidney particulate fractions as the
  activating system, was clearly mutagenic.  As in the experiment with synthetic TCVG,
  inhibition of renal beta-lyase or GOT reduced the mutagenicity of the bile samples
  (Vamvakas et al., 1989b).
        Green et al. (1990) reported the presence of N-acetyl-S-(l,2,2-trichlorovinyl)-L-
  cysteine in the urine of rats dosed with perchloroethylene by gavage and in rats and mice
  dosed by inhalation.  These investigators have also shown that renal cytosolic beta-lyase from
  rats, mice, and humans is capable of metabolizing TCVC.  Others also have  shown that the
  beta-lyase pathway is present in human proximal tubular cells and is responsible for
  activating glutathione or cysteine conjugates of a variety of chloro- and fluoroalkenes to
 reactive metabolites capable of binding to cellular macromolecules (Chen et al., 1990).
        In additional studies, TCVG has been found to induce unscheduled DNA synthesis in
 a porcine kidney cell line (Vamvakas et al., 1989a), and TCVG and N-acetyl-S-( 1,2,2-
 trichlorovinyl)-L-cysteine have both been found to be mutagenic in the Ames  test (Dekant et
 al., 1989).                                                          I
       The available data indicate that metabolism is a prerequisite for perchloroethylene
 mutagenicity.  The data do not support classifying the parent compound per se as a mutagen.
 Although certain metabolites of oxidative metabolism may be mutagenic (e.g., the
 chloroacetaldehydes including chloral hydrate), these positive data are predominantly limited
 to in vitro studies.  Moreover, perchloroethylene was assayed in the presence  of several
 types of metabolic activation systems (e.g.,  liver homogenates and intact hepatocytes) that
 would favor oxidative  metabolism and, under these conditions, predominantly  negative
 results were found.                                                   !
       Perchloroethylene may also be activated by a minor pathway involving conjugation
 with glutathione followed by renal processing of the S-conjugate.  This S-conjugate is a beta-
 lyase-dependent mutagen in the Salmonella/mammalian microsome assay. ! Mutagenic
metabolites formed in the kidney  could conceivably contribute to the tumors observed in
male rat kidneys.  However,  these mutagenicity studies of perchloroethylene metabolites
formed by the kidney are in vitro only.
                                           21

-------
       The mutagenicity studies on metabolites of perchloroethylene emphasize the need for
further studies concerning a mutagenic role for them hi perchloroethylene carcinogenesis.
                                             22

-------
                             5.  MOUSE LIVER TUMORS

 5.1.  CARCINOGENICITY BIOASSAY DATA AND EPA'S POSITION
        In carcinogenicity bioassays, perchloroethylene has been shown to cause a statistically
 significant increase in the incidence of hepatocellular carcinoma in both sexes of B6C3F1
 mice, following either oral gavage administration or inhalation exposure (NCI, 1977; NTP,
 1986a).
        In a study conducted by NCI (NCI, 1977), groups of 50 male mice received time-
 weighted average doses of 536 or 1,072 mg/kg of perchloroethylene in corn oil by
 intragastric gavage for 78 weeks (450 or 900 mg/kg for 11 weeks, then 550 or 1,100 mg/kg
 for 67 weeks). Groups of 50 female mice received time-weighted average doses of 386 or
 772 mg/kg of perchloroethylene in corn oil by gavage (300 or 600 mg/kg for  11 weeks, then
 400 or 800 mg/kg for 67 weeks).  Mice were dosed 5 days per week.   The perchloroethylene
 used in the study was greater than 99 percent pure, but impurities were not identified (NCI,
 1977;  U.S. EPA, 1985a). The test sample, however, was estimated to contain
 epichlorohydrin concentrations of less  than 500 ppm (U.S.  EPA, 1985a). It was considered
 unlikely, however, that the tumor response resulted from this low concentration of
 epichlorohydrin. Perchloroethylene caused statistically significant (p< 0.001) increases in the
 incidences of hepatocellular carcinoma hi both sexes of mice in both treatment groups when
 compared with untreated controls or vehicle controls.  The  tune to tumor was decreased in
 treated mice.
                                                               I
       Additional studies reported by the NTP confirmed the finding of hepatocellular
 carcinoma in B6C3F1 mice exposed to perchloroethylene.   Groups of 50 mice of each sex
 were exposed to perchloroethylene concentrations of 0, 100, or 200  ppm by inhalation
 exposure, 6 hours a day, 5 days per week, for 103 weeks.  Perchloroethylene caused dose-
 related statistically  significant increases hi the incidences of hepatocellular carcinoma in both
 sexes.
       The biologic significance of chemically induced mouse liver tumors, with respect to
identifying human hazard and using such tumor data in assessing cancer risk to humans, is a
subject of extensive debate. The controversy surrounding the liver tumor response in the
                                         23

-------
B6C3F1 mouse specifically is well recognized and has been ongoing for some time.  Several
meetings and symposia on the subject have been held, and numerous publications deal with
different aspects of the subject (e.g., Popp,  1984; Stevenson et al., 1990).  The EPA is aware
of the divergent scientific views regarding the predictive validity of mouse liver tumors in the
assessment of carcinogenic risk in general, and in the case  of perchloroethylene in particular.
The EPA extensively reviewed the issues concerning mouse liver tumors while it was
developing its guidelines for carcinogen risk assessment and has  kept abreast of the issues
since that tune.
       The relevance of mouse liver tumors to the assessment of carcinogenicity hi humans
has been questioned  because of:

       •   The high, and sometimes variable, background incidence of spontaneously
           occurring tumors in certain strains of mice, particularly the male B6C3F1 mouse
           used in the perchloroethylene studies conducted by NCI and NTP;
       •   The observation that liver cancer is a relatively uncommon cause of death in the
           United States (although not worldwide) (Pickle et al., 1987; Page and Asire,
           1985); and
       •   Some of the hypothesized mechanisms for mouse liver tumorigenesis that many
           scientists  believe would be unlikely to occur in humans.

       On the other hand, many scientists believe that mouse liver  tumors are as relevant as
 any other tumor type observed in laboratory test animals.  This  point of view concurs with
 the philosophy of using a sensitive model to detect a response in small numbers of test
 animals.  Also, certain proposed mechanisms, such as oncogene activation, involve steps that
 are comparable to those observed hi the development of other tumor types both in animals
 and in humans (McConnell, 1990).  At least 8 of the fewer than 30 known human
 carcinogens cause liver tumors hi mice, and most of these chemicals also cause other types
 of tumors hi rodents (IARC, 1987).
        Hepatocellular tumors are common end points hi rodent  carcinogenicity studies.  Of
 the chemicals  tested hi the NTP's bioassay program, 50 percent of those testing positive
 caused increased incidences of liver tumors hi mice. Most,  however, also caused other
                                           24

-------
 tumors in mice or rats as well (Maronpot et al., 1987; Haseman et al., 1984b). Only about
 5 to 6 percent of the compounds studied in the NTP carcinogenesis bioassays induced only
 mouse liver tumors (McConnell, 1990). Of the chemicals evaluated as carcinogens by the
 EPA, fewer than 10 percent have been found to cause only mouse liver cancer (Deal,  1990).
       At this tune, the Agency's position is that increased incidences of mouse liver tumors
 are considered evidence for human carcinogenic potential, although the evidence may be
 downgraded on a case-by-case basis according to chemical-specific data. The current EPA
 policy for evaluating mouse liver tumor data is described in the Guidelines for Carcinogen
 Risk Assessment published in 1986 (U.S. EPA, 1986b):             !
       An increased incidence of neoplasms that occur with high spontaneous
       background incidences (e.g., mouse liver tumors and rat pituitary tumors hi
       certain strains) generally constitutes "sufficient" evidence of carcinogenicity
       but may be changed to  "limited" when warranted by the specific information
       available on the agent (p. 1-7).                             I
                                                                 i
       For a number of reasons, there are widely diverging scientific views about the
       validity of mouse liver tumors as an indication of potential carcinogenicity in
       humans when such tumors occur in strains with high spontaneous background
       incidence and when they constitute the only tumor response to an agent.
       These  Guidelines take the position that when the only tumor response is hi the
       mouse liver and when other conditions for a classification of "sufficient"
       evidence in animal studies are met (e.g., replicate studies, malignancy; see
       section IV), the data should be considered as "sufficient" evidence of
       carcinogenicity. It is understood that this classification could be changed on a
       case-by-case basis to  "limited," if warranted, when factors such as the
       following are observed:  an increased incidence of tumors only in the highest
       dose group and/or only  at the end of the study; no substantial dose-related
       increase  in the proportion of tumors that are malignant;  the occurrence of
       tumors that are predominantly benign; no dose-related shortening of the time
       to the appearance of tumors; negative or inconclusive results from a spectrum
       of short-term tests for mutagenic activity; the occurrence of excess tumors
       only hi a single sex (pp. 1-5, 1-6).
       Thus, hi the absence of convincing evidence to the contrary, the Agency considers
increased incidences of mouse liver tumors in replicate studies to be "sufficient" evidence of
carcinogenicity.
                                           25

-------
       The 1987 EPA paper on the weight-of-evidence classification for perchloroethylene1

was hi keeping with the Agency's Guidelines for Carcinogen Risk Assessment.  The EPA

paper stated that:
       a strong carcinogenic response has been demonstrated in two separate
       experiments, hi different laboratories, using different routes of exposure,
       producing similar dose-related responses, increases the weight of the evidence
       that the response is indicative of a carcinogenic response in animals.  While
       this interpretation can be debated because the response is seen in mouse liver
       and is accompanied by some non-neoplastic pathology, the confirmatory
       finding as well as the nature of the response is viewed by many in the science
       community as "sufficient evidence" of an animal carcinogenic response, as is
       stated hi the Agency's Carcinogen Risk Assessment Guidelines.  Additional
       support for this view comes from the recent deliberations on the classification
       of perchloroethylene by IARC/(see footnote 5 in the conclusion section)/.... It
       is the position of the Agency,  therefore, that the new inhalation liver tumor
       data from the NTP study should add to the weight-of-evidence determination
       for perchloroethylene.


       The EPA paper and the most recent  letter from the SAB concerning perchloroethylene

 (personal communication) are consistent regarding statements about the mouse liver tumors.
 The SAB letter  in fact stated that the Board's consensus on the significance of mouse liver

 tumors was  "that mechanistic explanations  are not sufficiently well developed and validated
 at this time to change EPA's present approach expressed in its risk assessment guidelines for

 carcinogenicity."  The SAB concluded that


        the generation of mouse liver tumors by chemicals is an important predictor of
        potential risks to humans.  Of the several mechanistic models under
        consideration (including regenerative hyperplasia, oncogene activation and
        trihalomethyl radical formation) the  one most promising for immediate
        application to risk assessment is characterized by proliferation of peroxisomes,
        an intracellular organelle, hi the liver.
      staff paper sent to the SAB as an attachment to the August 3, 1987, letter from EPA's
  Administrator, written hi response to the formal comments submitted to the Agency by the
  Halogenated Organics Subcommittee regarding the public SAB review of the draft addendum
  to the HAD on perchloroethylene.

                                            26

-------
5.2. PEROXISOME PROLIFERATION AND PERCHLOROETHYLENE
       Beginning in 1986, additional information about the possible link between peroxisome
proliferation and liver cancer in B6C3F1 mice exposed to perchloroethylene has been
published (Odum et al., 1988; DeAngelo et al.,  1989; Goldsworthy and Popp, 1987).  These
newer data warrant an evaluation with respect to the interpretation of perchloroethylene
mouse liver tumor data as they relate to human health hazard.
       A chemically induced increase in numbers of hepatic peroxisomes, generally referred
to as peroxisome proliferation, has been suggested as the underlying mechanism through
which perchloroethylene induces hepatocellular carcinomas in B6C3F1  mice (Odum et al.,
1988).  Carcinomas are proposed to arise as  a result of oxidative damage to the cell, possibly
at the level of DNA, caused by elevated concentrations of hydrogen peroxide, a peroxisome
degradation product. Hydrogen peroxide is normally degraded by a peroxisomal catalase,
but the activity of this enzyme does not increase in a parallel fashion with peroxisomes and
other peroxisomal enzymes following perchloroethylene exposure. This enzymic unbalance
may result  in the accumulation of cytotoxic concentrations of hydrogen peroxide.  Although
DNA is identified as a potential ultimate target of oxidative damage, the mechanism is still
described as "epigenetic" or "nongenotoxic."  These terms are used in  this context to contrast
DNA-damaging events that are secondary to  other effects caused  by perchloroethylene from
DNA damage produced by a direct, primary interaction of perchloroethylene  or its
metabolites with DNA.                                           !
       Assuming that the observed perchloroethylene-induced peroxisomal proliferation and
liver hepatocellular carcinomas in the B6C3F1 mouse are related, one course  of reasoning
supports the hypothesis that the liver tumors could be unique to that species if this is the only
mechanism, therefore,  decreasing their predictive validity relative to human health hazard.
The central points of the supporting rationale are:                  \
       •  A major metabolite of perchloroethylene, trichloroacetic acid, is the peroxisome-
          inducing agent and, therefore, the cancer-causing agent in mouse liver.
       •  Because of a lower rate of metabolism and metabolic enzyme saturation at
          relatively low concentrations of perchloroethylene, rats are not as efficient as mice
                                          27

-------
           in metabolizing perchloroethylene to TCA, which explains the lack of a
           hepatocarcinogenic effect in rats. This also implies that a threshold level of TCA
           must be exceeded before peroxisomal proliferation and liver cancer can occur.

       •   Humans are even less efficient metabolizers of perchloroethylene than rats.

       •   Human liver cells would probably not respond to the peroxisome-proliferating
           activity of TCA even if sufficient levels of the chemical could be produced in
           humans (because when human liver cells are challenged in vitro with TCA at
           concentrations known to cause peroxisome proliferation in cultured mouse and rat
           hepatocytes, no evidence of increased peroxisomal  activity can be detected).

       •   Human hepatocytes are also not as responsive as rodents to other known
           peroxisome proliferators such as the hypolipidemic drugs and phthalate esters.


       The conclusion from the above rationale is that humans, because of an inability to

generate sufficient TCA levels from perchloroethylene metabolism and a general

unresponsiveness to peroxisome-proliferating agents, are unlikely to show a hepatocellular

carcinogenic  response to perchloroethylene, assuming that this is the mechanism of action in
mice.

       Many aspects of the above statements are supported to  some extent by experimental
data.
          TCA has been shown to be a major metabolite of perchloroethylene (Odum et al.,
          1988); TCA has been shown to cause peroxisome proliferation (as measured by an
          increase hi peroxisomal enzyme activity) in hepatocytes of mice and rats both in
          vivo and hi vitro after short-term exposure (Elcombe, 1985).  Further, TCA has
          been shown to be a hepatocellular carcinogen in the B6C3F1 mouse (Herren-
          Freundetal., 1987).

          Perchloroethylene oxidative metabolism approaches saturation at lower levels in
          rats than hi mice.  Saturation in rats occurs at atmospheric concentrations in
          excess  of 100 ppm (Ikeda et al., 1972).  Consequently,  at high atmospheric
          concentrations of perchloroethylene (i.e.,  > 100 ppm) mice generate relatively
          more TCA than do rats (Odum et al., 1988).  Following 6 hours of inhalation
          exposure to 400 ppm, the cumulative blood concentrations of TCA in mice were
          six to seven times greater than concentrations in rats; peak blood levels of TCA
          were found to be 13-fold higher in mice than in rats.
                                          28

-------
 These results show that there is a quantitative difference between rats and mice
 with respect to their abilities to metabolize perchloroethylene to TCA. Such a
 difference is consistent with the known species variability in responsiveness to the
 hepatocarcinogenic effects of perchloroethylene.  In view of the peroxisome-
 inducing activity of TCA in the rat, which for equivalent doses may be even more
 responsive than the mouse (Elcombe, 1985), and the belief that peroxisome
 proliferation is a necessary prerequisite to perchloroethylene-induced
 hepatocellular carcinogenesis,  these results also suggest that TCA per se  should be
 hepatocarcinogenic in rats if sufficient blood levels are achieved (Elcombe  1985).
 Although a two-year bioassay  was stated to be in progress (.Elcombe, 1985), the
 EPA is unaware of published  studies on the hepatocellular carcinogemcity of TCA
 hi rats.

 Saturation of human perchloroethylene metabolic processes has been reported to
 occur at perchloroethylene concentrations of  approximately 100 ppm to 400• ppm
 (US  EPA, 1985a; Ikeda et al., 1972; Ohtsuki et al, 1983).  Odum e al. (1988)
 summarized evidence suggesting that humans exposed to perchloroethylene would
 be "exposed to lower concentrations of TCA than mice or rats.

  In in vitro experiments conducted to compare rat,  mouse, and human hepatocytes
  with respect to susceptibility to TCA-induced peroxisome proliferation  mouse
  hepatocytes are more responsive than rat, and human hepatocytes have been found
  to be relatively unresponsive  (Elcombe, 1985).

  If peroxisome proliferation is required for perchloroethylene-induced liver
  carcinogenesis, reduced human hepatocyte responsiveness to TCA combined  with
  reduced ability to form TCA supports the hypothesis that perchloroethylene is
  unlikely to be carcinogenic in humans.  Inasmuch as this study was based on only
  two human livers, the  evidence cannot be considered persuasive at this  point,
  however  Considerable individual variation in function may be expected to exist
  between human livers, particularly in specimens from donors who may have been
  treated with a variety  of drugs.  More human liver samples need to be  examined
  before a convincing argument can be made.

.  There is some evidence that hepatocytes from humans and other primates are
   relatively unresponsive to a  variety of agents that cause peroxisomal proliferation
   in rodents, although this evidence is limited in quantity and scope.  Microscopic
   studies on liver biopsies from humans chronically dosed with hypolipidemic drugs
   suggest that these agents do not cause peroxisome proliferation in humans (Canot
   et al   1983- De La Inglesia et al., 1982; Hanefeld et al., 1983).  Biochemical
   data  'such as peroxisomal enzyme measurements, have not been reported,
   however  Nafenopin, a peroxisome inducer in rodents, was inactive in cultured
   marmoset hepatocytes (Bieri et al., 1988), but the drug did cause increased  cell
   proliferation.  Several in vitro studies also  have provided suggestive evidence that


                                    29

-------
           human hepatocytes are relatively unresponsive to hypolipidemic drugs and
                     ter plasticizers (Elc°mbe and Mitchell, 1986; Butterworto et al
                     e T? IT"™*' ** Sampl£ Size is lhnited to a few human "v'ers.
                   tion of the data is subject, therefore, to the same uncertainty as the in
           vitro TCA studies on human hepatocytes.
       Although evidence exists to support the hypothesis that mouse liver cancer associated

with exposure to perchloroethylene may be secondary to peroxisome proliferation, there are
points that run counter to the hypothesis, and thus its validity remains questionable.  For
example:
       '   If peroxisome proliferation is causally related to the induction of liver cancer one
          would expect to detect a quantitative relationship between the two events  That is
          potent peroxisome proliferators should also be potent hepatocarcinogens ' This   '
          does not appear to be the case.

          In a chronic 65-week study of DCA and TCA in male B6C3F1 mice Herren-
          Freund et aL (1987) found that DCA was a more potent hepatocarcinogen than
          1C A.  Equidoses (5 g/L in drinking water) resulted in a nearly threefold higher
          incidence of liver cancer in DCA-dosed animals than in TCA-dosed animals
          DeAngelo et al.  (1989), however, reported that TCA was more potent as a '
         peroxisome proliferator than was DCA in male B6C3F1 mice.   Nelson et al
         DPA I Sr?p°rd ^ TC/Produced Sreater Peroxisome proliferation than did
         DCA in B6C3F1 mice dosed for only 10 days.

         An even more pronounced lack of correlation was reported in a study of the
         peroxisome proliferators DEHP and Wy-14643 (Marsman et al., 1988)   At doses
         producing equivalent increases  in peroxisome volume density and peroxisomal
         enzyme activity for the two compounds, the liver lesions and tumors were
         produced only by Wy-14643 (100 percent  incidence). DEHP produced no liver
         lesions  in dosed rats.

         Studies on the Swiss mouse also raise questions about the connection between
         peroxisome proliferation  and cancer.  If mouse liver peroxisome proliferation in
         response to perchloroethylene is correlated to the induction of liver cancer one
         would expect strains of mice that exhibit peroxisome proliferation to also exhibit
         hepatocellular carcinoma  induction.  While the EPA is unaware of
         perchloroethylene bioassays being conducted in mouse strains other than B6C3F1
         and Strain A, the closely related hepatocarcinogen in B6C3F1 mice
         tachJoroethylene, has been studied in the Swiss mouse (Henschler et al  1984)
         Tnchloroethylene, which is metabolized to TCA, induces peroxisome proliferation

                                        30

-------
in Swiss mice (Elcombe, 1985). Hepatic peroxisome proliferation was induced as
measured by both an increase in peroxisomal enzymes and peroxisome density
volume.  Induction of the peroxisome marker enzyme, cyanide-insensitive
palmitoyl-CoA oxidase activity, increased linearly with increasing
trichloroethylene doses of 0.05-0.5 g/kg/day following 10-day exposure periods
(Elcombe,  1985).  TCA, a major metabolite of trichloroethylene, also induced
peroxisome proliferation in Swiss mice.  The daily oral administration of
trichloroethylene to Swiss mice for 18 months did not produce liver cancer in
either sex,  however (Henschler et al., 1984).  The doses  administered by
Henschler et al. were from three to five times higher than doses found by
Elcombe to induce peroxisome  proliferation in these mice.

The results  of recent studies have raised the possibility that genotoxicity may
occur independently of peroxisomal proliferation following exposure to
perchloroethy lene.

Recent studies have shown that the TCA and DCA metabolites of
perchloroethylene have DNA-damaging activity without increasing the levels of
peroxisomal enzymes.  Within  1 hour after a single dose of TCA or DCA at 0.5
g/kg, a significant increase in single-strand breaks hi DNA was detected with an
alkaline unwinding assay (Nelson et al.,  1989).  No increase  in peroxisomal
palmitoyl-CoA oxidase activity was detected for periods up to 24 hours after
dosing. This study raises the possibility  that genotoxicity, and thus potential
mutagenicity, may occur independently of peroxisome proliferation following
perchloroethylene exposure.  As discussed earlier,  other investigators have been
unsuccessful hi demonstrating single-strand breaks hi DNA with TCA and DCA
(Chang etal., 1989).

Other metabolites, such as  chloroacetaldehydes like chloral hydrate,  might also
contribute to liver tumorigenesis. A preliminary study in mice indicates that
chloral hydrate is a hepatocarcinogen in mice (Rijhsinghani et al., 1986).  Chloral
hydrate has been shown to produce aneuploidy and may be mutagenic (see section
4.2.).                                                 ;

More recently, Nelson et al. (1990) reported that TCA administration significantly
increased expression of the C-H-ras and  c-myc oncogenes in hepatocellular
carcinomas in B6C3F1  mice.   The data indicate that elevated expression of these
oncogenes may play an important role in the development of liver tumors in these
mice. As discussed by these authors, the consequence of increased c-myc
expression  hi most cells is loss of cellular differentiation. The oncogene does not
induce cell division but appears to play a permissive function in relation to cell
division.  These workers suggest that TCA-increased c-myc expression may
prevent initiated cells from differentiating, thereby  increasing their probability of
progressing to hepatocellular carcinoma.  Several investigators have reported a
                                 31

-------
          different pattern of oncogene activation in chemically induced mouse liver tumors
          compared with that observed in the tumors of untreated animals. This indicates
          that the chemicals do not simply promote spontaneous background tumors (Fox et
          al., 1990; Reynolds et al.,  1987; Oshimura et al., 1988).
       In summary, some evidence supports the hypothesis that perchloroethylene-induced
hepatic carcinogenesis may be related to peroxisome proliferation, but critical review of the
scientific literature reveals significant data gaps regarding the relationship between the
proliferative effect and neoplasia.  The recent demonstration of a peroxisome-proliferator-
activated receptor (Issemann and Green, 1990) should lead to increased understanding of the
mechanism of action for chemicals causing this phenomenon. Also, the recent demonstration
that the major metabolite of perchloroethylene,  TCA, causes the expression of the c-myc
oncogene in B6C3F1  mice requires experimental exploration.
                                          32

-------
                       6.  KIDNEY TUMORS IN MALE RATS
      The inhalational administration of perchloroethylene to male and female F344/N rats
and B6C3F1 mice produced dose-related increases in the incidences of nontumor
nephrotoxicity in both sexes of both species and a nonstatistically significant increase in the
incidence of proliferate lesions of the renal tubular cells  (tubular cell hyperplasia, adenoma,
and adenocarcinoma) in male rats (NTP, 1986a). A slight increase in renal tumors was also
observed in male Sprague-Dawley rats receiving perchloroethylene by gavage or by
inhalation in other studies (Maltoni and Cotti, 1986; Rampy et al., 1978).
       In the NTP studies, groups of 50 male and 50 female F344/N rats were exposed by
 inhalation to atmospheres containing 0, 200, or 400 ppm perchloroethylene for 6 hours a
 day  5 days per week for 103 weeks. Tubular cell hyperplasia was observed in male  rats
 (control 0/49, low dose 3/49, and high dose 5/50) and in one high-dose female. Renal tubule
 neoplasms were observed in male rats (control  1/49, low dose 3/49, and high dose 4/50).
        Although the incidence  of renal tubule neoplasms  in perchloroethylene-exposed male
 rats  was not statistically significant (p>0.05) relative to concurrent controls, the production
 of the lesions is considered to  be evidence of a carcinogenic effect in rats. This is supported
 by the following facts:                                          ,



                                                            ^
                      .  percem]).  Likewise, the overall historic* control mcidence  of renal
            tubule tumors in male F344/N vehicle controls in gavage studies  is 1/1,943 (0. 05
            nercent)   The incidence is even lower in female controls. This is supported by
            s«ou?«ml lule tumor incidence rates recorded for  other rat strams £.«..
             Osbcme-Mendel, males 0.3 percent; females 0 percent;  Goodman et al.,  1980).
             Tne ^oearance  of tubule neoplasms in 8% of perchloroethylene-dosed animals
             JowloSTd h$,lse groups combined) is convincing evidence of a treatment-
             related effect.                                       ,
         •   No malignant renal tubule neoplasms have been observed in any control rats
             elaSnefby me NTP.  This includes the chamber controls  from the performing
             Satory an?me untreated controls and the vehicle controls from gavage studies.
                                             33

-------
              Two of the tumors observed in high-dose animals in the NTP study were
              carcinoma^ The probability of two rare carcinomas appearing by^chTnS in a
              group of 50 animals has been calculated to be less than 0.00 "(US Epl 1987)
         In addition to the NTP study findings of renal tubule tumors in male F344/N rats
  Rampy et al. (1978) and Maltoni and Cotti (1986) reported slight increases in renal tumors in
  Sprague-Dawley rats dosed with perchloroethylene by inhalation and gavage, respectively
         There is good evidence that the tubule tumors are not unique to the administration of
  perchloroethylene.  The NTP has found low incidences of tubule neoplasms in rats dosed
  with other chlorinated ethanes and ethylenes (NTP, 1983, 1988; and unpublished results  cited
  in NTP, 1986b).  There is some evidence that nontumor pathology is not unique to
  perchloroethylene; however, there is also evidence that the nephrotoxicity observed with
  certain chemicals of this group,  such as pentachloroethane, may be different from that seen
  with others of these  compounds, such as trichloroethylene.
        The data support the conclusion that the chronic administration of perchloroethylene
 produces nephrotoxicity in both sexes of mice and rats and an increased incidence of
 proliferate lesions of the kidney tubules in male rats.  The use of these data to infer risk of
 carcinogenesis to humans, however, is  a focus of scientific debate. Of particular
 consequence in this debate is the possibility that the induction of renal tubule tumors by
 perchloroethylene may be unique to male rats, and therefore, is inappropriate for deducing
 potential human health hazard. This is  because evidence exists to suggest that renal effects
 induced in male rats by chemicals causing alpha-2u-globulin accumulation are unlikely to
 occur in any species not  producing alpha-2u-g,obulin or a protein with a structurally similar
binding domain, in the large quantities typically seen in the male rat.  Thus, if a chemical
                                           34

-------
 induces alpha-2u-globulin accumulation in hyaline droplets and a carcinogenic response in the
 male rat kidney, the tumor response may not constimte evidence of a carcinogenic hazard to
 humans.
        The EPA is presently developing criteria that will define a weight-of-ev,dence
 approach for evaluating, on a case-by-casc basis,  the role of a,pha-2u-g.obulin in rat kidney
 tumor formation (U.S. EPA,  1991).  A report (U.S. EPA, 199!) currently being developed
 by a technical panel of EPA's Risk Assessment Forum provides guidance on determnung
 when it is reasonable to presume that a renal tumor in male rats results from alpha-2u-
 globulin accumulation and on selecting appropriate procedures to  use in extrapohmon to
 humans under such circumstances.  The report also defines other situations that suggest a
  different approach and calls for research to clarify questions raised because of the existence
  of human proteins that may be structurally similar to alpha-2u-globulin.  Data on renal
  tumors in fl>e male ra, will fall into  several categories depending on whether the tumors are
  attributable solely to alpha-2u-globuUn accumulation, whether anofher mechanism apphes,
  whether several mechanisms are feasible, one of which involves  alpha-2u-globuhn, or
  whether the available information is inadequate to determine the  role of alpha-2u-globulm.
   For instance, if the perchloroethylene alpha-Zu-globuUn data are subsequently judged to be
   me only definitive explanation for the occurrence of male rat kidney tumors, mis tumor end
   point may not have  relevance  for human health hazard assessment. TOs can be further
   evaluated as the EPA's criteria for identifying chemicals inducing alpha-2u-globuhn
   accumulation become available to apply to the perchloroethylene-specific data.

    6 1   ALPHA-2U-GLOBULIN IN RENAL  CARCINOGENESIS IN MALE RATS
          A variety  of organic compounds studied by the NTP and others have been shown to
    produce sex- and species-specific lesions in the renal tubules of male rats in the form of
    hyaline droplet nephropamy (NTP, 1983, 1986b, 1987, 1990; Alden et al., 1985;
    MacNaughton and Uddin,  1984; Alden et a,.,  1984;  PhiUips ct al., 1987).  The accumulate
    of the protein alpha-2u-globu.in is believed to  be the reason for an excessive number of
    hyaline droplets  (Stonard et  a!., 1986; Olson et al., 1987). A  normal urinary protein tn the
j     male rat, alpha-2u-globulin is synthesized in me liver under hormonal control, bu, „ has no,
                                               35                     :

-------
   been detected in the liver of fema,e rats or in other species, including humans.  Among the
   chemicals tested so far in chronic animal bioassays, those tha, invoked mis specific type of
   protem droplet nephropathv in male rats also produced renal tubule tumors in male rats but
   dtd not produce renal  tubule tumors in other species tested. The renal tubule tumors appear
   to be the end product  in the following sequence of functional changes in the epitheHal cells of
   proximal tubules (UAREP,  1983; Alden e. al., 1984; Haider et a., 1984; Swenberg et al
   1989):
•  Excessive accumulation of hyaline droplets in proximal tubules
                      '^ '° ^ Cdl "'«''• ">'
                                                                        reDresentins,
                          of tubules within the renal medulla.                            Y'
        •   The chronic progressive nephropathy characteristically found in aging rats is
            exacerbated as a consequence of the induced nephrotoxicity.        g
        •   Renal tubule hyperplasia and neoplasia develop subsequently.

        A number of investigators hypothesize that the increased proliferate response caused
 by the chemically induced cytotoxicity results in clonal expansion of spontaneously initiated
 renal tubule cells and increased incidence of renal tumor formation (Trump et al   1984-
 Alden, 1989; Swenberg et al. , 1989). This line of reasoning leads to a conclusion that me
 acute and chronic renal effects induced in male rats by these chemicals will be unlikely to
 occur in any species not producing alpha-2u-globulin or a very closely related protein in the
 large quantities typically seen in the male rat (Alden, 1989; Borghoff et al.,  1990; Green et
 al., 1990; Flamm and Lehman-McKeeman,  1991).
       This proposed mechanism of tumorigenesis seems plausible and may provide an
adequate explanation of the specific susceptibility of the male rat to the induction of renal
tubule tumors by certain chemicals.  However, definitive links between alpha-2u-globulin
accumulation and tumorigenesis  in male rats must be established on a chemical-by-chemical
                                          36

-------
basis before it is reasonable to discount the significance of the tumor induced by a particular
chemical.  Guidance for evaluating data concerning alpha-2u-globulin and kidney tumors m
male rats is being developed by 
-------
        Alpha-2u-Globulin Response for Perchloroethylene Is R(
          Although the a,pha-2u-globulin response occurs in male rats exposed to
    per^oroethylene, it has been observed following only Ugh dose,  ^ ^
    tester inhaled doses of perch.oroetby.ene (up ,„ 400 ppm 6 hours per day  or
    wun arnmaU sacnficed within 18 hours of termination of the fina, exposure) in rats  but thTre
   was no evince of hyaline dropiet formation although .here may have been tin, for
   recovery  efore sacrifice.  ,t is noteworthy that the 400 ppm concentration was the same
                      , but fte age of the rats as weH as the ,ength of time Oaat ed
  between fina, exposure and sacrifice may expiain some of the differences.  Mineraiization in
  tne mner meduHa and papiiia of ^e Hdney, a characteristic trait of a-pha-^obuHn
  nephropathy, was not seen, however (NTP,  1986a).
        It is possible to longer term exposure to the 400 ppm concentration of

                       r* for "" hyaltoe ^ accumulation to «- ^ of -
           , 1990).  Alpha-2u-globu,in accumu,a«iOn can be demons.a.ed, however, after
 only short-term exposures (even a single administration) to severa, agents such as d-
  TssHlTt' ^^ gaS°ltae' "* "«*»*-»- (Charbonneau et a,., 1987;  NTP
 1988). lack of hyaline drop,et formation or increase in alpha-2u-globulin or signs of the
 charactenstic rena, nephropamy at the high-dose level of the  NTP inhalation study may
 -ndtcate a threshold effect and thus diminish the like,ihood tha, the rena, tumors associated
 w th exposure to perchloroethylene are induced through this mechanism (Green e, a,  ,990)
 Phannaco^eUc differences between oral and inhalation exposure may contribute to the
 observed discrepancies in some of the  results.
       The NTP did not report the presence of hyaline droplets in rats that had been exposed
to either 200 or 400 Ppm of percMoroethylene for up to 2 year,  These doses       "
                                         38

-------
associated with the production of renal tubule neoplasms in male rats.  The fact that the NTP
did not report the presence of hyaline droplets in either the 14-day, 90-day, or 2-year studies
is not definitive, however, because the NTP protocol at that time was not designed to detect
hyaline droplets or alpha-2u-globulin accumulation in the kidney (NTP, 1990).  Thus, the
procedures followed at the time of the study were not necessarily conducive to detecting
hyaline droplets. For example, in the chronic study of perchloroethylene, at least 1 week
elapsed between the final perchloroethylene exposure and the scheduled sacrifice of the
surviving animals.  It is possible that had the  hyaline droplets been present, they could have
regressed. Also, the nephropathy observed at the end of a 2-year bioassay could be difficult
to distinguish from the old-age nephropathy that occurs in these rats.  Other investigators
(Goldsworthy et al.,  1988; Green et al., 1990) have observed hyaline droplets containing
alpha-2u-globulin following high doses of perchloroethylene administered to male rats.
       In the NTP bioassay, however, the renal pathology reported is not entirely consistent
with the results generally found for chemicals where there is alpha-2u-globulin accumulation
 (NTP, 1986a; letter from Scot Eustis to William Farland,  1988). For example, as mentioned
 above, there was no mineralization in the inner medulla and papilla of the kidney, a frequent
 finding in bioassays of chemicals inducing alpha-2u-globulin accumulation (e.g., for
 pentachloroethane, the incidence of renal papillar mineralization is 8 percent in controls; 59
 percent,  low dose; 58 percent, high dose).  In addition, some aspects of toxic tubular
 nephropathy were also observed in female rats and male mice exposed to perchloroethylene.
        Perchloroethylene induces  alpha-2u-globulin accumulation and some of its associated
 nephropathy in male  rats, although the evidence for this exists only at high doses.
 Nevertheless, the hypothesis of hyaline droplet formation leading to renal tubule tumors in
 male rats is a valid proposal for the mechanism of tumorigenesis. The absence of evidence
  that chronic inhalational exposure to 200 or  400 ppm of perchloroethylene causes the
  accumulation of hyaline droplets and its associated nephropathy in the kidneys of male rats,
  considered along with data supporting other mechanisms, including possible genotoxicity
  discussed below, makes it difficult to conclude that perchloroethylene-induced renal tumors
  can be  attributed solely to this hypothesized species- or sex-specific mechanism.
                                             39

-------
  Chronically Induced Perchloroethylene Nonneoplastic Kidney Lesions Exhibit Neither
  Species Nor Sex Specificity.
         In contrast to most other chemicals inducing alpha-2u-globulin accumulation that have
  been tested by the NTP in chronic carcinogenicity bioassays, renal lesions occurring in
  animals exposed to perchloroethylene were not limited to the male rat.  Although the female
  rat did not develop any renal tubule tumors, the incidence of karyomegaly was significantly
  elevated in the female rat as well as in the male rat;  1 of 50 female rats exposed at the high
  dose developed tubule cell hyperplasia.
        In the mouse, "nephrosis" was observed at increased incidences in dosed females,
  casts were observed at increased incidences in dosed males and high-dose females, and
 karyomegaly of the tubular cells was observed at increased incidences in both sexes of
 treated mice.  The severity of the renal lesions was dose related, and one low-dose male had
 a renal tubular cell adenocarcinoma.
        In the NCI gavage study of perchloroethylene, toxic nephropathy, not detected in the
 control animals, occurred in both male and female  Osborne-Mendel rats  administered
 perchloroethylene. Unfortunately, the animal survival in this study was not adequate  to
 support any conclusions about perchloroethylene carcinogenicity.
        Other chlorinated ethanes and ethylenes produce nephrotoxicity and renal tubule
 tumors in laboratory animals as  well.  Hexachloroethane causes accumulation of hyaline
 droplets and  renal tubule tumors in male rats (NTP, 1989).  On the other hand,
 trichloroethylene, which was also tested by the NTP,  induces kidney tumors in male rats
 only (NTP,  1988b) but does not cause an accumulation of hyaline droplets or an increase in
 levels of alpha-2u-globulin (Goldsworthy et al., 1988). Consequently, kidney tumors induced
by this compound are not considered to be associated  with alpha-2u-globulin accumulation.
Perchloroethylene is structurally  closely related to trichloroethylene, and both chemicals have
been shown to be metabolized in the kidney to mutagenic compounds.
                                          40

-------
6.2. SUSTAINED CHRONIC NEPHROTOXICITY AS A POSSIBLE MECHANISM
INDEPENDENT OF ALPHA-2u-GLOBULIN ACCUMULATION
       Numerous compounds such as perchloroethylene, trichloroethylene, and
pentachloroethane have been reported to produce nephrotoxicity in male and female rats and
mice.  This toxicity, although appearing to be characteristic of chronic administration of
chlorinated ethanes and ethylenes, manifests itself differently with specific chemicals and
may include tubular cell cytomegaly, karyomegaly and pleomorphism, tubular cell dilation,
or the formation of granular casts.  Certain compounds cause kidney tumors in male mice
only (vinylidene chloride), male rats only (trichloroethylene), and both male rats and male
mice (chloroform).
       As previously discussed for the alpha-2u-globulin nephropathy, sustained kidney
damage may be a risk factor for tumorigenesis.  Thus, there may be a link between renal
toxicity and tumorigenesis, and  it is reasonable to suspect that renal tubule  neoplasia in male
rats may be influenced by perchloroethylene-induced cytotoxicity and subsequent cellular
 regeneration.  It also has been suggested that renal neoplasms induced by perchloroethylene
 may be secondary to renal cytotoxicity and subsequent cellular proliferation without regard to
 alpha-2u-globulin accumulation. If this is the case, renal tubule neoplasia in these
 experiments would not be expected to be a species- or sex-specific response to  chronic
 administration of perchloroethylene because the nontumor lesions appeared in both sexes of
 both species.  Perchloroethylene-induced cytomegaly and karyomegaly appeared in both rats
 and mice during the early phases of the NTP inhalation study, indicating that animals of both
 species  surviving to  the scheduled termination of the study had long-standing nephrotoxicity.
 If renal tubule neoplasia were directly consequent to this pathology, tumors would likely
 have been found in dosed female rats or male and female mice.  Goldsworthy  et al. (1988)
 determined that cell replication rates increased specifically in the histologically damaged
 tubule segments of male rats, but not in female rats, after perchloroethylene exposure.  Cell
  replication did not differ from  controls  in trichloroethylene-treated male or female rats,
  however. Because both trichloroethylene and perchloroethylene produce renal tubule tumors,
  but no  enhanced cell replication was  seen with trichloroethylene, it is difficult to conclude
                                             41

-------
  that perchloroethylene induces the renal tumors solely by a nephrotoxic mechanism apart
  from nephropathy associated with alpha-2u-globulin accumulation.
         The fact that there is little doubt that the kidney is a target organ for
  perchloroethylene and other chlorinated ethanes and ethylenes in mammalian species
  contributes to the overall concern regarding the kidney tumor end point.  Although
  nephrotoxicity may play a role, more supportive evidence is needed to define such a role.

  6.3.  A MUTAGENIC MECHANISM OF PERCHLOROETHYLENE-INDUCED
  CARCINOGENESIS IN MALE RATS
        The possible explanations of perchloroethylene-induced renal carcinogenesis discussed
  earlier have centered on nonmutagenic (epigenetic) mechanisms, because  mutagenicity studies
  of perchloroethylene have produced largely negative or only weakly positive results.  The
  early studies of genetic toxicology of perchloroethylene have centered on the effects of
  perchloroethylene per se and later on certain products of oxidative metabolism.  A secondary
  metabolic pathway for perchloroethylene (hepatic conjugation with glutathione and
  subsequent degradation by renal beta-lyases) has been discovered in rats (see sections 3 and 4
 on metabolism and mutagenicity;  also see Dekant et al., 1989; Vamvakas et al., 1989b).
 Perchloroethylene is conjugated with hepatic glutathione to form S-(l,2,2-trichlorovinyl)
 glutathione. The conjugative pathway in the liver appears to be the minor of two competitive
 pathways; its activity is thought to increase as the oxidative pathway approaches saturation.
 The glutathione S-conjugate metabolite thus formed in the liver is either excreted into the bile
 or  transported to the kidney where it is  acted on by gamma glutamyl transferee and
 dipeptidase to form its corresponding cysteine S-conjugate, S-(l,2,2-trichlorovinyl)-L-
 cysteine (Dekant et al., 1987). TCVC may undergo N-acetylation and be  excreted in the
 urine, or it may become a substrate for  renal beta-lyases that cleave TCVC to form a
 mutagenic fragment that probably  includes electrophilic acylating and alkylating agents.  The
 mutagenic activities of TCVG, in the presence of hepatic- and renal-activating systems, and
 TCVC, in the presence  of a renal-activating system, have been demonstrated in an Ames test
protocol.  The  conjugative pathways producing mutagenic metabolites in the kidney are
operative for trichloroethylene, a close structural analog of perchloroethylene (Anders et al.,
                                          42

-------
1988; Dekant et al.,  1989).  Trichloroethylene also induces renal tumors in male rats but is
not a'chemical that induces alpha-2u-globulin accumulation (NTP, 1988; Goldsworthy et al.,
1988; U.S. EPA, 1991).  The trichloroethylene conjugates that lead to mutagenic constituents
are dichlorovinylglutathione and dichlorovinylcysteine (Anders et al.,  1988; Dekant et al.,
1989).
       Although the beta-lyase enzymes necessary for the metabolism to the mutagenic
metabolite are found in the kidneys of rats, mice, and humans (Green et al., 1990;  Chen et
al., 1990), in vitro conjugation of perchloroethylene with glutathione by human liver was not
detected (Green et al., 1990).  If the human is incapable of conjugating perchloroethylene
with glutathione, this potential mechanism of carcinogenesis may not  be relevant to
projecting human health hazards associated with perchloroethylene.
        Green et al.  (1990) have reported species differences with respect  to
 perchloroethylene-glutathione conjugation. In vivo conjugation by rat liver occurred at a
 relatively low rate,  but this rate was five times greater than that observed for mouse liver.
 Using  a limited number of human liver samples, Green et al.  were unable to demonstrate
 conjugation of perchloroethylene.  Few liver samples were studied, however, and a
 conjugation rate tenfold lower than that observed for rats would fall below the limits of
 detection of the method employed.  Tenfold differences in enzyme activities within the
 human population are not uncommon. Consequently, it remains a distinct possibility that
 humans  may conjugate perchloroethylene, although probably  at a very low rate, as indicated
 by the low rate measured hi rodent tissue.
        The roles of this metabolic  pathway  in producing renal tumors in male rats and in the
  carcinogenic potential of perchloroethylene  in humans remain to be established. It has been
  pointed  out that the conjugative pathway is  minor and may be noticeably more active only
  when the oxidative pathway approaches saturation (Green et  al., 1990).  Human (and rat)
  oxidative metabolism of perchloroethylene is recognized as being saturable.  The  quantitative
  relationships between various degrees of saturation of the oxidative  pathway and concomitant
  importance of the  conjugative pathway require close scrutiny before it can be concluded that
  conjugation is unimportant in perchloroethylene metabolism  by humans, particularly since
  products include potential mutagens.  Moreover, human liver is known  to conjugate other
                                             43

-------
  chloroalkenes, related closely to perchloroethylene, with gutathione (see sections on

  metabolism and mutagenicity). For example, N-acetyl-S.(dichlorovmyl)-L-cysteine has been
  measured in urine of humans exposed to trichloroethylene in the occupational setting.

        Also, if the glutathione-beta-lyase pathway provides a mechanism for the induction of
 renal tumors,  it is difficult to explain why female rats and both sexes of mice did not exhibit

 renal tumors.  The metabolic processes required for the generation of the mutagenic

 intermediate are operative in both sexes of both species, albeit to a lesser extent in  female

 rats and both sexes of mice.  The male rat tumor rate was relatively low,  but the incidences

 in female rats  or male and female mice might be expected to be still lower.  The rates in

 female rats and both sexes of mice might be too low to be detected with the small numbers
 of animals subjected to testing.

       In summary, since the NTP discovery that chronic administration of perchloroethylene
 induces a low level of renal tubule tumors in male rats,  significant research has been

conducted to explain the mechanism of the carcinogenic effect.  This research has resulted in
at least three possible explanations:


      1.  The  tumors may be secondary to the renal accumulation of the low molecular
          weight protein,  alpha-2u-globulin.  Because only male rats product £±SSn
          *e tum0rs would have little or no predictive validity  with respect to 1C health
          hazard on a site- or mechanism-specific basis.  Perchloroethylene indu™dnev
          tumors at lower doses than those required to cause alpto-^Sto          *
          acamnilation, however.  The EPA is developing criteria thatwill define a weight-
         of-ev dence approach for evaluating, on a case-by-case basis, the role of alia 2u-
         globulrn in rat kidney tumor formation (U.S. EPA, 1991).

      2.  The chronic administration of perchloroethylene  produces nephrotoxicity and it
         has been suggested that tumor production is  secondary to susfcdned cyS«Sy
         and cellular regeneration.  Although certain  "nephrotoxicity" occurs HoSLs
         of rats and mice, implying that kidney tumors  would occur in rats and mfce of

         SSSTiS    CarT°gtCity bi°aSSayS' Cdl rePlication occurs * nTbut not
         in female rats  suggesting that any nephrotbxic mechanism would likely be
         mSech^mW1f alp?r2ut°bUlin accu^tion.  On the other hand  me
         mechanism of perchloroethylene tumorigenesis may be similar to  that of its
         structural analog, trichloroethylene.  Trichloroethylene induces kidney
                                         in
44

-------
      3
                                                                             >      .
                                                                                    m

          pathway is irrelevant with respect to human risk projection.

       Although there is some evidence to support each of the proposed mechanisms, there
are also significant quantitative and qualitative gaps in the supportive data.  The mode of
perchloroethylene-induced renal tumorigenesis in male rats is not yet understood.
                                              45

-------
                    7. MONONUCLEAR CELL LEUKEMIA IN RATS

         The NTP (1986a) reported that the chronic inhalational administration of
  perchloroethylene to male and female F344/N rats caused positive trends in the incidence of
  mononuclear cell leukemia (MCL) in both sexes.  Pairwise comparisons of tumor incidences
  m dosed and control groups of males (life table  analysis) revealed statistically significant
  increases  in both the low- and high-dose groups (controls, 28/50;  low dose, 37/50  p=0 046-
  high dose, 37/50, p=0.004; trend test p=0.004).  Analysis of the data for female'rats
  revealed a marginally significant trend (p=0.053) and a significant increase in the low-dose
 group and a marginally significant increase in the high-dose group (control, 18/50; low dose
 30/50, p=0.023; high dose,  29/50, p=0.053).
        Interpretation of these data is somewhat clouded by the fact that overall incidences of
 MCL m the concurrent chamber control groups were high relative to historical chamber
 control groups at the performing laboratory (males 28/50, 56 percent versus 117/250  47
 percent; females 18/50, 36 percent versus  73/249, 29 percent). The concurrent control
 group rates were also higher  than the  NTP historical rate for untreated  control groups  (males
 583/1,977, 29 percent and females 375/2,021,  18 percent).
       Because of these factors the NTP conducted supplemental analyses of the progression
 of the disease, the effect of perchloroethylene on  the time of onset of advanced MCL  and
the contribution of MCL to early deaths in control and dosed animals.  The results of these
supplemental analyses showed that:
' PercMor0eth^e Produ-d a dose-related
                                                                          increase in
         S°ratf ^^ eXP°SUre Significantly shorten^ the time to onset of MCL in

         Although there was no remarkable effect of perchloroethylene exposure on
         survival of female rats, there was an increased incidence of Tdvanced MCL in
         female rats that died before the scheduled termination of the smdy   Thus a more
         appropriate statistical analysis was conducted in which only the ScidenTes' of
         advanced MCL m rats were considered.  Significantly posftive trends and

                                        46

-------
           significant increases in the incidences of advanced MCL in both male and female
           rats in the high-dose groups were observed.
       In 1987, the EPA Science Advisory Board took exception to the use of these special
 analyses because they did not represent generally accepted approaches to evaluating increased
                                                                       i
 incidences of MCL.  According to the NTP report, however, the interpretation of MCL
 incidences in the perchloroethylene study was based on standard methods of data evaluation
 (NTP, 1986a).  The special analyses were  conducted to support, rather than establish, the
 interpretation.
       Under the conditions of the NTP study a carcinogenic effect of percfaloroethylene in
 male and female rats was evidenced by significant increases in the incidences of MCL in
 both sexes.  However, the usefulness of increased incidences of MCL in predicting human
 carcinogenic risk associated with exposure  to perchloroethylene has been questioned on
 several grounds:

 MCL Is  a Common and Variable Tumor That Occurs Spontaneously in F344/N Rats.
 Marginal Increases in Incidences Are of Questionable Biological Significance.
       MCL is recognized as  a common neoplasm in rats, and its rate of appearance in
 historical control groups is highly variable.  In the five contemporary inhalation studies
 conducted at the performing laboratory, the incidences  of MCL in chamber control male rats
 ranged from 32  to 68 percent.  In female chamber control rats,  the incidences ranged from
 22 to 36 percent.  A similarly high variability has been observed among the NTP untreated
 control groups (males, 10 to 60 percent; females, 6 to  38 percent).
       Concurrent controls represent the most appropriate groups to use for determining the
 statistical significance of observed differences between  experimental groups. It is
 recognized, however, that in the case  of spontaneous and highly variable tumors,
comparisons of treatment groups to historical control groups may be helpful in interpreting
experimental results (Haseman et al.,  1984a).  When the  overall rates of MCL hi male rats
 hi the perchloroethylene studies are compared with the  range of tumor incidences hi
historical controls, the perchloroethylene-treated  animals were essentially identical to those in

                                          47

-------
the historical control group with the highest incidence (74 percent in perchloroethylene-dosed
animals versus 68 percent in the historical control group). However, the incidences of MCL
in perchloroethylene-dosed female rats (60 percent and 58 percent) were elevated relative to
the highest incidence in historical control groups (38 percent).

The Pathobiology of MCL Is Too Poorly Understood to Allow the Tumors to Be Used to
Determine Human  Health Risk.
       MCL is a relatively well-defined and understood rodent neoplasm characterized by
infiltration of pleomorphic blastlike mononuclear cells in numerous organs. The disease per
se, which is splenic hi origin but later infiltrates the liver, lung, bone marrow, lymph nodes,
and other organs, is readily and unequivocally diagnosed by standard histopathological
techniques.  MCL has also been described as large, granular, lymphocytic leukemia and is
known to be a rapidly progressing and fatal neoplasm whose incidence is age related.  The
tumor is transplantable; its etiological factor is unknown.  It has been suggested that a
cellular oncogene may be responsible for the induction of MCL.
       Although the specific mechanism of leukemogenesis in rats is not understood, it is
interesting to  note early reports of toxicity of cysteine S-conjugates where S-(l,2,-
dichlorovinyl)-L-cysteine was implicated in induction of aplastic anemia and marked
biochemical alteration of DNA hi bone marrow, lymph nodes, and thymus in calves
(McKinney et al., 1957; Schultze et al., 1959; Bhattacharya and Schultze, 1971, 1972).  As
discussed earlier, the glutathione conjugate of perchloroethylene is hydrolyzed in the kidney
to the cysteine S-conjugate, a compound that can be cleaved to form a mutagenic fragment.
Humans as well  as rodents activate the conjugate via the beta-lyase pathway. Thus, the
possibility exists that the perchloroethylene S-conjugate, S-(l,2,2-trichlorovinyl)-L-cysteine,
may be involved hi inducing leukemia hi rats and may have the potential to produce blood
dyscrasias hi humans as well.

MCL Is a Rodent-Specific Tumor With No Human Correlate.
       MCL is  a neoplasm whose incidence  and progression can  be influenced by chemical
agents.  While human leukemias originate hi bone marrow, MCL is splenic in origin.
                                           48

-------
Discounting a rodent neoplasm simply because it has no exact human counterpart is not a
scientifically defensible reason, however. Site concordance is not a requirement for
relevancy in extrapolation of hazard potential, although its actual presence can strengthen
belief in a particular hazard; e.g., many aromatic amines are probable bladder carcinogens in
humans but are likely to produce Zymbal gland tumors rather than bladder tumors in rats.
The human does not develop Zymbal gland tumors.
                                             49

-------
                           8. SUMMARY AND CONCLUSIONS

         The Office of Health and Environmental Assessment has reviewed the data currently
  available regarding perchloroethylene carcinogenicity.  Because the human epidemiology data
  are inadequate for determining the carcinogenicity of perchloroethylene, the focus of this
  paper is on the animal data and other information as they are relevant to potential cancer-
  causing activity in humans.  Two questions now must be considered:  first, whether the data
  are judged as "sufficient" evidence of carcinogenicity in animals,  and second, whether the
  general assumption that "sufficient" animal data lead to an overall weight-of-evidence
  classification of B2, probable human carcinogen, holds up in the case of perchloroethylene.

 Is There Sufficient Evidence of Carcinogenicity in Animals?
       EPA's Guidelines for Carcinogen Risk Assessment provide criteria to follow in
 weighing the scientific evidence.  The Guidelines also permit the exercise of professional
 judgment throughout the process, with explanations for the judgment calls.  The positive
 animal evidence for perchloroethylene2 clearly meets the criteria for "sufficient" animal data
 as depicted in the Guidelines. Sufficient evidence of carcinogenicity indicates that there is an
 increased incidence of malignant tumors or combined malignant and benign tumors (1) in
multiple species or strains; (2) in multiple experiments (e.g., with different routes of
administration or using different dose levels); or (3) to an unusual degree in  a single
experiment with regard to high incidence, unusual  site or type of tumor, or early age at
onset.
 Perchloroethylene has been shown to cause multiple tumor end points-hepatocellular
 carcinomas m both sexes of mice, kidney tumors and some indication of gliomas in male
  H th     v                   and female rats'  The incidences of liver tumors in mice
and the leukemias in rats are statistically significantly elevated when compared with
controls.  The incidence of renal tumors and gliomas in male rats is significantly elevated
when compared with historical controls, and these tumors are biologically sigScant
                                            has been demo^ted byU inhalation and
                                          50

-------
       The perchloroethylene data meet, at least to some degree  each of the three

======
         OuideL criteria for  'sufficient- animal evidence because there are mufcp.e
            having different exposure routes and no appreciable d°™^°f ^
  would       *     ative of potent, human hazard and wou.d .ead ;
  Ividence dassificaUon of B2, prooaWe human carcinogen, When cons.dered w* the
    increased incidences of bom hepato^llular adenomas a          affirmative votes, one
    hepatocellular carcinomas in females. The p anel wr o    ^ s       carcinogenicity of
    negative vote) conclusions of some evince inl ^^^.^ MU leukemia.  The
    pe?chloroethylene, as shown by mcr^™*n^o™; tot there was clear evidence of
    panel concluded (five affirmative votes, four ne|M ve vo  ^          ^ ^^^
    carcinogenicity of P-te"*^^.^! mbular cell neoplasms. (Two
              o
                                             voting due to employment conflict of interest,
*There is agreement between the It-temationa, Agency for
    EPA that flie animal          '"
                                                         more similar to EPA's sufficient
                                                           "Possible" human carcinogen,
                                              51

-------
   "inadequate" epidemiologic data (U S EPA  tos*  D




      be answered is whether fte genera,                              ~ ^ ">*
                              reasonable
 Implications of Perchloroethylene Metabolism
       I. is generally considered tot the toxicity, mutagenicjtyi and
              e reside in reac,ive mua^ The ^ do „        d

              ^ as; mutagen' although — '-*-*• —
        c (e.g., TCVC, ohloroacetaldehydes including cUoral hydrate)  Severa,
Perchloroethylene metabolites have been shown to be cytotoxic .  „
TCA  DPA anH , •  t,                          cytotoxic, and certain metabolites (e g
1CA, DCA, and trichloroacetaldehyde) cause liver tumors in mice
                                    52

-------
                        — '


e oon
                          53

-------
Mouse Liver Tumors
case basis according ,o   uc
   tun™ data is
   (U.S. EPA,
                              liver
                     boiwsed for
                 —•
                                     are
                                         °" a
                                              .ouse
                                          publislled in

                     54

-------
         S  although evidence exists that supports the hypothesis of mouse iiver cancer
             second to peroxisome proliferation, the vaUdity ot the nypotnests remams

        *.  . , - .—— - — « :^ rrr a
peroxisome proliferation and also causes hepatocellular carcmoma » B6OF1 nuce bu
clL-eL relationship between the two effects has not been shown.  It „ not dear
 T,  ,  if any peroxisome proliferation actually plays in perchloroethylcne


 role the peroxisome proliferation phenomenon may have in tumongenes,
       1, severa, other proposed mechanisms need to be further mvest.ga.d. For
         the recent demons^atiou that the major metabolite of perchioroethylene  TCA,
  consideration as well.
                  was identified in a male mouse.  !n .e case of percMoroe.y.ene, U .

                                         55

-------
          rrricrr.r?       •
-A «.. tat cenam nonto.or pathology associated with the chronic administn,^ of


       n
      genera, .
o
of humans ,„ conjugate perchtoroethyiene. Ms i
      56
                                    is because

-------
such few liver samples were studied, and a conjugation rate tenfold lower than that observed
for rats would fall below the limits of detection of the method employed.  Tenfold
differences in enzyme activities within the human population are not uncommon.
       Although some evidence supports each of the proposed mechanisms, there also are
significant quantitative and qualitative gaps hi the supportive data. The mode of
perchloroethylene tumorigenesis hi male rats is not yet understood.

Leukemia in Rats
       Under the conditions of the NTP study, a carcinogenic effect of perchloroethylene hi
male and female rats was evidenced by significant increases hi the incidences of MCL in
both sexes.  However, the usefulness of increased incidences of MCL hi predicting human
carcinogenic risk associated with exposure to perchloroethylene has been questioned on
several grounds:  high spontaneous background incidences, use of special (supplemental
analyses to aid hi data interpretation, and the relevance of MCL in F344/N rats because this
type of leukemia does not occur in humans.
       The leukemia incidences were statistically significantly increased in both male and
female rats.   In both sexes, perchloroethylene caused a dose-related increase hi severity  of
MCL and shortened the tune to tumor in female rats.  There was an increased incidence of
advanced MCL in female rats that died before the scheduled termination of the study.
Supplemental analyses were based on standard methods of data evaluation and supported the
data interpretation.
       While human leukemias originate in bone marrow, MCL is splenic in origin. Despite
the fact that the disease occurs only in rats, it is a neoplasm whose incidence and progression
can be influenced by chemical  agents.  Discounting a rodent neoplasm simply because it has
no exact human counterpart is  not reasonable.  Many aromatic amines are probable bladder
carcinogens  in humans but are likely to produce Zymbal gland tumors rather than bladder
tumors hi rats.  The human does not have  a Zymbal gland.
                                          57

-------
 Conclusion
       There is no one specific mechanism that explains all of the tumor end points.  Of
 several modes of action that have been proposed for perchloroethylene tumorigenesis,
 different mechanisms have been hypothesized to cause the three distinct tumor types observed
 in rats and mice.  Data support certain of these mechanisms.  Scientific evidence also
 supports species specificity for particular mechanisms, but the incompleteness of the data
 limits the scope of the conclusions that can be deduced although the mechanisms are certainly
 plausible.  The modes of action  for perchloroethylene carcinogenesis are not yet well
 understood. With regard to the  kidney tumors, evidence exists for more than a single
 mechanism, all of which may play some role hi the tumor development independently or in
 concert.  A cause-and-effect relationship cannot be effectively demonstrated between the
 peroxisome proliferation and development of liver tumors in mice.   Several other
 mechanisms may contribute to hepatocarcinogenicity in mice. The possibility that there  may
 well be a rnutagenic component  in the development of the tumors, especially the kidney
 tumors, cannot be entirely ruled out.  The evidence that metabolic pathways in rodents may
 not occur in humans is not convincing.
       Because mechanisms of perchloroethylene  carcinogenesis are not understood well
 enough, each individual tumor type is viewed as contributing not just to the "sufficient"
 evidence in animals, but to the overall weight-of-evidence determination that
 perchloroethylene is  a probable human carcinogen.  In each case, reasonable doubt exists that
 the mode of tumorigenesis is only through mechanisms species specific to rodent strains.
 Other mechanisms are feasible that would not be specific to rodents.  All three tumor types
 can therefore be considered valid as indicators relevant to potential carcinogenicity in
 humans,  although some uncertainty exists concerning relevance to humans.
       The EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986b) suggest
that the weight of evidence increases with the increase in number of animal species,  strains,
sexes,  and number of experiments and doses showing a carcinogenic response, with the
 increase in number of tissue sites affected, with the occurrence of dose-response relationships
as well as statistical significance of the increased tumor incidence in treated compared with
control groups, when there is decreased tune to tumor occurrence or death with tumor, and
                                          58

-------
when there is a dose-related increase of malignant tumors.  All of these criteria are met in
some way.  Perchloroethylene causes at least three types of tumors in rodents, each of which
can be considered as contributing in some way to the concern for cancer-causing potential in
humans.  Indications of cancer-causing activity were seen in two species, in two sexes, by
inhalation and oral exposure, and are called "sufficient" animal evidence.  Although there is
some scientific uncertainty concerning relevance  to humans for some of the data, the totality
of the animal data for perchloroethylene is not only closer to the "sufficient" evidence
category but also can be considered relevant for  extrapolation of hazard potential to humans.
       Therefore, although the relevance of some of the data is less than certain, the
inclusive  animal data for perchloroethylene taken as a whole, along with  the considerations
of inadequate human data, information on metabolism, and mutagenicity  data on metabolites
can most logically be categorized as a Group B2 probable human carcinogen, although the
EPA's SAB does not fully agree.  It must be remembered that classifications refer only to the
weight of the experimental evidence that a chemical is carcinogenic and not to its potency of
carcinogenic action.  This paper has not addressed directly the quantitative estimation of risk.
Mechanistic considerations may justify special interpretation of the dose-response data with
respect to projecting human carcinogenesis risk.                          j
        The Agency's quantitative estimates of carcinogenic risk from exposure to
perchloroethylene have not been recently updated. Existing estimates are relatively simplistic
and are calculated using a nonthreshold  model that is  linear at low doses. The risk estimates
made using such a model are regarded as conservative and represent a plausible upper limit
for the risk such that the true risk is not likely to be higher than the estimate, but it is likely
to be lower and could even be zero.  Relative to other potential carcinogens evaluated by the
Agency,  perchloroethylene carcinogenic potency index ranks in the lowest quartile, so
perchloroethylene is not viewed as  a very potent carcinogen.
        Because it is generally accepted that the carcinogenic potential of perchloroethylene
 resides in its metabolites, the amount of the compound metabolized was  considered as being
 directly proportional to the dose to the target tissue and was thus factored into the calculation
 of the risk estimates as the dose rather than using the administered dose. Since amount of
 perchloroethylene metabolized was used as the dose and since metabolism is nonlinear, it is
                                             59

-------
 important to realize that the metabolism curve should be factored in when back calculating
 from the unit risk estimate to the risk to humans at high exposures.
        The upper-bound estimate of the incremental unit risk for inhalation of 1 /ig/m3 of
 perchloroethylene hi air is 4.8 X 10'7, and the upper-bound estimate of the incremental
 cancer risk due to 1 /zg/L of perchloroethylene in drinking water is 1.5 x 10"6 (U.S. EPA,
 1985a), based on hepatocellular carcinomas observed in oral studies in mice.  Calculations
 based on inhalation studies hi which hepatocellular carcinomas were detected in the same
 mouse strain resulted hi similar risk estimates (U.S. EPA, 1986a).  It is of interest to note
 that the air unit risk estimates calculated on the basis of leukemias in rats exposed to
 perchloroethylene via inhalation are also comparable to the estimates derived from the mouse
 studies. Thus, risk estimates calculated on the basis of six different inhalation data sets are
 comparable, ranging from 2.9  X 10 "7 to 9.5 X  10'7.  In addition, all of these values are
 comparable to the unit risk estimate of 4.8 X 10'7, derived from the oral gavage study  in
 mice,  a risk estimate that falls  approximately hi the middle of the range of estimates.
        An ED10 value (effective dose in 10 percent) has been calculated for
 perchloroethylene based on the the  geometric mean of tumor incidences from the six
 inhalation data sets.  The ED10 is the dose level expected to cause cancer in 10 percent of
 the population exposed.  The ED10 for oral exposure is 83 mg/kg/day. The ED10 for
 inhalation exposure is 2.9E-5 jug/m3.  The ED10  can be divided by the environmental level,
 or the exposure  level of concern, to give a margin of exposure.
       The  Agency's Science Advisory Board reviewed a draft of this report and deemed it
to be well written and of high scientific quality.  In an August 1991 letter to the EPA
Administrator, however, the Board  maintained as still being appropriate its previous advice
regarding the weight-of-evidence classification for perchloroethylene—that it lies on a
continuum between categories B2 and C.  This differs somewhat from the EPA position of a
B2 classification for perchloroethylene. The Board also adhered to its earlier opinion that,
from a scientific perspective, exposure should be  considered more important than
classification category hi determining potential threat to  human health and whether or not a
chemical substance should be regulated.
                                           60

-------
                                                                                                  1
                                     REFERENCES
 Alden, C.L. (1989) Male rat specific alpha 2u globulin nephropathy and renal tumorigenesis.
       In: Bach, P.H.; Lock, E.A., eds. Nephrotoxicity: in vitro to in vivo, animals to man.
       New York:  Plenum Press; pp. 535-541.

 Alden, C.L.; Kanerve, R.L.; Ridder, G.; Stone, L.C. (1984) The pathogenesis of the
       nephrotoxicity of volatile hydrocarbons in the male rat. In: Mehlman, M.A.;
       Hemstreet, G.P.; Thorpe, J.J.; Weaver, N.K., eds. Advances in modern
       environmental toxicology, Vol. VII. Renal effects of petroleum hydrocarbons.
       Princeton, NJ:  Princeton Scientific Publishers, Inc.; pp. 107-120.

 Alden, C.L.; Ridder, G.; Stone, L.; Kanerva, R.L. (1985) Pathology of petrochemical fuels
       in male rats: acute toxicity. In: Bach, P.H.; Lock, E.A., eds. Renal heterogeneity and
       target cell toxicity. New York, NY: John Wiley and Sons;  pp. 461-472.

 Anders, M.W.; Lash,  L.; Dekant, W. (1988) Biosynthesis and biotransformation of
       glutathione S-conjugates to toxic  metabolites. CRC Crit. Rev. Toxicol. 18: 311-341.

 Bartsch, H.; Malaveille,  C.; Barbin, A.; Planche, G. (1979) Mutagemc and alkylating
       metabolites of halo-ethylenes, chlorobutadienes and dichlorobutenes produced by
       rodent or human liver tissues. Arch. Toxicol.  41: 249-277.

 Beal, D. (1990) Use of mouse liver tumor data in risk assessment performed by the U.S.
       Environmental Protection Agency. In:  Stevenson, D.E.; McClain, R.M.;  Popp, J.A.;
       Slaga, T.J.;  Ward, J.M.; Pilot, H.C.,  eds. Mouse liver carciriogenesis: mechanisms
       and species comparisons. Progress hi clinical and biological research,  Vol. 331  New
       York, NY: Wiley-Liss; pp. 5-18.

Beliles, R.P.; Brusick, D.J.; Mecler, F.J. (1980) Teratogenic-mutagenic risk of workplace
       contaminants: trichloroethylene, perchloroethylene, and carbon disulfide. Prepared for
       National Institute for Occupational Safety and  Health, Cincinnati, OH, under Contract
       No. 210-77-0047. Kensington, MD: Litton Bionetics, Inc. Available from: National
       Technical Information Service, Springfield, VA; PB82-185075.

Bhattacharya, R.K.; Schultze, M.O. (1971) Properties of DNA isolated from tissues of
       calves treated with S-(l,2-dicholorvinyl)-L-cysteine. 1. Chemical and physical
       properties. Arch. Biochem. Biophys. 145(2): 565-574.

Bhattacharya, R.K.; Schultze, M.O. (1972) Properties of DNA treated  with S-(l,2-
       dichlorovinyl)-L-cysteine and a lyase. Arch. Biochem. Biophys. 153(1): 105-115.
                                          61

-------
Bhunya, S.P.; Behera, B.C. (1987) Relative genotoxicity of trichloroacetic acid (TCA) as
       revealed by different cytogenetic assays: bone marrow chromosome aberration
       micronucleus and sperm-head abnormality hi the mouse. Mutat. Res.  188: 215-221.

Bieri, P.; Stubli, W.; Waechter, P.; Muakkassah-Kelly, S.; Bentley, P. (1988) Stimulation
       of DNA synthesis but not peroxisome B-oxidation by nafenopin in primary cultures of
       marmoset hepatocytes. Cell Biol. Int. Rep. 12:  1077-1087.

Bignami,  M.; Conti, G.; Conti, L.;  Crebelli, R.; Misuraca, P.; Puglia, A.M.; Randazzo,
       R.; Sciandrello, G.; Carere,  A. (1980) Mutagenicity of halogenated alphatic
       hydrocarbons hi Salmonella typhimurium, Streptomyces coelicolor and Aspergillus
       nidulans.  Chem. Biol. Interact. 30: 9-23.

Blair, A.; Tolbert, P.E.; Grauman,  D.; Moran, F.X.; Vaught, J.;  Rayner, J. (1990) Cancer
       and other causes of death among a cohort of dry cleaners. Br. J. Ind. Med. 47:  162-
       168.

Borghoff, S.J.; Short, E.G.; Swenberg, J.A. (1990) Biochemical mechanisms and
       pathobiology of alpha 2u-globulin nephropathy.  Annu. Rev.  Pharmacol. Toxicol. 30:
       349-367.

Bronzetti, G.; Bauer, C.; Corsi, C.; Del Carratore, R.; Galli, A.; Nieri, R.; Paolini, M.
       (1983) Genetic and biochemical studies on perchloroethylene hi vitro and hi vivo.
       Mutat. Res. 116: 323-331.

Bronzetti, G.; Galli,  A.; Corsi, C.; Cundari, E.; Del Carratore, R.; Nieri, R.; Paolini, M.
       (1984) Genetic and biochemical investigation on chloral hydrate in vitro and hi vivo.
       Mutat. Res. 141: 19-22.

 Brown, D.P.; Kaplan, S.D. (1987) Retrospective cohort mortality study of dry cleaner
       workers using perchloroethylene. J. Occup.  Med. 29:  535-541.

 Butterworth, B.E.; Smith-Oliver, T.; Earle, L.; Loury, D.J.; White, R.D. (1989) Use of
       primary cultures of human hepatocytes hi toxicology studies. Cancer Res. 49: 1075-
        1084.

 Callen, D.F.; Wolf, C.R.; Philpot, R.M. (1980) Cytochrome P-450 mediated genetic activity
       'and cytotoxicity of seven halogenated aliphatic hydrocarbons hi Saccharomyces
        cerevisiae. Mutat. Res. 77:  55-63.

 Cariot, P.; Barrat, E.; Mejean, L.; Pointel, J.P.; Drouin, P.; Debbry, G. (1983)  Fenofibrate
        and human liver. Lack of proliferation of peroxisomes. Arch.  Toxicol. 53: 151-163.
                                            62

-------
 Cerna, N.; Kypenova, H. (1977) Mutagenic activity of chloroethylenes analyzed by
        screening system test. Mutat. Res. 46: 36 (Abstract).

 Chang, L.W.; DeAngelo, A.B.; Daniel, F.B. (1989) Evaluation of DNA strand breaks (SB)
        by chloroacetic acids (CAs) and chloroacetaldehydes (CADs) in vivo and in vitro
        Carcinogenesis 30  (Abstract 578).

 Charbonneau, M.; Short, B.G.; Lock, E.A.; Swenberg, J.A. (1987) Mechanism of
       petroleum-induced  sex-specific protein droplet nephropathy and renal cell proliferation
       in Fischer-344 rats: Relevance to humans. Hemphill, D.D., ed. Trace substances in
       environmental health, Vol. 21. Columbia, MO: University of Missouri; pp. 263-273.

 Chen, J.C.; Stevens, J.L.; Trifillis, A.L.; Jones, T.W. (1990) Renal cysteine conjugate beta-
       lyase-mediated toxicity studied with primary cultures of human proximal tubular cells
       Toxicol. Appl. Pharmacol. 103: 463-473.                   :

 Costa, A.K.;  Ivanetich, K.M. (1984) Chlorinated ethylenes: their metabolism and effect on
       DNA repair in rat hepatocytes. Carcinogenesis 5(12): 1629-1636.

 DeAngelo, A.B.; Daniel, F.B.; McMillan, L.; Wernsing,  P.; Savage, R.E., Jr. (1989)
       Species and strain sensitivity to the induction of peroxisome proliferation by
       chloroacetic acids. Toxicol. Appl. Pharmacol.  101: 285-298.
                                                                 i
 Degrassi, P.;  Tanzarella, C. (1988) Immunofluorescent staining of kinetochores in
       micronuclei: A new assay for the detection of aneuploidy. Mutat. Res. 203: 339-345.

 Dekant, W.; Vamvakas, S.; Berthold, K.; Schmidt, S.; Wild, D.; Henschler, D. (1986)
       Bacterial 0-lyase mediated cleavage and mutagenicity of cysteine conjugates derived
       from the nephrocarcinogenic alkenes trichloroethylene, tetrachloroethylene, and
       hexachlorobutadiene. Chem. Biol. Interact. 60: 31-45.

 Dekant, W.; Martens, G.; Vamvakas, S. (1987) Role of glutathione S-transferase-catalyzed
       conjugation versus cytochrome P-450 dependent phospholipid alkylation Drug Metab
       Dispos. 15(5): 702-709.

Dekant, W.; Vamvakas, S.; Anders,  M.W. (1989) Bioactivation of nephrotoxic haloalkenes
       by glutathione conjugation:  formation of toxic and mutagenic intermediates by
       cysteine conjugate /3-lyase. Drug Metab. Rev. 20(1): 43-84.    >

De La Inglesia, F.A.; Lewis, J.E.; Buchanan, R.A.; Marcus, E.L.; McMahon, G. (1982)
       Light and electron microscopy of liver in hyperlipoproteinemic patients under long-
      term gemfibrozil treatment.  Atherosclerosis 43: 19-37.
                                          63

-------
                                                                                               I
Elcombe, C.R. (1985) Species differences in carcinogenicity and peroxisome proliferation
       due to trichloroethylene: a biochemical human hazard assessment. Arch. Toxicol.
       8: 6-7.

Elcombe, C.R.; Mitchell, A.M. (1986) Peroxisome proliferation due to di(2-
       ethylhexyl)phthalate (DEHP): species differences and possible mechanisms. Environ.
       Health Perspect. 70:  211-219.

European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC). (1990)
       Tetrachloroethylene:  assessment of human carcinogenic hazard. ECETOC Technical
       Report No. 37. Brussels.

Flamm, W.G.; Lehman-McKeeman, L.D. (1991) The human relevance of the renal tumor-
       inducing potential of D-limonene in male rats: implications for risk assessment.
       Regul. Toxicol. Pharmacol. 13(1): 70-86.

Fox, T.R.; Schumann, A.M.; Watanabe, P.O.; Yano, B.L.; Maher, V.M.;  McCormick, J.J.
       (1990) Mutational analysis of the H-ras oncogene in spontaneous C57BL/6 x
       C3H/He mouse liver tumors and tumors induced with genotoxic and nongenotoxic
       hepatocarcinogens. Cancer Res. 50: 4014-4019.

 Galloway, S.M.; Armstrong, J.J.; Reuben, C.; Colman, S.; Brown, B.; Cannon, C.; Bloom,
       A.D.; Nakamura, F.; Ahmed, M.; Duk, S.; Rimpo, J.; Margolin, B.H.; Resnick,
       M A • Anderson, B.; Zeiger, E. (1987) Chromosome aberrations and sister-chromatid
       exchanges in Chinese hamster ovary cells: evaluations of 108 chemicals.  Environ.
       Molec. Mutagen. 10(10): 1-109.

 Goldsworthy, T.L.; Popp, J.A. (1987) Chlorinated hydrocarbon-induced peroxisomal enzyme
       activity in relation to the species and organ carcinogenicity. Toxicol. Appl.
       Pharmacol. 88(2): 225-233.

 Goldsworthy, T.L.; Smith-Oliver, T.; Loury, D.J.; Popp, J.A. (1988) Assessment  of
       chlorinated hydrocarbon-induced genotoxicity and cell replication in rat kidney cells.
       Environ. Molec. Mutagen. 11(11): 39.

 Goodman, D.G.; Ward, J.M.; Squire, R.A.; Chu, K.C.; Linhard, M.S.  (1979) Neoplastic
       and non-neoplastic lesions in aging F344 rats. Toxicol.  Appl. Pharmacol. 48: 237-
       248.

 Goodman D.G.; Ward, J.M.; Squire, R.A.; Paxton, M.B.; Reichardt, W.D.; Chu, K.C.;
        Linhart, M.S. (1980) Neoplastic and non-neoplastic lesions in aging Osborne-Mendel
        rats. Toxicol. Appl. Pharmacol. 55: 433-447.
                                           64

-------
Goodman et al. (1990) The pharmacological basis of therapeutics. 8th Ed. Oilman, A.G.;
       Rail, T.W.; Nies, A.S.; Taylor, P., eds. Pergamon Press; 1811 pp.

Green, T.;  Odum, J. (1985) Structure-activity studies of the nephrotoxic and mutagenic
       action of cysteine conjugates of chloro- and fluoroalkenes. Chem.  Biol. Interact. 54:
       15-31.

Green, T.;  Odum, J.; Foster, J.K.  (1990) Perchloroethylene-induced rat kidney tumors: an
       investigation of the mechanisms involved and their relevance to humans. Toxicol.
       Appl. Pharmacol. 103:  77-89.

Greim, H.; Bonse, G.; Radwan, Z.; Reichert, D.; Henschler, D. (1975) Mutagenicity in
       vitro and potential carcinogenicity of chlorinated ethylenes as a function of metabolic
       oxirane formation. Biochem. Pharmacol. 24(21): 2013-2017.

Gualandi, G. (1987)  Use of alpha-  and beta-tubulin mutants for the study of spontaneous and
       induced chromosomal mis-distribution in Aspergillus nidulans. Mutat. Res. 178: 33-
       41.

Haider, C.A.; Warne, T.M.; Hatoum, N.S. (1984) Renal toxicity of gasoline and related
       petroleum  napthas in male rats. In:  Mehlman, M.A.; Hemstreet, G.P.; Thorpe, J.J.;
       Weaver, N.K., eds.  Advances in modern environmental toxicology, Vol. VII. Renal
       effects of petroleum hydrocarbons.  Princeton, NJ: Princeton Scientific Publishers,
       Inc.; pp. 73-88.

Hanefeld, M.; Kemmer, C.; Kadner, E.  (1983) Relationship between morphological changes
       and  lipid-lowering action of p-chlorphenoxyisobutyric acid (CPIB) on hepatic
       mitochondria  and peroxisomes in man. Atherosclerosis 46: 239-246.

Haseman, J.K.; Huff, J.; Boorman, G.A. (1984a) Use of historical control data in
       carcinogenicity studies in rodents. Toxicol.  Pathol. 12:  126-135.

Haseman, J.K.; Crawford,  D.D.; Huff, I.E.; Boorman, G.A.; McConnell, E.E.  (1984b)
       Results from  86 two-year carcinogenicity studies conducted by the NTP. J. Toxicol.
       Environ. Health 14: 621-639.

Haworth, S.; Lawlor, T.; Mortelmans, K.; Speck, W.; Zeiger, E. (1983) Salmonella
       mutagenicity test results for 250 chemicals. Environ. Mutagen. 5(1): 1-142.

Henschler,  D. (1977a) Metabolism of chlorinated alkenes and alkanes as related to toxicity.
       J. Environ. Path. Toxicol. 1(2):  125-133.

Henschler,  D. (1977b) Metabolism and mutagenicity of halogenated olefins: a comparison of
       structure and  activity. Environ. Health Perspect. 21: 61-64.   ,


                                          65                    :

-------
Henschler, D.; Elsasser, H.; Romen, W.; Eder, E. (1984) Carcinogenicity study of
       trichloroethylene, with and without epoxide stabilizers, in mice. Cancer Res. Clin.
       Oncol. 107:  149-156.

Herren-Freund, S.L.; Pereira, M.A.; Knoury, M.D.; Olson, G. (1987) The carcinogenicity
       of trichloroethylene and its metabolites, trichloroacetic acid and dichloroacetic acid, in
       mouse liver. Toxicol. Appl. Pharmacol. 90: 183-189.

International  Agency for Research on Cancer (IARC). (1982) Evaluation of the carcinogenic
       risk to humans. IARC Monogr.  (Suppl) 4.

International  Agency for Research on Cancer (IARC). (1987) Overall evaluations of
       carcinogenicity: an updating of IARC monographs Volumes 1 to 42. IARC Monogr.
       Eval.  Care. Risk Chem. Hnr. 7 (Suppl): 1-440.

Ikeda, M.; Ohtsuji, H.; Imamura, T.; Komoike, Y. (1972) Urinary excretion of total
       trichloro-compounds, trichloroethanol and trichloroacetic acid as a measure of
       exposure to trichloroethylene and tetrachloroethylene. Br. J. Ind. Med. 29: 328-333.

Ikeda, M.; Miyae, Y.; Ogata, M.; Ohmore, S. (1980) Metabolism of trichloroethylene.
       Biochem. Pharmacol. 29(21): 2983-2992.

Issemann, I.; Green, S. (1990) Activation of a member of the steroid hormone receptor
       superfamily by peroxisome proliferators. Nature 347: 645.

Kafer, E. (1985) Tests which distinguish induced aneuploidy and crossing-over from
       secondary segregation hi Aspergillus treated with chloral hydrate or x-rays. Mutat.
       Res. 167: 9-34.

Kline, S.A.;  McCoy, E.G.; Rosenkranz, H.S.; Van Durren, B.L. (1982) Mutagenicity of
       chloroalkene epoxides hi bacterial systems. Mutat. Res. 101:  115-125.

Koch, R.; Schlegelmilch,  R.; Wolf, H.U. (1988) Genetic effects of chlorinated ethylenes hi
       the yeast Saccharomyces cerevisiae. Mutat. Res. 206(2): 209-216.

Lash, L.H.;  Nelson, R.M.; Van Dyke, R.A.; Anders, M.W. (1990) Purification and
       characterization of human kidney cytosolic cysteine conjugate /3-lyase  activity.  Drug
       Metab. Dispos. 18: 50-54.

Liang, J.C.;  Pacchierotti, F. (1988)  Cytogenetic investigation of chemically-induced
       aneuploidy in mouse spermatocytes. Mutat. Res. 201: 325-335.

Lock, E.A. (1988) Studies on the mechanism of nephrotoxicity and nephrocarcinogenicity  of
       halogenated alkenes.  CRC Crit. Rev. Toxicol. 19: 23-42.

                                           66

-------
MacNaughton, M.G.; Uddin, D.E. (1984) Toxicology of mixed distillate and high-energy
       synthetic fuels. In: Mehlman, M.A.; Hemstreet, G.P.; Thorpe, J.J.; Weaver, N.K.,
       eds. Advances in modern environmental toxicology. Vol. 7, Renal effects of
       petroleum hydrocarbons. Princeton, NJ: Princeton Scientific Publishers, Inc.;
       pp. 121-132.

Maltoni, C.; Cotti, G. (1986) Results of long-term carcinogenicity bioassays of
       tetrachloroethylene on Sprague-Dawley rats administered by ingestion. Acta
       Oncologica 1:11-26. As cited in Green, T. (1990) Chloroethylenes: a mechanistic
       approach to human risk evaluation.  Ann. Rev. Pharmacol. Toxicol. 30: 73-90.

Margard, W. (1978) In vitro bioassay of chlorinated hydrocarbon solvents. Battelle
       Laboratories.  Unpublished document for Detrex Chemical Industries.

Maronpot, R.R.; Haseman, J.K.; Doorman, G.A.; Eustis, S.E.; Rao, G.N.; Huff, I.E.
       (1987) Liver lesions B6C3F1 mice: the National Toxicology Program, experience and
       position. Arch. Toxicol.  (Suppl) 10: 10-26.

Marsman, D.S.; Cattley, R.C.; Conway, J.G.; Popp, J.A.  (1988) Relationship of hepatic
       peroxisome proliferation and replicative DNA synthesis to hepatocarcinogenicity of
       the peroxisome proliferators di(2-ethylhexyl)phthalate and [4-chloro-6-(2,3-xylidino)-
       2-pyrimidinylthio]acetic acid (Wy-14,643) hi rats. Cancer Res,, 48: 6739-6744.

Mazzullo, M.; Grilli, S.; Lattanzi,  G. (1987) Evidence of DNA binding cavity of
       perchloroethylene. Res. Commun. Chem. Pathol. Pharmacol. 58(2): 215-235.

McConnell, E.E. (1990) Mouse liver tumor: the problem. In: Stevenson, D.E.; McClain,
       R.M.; Popp, J.A.; Slaga, T.J.; Ward, J.M.; Pilot, H.C., eds. Mouse liver
       carcinogenesis: mechanisms and species comparisons. Progress hi clinical and
       biological research, Vol. 331. New York, NY: Wiley-Liss; pp. 1-3.

McGregor, D.B.; Brown, A.; Cattanach, P.;  Edwards, I. (1988) Responses of the L5178Y
       TK+/TK- mouse lymphoma cell forward mutation assay. 3.72 coded chemicals.
       Environ. Molec. Mutagen.  12: 85-154.

McKinney, L.L.; Weakley, F.B.; Eldridge, A.C.; Campbell, R.E.; Cowan, J.C.; Picken,
       J.C.; Blester, H.E. (1957) S(dichlorovinyl) L-cysteine: an agent causing fatal aplastic
       anemia hi calves. J.  Am. Chem. Soc.  51: 16946E.

Myhr, B.C.; Bowers, L.R.; Caspary, WJ. (1986) Results from the testing of coded
       chemicals hi the L5178Y TK+/- mouse lymphoma mutagenesis assay. Environ.
       Mutagen. 8(6): 58.
                                         67

-------
National Toxicology Program (NTP). (1983) Carcinogenesis bioassay of pentachloroethane
       (CAS No. 76-01-7) in F344/N rats and B6C3F1 mice (gavage study). NTP-81-34.
       NIH Publication No. 83-1788. NTP TR 232.

National Toxicology Program (NTP). (1986a) Toxicology and carcinogenesis of
       tetrachloroethylene (perchloroethylene) (CAS No. 127-18-4) in F344/N rats and
       B6C3F1 mice (inhalation studies). NIH publication No. 86-2567. NTP TR 311.

National Toxicology Program (NTP). (1986b) Toxicology and carcinogenesis studies of
       isophorone (CAS No. 78-59-1) in F344/N rats and B6C3F1 mice (gavage studies).
       NIH publication No. 86-2547. NTP TR 291.

National Toxicology Program (NTP). (1987) Toxicology and carcinogenesis studies of
       dimethyl methylphosphonate (CAS No. 756-79-6) hi F344/N rats and B6C3F1 mice
       (gavage studies).  NIH publication No. 87-2579. NTP TR 323.

National Toxicology Program (NTP). (1988) Draft. Toxicology and carcinogenesis studies of
       d-limonene (CAS No. 5989-27-5) hi F344/N rats and B^Fj mice (gavage studies).
       NIH publication No. 88-2902. NTP TR 347.

National Toxicology Program (NTP). (1989) Toxicology and carcinogenesis studies of
       hexachloroethane (CAS No.  67-72-1) hi F344/N rats  (gavage studies). NIH
       publication No. 89-2816. NTP TR 361.

National Toxicology Program (NTP). (1990) Final. Toxicology  and carcinogenesis studies of
       d-limonene hi F344/N rats and ^C3F1 mice (gavage studies). NTP TR 347.
       Washington, DC: U.S. Department of Health and Human Services.

Nelson, M.A.; Bull, R.J. (1988) Induction of strand breaks hi DNA by trichloroethylene and
       metabolites in rat and mouse liver hi vivo. Toxicol. Appl. Pharmacol. 94: 45-54.

Nelson, M.A.; Lansing, A.J.; Sanchez,  I.M.; Bull, R.J.; Springer, D.L. (1989)
       Dichloroacetic acid and trichloroacetic acid-induced DNA strand breaks are
       independent of peroxisome proliferation.  Toxicology 58: 239-248.

Nelson, M.A.; Sanchez, I.M.; Bull, R.J.; Sylvester, R.S. (1990) Increased expression of c-
       myc and cH-ras hi dichloroacetate- and trichloroacetate-induced liver tumors in
       B6C3F1 mice. Toxicology 64: 47-55.

Odum, J.; Green, T.; Foster, J.R.;  Hext, P.M.  (1988) The role of trichloroacetic and
       peroxisome proliferation hi the differences hi carcinogenicity of perchloroethylene in
       the mouse and rat. Toxicol.  Appl. Pharmacol. 92:  103-112.
                                         68

-------
Oesch, F.; Wolf, C. (1989) Properties of the microsomal and cytosolic glutathione
       transferases involved in hexachloro-l:3-butadiene conjugation. Biochem.  Pharmacol.
       38: 353-359.

Office of Science and Technology Policy (OSTP). (1985) Chemical carcinogens: review of
       the science and its associated principles.  Federal Register 50: 10372-10422.

Ohtsuki,  T.; Sato, K.; Koizumi, A.; Kumai, M.; Ikeda, J. (1983) Limited capacity of
       humans to metabolize tetrachloroethylene. Int. Arch. Occup. Environ. Health 51(4):
       381-390.

Olson, M.J.; Garg,  B.D.; Murty, C.V.R.; Roy, A.K. (1987) Accumulation of alpha-2u-
       globulin hi the renal proximal tubules of male rats exposed to unleaded gasoline.
       Toxicol. Appl. Pharmacol. 90(1): 43-51.

Oshimura, M.; Koi, M.; Ozawa, N. (1988) Role of chromosome loss in ras/myc-induced
       Syrian hamster tumors. Cancer Res. 48:  1623-1632.

Page, H.S.; Asire, A.J. (1985) Cancer rates and risks. NIH publication No. 85-691.

Phillips, R.D.; Moran, E.J.;  Dodd, D.E.; Fowler, E.H.; Dary, C.D.; O'Donoghue, J.
       (1987) A  14-week vapor inhalation toxicity study of methyl isobutyl ketone. Fund.
      Appl. Toxicol. 9: 380-388.

Pickle, L.W.; Mason, T.J.; Howard, N.; Hoover, R.; Fraumeni, J.F.  (1987) Atlas of U.S.
      cancer mortality among whites:  1950-1980. Washington, DC: U.S. Department of
      Health and Human Services Publication 46 (NIH 87-2900).

Popp, J.A., ed. (1984) Mouse liver neoplasia. Washington, DC: Hemisphere.

Rampy, L.W.; Quast, J.F.; Balmer, M.F.; Leong, B.K.J.; Gehring, P.J. (1978) Results of a
      long-term inhalation toxicity study: perchloroethylene hi rats. Toxicology Research
      Laboratory, Health and Environmental Research, Dow Chemical Company, Midland,
      MI. Unpublished.                                        ,

Reynolds, S.H.; Stowers, S.J.; Patterson, R.M.; Maranpot, R.R.; Aaronson, S.A.;
      Anderson, M.W. (1987) Activated oncogenes hi B6C3F1 mouse liver tumors:
      implications for risk assessment.  Science 237:  1309.

Rijhsinghani, K.S.; Abrahams, C.; Swerdlow, M.A.; Rao, K.V.; Ghose, T. (1986) Induction
      of neoplastic lesions in the livers of C57BL  X C3HF1 mice by chloral hydrate.
      Cancer Detect. Prev. 9(3-4): 279-288.                     ;
                                         69

-------
Ris, H. (1949) The anaphase movement of chromosones in the spermatocytes of the
       grasshopper.  Biol. Bull. 96: 90-106.

Russo, A.; Pacchierotti, F.; Metalli, P. (1984) Nondisjunction induced in mouse
       spermatogenesis by chloral hydrate, a metabolite of trichloroethylene. Environ.
       Mutagen. 6: 695-703.

Schultze, M.O.; Klubes, P.; Perman, V.; Mizuno, W.S.; Bates, F.W.; Sautter, J.H. (1959)
       Blood dyscrasia hi calves induced by S(dichlorovinyl) L-cysteihe. Blood 14:  1015.

Schumann, A.M.;  Quast, J.F.; Watanabe, P.O. (1980) The pharmacokinetics and
       macromolecular interactions of perchloroethylene in mice and rats as related to
       oncogenicity.  Toxicol. Appl. Pharmacol. 55: 207-219.

Shimada, T.; Swanson, A.F.; Leber, P.; Williams, G. (1985) Activities of chlorinated ethane
       and ethylene compounds hi the Salmonella/Tat microsome mutagenesis and rat
       hepatocyte/DNA repair assays under vapor phase exposure conditions. Cell Biol.
       Toxicol.  1: 159-179.

Singh, M.; Sinha,  U. (1976) Chloral hydrate induced haploidization in Aspergillus nidulans.
       Experientia 32: 114-115.

Singh, M.; Sinha,  U. (1979) Mitotic haploidization and growth of Aspergillus nidulans on
       media containing chloral hydrate. J. Cytol. Genet.  14: 1-4.

Solleveld, H.A.; Haseman, J.K.; McConnell, E.E. (1984) National history of body weight
       gain, survival, and neoplasia in the F344 rat. J. Natl. Cancer Inst. 72: 929-940.

Sora, S.; Agostini-Carbone, M.L. (1987) Chloral hydrate, methylmercury hydroxide and
       ethidium bromide affect chromosomal segregation during meiosis of Saccharomyces
       cervisiae. Mutat. Res. 190: 13-17.

Stevenson, D.E.; McClain, R.M.; Popp, J.A.; Slaga, T.J.; Ward,  J.M.; Pilot, H.C. (1990)
       Mouse liver tumor carcinogenesis: mechanisms and species comparisons. Progress in
       clinical and biological research, Vol. 331. New York, NY:  Wiley-Liss;  444 p.

Stonard, M.D.;  Phillips, P.G.N.; Foster, J.R.; Simpson, M.G.; Lock,  E.A. (1986) Alpha-
       2u-globulin measurement in rat kidney following administration of 2,2,4-
       trimethylpentane. Toxicology 41(2): 161-168.

Swenberg, J.A.; Short, B.; Borghoff, S.; Strasser, J.;  Charbonneau, M. (1989) The
       comparative pathobiology of alpha-2u-globulin nephropathy. Toxicol. Appl.
       Pharmacol. 97(1):  35-46.
                                          70

-------
 Tramp, B.F.; Lipsky, M.M.; Jones, T.W.; Heatfield, B.M.; Higginson, J.; Endicott, K.;
        Hess, H.B.  (1984) An evaluation of the significance of experimental hydrocarbon'
        toxicity to man. In: Mehlman, M.A.; Hemstreet, G.P.; Thorpe, J.J.; Weaver, N.K.,
        eds. Advances in modern environmental toxicology, Vol. VII. Renal effects of
        petroleum hydrocarbons. Princeton,  NJ: Princeton Scientific Publishers  Inc • DD
        273-288.

 Tramp, B.F.; Berezesky, I.K.;  Smith, M.W.; Phelps, P.C.; Elliget, K.A. (1989) The
        relationship  between cellular ion deregulation and acute and clironic toxicity Toxicol
        Appl. Pharmacol. 97(1): 6-22.

 Universities Associated for Research and Education in Pathology (UAREP). (1983)
        Hydrocarbon toxicity: acute, subchronic,  and chronic effects in relation to unleaded
        gasoline exposure of rodents with comments on the significance to human health.
        Universities  Associated for Research and Education in Pathology, Inc. Contract No.
        PS-6 UAREP (504-2)  with the American  Petroleum Institute.

 U.S. Environmental Protection Agency (U.S. EPA). (1985a) Health assessment document for
       tetrachloroethylene (perchloroethylene). Prepared by the Office of Health and
       Environmental Assessment, Washington, DC. EPA/600/8-82/005F. Available  from:
       National Technical Information Service, Springfield, VA; PB-85-249704/AS.

 U.S. Environmental Protection Agency (U.S. EPA).  (1985b) Health assessment document for
       trichloroethylene. Prepared by the Office  of Health and Environmental Assessment
       Washington, DC. EPA/600/8-82/006F. Available from: National Technical
       Information Service, Springfield, VA; PB-85-249696/AS.

 U.S. Environmental Protection Agency (U.S. EPA).  (1986a) Addendum to the health
       assessment document for tetrachloroethylene (perchloroethylene). Prepared by  the
       Office of Health and Environmental Assessment, Washington, DC. EPA/600/8-
       82/005FA. Available from: National  Technical Information Sei-vice,  Springfield  VA-
       PB-86-174489/AS.                                       ;
                                                               i
U.S. Environmental Protection Agency (U.S. EPA). (1986b) Guidelines for carcinogen risk
       assessment. Federal Register 51: 33992-34003.

U.S. Environmental Protection Agency (U.S. EPA). (1987) Evaluation of the carcinogenicity
       of unleaded gasoline. Prepared by the Office of Health and Environmental
       Assessment,  Washington, DC. EPA/600/6-87/001. Available from the National
       Technical Information Service, Springfield, VA; PB-87-186151/AS.

U.S. Environmental Protection Agency (U.S. EPA). (1989a, December 27)  Federal Register
       54(247): 53057-53065.
                                         71

-------
U.S. Environmental Protection Agency (U.S. EPA). (1989b, August 14) Federal Register
       54(155): 33418-33420.

U.S. Environmental Protection Agency (U.S. EPA). (1989c, May 22)  Federal Register
       54(97): 22062-22160.

U.S. Environmental Protection Agency (U.S. EPA). (1991) Alpha-2u-globulin: association
       with chemically induced renal toxicity and neoplasia in the male rat. Risk Assessment
       Technical Panel document draft.

U.S. National Cancer Institute (NCI). (1977) Bioassay  of tetrachloroethylene for possible
       carcinogenicity. U.S. Department of Health, Education, and Welfare Publication No.
       (NIH) 77-813. Bethesda, MD.

Valencia, R.; Mason, J.M.; Woodruff, R.C.; Zimmering, S. (1985) Chemical mutagenesis
       testing on Drosophila. II. Results of 48 coded compounds tested for the National
       Toxicology Program. Environ. Mutagen. 7: 325-348.

Vamvakas,  S.; Dekant, W.; Berthold, K.; Schmidt, S.; Wild, D.; Henschler, D. (1987)
       Enzymatic transformation of mercapturic acids derived from halogenated alkenes to
       reactive and mutagenic intermediates. Biochem. Pharmacol. 36(17): 2741-2748.

Vamvakas,  S.; Dekant, W.; Henschler, D. (1989a) Genotoxicity of haloalkene and
       haloalkane glutathione S-conjugates in porcine kidney cells. Toxicol. In Vitro 3(2):
    '   151-156.

Vamvakas,  S.; Herkenhoff, M.; Dekant,  W.;  Henschler, D.  (1989b) Mutagenicity of
       tetrachloroethane  in the Ames test: metabolic activation by conjugation with
       glutathione. J. Biochem. Toxicol.  4(1): 21-27.

Vamvakas,  S.; Dekant, W.; Henschler, D. (1989c) Assessment of unscheduled DNA
       synthesis hi a cultured line of renal epithelial cells exposed to cysteine S-conjugates of
       haloalkenes and haloalkanes.  Mutat. Res. 222(4): 329-336.

Walles,  S.A.S. (1986) Induction of single-strand breaks in DNA of mice by trichlorethylene
       and tetrachloroethane. Toxicol. Lett. 31: 31-35.

Warner, J.R.; Hughes, T.J.; Claxton, L.D. (1988) Mutagenicity of 16 volatile organic
       chemicals hi an evaporation technique with Salmonella typhimurium TA100.
       Environ. Molec.  Mutagen.  11(11): 111-112.

 Waskell, L. (1978) A study of the mutagenicity of anesthetics and their metabolites. Mutat.
       Res. 57: 141-153.
                                           72

-------
Yllner, S. (1961) Urinary metabolites of 14C-tetrachloroethylene in mice. Nature 191: 820.


Yoon, J.S.;  Mason, J.M.; Valencia, R.; Woodruff, R.C.; Zimmering, S. (1985) Chemical
      mutagenesis testing in Drosophila. IV. Results of 45 coded compounds tested for the
      National Toxicology Program.  Environ. Mutagen 7(3): 349-367.


Zhang, G., Stevens, J. (1989) Transport and activation of S-(l,2-dich!orovinyl)-L-cysteine
      and N-acetyl-S-(l,2-dichlorovinyl)-L-cysteine in rat kidney proximal tubules  Toxicol
      Appl. Pharmacol. 100: 51-61.
                                         73

-------

-------