xvEPA
            United States
            Environmental Protection
            Agency
            Office of Health and
            Environmental Assessment
            Washington DC 20460
EPA/600/8-82/005FA
April 1986
External Review Draft
            Research and Development
Addendum to the     Review
Health Assessment   Draft
Document for        (DO Not
Tetrachloroethylenecite or Quote)
(Perchloroethylene)

Updated
Carcinogenicity
Assessment for
Tetrachloroethylene
(Perchloroethylene,
PERC, PCE)
                        NOTICE

            This document is a preliminary draft. It has not been formally
            released by EPA and should not at this stage be construed to
            represent Agency policy. It is being circulated for comment on its
            technical accuracy and policy implications.

-------

-------
          DRAFT                                                       EPA/600/8-82/005FA
          DO NOT QUOTE OR CITE                                        April  1986
                                                                      Review Draft
f
                             ADDENDUM TO THE HEALTH ASSESSMENT DOCUMENT

                             FOR TETRACHLOROETHYLENE (PERCHLOROETHYLENE)



                                 Updated Carcinogenicity Assessment

                       for Tetrachloroethylene  (Perch!oroethylene, PERC, PCE)
                                               NOTICE

            THIS  DOCUMENT  IS  A  PRELIMINARY DRAFT.   It  has  not  been  formally  released  by
            the U.S.  Environmental  Protection Agency and should not at this  stage be
            construed to represent  Agency policy.   It  is being circulated  for  comment
            on its  technical  accuracy and policy implications.
                           Office of Health  and  Environmental Assessment
                                 Office of Research and Development
                                U.S. Environmental Protection Agency
                                          Washington, D.C.

-------
                                    DISCLAIMER
                                    i


                                    i
     This document is an ext&rn.al  draft  for  review purposes  .only and .-.doe-s

not constitute Agency .policy.   Mention  of trade names or commercta-i  products

does not constitute endorsement or recommendation  for use.
   The Health Assessment  (Document  for Tetraehloroethylene (Perch!oroethylene)
   (July 1985; .EPA-600/8-82-005F)  i?  .available front:

                      National  Technical  Information Service
                      5.285 Port Royal  Road
                      Springfield,  VA   22161
                      Telephone:   703-487-465U

   Order No.:  PB-85-249704
   Cost:  $22,95  (subject to  change)

-------
                                      CONTENTS
P P£*f f\ f P

Authors! Contributors* and Reviewers  !'*'*'  '  ''''        *  '  '    1'
                       *                      '*'  '  ''''        *  '  '

  1.  SUMMARY AND  CONCLUSIONS
 2.   INTRODUCTION  ......  .............


 3*  ?MriEWi!L!HE NATIONAL TOXICOLOGY  PROGRAM  INHALATION BIOASSAY
     vINIr, 1985; ...............  .....                     ^  i


     3.1.  ANIMALS STUDIED. .....                                       01
     3.2.  RESULTS OF THE RAT STUDY  ...'*'  '  '  '  .........     ,,
     3.3.  RESULTS OF THE MOUSE STUDY ...*'*  ........... '    o"in
     3.4.  SUMMARY ..... . .....  ...'.'.'.  ..........     o~Tc
                                                     *  *  * •  •  *  •  •  •  •    • -3 ™ *A 0

 4.  QUANTITATIVE RISK ASSESSMENT ..........  ........       4-r
     4.1.  DATA AVAILABLE FOR RISK CALCULATION
           4.1.1.  NTP Rat Inhalation Assay ..........            4 2
           4.1.2.  NTP Mouse Inhalation Assay ....!*!' 1 !*,*.]    4! 2

     4.2,.  POSSIBLE MECHANISMS LEADING TO A CARCINOGENIC RESPONSE
                                                                    • •    ft
     4.5.   CALCULATION OF METABOLIZED DOSE FROM INHALATION 'STUDIES.' .' .'    t-8

           4.5.1.   Direct Estimation from Metabolism Experiments. . . .    4-8

                   4.5.1.1.   Rat.  ............ .......    4-8
                   4.5.1.2.   Mouse ............ *!*.!*,*,"    4-11

           4.5.2.   Metabolized  Dose  Predicted  by  a  PB-PK Model.  . . .  .    4-12


                                                                    '  '    ^
                                                          .......    4-16

           4.7.1.   Based  on the Metabolized Dose  Estimated
                   Directly from the  Empirical  Data  .....               415
           4.7.2.   Based  on the Metabolized Dose  Calculated
                   from the PB-PK  Model  .........  .......     4_20

    4.8.   DISCUSSION  .......                                         ...
    4.9.   SUMMARY  OF QUANTITATIVE ASSESSMENT  .' .' *  ."  .*  .*  '  .*  *  '  *  '  '
5.  REFERENCES ............. •                                  ^
                                        ***••*•*•••*••••    D~i

APPENDIX A:  PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS
        -     AND THEIR APPLICATION IN THE QUANTITATIVE RISK
             ASSESSMENT OF PCE ..... ....... ...  .....    A-l


                                      iii

-------
                                    PREFACE
     The Office of Health and Environmental Assessment has prepared this adden-



dum to serve as a source document, for EPA use.  This addendum updates EPA's



July 1985 Health Assessment Document! for Tetrachloroethylene (perchloroethy-



lene, PERC, PCE) by providing a review of the findings of the draft National



Toxicology Program (NTP) Technical Report on the Toxicology and Carcinogenesi s



Studies of Tetrachloroethylene in F344/N rats and B6C3F1 mice (inhalation



studies).  The addendum discusses th|e ways in which the data impact the assess-
                                    I


ment of the weight-of-evidence for the carcinogenicity of PCE, and uses rele-



vant data to develop a revised unit risk estimate for inhalation exposure to



PCE.

-------
      ;                              PREFACE

     The Office of Health and Environmental Assessment has prepared this adden-
dum to serve as a source document for EPA use.  This addendum updates EPA's
July 1985 Health Assessment Document for Tetrachloroethylene (perch!oroethy-
lene, PERC, PCE) by providing a review of the findings of the draft National
Toxicology Program (NTP) Technical Report on the Toxicology and Carcinogenesis
Studies of Tetrachloroethylene in F344/N rats and B6C3F1 mice (inhalation'
studies).  The addendum discusses the ways in which the data impact the assess-
ment of the weight-of-evidence for the carcinogenicity of PCE,  and uses rele-
vant data to develop a revised unit risk estimate for inhalation exposure to
PCE.

-------

-------
                           1.  SUMMARY AND CONCLUSIONS

      In the July 1985 Health Assessment Document for Tetrachloroethylene (per-
 chloroethylene, PERC, PCE), the U.S. Environmental  Protection  Agency's  Office
 of Health and Environmental Assessment (OHEA)  summarized  the pertinent  evidence
 for the carcinogenicity of PCE, based primarily on  the  results  of  a  2-year  bio-
 assay conducted by the National Cancer Institute (NCI,  1977).   In  that  study,
 PCE induced a statistically significant increase in the incidence  of hepatocell-
 ular carcinomas in both high-  (1072  and 772 mg/kg/day)  and  low-dose  (536 and
 386 ng/kg/day)  male and female B6C3F1  mice administered PCE by  gavage for .a
 period  of 78 weeks.
      In a 1985  inhalation  bioassay conducted by  the National Toxicology Program
 (draft  report approved  by  the  Board  of  Scientific Counselors), groups of 50
 male and 50 female F344/N  rats  and B6C3F1 mice were  exposed to PCE (99.9% pure),
 6  hours/day,  5  days/week,  for 2 years.  The exposure concentrations used were
 0,  200,  and 400 pprn  for rats and 0,  100, and 200 ppm for mice.   At  the 200 and
 400 ppm exposure concentrations, PCE induced a marginally statistically signi-
 ficant  increase of rnononuclear cell  leukemia  in male rats.  PCE also  induced an
 increased  incidence  of  renal tubular cell adenomas and/or carcinomas  (combined)
 in male  rats.  While not statistically significant,  there  was  a dose-related'
 trend of  these  rare  tumors.  In female rats at  both  the  200 and 400 ppm  expo-
 sure concentrations, statistically significant  increased incidences of mono-
 nuclear  cell leukemia were observed.   In male  and female B6C3F1  mice, PCE
 induced a statistically significant  increased incidence  of hepatocellular
 carcinomas at both the 100 and the 200 ppm exposure  concentrations.  Although
mutagenicity studies of PCE have produced  inconclusive or  negative  results,  PCE
epoxide, a reactive metabolite of PCE,  has been  found to be  mutagenic.
                                      1-1

-------
     The NTP (1985, draft) inhalation bioassay demonstrated that PCE -can induce
                                    I

carcinogenic effects in both rats and mice through inhalation exposure,  The

earlier NCI (1977) bipassay provided positive evidence of h.epatocellalar car-

cinomas in mice administered PCE bylgavage,.  The.se findings, together with tihe

supporting evidence regarding the PCE epoxide (a presumed reactive metabolite

that has been shown to be mutayenic), constitute a "sufficient" level of evi-

dence for the carcinogenicity of P.C.E in animals, using the EPA's Proposed

Guidelines for Carcinogen Risk Assessment.  The evidence In rats arid mice, to-

gether with the inconclusive evidence in humans as reported in the 1985 tteailth

Assessment Document for Tetrachloroethylene,, results in the placement of .£CE in

EPA's weight-of-evidence Group B2, meaning that it should be considered a proba-

ble human carcinogen.               ;

     OHEA is aware of three issues related to the NTP (1985, draft)  inhalation

bioassay.  These issues are raised in a document presented to EPA by the Halo-

genated Solvent Industry Alliance (hisiA),, and focus on reported carcinogenic
                                    i
responses in male F344 rats, as well as on certain aspects of the conduct of

the bioassay.  The validity of a reported marginally statistically significant
                                    I
increased incidence of mononuclear cell leukemia in male F344 rats is in ques-

tion.  The finding is dependent on the criteria used for "staging" the leukemia
                                    i
and the approach used in assessing the variable incidence in the untreated

rats.  The reasonableness of these approaches, as used in data interpretation,

is at issue.  Also at issue is the significance of a reported finding of a low

incidence (not statistically significant) of rare tubular cell adenomas or

carcinomas in male F344 rats.  Lastly, alleged deficiencies In animal handling

and identification raise doubts about the reliability of the reported marginal

and low-level  incidences of tumors in male rats.

     The impact of the issues regarding the findings in male F344 rats relative
                                    I
                                    I
                                    i

                                      1-2

-------
to a weight-of-evidence evaluation is to add to the multiplicity of evidence

if the findings are considered valid; on the other hand, if the findings are

considered uncertain, the remaining evidence in male and female mice and female
        i
rats is likely to be judged "sufficient" according to EPA's Proposed Guidelines

for Carcinogen Risk Assessment.


     In this addendum, an inhalation unit risk estimate for PCE is  developed

using data from the NTP (1985, draft) inhalation bioassay—the first lifetime

inhalation study to show evidence of the carcinogenicity of PCE.  Carcinogenic

responses were observed in both rats and mice.  These data provide  a basis  for


directly estimating the risk of PCE by the inhalation route, and  an opportunity

to test the reasonableness of the previous inhalation risk estimate, which  was

developed using gavage data from the NCI bioassay (1977).


     It is generally recognized that the carcinogenic potential of  PCE resides

in its biologically reactive metabolites, such as the short-lived,  reactive PCE

epoxide intermediate, rather than the parent PCE compound  itself.   The rate of

metabolism of PCE is dose-dependent--meaning that the effective dose of active

metabolite is not proportional to the administered dose of PCE--and proceeds

according to Michaelis-Menten kinetics.   Covalent binding  and hepatotoxicity of

PCE are directly proportional  to metabolized dose.  The tumorigenic response

is also assumed to be directly related to metabolized dose.

     The ingestion and inhalation unit risk estimates reported in the 1985

Health Assessment Document for Tetrachloroethylene were extrapolated from a .

gavage cancer bioassay study in  mice.  Metabolism and kinetic information were

incorporated into these estimations, expressing dose in terms of  the amount

metabolized, or a "metabolized dose."  Thus, the inhalation  unit  risk estimate

for PCE reported in the 1985 document was based on tumor incidence  data from a

gavage study and the assimilated dose metabolized, accomplished by  incorporating



                                      1-3

-------
metabolism and kinetics information fpr both animals and humans.  On this
                                     I

basis, the inhalation unit risk is 4.|8 x 10-7/jig/m3.  The unit risk for


ingestion of 1 ug PCE/L in drinking water is 1.5 x 1CH>.

                                     t
     The new unit risk estimate for i'nhalation also utilizes the metabolized


dose-tumor incidence relationships derived from pertinent exposure concentra-


tion dose-metabolism relationship data in both animals and humans.  The line-
                                     I

arized multistage model is used for t'he low-dose extrapolation.  The details

                                     \
of this procedure are presented in the 1985 Health Assessment Document for


Tetrachloroethylene.  The typical uncertainty associated with extrapolating
                                     i

dose from animals to humans is reduced by the consideration of actual human


exposure concentration dose-metabolism relationship data, where urinary tri-


chloroacetic acid is taken as an index of the amount of PCE metabolized, in


addition to the use of comparable ani,mal data.  The inhalation unit risk esti-


mates calculated on the basis of animal inhalation carcinogenicity  studies


(rats and mice) are comparable to that previously derived from the  gavage
                                     j

carcinogenicity data in mice.  The upper-bound estimates of the incremental

                                              o
cancer risk due to lifetime exposure 'to I pg/m^ of PCE in air are calculated


for six sets of dose-tumor incidence data by three methods.  The  ranges of the
                                     i

six estimates  under each  of the three methods are as follows:   (a)  2.9 x 10"7


to 9.5 x 10-7/yg/m3; (b)  2.9 x ID-6 to 1.1  x 10-5/yg/m3; (c) 9.6  x  10"7 to


3.6 x 10-6/ug/m3.                    j


     The first range of estimates isjcalculated on the basis of metabolized
                                     i

doses that are derived directly from;a single-exposure balance  study  (assuming


that the total amount metabolized over a 72-hour period from a one-time expo-


sure  is comparable to the steady-state metabolized dose under the NTP multiple-


exposure cancer bioassay  study).  The second and third  range of estimates are
                                     I

calculated on  the basis of metabolized doses that are derived from  a  physiolog-


                                     : 1-4

-------
 ically-based  pharmacokinetic  (PB-PK)  model.   Such  models, which  piece  together
 the relevant  physiological  and  biochemical  information,  can  be useful  tools  for
 predicting  absorption, distribution,  biotransformation,  and  elimination of .-
 chemicals across  species when sufficient data are  available.  However, until
 more  data become  available, EPA's Carcinogen Assessment  Group (CA6) recommends
 that  the first  range of estimates be  used to represent the inhalation  risk for
 PCE,  since  there  are uncertainties with parameters used  in the PB-PK model
 that  were estimated from the same empirical data set used in the direct estima-
 tion, and the model cannot be considered to give a better estimate.  That is,
 the revised upper-bound estimate of the incremental cancer risk due to lifetime
 exposure of 1 pg/m3 of PCE in air ranges from 2.9 x 10~7 to 9.5 x 1Q-7.  This
 unit  risk estimate is applicable only for low-level exposures where the rela-
 tionship between the ambient air concentration and metabolized dose is linear.
 The upper-bound nature of these estimates is such that the true risk for humans
 is not likely to exceed these values and may be  lower.  Expressed in terms of
 relative potency, PCE ranks in the lowest quartile among 55  suspect or known
carcinogens  evaluated by  the CAG.
                                     1-5

-------

-------
       . ;                        2.  INTRODUCTION

     The July 1985 Health Assessment Document for Tetrachloroethylene (per-
ch! oroethylene, PERC, PCE) (U.S. EPA, 1985) summarized the evidence that PCE
administered by gavage induced malignant tumors pf the liver in both male and
female B6C3F1 mice (NCI, 1977).  The Health Assessment Document also cited the
results of several epidemiologic studies, which were characterized as providing
inconclusive findings.  According to the Agency's Proposed Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 1984), the carcinogenic evidence avail-
able as of July 1985 for PCE in animals was judged to be limited, and the
epidemiologic data inconclusive.  The overall weight-of-evidence for the car-
cinogenicity of PCE was classified as belonging in EPA Group C, meaning that
PCE should be considered a "possible" human carcinogen.  Quantitative human
unit risk estimations for ingestion and inhalation were developed using the NCI
mouse gavage study, relevant metabolism and kinetics data, and the linearized,
multistage model for low-dose extrapolation.
     Since publication of the July 1985 Health Assessment Document for Tetra-
chloroethylene, the National  Toxicology Program inhalation bioassay draft
report (NTP, 1985) on PCE carcinogenicity in rats and mice has become available.
The study findings have been  audited and validated for technical  accuracy,  and
the report has been peer-reviewed by the NTP Board of Scientific  Counselors.
Male F344/N rats in the study showed a statistically significant  increased
incidence of mononuclear cell leukemia and a dose-related trend for rare renal
tubular neoplasms.  Female F344/N rats showed statistically significant  inci-
dences of mononuclear cell  leukemia.  In addition, PCE induced a  statistically
significant increased incidence of hepatocellular carcinomas in both  male and
female B6C3F1 mice.
                                      2-1

-------
     This addendum reviews the new NTPj (1985, draft) inhalation bioassay of PCE

in detail, and discusses the ways in which the data impact the assessment of

the weight of evidence for the carcinogenicity of PCE.  This addendum also in-

cludes a section on unit risk estimation which incorporates relevant •inhalation
                                       i
metabolism and kinetics data.          '

-------
             .3.   REVIEW  OF  THE  NATIONAL TOXICOLOGY  PROGRAM  INHALATION
                       CARCINOGENICITY BIOASSAY  (NTP, 1985)

      A 2-year inhalation cancer bioassay of^PCE in F344/N rats and B6C3F1 mice
 has been performed by Battelle Pacific Northwest Laboratories, under the spon-
 sorship of the National  Toxicology Program {NTP, ;1.985).  The study provides
 positive evidence for the carcinogenicity .of PCE in bothiof the, above species.
 A description of the various aspects of the study'follows.;               -   ';
      The study utilized  high-purity PCE (99.9$)  obtained from the Dow Chemical
 Company.  The purity and identity analyses of the PCE were performed by the
 Midwest Research Institute.  The PCE was  stabilized with n-methylmorpholine
 (53 ppm)  to prevent decomposition.  The PCE was  ,then vaporized  'at 100°C to
 110°C,  diluted with air, and introduced  into the inhalation  chambers.   The
 concentrations in the chambers  were monitored 8  to  12 times  per exposure period
 by means  of a  Hewlett-Packard 5840A gas chromatograph equipped  with  a  flame
 ionization  detector.
 3.1.  ANIMALS  STUDIED
     The 8- to 9-week-old male  and  female F344/N rats and B6C3F1 mice  used in
 this study  were obtained from the  Charles River  Breeding Laboratories.   Groups
 of 50 rats:'of each ;sex were  exposed to air containing PCE at concentrations of i
 0  (chamber^controlE  200, or 400 ppm, 6 hours/day, 5,days/week for 103 weeks.  V:
 Groups of 50 mice of  each sex were also exposed to PCE at 0, 100, and 200 ppm. ; :
 These doses /were selected based on the results of an earlier 13-week study in  :
which incidence ;of death at 1,600 ppm and incidence of liver lesions at lower
concentrations had been observed for rats.  Incidence of death at 1600. ppm,  an.d
of hepatic and renal lesions at  lower concentrations, were observed for mice.
Rats and mice were housed individually.   The animals were given  a diet  (NIH07
                                      3-1

-------
 Open Formula) and water freely, except during the exposure period,  when  water


 was available.


      All  animals were observed two tijmes per day.  Clinical  signs were recorded


 once per month.  Body weights were recorded  once each  week for  the  first  13


 weeks of the study and once per month! thereafter.  Mean  body weights  were


 calculated for each group.  Moribund  janimals, as well  as  animals that survived
                                      I
 to the end of the study,  were killed.;  A necropsy was  performed on  all animals.


 Vital  organs were examined for any grjossly visible lesions.   Tissues  were


 preserved in 10% neutral  buffered  formalin,  embedded in  paraffin, sectioned,


 and stained with hematoxylin  and eosih,  with  other stains  being used  as needed.


 The tissue slides were examined microscopically  to determine pathology.


 3.2.   RESULTS OF THE RAT  STUDY       ;


      Figure 3-1 presents  the  growth curves for male and  female  rats exposed to


 PCE by inhalation for 2 years.   The mean  body weights of PCE-exposed male and


 female rats were comparable with those of controls.
                                      i

      Survival  data  for rats in  the 2-j^ear inhalation study are given Table 3-1


 and Figure 3-2.   The survival  of malejrats exposed  to 400 ppm was significantly
                                      I

 reduced (control, 23/50;  200  ppm, 20/50;  400  ppm,  12/50) but not that of  females
                                      i

 (control,  23/50;  200 ppm,  21/50; 400 ppm, 24/50).   Most of the unscheduled deaths


 (87%)  in the  400  ppm dose  group  occurred  late in the study (week 82  or later)


 and may have  been related  to  the high incidence of mononuclear cell  leukemia.


     Table  3-2  presents the incidence [of mononuclear cell leukemia  in male and
                                      i

 female rats exposed  to 200 and 400 ppm PCE.   Incidences of mononuclear cell
                                      i

 leukemia in male  rats were marginally'statistically significant:  control,


 28/50; 200  ppm, 37/50  (p = 0.046);  400 ppm, 37/50 (p =  0.046).  In  female rats,


the incidences of mononuclear cell  leukemia were also statistically  signifi-


cant:  control, 18/50; 200 ppm, 30/50j(p =0.014); 400  ppm, 29/50 (p = 0.022).
                                      i


                                      3-2

-------
    6OO.O
«,   450.0-
3E
    100.0
           >
    500.0
«0   460.0-
O
    400.0-
j_   360.0-
X
o
uj   300.0-
     100.0
                          i
                                                  MALE RATS
                                                D = UNTREATED
                                                O = 2OO  PPM
                                                AS 400  PPM
       —T~
        30
     45      90      76
WEEKS  ON STUDY
                                                            00
106
                    3
                      g
      Q@   88
                                  §8
                              FEMALE  RATS
                              D = UNTREATED
                              O = 200 PPM
                              A = 400 PPM
15      30      45      6O      75
           WEEKS  ON STUDY
                                                            90
                                     105
       Figure 3-1.   Growth curves for rats  exposed to PCE by
       inhalation for 2 years.
       SOURCE:   NTP,  1985.
                                3-3

-------
                          TABLE 3-1.  'SURVIVAL OF RATS
                  IN THE NTP 2-YEAR INHALATION BIOASSAY OF PCE
Group C
Males
Animals initially
in the study
Nonacci dental deaths
Died during termination
period
Killed at termination
Females
Animals initially
in the study
Nonacci dental deaths
Killed at termination13
ontrol 200 ppma

50 50
27 (54%) 3.0 (60%)
0 1 (2%)
23 (46%) 19 (38%)
50 50
27 (54%) 29 (58%)
23 (46%) 21 (42%)
4QQ ppm

50
38 (76%)
1 (2%)
11 (22,%)
50
26 (52%)
24 (48%)
£1 ppm is equivalent to 6908 yg/m3.
bTerminal kill period:  week 104.

SOURCE:  NTP, 1985.
                                      3-4

-------
   1.0


   0.9-

 > 0.8-

 «C

 u.
 O
   0.6-
 t-
 ^0.5H
 CD
 ffi
 O0.4-
 E
 a.
   0.3-
  0.2
  0.9
> 0.8-

oc
= 0.7-1
  o.e-
  0.5-
rt
a.
  0.3-
  O.2
   MALE  RATS
 O = UNTREATED
 0= *00 PPM
 A = 400 PPM
              15      SO      45      «0      76
                          WEEKS  ON  STUDY
 FEMALE RATS
D s  UNTREATED
O =  200  PPM
A =  400  PPM
                                         90     105
                     80      45      60      7'6
                         WEEKS  ON  STUDY
                                                105
Tl9nor ?"2'   KaP1an-Meier survival curves  for  rats exposed
 to PCE by inhalation  for 2 years.
 SOURCE:  NTP,  1985.
                             3-5

-------
                 TABLE; 3-2.   MQNONUCLpAR, CE.LL LEUKEMIA  IN RATS
                           EXPOSED TO pCE BY INHALATION
Group
   Control
  200 ppma          400: p;pma
Hales

Mononuclear cell
 leukemia
28/50 ($6%-).
37/50 (74%)
 p = 0.046
37/50: (74%)
 p = 0..046
Females

Mononuclear cell
 leukemia
18/50 (36%)
30/50 (60%)
 p. =• 0.014;
29/50 ('58%)
 p = 0.022
ap-values were calculated  using the Flisher Exact Test.,
                                      • 3-6

-------
  The diagnoses of mononuclear cell leukemia were classified as stage 1.(early),
  stage 2 (intermediate), or stage 3 (advanced and probably fatal).
       The following criteria were used for assigning these stages:
       Stajej, - Spleen not enlarged or only moderately enlarged, with  few '
       suspicious cells surrounding Malpighian corpuscles of the  spleen, and
       only  very few neoplastic  cells in the blood vessels of the  lung.
       Stage^ - More  severe than  in the  first  category;  spleen and/or liver
       only  moderately  enlarged, with neoplastic  cells  in  the  liver sinusoids
       and in  the  blood  vessels of  other organs.
       StajLl "  Enlarged spleen and  liver, with many neoplastic  cells in    -
      the liver sinusoids and in the blood  vessels of other organs.
 Table 3-3 summarizes the distribution of these three stages of mononuclear cell
 leukemia in the male and female rats studied.
      The strength of the evidence for carcinogenicity presented  here for  F344
 rats hinges on the resolution of  issues raised regarding the uncertainty  in   '
 the assignment of incidence within stages of mononuclear cell  leukemia  in  the
 treated  F344 rats and  the  high  and variable incidence  of mononuclear cell  leu-
 kemia in.F344 untreated rats, particularly  males.
      Table  3-4 presents  an  analysis  of  adenomas  and  carcinomas in  the rats  ex-
 posed to  PCE.   Tubular  cell  adenomas  or carcinomas were  observed  only in male
 rats,  but with  a  dose-related incidence that  is not  statistically  significant.
 Renal tubular  adenomas were:  in control, 1/49; 200  ppm, 3/49; 400 ppm, 2/50.,
 Carcinomas were:  in control, 0/49; 200 ppm, 0/49; 400 ppm, 2/50.   The his-
 torical incidences of renal  tubular adenomas in the NTP studies were  4/1720, or
 less than 0.2% ± 0.7%;  no malignant tubular cell carcinomas have  been observed
historically.  In addition,  dose-related increased incidences of  renal.tubular
karyomegaly  were observed in both  male and female rats.   Tubular  cell  hyper-
                                      3-7

-------
       TABLE 3-3.   CLASSIFICATION OF!NONONUCLEAR CELL LEUKEMIA IN RATS
                          EXPOSED TOJPCE BY INHALATION


Group
Males
Control
200 ppm
400 ppm
Females
Control
200 ppm
400 ppm
Number j
of rats [
with <
mononuclear \
cell leukemia

28
37
37

18
30
29
1
1


1

5
6
4

3
6
2
Stage

2

3
7
6

5
6
6


3

20
24
27

10
18
21
SOURCE:  NTP, 1985.
                                      3-8

-------
                 TABLE 3-4.  TUBULAR CELL ADENOMAS AND CARCINOMAS
               OF THE KIDNEY IN RATS EXPOSED TO PCE BY INHALATIONS
GrouP                            Control           200 ppma
                                                                      400 ppma
Males
Tubular cell
adenomas
Tubular. cell
adenocarcinomas
Tubular cell
adenomas or
adenocarcinomas
1/49 (2%)
0/49 (0%)
1/49 (2%)
3/49 (6%)
0/49 (0%)
3/49 (6%)
p = 0.259&
2/50 (4%)
2/50 (4%)
4/50 (8%)
p = 0.070&
Females


Adenomas611                    °/5°  (°%)          °/49  (°%)         0/5°

Tabular cell                    0/50  (0%)          0/50  (0%)         0/50  (0%)
 adenocarcinomas                                                           v   '


"adenomas"^                    °/5°  (°%)          °/49  (°%)         0/5°  ^
 adenocarcinomas


amstorical incidence at the Battelle Pacific Northwest Laboratories:  0/100-
 historical incidence in NTP studies:  4/1,720, 0.2% ± 0.7%; no malignant    '
 tubular cell  tumors were observed.

bp-values were calculated for tubular cell  adenomas or carcinomas using the
 Msher Exact  Test.

SOURCE:  NTP,  1985.
                                      3-9

-------
plasias were also observed;'these may be preneoplastic lesions.
      In the rats, there were no statistically significant increased incidences
                                    I
of tumors  In other organs examined.; In particular, no neoplastic changes were

observed in the respiratory tracts of either male or female rats.  However,
there was  an increased incidence ofjsquamous cell metaplasia in the nasal cavi-
ties of the exposed male rats  (0/50), control; 5/50, 200 ppm; 5/50, 400 ppm).
3.3.  RESULTS OF THE MOUSE STUDY
                                    i
                                    I
     Figure 3-3 presents the growth!curves for male and female mice exposed to
                                    i
PCE by inhalation for 2 years.  As indicated in the figure, there was no signi-

ficant change in the mean body weights of dosed animals as compared with con-
trols.  The numbers' of mice surviving the 2-year inhalation study of PCE are
                                    [
shown in Table 3-5 and Figure 3-4.  [The survival of male mice was significantly

reduced at both 100 and 200 ppm PCE exposures, as compared with controls:  low
dose, 25/50 (50%); high dose, 32/50j(64%); control, 46/50 (96%).  In female
                                    I
mice, survival  was reduced at 200 ppm as compared with controls:  low dose,
31/50 (62%); high dose, 17/50 (34%); control, 36/50 (72%).  Survival of male
control mice was quite high.  The high early mortality observed in female mice
[31/50 (60%)] may be attributed to line high incidence of hepatocellular car-
cinomas.                            ;
     The incidences of liver tumors 'in the male and female B6C3F1 mice exposed
                                    i
to PCE are summarized in Table 3-6.  In the male mice, PCE induced statisti-
                                    i
cally significant increased incidences of hepatocellular carcinomas at both the
100 and 200 ppm exposure levels, as [compared with controls.  These tumor inci-
                                    I
dences were:  low dose, 25/49 (51%) (p < 0.001); high dose, 26/50 (52%)  (p <
                                    r
0.001); control, 7/49 (14%).  Hepatocellular adenomas alone were not statistic-
                                    I
ally significant at either the 100 ppm or the 200 ppm exposure concentration.
                                    i
The combined incidences of hepatocetlular adenomas and carcinomas were sta-
                                    i                               .      •
                                      3-10

-------
  z
  < 35.0
  ac
  o
 X
 o
 IU
    30.0-
 >-  25.0
 O
 o
 <  20.0-
 IU
    15.0
            D
    40.0
 CO

 <  35.0
 OC
 O
    30.0
>  26.0
Q
O
CD
,<  20.0
ui
   15.0
                                           8°
                                                    o
                                          MALE MICE

                                        n s  UNTREATED
                                        O «  1OO PPM

                                        A B  200  PPM
         15      30      45      «0      76

                    WEEKS ON STUDY
                                                           90
                                                                  105
                                               S
                                                 a
                        A    O _
                   O  D O  O   °

                   O  A
                                        FEMALE MICE

                                       D•UNTREATED

                                       OB 100 PPM
                                       A K 200 PPM
0       16      30      45      «0      76

                   WEEKS ON  STUDY
90
                                                                 105
       ear'   Gr°Wth curves  for  mice exposed to  PCE by inhalation


SOURCE:   NTP,  1985.
                                3-11

-------
                           TABLE 3-5.   SURVIVAL OF MICE
                   IN THE  NTP  Z-YEAR IINHALATION BIOASSAY OF PCE
Group
Control
100 ppm
200 ppm
Males                               ,
                                    i
Animals initially  in  study        50 i
                                    i
Nonaccidental deaths               3 [(6%)

Animals missexed                   1 |(2%)
                                    i
Killed at termination             46  (92%)
                                    i

Females

Animals initially  in  study        50
                                    I
Nonaccidental deaths              11 !(22%)

Accidentally killed                2 |(4%)

Animals missexed                   1 j(2%)
                                    i
Killed at termination             36 |(72%)
                                    i

aTwo mice died during termination  period.
                   50

                   25 (50%)

                    0 (0%)

                   25 (50%)




                   50

                   17 (34%)

                    2 (4%)

                    0 (0%)

                   31 (62%)
                   50

                   18 (36%)

                    0 (0%)

                   32 (64%)




                   50

                   30 (60%)

                    1 (2%)

                    0 (0%)

                   17 (34%)a
                                      3-12

-------
  •.9
   O.7-
M.
O

I- O.C +
  0.6-
O
c
  0.4-
  O.3

  0.8-
-0.8-
  0.7-
t o.«H
  O.6-
  0.4-
  0.34
     0
                                              *1
                                                ^
   MALE MICE
O =*  UNTREATED
O=100  PPM
A =  200  PPM
                                                             00
             16      SO       46      CO      76
                          WEEK*  ON  STUDY
                                        00      105
             n -----
 FEMALE MICE
D = UNTMEATEO
0=10O  PPM
A = 200  PPM
       *
-------
          TABLE 3-6.   LIVER TUMORS IN[MICE EXPOSED TO PCE BY INHALATION
 Group
 Control3
 100
 200 ppma»b
Males

Hepatocellular  adenoma


Hepatocellular  carcinoma
Hepatocellular  adenoma
 or carcinoma
Females
11/49j(22%)


 7/49 j(14%)


16/49 l(33%)c
 8/49 (16%)
 N.S.

25/49 (51%)
 p<0.001

31/49 (63%)d
 p=0.002
18/50 (36%)
N.S.

26/50 (5.2%)
 p<0.001

40/50 (80%)e
 p<0.001
Hepatocel
Hepatocel
Hepatocel
or carci
lular
lular
lular
noma
adenoma
carcinoma
adenoma
3/48
(6%)
1/48 (2%)
4/48
™ 1 rt G ft O in rtrtm rt a 4* /*\ t^^ r*^\»-i*%/^f»j-***4" 4- W « mi i.-«l*. »»
(8%)
6/50
N.S.
13/50
p<0.
(12%)
(26%)
001
17/50 (34%)f
, p=0.001
2/50
N.S.
36/50
p<0.
38/50
p<0.
(4%)
(72%)
001
(76%)
001

 — —,^..,.,.v.~w. o  i v.^1 v^jtm,  one  nuiiiuci 3  of  animals  examined  histologically.
Dp-values were calculated using  the  Fiisher  Exact  Test.
^Controls, male -  2 mice had both adehoma and  carcinoma.
°Low dose, male -  2 mice had both adenoma and  carcinoma.
=High dose, male - 4 mice had both adenoma  and carcinoma.
TLow dose, female  - 2 mice had both  adenoma and carcinoma.
N.S. - not significant.

SOURCE:  NTP, 1985.                  \
                                      3-14

-------
 tistically significant as  compared  with  the  controls.   The  incidences were:
 low dose,  31/49  (63%)  (p = 0.002);  high  dose,  40/50  (80%) (p  <  0.001); control,
 16/49 (33%).
      In female mice,  PCE induced  a  statistically  significant  increased inci-
 dence of hepatocellular carcinomas  at  the  100  and 200 ppm exposure concentra-
 tions, as  compared  with the controls.  These incidences were:   low dose, 13/50
 (26%) (p =< 0.001); high dose,  36/50  (72%) (p  < 0.001); control, 1/48 (8%).
 Incidences of  adenomas were not statistically  significant.  However, the com-
 bined incidences of adenomas and  carcinomas were statistically  significant as
 compared with  controls.  The incidences  were:  low dose, 17/50  (34%) (p =
 0.001);;high dose, 38/50 (76%)  (p < 0.001); control, 4/48 (8%).  Carcinogenic
 effects  were not seen  in any of the other organ sites, including the respira-
 tory  system.   In addition,  dose-related  increases in liver degeneration,
 hepatocellular necrosis, and hepatocellular nuclear inclusion were observed
 in  both  male and female  mice exposed to  PCE.   Dose-related renal tubular  cell
 karyomegaly was observed in the kidneys  of both male and female mice, but no
 dose-related tubular cell carcinomas were found in the mice.
      The NTP mouse inhalation bioassay further confirms the  earlier National
 Cancer  Institute study in mice (NCI, 1977)  in which  PCE induced a statistically.
 significant increased  incidence of hepatocellular carcinomas by gavage.
      The NTP (1985, draft)  study was audited  by Dynamac  Corporation  for complete-
 ness, consistency,  accuracy, and for procedures consistent with good  laboratory
 practice requirements.  The audit  revealed  no major  problem  in the  conduct  of
the study or in the collection  and documentation  of  the  experimental  data.
 3.4.  SUMMARY
     In the NTP (1985,  draft)  inhalation  bioassay, groups of 50 male  and  50
female F344/N rats  and B6C3F1 mice were exposed to PCE  (99.9%  pure),  6 hours/

                                      3-15

-------
day, 5 days/week, for 2 years.  The exposure concentrations used were 0, 200,
and 400 ppm for rats and 0, 100, and 200 ppm for mice.  At the 200 and 400 ppm
exposure concentrations, PCE induced |a marginally statistically significant
                                     j
increase of mononuclear cell leukemia in male rats.  PCE also induced an in-
                                     t
creased incidence of renal tubular cell  adenomas or carcinomas (combined) in
male rats.  This incidence shows a do[se-related trend, although it is not
statistically significant.
     In female rats at the 200 and 400 ppm exposure concentrations, a statis-
tically significant increased incidence of mononuclear cell leukemia was ob-
served.  In male and female B6C3F1 mice, PCE induced a statistically signifi-
cant increased incidence of hepatocelflular carcinomas and/or adenomas at both
the 100 and the 200 ppm exposure concentrations.
                                      3-16

-------
        •                 4.  QUANTITATIVE RISK ASSESSMENT

      This section updates the inhalation unit risk estimate for PCE reported in
 the July 1985 Health Assessment Document for Tetrachloroethylene by using data
 from the NTP (1985) inhalation bioassay, which is the first lifetime inhalation
 study to show evidence of the carcinogenicity of PCE in mice and rats.   These
 data provide a basis for directly estimating the risk of PCE by the inhalation
 route, and  an opportunity to test the reasonableness of the July 1985  risk
 estimate, which was developed using the gavage data  from the NCI bioassay
 (1977).'  The inhalation  unit risk estimate  for PCE based on tumor incidence
 data from a gavage  study in  mice  and the assimilated dose metabolized,  was
 accomplished by utilizing metabolism information  for both animals and humans,
 expressing  dose as  the amount  of  PCE metabolized.  On  this  basis the reported
 inhalation  unit  risk  is  4.8  x  10-7/yg/m3.   This estimate was  found  to be
 comparable  to three other estimates  using different  data bases  and/or different
 methods  of  calculation.   The details  of  these  calculations  are  presented  in the
 Health Assessment Document (U.S.  EPA, 1985, pp. 9-53 .to  9-61).
     The  new unit risk estimate for  inhalation also  utilizes metabolized  dose-
 tumor  incidence  relationships derived from exposure dose-metabolism relationship
 data in both animal  and human experiments.  The uncertainty associated with ex-
 trapolating  from animals to humans is reduced by the use of actual human expo-
 sure dose-metabolism relationship estimates, using urinary trichloroacetic acid
 as a measure of metabolism, in addition to the use of comparable animal  data.
     For comparison, we have also used a physiologically-based pharmacokinetic
model to predict the amount of PCE metabolized in the NTP bioassay animals.
4.1.  DATA AVAILABLE FOR RISK CALCULATION
     The NTP (1985)  inhalation bioassay  shows evidence of the carcinogenicity
                                      4-1

-------
 of PCE in rats (mononuclear cell leukemia) and mice (hepatocellular adenomas
 and carcinomas).  The assays for botlh species have been selected  for estimation
 of carcinogenic risk to humans for inhalation exposure.
 4.1.1.  NTP Rat Inhalation Assay    !
      Long-tern (103 weeks) carcinogenicity studies were conducted  by inhalation
 exposure (0, 200,  and 400 ppm, 6 hours/day, 5 days/week)  on  groups  of 50 male
 and 50 female Fischer 344/N rats.  The animals were placed on study at 8 to 9
 weeks of age and  killed at 112 to lib weeks of age.
      Table 3-2 shows the overall  incidence rates  for mononuclear cell  leukemia.
 Although the survival  of male  rats  in the  hiyh-dose  group is significantly
 reduced when compared to the control  group, the difference appears  to be signi-
 ficant  only  at the late stage  (after! 80 weeks)  of  the experiment.   Because the
 survivalship differs only at the  Iat4 stage of the experiment and a  large pro-
                                     [
 portion of leukemias were  found at  the  termination of the bioassay, it is more
 appropriate  to use the incidence  rat6 data,  rather than the time-to-death data,
 to calculate the lifetime  cancer  risk.
 4.1.2.   NTP  House  Inhalation Assay   '
        -              _         .   .
      Lifetime  (103 weeks)  carcinogenicity  studies were conducted by inhalation
 exposure  (0, 100,  and 200  ppm, 6  hourfs/day, 5 days/week) on groups of 50  male
                                     t
 and 50  feniale  B6C3F1 mice.   The animals were placed on  study  at  8 to 9 weeks
                                     I
 and killed at  112  to 113 weeks.   Exposure to PCE at 100 or 200 ppm reduced
 survival of male mice, and at 200 ppm reduced survival  of female mice. Most
 of the deaths  in treated animals  occurred after week 82, and  there  is evidence
that the unscheduled deaths may have been influenced by  the high incidence of
 hepatocellular carcinomas.  However, !for the same  reasons as  discussed for
rats,  the incidence data are considered more appropriate for  risk calculation
than the time-to-death data.  Table 3|-4 shows the  overall  incidence  rates for

                                      4-2

-------
 hepatocellular adenomas and carcinomas/which, combined, show a dose-related
 incidence of hepatocellular neoplasms in both male and female mice.
 4.2.  POSSIBLE MECHANISMS LEADING TO A CARCINOGENIC RESPONSE FOR PCE
      Possible mechanisms proposed for carcinogenesis include direct interaction
 of a chemical  or its metabolites with DNA,  long-term tissue injury, stimulation
 of cell, proliferation, immunosuppression,  hormonal  imbalances,  or release of
 altered cells  from growth control  (Weisburger and  Williams, 1980, 1981).   The
 mechanism(s)  by which PCE causes cancer  is  not known and therefore a matter  of
 some speculation.   The current  knowledge of PCE  metabolism  and  the acute  cell-
 ular toxicity  directly related  to  reactive  metabolites  suggest  different  cell-
 ular processes  that  may possibly lead  to carcinogenic activity.   One process
 may  involve cell death induced  by  the  cellular toxicity  of  PCE  followed by the
 consequent stimulation of DNA replication associated with cell multiplication,
 resulting in an  indirect  or promoting  effect.  The  continuing process of  DNA
 replication results  in increased proportions  of  single-strand DNA, which  is
 more susceptible to  irreversible binding by  reactive intermediates than is
 double-strand DNA.   Alternatively, it  is possible that PCE  acts by stressing
 the  fidelity of  replication, thus  increasing  the possibility of introducing DNA
 transcript errors.   A  general  process proposing that necrosis constututes a
 necessary precursor  for PCE liver  carcinogenicity is consistent with a thresh-
 old  effect at the necrotizing doses at which all  death and turnover occur.  It
 is also consistent with the negative or inconclusive results of the majority  of
mutagenicity assays.
     Peroxisome proliferation  has been linked with  cancer in rodents (Moody and
Reddy, 1978;  Reddy et al., 1980; Prout et al., 1985; Green and Prout, 1985).   •
Trichloroacetic acid (TCA) resulting  from the biotransformation  of PCE has been
shown to induce liver peroxisome proliferation in mice but not in  rats (Green,
                                      4-3

-------
 1986).   It has been hypothesized that the species difference in  the hepatocar-

 cinogenicity of PCE seen between rats and mice is possibly due to a species

 difference in peroxisome proliferation resulting from higher TCA blood  levels
                                      I ,
 in the  mice (Green, 1986).            |

      Increased reactive oxygen  species formed  by the  peroxisomes may  result  in

 DNA damage, or the peroxisome proliferation  may result in  formation of  the low

 amounts of DNA alkylation that  have  been  observed following  exposure  to PCE.

 However, a cause-and-effect  relationship  between liver peroxisome proliferation

 and PCE hepatocarcinogenicity in mice! has not  been  established.   This mechanism

 does not account  for the mononuclear cell  leukemias and  renal  tubular neoplasms

 seen in rats.

      Another general  process  that may[ lead to  carcinogenicity  involves  the
                                      f
 metabolic production of active  PCE metabolites  (tetrachloroethylene oxide) and
                                      I
 their direct interaction with DNA.   This  genotoxic  process could  be expected
                                      1
 to  be modulated by various' homeostatic  mechanisms,  such  as DNA repair and im-

 munological  surveillance, but it  is  generally  regarded as  lacking a threshold.

 This  mechanism is  supported by  the mujbagenicity  of  the PCE-epoxide  and  the fact

 that  DNA binding of  PCE  metabolites  has been observed, even though  only at low

 doses.

      In  the  absence  of  definitive evidence solely supporting any specific one

 of  the  likely  processes  operative in the  carcinogenic activity of PCE, the risk

 assessment performed by  the Carcinogenic Assessment Group considers the process

which is  generally accepted—the genotoxic mechanism—to be associated with the

 greatest  risk.  The  risk of PCE carcinogenicity by this process,  with its lack

 of  threshold,  is appropriately estimated by a mathematical  model  predicting

zero incremental risk only at zero exposure.  In addition,  at the present time,

the lack of sufficient understanding about the mechanisms of action involved

                                      4-4

-------
 in carcinogenesis does not allow for a distinction between chemicals acting
 directly with DNA and those which do not, nor can those which have not been
 shown to be genotoxic be considered to have identifiable population thresh-
 olds or to be safer than those which are considered to be genotoxic (Pereira,
 1984).  More research is required on the mechanism(s)  of action  of PCE before
 a threshold model  would be appropriate for carcinogen  risk assessment  (Pereira,,
 1984).
 4.3.  RELEVANCE  OF METABOLISM  TO QUANTITATIVE RISK ASSESSMENT
      The general  considerations  involved in the  scaling  of toxicologic effects,
 including  carcinogenicity,  among species (including  man)  have been  outlined  for
 PCE  in the Health  Assessment Document  for Tetrachloroethylene (U.S.  EPA, 1985),
 which also presented  information on  the  kinetics,  metabolism,  and covalent
 binding properties of PCE  in rat and mouse  species  and metabolism and  kinetics
 in humans.   This  information may be  summarized as  follows:
      (1) .There  is evidence that the liver  and kidney toxicity and the  carci-
 nogenic  potential  of  PCE and other halogenated ethylenes are  dependent  on the
 metabolic  conversion  of these  compounds  to  reactive epoxides  and other  bio-
 logically: reactive  intermediates  of metabolism (Figure 4-1) (Bolt et a!., 1982;
 U.S.  EPA,  1985, Chap.  5).
      (2)   The reactive metabolites .of PCE bind irreversibly to cellular macro-
 molecules jm vitro  and jm \nvo_ (Costa and Ivanetich, 1980; Pegg et a!., 1979;
 Schumann et al., 1980).  Significant binding to hepatic DNA has not been
 demonstrated to occur in. vjrvo  (Schumann et al., 1980).   Irreversible binding
to other cellular macromolecules  (proteins and lipids)  in vivo occurs indepen-
dently of route of administration (gavage or inhalation)  but proportionally
to the amount metabolized (Pegg et al., 1979; Schumann  et al., 1980).  Indices
of hepatocellular toxicity and  cellular damage have been  shown to correlate     '
                                      4-5

-------
                                            CHLORIDE
                                           MIGRATION
   DECARBOXYLATION
                                                            +H Cl
                                                       TCA
CO
 Figure 4-1.   Postulated scheme for the metabolism of PCE.
 SOURCE:  U.S. EPA, 1985.
                            4-6

-------
 linearly with  metabolism (Buben  and  O'Flaherty,  1985).
      (3)  The  metabolism and  kinetics  of PCE  have been  studied  in  three  species:
 mice,  rats,  and  humans.   In each of  these species there is  evidence  that  metabo-
 lism of PCE  after oral  (mice,  rats)  or inhalation exposure  (mice,  rats,  humans)
 is  rate-limited  and  proceeds  according to Michaelis-Menten  kinetics  (Filser  and
 Bolt,  1979;  Pegg et  al.,  1979; Schumann et al.,  1980; Buben  and O'Flaherty,  '
 1985;  Ohtsuji  et al.,  1983; U.S.  EPA,  1985, Chap.  5).   There is no comparative
 experimental evidence  that the metabolic pathway  for PCE qualitatively differs
 for  the  above-mentioned  species.   The  metabolic  capacity for the metabolism  of
 PCE  by mice  and  rats has  been found  to be proportional  to body surface area  for
 these  two  species  (U.S.  EPA,  1985, Chap.  9).
      (4)   Elimination  of  PCE  following inhalation  exposure  in humans is tri-
 phasic,  with a terminal  first-order  elimination half-time of 55 to 65 hours;
 for  the  rat, the dominant half-time  of elimination is 7 hours after  inhalation
 or oral  exposure.  These  long half-times  indicate  that with  chronic daily
 exposures  in humans or animals (conditions of carcinogenicity bioassays) com-
 plete  elimination  (by metabolic and  pulmonary routes) would  not occur within
 the  24-hour dosing interval and that body  accumulation would occur until a
 plateau of equilibrium is reached between  body concentration and inhalation or
 oral exposure dose.
 4.4.   INHALATION EXPOSURE IN HUMANS
     PCE is readily absorbed through the lungs into the blood by first-order
 passive diffusion processes.   Two major processes account for the known eli-
mination of PCE from the body:  1) pulmonary excretion  of unchanged PCE,  and
2) metabolism of PCE to urinary metabolites.  The metabolism of PCE appears
to be very limited in humans,  as  it is  in experimental  animals.   Only a small
percentage of the estimated amounts absorbed by inhalation  are metabolized to
                                      4-7

-------
 trichloroacetic  acid  (TCA)  and  other  chlorinated metabolites  found  in  the urine
                                      |
 (U.S.  EPA,  1985, Chap.  5).   Saturation  of  metabolism  has  been estimated to

 occur  in  humans  at  relatively low  exposure concentrations  (between  100 and 400

 ppm PCE  in  inspired air).   Estimates  of the extent of PCE  metabolism in humans

 have been made from balance studies by  measuring urinary metabolites after

 accounting  for a retained  inhalation  exposure  dose (U.S.  EPA, 1985, Chap,. 5).
                                      I
 However,  there are  several  difficulties associated with balance studies in

 humans.   For  example, problems  are encountered'in obtaining an accurate measure-
                                      i
 ment of the retained dose of PCE from!inhalation exposure.  In addition, the

 available balance studies  in humans are incomplete.   One shortcoming in some of

 the  older PCE studies has been  the imprecision of the analytical methodologies
                                      !
 used;  for example,  the  Fujiwara reaction for measuring metabolites.  Also,

 urinary sampling may not have been sufficient  considering  the long half-life of

 PCE  metabolites.  In addition,  individual  variations  among subjects can be
                                      i
 relatively high.  Only  urinary  TCA was  used in these  studies as a measure of

 metabolism, and  the possibility exists  that significant amounts of metabolites

 other than TCA are  excreted  in  the urine or by other  routes.  It has been

 proposed  that a  much higher  percentage  of  the  dose is actually metabolized, and

 that a so-far unrecognized  pathway may  exist for PCE  and for other metabolites
                                      i
 which have not yet  been identified,  However, measuring the amount of urinary
                                      I
 TCA  is one approach that may be used to compare the metabolized dose of PCE

 among species.   The metabolized dose may be considered to be the effective dose

 in organ toxicity and carcinogenicity.

 4.5.  CALCULATION OF METABOLIZED DOSE SFROM  INHALATION STUDIES

 4*5.1.  Direct Estimation from Metabolism Experiments

4.5.1.1.  RaJ>-Pegg et al.  (1979)  and Schumann et al. (1980) determined body

burdens of Sprague-Dawley rats and B6q3Fl mice after  6-hour inhalation  expo-

                                      I
                                      4-8

-------
 sures.   The animals  were  exposed  to  14C-PCE  and  the  amount  metabolized  as  a
 fraction of the  body burden  was estimated  from  radioactivity  in  urine,  feces,
 expired  air, etc., and  unchanged  14C-PCE  in  exhaled  air.  Their  results for
 exposed  rats are given  in Table 4-1.
      Total  recovery  of  14C-radioactivity from 6-hour inhalation  of  10 ppm  an,d
 600  ppm  l^C-PCE  1n rats was  nearly directly  proportional to exposure concen-   -
 tration;   However, the  absolute amount metabolized (10  ppm, 0.47 mg-equiva-
 lents; 600  ppm,  9.11 mg-equivalents  PCE) was not  proportional to exposure  con-
 centration  and decreased  both  in  absolute  proportions and as  a percentage  of the
 total assimilated dose  (10 ppm, 32%;  600 ppm, 12%).  Thus,  these data from rats
 confirm  in  this  species limited and  saturable metabolism of PCE consistent with
 dose-dependent Michael1s-Menten kinetics,  as has  been indicated by  kinetic in-
 halation  studies of  Filser and Bolt  (1979).  These data are also in accord with
 the  demonstration by Buben and O'Flaherty  (1985)  of  dose-dependent  Michaelis-
 Menten kinetics  for  PCE in mice.
     The  relationship between the rate of  PCE metabolism in adult male  Sprague-
 Dawley rats  (or  the  amount of PCE metabolized per unit body weight  per  6-hour
 exposure, M) and exposure concentration dose, d (ppm), can  be expressed by a
 Michaelis-Menten-like equation

                                M = V x d/(K + d)

where V and K are constants to be estimated from the  data.   In order to utilize
the human data (metabolites in urine) for interspecies extrapolation,  the amount
of urinary metabolites in animals  is  used to establish the  dose-metabolism rela-
tionship.  Using the  data of urinary  metabolites for  rats  and  assuming  Fischer
344 rats  metabolize  PCE  qualitatively similar to Sprague-Dawley  rats with inha-
                                      4-9

-------
                  TABLE  4-1.   RECOVERY  0F  14C-PCE  RADIOACTIVITY
                      AFTER  INHALATION  EXPOSURE  FOR  6  HOURS
                     TO  SPRAGUE-DAWLEY  RATS AND  B6C3F1 MICE*
                              Rats  (avl  of  3)a
                         10  ppm         |      600 ppm
                             (mg-equivalents  PCEb)
Mice (av. of 3)

    10 ppm
Expired unchanged
Metabolized
14co2
Urine
Feces
Carcass

Total
1.008 (68%)C
1

0.053
0.275 ;
0.076
0.063
0.467 (32%)c i
1.475
68.39 (88%)c

0.54
4.54
2.36
1.67
9.11 (12%)C
77.50
0.048 (12%)c

0.032
0.285
0.027
0.012
0.356 (88%)c
0.404
aAverage experimental animal weights (drams):  250, rat; 24.5, mouse.  The
 animals were exposed to 10 or 600 ppm ^C-PCE for 6 hours and maintained in
 metabolism cages for 72 hours after exposure.
"Each value represents the mean of data from three animals.
cNumbers in parenthesis indicate the fraction of total  administered dose.
                                       i
SOURCES:  Adapted from Pegg et al., 197;9 and Schumann et al.,  1980.
                                      4U10

-------
  latloh exposure (Table 4-1), the dose-metabolism relationship is determined to
  be
                          M (mg/kg)  = V x d (ppm)/(K + d)

  with  V and  K estimated to be 24.64 mg/kg and  213.96  ppm,  respectively.
  This  relationship  is  used to estimate the amount  of  PCE metabolized  for  a  6-
  hour  exposure  at 200  or 400  ppm  in  the  NTP  (1985)  rat bloassay.   The amounts
  metabolized  (urinary  excretion)  are  11.90 mg/kg/day  and 16.05 mg/kg/day  for
  6-hour  exposures,  corresponding, respectively, to 200 and 400 ppm ambient  air
  concentrations.  It may be noted that doubling the daily inhalation exposure
  concentration  from 200  to 400 ppm results in only a  30% increase in the amount
  of PCE metabolized.  The two doses represent approximately 48% and 65% of
  total  metabolic capacity.  These observations may, in part,  explain the fail-
 ure to observe an exposure dose-leukemia response relationship for this rat
 bioassay.
 4.5.1,2.  Mouse-The  dose-dependent Michaelis-Menten  kinetics  for the metabol-
 ism of PCE by mice  after oral administration have been amply demonstrated by
 Buben  and O'Flaherty  (1985).   For inhalation exposure, however,  few data  are
 available pertinent to the conditions of the NTP  (1985) bioassay,  from  which to
 derive a comparative relationship between  exposure concentration  and  metabol-
 ism, as  was done  in the  Buben and O'Flaherty (1985) oral study.
     Watanabe and associates  (Pegg et  al., 1979; Schumann et al.,  1980) assess-
 ed  the  amount of PCE metabolism accrued  from a 10 ppm, 6-hour exposure to
 B6C3F1 mice (Table 4-1).  The metabolism of  PCE in mice was not determined  at
 other exposure  concentrations.  The exposure levels, of the NTP inhalation bio-
assay in mice were 10- and 20-fold higher.  At the 10 ppm exposure level, the
absolute amount of metabolites in urinary excretion is comparable between the
                                      4-11

-------
two species; however, expressed in mg/^kg, the amount of PCE metabolites is



about 10 times greater for mice than for rats (11.6 mg/kg vs. 1.1 mg/kg).  This



ratio should not be expected to hold at higher concentrations because at higher



concentrations the metabolic rate is cjominated by Vmax in the Michaelis-Menten



equation.  Theoretically, the ratio of Vmax between two species equals unity if

                                      i

Vmax is expressed in terms of metabolic rate as represented by mg/W2' , or, if



Vmax is expressed in terms of mg/kg, is equal to (W1/W2)1/3 where W's are body

                                      i

weights for animals of each species.  Therefore, at high doses one might expect



to see that mice metabolize about twice the amount of PCE, in terms pf mg/kg,
                                      i


as do rats [i.e., (0.25/0.0245)1/3 = 2.2].  This ratio is supported by the



observation from oral studies (Table 4-2) that at 500 mg/kg, the ratio of the
                                      i


absolute amount of urinary PCE metabolites between mice and rats is 0.267,



or 2.7 on a mg/kg basis  [i.e., (1.53/0.0245)7(5.72/0.25) = 2.7].  The amount



metabolized of a 500 mg/kg oral dose in rats is approximately equal to the



amount metabolized of a  600 ppm inhalation exposure in this species.  To calcu-

                                      i

late the metabolized dose for the NTPjmice, it is assumed that mice metabolize



approximately 5 times  (in terms of mg/kg) more than rats at the cancer bioassay



doses of 100 and 200 ppm on the basis[that mice metabolize greater amounts  (in
                                      I


terms of mg/kg) than rats, ranging from 10 times greater at 10 ppm to 2 to  3
                                      i


times greater at higher  doses.  There-Tore, the metabolized doses of  PCE for



mice, based on urinary metabolites, corresponding to 100 and 200 ppm  6-hour



exposures are, respectively,  39.24 mg/kg/day and 59.50 mg/kg/day—or  approxi-

                                      I

mately five times the  values  calculated from the dose-metabolism relationship



for  rats, when expressed in terms of mg/kg.
                                      j
                                      i

4.5.2.   Metabolized  Dose Predicted by a PB-PK Model
      In addition to the (urinary)  metabolized  doses  that  are  calculated direct-
                                      I


 ly from the empirical  data and are prjesented  in  the  previous  sections, we  have




                                      : 4-12

-------
CD
I—
00
UJ
OO
o
Q

— 1
- -
O LU
' — i __]
Q 3
— (
1—
i — i
OO
o
Q_
CO
i — i
Q
•
CM
1
LU
— 1
CQ
[
















* — •
OO

**-
o

re
cu
o
•i —
s:






^— .
i —
re
E

-5^ C
-•-. re
E "cr
O
O LO
LO CM

CM
r— 1
LU
O
Q-

(/) ,
4-> re
c: E
CU -r-
re re
•r- £_
=3 0)
cr ex
cu
i
cn
E
















re
OO

4-
0
.*
re

in
4J
re
a;




, — *
re
CD E
~-» c:
cn re
E "^^
CD
LO LO
CM
I— 1



LU
O
Q_
to re
4-> E
O! C
i — re
re
> £_
•i- O)
Z3 CX
cr
o>
i
CO
E







1
re
CD.r-
~-s. re
E cn
t-H
LO
CM
O
^"-" **
LU
o
D_
1/1 i —
-t-> re
 «
OO O «— I O O r— t O
t 1












., — ^
S-S. ^^p
£ 2
f~~- r^ CM CM ^H CM CT>
'~O LO r-. CO •=*• LO rH
* * • ' • • • •
O CD LO ^}- rH CM OO
<— < r-H CM
' 1— — 1











s? T
>— ' cVi
r~- CM
^" r-^ O LO co CD •=*•
f~ O "=3- I-H CD r-~ «d-
1—1 O O O O O CM
CD CD O CD O O O






•O CU co
O) N (^
cn -r- CM cu t/i re
•o cr i — CD c: cu o
cu re O o •!- o c_
£ — c_c<3- c_cure i —
•i— o re r-n z3 LL. o re
ex cr -)-> i>
x 3 _cy o
l_LJ ^i_ ^_

















.
Cl)
to
o
E
„
LO
• •
•=*• o
CM CO
cn
4->
re «
^~ *
i —
o
LO 4->
CM CU

to re
E E
re o
£— -C
cn o
	 	 CO

•)-> "O
cn re
cu cn
5 r-
cn
're
E •>
•r- •
re re
experimental
Tom Pegg et
vt/ *t^
cu -o
CD CU
re 4^
£- CL
cu re
> T3
re •<
p_
o
OO
T3 LU
CU O

re rs
CQ CD
re co
4-13

-------
also calculated the amount metabolized by using the physiologically-based phar-

macokinetic (PB-PK) model.  An explanation of the construction of the model  is

given in Appendix A.  The PB-PK model I predicts only the total amount metabo-

lized, not the urinary metabolites as j previously calculated^.  The metabolized
                                      I
doses predicted by the PB-PK model are comparable to those calculated directly
                                      i
from the empirical data (Table 4-3), considering the fact that urinary metabo-

lites account for about 50% to 80% of the total metabolites and the fact that

the metabolized doses for mice are slightly underestimated (approximately three

times), by the PB-PK model, as discussed in section A.3.2.2. of Appendix A.
  _
4."6.  HUMAN METABOLIZED DOSE AT LOW-LtVEL AMBIENT EXPOSURE
                                      i
     Estimation of the extent of metabolism in humans have been made from sev-

eral balance studies by accounting for a retained dose after inhalation expo-

sure through measuring metabolites excreted in the urine.  The difficulties

associated with these balance studies Jin humans have already been mentioned  in

section 4.3.  However, urinary TCA measurements can reasonably be used to make

species comparisons of metabolized dose.  The data in humans are represented

here by studies selected on the basis of the lower exposure levels more com-

mensurate with anticipated environmental concentrations.

     A study by Bolanowska and Golacka (1972) provides some information on the

relationship between the air concentration of PCE and the amount metabolized in

urinary excretion.  In this study, five subjects were exposed to 390,000 pg/m3

of PCE (approximately 50 ppm) for 6 hours.  The excretion rate of metabolites

in urine (mg/hr) is presented graphically over' the 20-hour period during and

after exposure.  The total amount of ryietabolites in urine (area under curve)

is estimated to be about 13 mg.  The area under the curve from 20 hours to
                                      i
infinity is calculated by C x ti/2/0.693, where C is the concentration at the
                                      t
last sampling time (i.e., 20 hours) and ti/2, the half-life, is assumed to be


                                      4-14

-------
             TABLE 4-3.  COMPARATIVE ESTIMATES OF METABOLIZED DOSES
                      IN THE NTP  (1985) INHALATION BIOASSAY
                                                                  Total
                                                                metabolites
                   Exposure         Direct estimate of          estimated by
                concentration       urinary metabolites         PB-PK model
SPecies             (PPm)               (mg/kg)                   (mg/kg)*
Rats

Mi ce

200
400
100
200
11.90
16.05
39.24
59.50
19.39
28.57
19.77b
30.00b
 The PB-PK model  can predict only the total amount of metabolites.  The values
 are total metabolized dose over 72 hours after a single 6-hour exposure.
 These values are presented here for the purpose of comparison only: they are
 not used for risk calculation.                                         J
DThe metabolized  doses for mice are slightly underestimated (approximately 3
 times) due to the use of under-valued physiological  parameters, in particular
 the alveolar ventilation rate (see section A.3.2.2.  in Appendix A).
                                     4-15

-------
 100 hours.   Under the assumption that the amount metabolized is linearly pro-

 portional to the air concentration below this concentration, the amount meta-

 bolized  (in  urine) associated with 1 pg/m3 of PCE in air is 13 mg/390,000 =

 3.33 x 10~5  mg  for a 6-hour exposures.

     Another study that supports the above estimate is that of Fernandez et al,

 (1976) in which humans were exposed! to PCE vapor at 150 ppm for 8 hours.  The
                                   i

 quantity of  PCE metabolized and eliminated in urine over 72 hours was about

 25 mg.  This is clearly less than the total amount metabolized from the 8-hour
                                   I
 exposure because of the long half-ljife in the final phase of urine elimination.

 As an approximation, the (urinary) metabolized dose associated with 1 pg/m3

 of PCE in air is calculated to be
                                   i

     1 ug/m3 x (25 mg/150 ppm) x (6 hr/8 hr)/6908/ug/m3/ppm = 1.8 x 10-5
mg
for a 6-hour exposure, given the fatt that 1 ppm = 6908 yg/m3.  This value is

less than the value of 3.33 x 10~5 mg calculated on the basis of Bolanowska and

Golacka (1972).  The value 3.33 x lO'5 mg/6 hours will be used for risk calcu-

lation because it is obtained from a lower concentration (50 ppm), where the

assumption that the dose-metabolism:relationship is linear is not unreasonable.

4.7.  CALCULATION OF UNIT RISK
                                   i
4:7.1.  Based on the Metabolized Dose Estimated Directly from the Empirical
        Data

     In this section, the incremental lifetime cancer risk over background due

to 1 yg/m3 of PCE in ambient air (iie., unit risk) is calculated.   The calcu-
                                   i

lation procedure, the interpretation, and the appropriate use of the concept  of

unit risk were presented in the Health Assessment Document for PCE (U.S.  EPA,

1985).  The data used for the risk calculation are given in Tables 4-4 and 4-5.


                                      4-16

-------
                        TABLE-4-4.   NTP METABOLIZED DOSE
                   AND  INCIDENCE  OF  MONONUCLEAR CELL LEUKEMIA
                               IN  FISCHER 344 RATS

Group
Males


Exposure
concentration
(ppm)
0
200
400
Human equivalent
metabolized dose
(mg/W2/3/day)a
0
6.26
8.45

Incidence
rate
28/50
37/50
37/50
Females            0                 0                         18/50

                 200                 5.81                      30/50

                 400                 7.84                      29/50
aMetabolized dose = M(mg/kg/day) x (5/7) x wl/3, where M(mg/kg/day)  is the
 direct.estimate of the dose metabolized taken from Table 4-3-  the factor 5/7
 reflects that animals were exposed only 5 days/week;  and W = 0.40 kg fon male
 rats and 0.32 kg for female rats.  For convenience, the notation  mg/W2/3 is
 used to indicate the amount metabolized in terms of mg per two-thirds of body
 weight, which is assumed to be proportional  to the body surface area.
                                     4-17

-------
                        TABLE 4-5.  NTP METABOLIZED DOSE
                     AND INCIDENCE OF'LIVER TUMORS IN MICE
Group
  Exposure
concentration
   (ppm)
Males
                0

              100

              200
    Human    i
  equivalent I
  metabolized
    dose
(mg/W2/3/day)a
                    0

                    9.37

                   14.21
                                                        Incidence rate
Carcinomas   Adenomas/carcinomas
                       7/49b

                      25/47

                      26/50
                      16/49

                      31/47

                      40/50
Females
0
100
200
0 ;
8.92
13.52
1/46
13/42
36/47
4/46
17/42
38/47
Metabolized dose = M(mg/kg/day) x (5/7) x W1/3, where M(mg/kg/day) is the
 direct estimate of metabolized dose taken from Table 4-3; 5/7 is a factor
 to adjust for 5 days exposure in a we'ek; and W. = 0.0374 kg for male mice
 and 0.0322 kg for female mice.       \
bOnly animals surviving beyond 60 week's are included in the denominator.  The
 first death from liver tumor occurred at 60 weeks.
                                      4-18

-------
 The NTP bioassay doses for rats  and  mice  have  been  converted  into  metabolized
 doses using the dose-metabolism  relationship  previously  established.   The
 metabolized dose, expressed in mg/W2/3/day, where W is the  body weight  in  kg,
 is  assumed to  be "equivalent"  (equally  potent)  among  species.  The carcinogenic
 potencies  (slopes)  of PCE  are  calculated  using  the  multistage model on  the
 basis of "equivalent" doses and  their corresponding tumor incidence rates.   The
 slope Is defined as the 95% upper  confidence limit  of the linear parameter  in a
 multistage model.  These slope estimates  are presented in Table 4-6, and have
 been  used  for  calculating  the  unit risk of PCE.  It should  be noted that these
 slope values are expressed  in terms  of  "equivalent" doses.  These  values are
 only  used  in the intermediate  steps  of  the derivation of the risk  to humans due
 to  1  ug/m3 of  PCE in  air.   The slope estimates calculated on the basis  of six
 different  data  sets are comparable,  ranging from 3.64 x 10-2/(mg/w2/3/day) to
 1.21  x  10-1/(mg/W2/3/day).   It is possible to express slopes in terms of ppm
 ambient  air concentrations  by using the dose-metabolism relationship previously
 determined.  Since  such a conversion does not affect our derivation of  risk to
 humans,  it suffices to point out that slope values  per ppm for mice are greater
 than  those for  rats because the  rate of metabolism  is greater in mice than in
 rats.
      To  calculate the human risk due to 1 yg/m3 of PCE in air, it is neces-
 sary  to  estimate  the corresponding amount metabolized in  humans.   As previously
estimated, the amount metabolized is  3.33 x 10'5 mg  when  an  individual  is ex-
posed to 1 yg/m3  of PCE in  air for 6  hours.   As an  approximation,  the  amount
metabolized corresponding to a continuous  (24  hours) exposure  to  1  yg/m3 of
PCE in air is 3.33 x 10-5 mg x (24 hours/6 hours)  =  1.33  x  1(H.ing/day,  or  7.83
x lO-6 mg/W2/3/day for a 70-kg person.
     An upper-bound estimate of the incremental  risk due  to  1  yg/m3 of  PCE  .
                                      4-19

-------
(i.e., unit risk) is calculated as 7.83 x 10~6 x slope.  The risk estimates

calculated on the basis of six different data sets (Table 4-6) are comparable,

ranging from 2.85 x 10"? to 9.47 x 10["7, with a geometric mean equal  to 5.78

x 1Q~7.  All of these values are comparable to our previous unit risk estimate
                                     l
of 4.8 x 10~7, which is based on hepajtocellular carcinomas in mice in the NCI

(1977) gavage study.

4.7.2.  Based on the Metabolized Dosej Calculated from the PB-PK Model
                                     i
     For comparison, the inhalation unit risk for PCE is also calculated on the

basis of metabolized dose derived frorn the PB-PK model, which has gained in-
                                     i
creasingly broad acceptance over the [last few years as an effective tool for the

study of the absorption, distribution, biotransformation, and elimination of

chemicals in the body.  However, due jto the lack of data, some of the parameters

that are used in the modeling (including the Michaelis-Menten kinetic parameters

Vm and Km) must be estimated from thej same data that were used in the approach

presented in the previous section.  For this reason, the model so derived is

not necessarily more reliable than th'at approach to predict the amount metabo-

lized in the NTP bioassay.  The metabblized dose (mg/W2/3/day) in the NTP study,

and the corresponding tumor incidence! rate used in the risk calculation, are
                                     i
presented in Tables 4-7 and 4-8.  Tab^le 4-9 presents the resultant unit risk

estimates calculated from the six difjferent data sets.  The risk calculated by
                                     I
this approach is also comparable to those calculated previously (Table 4-6),

considering the fact that the risk calculated on the basis of mouse data is
                                     I

overestimated due to the underestimation of the metabolized dose by the PB-PK

model (see section A.3.2.2.  in Appendix A), in particular, when the risk

estimates (in terms of mg/W2/3) in the second columns of Tables 4-6 and 4-9

are compared.  The risk estimates in the second columns are calculated directly

from animal data, and thus do not reflect the uncertainty associated with the
                                      4-20

-------
                         TABLE  4-6.   UNIT  RISK ESTIMATES
                 CALCULATED ON  THE  BASIS OF  DIFFERENT DATA SETS
 Data  set
   Risk due to
one unit (mg/W2/3)a
   of urinary
metabolized dose
Risk due to
 1 pg/m3
of PCE in air
Rat  (leukemia):

     Males

     Females

Mouse  (liver carcinomas):

     Males

     Females

Mouse  (liver adenomas/

      carcinomas):

    Males

    Females

Geometric mean:
    1.21 x 10-1

    1.01 x 10-1



    6.67 x 10-2

    3.64 x ID'2
    1.08  x  10-1

    5.05  x  ID"2

    7.38  x  10-2
 9.47 x 10-7

 7.91 x 10-7



 5.22 x 10-7

 2.85 x 10-7
 8.46  x  10-7

 3.95  x  10-7

 5.78  x  10'-7
aThe metabolized dose expressed in mg/w2/3 (wnere w is body weignt in kq) is
 assumed to be equivalent, i.e., equally potent, among species.
                                     4-21

-------
         TABLE 4-7.  NTP RAT METABOLIZED DOSE PREDICTED BY PB-PK MODEL
         AND THE CORRESPONDING MONONUCLEAR CELL LEUKEMIA INCIDENCE RATE
                                           Human
                  Exposure            j  equivalent
               concentration         metabolized dose
Sex                (ppm)              !(mg/W2/3/day)a       Incidence rate
Males                                 i
                      0               |    0                   28/50

                    200               i   10.95     .           37/50
                                      i
                    400                  13.68                37/50



Females
                      0                   0                   18/50
                                      I
                    200                  10.16                30/50

                    400               1   12.70                29/50
aMetabo!ized dose (mg/W^/^/day) = d(mg/kg/day) x W^/3, where d - 14.86 mg/kg/
 day and 18.57 mg/kg/day, respectively, for the 200 ppm and 400 ppm groups.
 The body weight W is 0.4 kg for male [rats and 0.32 kg for female rats.  For
 convenience, d is derived on the basis of a 0.35 kg, rat, rather than calcula-
 ted separately for 0..4-kg and 0.32-kg rats.
                                      ,4-22

-------
        TABLE 41L xTP M°USE METABOLIZED DOSE PREDICTED BY PB-PK MODEL
                AND THE CORRESPONDING LIVER TUMOR INCIDENCE RATE
SeX
Males
Females
            Exposure

         concentration

            (ppm)
             100


             200






               0


             100


             200
                              Human

                           equivalent

                           metabolized
                                                     Incidence rate
                                                         ~
ion dose
(mg/W2/3/day)a
0
4.49
6.55
0
4.27
6.23
Carcinomas
7/49
25/47
26/50
1/46
13/42
36/47
Carcinomas/adenomas
16/49
31/47
40/50
4/46
17/42
38/47
                                     4-23

-------
   TABLE 4-9.  UNIT RISK ESTIMATES BASED ON PB-PK PREDICTED METABOLIZED DOSE
Data base
   Risk due to
one unit (mg/W2/3)
   of total
 metabolites
                                                        Risk to humans
                                                        due to 1 yg/rn3
                                                         of PCE in air
Method Ia
Method IIb
Rat (leukemia):

    Males                      7.32 x 10--2

    Females                    6.13 x 10~2

Mouse (liver carcinomas)0:

    Males                      1.42 x 10-1

    Females                    7.74 x 10~2

Mouse (liver adenomas/
      carcinomas):

    Males                      2.30 ^ 10'1
                                     j
    Females                    1.05 % 10"1
                                    . i
 Geometric mean:               1.03 x 10-1
                        3.40 x 10-6

                        2.85 x 10-6




                        6.60 x 10-6

                        3.60 x 10-6





                        1.06 x ID'5

                        4.88 x 10-6

                        4.77 x ID"6
                1.15 x 10-6

                9.62 x 10-7



                2.23 x 10-6

                1.21 x 10-6
                3.60 x lO-6

                1.65 x 10-6

                1.61 x 10-6
aMethod I uses the human metabolized^dose actually predicted by the PB-PK
 model.  For a 70-kg person, the metabolized dose (at steady-state) corre-
 sponding to 1 ug/m3 of PCE in air is 7.89 x 10~4 mg or 4.65 x .10-5 mg/W2/3.
 The risk is calculated by multiplying 4.65 x lO'5 to the values in the second
 column.                             i
bMethod II uses the metabolized dose [calculated from the direct observation of
 humans by Bolanowska and Golacka (1972).  As discussed in section 4.6.1., the
 metabolites in the urinary excretion were estimated to be 7.83 x 10~6 mg/w2/3
 for a 70-kg person exposed to 1 ug/m3 of PCE in air.  If the urinary excre-
 tion is assumed to account for 50% of the total metabolites, the metabolized
 dose corresponding to 1 ug/m3 of PCE in air would be 1.57 x 10-5 mg/w2/3.
 This value is multiplied in the second column to obtain the risk.  The assump-
 tion about the percentage of urinary! metabolites in the total  metabolites is
 based on the observation in rat and mouse studies where urinary metabolites
 account for more than 50% of total  metabolites.
cThe unit risks calculated on the basis of mouse data are somewhat over esti-
 mated because the metabolized dose was underestimated (see section A.3.2.2.
 in Appendix A for explanation).
                                      4-24

-------
 dose-metabolism relationship for humans.
 4.8.  DISCUSSION
      In calculating the dose-response relationship for PCE,  the amount  metabo-
 lized is considered to be an effective dose.   The use of this  surrogate effec-
 tive dose may not eliminate the uncertainty associated with  the low-dose ex-  :
 trapolation because the dose actually reaching the receptor  sites  may not be
 linearly proportional  to the total  amount  metabolized, and the  shape of the
 dose-response curve is still  unknown.   However,  it seems reasonable to  expect
 that the uncertainty with regard  to the low-dose  extrapolation  would be some-
 what reduced by the use of the  metabolized dose  because  the  metabolized dose
 better  reflects the dose-response  relationship,  particularly in  the high-dose
 region.   To extrapolate from animals  to humans,  the metabolized  dose per  body
 surface  area  (which is  proportional to W2/3) is  assumed  to be equivalent  (i.e.,
 equally  potent)  among  species.  This  assumption appears  to be reasonable,  since
 it  is partially supported  by  the fact that  the carcinogenic  potency (i.e.,
 slope) values  are comparable  between  rats  and mice when  these values are  ex-
 pressed  in  terms of mg/W2/3.  it" is interesting to note  that the use of meta-
 bolic and kinetic information, when available, improves  the  accuracy of risk
 estimation, as  evidenced by the fact that the PCE  inhalation unit risk,  calcu-
                                                                 /
 ulated previously on the basis of gavage data, is shown to be a reasonable
 estimate when the newly released inhalation bioassay data are used.
     The CAG has also employed the PB-PK model  to calculate the amount  metabol-
 ized in the NTP bioassay.  This modeling technique provides insight into vari-
ous assumptions used in the risk assessment.  The most serious  limitation for
the PB-PK modeling  is that the metabolism data reported in the  literature do
not represent periodic  sampling over an interval  of time.  A detailed discussion
dealing with PB-PK modeling is presented in Appendix A.  The  risk calculated by
                                      4-25

-------
this approach is also comparable to the previous risk estimates, considering


the fact that the risk calculated on the basis of mouse data is overestimated
                                     f


due to the underestimation of the metabolized dose by the PB-PK model  (see


section A.3.2.2. in Appendix A); in particular, when one compares the risk

                                     i

estimates (expressed in terms of mg/W2/3) in the second columns of Tables 4-6
                                     i

and 4-9, which are calculated from the animal data.


4.9.  SUMMARY OF QUANTITATIVE ASSESSMENT
                                     L

     This section updates the inhalation unit risk estimate for PCE, using data
                                     i

from the NTP (1985, draft) bioassay, which is the first lifetime inhalation


study to show clear evidence of the carcinogenicity of PCE in both rats and


mice.  These data provide a basis foridirectly estimating the risk of PCE by
                                     i

the inhalation route, and an opportunity to test the reasonableness of the

                                     [

earlier risk estimate which was calculated using tumor incidence data from the


NCI gavage cancer bioassay (1977) and the assimilated dose metabolized.  The


inhalation unit risk estimate reported in the 1985 Health Assessment Document


for Tetrachloroethylene was 4.8 x 10~7/ug/m3.



     The new unit risk estimate for inhalation also utilizes the metabolized
                                     i

dose-tumor incidence relationships derived from pertinent exposure concentra-
                                     I

tion dose-metabolism relationship data in both animal and human experiments.
                                     t
                                     t

The uncertainty associated with extrapolating from animals to humans is reduced


by the consideration of actual human exposure concentration dose-metabolism re-


lationship data where urinary TCA is taken as an index of the amount of PCE


metabolized, in addition to the use of comparable animal  data.  It is of con-


siderable interest to note that the air unit risk estimates calculated on the



basis of animal inhalation carcinogenicity studies (rats and mice) are compara-


ble to estimates derived previously from the oral gavage cancer data in mice,


when a metabolized dose is used for the dose-tumor incidence relationships.




                                     ^4-26

-------
 Furthermore, the  risk per unit of metabolite is found to be comparable between
 rats  and mice  even though the risk for mice corresponding to a unit of ambient
 air concentration is higher than that for rats.  The implication is that the
 cancer risk might be overestimated if it were assumed that PCE in the ambient
 air is equally potent among species.  The upper-bound estimates of the incre-
 mental lifetime cancer risk due to lifetime exposure to 1 yg/m3 of PCE in air
 are calculated by three methods for six sets of dose-tumor incidence data.  The
 range of the six risk estimates under each of the three methods are as follows:
 (a) 2.9 x ID'7 to 9.5 x 10-7/yg/m3; (b) 2.9 x IQ~6 to 1.1 x 10-5/pg/m3; and
 (c) 9.6 x ID'7 to 3.6 x 10-6/yg/m3.
     The first range of estimates is calculated on the basis of metabolized
 doses that are derived directly from the single-exposure metabolism study,
 assuming that the total  amount metabolized over the 72-hour period from the
 single exposure is comparable to the steady-state metabolized dose under the
 NTP multiple-exposure pattern.  The second and third ranges of estimates are'
 calculated on the basis of metabolized doses that are derived from the physi-
 ologically-based pharmacokinetic (PB-PK) model.  Such models, which piece
 together the relevant physiological  and biochemical  information,  can be useful
 tools for predicting absorption, distribution, biotransfonnation,  and elimina-
 tion of chemicals across species when  sufficient data are available.   However,
 until  more data become available,  the  CAG recommends that the first range of
 of estimates be used to represent  the  inhalation risk for PCE,  since there are
 uncertainties with parameters  used in  the PB-PK model  that  were estimated from
 the same empirical data  set  used in  the direct estimation,  and  the model  cannot
be considered to give a  better estimate.   That is,  the upper-bound estimate of
the incremental cancer risk  due  to lifetime  exposure of 1 yg/m3 of PCE in air
 ranges from 2.9 x 10'7 to  9.5  x  lO"7.   This  range includes  the  previous  estimate
                                      4-27

-------
of 4.8 x 10~7/yg/m3 derived from the NCI (1977) gavage data.  Thus, the
inhalation risk estimated from gavagejcarcinogenicity data is supported by the
inhalation risk developed from the use of inhalation carcinogenicity data.
                                      4-2,8

-------
                                  5,  REFERENCES
 Bolanowska, W . ;  Golacka, J.  (1972)  Absorption and elimination of tetra-

                 ?ne 'r hUmaPS Under exP^imental  conditions.  Medycyna Pracy
               -119.  English translation.
 Bolt, H.M.; laib, R.J.; Filser, J.G.  (1982)  Reactive metabolites and car-

      cinogemcity of halogenated ethylenes.   Pharmacology 31(l):l-4.



 Buben, JR.;  O'FJaherty,  E.J.   (1985)   Delineation of the role of metabolism

      in the hepatotoxicity of  trichloroethylene and perch! oroethylene: a

      dose-effect study.  Toxicol.  Appl .  Pharmacol . 78:105-122.



 Costa  A.K.;  lyanetich   K.M.   (1980)  Tetrachloroethylene metabolism by the

      hepatic  microsomal  cytochrome P-450 system.   Biochem.  Pharmacol. 20:2863
 Fernandez   T.;  Guberman   E.;  Caperos,  J.   (1976)   Experimental  human  exposures


     Am.^^yg0:^1.6^  (K^lsT '^  '"  ^  ^  ^Mm"
                                                    of  halogenated  ethylenes  1n

Green  T; Prout  M.S.   (1985)   Species differences  in  response  to trichloro-
Green, T.    (1986)   To be published.



Moody, D.C.; Reddy  O.K.  (1978)  Hepatic peroxisome  (microbody)  proliferation

     45:497-504. plasticizers and related compounds.  Toxicol. Appl. Pharmacol.




National  Cancer Institute (NCI).  (1977)  Bioassay of tetrachloroethylene for

     possible carcinogemcity.  U.S. Department of Health, Education, and

     Welfare Publication No. (NIH) 77-813.



National  Toxicology Program (NTP).  (1985, Aug.)  NTP technical report on the

     toxicology and carcinogenesis studies of tetrachloroethylene in F344/N

     rats and B6C3F1 mice (inhalation studies).  Draft.



Ohtsuji,  T.; Sato,  K.;  Koijumi,  A.;  Kumai, M.; Ikeda, M.  (1983)  Limited
     r>^nar>i't-\//-\-Fl-»f«m-»*>^.4-«.	j__i_~i_?    j_j_   11     ...
                                   ze tetrachloroethylene.  Int. Arch. Occup.
     ttrhiMiif-H-; Watanabe' P-G-  (1979)  Disposition of
     tetrachloro(iltC)ethylene following oral  and inhalation exposure in rats.

     Toxicol. Appl. Pharmacol. 51:465-474.
                                      5-1

-------
Pereira, M.  (1984)  The genotoxic/epijgenetic distinction: relevance to cancer
     policy.  Env. Res. 34:175-191.   !

Prout, M.S.; Provan, W.M.; Green, T.  i(1985)  Species differences in response
     to trichloroethylene.  I.  Pharmacokinetics in rats and mice.  Toxicol.
     Appl. Pharmacol. 79:389-400.

Ramsey, J.C.; Anderson, M.E.  (1984)  JA physiologically based description of
     the inhalation pharmacokinetics of styrene in rats and humans.  Toxicol.
     Appl. Pharmacol. 73:159-175.     !
Reddy, O.K.; Azarnoff, D.L.; Higuita,
C.E.  (1980)  Hypolipidemic hepatic
     peroxisome proliferation form a novel class of chemical carcinogens.
     Nature 283:397-398.

Schumann, A.M.; Quast, T.F.; Watanabe,1 P.G.  (1980)  The pharmacokinetics of
     perch!oroethylene in mice and rats as related to oncogenicity.  Toxicol.
     Appl. Pharmacol. 55:207-219.     !

U.S. Environmental Protection Agency, j (1984, Nov. 23)  Proposed guidelines for
     carcinogen risk assessment.  Federal Register 49(227):46294-46301.
                                      I
U.S. Environmental Protection Agency. ; (1985, July)  Health assessment document
     for tetrachloroethylene.  Final report.  EPA-600/8-82-005F.

Weisburger, J.H.; Williams, G.M.  (1980)  Chemical carcinogens.  In:  Toxicol-
     ogy: the basic science of poisons (Doull, J.; Klaassen, C.D.; Amdur, M.D.,
     eds.).  2nd ed.  New York, NY: Macmillan, pp. 84-138.
                                      i
Weisburger, O.H.; Williams, G.M.  (1981)  Carcinogen testing.  Current problems
     and new approaches.  Science 214:'401-407.

-------
                                   APPENDIX A

                  PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS

        AND THEIR APPLICATION IN THE QUANTITATIVE RISK ASSESSMENT OF PCE*



     Physiologically-based pharmacokinetic (PB-PK) models have gained increas-

ingly broad acceptance in recent years as effective tools for studying the

absorption, distribution, biotransformation, and elimination of chemicals in

the body.  The most desirable characteristic of these models is their ability

to piece together all of the relevant physiological and biochemical  information

to formulate a tool  for predicting absorption, distribution, biotransformation,

and elimination across species.   In calculating the unit risk estimate for PCE,

the PB-PK approach is used to supplement the risk calculations that  are done on

the basis of more directly estimated metabolized dose.   The PB-PK model to be

developed here is only applicable to inhalation exposures.   The PB-PK approach

is not considered to be superior to the approach that was used previously in

Section 4.6.1. for the following reasons:

     (1)  Some of the parameters used in the PB-PK model  must be estimated from

          the same set of data that is used in the crude approach.   There are

          uncertainties with the parameters used in the model.  Therefore, the

          amount of metabolites  predicted  by the PB-PK  model  is not  necessarily

          more reliable.
'Dr.  Jerry Blancato  of the Exposure  Assessment  Group  provided  valuable  advice
 during the preparation of this  Appendix,  and Dr.  Curtis  Travis  of  the  Oak  Ridge
 National  Laboratory provided  valuable  information regarding his  work on  PCE.
 Some of the data used in  this Appendix are  taken  from  a  report  entitled  "Inter-
 pretation of Metabolic Data  in  Exposure Analysis," conducted  by  Oak Ridge
 National  Laboratory and sponsored by the  Exposure Assessment  Group of  the  U.S.
 Environmental  Protection  Agency.
                                      A-l

-------
      (2)  Available data on the dose-metabolism relationship are primarily for
                                    i
          rats, with limited data avjailable for mice and humans.  Therefore,
                                    I
          the uncertainty associated with species conversion is somewhat re-

          duced by the direct use ofj these data.

      The use of the PB-PK approach in calculating the unit risk estimate for

PCE has provided insight into the problems of using metabolism/kinetic data in

quantitative assessment, and has pointed out a need for further research.

A.I.  DESCRIPTION OF THE PB-PK MODEL' FOR INHALED PCE

      PCE in vapor form in air is readily absorbed through the lungs into blood

by first-order diffusion processes.  Pulmonary uptake of PCE is largely deter-

mined by the ventilation rate, duration and concentration of exposure, solu-

bility in blood and body tissues, and metabolism.  PCE is eliminated by pulmo-

nary  excretion and by metabolism (primarily in the liver).  Figure A-l depicts

the PB-PK model used to simulate PCE; distribution and metabolism over time.

This model was used by Ramsey and Aniderson (1984) to study the behavior of
                                    i
inhaled styrene.  Abbreviations usedi in the figure are given in Table A-l.
                                    i
     The model in Figure A-l is described by a system of differential equations

which quantify the rate of change of < PCE concentration in each tissue group

over time.  Pulmonary uptake and elimination 'are described by the following

equation of mass balance of PCE entering and leaving the lungs:
                               - Ca)|= Qt(Cart - Cven)
This equation assumes a concentration equilibrium between arterial  blood and

alveolar air.  By substituting the relationship N = Cart/Ca into the equation

above, we have
                                      A-2

-------
                 'van
                "vf
                'vr
                'vp
                'vl
                          Alveolar Space
                            Lung Blood
   Fat Tissue

     Group
                          Richly  Perfused

                           Tissue Group
Poorly Perfused

 Tissue  Group
     Liver

(Metabolizing)

 Tissue  Group
                                    m
                      "H;
                                                 '•rt
                                                 'art
                        'art
                                                Q,
                                                 'art
                                                 'art
                                   K
                                         Metabolites
                                    m
Figure A-l.  Diagram of the physiologically-based pharmacokinetic
model.  Abbreviations are defined in Table A-l.
                                A-3

-------
                  TABLE A-l.  ABBREVIATIONS USED IN FIGURE A-l
Qa:  Alveolar ventilation rate (L/min)
Qt:  Cardiac blood output (L/min)     !
                                      |
Q^:  Blood flow rate to liver (metabolizing) tissue group (L/min)
Qf:  Blood flow rate to fat tissue group (L/min)
Qr:  Blood flow rate to richly-perfuse|d tissue group (L/min)
Qp:  Blood flow rate to poorly-perfused tissue group (L/min)
Y£, Vf, Vp, and Vp:  Volumes of tissue^ groups (L) corresponding, respectively,
     to liver, fat, richly-perfused, and poorly-perfused tissue groups
 N:  Bloodrair partition coefficient
P£, Pf, Pr, and Pp-.  Tissuerblood partition coefficient corresponding,
     respectively, to liver, fat, richly-perfused, and poorly-perfused tissue
     groups
m
     Maximum velocity of metabolism (m'g/min)
     Michael is constant (mg/L blood)
     Amount metabolized (mg)
    :  Concentration in arterial blood (mg/L)
Cven:  Concentration in venous blood (mg/L)
                                      I
Cjj., Cf, Cp, and Cp:  Concentrations in tissue groups corresponding, respec-
     tively, to liver, fat, richly-perfused, and poorly-perfused tissue groups
Cj:  Concentration in inhaled air (mg/L)
Ca:  Concentration in alveolar air (mg!/L)
                                      A-4

-------
                  '     cart  =  (Qacl  + Qtcven)  /  (Qt

 where  venous  blood  concentration  is given  by
               Cven  =  (Q^^/P^  + QfCf/Pf  +  QrCr/Pr  +  QpCp/Pp)  /  Qt

      The metabolism of  PCE  occurs mainly in  the  liver, with a rate of metabo-
 lism  characterized  by the Michaelis-Menten type  equation
                         dA
     The PCE concentrations in each of the four tissue groups are described by
the following equations:
Metabolism tissue group:
                               = o   (c    - c
                                 g   (       C
                                                     dAm
                           dt~    *   art
Non-metabolism tissue groups:
                           dC,-
where subscript i represents fat, richly-perfused, and poorly-perfused tissue
groups.
A.2.  PARAMETERS USED IN THE MODEL
     Table A-2 presents the physiological  and biochemical  parameters used in
the model.  How these parameters were obtained is described below.
                                      A-5

-------
TABLE A-2.  PHYSIOLOGICAL AND BIOCHEMICAL PARAMETERS USED IN THE MODEL
Parameters
Body weight (kg)
Alveolar ventilation rate
Qa
Blood flow rate
Qt
QA

Qf
Qr
Qp
Tissue volume
VA
Vf
vr
VP
Partition coefficient
N
PA
Pf
Pr
PP
vm
Km
Rats
I
0.35

0.083

0.104
0.0389
I
0.00920
' 0.0434
' 0.0126

i 0.0140
i 0.0315
0.0150
0.2550

18.9
3.719
108.994 -
3.719
•• 1.058
0.00586
2.9378
Mice
0.04

0.018

0.0225
0.00836

0.00198
0.00940
0.00272

0.00160
0.00359
0.00199
0.02.920

16.9
4.159
121.893
4.159
1.183
0.00138
2.9378
Humans
70

3.421

4.31
1.607

0.387
1.800
0.506

0.8000
6.2970
3.4980
51.085

10.3
3.719
108.994
3.719
1.058
0.2004
2.9378
•
A-6
1

-------
A.2.1.  Physiological Parameters
     These parameters are taken from Ramsey and Anderson (1984), with adjust-
ment for the difference in body weight.  This adjustment is performed by assum-
ing that within a given species the ventilation and blood flow rates are pro-
portional to the two-thirds power of body weight, and the volume of each tissue
group is directly proportional to the body weight.  For example, the alveolar
ventilation rate for a 0.35-kg rat is calculated by

                    Qa = 0.075 x (0.35/0.3)2/3 = 0.083 L/min

where 0.075 L/min is the ventilation rate for a 0.30-kg rat that is given in
Ramsey and Anderson (1984).
A.2.2.  Partition Coefficients
     The parameters for rats and mice were provided by Dr.  Curtis Travis of the
Oak Ridge National  Laboratory, which is the contractor for  the EPA project,
"Interpretation of Metabolic Data in Exposure Analysis," sponsored by the
Exposure Assessment Group, Office of Health and Environmental  Assessment.
It is our understanding that these parameters were obtained by Dr. Melvin E.
Anderson of the Air Force Medical Research Laboratory at Wright-Patterson Air
Force Base, using the equilibrium vial  technique.  The tissue:blot)d concentra-
tion ratio was calculated by taking the ratio of two concentration ratios,
air/blood and air/homogenate tissue.  The tissuetblood coefficients for humans
are assumed to be identical to those of rats.
A.2.3.  Metabolism Rate Constants, Vm and Km
     Rats:  The constants Vm and Km were estimated by least-square optimization
using the system of differential  equations and data from a  metabolism study by
Pegg et al. (1979).  As presented in Table 4-1, the amounts metabolized over 72
                                      A-7

-------
 hours  for  a  0.25-kg  rat  exposed  to  PCE  at 10 ppm and 600 ppm for 6 hours  were
                                     t

 recorded to  be,  respectively,  0.467 mg  and 9.11 mg.   These two data points


 (72  hours, 0.467 mg)  and (72  hours, 9.11  mg),  were  used  to fit the system of


 differential  equations,  described previously,  with  Vm and Km as unknown para-
                                     i

 meters.  The resultant estimates are[Vm = 0.004685  mg/min and Km = 2.9378 mg/L.


 For  the NTP  bioassay  rats,  which had (approximately  0.35  kg body weight, the


 estimates  are Vm = 0.004685 x  (0.35/0.25)2/3 = 0.00586 mg/min, with Km the same,


 independent  of body  weight.


     Mice:   An approach  similar  to  trie  one used for  rats  cannot be used for


 mice because there is only  one data point for  mice;  the  amount metabolized over


 72 hours was  0.356 mg when  mice  were [exposed to 10  ppm of PCE for  6 hours (see
                                     I
 Table  4-1).   To  derive Vm and Km for Imice, we  assume that the ratio of Vm


 between two  species  is equal to  the two-thirds  power of  the  ratio  of their body


 weights, and  that  Km  is  the same acro|ss species.  For  a  0.04-kg mouse,
                                     i


                                     i

                 Vm = 0.004685 x  (0.04;/0.25)2/3  =  0.0013S  mg/min



 and



                 Km = 2.9378 mg/L     :





These  results  are  presented in Table A-2.


     Humans:    The  metabolism constants for  humans were calculated  in a similar
                                     r
way  as the constants for mice.   The  rjesults  are presented  in Table A-2.


A.3.  RESULTS AND  APPLICATIONS TO RISJC ASSESSMENT


A.3.1.   Results  Under a Single Exposure Scenario


     Figure A-2  describes the amount metabolized over an 80-hour (4,680-,min)
                                     i

period for a 0.35-kg rat exposed to 200 ppm  and 400 ppm of PCE for 6 hours.


                                     ! A-8

-------
                                                                  o
                                                                  o
                                                                  
  CL
  Q-

 O
 O
                                                                                -o
                                                                                c
                                                                                fO


                                                                                a.
                                                                                Q.

                                                                                o
                                                                                o
                                                                                CM
                                                                               •o
                                                                                CD
                                                                                (/)
                                                                                O
                                                                                Q.
                                                                                X
                                                                                0)
 LT)
 CO
                                                                               O

                                                                               "3
                                                                               "C
                                                                               Ol
                                                                               NJ
                                                                               O
                                                                               -Q
                                                                               03
                                                                               -M
                                                                               Ol
(Biu)   p»Z||oq«)»ui  lunouiy
   l-D
  •

C\J  £_
 I  O
=C 4-


 CU UJ
 £- O
 13 0.
 C75
•r- t-
LL-  O
                        A-9

-------
Fron this figure, it is seen that the|curve for the 200-ppm group reaches

asymptotic value at about 72 hours, whereas the curve for the 400-ppm group

approaches its asymptotic value at on beyond 80 hours.  Thus, the assumption
                                     I

that the metabolites collected over the 72-hour period represent the total

amount metabolized is justified only for the 200-ppm group and represents,

at most, a very crude approximation for the higher dosed group.  Figures A-3

through A-7 depict blood and tissue cbncentration curves over an 80-hour

(4,680-min) period.  These curves indicate that PCE is rapidly removed in all

of the tissue groups except the fat tissue group.

A.3.2.  Results Under the NTP Multiple Exposure Scenario
                                     I
     Animals in the NTP bioassay were1 exposed to PCE 6 hours/day, 5 days/week.
                                     i
It would be of interest to investigate whether the amount metabolized reaches

steady state during the weekday exposure period and whether the metabolites are

eliminated over the weekend.         i
                                     i
A.3.2.1.  Results for Rats—From Table A-3, it is seen that the amount meta-
          	           i
                                     i
bolized per day reaches steady state [quickly at the second day of exposure.

The steady-state values are about 5.2; mg/day and 6.5 mg/day, respectively,
                                     I

for the 200-ppm and 400-ppm dosed grolips.  Although the amounts metabolized
                                     I
in the first and the last two days of} the week are less than the. steady-state
                                     i
values, these values increased from t|he first week to the second week, and thus
                                     i

are expected to increase to larger values than the values presented in Table
                                     i
A-3.  As an approximation, the steady-state values are taken as the effective

daily dose rate for the risk calculation (see Table 4-7).

A.3.2.2.  Results for Mice—The results for mice, are more uncertain than  for

rats because the metabolism constants: Vm and Km must be converted from those

for rats.  Using the converted Vm and Km for mice, the amount metabolized for a

0.025-kg mouse that is exposed to 10 |ppm of PCE for 6-hours is predicted  to be
                                     i
                                      A-10

-------
 c
i


 E
»-
                                                                                           I
                                                                                          Lf)
                                                                                          00
                                                                                           •
                                                                                          o

                                                                                          ra

                                                                                          £_
                                                                                          O
                                                                                          O
                                                                                          o

                                                                                          .0  l/l
                                                                                              t_
                                                                                          tO  Z3
                                                                                          3  O
                                                                                          O -C
                                                                                          C
                                                                                            Q.
                                                                                          C  CL
                                                                                          O)
                                                                                          O O
                                                                                          C O
                                                                                          O C\J
                                                                                          U
                                                                                            4->
                                                                                         LU  fO
                                                                                         oo  o
                                                                                          I  +->
                                                                                         Ol  O)
                                                                                         t_  m
                                                                                         3  O
                                                                                         en  CL
                                                                                        •i-  X
                                                                                        U_  OJ
(1/8 (JU)   UO|}BJ}U»3UOO
                          A-ll

-------
                                                       o
                                                       oo
                                                        3
                                                                          •r- O
                                                                          -I-3 iC
                                                                          •i- O
                                                                           N 4-
                                                                           O Q.
                                                                          .0 Q.
                                                                           03
                                                                          •t-> O
                                                                           O) O
                                                                              T3
                                                                              d
                                                                              to
     Q.
     Q.


    O
    O
   -CM
                                                                           CT> "3

                                                                          • — ^ t i l
                                                                              O
                                                                           C.  Q.
                                                                           O
                                                                          •r-  O
                                                                          +->  -M
                                                                           (O
                                                                           £_  TD
                                                                          -P  CL)
                                                                           C  CO
                                                                           dJ  O
                                                                           O  Q-
                                                                           C  X
                                                                           O  (1)
                                                                           o
                                                                              -p
                                                                          uj  ro
  i  co
 
-------
                                                                            o
                                                                            00
                                                                              O
                                                                                                 JC
                                                                                              CD
                                                                                                  CL
                                                                                                  Q.
                                                                                                 O
                                                                                                 o
 [__> fi_
 C Q.
 O)
 O O
 C O
 O CM
 O

UJ fO

Q. UJ
    CJ
    Q.

un o
 I  4->

    T3
 OJ O)
 t- 00
 rs o
 en Q.
•r- X
U_ O)
(T/BUJ)   UO|}B JJU9DUOO
                                 A-13

-------
                                                                                Cl
                                                                                e>
                                                                                c:

                                                                                e.
                                                                                             re
                                                                                             t_
                                                                                              I
                                                                                             Lf>
                                                                                             ro
                                                                                              •
                                                                                             O

                                                                                             re

                                                                                             £_
                                                                                             o
                                                                                             4-

                                                                                             CU

                                                                                             CO
                                                                                             CO
O)
CO
CD
Q_
                                                                                             _ 1  Q.
                                                                                             ~-~. Q.
                                                                                              CD
                                                                                              O  T3
                                                                                             •i—  C
                                                                                             4-J  fO
                                                                                              re
                                                                                              t-  E
                                                                                             4->  Q.
                                                                                              C  Q-
                                                                                              cu
                                                                                              o  o
                                                                                              c:  o
                                                                                              O  
                                                                                             uu  re
/ 0 ui )    u o i j w j j u • a u o
                                                                                              
-------
                                                                                               o
                                                                                               oo
                                                                                               «0
                                                                                                                CD

                                                                                                                 I

                                                                                                               OO
                                                                                                                - •
                                                                                                               o

                                                                                                                (D
                                                                                                                O
                                                                                                               <4-

                                                                                                                0).
co
 •
10
(N
                                           CM
                                           O
CM

1 I
/ 1
/ 1
//
/ 1
•//
Q / /
o°/ /
V / /
/ /
/ /
/ /
/ */
/ *I
/ **• 1
/ I
s o /
s °/
x 
o _
..3 <«-'
0 c oJ .
^* ~~ Q. OO
« E >,=
CM ~ J|
• §^
E
H =£

-— E
	 1 Q.
^°-
E 0
-— -o
«^-
c:
o -a
4_> n3
ro
4-J Q.
c: Q.
OJ
o o
c: o
0 g~
O UJ (X3
CJ
D- UJ
                          (I/Bui)    u o 11» J »u • o u o
I""-.  O
 I  4-5
•<.
•   T3
O>  OJ
£_  00,
3  O
cn  a.
•i-  X
                                                      A-15

-------
TABLE A-3.  DAILY METABOLITES  (mg)
I-UK t
EXPOSED TO PCE AT THE

Week Day
1 Mon.
Tues.
Wed.
Thur.
Fri.
Sat.
Sun.
2 Mon.
Tues.

Wed.
Thur.
Fri..
Sat.
Sun.
VU.-50-Kg KAI
IJJTP BIOASSAY
| Exposure
' 200 ppm
4.575
5,017
5.134
5.166
5.175
2.260
1 0.670
4.6.14
i 5,032
\
5.137
5.167
I 5,175
2.260
I
0,670
EXPOSURE PATTERN
conGentration (mg)
400 ppm
5.94,6
6.346
6,439
6.467
6.474
3.79.1
1,409
:5.987
6,35.2

6.441
6.467
6,475
3.792
1.413
! . -
               A-1.6

-------
0.10 mg, which is only one-third of that observed in the metabolism study by
Pegg et al.  (1979).  The reason for the underestimation becomes obvious when
one examines the following calculation.  For a 0.025-kg mouse, the alveolar
ventilation  rate (given by Ramsey and Anderson, 1984) is 0.0132 L/min.  At 10
ppm (i.e., 0.06908 mg/L), 6-hour exposure, the total uptake (including the
amount metabolized) is no more than

        '   0.0132 L/min x 0.06908 mg/L x 6 hr x 60 min/hr = 0.328 mg

which is even less than the metabolized value of 0.356 mg observed by Pegg et
al. (1979).  Thus, the underestimation is probably due to use of undervalued
physiological parameters; in particular, the ventilation rate.  Although it is
possible to adjust the ventilation rate and other parameters in the model, we
choose not to do so because the difference between the observed and predicted
metabolized dose is not large.  However, one should keep in mind that the risks
calculated on the basis of mouse data (using the PB-PK model) may be overesti-
mated due to the underestimation of the metabolized dose.  Table A-4 presents
the amount metabolized for mice.
     The NTP mice reached steady state quickly, with steady-state values at
0.74 mg/day and 1.07 mg/day, respectively,  for the 100-ppm and 200-ppm dosed
groups.   However, the mice appeared to metabolize faster than the rats, as
evidenced by the fact that the metabolized  dose almost Reached steady state on
the first day of exposure.   Furthermore, the mice also eliminated faster than
the rats, as evidenced by the fact that the metabolites  were almost totally
removed  over the nonexposed weekend.   Therefore,  it is more appropriate to use
as the daily dose,  the averaged amount metabolized over  7 days.  That is,  the
effective doses for the 100-ppm and 200-ppm dosed groups are respectively 0.54

                                      A-17

-------

EXPOSED
Day
Mon.,
Tues .
Wed.,
Thur.
Frt-
Sat.
Sun.
j
i

TABLE A-4»» DAILY METAMILHES (ing:)1
FOR A. 0.04-kg HOUSE
TO* PCE AT THE Np BEOASSA^ EXPOSURE" PATTERN-
i
1
Exposure
i
100 ppm
0.710
0.740
O.J42:;
O.J43-
0.743
0.084!
0.006'
; concenlrra'tfonr (mg)
ZOO; ppm
1.027
1.064
1..067'
1.067
1..067'
0.,17:S
0:313
A-18-

-------
mg/day and 0.78 mg/day.  These effective daily doses have been used in the risk
calculation  (see Table 4-8).
A.3.3.  Results for Human Exposure
     In the  previous risk calculation (sections 4.5 and 4.6.1), the study by
Bolanowska and Golacka (1972) was used to estimate the human dose-metabolism
relationship.  In that study, five subjects were exposed to 390,000 ug/m3
of PCE for 6 hours.  The metabolites in the urinary excretion were estimated
to be 13 mg.  Assuming that the urinary excretion accounts for 50% of the total
metabolites, one would expect that the PB-PK model predicts 26 mg of the total
metabolites.  This assumption is based on the observation from rat and mouse
studies that urinary metabolites accounted for at least 50% of total metabo-
lites.  The amount of metabolites actually predicted by the PB-PK model under
the above exposure scenario is 64.34 mg, which is about 2.5 times greater
than the expected value of 26 mg.  Therefore, if the Bolanowska and Golacka
study is considered adequate, the PB-PK model may somewhat overpredict the mag-
nitude of human metabolites and thus lead to an overestimation of human risk by
a factor of about 2 to 3.  However, it is remarkable that the PB-PK model could
predict with such accuracy despite the fact that the metabolism for humans must
be estimated from that of rats.
     To calculate the unit risk for humans, it is necessary to calculate the
amount metabolized associated with a continuous (24-hour) exposure to 1 ug/m3
of PCE in air.   Table A-5 presents the consequences of the PCE exposure over a
2-week (14-day) period during which the steady state is reached.   The steady-
state value of  the amount metabolized is 7.9 mg/day.  This value  was used to
calculate the unit risk of PCE in Table 4-9.
     Table A-6  presents the consequences of continuous  (24 hours/day) exposure
to PCE at higher concentrations.   The values presented  in Table A-6 represent
                                      A-19

-------
TABLE A-5.  AMOUNT  METABOLIZED, PER DAY FOR A 70-kg PERSON

Day
1
2
3
4
5
6
7
8
9

10
11
12
13
14
Amount of metabolite (m.g:) at end of each day
4.36
x 10-4
6.108 x 10-4
6.43
6.92
7.25
7.47
7.62
7.71
7.78

7.82
7.85
7.87
7.88
7.88
x 10-4
x 10-4
x 10-4
x 10-4
x 10-4
x 10-4
x 10-4

x 10-4
x 10-4
x 10-4
x 10-4
x 10-4 ;

                            A-2.0

-------
 TABLE A-6.  STEADY-STATE METABOLIZED DOSE  (rug/day)
                 FOR A 70-kg PERSON
         CONTINUOUSLY EXPOSED TO PCE IN AIR
 Ambient air                     Amount
concentration                  metabolized
   (ppro)                          (mg)


    150                          221.96

    125                          211.71

    100                          197.82

     75                          178.02

     50                          147.76

     25                           96.89

      1                            5.37
                        A-21

-------
the amount metabolized per day 14 days after first exposure; these values are


considered to be steady-state values.


A.4.  COMMENTS                       ;


     As demonstrated in this Appendix, the PB-PK model can be a useful to.ol
                                     I

for quantitative risk assessment.  It: provides not only the estimation of the


metabolized dose in the NTP bioassay,* but also an insight into, the validity of


various assumptions that must be made in risk assessment, such as the validity


of using data from a single-exposure experiment in place of multiple-exposure


bioassay data.  For the PCE PB-PK modeling, we find the following data to be


desirable:
                                     i
    (1)  The tissue concentrations and the amount metabolized over time are
                                     i

    useful.  For PCE, only metabolism! data at 72 hours are available.  The


    measurements at several time points and at different exposure concentra-
                                     i .
    tions are necessary for estimating the parameters and verifying the ade-


    quacy of the model constructed.  ;


    (2)  The physiological and biochemical parameters for the animals used in


    a metabolism/kinetic study are useful.  Although the physiological para-


    meters for most species are known to be within certain ranges, the use of

                                     t
    exact parameter values for the animals could greatly improve the accuracy


    of the prediction by the model.  We feel that work should be initiated to


    develop a methodology that better quantifies key metabolic parameters in
                                     j
    humans.                          !
                                      A-22

-------

-------
m
-o
o>
o
o

00

CO
ro

o
o
in
                                                                                           3
                                                                                           O

                                                                                           "
                                                                                           > me

                                                                                              i.1-
                                                                                              -i 
                                                                                              S «
                                                                                              Q) (A
                                                                                              O
                                                                                             ,
                                                                                              a
                                                                                           3 S 5
                                                                                           *   3
                                                                                           *•   5'
                                                                                           en   o
                                                                                           ro   3
                                                                                           2   3
                                                                                           oo   «


                                                                                                I

                                                                                                30
                                                                                                0>
                                                                                                
-------