EPA/600/8-fi^/nn
! 983
HEALTH "l.V.; ASSESSMi'.M A-WJACsl !'0R
2, J , 7 . «<- iETilAoHLOKCi) I B:- SV.O-£-l> ICX IN
by
O.^baas Muker'.ee
Environmental Criteria and As^esynert Orti^e
U.S. Envirornncr.tal Prntfftior. A^en.y
Cincinnati, OH 45268
Charles H. Ris and John Sohau.r,
Office of Health due Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC 20460
EPA Project l)ff:c»-r
Dehrias Muker; :c
KNVIIU5XMENTAL CRITERIA A>. ASSESSMENT OFFICE
07?ICC OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION ACSNCY
CINCINNATI, OH 4*268
-------
1 «tBORT NO.
EPA-600/8-85-013
2.
3 *€C HINT'S ACCESS ON NO
P286-10067*
« TiTke ano subtitle
Health Risk Assessment Approach for
2,3,7 »B-Tetrachlerodlbenzo-£-Dioxln
t> arponr l-'ate
June l?6b
l H»to«MiNG ORGANIZATION CODE
7 AUTHOMtS) ^ )
Dr. Debdas Mukerjee, Charles H. His and John Scha-.:m
• HiRFORMlNQ ORGANIZATION HEPCRT NO
V Ptft'ORMlNaORGANIZATION *A». E ANC AODRESS
Environmental Criteria and Assessment Office
2U-S. EPA, Cincinnati, OH 45268
Crfice of Health an 4 Environrvnral Assessment
_ U-S. EPA. Cincinnati. 'JH .
12. SPONSORING AGENCY NAME ANC ADCESS
Environmental Criteria and Assessment Office
Office of Research and Development
J.S. Environmental Protection Agency
Cincinnati, 0:i i5:68
TECHNICAL REPORT OATA
fPitau rtau o> tht mtrtc t*.'urt conpltung;
'>0 FMOOAAW tLEVlf^T NO
TiACT'OHAnT NO
13 tr'f O* •eeo«T ANO PERIOD COVtREO
14 SPONSORING AGENCY CODE
r.PA/600/22
IS. SUPPLEMENTARY NOTkS
A69TRACT
2,3,7.8-Tctrachlorodibenzo-£-dioxin (TCDD) is one of the meet toxic and
environmentally stable pollutants. In addition to vf.. -is toxic effects,
TCDD has beer, found to can qp teratogenic, fe-ocidal, reproductive and
carcinogenic effects in animals. In huxans it r.dverseiv affects various
organ systems and is probably carcinogenic as well. This reoort documents
the nethodologies utilized by the United States Environmental Protection
Agency in its develupcient of health risk assessment from expuiury to TCDD.
17.
KEY wonOS AND OOCUMEKT ANALVfS
OfSCR'PTORS
b lOfNTl- t oS'CPt N ENCEO Tl "MS
COSATIIn
18. OlSTBHU'lONSTATtMl NT
Public
19 SkCU'RIT* CLAS* fcrr
Unciassified
21 no o> PA ,kS
GO
|?0 SEClmty Class lm
I Unclassitied
fpa *•»» 2270-1 (¦•». 4-7T) >ki«iou> icfiONiioitc.tM
-------
NOTICE
This document has been revised In accordance with the U.S. Environ-
mental Protection Agency policy and approved for publication, hentlon of
trade names or commercial products does not constitute endorsement or recom-
mendation for use.
-------
PREFACE
ms document was compiled fror the following documents prepared Dy the
U.S. iPA. Office of Health -
dtoxln {£PA 440/5-04-007)
?) Health Assessment Do sent for Polychlorlnated J1benzo-p-<31ox1ns
UPA 6C0/8-64-014A; E-diox1ns |ECA0-C1N-~'J04; Program Of f i :e Oraft)
4) Drinking Water Criteria Document for 2,3,7,0-letrachlorodlbemo-
£-d1ox1n (ECAO-CIN-40i; Program Office Draft)
5>) Guidelines for Developing Water QjdUty Criteria Documents
(federal Register, Volume 45, No. 231; November 2d, I960}
6) Proposed Guidelines for Mutagenicity Risk Assessment (Federal
Register, Volume 49, No- 227; November 23, 1984)
This document has been compiled by the following:
Defcdas Mukerjee Charls; H. ;*1s
Environmental Criteria and Carcinogen Assessment Group
Assessment Office Washington, DC
Cincinnati, OH
John Schauni
Exposure Assessment Group
Washington, nc
ill
-------
While tne material contained 1n this document was drawn from the six
source documents listed m the previous page an acknowledgement 1s hereby
given to ttose IPA employees whose scientific assessment work has been
featured.
Cancer Toxicology, Or. Charles Hlremath, Carcinogen Assessment
Grouo
• Cancer Epidemiology, Or. David Bay 11ss. Carcinogen Assessment
Gr j up
Risk Assessment, 0. . Steve Bayard, Carcinogen Assessment Group
Mutagenicity, Or. Sht-lla Rosenthal, Reproductive Effects
Asse
-------
TABLE OF CONTENTS
Page
' i
V.'u ',M .SMf JT AI'PR'J/^CH rOR CH'CjNJC ::XPCSUF.E 3
2.1. HoMAf HtAL'.-I R.!fA Af/^E'o^EN'i 8V>tP ON CAftC'NOGENiCITY
OA! n 3
2.1.1. Cjrc'nogenv Ity 'lata AvjUj'o'f for 'COfJ 3
2.1.2. "ualHatlvt Cr«Urla for f-eterfilnl'tg Strength
of Carcinogenic Jv1der.ee 7
2.1.3. Approach 'or Quantitative Cancer Risk Assessment. 10
2.1.4. Cancer R>sk Estimates fc* 1 COO via Oral F.outf-
ar.o Derivation of the Amnlent Water QuaMty
Criteria for TCDD . . 16
2.1.5. Cancer Unit Risk Estimates for TCDD via the
Inhalation Route 25
2.1.6. Determination of Relative Potency 27
2.2. COMPARISON OF EPA'S RISK ASSESSMENT APPROACH WITH
OTHtR METHODS 27
2.2.1. Alternative Methodological Approaches Utilized
by EPA tor Cancer Risk Assessment for 1COO. ... 27
2.2.2. Comparison of EPA's Carcinogenic Potency and
Criteria with Approaches Followed by ^DA nd CDC. 28
2.2.3. Comparison of Approaches by FDA and CDC fo:
Estimating So* 1 Ingestion Exposure 31
2.3. HUMAN HEALTH RISK ASSESSMENT BASED ON CHRONIC NOM * INO-
GEN1C TOXICITY DATA (INCLUDING REPRODUCTIVE EFFEC1\ rMA). 36
2.0.1. Honcardnogenlc Toxicity Oata Available for
Chronic Exposure to TCOD 38
2.3.2. Estimation of Acceptable Dally Intake for
TCOD (for Comparative Purooses Only}
2.4. HUMAN HEALTH RISK ASSESSMENT BASED ON MUTAGENICITY DATA. . 42
2.4.1. Mutagenicity Data Available on TCDD 42
2.4.2. Qualitative Mutaqenlc Risk Assessment
Methodology 44
3. RIS< ASSESSMENT APPROACH TOR ACUTt EXPOSURE TO 2,3,7,8-TCDD ... 47
3.1. TOXICITY DATA AVAILABLE FOR ACUTE EXPOSURE TO TCDD .... 49
3.?. ESTIMATION OF 1-DAY HEALTH ADVISORY 52
v
-------
Page
4. RISK ASSESSK?NT APPROACH FOR SHORT-TERM (SU6CHR0NIC) EXPOSURE
TO 2,3,7,8-KOO 53
4.1. TOXICITY OATA AVAILABLE FOR SUBCHRONIC EXPOSURt TO TCOO. . 54
4.2. ESTIMATION OF 10-DAY HEALTH AQVISORY 5S
5. REFERENCES 5?
v 1
-------
LIST OF ABBREVIATIONS
ADI
Acceptable dally Intake
AHH
Aryl hydroxycarbon hydroxylase
bw
Body weight
BCF
Bloconceniratlon factor
DMBA
Dltnethylbemanthracene
DNA
Deoxyribonucleic add
fll
Frank-effect level
GI
Gastrointestinal
HA
Health Advisory
LOAEL
Lowest-observed-adverse-effect level
LOEL
Lowest-observed-effect u-vel
3-NC
3-Hethylcholanthrene
NOAEL
No-observed-adverse-effect lev?1
NOEL
No-observed-effect level
ppb
Parts per billion
PPt
Parts per trillion
R
Bloconrentratlon factor
UF
Ur,certainty factor
yc
Water consumption
-------
1. INTRODUCTION
2,3,7,8-Tetrachlorod1ben?o-|>-d1ox1n (TCDD), an Isomer of a large series
of chlorinated aromatic hydrocarbons classified as polychlorlnated dlbenzo-
£-d1ox1ns, 1s one of the nost toxic and highly stable compounds found 1n the
envlronnent. It Is a highly toxic compound with demonstrated acute, sub-
chronic and chronic effects In an;.ma1s and man. Acute and subchronlc expo-
sures of TCDD adversely affect the skin, liver, nerve and Irmune systems
It H teratogenic* fetotoxic and reduces fertility 1n laboratory animals.
It 1s a proven animal carcinogen and 1s probably carcinogenic 1n humans.
This toxic substance Is not produced commercially and has no industrial
use. It Is produced as an unwanted contaminant during the manufacture of
chlorobenzenes, chlorophenols and their derivatives such as 2,4,5-trIchloro-
phenoxyacetk add (2,4,5-T) and Z-(2#4,5-tMchlorophenoxy) propionic acid
(sllvex). The disposal of spent chemical manufacturing wastes 1n landfills
or 1n uncontrolled disposal operations creates the opportunity for environ-
mental hazards to develop. It has been detected 1n many spedes of com-
mercial and noncommercial fish from several rivers and lakes In the United
States, fly ash from municipal waste Incineration and soot from samples of
a transformer fire 1n an office building have been found to have TCOD as a
contaminant. A1r filter samples collected during an Industrial Mre have
also been found to contain TCDO. The release of TCDD Into the atmosphere Is
due to uncontrolled endothernlc reactions during the production of 2,4,5-T
and other thlorgphenol compounds. Another potential source of TCOO in the
atmosphere 1s spraying of herbicides containing 2,4,5-T.
TCDD 1s resistant to blodegradatlon reactions 1n the soil. The half-
Ufe of IlDD 1n soil has been found to be more than a decade. It binds
avidly to the organic content of soil.
1«78A -1- 04/00/85
-------
The public health concern for human exposure to TCD9 Is due to the
multiplicity of adverse health effects, which can result from very low
levels of exDosure and increasing knowledge about Us wldespiead distribu-
tion 1n the environment. These considerations combined ylth Its extrene
stability 1n the environment make TCOD a potential major environmental
concern.
U.S. EPA with Hs numerous regulatory activities has prepared several
health assessment documents on dloxlns that Inventory the data from world
scientific literature ar.d provide Interpretations for hazard and risk
analysis estimations. Underpinning the health assessment activity, are
assessment methodologies for toxic effects consisting of carcinogenic and
noncarctncgenlc effects from chronic, subchroMc and acute exposures.
This report documents the health assessment methodologies utllWed by
U.S. EPA 1n Its development of TCOD health assessments and presents the key
scientific information es:enMal for risk analysis. The scientific data
base as reported by reference 1n this report 1s current through May, 1904.
1876A
04/04/85
-------
2. RISK ASSESSMENT APPROACH FOR CHRONIC EXPOSURE
2.1. HUNAN HEALTH RISK ASSESSMENT BASEO ON CARCINOGENICITY DATA
2.1.1. Carcinogenicity Data Available for TCOD. In a preliminary study
by Van Miller (i9Mft,b), 2,3,7,8-KDD was tested for Lar c inogenlcl ty follow-
ing oral administration to rats. Ai the five highest dietary levels, 0.005,
0.05, 0.5, 1.0 and 5.0 ppb, which allowed long-tern survival of the animals,
an increase in the Incidence of total tumors was observer There were no
tumors In animals at an exposure level of 0.001 ppm and in the control
animals. This study, however, provides only suggestive evidence of a carci-
nogenic response since no 1ncpease m *He ipeclf^c tumors was detected and
the group sizes, ~10 animals/group, were too small for an assessment of a
treatment-related response. Koctba et al. (1978D), in a more extensive
study, detected a positive carcinogenic response. In this study, the
estimated Intake of 2,3,7,B-TCOD from the diet was 0.0, 0.001. 0.01 and 0.1
ug/kg/day. In the high-dose group, both male and female animals had
significant increases in sUe-spec1fU tumors. The target organs and tumor
types 1n male animals were squamous cell carcinomas of the tongue and of the
hard palate, and adenomas of *he adrenal cortex, while 1n female animals the
target organs and tumor types were hepatocellular carcinomas, squamous eel!
carcinomas of the tongue and of the lung. The data demonstrate «.hat dietary
exposure to 2,2,7,8-TCOD at levels that produce a dally dose of 0.1 pg/kg
produce increased tumor incidences In both male and female rats.
Under the National Toxicology Program, 2,3,7,6-TCDD was tested for
carcinogenicity m rats following administration by gavage (NIP, 1980a).
Both male and female animals were exposed to weekly doses of 0.0, 0.01, 0.05
and S pgAg bw. The only tumors that appeared to be treatment-related
were follicular cell adenomas or carcinomas of the thyroid In male animals,
1878A
-3-
UV31/85
-------
and neoplastic nodules or hepatocellular carcinomas of the liver in female
animals. The Incidence of these turcors was significantly greater than
control 1n the high-dose groups, and the incidences of both tumors showed a
posHlve dose-related trend. Unier the conditions of this assay, 2,3,7,6-
TCQO was concluded to be carcinogenic 1n both male and female rats.
Further studies 1n mice exposed by gavage have provided support for the
carcinogenicity of 2,3,7,8-TCUO. Tcih et al. (1979) exposed male mice to
2,3,7,8-TCQD at doses of 0.0, 0-0U7, C.07 and 7.0 pg/kg/week 1n a study to
determine ./hether 2,4,5-TCPE, Us contaminant 2,3,7,8-ICDD, or both were
carcinogens. At the 0.7 ug/kg/ueek level there was o significantly
Increased Incidence of liver tumors. Liver tumors were not significantly
Increased In the high-dose group; however, early mortality In this group may
have precluded observing late oevelop1n$ tumors. Similar Increased Inci-
dences of liver tumors were observed In the NTP (1980s) study An the hlgh-
dosv male m'ce exposed to 0.5 jig/kg/week and 1n the high-dose female mice
exposed to 2 pg/kg/week of 2,3,7,8-TCDD by gavage. Female mice also had
ar. Increased Incidence of follicular cell adenomas of the thyroid. In both
studies, 2,3.7,8-TCDD was carcinogenic to mice with effective dose* ranging
between 0.5 and 2 tig/kg/day depending on sex and the Individual study.
The mo ie skin two-stage tumcrlgen1c1ty model has also been used to test
the carcinogenic potential of 2,3,7,8-TCDD. Following long-term dermal
application 3 times/week of 2,3,7,8-TCOO at levels of 0.01 and 0.005
^'application to male and female nice, respectively, there was an
Increased Incidence of skin t;;mors only 1n female mice (NTP, 1980b). Along
with the Ino'.catlon that 2,3,7,8-lCOD was a complete carcinogen 1n this
system, OlG^ovjnnl at al. <1977) reported tnat 2,3,7,8-TCDD was also a tumor
Initiator 1n mouse skin. The ability of 2,3,7,6-TCDD to Initiate tumors.
1878A
05/31/85
-------
however, has yet to ba confirmed since appropriate vehicle and promotion
only control groups were not Included. Attempts to demonstrate tumor
promoting activity with 2,3,7,8-TCDD on mouse skin have produced negative
results 1n some assays (HTP, 1980b; Berry et al.. 1978. 1979); however,
Poland et al. (198?) reported that 2,3,7.8-TCOD was a tumor promoter when
tested on the skin of mice homozygous for the "hairless* trait but not In
mice heterozygous for this recessive trait. P1tot et a1. (1980) also
reported that 2,3,7,8-TCDD was a promoter for DEN-lnHiated hepa^carcino-
genesis 1n rats following parenteral administration of the compounds. On
nouse skin, 2,3,7,8-TCOO was a complete carcinogen and possibly a tumor
Initiator, while no tumor promoting activity could be attributed to 2,3,7,8-
TC00 In the assays. In rat liver initiated with DEN, 2,3,7,8-TCDD was a
(umor promoter.
In studies of the Interaction of 2,3,7,8-TCDO with other chemical car-
cinogens, Kouri et al. (1978) reported that 2,3,7,8-TCDD was a rocarclnogen
with 3-NC when administered by subcutaneous Injection. In the mouse skin
bloassay, Initiation with simultaneous administration of 2,3,7,8-TCDD and
MBA, however, did not affect tumor ylelo (01G1ovann1 et al., 1977).
Similarly, no effect was observed when 2,3.7,8-TCDD was administered either
Immediately before (5 minutes) or 1 day after DN8A Initiation {Berry et al.,
1979; D1G1ovann1 et al., 1977, 1979; Cohen et al., 1979). When treatment
wm 2,3.7,8-TCDD occurred 1-10 days before 0N8A Initiation, 2,3.7,8-TCDD
demonstrated a potent antlcarclnogem? action. Although 1-5 days prior
exposure to 2,3,7,8-TCOO Inhibited tumor initiation by 8aP, 3-MC and BaP-
dlol-epoxlde, the tumor Initiating ability of the latter compound was also
Inhibited when 2,3,7,8-TCDO exposure occurred either S minutes before or 1
day after Initiation (L1G1ovann1 et al., 1980). The Increased AHH activity
1876A
-S-
10/12/64
-------
resulting from ?,3,7,8-TC0D exposure may c^count for the antlcarclnogenlc
activity by altering the metabolism of the Initiating compound; however,
DlGlovannl et al. (1980) suggested that the Inhibition of the initiating
activity of BaP-d1ol-epox1tie 1 day after Initiation Indicates that more than
one mechanism participates 1n the antlcarclnogenlc activity of 2,3,7,8-TCDD.
Two Swedish case-control studies report a highly significant association
of soft-tissue sarcoma with exposure to phenoxy acid or chlorcphenols (or
both) (Harder and Sandstrom, 1979; Eriksson et al., 1979, 1981). They do
not, however, pinpoint the risk to the dloxln contaminants. In fact, 1n one
study, the risk was found to extend to phenoxy acids free of dloxln Impuri-
ties. In that study, the risk Increased to 17 when phenoxy adds known to
contain dloxln Impurities were considered (polychlorlnated dibenzodloxins
and dlbenzofurans). The extent of observer bias and recall bias Introduced
Into these studies by the employment of an undesirable methodology (self-
administered questionnaires) 1s probably not of sufficient magnitude to have
produced the highly significant risks found 1n the studies. However, the
possibility exists that these biases could have played a role 1n the
determination of these risks, and consequently, the data must be considered
limited for the carcinogenicity of phenoxy acid herbicides and chlorophenols
1n the absence of confirmatory studies.
Later studies thjt did not reveal a significant excess risk of sofi-
tissue sarcoma have severe problems with their methodologies (Eriksson et
al., 1981; Cook et al.. 1980). These problems make later studies Inadequate
to evaluate the risk of soft-tissue sarcomas from exposure to phenoxy acids
and/or chlorophenols and. consequently, t»iox1n.
1878*
-6-
05/31/85
-------
Therefore, the Swedish case-control studies provide United evidence for
the carcinogenicity of phenoxy acids and chlorophenols 1n humans. However,
with respect to the dloxln Impurities contained within, the evidence for
human carcinogenicity based on the epidemiologic evidence 1s only sugges-
tive, because H is difficult to evaluate the risk of dloxln exposure 1n the
presence of the confounding effects of phenoxy adds and chlorophenol.
Substantially wpiker evidence exists Incriminating 2,4,5-T or 2,3,7,8-
TCUD (or both) as the ca'jse of malignant lymphoma (Hardell, 1979; Hardell et
a1.» 1980; Hardell and Eriksson, 1981; Edllng and Granstam, 1979) and
stomach cancer (Axelson et a1., 1980; Thless and frentzel-Beyme, 1977) 1n
humans.
2.1.2. Qualitative Criteria for Determining Strength of Carcinogenic
Evidence. The approach followed by the International Agency for Research
on Cancer (I ARC, 1982) has been utilized by EPA to determine th» strength of
cardnogc-nlc evlder.ee for a chemical. A chemical under study 1s classified
1n the following categories, according to the strength of scientific
evidence for Its carcinogenic properties 1n experimental an1m«l test systems:
¦Sufficient Evidence" of Carcinogenicity. There Is an In-
creased Incidence of malignant tumors 1) 1 ri multiple- species
or strains; 2) m multiple experiments (preferably with dif-
ferent routes of administration or using different aose
levels); or 3) to an unusual degree with reyaro to Incidence,
site, or type of tumor, cr age at onset. Additional evidence
may be provided by data concerning dose-response effects as
well as Information on mutagenicity or chemical structure.
"Limited Evidence" of Carcinogenicity. Data suggest a carci-
nogenic effect but are limited for the following reasons:
1) the studies refer to a single species, strain or experi-
ment; 2) the experiments are restricted due to Inadequate
dosage levels, Inadequate deration of exposure to the agent,
Inadequate period of follow-up, poor survival, too few
animals, or Inadequate reporting; or 3) the neoplasms produced
often occur spontaneously or are difficult to classify as
malignant by histological criteria alone (I.e., lung and 1iv«r
tumors In mice).
18T8A
-7-
05/31/85
-------
'Inadequate Evidence." Because of major qualitative or quan-
titative limitations, the studies cannot be interpreted as
showing either the presence or the absence of a carcinogenic
effect.
"Negative Evidence." Within the limits of the toats used, f.he
chemical 1s not carcinogenic. The number of negative studies
1s limited since, 1n general, studies showing no effect are
less likely to be published than those suggesting carcino-
genicity.
"No Oata." Data are not available.
When available human data are factored Into the evaluation of carcino-
genicity, 1t 1s found that there are three main sources of evidence for
carcinogenic j v ,iumans:
1. Cast . j-ts of individual cancer patlen' who wert exposed lo
the che. <.
2. Descriptive epidemiological studies 1n which Individual expo-
sures to the chemical or group of chemicals was found to be
associated with an Increased risk of cancer.
3. Analytical epidemiological (case-control 1n cohort) studies 1n
which nd1vlo;-al exposures to the chemical or group of chemi-
cals was found to be associated with an Increased risk of
cancer.
In evaluating human evidence, three criteria must be met before a causal
association may be Inferred between exposure and human cancer:
1. There 1s no Identified bias that could explain the association.
2. The possibility of confounding has been considered and ruled
out as explaining the association.
3. The association Is unlikely to be due to chance.
In general, although a single study may be Indicative of a cause-effect
relationship, confidence in inferring a causal association Is Increased when
several independent stuoles are concordant 1n showing the association, when
the association 1s strong, when there Is a dose-response relationship, or
*878A
-8-
05/31/8*
-------
when a reduction in exposure 1s followed by a reduction 1n the evidence of
cancer. Therefore, when evaluating humen data, the IARC we1ght-of-evidence
categories are as follows:
¦Sufficient Evidence" of Carcinogenicity. The data Indicate a
causal association between exposure and hjman cancer.
"Limited Evidence" of Carcinogenicity. The data indicate a
possible carcinogenic effect 1n humans, although the data are
not sufficient to demonstrate a causal association.
"Inadequate Evidence" of Carcinogenicity. The data are quali-
tatively or quantitatively Insufficient to allow any cone' .slon
regarding carcinogenicity for humans.
Ir. the absence of sufficient evidence from human studies, evaluation of
th* carcinogenic risk to humans 1s based on consideration of the epidemio-
logical and experimental animal evidence together.
As a final rank'ng step, the chemicals, groups of chemicals, or Indus-
trial processes are placed Into one of three groups.
Group 1. The chemical, group of chemicals, or Industrial process
1s carcinogenic for humans. This category 1s used only
when there wu sufficient evidence to support a causal
association between the exposure and cancer.
6roup 2. The chemical or group of chemicals 1s probably carcino-
genic for humans. This category Includes chemicals for
wMch the evidence of human carcinogenicity 1s almost
sufficient as well as chemicals for which 1t 1s only sug-
gestive. To reflect this range, this category has b*en
divided Into higher (Subgroup A) and lower (Subgroup B)
degrees of evidence. The lata from experimental animal
studies plays an Important role 1n assigning chemicals to
Group 2, particularly Subgroup 0.
Group 3. The chemical or group of chemicals cannot be classified
as to Us carcinogenicity for hurvns.
Using the criteria developed by IARC (1962), a welght-of-evllence rank-
ing for the carcinogenicity data of 2,3,7,0-TCDD can be made. Because of
the induction of hepatocellular carcinoma 1n two strains of female rats and
1878A
-9-
05/31/BS
-------
both sexes of o.it mouse strain, along with the Induction of thyroid tumors,
subcutaneous fibrosarcomas, and lung and tongue tumors In both rats and
n'ce, the evidence of carcinogenicity for 2,3,7,8-TCDQ 1n animals would be
regarded as 'sufficient" if the classification system of the XARC were used,
lhe demonstration of a promotion effect 1n rat liver after 1n* tlatlon with
dlethylnltrosamlne and a cocarclnogenlc response when ?,3,7,B-TCDD was
Injected simultaneously with 3-methyl chloranthrene further supports the
"sufficient" classification In animals.
The human evidence for the carcinogenicity of ?,3,7,8-TCD0 alone 1s
regarded as "inadequate" using the IARC classification because of the
difficulty of attrlput1ng the effects to 2,3,7,8-TCOD, which occurred a, an
Impurity 1n the phenoxyacetlc acids and chlorophenols to which the people
were exposed. However, the human evidence for the carcinogenicity of
chlorinated phenoxyacetlc herbicides and ch icrophenols with chlorinated
dlbenzodlcxln *.-d dlbeniofuran Impurities 1s "limited" according to the IARC
criteria.
The overall ev'dence of carcinogenicity, considering both animal and
human studies, would place 2,3,7,8-TCDD alone 1n IARC s Group ?B and
2,3,7,8-TCDD 1n association with the phenoxy herbicides and chlorophenols in
Group ?A category (IARC, 198?). The IARC regards chemicals In both cate-
gories as probably carcinogenic in humans.
2.1.3. Approach for Quantitative Cancer Risk Assessment. The f.ata used
for a Quantitative estimate 1s one or both of two types: 1) lifetime animal
studies and 2) human studies where excess cancer risk has been associated
with exposure to the agent. In animal studies 1t 1s assumed, unless
evidence exists tr. the contrary, that 1f a carcinogenic response occurs at
the dose levels used 1n the study, then respi ses will also occur at all
lower doses with an Incidence determined by an extrapolation model.
1878A -10- OD/31/85
-------
Therd 1s no solid sclentViic oasis for any mathematical extrapolation
model that rebates carcinogen exposure to cancer risks at the extremely low
concentrations that must be dealt with 1n evaluating environmental harards.
For practical reasons, such low levels of risk cannot be measured directly
either by animal experiments or by epidemiologic studies. Therefore* scien-
tific Judgement depends on the current understanding of the mechanisms of
carcinogenesis for guidance as to which risk model to use. At the present
time, the dominant view of the carcinogenic process involves the concept
that most canter-causing agents ahc cause Irreversible damage tc DNA.
There 1s reason to expect that the quantal type of biological response Is
associated with a linear nonthreshoid dose-response relationship. This is
particularly true at the lower end of the oose-response curve; at higher
doses, there can be an upward curvature, probably reflecting the effects of
multistage processes on the mutagenic response. The linear nontNrestiolJ
dose-response relationship 1s also consistent with the relatively few
epidemiologic studies of cancer responses to specific agents tha; contain
enough information to make the evaluation possible (e.g., radiation-induced
leukemia, breast and thyroid cancer, skin cancer Induced by arsenic 1n
drinking water, liver cancer Induced by anatoxins the diet). There is
also some evidence from animal experiments that 1s consistent with the
linear nonthreshold model (e.g., INer tuoors Induced 1n mice by 2-acetyl-
amlnofluorene In the large-:cal? EDD1 study at the National Center for
ToxUologlcal Research, and the initiation stage of the two-stage carcino-
genesis model in rat liver and mouse skin).
Because of Us scientific basis, although limited, the linear nonthresh-
old model has been adopted by EPA as the primary basis for upper bound risk
extrapolation in the low-dose region of 1 he dose-response relationship.
1878A
-11-
05/31/85
-------
The risk estimates made with this model should be regarded as conservative,
representing the most plausible upper limit for the risk {I.e., the true
risk is not likely to be higher than the estimate, but 1t could be lower).
The inaihematlcal formu1ation chosen to describe the linear nonthreshuld
<1ose-response relationship at low doses 1s the linearized multistage model.
Tnls modi'l employs enough arbitrary constants to be able to Mt almost any
monotonUeliy Increasing dose-response data, and 1t Incorporates a procedure
for est'mating the largest possible linear slope (In the 95% confidence
I'.tnlt :ense? at low extrapolated doses that 1s consistent, with the data at
all dose levels of the experiment. The following is a description of the
Low-Cose Anl.w 1 Extrapolation Model:
Let P(d) represent the lifetime risk (probability) of cancer at dose d.
The multistage model has the form
where
tquivalently.
where
P(d) - 1 - exp [-(qQ ~ q1d ~ q2da ~ ... + qkdk>]
q > 0. 1 - 0. 1, ?. ..
Pt Id) - 1 - exp L{q1d «¦ q.,d» ~ ... ~ q^dS]
pt{d) .
1 1 • P(0)
1s the extra risk over background rate at dose d.
The point estimate o* *~»» coefficients q^, 1-0, 1, 2 .... and .on-
sequently, the e\tra risk function, P^(d), at any given dose d, 1s calcu-
lated by max mlzlng the Hke'lhood function of the data.
1878A
-12-
Q*/04/85
-------
The point estimate and the 95% upper confidence Unit of the extra risk,
P^(d), are calculated by using the computer program 6L0BAI 79 developed by
Crump and Watson (1979) At low doses, upper 95X confidence Units on the
extra risk anc lower 95% confidence limits on the dose producing a given
risk are determined from a 95% upper confidence limit, on parameter
qj. Whenever is >0, at low doses the extra risk Pt(d) has approxi-
mately the form P^(d) ¦ q^* x d. Therefore, q^* x d Is a 95% upper
confidence Unit on the extra risk and R/q^* 1s a 95% lower confidence
limit on the dose producing an extra risk of R. let be the maximum
value of the 1og-11ke11hood function. The upper limit, q/, 1s calculated
by Increasing q^ to a value q^# such that when the 1og-l1ke1Kood 1s
remax1m1zed subject to this fixed value for the linear coefficient,
the resulting maximum value of the 1og-11ke1Ihood lj satisfies the equation
2 (Lq - L1) . 2.70554
where 2.70554 1s the cumulative 90% point of the ch1-square distribution
with one c'egree of freedom, which corresponds to a 95% upper limit (one-
sided). This approach of computing the upper confidence limit for the extra
risk A(d), 1s an Improvement on the Crump et al. ( 1977 ) model. The upper
confidence limit for the *xtra risk calculated at low doses 1s always
linear. Vhls 1s conceptually consistent with the linear nonthreshold
concept discussed earlier. The slope, q^. 1s taken as an upper bound of
the potency of the chemical 1n Inducing cancer at low doses. [In the
section calculating the risk estimates, Pt(d) will be abbreviated as P.]
In fitting the dose-response model, the number of terms 1n the poly-
nomial 1s chosen equal to (h-1), where h Is the number of dose groups in the
experiment, Including the control group.
1878A
-13-
10/12/84
-------
Whenever the multistage model does not fit the data sufficiently well,
data at the highest dose are deleted and the viodel Is refit to the rest of
the data. This 1s continued until an acceptable fit to the data is
obtained. To determine whether or not a fit 1s acceptable, the chl-square
statistic
KJ £ <»1 - *1P1>'
" 1., N^d-PO
1s calculated where Is the number of animals In the dose group,
1s the number of animal 1n the 11*1 dose group with a tumor response,
Is the probability of a response in the 1th dose group estimated by
fitting the multistage model to the data, and h 1s the number of remaining
groups. The fit 1s determined to be unacceptable whenever X1 1s larger
than the cumulative 99% point of the chl-square distribution with f degrees
of freedom, wh°re f equals the number of dose groups minus the number of
nonzero multlslage coefficients.
2.1.3.1. SELECTION OF CARCINOGENICITY DATA — for some chemicals,
several studies 1n different animal species, strains and sexes are avail-
able, each run at several doses and different routes of exposure. A choice
must be made as to which of the data sets from several studies to use 1n the
model. The procedures used 1n evaluating the^e data are corslstent with the
approach of making a maximum-likelihood risk estimate. They are listed
below as follews:
1. The turner Incidence data are separated according to organ sites
or tumor types. The set of data (I.e., dose and tumor Inci-
dence) used In the model 1s the set where the Incidence 1s sta-
tistically significantly higher than the control for at least
one t.*st dose level and/or where the tumor Incidence rate shows
a statistically significant trend with respect to dose level.
187BA
-H-
10/12/84
-------
The data set that gives the highest estimate of the lifetime
carcinogenic risk, qi*. Is selected 1n most cases. However,
efforts are made to exclude data sets that produce spuriously
high risk estimates because of a sraaV number of animals. That
1s, If two sets of data show * similar dose-response relation-
ship, and one has a very small sample size, the set ot data
having th* larger sample size 1s selected for calculating the
carcinogenic potency.
2. If there are two or more data sets of comparable size that are
Identical with respect to species, strain, sex and turssr sites,
the geometric mean of qi + , estimated from each of these data
sets, 1s used for risk assessment.
3. If two or more significant tumor sites are observed 1n the same
study, and 1f the data are available, the number of animals
with at least one of the specific tumor sites under cons^dera-
tion 1s used as Incidence data in the model.
2.1.3.2. DERIVATION OF HUMAN EQUIVALENT D0SA6ES — It Is appropriate
to correct for metabolism difference between species and any variation in
adsorption factors by different routes of administration.
Following the suggestion of Mantel a-id Schnelderman ( 1 97 7 ), 1 t 1s
assumed that rug/surface area/day 1s an equivalent dose between species.
Since, to a close approximation, the surface area 1s proportional to the 2/3
power of the weight, as would be the case for a perfect sphere, th? exposure
In mg/day per 2/3 power of the we^ht Is also considered to be equivalent
exposure. In an animal experiment, this equivalent dose
-------
Then, the lifetime average exposure 1s
. 1e x "
" Le x «2/3
2.1.3.3. ESTIMATION OF THE UNIT RISK FROM ANIMAL STUDIES — The risk
5/3
associated with d mg/kg /day 1s obtained from GLOBAL 79, and for most
cases of Interest to risk assessment, can be adequately approximated by
P(d) = 1 - exp (-q^*d). A "unit risk" 1n units X Is simply the rls*
corresponding to an exposure of X * 1. To estimate this value, we simply
2/3
find the number of mg/kg /day co responding to one unit of X and substi-
tute this value Into the *bove relationship. Thus, for example, 1f X 1s in
units of ug/m* In the air, then for case |1), d « U.2S x 701'3 x
2/3
10 * mg/kg /day, and for case \2), d ¦ 1, when yg/m" Is the unit
used to compute parameters 1n animal experiments.
If exposures are given 1n terms of ppm In aV. , we *aay simply use the
fact that
1 ppm 1 2 x molecular weight (gas) mq/m»
molecular weight (air)
Note that an equivalent method of calculating unit risk would be to use
th
mg/kg/day for the animal exposures and then to Increase the ) polynomial
coefficient by an amount
'W*'3 3 - 1. 2 k.
and use the mg/kg/day equivalents for the unit risk values. In the section
calculating the unit risks from animal data, the final q-j* will always bt
the upper-limit potency estimate for humans.
If the duration of the experiment, L^, 1s less than the natural life-
span of the test animal, L, the slope, q^*, or more generally the expo-
nent, g(d). Is Increased by multiplying a factor (L/L^)1. It Is assumed
that If the average dose, d, 1s continued, the age-spec1F1c rate of cancer
will continue to Increase as a constant function of the background rate.
1676A -16- 10/09/84
-------
The age-sp»c1f1c rates for humans Increase at least by the second power of
the age, and often by a considerably higher power, as demonstrated by Doll
(1971). Thus, the cumulative tumor rate would be expected to Increase by at
least the third power of age. Using this fact, 1t can be assumed that the
slope, q or more generally the exponent, g(d), would also Increase by
at least the third power of age. As a resuU, 1f the slope q^ [or g(d)]
1s calculated at age Le, It can be expected that If the experiment had
been continued for the full lifespan, L, at the given average exposure, the
slope q • [or g(d)] would have been Increased by at least (L/l^)1.
This adjustment 1s conceptually consistent with the proportional hazard
model proposed by Cox (197?) and the t1me-to-turnor model considered by Crump
and Watson (1979), where the probability of cancer by age t and at dose d 1s
given by
P(d.t) * 1 - exp [-f(t) x y(d)].
2.1.3.4. INTCftPfiCTA1ION OF QUANTITATIVE ESTIMATES — For several
reasons, the unit-risk estimate based on animal bloassays 1s only an approx-
imate 1nd1ca Jon of 'he absolute risk 1n popala' ions exposed to known car-
cinogen concentrations. First there are Important species differences 1n
uptake, metabolism and organ distribution of carcinogens, as veil as species
differences 1n targe: sUe susceptibility. Immunologic responses, hormone
function, dietary factors and disease. Second, the concept of equivalent
doses fur humans compared with animals on a mg/surface area basis 1s virtu-
ally without experimental verification regarding carcinogenic response,
finally, human populations are variable with respect to genetic constitution
and diet, Uvvng environment, activity patterns and other cultural factors.
The unit-risk estimate can give a rough Indication of the relative
potency of a given agent compered wUi*. other cardnojens. The comparative
1878A
-17-
10/12/84
-------
potency of different agents ts more reliable when the comparison 1s based on
studies 1n the same test species, strain and sex and by the same route of
exposure, preferably by Inhalation.
The quantitative aspect of the carcinogen risk assessment 1s included
here because K may be cf use In the regulatory decision-making process
(setting regulatory priorities, evaluating the adequacy of technology-based
controls, etc.). However, It should be recognized that the estimation of
cancer risks to humans at low levels of exposure 1s uncertain. At be:* the
linear extrapolation model us?d here provides a rough but plausible estimate
of the upper 11m1t of risk (I.e., It 1s not likely that the true risk would
be much more than the estimated risk, but 1t could very well be considerably
lower). The risk estimates presented 1n subsequent sections should n^t be
regarded as an accurate representation of the true cancer risks even w>
-------
assumed to be equal to one. For a uniform diet, the weight of the food con-
sumed 1s proportional to the calories required, which 1n turn 1s proportion-
al to the surface area, or 2/3 power of the weight. Water deifiands are also
assumed to be proportional to the surface area, so that
ih « ppm x x r
or
ppm
rW2/3
As a result, ppm in the diet or water Is often assumed to be an equivalent
exposure between species. However, this 1s not felt to be justified since
the ca1or1es/kg of food 1s very different 1n the diet of man 1n comparison
with that of laboratory animals, primarily because of moisture content
differences. Consequently, the amount of drinking water required by each
species also differs because of the amount of moisture 1n the food. There-
fore, we use an empirically derived factor, f - F/W, which 1s the fraction
of a species' body welgiit that 1s consumed/day as food. The following rates
are used:
Fraction of Body
Weight Consumed as
Species W *Food *Water
Nan 70 0.028 0.029
Rats 0.35 0.0S 0.07B
Mice 0.03 0.13 0.17
Thus, when exposure 1s given as a certain dietary or water concentration 1n
2/3
ppn, the exposure 1n ny/W 1s
BMJL-t PP" ' t » W . ppn x f x „l/3
rw2/3 y2/3 w2/3
1B78A
-19-
10/12/84
-------
When exposure 1s given 1n terms of mg/kg/day > m/Vlr - s, the conversion U
slmplv
- s x «V3
ry2/3
The positive animal cancer data available for calculating a unit-risk
estimate for 2.3,7,8-TCQD Include the following:
1. The Koclba et al, (1978a) diet study on Sprague-Dawley rats.
Spartan substrain. Significant cancers In the males Included
stratified squamous cell carcinomas of the tongue and squamous
cell carcinomas of the nasal turbinates and hard palate.
Significant cancers 1n the females Included lung, nasal turbi-
nate and hard palate cancers, and liver tumors. As with the
males, the total number of animals with at least one of these
significant tumors was recorded.
The NCI gavage study (NTP, 1980a) 1n Osborne-Kendel rats and
86C3F1 mice.
a. 2,3,7,8-TCOD 1n male rats caused an Increase 1n follicular
cell adenomas and carcinomas combined of the thyroid. How
ever, these tumors were not considered biologically signifi-
cant fur risk assessment purposes. In fences, the combined
neoplastic nodules and hepatocellular carcinoma, were consid-
ered significant, and these data were used. The adrenal
cortical aCenomas or carcinomas were not considered biologi-
cally significant.
b. 2,3,7,6-TCCD In male mice caused an Increase In hepatocellu-
lar carcinomas and 1n combined hepatocellular adenomas and
carcinomas. In female mice, 2,3,7,8-TCOD caused an Increase
In subcutaneous tissue fibrosarcomas, lymphomas or leukemlas
of the hematopoietic system, liver hepatocellular carcinomas
and edenomas, and thyroid follicular cell adenomas.
The above data have been fitted to the linearized multistage model
described 1r. th* inethodology section. The data from which the steepest
slope factor (q.,*) (I.e., greatest potency) was calculated were from at<
Independent pathologist's (Dr. R. Squire) review of t'>e Dow Chemical Company
llfetVue rat feeling study. This factor 1s
q^# • 4.25 x 10* (mg/kg/day)"1
1878A
-20-
03/05/85
-------
based on the tumors In female Sprague-Oawley rats. For the purpose of these
calculations, the largest dose group In the study was eliminated because
Inclusion of all of the dose groups resulted 1n a poor fit of the mode 1
{p<0.01). Early increased mortality 1n the high-dose group was also adjust-
ed for by eliminating animals th*t died Curing the first year, so that the
first tumor* considered were those detected during the 13th month of the
study. The results yield acceptable fits of the data wlthou' dropping the
responses at the highest dose levels. The slope estimates for the Kodba
and Squire analyses, 1.51xlOv arid 1.61x10* {mg'kg/aay)**1, were
averaged by taking the geometric mean, and the final estimate thus becomes
q-j* - [{1.51 x 10*) x (1.51 x 10*)]1/2 - 1.56 x 10* Kj/kg/ day)"i.
This up?er-l1m1t estimate represents a range of uncertainty that 1s related
as much to the f1 ting procedure as to the model Itself. The dropping of
the highest dose-response data and the resulting Increased 95% i>pper-l1m1t
slope estimate based on the Squire analysis, can be defended on the basis
that the highest dose data 1n this bloassay Is "i00 times that of the lowest
and wcjld, therefore, contain very UUIe Information about the shape of the
dose-rcsponse curve at low dose levels. It could also b? argued on the
basis of a saturation effect of either dose or response; the data can parti-
ally support either hypothesis. An adjustment of the multistage model
needed *o Incorporate such an effect or effects, howsver, 1s felt to be
unwarranted by the spars 11y of the supporting evidence- As alternative,
to Incorporate th*s uncertainty, a range of 95% upper-limit estimates of
q^* • 9.0x104 to 4.25x10* (mg/kq/dayhas been chosen to accommo-
date this unusual data set.
1878A
-21-
03/0o/8i
-------
In order to estimate a unU risk for a 1 pg/i concentration 1n
drinking water, the following conversion 1s used:
1 yg/kg/day x 70 kg x 10s ng/wg x l day/2 i - 3.5 x TO4 ng/t
based on human consumption of 2 t water/day for a lifetime. Therefore,
unit risk corresponding to 1 ng 2,3,7,6-TCDD/l water U
-------
The use of fish consumption as an exposure factor requires he Quantita-
tion of pollutant residues 1n the edible portions of the Ingested species.
Accordingly, BCFs are used to relate pollutant residues 1n aquatic organisms
to the pollutant concentration 1n the ambient waters 1n which they reside.
To estimate the average per capita Intake of a pollutant from consump-
tion of contaminated f 1 sh and shellfish, the results of a diet survey were
analyzed tc calculate the average consumption of freshwater and estuarlne
f1sh and shellfish. A species 1s considered to be a consumed freshwater or
»stuanne fish and shellfish species 1f at some stage In Its 11fe cycle, 1t
Is harvested from fresh or estuarlne water for human consumption 1n signifi-
cant quantities.
Three different procedures are used to estimate the weighted average BCF
depending upon the lipid solubility of the chemical and the availability of
bloconccntration data.
for 11pld-soluble compounds, the average BCf 1s *alculated from the
weighted average percent lipids in the edible portions of consumed fresh-
water and estuarlne fish and shellfish, whUh was calculated from data on
consumptlor f each species and Its corresponding percer.t lipids lu be 3.OX.
Because the steady-state BCFs for Hp1d-soluble compounds are proportional
to percent lipids, BCFs for f 1 sh and shellfish can be adjusted to the
average percent lipids for aquatic organisms consumed by American"., lor
many 11p1d-soluble pollutants, there exists at least one BCF for which the
percent lipid value was measured for the tissues for which ttit BCF 1s
determined.
With 3.0% as the weighted average percent lipids for freshwater and
estuarlne fish and shellfish 1n the average diet, a BCF, and a corresponding
percent 11p1d value, the weighted average BCF can be calculated.
187BA
-23-
02/27/85
-------
In those cases where an appropriate BCF 1s not available, the equation
"Log beF • (0.85 Log P) - 0.70" can be used to est1r»ate the BCF for aqu -ic
organisms containing about 7.6% lipids from the octanol/water partition
coefficient P. An adjustment for percent lipids 1n the average diet versus
7.X 1s made 1n order to derive the weighted average BCF.
For nonllpld-soluble compounds, the available BCFs for the edible
portion of consumed freshwater and estuarine fish and shellfish are weighted
according to consumption factors to determine a weighted BCf representative
of the average diet.
Therefore, the total Intake 1 can b* written as the sum of two terns:
l(mg/day) - C{mg/I} x R(1/kg fish) x 0.0065 kg fish/day * C(mg/t x 2 I'day)
- c(2 ~ o.noesR)
where C represents the water concentration in mg/t and R ^s the bloconcen-
tratlon factor. Therefore, the water concentration 1n ng/i corresponding
to a lifetime risk of 10~* for a 70 kg person 1s calculated by the formula:
... 70 x 10'*
Water concentiatlon .
qi*(2 ~ 0.0065 R|
for 2,3,7,8-TCDD, the calculated or estimated values for q^* and R are:
1.56 x 10*
qi* «
mg/kg/day
ft = SCiuO (U.S. EPA, 1984a}
and therefore
70 x 10"»
water concentration
1.56 x 10* [2 ~ 0.0065(5000)j
« 1.3 x 10"iO mg/l P 10"* risk level
1B7BA
-24-
05/31/85
-------
2.1.5. Cancer Unit Risk Estates for 2,3,7 8-TCDD via *ht Inhalation
Rout*. The positive animal cancer data available for calculating a unit-
risk estimate for 2,3,7,8-TCOD are as follows:
1. The Koc'ba et al. (^970a) diet study on Sprague-Dawley rats,
Spartan substrain. Significant cancers 1n the males Included
stratified squamous cell carcinomas of the tongue and squamous
cell carcinomas of the nasal turbinates anJ hard palate.
Significant cancers 1n the females Included lung, nasal turbi-
nate and hard palate cancers, and liver tumors have been
observed. As with the males* the total number of animals with
at least one of these significant tumors was recorded.
?. The NCI gavage study (NTP, 1980a) 1n Osborne-Nendel rais and
B6C3F1 mice.
a. 2,3,7,8-TCDD 1n male rats caused an Increase 1n fol-
licular cell adenomas and carcinomas combined of the
thyroid. However, these tumors were not considered
biologically significant for risk assessment purposes.
In females, the combined neoplastic nodules and hepa-
tocellular carcinomas were considered significant, and
these data were used. The adrenal cortical adenomas
or carcinomas were not considered biologically signif-
icant .
b. 2,3,7,8-TCDO in male mice caused an increase 1n hepa-
tocellular carcinomas and 1n combined hepatocellular
adenomas and carcinomas. In female mice, 2,3,7,8-TCOD
caused an increase In subcutaneous tissue fibrosar-
comas, lvnphomas or leukemlas of the hematopoietic
system, and liver hepatocellular carcinomas.
The above data have been fit with the linearized multistage model
described 1n the methodology section.
The data from which th? largest slope factor was calculated were
from an Independent pathologist's (Or. R. Squire's) review of the Dow Chemi-
cal Company (Koclba et al., 1978a) lifetime rat feeding study. This factor
q^ • 4.25 x 10# (mg/kg/day)'1.
This unit-risk estimate from an oral study must be transformed before an
estimate can be mad* of the effect from exposure 1n the ambient air.
1B78A
-25-
05/31/85
-------
Exposure will be assumed to occur only through respiration of dloxln-
contanlna .ed particulates. The amount of exposure depenOs on the particu-
late size distribution. Based on the report of the Tas.' Group on Lung
Dynamics (1966), H can be assumed that 100% of particulates of -0.1 micron
tn size pass the nasopharyngeal (upper respirator) tract) barrier and are
deposited on the tracheobronchial and alveolar passages. For the larger-
stie particles, the percentage deposition of S-m1cron particles In the lover
respiratory tract 1s not >30%. Even those lawyer particles ret3lned by the
upper respiratory tract, however, may be swallowed and eventually absorbed
by Ingestion. In the absence of any specific data on the size distribution
and eventual fate of the particulates, the Information developed by the
International Coirwilsslon on Radiological Protection, Connlttee 2, will be
used. The Committee developed the following estimates for retention of
particulate matter In the lungs, for non-read1ly soluble compounds. 25%
will be exhaled, 50% will be deposited In the upper respiratory passages and
subsequently swallowed, and the final 25% will be deposited 1n the lungs
(lower respiratory passages). Of this final 25%, half 1s eliminated fror#
the lungs and swallowed 1n the first 24 hours, making a total of 62.5%
swallowed; the remaining 12.5% remains 1n the lung alveoli for long periods
of time, with some eventually being transferred to pulmonary lymph nodes.
If we take a worst-case estimate and assume that all of the swallowed
material 1s eventually absorbed Into the body, then we can assume total
uptake of 75% of the lnh?led material. Ue further assume a breathing rate
of 20 mVday for a 70 kg man. Based on the above assumptions and the fact
that 1 pg Is equal to 10"* mg, the lifetime cancer risk for an ambient
concentration of 1 pg/m* of 2,3,7,8-TCDO 1s 3.3x10"* as calculated below:
q^* (resp) - 1.56x10* (mg/kg/day)"1 x IxlO"* mg/pg x 0.75 x 20 m»/70 kg
or q^* (resp) - 3.3x10~4 ipg/m')"1
18?8A -26- 05/31/85
-------
2.1.6. Determination of Relative Poienc/. One of the uses of unit risk
1s to compare the potency of carcinosis. To estimate the relative potency
on a per-mole basis, the unit risk slope factor 1s multiplied by the molecu-
lar weight, and the resulting ni>mber is expressed 1n terms of '.trAol/kg/
day)"1. Th'is 1s called the relative potrnc> Index.
When human data are available for a compound, they have been used to
calculate the Index. When no human data are available, animal oral studies
are selected over animal Inhalation studies because most of the chemical?
have an'.mal oral studies; this allows potency comparisons by route
The potency Index for 2,3,7,8-TCDD based on tongue, lung, and nasal
turbinate ond hard palate tumor*, 1n the female rat 1n the now 2,3,7,8-TCDD
feeding study (Kodba ei a!., 1978a) 1s 5x10* (mMol/kg/day)"1. This 1s
derived as follows: the 95% tpper-Umlt slope estimate from the Dow study
uslrg the Squire review d?ta 1s q^# - 1.56x10* (mg/kg/day}-1, multi-
plying by the nolecular weight of 322 yields a potency Index of 5* 107
Rounding off to the nearest order of magnitude gives a value of 10a.
Ranking of the relative potency Indices Is subjec; to the uncertainty of
comparing estimates of potency of different chemicals based on different
routes of exposure to different species, using studies of different quality.
Furthermore, all the Indices are based on estimate-; of low-dose risk ur1ng
linear extrapolation from the observational range. Thus, these Indices are
not valid for the comparison of potencies in the experimental or observa-
tional range 1f linearity doe: not exist there.
2.2. COMPARISON OF CPA'S RISK ASSESSMENT APPROACH WITH OTHER METHODS
2.2.1. Alternative Methodological Approaches Utilized by EPA for Cancer
Risk Assessment for TCOD. The methods used for quantitative assessment are
consistently conservative (I.e., tending toward high estimates cf risk).
1878A
-27-
04/04/85
-------
The most Important part of the inethodology contributing to tMs conservatism
1n tMs respect Is the linear nonthreshold extrapolation model. There are a
variety of other extrapolation models that could be used, all of which would
give lower risk estimates. These alternative models have not been used 1n
the following analysis but car. be found 1n i?K report S0O/8-84-014A. The
models presented are the one-hit, problt and We 1 bull models. It Is felt
that with the limited data available from these anlma1 bloassays, especially
at the high dosage levels required for testing, almost nothing 1s known
about the true shape of the dose-response curve at low environmental levels.
The risk estimates obtained by use of the linear nonthreshold model are
upper limits, and the true risk could be lower.
Another alternative method Involves the choice of animal bloassay as the
basis for extrapolation. The present approach 1s to use the most sensitive
respon Alternatively, the average responses of all of the adequately
test' .oas'.ay animals could be used.
Extrapolations from animals to humans could also be done on the jasts of
relative wdghts rather than surface are/
-------
wUh 2,4,7,8-TCOD does not pose an unacceptable risk tc public health (FDA,
1981). EPA has reviewed the recent testimony before Congress of Or. S.A.
Wilier (FOA, 1983}, discussing cancer risk associated wUh Ingestion of
these fish. The FOA estimate of the 95% upper-limit carcinogenic potency
factor for 2,3,7,8-TCDD 1s q^-l.75x10* (mg/kg/day)-1, which 1s less
potent than EPA's estimate of q^-1.56x10* (mg/kg/day I"1 by a factor
of 9. Even though both Agencies used the same data base (Kodba et al.,
1978a) and risk extrapolation modei, some subtle differences 1n methodology
exist that account for this factor of 9. The major part of this difference
1s a factor of 5.38 that EPA uses for rat-to-man extrapolation on the
assumption that dose per unit body surface area, rather than dose per unit
bcdy weight, 1s an equivalent dose between species (45 FR 79351). Most of
the remaining factor of ~1.7 1s Cue to the FDA's use of the Koclba's hlstn-
pathologlcal diagnosis alone, without Including that of Squire, and EPA's
adjustment of Its calculations to compensate for the high early mortality
observed In the Kodba et al. (1978a) study.
FOA and EPA also differ In their assessment of human exposure to
2,3,7,8-TCOO 1n fish, 1n keeping with their respective regulatory
approaches. EPA calculates water quality criteria to protect a body of
water as though 1t were the direct source of 100% or a human population's
average dally Intake of water and/or freshwater and ustuarlne fish or shell-
fish. Tne concentration of a pollutant 1n the tissues of all such fish or
shellfish 1s further assumed to be determined by the water concentration and
the 6CF of the pollutant. FOA, on the other hand, premised Us exposure
assessment on the assumption that only limited amounts of fish having
2,3,7,8-TCDD levels at or near the advisory level will actually be consumed.
1B76A
-29-
02/27/85
-------
For example, FDA assumed that for this substance, significant contamination
problems were United to bottom feeders such as catfish and carp. However,
available data Indicate that other species, especially trout and salmon,
taken from some areas of the Great Lakes may also have tissue residues of
?,3,7,8-TCDD 'hat exceed 25 ppt. It al o assumed that actual average resi-
due levels In the flesh of bottom-feeding species reaching the market would
not exceed one-third of the advisory level (I.e., -8 ppt) and further, thH
for most Individuals, 90% of the fish consumed would consist of other
species showing no measurable contamination, or would be taken from uncon-
tamlnated areas. Under these assumptions, and using an upper 90 percentile
value for freshwater fish consumption of IS.7 g/day, the FDA potency esti-
mate yields an upper-limit risk estimate of 2.86x10'* for consumers of
these fish. lf the * exposure assumptions were used with CPA's potency
estimate a somewhat higher upper limit risk of 2.92x10*' would result.
The Center for Disease Control (CDC) has also calculated an upper-limit
potency value for 2,3,7,8-TCDD (Klmbrough et al., 1984). The CDC estimate
1s based on the Squire hlstopathologlcal results, and, like that of FOA,
extrapolates from rat to man on i basis of dose equivalence per unit body
weight. The COC difference from both the EPA and FDA approaches is that th<
curve fit was done, not on administered dose, but on liver concentration at
terminal sacrifice. Also, like FDA, CDC did not adjust for high ear'iy
mortality. The final resuH H that the CDC 95% upper-limit potency value
estimate when converted back to administered dose 1s q^*°3.6x104
(mg/kg/day)"1, which 1s more potent by a factor of 2 than that of FDA and
less potent by a factor of 4 than that of the EPA. The difference between
the EPA and FDA risk estimates results from the difference 1n potency esti-
mates and the use hy FOA of an Average human body weight of 80 kg vs. 70 kg
used by EPA.
1878A -30- 05/31/65
-------
In January 1964 the three Agencies met to review the differences 1n
carcinogenic potency estimation. The three Agencies agreed that they were
using virtually the same methodologies for potency estimation although there
were differences 1n some assumptions used, further, there was agreement
that correction for mortality 1s appropriate, making the differences less
between the EPA estimate and the other estimates. Lastly, the Agencies
agreed that the remaining differences are within thp range of uncertainty
inherent 1n the risk assessment process.
2.2.3. Comparison of Approaches by EPA and CDC for Estimating Soil
Ingestion Exposure. This section explains the differences between the
approaches used by CDC and FPA 1n estimating soil Ingestion exposure to
TCOO. Both Agencies have also analyzed other exposure pathways such as dust
Inhalation, dermal absorption, fish Ingestion, and beef/dairy products
Ingestion. Th15 explanation addresses only soil Ingestion, however, because
U appears to be fliost Important In norma* residential situations and was the
most Influential pathway affecting CCCs selection of 1 ppo as the level of
concern for TCOO 1n residential soils. In actuality, Inhalation contributed
another 0.5% and dermal another 4.0%, based upon the Ingestion value, to the
total CDC exposure estimate.
2.2.3.1. CDC APPROACH (SOIL INGESTION ONLY) -- This discussion 1s
based on a paper by Kimbrojgh et ai. (1984).
The CDC approach Is based on the following equation;
Tutdl 25,550
lifetime - 2 (K'DDt) (IN6t) (G!) (SEASt)
Dcse t * 1
1818A
-31-
04/04/05
-------
where
TCDOt • Concentration of TCDD 1n soil as a function of time
INGt - Ingestion rate of soil as a function of time
61 - G1 tract absorption fraction
SfcASt • Seasonal multiplier as a function of time
t - time
The values for each of the above parameters are explained below.
TCODt - COC assumes that TCDD 1n soil wm degrade with a half-life
of 1? years
Assuming first-order kinetics:
TCDDt . Co exp {-kt)
where
Co ¦ Initial concentration of TCDD 1n soil
k « degradation rate constant
> 0.00016 days"1
t «- time (days)
COC assumes soil Ingestion rate varies with age as shown below:
_ Age Group Ingestion Rate (Q/dav)
0-9 months 0
9-18 months 1
1.5-3.5 years 10
3.5-5 years 1
>5 year; 0.1
LUJ assumes that 30% of the TCDD 1n soil will be absorbed Into the body.
The seasonal multiplier 1s assumed to equal one 1n the warm months
(I.e., Apr 11-September) and zero 1n the colder months (October-March).
Because o* the lack of data in this area, COC supported the Ingestion and
seasonal multiplier assumptions on the basis of best scientific Judgment
rather than specific references.
1878A
-32-
05/31/85
-------
Substituting the above factors Into the orlglna' equation yields a total
lifetime dose of 1100 Co (ng of TCDD) where Co 1s 1n units of ng TCDO/g soil
or ppb. This 1s converted to exposure by dividing by 70 kg body weight and
70 years or 25,550-day lifetime, yielding an average dally exposure of
6.15xl0~4 Co (ng/kg/day).
2.2.3.2. EPA APPROACH (SOIL IK6ESTI0N ONLY) -- This discussion 1s
based on a paper by John Schaun, U.S. EPA (1964b).
The EPA approach 1s based en an equation conceptually similar to the one
used by CDC;
Total
lifetime ¦ (ING) (GI) (ED) (Co) (DEF)
Dose
whers
ING ¦ Ingestion rate
G1 * Gl trad absorption fraction
E0 c Exposure duration
Co ¦ Initial concentration of TCDD 1n soil
DEf « Degradation effects ratio
The parameter value assumptions are discussed below:
The amount of TC00-contam1nated soil that children may Ingest as a
result of normal playing around their home 1s very difficult to estimate.
The Ingestion rates will depend on the mouthing and pica tendencies of the
children.
8ased on measurements of the amount of so11 found on children's hands
and observation* of mouthing frequencies, lepow et al. (1975) estimated that
children could Ingest at least 100 mg of soil per day. This estimate does
not account for direct Ingestion of soil, which could increase dally inges-
tion rates to 5 g/day (Chlsolm, 1982). EPA has adopted this range as a
fKst approximation unless s 1 te-spec 1 f 1 c data are available suggesting more
appropriate values.
1678A
-33
04/04/05
-------
Polger anti Schlatter (1980) found that 13.8-18.2% of the TCOO adsorbed
to son for 8 days reached the liver In 24 hours. Assuming that this repre-
sents 70% of the body burden (Fries and harrow, 1975), the total GI tract
absorption is 20-2bX. NcCornell et al. (1584) also found that the absorp-
tion of TCDO from soil In the GI tract wm "highly efficient" In test
antoals.
For this pathway, the etposure duration represents the number of days
that a child consumes contaminated soil. Obviously this number can vary
trefnendously depending en Individual behavior patterns, access to contami-
nated areas, soil conditions, etc.
The children studied i>y lepow et al. (1975) ranged from 2-6 years old.
lacking other data, It was assumed that this represents the ages for which
mOwtMr'S tendencies and the related lack of understanding of personal
hygiene will cause the most ilg»>lf leant sol i Ingestion.
In a residential setting, behavior patterns and seasonal conditions will
most Influence this parameter. Children who enjoy playing outdoors could
contact soil very frequently. In warm climates such people could contact
soil every oay. In the coldest parts of the United States, such as
Minneapolis, the soil 1s frozen an average of 118 days/year (Baker, 1984).
Although other types of Inclement weather, lllne?*, travel and other factors
could reduce the duration period, no data could be found clearly connecting
these phenomena to the potential for soil contact. Accordingly, the
exposure Juration was assumed to last 247-36? days/year from ages 2-0 for a
total of 1240-1830 days.
The degradation of TCDD 1n soil 1s difficult to measure. Most investi-
gators have found that 1t 1s generally resistant to biological and chemical
degradation, but susceptible to photolytlc degradation (U.S. EPA, 1984a).
1878A
-34-
05/31/83
-------
Young (1983) measured the half-Hfe of TCDD
-------
2.2.3.^. SUMMARY Of DIFFERENCES — Table 1 summaries the differences
between the parameter values chosen by EPA and CDC. The cumulative effect*
of these differences cause EPA's exposure estimate to range froir approxi-
mately one-tenth of CDC's estimate to 10 times COC's estimate. Thus, it
shou'id be noted that the EPA range bounds the CDC estimate by ±1 order of
magnitude. A.so, 1t should be noted that the matr.ymatUal relationships
used to menlpulate these parameters are conceptually the same 1n the CPA and
CDC approach.
2.3. HUMAN HEALTH RISK ASSESSMENT BASED ON CHRONIC NONCARCINOMNIC TOXIC-
ITY OATA {INCLUDING REPRODUCTIVE EFFECTS DATA)
Human health risk assessment from chronic exposure to a che:n1cal
Involves extrapolation of data on adverse responses obstrveu 1n human
epidemiologic studies or experimental animals U pi jject acceptable risk
in humans. Since the level af acceptable risk tn humans should *>e very low,
the degree of extrapolation from high-dosed animal data to projected
Jow-dose human exposure tends to be gre.Jt.
For estimating human health r»sk *rom chronic exposure to a noncardnc-
genlc toxicant, the following graces of threshold toxic responses from
animal experimentations are generally considered;
• No-Observed-Effect Level (NOEL)
• No-Observed-Adverse-EfTect Level (NOAEL>
• Lovest-Observed-Effeet Level UOEl)
• Lowest-Otserved-Adverse-'.ffeci L*ve'i (LOAFL}
VHh a knowledge of projected risk in humans, the ADI of the noncarcinu-
genu chenlcal can be estimated from -lata reflecting one of the abrve toxtc
responses. ADI for a chemical under study *v- calculated from total exposure
data that Include contributions from the del and air.
>878A
-36-
05/31/85
-------
TABLE 1
Susaary o* Parameter Value Differences
f atior
19k
CDC
IPA/COC tatlo
Total ^nges*.eti dose (ngj*
Traction absorbed
Body weight (kg)
Llfetim (years)
Exposure (ng/kg/day)*
100 Co to S200 Co
0.2-0.;*
14
70
5.9x10~» Co to 6.7xl0"» Co
3700 Co
0.3
70
*t0
6.1x10"* Co
0.027-2.S
0.67-0.87
0.2
1
0.10-11.0
*Assumes Co 1s 1n ppb units
-------
In deriving AD I from experimental data considerable scientific Judgment
Is utilized In selecting the appropriate uncertainty factor. Uncertainty
factor {or safety factor) H an Indication of degree of uncertainty that 1s
considered during extrapolation of experimental data, primarily from animal
bloassays, to human system. factors considered 1n selecting the appropriate
uncertainty factor 'n,1ude t.tether the data have been derived from human or
animal experimentations; th* number, species and various aspects tested 1n
case of anlatil experimentation*; quality and utilization of controls both
vehicle and positive; do
-------
In chronic toxicity studies of rats and mice, it was, again, the liver
that appeared to be the mosl sensitive organ. Changes 1n the liver of rats
included Initially fatty Infiltration, and necrosis at higher doses.
According to the rat studies, O.oni Mg/kg/day was a NOF:, while 0.05 and
0.1 yg/kg/day were the NOAEl and PEL, respectively, for liver damage
(Koclba et al., 1978b, 1979: NTP, 1980a). In mice, a NOEL was not deter-
mined, and the lowest doses tested, 0.0015 and 0.006 »ig/kg/day, produced
liver damage 1n male and female B6C3F1 nice (NTP, 1980a), while the lowest
dose i.sted 1n Swiss mice, 0.001 vg/kg/day, produced amyloluosU of the
kidney, spleen and liver (Toth et al., 1978, 1979). In nonhuman primates,
chronic exposure to 2,3,7,8-TCDO 1n the diet at 50 or 500 ppt resulted 1n
hair loss, edema and pancytopenia (Allen et a1.t 1977; SchsnW et al.,
1979). Data were not available to determine a NOEL for monkeys. The guinea
pig, the most sensitive species to the acute toxic effects of 2,3,7,8-TCDO,
has not been used 1n a chronic b1oassa>.
2,3,7,8-TCDD has been demonstrated to be teratogenic 1n all strains of
mice tested (Courtney et al., 1970; Courtney and Moore, 1971; Courtney,
1976; Neubert and Plllman, 1972; Smith et al., 1976). Poland and Glover
(1980) demonstrated tnterstraln differences In the Induction of terata.
Responsive mice, containing high levels of the £t\ receptor, are highly
susceptible to the effects of 2,3,7,8-TCDO 1n producing clert palate, where-
as the nonresponslve mice, which contain low (or 0) levels of the Ah
receptor protein, are resistant to this teratogenic effect of 2,3,7,8-TCDD.
The most common malFormations observed were cleft palate and kidney cnom-
ailes; however, occasionally, other malformations have been reported. Smith
et al. (1976) reported that no terata were observed at 0.01 vg/kg while
cleft palates were increased at 0.1 vg/kg and su^sted that 0.1 >»g/kg
was the minimum effective dose for mice. In rats, 2,3,7,8-TCDO has produced
18? BA -39- 05/31/85
-------
fetotoxtc effects And kidney malformations (Courtney and Moore, 1971; Spars-
ehu et al., 1971; Khera and Ruddlck, 1973; G1av1n1 et al., 1982a; Hurray et
al., 1979}. Fetotoxldty has also be?n observed 1n rabbits (Glavlnl *t al.,
1982b; Norman et al., 1976), ferrets (Muscarella et al., 1982) and monkeys
(Schantz et al.. 1979; Allen et al., 1979; Barsottl et al.. 1979; NcNulty,
1978). The three-generation study of Murray et al. (1979) reported a NOAEL
for rats of 0.001 uv/k$/ddy with the next higher dose of 0..-1 yg/kg/day
demonstrating Increase* 1n anomalies. Nlsbet and Paxton (190?) used differ,
ent statistical methods to evaluate this three-gvneratlori study did con-
cluded that the lowest dose produced effects was actually a 10All. This
LOAEL for reproductive effects 1s similar to the LOAEL for Mver damage and
amyloidosis observed In chronic toxicity studies 1n mice (NTP, 1980a; Toth
et al., 1978, 1979).
There seems to be general agreement that human exposure to 2,2,7,8-TCDD,
whether acute or chronic, leads to chloracne, altered liver function,
hematological abnormalities, porphyria cutanea tarda, altered pigmentation,
hirsutism and some peripheral neuropathy (Poland et al.. 1971; Ott et al,
1980; Pazderova-VeJIupkova et a1., 1981; Singer et ai., 1982). A number of
studies, mostly correlation studies, have been conducted on groups of
persons exposed to 2,3,7,8-TCDD to assess reproductive hazard. Although
some studies have shown a positive association between exposure to 2,4,5-T
and birth defects or abortions (Hanlfy et al., 1981) other studies have
failed to demonstrate any association. In Investigations concerning
potential exposure to 2,3.7,8- TCDO through the manufacture of TCP. there
has been no positive substantiated association between exposure and repro-
ductive difficulties. In these studies, exposure was always to a mlr.ture of
compounds, with 2.3,7.8-TCDO being a contaminant. Hence, 1t 1s not possible
1878A
-40-
05/31/85
-------
to IttMbute with certainty any positive finding to 2,3,7,8-TCDO. It 1s
Also possible, since levels of 2,3,7.0-TCDO contamination of 2,4,5-T and TCP
were only estimated, that the negative results reflect exposure levels *00
low or study designs too Insensitive to elicit a detectable response.
2.3.2. Estimation of Acceptable Dally Intake for TCOD {for Comparative
Purposes Only). 2,3,7,8-TCQD 1s an unusually toxic compound with demon-
strated acute, subacute and chronic effects m animals and man. Acute or
subchronlc exposures to 2,3,7,8-ICOD can adversely affect the skin, the
liver, the nervous system and the tmune sy*;em.
2,3,7,8-TCOD displays an unusually high degree of reproductive toxicity.
It 1s teratogenic, fetotoxlc and reduces fertility. In a three-generation
reproductive study, Murray et al. (1979) reported a reduction 1n fertility
after dally dosing at 0.1 or 0.01 wg 2,3,7,8-TCDOAg 1n the and
generations of Sprague-Oawley rats. In addition, equivocal adverse effects
were seen at the lowest dose (0.001 pg/kg/day); this dose, herefore,
represents a I CAE L. Schantz et a1 . (1979) found reductions 1n fertility and
various other toxic effects 1n rhesus monkeys fed a SO ppt 2,3,7,t -TCOD diet
for 20 months. This corresponds to a calculated dally dose of 0.0015
2,3,7,8-TCQD/kg/day. These results suggest that monkeys may be somewhat
more sensitive than rats, since tiie effects 1n monkeys were more severe ano
not equivocal.
A toxldty-based criterion for 2,3,7,8-TCDD has been calculated for com-
parison with the cancer-based criterion (U.S. EPA, 1984a). Since the study
by Schantz et al. (1979) supports the findings of Murray et al. (1979), It
seems reasonable to determine an AO I based on th«j LOAfl. If one selects an
uncertainty factor of 100 based on the existence of lifetime animal studies
and the knowledge that there 1s suggestive evidence of reproductive effects
1878A
-41-
P5/31/85
-------
1n man (Han'ify et al., 1981), as per U.S. EPA (1980) guidelines, and then an
additional 10 because of a LOAEl 1s used as the bcsls of this calculation.
The AOI thus calculated would be:
A0I c lOlLvq/kq/day (LOAEL) . w,kg/dey.
100 x 10
According to the nethods published by U.S. EPA (I960), .in additional uncer-
tainty factor between 1 and 10 must be used becau:e the calculation 1s based
on a LOAEl. An uncertainty factor of 10 was chosen because of the adverse
effects seen 1n rhesus monkeys at 0.0015 yg/kg/da*. despite the equivocal
nature of the effects 1n rats seen at the 0.001 w9/kg/day ~os& level. The
AOI for a 70 kg nan would be 7.0xi0~s wQ TCPD/day.
2.4. HUMAN HEALTH RISK ASSESSMENT BASED ON MUTAGENICITY DAU
2.4.1. Mutagenicity Data Available on TCOO. A limited number of Initial
studies on the mutagenicity of 2,3,7,8-TCOO in bacteria reported* positive
results 1n S. tvphlmurium strain TA1532 1n the absence of a fnawmallan neta-
bollc activation system (Hussaln et a 1 ., 1972; Seller, 1973). More recent
attempts to repeat these results with strain TA1532 or reU/ced strains have
failed (Selger and Neal, 1981; Nebert Ft al., 1976; Gilbert et al., 1980;
HcCann, 1976). These authors also reported no Increase 1n mutation rate
when 2,3,7,8-TCOD was tested 1n the presence of a mammalian metabolic
activation system. In other Vn vitro assays, 2,3,7,8-TCDD Ms produced a
poslt've response 1n reunion to streptomycin Independence in E. coll Sd-4
cells and questlonatle positive response with prophage Induction in E. col 1
K-39 cells (Hussaln et al., 1972). Also, 2,3,7,0-TCDD has been reported to
be mutagenic 1n the yeast S. cerevlslae 1n both the In v1fcro assay with S-9
and the host-noJIated assay (Bronzettl et al., 1983). Rogers et al. (1982)
also reported positive mutagenicity results 1n the mouse lymphoma assay
system. In the E. co 11 studies, the poor survival of the cells or the
1878A
-42-
04/08/65
-------
Interference of the vehicle solvent, DHSO, with the assay makes the evalua-
tion of the studies difficult. With the data available. 1t 1s not possible
to resolve the conflicting reports on the mutagenic potential of 2,3,7,8-
TCDO.
Overall, the data Indicate little potential for the Interaction of
2,3,7,8-lCDD with nucleic acids or the ability of 2,3,7,8-TCOO to produce
chromosomal aberrations.
-------
The differences among the results reported could be due to several
factors, such as treatment protocols, solubility problems, purity of the
samples tested and the hlg'i toxicity of 2,3,7,6-TCOD. T!i1s chemical may be
a weak mutagen, but because It 1*. very toxic, the dose range for detecting a
positive genetic effect may be very narrow. Therefore, additional experi-
mentation 1s necessary before any conclusive determination can be made.
Suggested further testing Includes additional studies of the ability of
?,3,7,8-TC0D to Induce forward nutation* 1n mammalian cells 1n culture,
additional yeast and bacterial stud'.^s and the sex-Unked recessive lethal
test 1n Drosoohlla.
EPA follows the procedures outlined In the Proposed Guidelines for
Mutagenicity Risk Assessment (1984) for conducting qualitative and quanti-
tative mutagenicity risk assessments. In the case of 2,3 7.Q-TC0D, the
Information available 1s too limited and conflicting for applying quantita-
tive risk assessment methodologies. Qualitatively, the presently available
data provide limited evidence that TCDD poses heritable rlsx for humans.
The following 1s a description of the methodology EPA use: 1n qualitatively
assessing mutagenic risk.
Pertinent Information regarding the mutagenicity of PeCDOs and HxCDPs
were not located 1n the available literature.
2.4.2. Qualitative Mutagenic Risk Assessment Methodology. The evidence
for a chemical's ability to produce mutations and to Interact with the
germinal target are Integrated Into a we"ight-of-evidence judgment that the
agen*. may pose a hazard as a potential human germ-cell mutagen. All Infor-
mation bearing on the subject, whether indicative of potential concern or
not, are evaluated. Whatever evidence may exist from humans 1s also
factored Into the assessment.
1879A
-44-
05/31/65
-------
Bacteria and eukaryotes, '.r.cVjd'.r.j fungi, plants, Insects and various
namrdlian systems, are comnonly utilized for assaying potential mutagenic
anti cytogenetic aberrations that are due to exposure to a test chemical.
Table 5-1 1n the NAS Connlttee on Chen;1cal Environmental Mutagens report
(1962) lists these assays.
Information available will vary greatly from chemical to chemical
because there are many mutagenicity test systems, and there has been no
systematic attempt to develop Information on all chemicals of concern. The
responses noted for different tests may also vary from chemical to chemical
since often one does not find consistent positive or negative results across
all tests. ChemlcaU may show positive effects for sorne endpolnts 1n some
test systems, but negative responses 1n others. Each review takes Into
account the limitations in the testing and In the types of responses that
may exist.
Certain responses In tests that do not measure well-defined mutagenic
endpolnts (I.e., SCE Induction 1n maaiwllan germ cells) or germ-cell tests
1n higher eukaryotes (i.e., Droscphlla tests) may provloe a basis for
raising the weight of evidence from one category to another.
Sufficient evidence for potential human germ-ce- nutagenlclty Includes
cases m which positive responses arc demonstrated In a mamr*al1an ger-.-cell
test. Also, In general, sufficient evidence exists when there 1s confirmed
mutagenic activity In other test system* (positive responses 1n at le**t two
different test systems, at least one of which 1s 1n mammalian cells*, and
there 1s sufficient evidence fir germ-cell interaction.
Suggestive evidence encoi::OuSses a we1ght-of-ev1dence category between
sufficient and limited that Includes cases In which there Is some evidence
for mutagenic activity and for Interaction with germ cells.
187BA
-45-
05/31/85
-------
Limited evidence for potential human germ-cell nutager.Uity exists when
evidence H available only for mutagenicity tests (other t'an manvnallan germ
cells) or only for chemical Interactions In the gonad.
1878A
-46-
04/04/85
-------
3. RISK ASSESSMENT STRATEGY FOR ACUTE EXPOSURE TO 2>3,7.8-TCOD
Methodology for the assessment of health hazards associated with expo-
sure of human populations to chemical contaminants 1n the environment has
been developed recently by organizations such as the U.S. EPA and the
National Academy of Sciences (NAS). Since the purpose of these assessments
was to derive health-based criteria for pollutants 1n ambient and drinking
waters, which would designate levels of the pollutant that would be safe for
humans even with lifetime exposure* the emphasis of these methodologies has
been placed on prediction and assessment of effects from chronic exposure.
When less than chronic exposure 1s addressed, 1t 1s usually In reference to
extrapolating health effects data obtained from short duration exposures to
anticipated effects and associated doses following chronic exposure.
With today's new awareness of chemical hazards in the environment, 1t
has become apparent that situations exist where lifetime exposure to a
pollutant would not be anticipated. Examples of such situations Include
contamination of a drinking wMer system, which 1s followed by remedial
action to eliminate the contaminant, or exposure to toxicants from dump
sites during clean-up operations. As a result of this need. Initial steps
are being taken to develop hazard assessment metnodology concerned with less
Ihan lifetime exposures. As Indicated by the Office of Drinking Water, the
development of criteria (HAs) allowing greater levels of contamination of
water 1f exposure 1s assumed to be of short duration does not mean that the
prerence of these higher levels of contaminants 1s condoned. Rather, the
HAs are prepared 1n order to provide useful Information on the potability of
a drinking water supply 1n cases where contamination occurs.
167BA
-47-
05/31/05
-------
An approach similar to that described for derivation of an ADI for
chronic exposure 1s used for determination of a 1-day HA. The 1-day HA it
Inter.ded to represent the concentration 1n water that will not cause adverse
health effects after exposure for a single day 1n a 10 kg child consuming
1 l of water/day or a 70 leg adult consuming 2 l of water/day. The 10 kg
child 1s used because a child consumes a greater amount of water/body weigh;
than an adult and represents a more sensitive merrber of the population. If
sufficient data are available to differentiate between the sensitivity of
children and adults, separate assessments will be conducted. Since HAs are
developed only for transient periods of exposure rather than for lifetime,
the carcinogenic potential of a compound and the estimated risks associated
with exposure to that compound are not taken Into consideration. If the
qualitative evaluation of carcinogenicity data Indicates that a chemical
might be a potent carcinogen, an HA for this chemical would be provided only
with great caution.
An HA 1s based upon the identification of an adverse health effect
associated with the most sensitive noncarclnogenic endpolnt of toxicity.
The induction of this health effect 1s related to a particular dose of the
substance given over a specified period of time 1n a human or animal study.
To estimate a level at which no adverse effects would be expected to occur
1n members of the human population, an appropriate UF applied to the LOEl
or NOACL Identified 1n these studies. For a 1-day HA, studies of durations
from 1-14 Jays are used.
The 1-day HA 1s calculated by dividing the dose (d, 1n mg/kg bw) oo-
talned from a study that most clcsely defines a NOAEl for the most sensitive
1878A
-48-
02/27/85
-------
endpolnt 1n the most sensitive species t>y a UF using the following guide-
lines:
IP - for a human no-ooserved-adverse-effect level
103 - for a human adverse-effect level
100 - for an animal no-observed-adverse-effect level
10JO - for an anlnal adverse-effect level (lowest)
The HAs are then calculated as follows:
HA « (c! x bw) ~ {UF x wc)
where
bw > body velght, 10 kg for child
70 kg for adult
wc * water consumption, 1 l for child
2 i for adult
Acute lethal data will not be used to <2*r1ve an HA.
Pharmacokinetic data will be considered 1n a manner slmMar to that
described for derivation of an A0I from chronic exposure. The data avail-
able may actually be more relevant in assessing the appropriateness of acute
studies, sine roost pharmacokinetic studies are of short duration. As dis-
cussed tirptfiously, when pharmacokinetic data are similar between species,
spectes-to-sper1es extrapolation can be performed with greater confidence,
likewise, pharmacokinetic data ran provide confidence that a route-to-route
extrapolation 1s appropriate cr snggest that caution 1s required in perform-
ing such manlpulatlcns.
3.1. TOXICITY DATA AVAILABLE FOR SHORT-TERM ACUTE tXPOSURE TO 2,3,7,8-TCDD
The characteristic effects of exposure to 2,3,7,8-TCDD are thymic
atrophy antf weight lo:>s. In rats and rabbits, and to a lesser extent in
guinea pic;s and monkeys, liver damage 1s a major pathological symptom.
1P78A
-49-
02/27/85
-------
Death U preceded by a prnlanged period of weight loss, during which severe
deterioration of the animals 1s observed; hrwever, no specific lesion has
been Identified as the cause of death. Oeath occur* only after 2-3 weeks
following an acute exposure. This unusual characteristic of 2,3,7.8-TCOO
toxicity has resulted In many short-term studies that report only minor
effects at doses near, or sometimes many-fold greater than, the LD5Q.
2,3,7,8-TCDD 1s also an Imune suppressant in mice, rats and guinea pigs.
The acute toxicity of 2,3,7,8-TCOO 1s extremely species-dependent.
Acute oral LDggS ranging from 0.6 vgAg bw for male guinea p1 gs (Schwelz
et al., 1973) to 5051 yg/kg bw for hamsters (Henck et a 1«• 1981} have been
reported. The relative sensitivity of man, compared with other species, to
2,3,7,8-TCDD toxicity cannot be determined fr*n the existing data.
four studies were found that Identified HOAELs or l.OAEls that could br
useful In the derivation of an HA (Harris et al., 1973; Hadge, 1977; Smith
et al., 1981; Turner and Collins, 1983).
Harris et al. 41973) administered a single oral dose of 2,3,7,8-TCDD 1n
acetone:corn oil to groups of CD rats of mixed si-x. Weights were determined
at least once each week. Rats given SO or 100 ng/kg t>w demonstrated a
decreased weight gain and Increased mortality. In the high-dose group,
mortality approached SOX with a mean time Interval until death of 18.3 days.
A dose of 2S pg/kg bw, the LOAEL 1n this study, resulted 1r a decreased
body weight In females at 1 week postdostng and a decreased rate of weight
gain In males for 2 weeks postdoslng. After 2 weeks, both male and female
rat; gained weight at the same rate cs the controls. Doses of 1 or 5
v9/kg bw had no effect on body weight.
Hadge (1977) Investigated the effect of 2,3.7,8-TCOD on Intestinal
absorption 1n CD-I mice. Mice were given single oral doses of 10, 25, 75
1878A
-50-
02/27/85
-------
'SO, 200 or 300 pg 2,3,7,0-TCOD/kg bw. Absorption of 0-glucose, 0-galac-
tose, t-argen1ne and L-h1st1d1ne was measured 7 days later, using the evert-
ed Intestinal sac technique. Absorption of D-glucose was decreased at all
dose levels; however, absorption of the other compounds was not affected oy
any of the treatment. The decrease 1n D-glucose absorption was dose-related
over the range of 0-75 tig/kg bw. In this study, 10 pg/kg bw constituted
a L0AU.
S.n)th et al. (1981) Investigated the effect of 2,3.7,8-TCDD on hepatic
porphyrin levels 1n C57BV10 and DBA/? mice. A single oral dose was admin-
istered 1n arachH oil (0, 5, 15, 50, 75. 150, 300, 600 or 1200 wg/kg bw)
and hepatic porphyrin levels were determined at Intervals for up to 1?
weeks. There were large strain differences 1n susceptibility porphyria
Induction, with the C57BV10 strain being ~20 t'mes as sensitive as the
DBA/2 strain. In this study, the lowest d.se that Induced porphyria was 50
ug/kg bw. Thus, 50 »g/kg bw was a LOAti and 15 wg/kg bw represented a
NOAEL.
Turner and Collins (1983) administered s1r.;1e oral doses of 0.1, O.S,
2.5, 12.5 or 20 ^g/kg bw of 2,3,7,G-TCuD to group? of 4-fc female guinea
pigs. Survivors were killed 42 days after dosing and examined for histo-
pathologic changes In the 11*er. four of the b animals 1n the Mghe*t dose
group and 1 of 5 1n the 12.5 pg/kg group died before the end 5f the
observation period. N1ld histopathologic changes including steatosis (fatty
change), focal necrosis, and cytoplasmic degeneration were noted 1n animals
from all treated groups, but not In controls, and were not dose-related.
All of the IQAtLs and MOAEls determined for rats and mice are abo^e the
LOjq for guinea pigs (0.6-2.1 ug/kg). Although no NOfL or NOALL is
available for guinea pigs, a 10AEI of 0.1 vg/kg can be derived frow the
study of Turner and Collins (1983).
1878A -51- 02/27/85
-------
3.2 ESTIMATION OF 1-OAY HEALTH ADVISORY
The data on very short-terr? exposures. 1-14 days, ate sufficient for the
derivation of an HA. As discussed in Section 3.1., the available studies
established LOaELs for i ats <>nd nice that are greater than the LD^q for
guinea pigs. A NOAEL 1s not a^d1>ab1e for guinea ?1gs, but a LOAEL of 0.1
ti^/kg can be derived from the study of Turner and Collins ( 1983). This
LOAEL can a used to calculate a 1-day Health Advisory, using a UF of 1000
for an anlTial LOAFL:
HA ¦ dob x bw + (I'F x wc)
where
bw « body weight, 70 kg for adult
10 kg for child
wc •> water consumption, 2 l *or adult
1 i for chiid
Thus, for an adult:
1-day HA (adult) - 0.1 yg/kg x 70 kg bw) * (1000 x 2 I)
- Q.OOjS vg/l (0.007 wg/day or 0.0001 wQ/'kg bw/day)
For a chiid:
1-day HA (child) - (0.1 Mg/kg bw x 10 kg bw) ~ (1000 x 1 I)
• 0.001 (0.001 pg/day or 0.0001 yg/kg bw/day)
1876A
-52-
02/27/85
-------
4. RISK ASSESSMENT APPROACH FOR SHORT-TERM (SUBCHRONIC)
EXPOSURE TO 2,3,7,B-TCD0
In situations where exposure to a toxicant will contlnuv for a few days,
a 10-day HA 1s used as the basis for decision-making on the safety of drink-
ing water. To provide a margin o? safety for cumulative effects of the
toxicant, studies of 30-90 days duration In humans or experimental animals
are used tc der1v« the 10-day HA rather than a study of *10 days' (Juration.
Using studies of appropriate duration, the 10-day HA 1s calculated by the
same method as the 1-day HA. It would be anticipated that the advisory
would become progressively more conservative as the study used approached 90
days. This short-term advisory Is also calculated for Doth the 10 kg child
and the 10 kg nen.
In the abser«.e of appropriate studies, the 10-day HA can also be calcu-
lated by dividing the 1-day HA by '.0 This method should provide protection
from toxicants that have cumulative effects. It should be noted that a
1-day HA cannot be calculated by multiplying the 10-day HA by 10. This form
of extrapolation has no scientific basis, and nonlinear associations between
effect and total dose may result 1n highly toxic doses at the level of the
projected 1~day HA.
As 1s apparent from the d'«uuss1on on the 1- and 10-day HAs, the method-
ology for determining safe levels of xenoblotlcs for less than lifetime
exposures has not heen fully developed. One of the greatest areas of
concern Is how to address data fcr a chemical that 1s known tc be an animal
carcinogen. The development and validation of tnls hazard assessment
methodology will be of concern to regulatcry agencies In the future.
1B?*A
-53-
02/27/85
-------
4.1. TOXICITY DATA AVAILABLE FOR SUBCHRONIC EXPOSURE TO TCOO
There H less Information on species variability In response to sub-
chronic dosing than was available for acute exposures, in a subchronlc
gavage study (NTP, 1980a), BbC3M mice were 10 times more susceptible to the
Induction of toxic hepatitis than were Osborne-Mc.idei rats. In other
studies, the lowest reported LOAEl 1n rats (0.001 yg/kg bw/day) (Murray et
a 1 ., 1979) 1s similar to a L0AEL that has been reported for guinea pigs
(0.0057 pg/kg bw/day) {Vos et al., 1973). Hurray et al. (1979) Investi-
gated the teratogenic and reproductive effects of 2,3,7,8-TCDO 1n Sprague-
Oawley rats 1n a three-generation study. The animals were maintained on
diets that resulted in doses of 0, 0.001, 0.01 or 0.1 uy 2,3,7,8-TCDn/kg
bw/day. The fg rats were maintained on the treatment diets for 90 days
before the Initial mating. This group was mated twice, resulting 1n the
f^ and f1B generations. The f^ and f^ rats were mated at -130
days of age, producing the f^ and Utters, respectively. Under these
conditions, doses of 0.1 and 0.01 pg/kg bw/d-iy produced reduced fertility
and fetal survival. The authors reported a HOAEl of 0.001 yg/kg bw/day.
Nlsbet and Paxton (1962) reevaluated this study, using different statistical
methods. They concluded that the 0.001 vg/kg bw/day dose level resulted
In a significantly reduced gestational Index, decreased fetal weight,
Increased 11ver-to-body weight ratios and an Increased incidence of dilated
renal pelvis. Thus, they classified 0.001 vQ/kg bw/day as a 10AEI rather
than a HOAEL. Since at least some f tnese effects may be the result of
exposure during gestation, 1t Is appropriate to consider this study 1n the
calculation of short-term as well as long-term has; however, the difficulty
1n separating the effects from short-term exposure during gestation from
effects that are due to longer-tern exposure makes this study Inappropriate
for the derivation of a 10-day HA.
1676A
-54-
02/27/85
-------
Vos el al. (1973) administered 8 weekly doses of 0.006, 0.04, 0.2 or 1.0
u3 2,3,7.B-TCD0/kg bw (0.0011, 0.0057, 0.029 or 0.14 wg/kg bw/day) In
ac«tone:cc.rn oil to groups of 10 female Hartley guinea pigs by gavage. Body
weights were determined weekly. The effects of 2,3,7,B-TCDD on the Irmune
system were determined by measuring the response to a subcutaneous injection
of tetanus toxoid (humoral 1«munHy) and the delayed-type hypersensitivity
to tuberculin (cell-mediated IntnunUy). The lowest dose at which an effect
on the Innune response was observed was 0.0057 wg/kg bw/day; the NOAEL was
0.0011 yg/kg bw/day.
4.2. ESTIMATION OF 10-DAY HEALTH ADVISORY
Two studies havr reported adverse effects from subchronH exposures to
very low levels (<0.01 wg/kg b*/day) of 2,3,7,8-TCDD (Hurray et al., 1979;
Vos et al., 1973). The usefulness of the Murray et al. (1979) study tor the
development of a 10-day HA 1s limited by the design, which makes It diffi-
cult to distinguish between effects produced by .hort-term exposure during
gestation and effects from longer-term exposure*.
The study by Vos et 4l. (1973) defines a NOAEL of 0.0011 yg/kg bw/day
and a LOAEL of 0.00S7 wg/kg bw/day m guinea pigs, the most susceptible
species to the acute effects of 2,3,7,8-TCDO. These values agree well with
the KOAEL/IQAU derived from the Hurray et al. (1979) study. Using an UF of
100 for an animal MOAEl, a 10-day HA can be calculated from the repnrted
NOAEL for guinea pigs (0.0011 yg/kg bw/day):
HA - (dose x bw; + (UF x wc)
where
bw • body weight, 70 kg for adult
10 kg for ch1Id
wc > water consumption, 2 i for adult
1 I for child
1B7BA
-55-
02/27/35
-------
Thus, for an adult:
10-day H* (adult) - (0.0011 vg/kg bw x 70 kg bw) ~ (100 x 2 1)
• 0.00039 pg/l (0.00076 vg/tfay or 0.000011 vg/kg bw/day)
For a child:
10-day HA (child) - (0.0011 yg/kg bw x 10 kg bw) + (100 x 1 1)
• 0.00011 v
-------
S. REFERENCES
Alien, J.R., O.A. Bar so 111, L.K. lambrecht and J.P. Van N11 ler. 1979.
Reproductive effects of nalogenated aromatic hydrocarbons on nonhuman
primates. Ann. NY Acad. Scl. 320: 419-425.
Allen, J.R., O.A. Barsottl, J.P. Van Miller, L.J. Abrahamson and J.J.
LalUh. 1977. Morphological changes In monkeys consuming a diet containing
low levels of 1C0D. Food Cosmet. Toxicol. 15: 401-410.
Axelsor, 0., L. Sundell, K. Anderson, C. Edllng, C. Hogs tent and H. IC11 ng.
1980. Herbicide exposure and tumor moitallty: An updated epidemiologic
Investigation on Swedish railroad workers. Scand. J. Work Environ. Health.
6: 73-79.
Barsottl, 9.A., L.J. Abrahamson ang J.R. Allen. 1979. Hormonal alterations
In female rhesus monkeys fed a diet containing 2,3,7,8-tetrachlorod1benzo-£-
dloxln. Bull. Environ. Conlam. Toxicol. 21: 463-469.
Berry, D.L., J. D1G1ovann1, H.R. Juchau. W.M. Bracken, 6.L. Reason anc T.J.
Slaga. 1978. Lack of tumor-promoting ability of certain environmental
chemicals 1n a two-stage mouse skin tumoMgenesIs assay. Res. Comun. Chem.
Pathol. Pharmacol. 20(1): 101-106.
1078A
57-
03/05/80
-------
Berry, O.L., T.J. Slaga, J. D1G1o*rann1 and M.R. Juchau. 1979. Studies with
chlorinated dlbenzo-p-d1oxlns, polybromlnated blphenyls, and polychlorlnated
blphenyls 1n a two-stage system of mouse skin tumor1genes1s: Potent anti-
carcinogenic effects. Ann. NY Acad. Stl. 320: 405-414.
Bronzettl, 6., C. Bauer, C. Corsl, R. Del Carratare, R. Neerl and H.
Paollne. 1983. Mutagenicity study of TCDO and ashes from urban Incinerator
¦In vitro' and "In >'1vo" using yeast 07 strain. Chemosph^re. 12: 549-553.
ChWolm, J. "982. Personal corrvnunlcaMon. November.
Cohen, G.M., W.M. Bracken, R.P. Iyer, O.L. Berry, J.K. Selkirk and T.J.
Slaga. 1979. Antl^ardnogenlc effects of 2,3,7,8-tetrach1orod1benzo-{>-
dloxln on benzo(a4pyrene and 7, i2-d1methyltenz{a)ant*racene tumor Initiation
and Us relationship t. DNA binding. Cancer Res. 39(10): 40?7-4033.
Convnltiee on Chemical Environmental Mutagens. 1982. 'dentlfylng and
estimating the genetic Impact of chemical environmental mutagens. National
Academy Press, Washington, 0C.
Cook, R.fc., J.C. Townsend, M.G. Oil and L.6. S1lv«rste1n. I960. Mortality
experience of employees exposed to 2,3,7,8-tetrach1orod1benzo-£-d1ox1n. J.
Occup. Med. 22(8): 530-532.
Courtney, K.D. 1976. Mouse teratology studies with chlorodlbenzo-g-
dloxlns. Bull. Environ. Contam. T0x1col. 1Mb): P674-681.
1878A
-58-
02/27/85
-------
Courtney, K.D. and J.A. Moore. 1971. Teratology studies with 2,4,S-T and
2,3,7,8-TCOD. Toxicol. Appl. Pharmacol. 20: 396-403.
Courtney, K.O., O.W. Gaylor, K.D. Hogan, M.L. Fa Ik« R.R. Bates and I.
Mitchell. 1970. Teratogenic evaluation of 2,4,5-T. Science. 166: 864-866.
Cox, C.R. 1972. Regression model and Hfe tables. J. Roy. Stat. Soc.
34: 187.
Crump, K.S. and W.W. Watson. 1979. GLOBAL 7'); A Fortran program to extrap-
olate dlchotomous animal card logentcHy data to low dose. Natl. Inst, of
Environ. Health Science, Contract No. I-ES-2123.
Crump, K.S., H.A. 61vess and L.L. Deal. 1911. Confidence Intervals and
test of hypothesis concerning dose-response relatl >r.s Inferred from an 1 fr*a 1
carcinogenicity data. Biometrics. 33: 437-451.
Cielzel, i. and J. Klraly. 1976. Chromosome examinations 1n workers
producing Klorlno! ano Bumlnol. in: The Development of \ Pesticide as a
Complex Scientific Task, L. Bankl, Ed. Ned1c1na: Budapest, p. 239-256.
(Cited In NRCC, 1981a)
D1G1ovann1, J., A V1aje, D.l. Berry, T.J. Siaga and M.R. Juchau 1977 .
Tumor-1n1t1at1ng ibiUty of 2,3,7,8-tetrachlorodibenio-£-d1ox1n (TCDD) ana
arocMor 1254 1n the twc-stage system of mouse skin carcinogenesis. Bull.
Environ. Contain. Toxicol. 18(5): 552-5S7.
1878A
-59-
02/27/85
-------
01S1ovanr,1, J., D.L. Berry, fl.R. 3jcr.au and T.J. Slaga. 1979. 2,3,7,8-
Tetr«ch1orodibenzo-{)-d1ox1n: Potent ant1carc1nogen1c activity 1n CO-1 mice.
Blochem. 81 op^iys. 8es. Commun. 66(3): 577-584.
D161ova.in1, J., D.L. Berry, G.L. Gleason, r».S. Klshore and T.J. Slaga.
19b0. Time-dependent Inhibition by 2,3,7,8-tetrachlorodtbenzo-])-d1ox1n of
sMn tumorlgenesU w1lh polycycUc hydrocarbons. Cancer Res. 40(5):
15BO-1587.
Dell, R. 1971. The age of distribution of cancer: Implications fjr models
of carcinogenesis. J. Roy. Stat. Soc. A134: 133-166.
EdUng, C. and S. Granstam. 1979. XXXVI. National Conference cf the
Medical Society sponsored by the Swedish Medical Society at Stockholmsmasan,
AW1f1, Stockholm. December 5-8, 1979. Summaries transaction of the Swedish
rtsdUal Society. 88(3): 73-74. (translation)
Eriksson, Hardell, N.O. Be^g, T. Holler and 0. Axelson. 1979. Case-
control study on malignant mesenchymal tumors of the soft-tissue and expo-
sure to chemical substances. lakart'dMngen. 76: 3872-3875. (translation)
Eriksson, I.. Hardell* N. Q'Berg, T. Moller and 0. Axelson. 1981 - Soft-
tissue sarcomas and exposure to chemical substances: A case-referent study.
Br. J. Ind. Med. 38: 27-33.
FDA (Food and Drug Administration). 1981. FOA advises Great Lakes States
to monitor dloxln-contamlnated fish. FDA Tox. Paper, August 28. in: Food
«nd Drug Cosmetics, p. 41, C21. Convnerce Clearinghouse, Inc. September 8.
1878A -60- 02/27/85
-------
FOA (Food and Drug Administration). 1983. Statement by Sanford A. Miller,
Director, Bureau of Foods, FDA, before Subcommittee on Hater Resources,
Agricultural Research and Environment, U.S. House of Representatives,
June 30.
Fries, G.F. and G.S. Harrow. 197S. Retention and excretion of 2,3,7,8-
:etraeh1crod1benzo-p-d1ox1n by rats. J. Agrlc. Food Chem. 23(2): 265-269.
Gelger, L.E. and R.A. Neal. 1981. Mutagenicity testing of 2,3,7,8-tetra-
cn1orod1benzo-p-d1ox1n in h1st1d1ne auxotrophs of Salmonella typh1muMui>.
Toxicol. Appl. Pharmacol. 59(1): 125-129.
GUvlnl. E., M. Pra11 and C. Vlsmara. 1982a. Effects of 2,3,7,8-tetra-
chlorod1benzo-£-d1ox1n administered to pregnant rats during the prelmplanta-
tlon period. Environ. Res. 29(1): 185-189.
61av1n1, E., M. Pratl and C. Vlsmara. 1982b. Rabbit teratology study with
2,3,7,8-tetrach1orod1benzo-£-d1ox1n. Environ. Res. 27(1): 74-78.
611bert, P., G. Sa1nt-Ruf, F. Poncelet and H. herder. 1980. Genetic
effects of chlorinated anilines and azobenzenes on Salmone!ta tvohlmurlun.
Arch. Environ. Contam. Toxicol. 9(5): 533-5*1.
Green, S. and F.S. Moreland. 1975. Cytogenetic evaluation of several
dloxlns In the rat. Toxicol. Appl. Pharmacol. 33: 161.
1878A
-61-
02/27/85
-------
6reen, $., F. Noreland and C. Sh?u. 1977. Cytogentc effect of 2.3,7,8-
tetrach1orod1ben20-p-12: 349-351.
Hardell, L. 1979. HaMgnart lymphoma of histiocytic type and exposure to
phenoxyacet'c acids or chlcrophenols. Lancet. 1 : 55-56.
Hardell, L. and M. Eriksson. 1981. Soft-tissue sircomas, phenoxy herbl-
cVJes, and chlorinated phenols. Lancrt. 2: 8240.
Harden, I and A. Sandstrom. 1979. Case-control study: Soft-tissue sar-
comas and txposu'e to phenoxya;et1c acids or chlorophenois. 9r. J. Cancer.
39: 711-717.
Hardell, l., M. Eriksson and P. lenner. 1980. faHgnatit lymphoma and
exposure to chem1c£l sjbstances, especially organic solvents, chlorophenois,
and phenoxy adds. LakarUdn'ingen. 77: 208-210. (translation)
Harris, H.H.* J.A. Moore, J.G. Vos and 6.N. Gupta 1973. General bio-
logical effects of TCOD 1n laboratory animals. Environ. Health. Perspect.
5: 101-109.
1878A
04/08/85
-------
Menck, J.W.. M.A. Ntw, R.J. Koclba and K.S. Rao. 1981. 2,3,7,8-Tetra-
chiorod1benio-£-d1ox1n: Acute oral toxicity 1n namsters. Toxicol. Appl.
Pharmacol. 59; 405 407.
Hussaln, S., L. Ehrenbe/g, 6. lofroth and T. GeJ vail. 1972. Mutagenic
effects of TCDD on bacterial systems. Amblo. 1: 3?-33.
1 ARC (International Agency for Research on Cancer). 198?. IARC Monographs
on the Evaluation of the Carcinogenic Risk ot Chemicals to Humans. Vol.
1-29, Supplement 4. IAR'J, Lyon, France.
Khera, K.r,. and J.A. Ri-dd'ck. 19/3. Po1ych1orod1benzo-p-d1ox1ns: Perinatal
effects and the c!om1,iant lethal test 1n WUtar rats, in: Chlorodtoxlns --
OrlgVis and Fat'*, r..H. B1a1r, Ed. Adv. Chem. Ser. No. 120. Am. Chem. Soc.,
Washington, 0C. p. 70-84.
Klmbrough, R.. H. Falk, P. Stehr and G. Fries. 1984. Health Implications
of 2,3,7,0-TCDD contamination of residential soil. J. Toxicol. Environ.
Health. 14(1): 47-93.
Koc
-------
Kodba, R.J., J.E. B*yer, R.N. Carreon and P.J. fiehrlng. 1979. Long-term
toxicologic studies of 2,3,7,8-tetrach1orod1benzo-j>-dtox1n {TCOD) 1n labora-
tory animals. Ann. HY Acad. Sc1. 320: 397-404.
Kondorosl, A., I. Fedorcsak, F. Solymosy, L. Ehren^erg and S. Osterman-
6o1kar. 1973. Inactlvatlon of QBRNA by electrophlles. Mutat. Res. 17:
149-161.
Kauri, R.E., T.H. Rude, R. Joglekar, et al. 1978. 2,3,7,8-Tetrachljrocll-
benzo-£-d1ox1n as cocarclnogen causing 3-methylcholanthrene-1nU1ated subcu-
taneous tumors In mice genetically "nonresponslve" at Ah locus. Cancer Res.
30(9): 277 7-2783.
lepow, N., et al. 1975. Investigation Into sources of Had In the environ-
ment of urban children, fnvlron. Res. 10: 415-4?b.
lopMeno, N., I. Sbrana, 0. Rusdano, D. Lasclalf?.r1 and T. Lar1. ',982. in
vivo cytogenetic studies on mice and rats exposed to 2,3,7,8-tetrachlorodl-
benzo-p-d1ox1n. |n: Chlorinated D1ox1ns and Related Compounds. Impact on
the Environment, 0. Hutzlnger, R.W. Fre1, i. NeMan and P. PocchlaM, Ed.
Pergamon Press, NY. p. 419-426.
Madge, U.S. 1977. Effects of trlchlorophenoxyacetlc aclo and chlorodloxlns
on small Intestinal furctlon. Gen. Pharmacol. 8: 319-324.
Mantel, N. _nd M. A. Schneloerman. 1977. Estimation of 'safe' levels, a
hazardous undertaking? Cancer Res. 35: 1379-1386.
1878*
-64-
04/08/85
-------
HcCann. 1976. Unpublished study. (Cited 1n Wa&som et al., 1978)
MeConnell, I., 6., luder, R. Rumbaugh, et al. 1984. Oloxln 1n $o11:
Bioavailability after Ingestion hy rats and guinea pigs. Science. March 9.
MrNulty, y P. 1978. Direct testimony before the administrator, U.S.
r.nv1ronmental Protection Agency, FIFRA Oocket No. 415, EPA exhibit No. 106.
Motti;ra, A., 6. Ze1, F. Nuzzo, et al. 1981. evaluation of results of
chromosome analyses on lymphocytes of TCDD exposed subjects after the Seveso
accident. Mutat. Res. 65(4}: 238-239.
Murray, F.J., F.A. Smith, K.O. NUschke, C.G. HumUton, R.J. Koclba and B.A.
Schwetz. 1979. Three-generation reproduction study of rats given 2,3,7,8-
tetrach1orod1benzo-2-d1ox1n (TCDD) 1n the diet. Yoxlcol. Appl. Pharmacol.
50: 241-251.
Muscarella, 0., 0. Dennett, J.G. 8*b1sh and D. Noden. 1982. The effects of
2,3,7,8-tetrach1orodlbenzo-£-d1ox1ns on fetal development 1n the ferret.
Toxlcologlst. 2: 73.
Nebert, D., S. Thorglersson and J. Felton. 1976. Genetic differences m
mutagenesis, carcinogenesis, and drug toxicity. In: In vitro Metabolic
Activation 1n Mutagenesis Testing, F. de Serres, J. rolets, J. Bend and R.
Ph1l-ot, Ed. Elsevier/North Holland Biomedical Press, Amsterdam,
p. 105-124.
1878A
-65-
02/27/85
-------
Neubert, 0. and I. DUImann. 1972. Emtoryotoxlc effects 1r nki treated
vUh 2,4,5-Irlchlorophenoxyacet1c add and 2,3,7,B-tetrach1orod1ben*o-£-
dloxln. Arch. Pharmacol. 272(3): 243-264.
Nlsbet, I.C.T. and H.B. Paxton. 1982. Statistical aspects of three-genera-
tion studies of the reproductive toxicity of TOO and 2,4,5-T. An. Stat.
Vol. 36(3); 290-298.
NTP (National Toxicology Program). 1960a. B1o<*ssay of 2,3,7,8-tetrachloro-
d1benzo-&-d1ox1n for possible carcinogenicity (gavage study). OHHS tubl.
No. (NIH) 62-1765. Carcinogenesis Testing Program, NCI, NIH, Bethesda, MD,
and National Toxicology Program, RTP, Box 12233, NC.
NTP (National Toxicology Program). 1980b. Bloassay of 2,3,7,8-tetrachloro-
d1benzo-£-d1ox1n for possible carcinogenicity (Dermal study). CHHS Publ.
No. (NIH) 80-1757. Carcinogenesis Testing Drogram, NCI, NIH, flethesda, MO.
and National Toxicology Program, RTP, Box 12233, NC.
Ott, M.6., B.B. Holder and R.D. Olson. 1980. A mortality analysis of
employees engaged 1n the manufacture of 2,4.5-tr1chlorophenoxyacet1c acid.
J. Occup. Med. 22(1): 47-50.
Pazderova-Vellupkova, J., M. Hemeova, 0. Pickova, L. Jirasek and t. lukas
1981. The development and prognosis of ttironlc Intoxication by tetrachloro-
d1bento-g-d1o\tn 1 r, men. Arch. Environ. Health. 36(1): 5-11.
1878A
-vS6-
03/05/95
-------
PUot, H.C., T. Goldsworthy and H. Poland. 1980. Promotion by 2,3,7,8-
tetrach1orod1benio-[>*dlox1n of hepatocardnogenesls from d1ethy1n1trosam1ne.
Cancer Res. 40: 3616-3620.
Poiger, H. and C. Schlatter. 1980. Influence of solvents and atsorbenM ci.
dermal and Intestinal absorption of TCOD, Food Cosmet. Toxicol. 16(5):
477-481.
Poland, A. and I. Glover. 1979. An estimate of the maximum \n vivo cova-
lent binding of 2,3,7,8-telrachlorod1benzo-p-d1ox1n to rat liver oroteln,
rlbcsoma1. RNA and DNA. Cancer Res. 39(9): 3341-3344.
Poland, A. and i. Glover. 1980. 2,3,7.8-Tt»trach1orod1benzo-£-d1ox1n:
Segregation of toxicity with the Ah locus. Mol. Pharmacol. 17(1): B6-94.
Poland, A., 0. Palen and E. Glover. 1982. Tumour promotion by TCOD 1n skin
of HRS/J mice. Nature. 300(5889): 271-273.
Poland, A.P., D. Smith, G. Better and °. Posslek. 1971. A health survey of
workers 1n a 2,4-0 and 2,4,5-T plant with special attention to chloracne,
porphyria cutanea tarda, and r>>ycho1isj1c parameters. Arch. Environ. Health.
22: 316-327.
Regg1an1, G. 1990. Acute human exposure to TCOD 1n Seveso, Italy. J.
Toxicol. Environ. Health. 6(1): 27-43.
1878A
-67-
03/05/85
-------
Roger*, A.M., M.E. Anderson and K.C. Back. 1982. Mutagenicity of 2,3,7,8-
tetrach1orod1bemo-p-d1ox1n and perfluoro-n-decanolc add 1n LSI78y mouse
lymphoma cells. Mutat. Res. 105: 445-449.
Schanti, S.I., D.A. Barsottl and J.R. Allen. 1979. Tox1colog1cal effects
produced 1n nonhuman primates chronically exposed to fifty parts per tril-
lion 2,3,7,8-te*rachlorod1ben2o-£-o1ox1.i (TCOD). Toxicol. Appl. Pharmacol.
48; MBO.
Schweti, B.A., J.K. Norrls, G.L. Sparschu. et al. 1973. Toxicology of
chlorinated dlbenzo-p-dioxlns. Environ. Health Perspect. 5: B7.
Seller, O.P. l$73, A survey on th? mutagenicity of various pesticides.
Exper1ent1a. 29: 622-823.
Slnper, R., H. Hoses, J. Valcluvas, 3. Lllls and I.J. Sellkoff. 1982.
Nerve conduction velocity studies of workers employed 1n the manufacture of
phenoxy herbicides. Environ. Res. 29: 297-311.
Smith, A.H., J.E. Francis, S.O. Kay and O.B. Grelg. 1981. Hepatic toxicity
and uroporphyrinogen decarboxylase activity following a single dose of
2,3,7,B-tetrach1orod1benio-£-d1ox1n to mice. Blochem. Pharmacol. 30(2C|.
2B25-2B30.
Smith, F.A., B.A. Schwetz and K.D. Nltschke. 1976. Teratogenicity of
2,3,7,8-tetrach1orod1benzo-g-d1ox1n In CF-1 mice. Toxicol. Appl. Pharmacol.
38(3): 517-523.
1878A
-68-
03/t=:/85
-------
Snyder, U.S., H.J. Cook, E.S. Nasset, L.R. Karhausen, 6.P. Howe11$ and T.H.
Upton. 1975. Report of the Task 6roup on Reference Han. International
Contn1ss1on cf Radiological Protection No. 23. Pergaranon Press, NY.
Sparschu, G.L., F.l. Dunn, R.N. Llsowe and V.K. Rowe. 1971. Effects of
high levels of 2,4,5-trlchlorophenoxyacetU add on fetal development In the
rat. Food Cosmet. Toxicol. 9(4): 527-530.
Task Group on Lung Dynamics. 1966. Deposit'on and retention models for
Internal dosimetry of the human respiratory ra
-------
Turner, J.N. and O.N. Collins. 1983. Liver morphology In guinea pigs
administered either pyrolysls products of a polychlorlnated biphenyl trans-
former fluid or 2,3,7,8-tetrach1orod1ben2o-£-d1ox1n!». Toxicol. Appl.
Pharmacol. 67: 417.429.
U.S. EPA. I960. Guidelines and Methodology Used 1n the Preparation of
Health Effects Assessment Chapters of the Consent Decree Water Quality
Criteria. Federal Register. 45: 79347-79357.
U.S. EPA. 1984a. Ambient Water Quality Criteria document for 2,3,7,8-
Tetrach1orod1benzo-£-d1ox1n. Office of Water Regulations and Standards.
EPA 440/5-84-007.
U.S. EPA. 1984b. Risk Ana1y>1s of TCOD-Contamlnated Soil. J.I. Schaum.
NTIS-PBB5-146704/AS.
U.S. EPA. 1984c. Proposed Guidelines for Mutagenicity Risk Assessment:
Request for Comments. Federal Register. 49; 46314-46321.
Van Miller, J.P., J.J. Lallch and J.R. Allen. 1977a. Increased Incidence
of neoplasms In rats exposed to low levels of 2,3,7,8-tetrach1orod1bemo-£-
dloxln. Chemosphere. 6(10): 625-632.
Van Miller, J.P., J.J. Lallch and J.R. Allen. 1977b. Increased Incidence
of neoplasms 1n rats exposed to low levels of ?,3,7,8-tetrachiorod1benzo-j>-
dloxln. Chemosphere. 6(9): 537-544.
1878A
-70-
05/31/85
-------
Vos, J.fi., J.A. Moore and J.6. Z1nkl. 1973. Effect of 2.3.7,8-tetrachloro-
d1benzo-£-d1ox1r. on the Iwiune system of laboratory animals. Environ.
Health Perspect. 5: 149-H2.
Young, A.L. 1983. Long-term studies on the persistence and movement of
TCOD In a natural ecosystem. £n: Human and Environmental Risks of Chlori-
nated D1ox1ns and Related Compounds, R.E. Tucker et al., Ed. Plenum Pub-
lishing Corp., NY. p. 173-190.
1878A
-71-
02/27/85
------- |