OPTS-TECHNICAL I^ORKITION CENTER
United States EPA-600/6-81-003
Environmental Protection ' , *
Agency September 12, 1980
EPA Research and
Development
RISK ASSESSMENT ON
(2,4,5-TRICHLGROPHENOXY) ACETIC ACID (2,4,5-T)
(2,4,5-TRICHLOROPHENOXY) PROPIONIC ACID-.
2,3,7,8-TETRACHL0R0DIBENZ0-P-DI0XIN (TCDD)
Prepared for
Office of the General Counsel
U.S. Environmental Protection Agency
Prepared by
Office of Health and
Environmental Assessment
Washington DC 20460
Carcinogen Assessment Group
-------
0P1S-2ECKBICAL lliXOHMAl'ION CENTER
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completingj
1. REPORT NO. 2.
EPA-500/6-81- 003
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Risk Assessment on (2,4,5-Trichlorophenoxy) Acetic Acid
(2,4,5-T)-, (2,4,5-TrichloroDhenoxv) Proninnir Arid
(Silvex), 2,3,7,8-Tetrachlorodibenzo-P-Dioxin (TCDD^
5. REPORT DATE
September 12, 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Carcinogen Assessment Group
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Carcinogen Assessment Group
Office of Health and Environmental Assessment
Environmental Protection Agency
Washington, D.C. 20460
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
In-house
12. SPONSORING AGENCY NAME AND ADDRESS
Office of the General Counsel
Environmental Protection Agency
Washington, D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Response/Assessment
14. SPONSORING AGENCY COOE
15. SUPPLEMENTARY NOTES
IS. ABSTRACT
Carcinogenic responses have been induced in mice and rats at low doses of
TCDD. TCDD has been shown to be a cancer promoter. These results, together
with the strongly suggestive evidence in epidemiology studies, constitute
substantial evidence that TCDD is likely to be a human carcinogen. It
appears that TCDD is a more potent carcinogen than aflatoxin Bi which is one
of the most potent carcinogens known. The levels of TCDD (contained as an
unavoidable contaminant of the 2,4,5-T) used in the 2,4,5-T studies
apparently were too small to produce an observable response in those
experiments. The lack of a statistically significant tumor incidence in most
of the studies on the 2,4,5-T product may be attributed to the very low
levels of TCDD in the product relative to the levels at which it produced
carcinogenic effects in rats and mice, as well as to deficiencies of those
studies. However, since TCDD is a carcinogen, any product containing TCDD,
including 2,4,5-T and silvex, can be considered to pose a human carcinogenic
hazard. Furthermore, a rat study on specially purified 2,4,5-T provides
highly suggestive evidence that essentially pure 2,4,5-T may be a human
carcinogen. Quantitative assessments have been calculated for the
carcinogenic risk posed to humans.
17. KEY WORDS ANO DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENOED TERMS
c. COSATl Field/Group
IS. DISTRIBUTION STATEMENT
NTIS - Release to Public
19. SECURITY CLASS (Tliis Report)
UNCLASSIFIED
21. NO. OF PAGES
276
20. SECURITY CLASS (This page J
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
-------
Page
104
106
106
106
109
109
110
110
110
111
115
116
116
120
130
131
132
133
134
135
137
CORRECTIONS TO CARCINOGEN ASSESSMENT GROUP'S RISK ASSESSMENT
ON 2,4,5-T, SILVEX, AND TCDD
(Dated September 12, 1980)
Line
Present
Should Be
1
18-19
18
21
18
18
9
19
last
4
Table 49
Table 59
7
11
10
6
9
9
ae
that apply 2,4,5-T
the applicators
Pg. 13
exposures
4.7 x 10-4
high consumer group
as high as or
4.7 x lO"4
were
Females
Revised Table attached
Revised Table attached
8.4 x 105
210-4
4.8 x 10-3
Local population*
pg/kg/bw/day
4.7 x 10-4
are
(omi t)
the 2,4,5-T applicators
Pg. 14
exposure
4.5 x 10-4
local population
(omi t)
4.5 x lO"4
is
delete footnote b
Females3
Delete footnote and
replace with:
aSubcutaneous
combined fibroma or
fibrosarcoma not
significant
8.4 x lO"5
< 10-4
5.2 x 10-3
(omit *)
pg/kg bw/day
4.5 x 10-4
-------
TABLE 49. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL DATA QY STUDY AND PATHOLOGIST
LINEAR PARAMETER qj, MAXIMIZED TO GIVE UPPER 95% LIMIT q*
Compound TCDD
Study Koclba - Dow
Sex-spec1 es Male rat
Weight (wa) 600 gm
Tumor sites (one or more)....Tongue - squamous cell carcinomas
Nasal turbinates/hard palate - stratified squamous cell carcinoma
Pathologist - Koclba
Exposure level (mg/kg/day) 0
1 x lO"6
1 x 10"5
1 x 10"4
+r/n 0/76
2/49
1/49
3/42
+r = number of animals with one or more
n = total number of animals examined
of the tumors
Estimated
multistage parameters qo
<11 H2
13
Goodness of fit
When all dose groups
are used 1.40 x 10-2 1.
10 x 103 0
5.86 x 1010 3.01 x
103 3.34 (d.f.=2)
When the highest dose
group is not used
Above fit is satisfactory
When the two highest dose
groups are not used
q|~ThF maximum linear component from the model with adequate godclrTess oTTit (P > 0.01) = J.01 x 10^
^h = (70/wa)l/3 = 1.47 x 104, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 59. HUMAN SLOPE ESTIMATES
Compound
Species
Study
Sex
Pathologist Human Slope Estimate q£
TCDD
Rat
Dow
Male
Kociba
1.47
X
104
Squi re
1.73
X
104
Female
Kociba
2.52
X
105
Squire
4.25
X
105
NCI
Male
Female
NCI - Reviewed
NCI - Reviewed
2.43 x 104
3.28 x 104
h-*
CJ
o
Mice NCI
Male
Female
NCI - Reviewed
NCI - Reviewed
1.33 x 105
4.56 x 104
2,4,5-T Rat Dow
Male
Kociba
Squire
1.65 x 10_:
1.02 x 10"
*Values used in risk analysis
-------
THE CARCINOGEN ASSESSMENT GROUP'S
RISK ASSESSMENT ON
(2,4,5-TRICHLOROPHENOXY)ACETIC ACID (2,4,5-T)
(2,4,5-TRICHL0R0PHEN0XY)PR0PI0NlC ACID (SILVEX)
2,3,7,8-TETRACHL0R0DIBENZQ-P-DIQXIN (TCDD)
PARTICIPANTS
Elizabeth L. Anderson, Ph.D.
Larry D. Anderson, Ph.D.
Steven Bayard, Ph.D.
David Bayliss, M.S.
John R. Fowle III, Ph.D.
Bernard H. Haberaan, D.V.M., M.S.
Charalingayya B. Hiremath, Ph.D.
Chang S. Lao, Ph.D.
Robert McGaughy, Ph.D.
Charles Poole, M.P.H.
Dharm V. Singh, D.Y.M., Ph.D.
Todd W. Thorslund, Sc.D.
Peter Voytek, Ph.D.
rman
_Dwri rman
Septesrber 12, 1980
-------
NOTICE
The Carcinogen Assessment Group (CAG), located in the Office of Health
and Environmental Assessment of EPA1s Office of Research and Development, is
a small group of scientists who perform an advisory assessment function for
EPA's regulatory offices. The CAG analyzes existing scientific data and
furnishes the regulatory offices with an evaluation of the carcinogenicity
and levels of carcinogenic risk associated with chemicals in various exposure
situations, as best can be determined from currently available scientific
data.
The CAG reports are prepared for internal Agency use in response to EPA's
regulatory office needs. They range from brief chemical profiles to very
extensive evaluations, depending upon the nature of a request. The reports
are used by the regulatory offices for regulatory decision making as
appropriate. The reports are revised and edited based on regulatory office
needs and the availability of resources.
This document was prepared at the request of the EPA Office of the
General Counsel.
-------
CONTENTS
SUMMARY AND CONCLUSIONS. 1
Qualitative Risk Assessment 1
Quantitative Risk Assessment of 2,4,5-T, silvex, TCDD 6
QUALITATIVE RISK ASSESSMENT
I. Introduction 8
II. Metabolism TO
Metabolism of (2,4,5-Trichlorophenoxy)Acetic Acid
(2,4,5-T) 10
Metabolism and Storage of
2,3,7,8-Tetrachlorodibenzo-P-Dioxin (TCDD) 11
Aryl Hydrocarbon Hydroxylase (AHH)
Induction Studies With TCDD 13
Covalent Binding of TCDD with Macromolecules 15
III. Mutagenicity 17
Mutagenicity of 2,4,5-T 17
Mutagenicity of TCDD 21
Conclusion 23
IV. Toxicity 24
Animal Toxicity 24
Toxicity of 2,4,5-T 24
Toxicity of TCDD ' 25
Toxicity of 2,4,5-T, 2,4,5-Trichlorophenol,
and TCDD in Humans 27
V. Carcinogenicity 29
Carcinogenicity of 2,4,5-T in Mice 29
Muranyi-Kovacs et al. (Oral) Mouse Study 29
Muranyi-Kovacs et al. (Subcutaneous) Mouse Study 31
Innes et al. (Bionetics Laboratories) (Oral) Mouse Study .... 33
Innes et al. (Bionetics Laboratories)
(Subcutaneous) Mouse Study 35
Carcinogenicity of 2,4,5-T in Rats 36
Kociba et al. (Oral) Rat Study 36
Leuschner et al. (Oral) Rat Study ^ 44
i
-------
Carcinogenicity of (2,4,5-Trichlorophenoxy)Propionic Acid (Silvex) . 43
Innes et al . (Bionetics Laboratories) (Oral) Mouse Study .... 48
Innes et al. (Bionetics Laboratories)
(Subcutaneous) Mouse Study 50
Dow Chemical Company (Oral) Rat Study 51
Dow Chemical Company (Oral) Dog Study 52
Carcinogenicity of TCDD in Rats and Mice 53
Kociba et al. (Oral) Rat Study 53
National Cancer Institute (Oral) Rat Study 60
Van Miller et al. (Oral) Rat Study '. 63
Toth et al. (Oral) Mouse Study 67
National Cancer Institute (Oral) Mouse Study 70
Other Related Studies 73
Pitot et al. Promotion Study in Rats 73
National Cancer Institute Skin Painting Study
in Mice 75
Berry et al. Skin Painting Study in Mice 77
Cohen et al. Skin Painting Study in Mice 78
Kouri et al. Mouse Study 78
Estimation of TCDD Levels in 2,4,5-T Studies 84
Potency of TCDD 87
Summary of Laboratory Animal Studies on 2,4,5-T, Silvex, and TCDD. . 88
VI. Epidemiologic Studies 90
QUANTITATIVE RISK ASSESSMENT
I. Introduction 102
II. Estimation of the Dose-Response Model 104
III. Risks for Applicators 106
Forestry 107
Range and brush control 107
Rice-weed control 107
Rights-of-way brush and weed control 108
IV. Risks Due to Dietary Exposure 108
Beef and Milk 108
Deer and Elk 109
Rice 110
V. Summary 110
i i
-------
REFERENCES 137
APPENDICES 145
A. Dose-related Mortality Estimates in Kociba's TCDD Rat Study (Tables).
B. Pathologic Evaluations of Selected Tissues from the Dow Chemical
TCDD and 2,4,5-T Rat Studies by Robert A. Squire, Associates, Inc
(Summary Tables)
C. leuschner Histopathologic Testicular Tumors in Rat (Historical
Control Data)
D. Leuschner Histopathologic Report on Tongue in 2,4,5-T Rat Study . . .
E. Memo from Wade Richardson Concerning the Telephone Conversation
with Leuschner
F. Memorandom and report from Dr. David Severn, Hazard Evaluation
Division, Office of Pesticide Program Exposure. Assessment of
2,4,5-T, Silvex and TCDD
G. Methods for Determining the Unit Risk Estimates for Air Pollutants. .
i i i
-------
SUMMARY AND CONCLUSIONS
QUALITATIVE RISK ASSESSMENT
(2,4,5-Trichlorophenoxy)Acetic Acid (2,4,5-T)
(2,4,5-Trichlorophenoxy)acetic acid, widely known as 2,4,5-T is used as a
vegetation growth regulator and herbicide. "Agent Orange," a defoliant used
extensively by the U.S. Army in Vietnam, is a mixture of equal amounts of
2,4,5-T and (2,4-dichlorophenoxy)acetic acid. In 1970, amid growing concern
about the teratogenic effects of 2,4,5-T, the EPA cancelled the registration of
the compound for uses "around the home, recreation areas, and similar sites" and
"in crops intended for human consumption." Before some uses were suspended in
1979, it was used primarily to clear vegetation along power!ines, highways,
pipelines, and railroad rights-of-way, and on range, pasture, and forestlands.
The commercial preparation of 2,4,5-T contains 2,3,7,8-tetrachlorodibenzo
-p-dioxin (TCDO) as an unavoidable impurity present at a concentration of
approximately 0.05 ppm. TCDO is considered extremely toxic.
2,4,5-T is readily absorbed by several mammalian species, including man, and
is excreted unchanged - mostly in urine.
The available information about the mutagenic activity of 2,4,5-T is
considered to be limited. 2,4,5-T is indicated to be a weak mutagen in
Drosophila and, under acidic conditions, showed mutagenic effects in
Saccharomyces cerevisiae.
Tests for the chronic carcinogenicity of 2,4,5-T were performed by several
investigators. Two studies were carried out with Sprague-Dawley rats, one by
the Dow Chemical Company (Kociba et al. 1979) and one by F. Leuschner (1979),
Laboratorium fur Pharmakologie und Toxikologie, Hamburg, Germany. The Dow study
showed an increased incidence of carcinoma of the tongue in male rats dosed with
1
-------
specially purified 2,4,5-T at 30 mg/kg/day. This incidence as reported by the
authors of the study, is marginally statistically significant (F = 0.063) when
compared to controls. In addition, there was a significant dose-related linear
trend by the Cochran-Armitage test. When compared to historical controls, the
incidence of this tumor, as reported by the authors, is statistically
significant (P < 0.001); however, when the tongue tissues from this study were
reexamined by Dr. Squire, he found one additional tongue carcinoma in male rats
treated at the high dose with 2,4,5-T which increased the statistical
significance to P = 0.025. In a recently completed study by F. Leuschner, an
increased incidence of interstitial cell tumors of the testes was observed when
compared with matched controls. However, this increase is not statistically
significant when compared to historical controls. The results of the Kociba et
al. study provide highly suggestive evidence of the carcinogenicity of
essentially pure 2,4,5-T.
In mice, two studies by Muranyi-Kovacs et al. (1976, 1977) and two studies
by Innes et al. (1969) (Bionetics Laboratories 1968) have not provided positive
evidence of oncogenic effects of 2,4,5-T. However, several deficiencies in
these studies make them inadequate to assess the lack of oncogenicity of
2,4,5-T.
In summary, the Do* study in rats provides highly suggestive evidence of the
carcinogenicity of 2,4,5-T, while the Leuschner study showed only equivocal
results. The mouse studies were too insensitive to be considered valid negative
studi es.
(2,4)5-Trich1orophenoxy)Propionic Acid (Silvex)
Silvex, like 2,4,5-T, contains the highly toxic TC0D. Uses of silvex are
similar to those of 2,4,5-T. Chronic carcinogenicity studies have been
performed on mice and rats and a 2-year study has been conducted on dogs. Innes
2
-------
et al. (1969) (3ionetics Laboratories 1968) conducted two studies using mice,
one oral and the other subcutaneous. These studies were found to be inadequate
to assess the carcinogenicity of silvex.
Dow Chemical Company performed two feeding studies, a 2-year feeding study
on rats and a two year feeding study on dogs which were summarized by Mullison
(1966) and Gehring and Betso (1978). These have been found to be inadequate to
rule out the carcinogenicity of silvex.
2,3,7,8-Tetrachlorodibenzo-P-Dioxin (TCDD)
Probably one of the most toxic chemicals known to man is
2,3,7,8-tetrachlorodibenzo-p-dioxin. The major source of its environmental
contamination is from the pesticidal uses of 2,4,5-T, 2,4,5-trichlorophenol, and
silvex.
In smaTl amounts, TCDD is a potent inducer of arylhydrocarbon hydroxylase in
mammals. This is a complex enzyme system that consists of epoxidase,
epoxidehydratase, and glutathione transferase. The enzyme epoxidase is known to
mediate the formation of epoxides, which are potentially active carcinogenic
metabolites. TCDD can be metabolized in mammalian species via the epoxide to
dihydodiol and further conjugates with glutathione. Persistent residues of
TCDD were found in liver and fat in a 2-year feeding study in rats. Significant
covalent binding of TCDD to protein has been demonstrated by two investigators.
Covalent binding of TCDD with DNA is less significant in liver cells.
Currently available studies on the mutagenicity of TCDD are inconclusive.
Two bacterial systems, Escherichia coli and Salmonella typhimurium (without
metabolic activation), exhibited positive mutagenic activity. However, in
another study of Salmonella typhimurium (with and without metabolic activation),
the results were negative.
3
-------
There are several cancer bioassay stu'ies of TCDD: 1) a Dow Chemic?1
Company (Kociba et al. 1973) study in mal and female Sprague-Dawley (S> rtan
substrain) rats; 2) the Van Miller et al. (1977) study in male Sprague-Dawley
rats; 3) the Toth et al. (1979) study in Swiss mice; 4) the National Cancer
Institute (1980a, b) studies in rats and mice; 5) the Pitot et al. (1980)
promotion study in rats; and 6) the Kouri et al. (1978) cocarcinogenicity study
in mice.
The study by the Dow Chemical Company of male and female Sprague-Dawley rats
fed TCDD in doses of 22 ppt, 210 ppt, and 2200 ppt revealed a highly
statistically significant excess incidence of hepatocellular carcinomas in
female rats at the highest dose level and hepatocellular carcinomas and
hepatocellular hyperplastic nodules in female rats at the middle dose level, as
compared to the controls. In addition, there was a significant increase in
carcinomas of the hard palate/nasal turbinates in both high dose males and
females, of the tongue in males, and of the lung in females. The Van Miller et
al. study also showed some evidence of a carcinogenic response in the liver and
lungs of male Sprague-Dawley rats at dosages of 1000 and 5000 ppt, even though
the study used a relatively small number of animals. The Toth et al. study
provides suggestive evidence that TCDD induced an increased incidence of liver
tumors in male mice (females were not tested) receiving 0.7 ug/kg/week by
gavage.
In the National Cancer Institute rat study (1980a), male and female
Osborne-Mendel rats were administered TCDD by gavage at three dose levels 0.01,
0.05, and 0.5 ug/kg/week. TCDD induced statistically significant increases of
hepatocellular carcinomas, subcutaneous fibrosarcomas, and adrenal cortical
adenomas in high dose female rats. TCDD also induced significant increases in
thyroid tumors at low, middle, and high doses in male rats.
4
-------
In a companion mouse study by the National Cancer Institute (1980a), male
and female B6C3F1 mice were given TCDD by gavage at dose levels of 0.01, 0.05,
and 0.5 ug/kg/week for males and 0.04, 0.2, and 2.0 ug/kg/week for females.
TCDD induced statistically significant increased incidences of hepatocellular
carcinomas in the high dose males and females, and thyroid tumors, subcutaneous
fibrosarcomas, and histiocytic lymphomas in females.
In a study by Pi tot et al. (1980), TCDD has been shown to be a potent liver
cancer promoter. In a study by Kouri et al. (1978), TCDD has been shown to be a
cocarci nogen.
Epidemiologic Studies
Several epidemiologic studies have been conducted which are relevant to the
assessment of the carcinogenicity of 2,4,5-T, silvex, and TCDD. Two Swedish
epidemiological case-control studies (Hardell and Sandstrom 1979, Erikson et al.
1979) reported a very strong association between soft tissue sarcomas and
occupational exposure to phenoxyacetic acid herbicides and/or chlorophenols.
These studies indicated approximately five to sevenfold increases in the risk of
developing soft tissue sarcomas among people exposed to phenoxyacetic acids only
in comparison to people not exposed to these chemicals. Another Swedish
case-control study (Hardell et al. 1980) provides suggestive evidence of an
increased risk of developing lymphomas resulting from occupational exposure to
phenoxyacetic acids.
Two cohort studies, one by Axelson et al. (1980) and the other by Thiess and
Frentzel-Beyme (1977) provide suggestive evidence that phenoxyacetic acids
and/or TCDD increases the risk of stomach cancer in humans.
Four other cohort studies by Ott et al. (1980), Riihimaki et al. (1978),
Zack and Suskind (1980), and Cook et al. (1980) did not indicate cin increased
5
-------
risk of stomach cancer, but three of these studies were of relatively low
statistical power, and the fourth (Riihimaki et al. 1977) has certain
inconsistencies requiring clarification.
In summary, carcinogenic responses have been induced in mice and rats at
very low doses of TCDD. In addition, TCDD has been shown to be a potent cancer
promoter. These results, together with the strongly suggestive evidence in
epidemiologic studies, constitute substantial evidence that TCDD is likely to be
a human carcinogen. In addition, on the basis of the Dow study on TCDD, it
appears that TCDD is a more potent carcinogen than aflatoxin B]_ which is one
of the most potent carcinogens known. The levels of TCDD (contained as a
contaminant of the 2,4,5-T) used in the 2,4,5-T studies apparently were too
small to produce an observable response in those experiments. The lack of a
statistically significant tumor incidence in most of the studies on the 2,4,5-T
product may be attributed both to the very low levels of TCDD in the product
relative to the levels at which it produces observable carcinogenic effects in
rats and mice, as well as to the deficiencies of those studies. However, since
TCDD is a carcinogen, any product containing TCDD, including 2,4,5-T and silvex,
can be considered to pose a human carcinogenic hazard. In addition, a rat study
on specially purified 2,4,5-T provides highly suggestive evidence that
essentially pure 2,4,5-T may be a human carcinogen.
QUANTITATIVE RISK ASSESSMENT OF 2,4,5-T, SILVEX AND TCDD
A quantitative assessment has been calculated for the carcinogenic risk
posed to humans by the use of the herbicides 2,4,5-T and silvex. While there is
no evidence for carcinogenicity of silvex, the evidence for 2,4,5-T is highly
suggestive, and that for the contaminant TCDD is substantial. Furthermore,
TCDD is highly carcinogenic to animals.
6
-------
The assessment of risk from TCDD exposure covers only the herbicide
applicators and dietary exposure to beef, milk, deer, and elk. For unprotected
workers, the upper limits of lifetime risk of induced cancers are in many cases
as high as or in the 10" ^ range. For the general population exposed to beef
contaminated with TCDD, the upper limit of risk for the estimated exposure is
2.4 x 10"®. For local populations consuming only beef which is contaminated
with TCDD, the risk is much greater, as high as 1.9 x 10-4 for the estimated
exposure. For local populations consuming only milk and other dairy products
which are contaminated with TCDD, the risk is 4.7 x 10"^. For deer and elk
meat contaminated with TCDD, risks to the local population are no greater than
lO"4 for 12 meals a year.
The upper limit of dietary risk associated with estimated exposures to
2,4,5-T in contaminated rice and milk were in the 10"7 range for a high
consumer eating only contaminated rice or an average consumer drinking only
contaminated milk.
7
-------
QUALITATIVE RISK ASSESSMENT
I. INTRODUCTION
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is an extremely toxic contaminant
that forms when tetrachlorobenzene is hydrolyzed in an alkaline ethylene glycol
solution to produce 2,4,5-trichlorophenol. The amount of TCDD produced
increases with an increase in the temperature of the reaction. The
2,4,5-trichlorophenol is used as an intermediary in the production of
(2,4,5-trichlorophenoxy)acetic acid (2,4,5-T) and
(2,4,5-trichlorophenoxy)propionic acid (si!vex). Therefore, the TCDD
contaminates both products to the same extent (see Figure 1, below).
TCDD can also occur in other chlorinated phenols and in the chemicals
synthesized from them. TCDD does not occur naturally in the environment, but
exists only as a contaminant of other chemicals.
Figure 1. Formation of TCDD, 2,4,5-trichlorophenol, 2,4,5-T, and silvex.
In the 1960s, the TCDD content in commercial 2,4,5-T and
2,4,5-trichlorophenol ranged from 5 to 50 ppm. By the early '70s, the
manufacturers had set a limit of 0.1 ppm TCDD contamination in their products.
tetrachlorobenzene >2,4,5-trichlorophenol
(TCDD)
(TCDD)
8
-------
The structure of the four compounds is shown in Figure 2 below.
OH
I
ci.
CI
0 - CH2 - COOH
-CI
C1-\^
c'i
2,4,5-trichlorophenol
(2,4,5-TCP)
(2,4,5-trichlorophenoxy)acetic acid
(2,4,5-T)
0 - CH2 - CH2 - COOH
-CI
C1J
I
CI
2,3,7,8-tetrachlorodibenzo-p-dioxi n
(TCDD)
(2,4,5-trichlorophenoxy)propionic acid
(si 1 vex)
Figure 2. Structure of TCDD and TCDD-containing compounds.
2,4,5-T is used as a growth regulator and herbicide. The herbicide "Agent
Orange," used extensively by the U.S. Army as a defoliant in Vietnam, is a
mixture of equal amounts of 2,4,5-T and (2,4-dichlorophenoxy)acetic acid. In
1970, amid growing concern about the teratogenic effects of 2,4,5-T, the EPA
cancelled registration of the compound for uses "around the home, recreation
areas, and similar sites" and "on crops intended for human consumption." Until
EPA suspended certain uses in 1979, it was used primarily to clear vegetation
along powerlines, highways, pipelines, and railroad rights-of-way, and on range,
pasture, and forestlands.
9
-------
II. METABOLISM
METABOLISM OF (2,4,5-TRICHL0R0PHEN0XY)ACETIC ACID (2,4,5-T)
The metabolic fate of 98% pure 2,4,5-T was studied in beagle dogs and adult
Sprague-Dawley rats following a single oral dose of the chemical (Piper et al.
1973). The absorption of 2,4,5-T appeared to follow first order kinetics in
rats and dogs. The rate at which the compounds cleared from plasma was also of
the first order in rats, but in dogs, the clearance rate was much more complex
than first order.
The tl/2 values for clearance of ^C-activity from the plasma of rats
given doses of 5, 50, 100, or 200 mg/kg were 4.7, 4.2, 19.4, and 25.2 hours,
respectively. The volume of distribution also apparently increased with dose.
In dogs given 5 mg/kg, the tl/2 values for clearance from plasma and elimination
from the body were 77.0 and 86.6 hours, respectively.
Essentially all of the 2,4,5-T was excreted unchanged in the rats' urine,
except for a small amount of one unidentified metabolite that was detected only
in rats administered the two highest doses. Urinary excretion accounted for
most of the 2,4,5-T eliminated from the body in rat ; little was found in the
feces.
In dogs, a greater percentage of 2,4,5-T was excreted in the feces than in
the urine. Three unidentified metabolites of 2,4,5-T were detected in the
urine, but there may have been fecal contamination, so the source is equivocal.
The slower elimination of 2,4,5-T in the body of the dog may account for its
greater metabolic alteration. The authors suggest that the kidney possesses a
saturable active transport system for 2,4,5-T and this transport system has a
greater capacity in adult rats than in dogs. The longer half-life of
elimination and the metabolic degradation of 2,4,5-T in dogs may'explain why
10
-------
2,4,5-T is more tc ic to dogs than to rats.
Five male human volunteers ingested a single 5 mg/kg dose of 99% pure
2,4,5-T containing 0.05 ppm TCDD (Gehring et al. 1973). The plasma concentration
of 2,4,5-T increased rapidly and peaked at 57 ug/ml following 7 hours of
administration. The subsequent clearance rates from the plasma and body were of
first order, situated numerically between the rates for dogs and for rats. The
2,4,5-T was actively secreted in the urine. It was concluded that 2,4,5-T is
eliminated fairly unchanged from the human body. The volume distribution in
humans was smaller than for test animals. In humans, 65% of the compound
remaining after 24 hours was present in plasma, and 99% of this was reversibly
bound to protein.
In conclusion, 2,4,5-T is readily absorbed by several mammalian species
including man, and excreted mostly in the urine.
METABOLISM AND STORAGE OF 2,3,7,8-TETRACHLGR0DIBENZ0-P-DI0XIN (TCDO)
In a 1976 study by Rose et al., Sprague-Oawley rats were given either a
single oral dose of 1.0 ug ^C-TCDO/kg (98% pure with 2%
trichlorodibenzo-p-dioxin) or repeated oral doses of 0.01, 0.1, or 1.0 ug
l4C-TCDD/kg/day, 5 days per week, for 7 weeks.
The authors monitored the fate of l^C-TCDD jn rats after single oral
administration and found that, on the average, 83% of the dose was absorbed.
Twenty-two days after the single oral dose, concentrations of ^C-activity
were retained mainly in the liver (1.26% of dose) and fat (1.25% of dose). The
half-life of following a single oral dose was 31 + 6 days, which followed
first order kinetics. Most of the 14C-activity was detected in feces and not
in urine or expired air, which indicates that TCDDand/or its metabolites are
eliminated via the bile.
11
-------
The authors also monitored the fate of the l^C-TCDD ingested repeatedly.
Following the administration of all doses of l^C-TCDO, the average dose
absorbed was 8655. The rats were killed at 1, 3, or 7 weeks, and their liver,
fat, kidney, thymus, and spleen tissues were examined for ^C-activity.
Activity was primarily localized in the liver and fat, with radioactivity in the
liver being five times greater than in fat. The accumulation of the isotope in
both types of tissue followed first order kinetics. With continuous
administration, the concentrations of l4C-radioactivity in both tissues
approached plateau levels by 7 weeks (73.8% of steady state values). By 13
weeks, 93% of the steady state level had been reached; the rate of accumulation
of radioactivity was independent of the dose level administered over the dose
range of 0.01 to 1.0 ug TCDD/kg/day.
The half-life of elimination of ^C-activity in rats was 23.7 days. The
l^c-TCDD radioactivity was excreted primarily in the feces, with some of it in
an altered chemical form, presumably having been excreted from the liver via
bile. Significant amounts of were ais0 found in urine, particularly
in female rats. Female rats were given 0.001, 0.01, and 0.1 ug TCDD/kg of body
weight for 2 years (Kociba et al. 1978). After being killed, the liver and fat
tissues were analyzed for TCDD content. The chemical analysis of liver and fat
are shown in the Table 1 below. These results reveal the dose-dependent
accumulation of TCDD after long-term exposure.
12
-------
TABLE 1. CONCENTRATIONS OF TCDO IN RAT LIVER AND FAT
AFTER 2 YEARS OF FEEDING
Dose
Concentration
Concentrations
in liver3
in fata
0.001 ug/kg
540
540
0.01 ug/kg
5,100
1,700
0.1 ug/kg
dna»
-------
(3-MC) in 14 mouse strains with a high carcinogenic index, a measure of
sensitivity to chemical carcinogens which are metabolically activated to a
carcinogenic intermediate, and found that the AHH system plays a role in this
activation.
TCDO is reported to produce effects on microsomal mixed-function oxidases
that are very similar to those produced by 3-MC (Poland and Glover 1974). Many
compounds have been shown to induce certain mixed-function oxidases, but
enzyme-inductive plus enzyme-suppressive effects are peculiar to inducers like
3-MC and TCDD. The mechanisms involved in regulation of the mixed-function
oxidases vary with the organ and species (Hook 1975). The potency of TCDD in
inducing hepatic AHH is 3 x 10^ that of 3-MC (Poland and Glover 1974), and is
40 to 60 times that of 3-MC in inducing AHH activity in cultured human
lymphocytes (Kouri and Ratrie 1974).
There are genetic differences in AHH inducibility in humans and in mice
(Poland and Glover 1976b, Kouri and Ratrie 1974). Poland and Glover (1976b)
inferred from mouse data that the hepatic cytosol species that binds TCDD is the
receptor for the induction of hepatic AHH activity, and the mutation in
non-responsive mice results in an altered receptor with a diminished affinity
for inducing compounds.
In conclusion, TCDD is a potent inducer of arylhydrocarbon hydroxylase in
mammalian species. This is a complex enzyme system which consists of epoxidase,
epoxidehydratase, and glutathione transferase. The enzyme epoxidase is known to
mediate the formation of epoxides, which are potentially reactive carcinogenic
metabolites. TCDD undergoes metabolic transformation in mammalian species;
however, its persistent residues were found in liver and fat after 2-year
feeding studies in rats.
14
-------
COVALENT BINDING OF TCDD WITH MACROMOLECULES
There are two relevant studies that deal with the interaction of
2,3,7,8-tetrachlorodibenzo-p-dioxin with macromolecules. In the first study by
Guenthner et al. (1979), covalent binding of TCDD metabolites to cellular
macromolecules was measured in vitro after incubation of tritiated TCDD with
methylcholanthrene-induced B6C3F1 mouse microsomes, NADPH, and deproteinized
salmon DMA. The ratio of amount of DNA to the amount of protein in the reaction
vessel was 4:1. After incubation, the DNA was reisolated and treated with
DNase, phosphodiesterase, and alkaline phosphatase. TCDD metabolite-nucleoside
adducts were isolated by sephadex LH20 column chromatography. The
radioactivity equivalent to TCDD that binds with DNA was 0.074 p mole/mg. When
DNA was incubated with proteinase before being applied to the sephadex column,
more than 80% of the covalently bound TCDD metabolites were removed, leaving
only 0.016 p mole/mg of TCDD-equivalent radioactivity bound to DNA.
The amount of covalently bound TCDD equivalent to microsomal protein was
20.6 p moles/mg, indicating this binding occurred approximately 1,000 to 2,000
times more readily than the binding to DNA.
In the second study, Poland and Glover (1979) examined the in vivo covalent
binding of TCDD (or metabolites) to rat liver macromolecules. In this study,
tritium labeled 3[h]TCDD, 95% chemically pure, was used (the impurity
consisted of radiolabeled trichloro- and pentachlorodibenzo-p-dioxin). A dose
of 7.5 mg/kg [1,6 3h]TCDD with specific activity of 39 Ci/mmole was
administered intraperitoneal^ to Sprague-Dawley rats (approximately 90
uCi/rat). The dose level and duration of the experiment was selected on the
basis of an acute toxicity study to obtain highest hepatic concentrations
without substantial hepatic toxicity. The livers of the animals were pooled and
15
-------
homogenized, and the homogenate was processed for fractionation of protein,
ribosomal RNA, and DNA. These fractions were analyzed for radioactivity. The
maximum unextractable radioactivity from liver protein fraction was 60 p moles
of TCDD/mole of amino acid residue. The radioactivity associated with ribosomal
RNA and DNA was very low. That associated with microsomal RNA corresponded to
only 12 p moles of TCDD/mole of nucleotide residue, and that associated with DNA
corresponded to only 6 p moles of TCDD/mole of nucleotide residue.
These two studies essentially demonstrated that TCDD was transformed to a
reactive electrophile metabolite and showed significant covalent binding with
cellular proteins, but less significant binding with DNA.
16
-------
III. MUTAGENICITY
MUTAGENICITY OF 2,4,5-T
The mutagenicity of 2,4,5-T was evaluated by Ercegovlch et al. (1977),
employing the procedure of Ames using five strains of SamoneHa typhimurium
without activation. The authors concluded that 2,4,5-T is non-mutagenic.
Anderson and Styles (1978) reported that 2,4,5-T at concentration ranges
from 4 to 2500 ug per plate did not cause reversions in any of the four strains
of Samonella typhimurium (TA 1535, TA 1538, TA 98, and TA 100) with or without
microsomal activation. Several other investigators have reported negative
responses with 2,4,5-T in bacterial test systems which have been summarized in a
review by Grant (1979). Zetterberg (1978) found that 2,4,5-T increased the back
mutation frequency in the histidine defective strain of Saccharomyces cerevisiae
at pH values below 4.5, by approximately 300 fold at 40 mg/ml and 5000 fold at
60 mg/ml. However, the percent of survivors at the lower concentration was less
than 5% and at the higher concentration less than 0.1%. The author concluded
that 2,4,5-T is unlikely to cause mutations in a near neutral environment but
oral administration may increase the risk" of somatic mutation in the gastric
tract where pH values are as low as 1.2. The 2,4,5-T used in these studies
contained less than 1 ppm dioxins.
Majumdar and Golia (1974) fed Drosophila melanogaster males 1000 ppm 2,4,5-T
for 15 days and found a small increase in the percentage of sex-linked recessive
lethals by 0.61" over controls values of 0.05%. The herbicide was reported to
contain no detectable amount of dioxin. Similar findings by Magnusson et al.
(1977) also showed 2,4,5-T to be weakly mutagenic in Drosophila. In a parallel
experiment, the known mutagen ethylmethanesulfonate at 250 ppm increased the
incidence of sex-linked lethals by 13.65%. The CAG evaluated the negative
17
-------
report by Vogel and Chandler (1974) concerning the mutagenicity of 2,4,5-T in
the Drosophila melanogaster sex-linked recessive lethal tests as not providing
sufficient data to classify the compound as postive or negative. The number of
chromosomes analyzed by Vogel and Chandler was not large enough for a meaningful
test. The negative report by Rasmuson and Svahlin (1978) is also inadequate to
evaluate 2,4,5-T as a non-mutagen in Drosophila. These authors developed a test
for detecting somatic cell mutagenesis and tested EMS, 2,4-D, and 2,4,5-T using
it. EMS and 2,4-D were reported to be positive at concentrations of 500 and 25
ppm, respectively, but 2,4,5-T was reported to be negative. Since these results
were obtained from the first set of experiments in a new test system, it is not
possible to compare them in a meaningful way to other tests such as the
sex-linked recessive lethal test in Drosophila melanogaster. Furthermore, the
report suffered from other deficiencies such as a lack of information concerning
compound purity and dose-response.
Fujita et al. (1975) reported chromosomal abnormalities in in vitro
cytogenetics studies of human lymphocytes exposed to 10"^ to 10"^M 2,4,5-T.
Chromosome breaks, deletions, and rings were observed. Chromatid breaks
increased with increasing concentrations of 2,4,5-T. It was not possible to
distinguish whether this was a toxic effect or a potential genetic effect.
Yefimenko (1974) reported on an acute and chronic exposure to butyl ether
2,4,5-T in in vivo cytogenetics tests on gonadal and somatic tissue in male
albino rats. Twenty-four hours after a single oral administration at doses of
1, 0.1, and 0.01 ug/kg, structural damage to bone marrow cell chromosomes was
observed either as breaks or as true aberrations or rearrangements.
Chronic-exposure effects to the gonads were observed after exposure for
2-1/2 months to a dose of 0.1 ug/kg. The following effects were observed at the
termination of the experiment (7 months): testicular atrophy, decreased sperm
18
-------
count, desquamated tubules, and aberrant cells in the germinal epithelium.
These effects persisted after exposure was terminated.- Chromosomal aberrations
were also observed during chronic dosing. The authors' methodology appears to
be inadequate, however, and thus no valid conclusions can be drawn from this
study. Majumdar and Hall (1973) reported that intraperitoneal injections of
2,4,5-T (containing no measurable amount of TCDD) into gerbils at concentrations
of 350 mg/kg for 5 days produced 8.2, 4.6, and 1.8 percent incidences of
chromatid gaps, chromatid breaks, and fragments, respectively, in bone marrow
cells. Control values were given as 1.0% for gaps, 0.2% for breaks, and 0.2%
for fragments. When the animals were treated at lower doses, no significant
increases in chromosomal abnormalities were observed. Jensen and Renberg (1976)
performed cytogenetic tests on mice injected with 2,4,5-T at 100 mg/kg. They
reported no increase over control values in incidences of micronuclei in
polychromatic or normochromatic erythrocytes, or polychromatic cells 24 hours or
0 days after the injection of the chemical. They were, unable to confirm the
cytogenic effect reported by. Majumdar and Hall (1973), but pointed out that they
used extremely high doses which might cause toxic effects leading to cell death
and chromosomal fragmentation.
Renner (1979) reported that 2,4,5-T induces a weak positive response in the
SCE test using Chinese hamster bone marrow cells. Four SCE's per cell were
observed in the control animals compared to 7/cell at 100 mg/kg and 8/cell at
250 mg/kg. This report cannot be evaluated, however, because no information is
provided concerning the route of administration, the number of animals used, the
number of cells scored per animal, the purity and source of the compound, and
whether or not the test was repeated.
Kilian et al. (1975) examined lymphocytes for chromosomal aberrations in
industrial workers exposed to 2,4,5-T in a Midland Michigan plant and compared
19
-------
them with a control group of workers prior to employment at the plant. They
reported that there was no significant differences in the aberration rates among
the control group (84 people) and those exposed for less than a year (16 people)
or more than a year (17 people). However, the study did not indicate
concentrations of 2,4,5-T workers might have been exposed to, and for each
subject only about 20 cells were scored for chromosomal aberrations.
20
-------
MUTAGENICITY OF TCDD
Hussain et al. (1972) n .orted positive results in-three microbial test
systems using a 99% pure TCDD sample obtained from the Food and Drug
Administration (FDA). Reversion to streptomycin independence in Escherichia
coli Sd-4 occurred with high frequency at a concentration of 2 ug
TCDD/ml. Reversion at the histidine locus of Salmonella typhimurium TA 1532
occurred at concentrations between 2 to 3 ug/ml. This indicates that TCDD
produces frameshift mutations by intercalation between base-pairs of DNA. A
doubling in the frequency of prophage-induction was observed in E. coli K-39
exposed to TCDD. These studies were not performed with metabolic activation,
indicating that TCDD is a direct-acting mutagen.
Seiler (1973) classified TCDD as a strong mutagen (where the ratio of number
of revertants from treated plates per 10® bacteria divided by the number of
spontaneous revertants per 10® bacteria is greater than 10) in the TA 1532
Salmonella strain which detects revertants through frameshift mutations.
However, this report did not.give the source or purity of TCDD, the
concentration used in the assay, the toxicity of the compound where mutagenic
activity occurs, or whether microsomal activation was necessary.
However, McCann (personal communication) tested TCDD to be negative in the
standard plate test with strain TA 1532, with and without microsomal activation,
and Nebert et al. (1976) also reported that TCDD was not mutagenic in the
Salmonella in vitro assay. The differences between these laboratory results and
those discussed above could be due to several factors such as treatment
protocols, solubility problems of TCDD, and the high toxicity of this compound.
The Food and Drug Administration conducted a somatic in vivo cytogenetics
screening study on TCDD in rats and got negative results (Green 1975). Separate
experiments were performed with five multiple intraperitoneal doses or a single
21
-------
oral dose regimen with sacrifice at 1 or 29 days. Toxicity, as indicated by
slight body weight loss, was observed in the multiple dose study only at the
highest dose used. This indicates that the dose levels may have been too low.
Khera and Ruddick (1973) dosed male rats orally with 4 or 8 mg/kg/day TCDD for 7
days. These doses were acutely toxic and 20 survivors at the lower dose and 6
survivors at the high dose were mated after treatment seven times at 5-day
intervals. Reproductive values indicated no occurrance of dominant lethal
mutations. Green et al. (1977) studied the cytogenetic effects of TCDD on rat
bone marrow cells. Male and female animals received 0.25, 1.00, 2.00, and 4.00
mg/kg TCDD by gavage twice a week for 13 weeks. The authors examined bone
marrow cells at the end of treatment (approximately 50 cells per animal) for
abnormalities. They concluded TCDD produces chromosomal aberrations in bone
marrow cells but the effect is not one of great magnitude.
Chromosome analyses on 12 hospital patients exposed to TCDD in a July 1976
Seveso, Italy factory accident (Department of Health, Education, and Welfare,
1976) were examined for chromosomal lesions (gaps, chromatid and chromosome
breaks, and rearrangements). These analyses presented at the DHEW-Subcommittee
on Environmental Mutagenesis meeting, October 12, 1976 were of somatic cells
from males and females ranging in age from 2 to 28. One patient had 19" cells
(presumed blood cells) that were classified as having chromosomal lesions,
another had 10*. The remaining patients had values comparable to control levels
of 5 to 7%. Results from chromosome analyses of maternal peripheral blood,
placenta, and fetal tissue in 17 women exposed to TCDD (amounts not given) who
underwent spontaneous abortions were inconclusive. Reggiani (1977) reported
that the frequency of spontaneous abortions in the Seveso zone did not
significantly change nor did the incidence of malformations as a result of
exposure to TCDD, evfen at the regions where the exposure was estimated through
22
-------
soil analysis to be greater than 10 ug/kg. Similar conclusions were reached by
Tuchmann-Duplesis (1977). Reports by both Reggiani (1977) and Tuchmann-Duplesis
(1977) state no increase in abnormal cytological changes in tissues of aborted
fetuses or in maternal blood in the Seveso zone during the exposure incidence to
TCDD. However, these findings are poorly documented and complete experimental
procedures and design used to evaluate the data were not available.
Furthermore, it appears from these reports that only gross macroscopic
alterations were sought and not microscopic lesions which are more difficult to
assess. Such lesions are very dangerous in that they may survive and be carried
to future generations.
CONCLUSIONS
There is some evidence that 2,4,5-T appears to be a weak mutagen causing
point mutations. The best evidence for this is in Drosophila and Saccharomyces
cerevisiae. However, evidence in Saccharomyces cerevisiae indicates the potency
of the mutagenic effect may be related to the ionization of the carboxyl group
of 2,4,5-T and is increased under more acidic conditions. At the present time,
epidemiological evidence and cytogenetic studies for mutagenicity concerning
TCDD are inconclusive. Also, the reported effects of TCDD as a "frameshift
mutagen" are inconsistent. Because TCDD is structurally similar to acridines
which produce frameshift mutations by intercalation in the DNA base-pairs, it is
recommended that the ability of TCDD to induce forward mutations in systems such
as mammalian cells in culture and the sex-linked recessive lethal tests in
Drosophila be examined. Also, it is recommended that the mutagenic activity of
TCDD be re-tested in bacteria using a series of both strains which detect
frameshift and base-pair mutations.
23
-------
IV. TOXICITY
ANIMAL TOXICITY
Toxicity of 2,4,5-T
Oral LD50 levels for 2,4,5-T as referenced by the National Institute of
Occupational Safety and Health (1976), are shown in Table 2 below.
TABLE 2. ORAL LD50 LEVELS FOR 2,4,5-T
Species
LD50
Dog
100 mg/kg
Rat
300 mg/kg
Guinea pig
381 mg/kg
Mouse
389 mg/kg
In a study cited by Rowe and Hymas (1954), no adverse effects were seen in
dogs given 2,4,5-T by oral administration five times a week, for 90 days, at
doses of 25 and 10 mg/kg. However, at a dose level of 20 mg/kg, all four dogs
died and showed mild liver and kidney changes. A Dow Chemical Company internal
report (1971) cited by EP.A's 2,4,5-T Advisory Committee summarized a study using
2,4,5-T containing 0.5 ppm TCDD. The 2,4,5-T was fed to male and female rats
for 90 days at dose levels of, 0 to 100 mg/kg/day. No toxic effects were
observed at doses of 30 mg/kg or lower. At a dose level of 100 mg/kg/day, some
hematological effects and weight loss were noted, but toxic effects were
described as minor and inconsistent.
24
-------
Toxicity of TCDD
TCDD is one of the most toxic chemicals known to man. Oral LD50 values,
shown in Table 3, range from 0.6 ug/kg orally for the male guinea pig to 275
ug/kg dermally for the rabbit. Deaths typically occur about a week or more
after treatment.
Poland et al. (1971) cite a study in which rapid death in guinea pigs
followed dermal application of the tarry residues from TCDO synthesis. When
rabbit ears were painted with soil extracts contaminated with TCDD,
hyperkeratosis and liver pathology were observed in the rabbits (Kimbrough
1974).
Kociba et al. (1978) conducted a 2-year chronic toxicity and oncogenicity
study of TCDD in rats. In this study, the animals were maintained for 2 years
on diets supplying 0.1, 0.01, and 0.001 ug TCDD/kg/day. Aside from carcinogenic
effects, ingestion of 0.1 ug/kg/day caused increased mortality, decreased weight
gain, slight depression of erythroid parameters, increased urinary excretion of
porphyrins and delta-aminolevulinic acid, along with increased serum activities
of alkaline phosphatase.
In chronic and acute oral TCDD toxicity studies on several animal species,
the liver, thymus, and spleen have consistently been the target organs. Liver
damage, including necrotic and degenerative changes, lipid accumulation, and
increased liver weight, have been observed in mice, rats, and guinea pigs
following TCDD treatment (Vos et al. 1974, Jones and Greig 1975, Gupta et al.
1973, Goldstein et al. 1973, Kimmig and Schultz 1957). Liver damage was
markedly greater in rats receiving a comparable dose (Gupta et al. 1973). It
has been suggested that the fatty liver observed in mice may result from the
starvation and loss of body weight that occur following TCDD treatment, or may
be due to the induction of mixed-function oxidases (Jones and Greig 1975).
25
-------
TABLE 3. LETHALITY OF 2,3,7,8-TETRACHL0R0DIBENZ0-P-DI0XINa» b
Species
and sex
Route of
admi ni strati on
Time of
death, days
post administration
LD50
ug/kg
Dose
ug/kg
Number
deaths/
number
treated
Rat, male
Oral
9-27
22
8
16
32
63
0/5
0/5
10/10
5/5
Rat, female
Oral
13-43
45
Guinea pig!
male
Oral
5-34
0.6
Guinea pig,
mal ec
Oral
9-42
2.1
Rabbit, mixed
Oral
Skin
6-39
12-22
115
275
Intraperitoneal
6-23
(all doses)
32
63
126
252
500
0/5
2/5
2/5
2/5
3/5
Dogs, male
Oral
9-15
(all doses)
300
3000
0/2
2/2
Dogs, female
Oral
—
30
100
0/2
0/2
Mice, male
Oral
—
114
—
Responses to individual doses are given in those cases in which an LD50
could not be calculated.
bAll values are from Schwetz et al. (1973), except those for male mice,
which are from Yos et al. (1974).
CA sample that was more than 99% pure was used. All other tests except
the mice study used TCDD that was 91" pure.
26
-------
Atrophy of the thymus and spleen has also consistently be n found in
laboratory animals (Vos et al. 1974, Kociba et al. 1975, Gupt et al. 1973).
Yos et al. (1973) reported that cell-mediated immunity was suppressed in guinea
pigs and mice in TCDD-induced lymphoid depleted thymuses. Thigpen et al. (1975)
found that mice receiving 1 ug/kg or more of TCDD by stomach tube once a week
for 4 weeks had increased susceptibility to Salmonel1 a infection. Female
monkeys fed TCDD for 9 months showed hypocellularity of the bone marrow and
lymph nodes as well as hypertrophy, hyperplasia, and metaplasia of the bronchial
tree, epithelium, bile ducts, pancreatic ducts, and salivary gland ducts (Allen
et al. 1977.).
Other effects of TCDD ingestion include suppression of reproductive function
in rats (Kociba et al. 1975) and disturbance of the hematopoietic system with
occasional hemorrhaging in monkeys, rats, and mice (Allen et al. 1977, Kociba et
al. 1975, Vos et al. 1974). TCDD interferes with the biosynthetic pathway of
heme by inducing delta-ami no!evulinic acid synthetase ( -ALA), which results in
hepatic porphyria in mice and rats (Goldstein et al. 1976). Increased urinary
excretion of uroporphyrins has been observed in rat feeding studies (Kociba et
al. 197", Goldstein et al. 1976).
TOXICITY OF 2,4,5-T, 2,4,5-TRICHLOROPHENOL, AND TCDD IN HUMANS
The most consistently reported toxic effect of 2,4,5-T,
2,4,5-trichlorophenol, and TCDD to humans is chloracne, a disfiguring and
long-term dermatitis. This has occurred in 2,4,5-T factory workers (Bauer et
al. 1961, Poland et al. 1971), 2,4,5-trichlorophenol workers (Kimmig and Schulz
1957, Bauer et al. 1961, Bleiberg et al. 1964, Goldmann 1972), and laboratory
workers accidentally exposed to TCDD (Oliver 1975). It has also been observed
in exposed populations following the accidental production of TCDD in exothermic
27
-------
reactions at chemical plants (Hay 1976, Kimmig and Schulz 1957, May 1973). In
the 2,4,5-trichlorophenol plant accident in Seveso, Italy (1976), 300 to 500
grams of TCDD are believed to have been deposited in the most contaminated
areas, with lesser amounts in surrounding areas (Hay 1976). In the other
incidents, the level of TCDD present is not estimated.
There are scattered reports of hepatotoxic effects including abnormal liver
function tests and pathological changes in the liver (Kimmig and Schulz. 1957,
Bauer et al. 1961, Bleiberg et al. 1964, Poland and Smith 1971). Eleven of 14
men exposed to TCDD during an exothermic reaction at a 2,4,5-trichlorophenol
plant had abnormal liver function tests, but after 10 days without TCDD
exposure, most tests were normal. Bleiberg et al. (1964) reported that 11 of 29
workers at a 2,4,5-trichloropenol plant had porphyria, but in a study of the
same workers 6 years later, Poland et al. (1971) found no overt clinical cases
of the disease. They suggested that the change may have been due to increased
attention to worker safety or to a decrease in TCDD contamination.
Bauer et al. (1961) found one case of bloody urine in exposed workers.
Hemorrhagic cystitis was reported in a 6-year-old girl who was playing in a
horse area contaminated with 2,4,5-TCP (TCDD concentration of 31 to 33 ppm)
(Carter et al. 1975). Other organ system effects that have been reported are
gastrointestinal tract disturbances (Kimmig and Schulz 1957, Poland et al.
1971), neurological disturbances (01 iver. 1975), and respiratory and cardiac
disorders (Bauer et al. 1961). Psychological changes, including emotional
instability, lethargy, diminished libido, and high manic scores on psychological
tests, have also been noted (Poland et al. 1971, Oliver 1975, Bauer et al.
1961).
28
-------
Y. CARCINOGENICITY
CARCINOGENICITY OF 2,4,5-T IN MICE
Muranyi-Kovacs et al. (Oral) Mouse Study (1976)
Inbred C3Hf and XVII/G strains of mice were used. They were given 100
mg/liter of (2,4,5-trichlorophenoxy)acetic acid (2,4,5-T) in drinking water for
2 months, beginning at 6 weeks of age. (The 2,4,5-T product contained less than
0.05 ppm of 2,3,7 8-tetrachlorodibenzo-p-dioxin.) Thereafter, mice were given
2,4,5-T mixed with a sterile, commercial diet (UAR 1136) at concentrations of 80
ppm. It was not stated whether these levels represented maximum tolerated
values. However, the authors indicated that this dose was 1/40 of the LD5Q.
The mice were examined weekly for their general health and for the presence
of tumors. They were allowed to die or* were killed in extremis. Complete
necropsies were performed and grossly altered organs were examined
histologically. The urinary bladder was distended with fixative in mice
suspected of having lesions.
C3Hf control male mice survived an average of 630 days; treated male mice,
511 days (P = 0.001); control females, 680 days; and treated females, 620 days.
Survival times for XVII/G control male mice were 521 days; for treated male
mice, 583 days; control females, 569 days; and for treated females, 641 days
(P = 0.01).
Tumor presence in C3Hf female mice ingesting 2,4,5-T is indicated in Table
4. The results show that 12 of 25 C3Hf female mice (48") ingesting 2,4,5-T
developed tumors of all types, as compared to 9 of 44 control female mice (21*)
(P = 0.03). No other strain-sex combination yielded statistically significant
values, as evidenced by the data in Tables 4 and 5. Benign and malignant tumors
were considered together in this study. The authors stated that the "hepatomas"
29
-------
TABLE 4. TUMORS IN C3Hf MALE AND FEMALE MICE INGESTING 2,4,5-Ta
Dose (ppm)
Sex
Lung
Liver
Leukemia
Other
Total
No. of mice
with tumors
0
M
2
19
0
lb
22
21/43 (49%)
80 ppm
M
0
10
2
ic
13
12/22 (555)
0
F
5
3
1
0
9
9/44 (21%)
80 ppm
F
0
4
3
6d
13
12/25 (48%)
^Effective number of mice surviving longer than 300 days or developing a
tumor before 300 days of age.
^Pleomorphic salivary gland tumor.
cFibrosarcoma; not included are one hyperplastic lesion of the urinary
bladder and one hyperplastic lesion of the forestomach.
^osteogenic sarcoma, 2 sarcomas, 2 cutaneous tumors, and 1 tumor of the
cervix.
TABLE 5. TUMORS IN XVII/G MALE AND FEMALE MICE INGESTING 2,4,5-Ta
Dose (ppm)
Sex
Lung
Liver
Leukemi a
Other
Total
No. of mice
with Tumors
0
M
22
4
0
lb
27
25/32 (78%)
80
M
14
0
1
ic
16
15/20 (75%)
0
F
20
0
2
2d
24
21/40 (53%)
80
F
15
0
1
0
16
16/19 (84%)
Effective number of mice are mice surviving longer than 300 days or
developing a tumor before 300 days of age. In the XVII/G male mice, there was
no significant difference between the number of tumor-bearing mice among treated
animals as compared with controls, as shown in the table above.
bl forestomach tumor.
C1 urinary bladder papilloma; not included are 2 hyperplastic lesions of
the urinary bladders.
d2 hemangiomas.
30
-------
and lung tumors, which were carcinomas and alveologenic adenomas, occurred in
the same proportions in control and treated mice. Treated C3Hf females had
several tumors at sites not found in the controls. The authors reported a
significant increase in total tumors in one strain and one sex of rats at one
dose level. In reaching this conclusion, they used the Peto method and
distinguished between incidental and nonincidental tumors.*
To clarify questions concerning the design, execution, and interpretation of
this study, the CAG communicated with the principal author at the Curie
Foundation, Marseilles, France. From this discussion and from the published
account of this discussion it is concluded that: 1) this study was very
insensitive because insufficient numbers of animals were used in the treatment
groups; 2) the care of the animals was inadequate; 3) because the dose used, 80
ppm, was only 1/4-0 of the LD50, and appears to be less than the maximum
tolerated dose; 4) histologic examination of all animal tissues was not
performed; and 5) only macroscopically altered tissues were examined
histologically. In addition, the author recommended that more adequate studies
be conducted in a greater number of species. Because of the severe deficiencies
in the study, the CAG concluded that this study does not provide significant
evidence for either the carcinogenicity or non-carcinogenicity of 2,4,5-T.
Muranyi-Kovacs et al. (Subcutaneous) Mouse Study (1977)
In this study, the authors administered 2,4,5-T to two strains of mice, C3Hf
and XVII/G. Subcutaneous injections were given at 10 mg/kg of body weight in an
xThese results are not considered to be evidence of an oncogenic response
because there is no valid basis for grouping tumors at all sites or for
distinguishing between incidental and nonincidental tumors. The author did not
report any increases in tumors for any specific target site.
31
-------
aqueous solution on days 1, 3, 6, and 10 of the animals' lives. Th< sex,
strain, number of animals, survival time, and animals with ^umors at j shown in
Table 6.
TABLE 6. SURVIVAL TIME AND TUMOR INCIDENCE IN 2,4,5-T
TREATED MICEa
Strai n
Sex
Dose
Number of
Survival time
Percent of animals
mg/kg
animals
tumor-beari ng
with tumors
XVII/G
M
0
32
25/32
78
M
4 x 10
15
13/15
87
XVII/G
F
0
40
21/40
53
F
4 x 10
15
4/15
25
C3H/f
M
0
43
21/43
49
M
4 x 10
11
4/11
36
C3H/f
F
0
44
9/44
21
F
4 x 10
14
3/14
21
dTCDD content of 2,4,5-T was less than 0.05 ppm.
As indicated, there is no observed increased incidence of tumor-bearing animals
as compared to the treated animals of both sexes and strains. However, this
study was so incompletely reported that the details of the methodology cannot be
discerned. In addition, this study was very insensitive because insufficient
numbers of animals were used in the treatment groups, and only a few
subcutaneous doses were administered. Therefore, these studies do not provide
significant evidence for either the carcinogenicity or non-carcinogenicity of
2,4,5-T.
32
-------
Innes et a!. (Bionetics Laboratories 1968) (Oral) Mouse Study (1969)
The maximum tolerated dose of 2,4,5-T* was given to two hybrid strains of
mice, (C57BL/6 x C3H/Anf)F]_, B6C3F1 designated as "strain X," and (C57B/6 x
AKR)Fi, B6AKF1 designated as "strain Y." There were 18 treated mice and 18
untreated control mice of each strain and each sex. Each day, beginning at 7
days of age, 21.5 mg/kg of 2,4,5-T in 0.5% gelatin was administered by stomach
tube. After weaning at 28 days of age, 60 ppm of 2,4,5-T was mixed directly in
the diet and provided ad libitum. Treatment was continued for approximately 18
months.
At this time mice were killed and grossly examined both internally and
externally in the areas of the neck glands and the thoracic and abdominal
cavities. Histologic examination of major organs and all grossly visible
lesions was performed. Thyroid glands were not examined. The postmortem
results are given in Tables 7 and 8.
The results of the oral mouse study indicate that there was no significant
difference between the 2,4,5-T-treated and control groups of mice with respect
to tumors at specific sites, or total number of tumor bearing animals. This
study, however, does not provide significant evidence for the
non-carcinogenicity of 2,4,5-T because of certain defects in its design. The
use of small numbers of animals and the duration of the study, which was only 18
months rather than the entire lifetime, made the study relatively insensitive
for detecting an oncogenic effect.
* The Bionetics study did not report the level of TCDO contamination in the
2,4,5-T used. The 2,4,5-T used in a reproductive study conducted at
approximately the same time as the Bionetics study was reported to contain 30
ppm TCDO. It is possible that the contaminant of 2,4,5-T used in the Bionetics
study was the same as that of the 2,4,5-T used in the reproductive study.
However, this conclusion is far from certain without actual chemical analysis of
the 2,4,5-T used in the Bionetics study.
33
-------
TABLE 7. MALE AND FEMALE MICE INGESTING 2,4,5-T WITH TUMORS OF VARIOUS ORGANS
TUMOR TYPE
Reticulum Pulmonary
Strain Oose (ppm) cell sarcoma adenoma & carcinoma Hepatoma
Male
Female
Male Female
Male Female
"X"
matched
0
0/15
1/18
2/15 1/18
3/15 0/18
"X"
pooled
0
5/79
4/87
5/79 3/87
8/79 0/87
"X"
60
1/18
0/18
1/18 1/18
4/18 0/18
ii y ii
matched
0
0/18
1/15
3/18 0/15
0/18 0/15
ii y ii
pooled
0
1/90
3/82
10/90 3/82
5/90 1/82
ii y ¦<
60
2/18
1/18
0/18 0/18
1/18 0/18
TABLE 8.
MALE AND
FEMALE MICE
INGESTING
2,4,5-T WITH TUMORS
AT ALL SITES
Mai e
Female
"X"
matched
0
5/15 (33%)
2/18 (11%)
"X"
pooled
0
22/79 (28%)
8/87 (9%)
"X"
60
6/18 (33%)
1/18 (6%)
ji y 11
matched
0
3/18 (17%)
1/15 (7%)
ii Y >¦
pooled
0
16/90 (18%)
7/82 (9%)
ii y
60
3/18 (17%)
2/18 (11%)
34
-------
Innes et al. (Bionetics Laboratories 1968) (Subcutaneous) Mouse Study (1969)
2,4,5-T in dimethysulfoxide (DMSO) was given as a single subcutaneous
injection (215 mg/kg) to two strains of male and female mice (same strains as in
the oral study) at approximately 28 days of age. The mice were observed for
approximately 18 months. At that time mice were killed and examined grossly,
both internally and externally, in the areas of the neck, glands, and thoracic
and abdominal cavities. Histologic examinations of all major organs, as well as
all grossly visible lesions, were made. Thyroid glands were not examined. The
authors stated that histopathologic data did not show a statistically
significant difference between the 2,4,5-T-treated and control groups either
with respect to tumors at specific sites, or total number of tumor-bearing
animals. However, this study suffered from the same deficiencies as the Innes
et al. oral study. In addition, single subcutaneous dose studies are considered
to be highly insensitive for detecting an oncogenic response. Therefore, the
CAG does not consider this study to provide significant evidence of the
non-oncogenicity of 2,4,5-T.-
35
-------
CARCINOGENICITY OF 2,4,5-T IN RATS
Kociba et al. (Oral) Rat Study (1978, 1979)
The cancer bioassay study of 2,4,5-T in rats (TCDD content was not
detectable in 2,4,5-T with a detection limit of 0.33 ppb) was performed by
Kociba et al. (1978, 1979). In this study, one group of 86, and three groups of
50 Sprague-Oawley (Spartan substrain) rats of each sex were administered 0, 3,
10, and 30 mg/kg body weight/day, respectively, via the diet, for periods up to
2 years. There is some question as to the actual dosage. Adequate information
is lacking to show that the amount of 2,4,5-T found in the food by chemical
analysis actually remained constant in a given feed lot during the entire period
of consumption of that given lot. Based on the analytical data, the actual
doses given were slightly lower than the nominal doses.
There are certain aspects of this study which reduced its sensitivity for
detecting a carcinogenic response. First, there was a very high early mortality
among all groups of males and females (Tables 9 and 10). The mortality data
show that at the termination of the study more than 50% of the females had died
in the control group as well as in each of the treated groups (the mortality was
76% in the 10 mg/kg female group). Among males, mortality was approximately 92%
in the controls and ranged from 78% to 92% in the treated groups. Very high
mortality in both males and females was observed as early as 21 months. This
early mortality is important because it reduces the number of animals at risk
for late developing tumors.
The second factor which reduced the sensitivity of this study was the
relatively high incidence of spontaneous tumors in some organ sites in the
controls. For example, among 86 control males, three hepatocellular carcinomas
and four hepatocellular neoplastic nodules were found.
The data reported by Kociba et al. are presented in Table 11J
36
-------
TABLE 9. CUMULATIVE MORTALITY DATA OF MALE RATS MAINTAINED ON DIETS
CONTAINING 2,4,5-T FOR 2 YEARS
Dose Level (mg/kg/day)
Total no. of
rats studied
86
50
50
0
30
10
3
No. dead
No. dead
No. dead
No. dead
{% dead)
(% dead)
(% dead)
(% dead)
Original no.
in group
86
50
50
50
Days on test
0-30
0
0
0
0
31-60
0
0
0
0
61-90
1(1.2)
0
0
0
91-120
1(1.2)
0
0
0
121-150
1(1.2)
0
0
0
151-180
1(1.2)
0
0
0
181-210
1(1.2)
0
0
0
211-240
1(1.2)
0
0
0
241-270
1(1.2)
0
1(2.0)
0
271-300
2(2.3)
0
1(2.0)
0
301-330
2(2.3)
0
1(2.0)
0
331-360
2(2.3)
0
1(2.0)
1(2.0)
361-390
2(2.3)
2(4.0)
2(4.0)
2(4.0)
391-420
5(5.8)
2(4.0)
2(4.0)
3(6.0)
421-450
6(7.0)
2(4.0)
4(8.0)
4(8.0)
451-480
9(10.5)
4(8.0)
9(18.0)
6(12.0)
481-510
10(11.6)
6(12.0)
12(24.0)3
10(20.0)
511-510
16(18.6)
8(16.0)
22(44.0)3
12(24.0)
541-570
23(26.7)
11(22.6)
24(48.0)3
14(28.0)
571-600
32(37.2)
16(32.0)
29(58.0)3
23(46.0)
601-630
47(54.6)
19(38.0)
37(74.0)3
30(60.0)
631-660
67(77.9)
24(48.0)a
38(76.0)
32(64.0)3
661-690
74(86.0)
27(54.0)a
42(84.0)
34(68.0)3
691-720
77(89.5)
32(64.0)3
45(90.0)
38(76.0)3
721-728
79(91.7)
39(78.0)a
46(92.0)
40(80.0)3
50
dStatistically significant difference from control values by Fisher's
Exact Probability Test, P < 0.05.
37
-------
TABLE 10. CUMULATIVE MORTALITY DATA OF FEMALE RATS MAINTAINED ON DIETS
CONTAINING 2,4,5-T FOR 2 YEARS
Dose Level
(mg/kg/day)
U
30
1U
3
No. dead
No. dead
No. dead
No. dead
{% dead)
(% dead)
{% dead)
(% dead)
Original no.
in group
86
50
50
50
Days on test
0-30
0
0
0
0
31-60
0
0
0
0
61-90
0
0
0
0
91-120
0
0
0
1(2.0)
121-150
0
0
0
1(2.0)
151-180
0
0
1(2.0)
1(2.0)
181-210
1(1.2)
0
1(2.0)
1(2.0)
211-240
1(1.2)
0
1(2.0)
1(2.0)
241-270
1(1.2)
0
1(2.0)
1(2.0)
271-300
1(1.2)
1(2.0)
1(2.0)
1(2.0)
301-330
2(2.3)
1(2.0)
3(6.0)
2(4.0)
331-360
2(2.3)
2(4.0)
3(6.0)
2(4.0)
361-390
2(2.3)
3(6.0)
3(6.0)
3(6.0)
391-420
4(4.6)
3(6.0)
4(8.0)
3(6.0)
421-450
5(5.8)
3(6.0)
5(10.0)
3(6.0)
451-480
9(10.5)
7(14.0)
8(16.0)
5(10.0)
481-510
12(14.0)
9(18.0)
13(26.0)3
7(14.0)
511-540
18(21.0)
11(22.6)
14(28.0)
10(20.0)
541-570
20(23.2)
12(24.0)
15(30.0)
12(24.0)
571-600
25(29.1)
15(30.0)
16(32.0)
16(32.0)
601-630
29(33.7)
18(36.0)
21(42.0)
20(40.0)
631-660
34(39.5)
24(48.0)
25(50.0)
23(46.0)
661-690
41(47.7)
26(52.0)
33(66.0)a
26(52.0)
691-720
46(53.5)
27(54.0)
35(70.0)3
29(58.0)
721-732
46(53.5)
28(56.0)
38(76.0)a
29(58.0)
Total no. of
rats studied
86
50
50
50
dStatistically significant difference from control values
by Fisher's
Exact Probability Test, P < 0.05.
38
-------
TABLE 11. STRATIFIED SQUAMOUS CELL CARCINOMA OF THE TONGUE OF SPRAGUE-DAWLEY
RATS FED WITH PURIFIED 2,4,5-T
Kociba
2,4,5-T
2,4,5-T dosage in mg/kg/day
Controls 3D
(P-value)a
10
7
Test for Trend'3
Males
1/83
4/49 0/46
(P = 0.063)
1/50
< 0.03
Females
0/83
1/49 0/48
(P = 0.371)
0/48
N.S.c
dP values determined by Fisher's Exact Test (one-tailed).
^Cochran's test for trend, one-tailed, scoring = 0, 1, 2, 3.
CN.S. = not significant at P = 0.05.
The increase in squamous cell carcinoma of the tongue in males at the 30
mg/kg/day dose level is marginally statistically significant (P = 0.063). Also,
the dose-related trend for the incidence of tongue tumors in males is
statistically significant in the Cochran-Armitage Test (P < 0.03).
Examination of male Sprague-Dawley rats in the Dow studies (Spartan
substrain) for historical controls found the following incidence of squamous
cell carcinomas of the tongue as illustrated in Table 12 (taken from selected
Tables provided to EPA by Dow which summarize the results of six Dow studies).
39
-------
TABLE 12. TONGUE TUMOR INCIDENCE IN HISTORICAL CONTROLS
Tongue carcinomas/
no. examined
y
o
1. Kociba TCDD study
0/76
0
2. Kociba 2,4,5-T study
1/83
1.2
3. Dow Historical Control
Study No. 4
1/63
1.6
4. Dow Historical Control
Study No. 5
1/75
1 .3
5. Dow Historical Control
Study No. 6
1/49
2.0
Total
4/346
1.2
A comparison of the 4/49 tongue carcinomas vs. 4/346 historical controls
yields a value of (Z = 3.26) P < 0.001, one-tailed test.
Dr. Robert Squire, pathologist at the Johns Hopkins School of Medicine and
consultant to the Carcinogen Assessment Group, performed an evaluation of
histopathological slides from Dow Chemical Company's 2-year rat feeding studies
of 2,4,5-T (see A^iendix 3). A comparison of his findings to Dr. Kociba's
histopathological evaluation of tongue lesions are summarized in Table 13. The
histological analysis of Dr. Robert Squire revealed the occurrence of squamous
cell carcinomas of the tongues in 5 of 48 high dose males as compared to the 4
of 49 high dose males reported by Dr. Kociba. Dr. Squire and Dr. Kociba both
found one tongue carcinoma out of 83 controls. The histopathologic evaluation
of the extra tongue carcinoma observed by Dr. Squire in the high dose group was
confirmed by Dr. Goodman, pathologist, who reviewed six slides with tongue
lesions in male rats from the Dow Chemical study. Five of these were squamous
40
-------
cell carcinomas of the tongue in high dose males and 1 was in a control male
(Goodman 1980).
The increase in squamous cell carcinomas of the tongue in males at the 30
mg/kg/day dose level is statistically significant (P = 0.025) compared to
matched controls when using Drs. Squire's and Goodman's diagnoses. These
results provide highly suggestive evidence of the carcinogenicity of essentially
pure 2,4,5-T.
41
-------
TABLE 13. ORS. SQUIRE'S AND KOCIRA'S REVIEW OF DOW 2,4,5-T ORAL RAT STUDY (8/15/80)
Sprague-Dawley Rats - Spartan Substrain (2 yrs.)
MALES
Tissue and Diagnosis Dose Levels (mg/kg/day)
0 3 10 30
(control)
SKSKSKS K
Tongue
Squamous cell carcinoma 1/83 1/83 1/50* 1/50 0/46* 0/46 5/48 4/49
(P = 0.025) (P = 0.063)
*Dr. Squire examined all slides from the middle and low dose described by Dow (original report) as
exhibiting any lesions, but did not review tongue slides that Dow described as having no lesions. The
incidence numbers for low and medium dose levels in this tble represent lthis combined review incidence
(i.e., Dow*s tongue diagnoses confirmed by Dr. Squire).
S = Dr. Squire's histopathologic evaluation
K = Dr. Kociba's histopathologic evaluation
-------
The question arises whether these squamous cell carcinomas of the tongue
could have been induced by any TCDD contamination which was present below the
level of detection. Assuming TCDD was present at the level of detection (0.33
ppb), the amount of TCDD daily intake in the 2,4,5-T was estimated at less than
10 pg/kg/day. A second long-term TCDD study by Kociba (1978) on TCDD in
Sprague-Dawley rats, also showed increased squamous cell carcinoma of the tongue
in males. The results from the TCDD study are shown in Table 14.
TABLE 14. KOCIBA (1978) STUDY ON TCDD IN MALE SPRAGUE-DAWLEY RATS
pg/kg/day TCDD
Site Control 100,000 10,000 1,000
Tongue-stratified 0/7 6a 3/50 1/50 1/50
squamous cell carcinoma
Fisher's Exact Test (one-tailed) P = 0.06 N.S.& N.S.b
Test for trend exact test P = 0.01
*0nly 76 of 85 tongues were examined microscopically.
bN.S. = not significant at P = 0.05.
Two exact probability tests both show statistical significance at the P =
0.06 level. The high dose response of 3/49 tumors at 100 ng/kg/day is
significant at the P = 0.06 level, and the exact test for trend has a P-value
= 0.01. Thus, the Kociba TCDD study provides suggestive evidence of a
carcinogenic effect in the tongues of males.
A comparison of the two Kociba studies at comparable TCDD dose levels for
comparable effects can only be made approximately. At 30/mg/kg/day 2,4,5-T, the
43
-------
TCDD would have been present, if present at all, at a level below 10 pg/kg/day.
This group showed in response of 4/49 (using Kociba's diagnoses). For the TCDD
study, the comparison of 1/50 at 1,000 pg/kg/day TCDD vs. 5/49 (at 10
pg/kg/day), was not statistially significant (P = 0.36, two-tailed test). This
leads us to conclude that the carcinogenic effect on the tongue exhibited in the
2,4,5-T study was at least partly and probably totally due to the 2,4,5-T and
not to the TCDD contaminant. But in another way, the TCDD dose required for
3/50 tongue tumors was 100,000 pg/kg/day.
Based on the dose response slope of tongue tumors from the TCDD study, the
estimated number of tongue tumors from the 10 pg/kg/day TCDD contaminant in the
2,4,5-T study would have been 1 x 10-5. £Ven considering the historical
control group experience, the expected number would be less than 0.5.
Thus, based on the above analysis, if the TCDD contamination level of
2,4,5-T in the Kociba study was not greater than 0.33 ppb, then this study of
2,4,5-T provides highly suggestive evidence of the carcinogenicity of 2,4,5-T in
rats.
Leuschner et al. (Oral) Rat Study (1979)
Leuschner et al. (1979) investigated the chronic effects of commercial grade
2,4,5-T (containing 0.05 ppm TCDD) on Sprague-Dawley (SIV50) rats. The study
used four groups of 60 male and 60 female rats from the F]_ generation of a
three-generation reproduction study in which the dams received 2,4,5-T at 0, 3,
10, and 30 ma/kg body weight/day in the diet. From 6 weeks (42 days) of age,
rats in the four groups were placed on the same feeding regimen as their
mothers. The treatment continued for the duration (130 weeks) of the
experiment. Three groups received 2,4,5-T dissolved in acetone, which was
poured over a small quantity of feed and then mixed after the evaporation of the
\
acetone. One group, identified in the report as the pre-mix controls, was given
44
-------
only acetone in the diet. A fresh diet was prepared every 7 days.
Additional groups of 60 male and 60 female Sprague-Dawley rats served as
untreated controls. Rats in this group were supplied at 6 weeks of age by the
same source that had supplied the Fq generation of the three-generation study.
During the experiment, clinical signs, body weights, and consumption of food and
water, were monitored at regular intervals. Urinalyses were performed and
hematological and clinical chemistry parameters were determined for 10 rats from
each group at regular intervals. The same rats were used for measurements
throughout the experiments; the authors found no effects attributable to 2,4,5-T
in any of these observations. At 13 weeks, 10 rats were sacrificed from each
group and examined leaving 50 animals of each sex for long-term exposure. Rats
that died, were moribund, or killed during the experiment, and all surviving
rats killed after 130 weeks, were necropsied. All major tissues of all animals,
except for tissues of the survivors dosed at 3 mg/kg/day, were examined
histopathologically.
The authors reported that they found no evidence that the test compound had
a toxic or carcinogenic effect on either male or female rats. The type and
incidence of lesions observed were considered normal in old-age breeding rats of
the test strain. However, a statistically significant increase in interstitial
cell tumors of the testes in the high dose group of males (P = 0.014), as well
as a significant dose-related trend (P <0.01) for these tumors was observed
when comparison is made to the incidence of these tumors in the pre-mix control
animals (Table 15). The significance of these results disappeared when
comparison was made to the untreated control group, which had an incidence of
testicular tumors higher than that in the high dose group. The incidence of
testicular tumors in the untreated controls (22/50 or 445) is very significantly
higher (P < 0.01, using a one-tailed Fisher Exact Test) than that in the pre-mix
45
-------
controls {12%). Because of this difference, the two control groups cannot be
pooled for comparision with the treated groups. Clearly, a compound-related
increase can be shown only if the pre-mix control group alone is used as a point
of comparison with the treated groups.
Because of the experimental design, the pre-mix controls are the appropriate
comparision group. However, in this situation there is a question of whether
the pre-mix or untreated controls manifest an atypical spontaneous rate for
testicular tumors. The results are not statistically significant when the
incidence of testicular tumors in the high dose group is compared to the
incidence of testicular tumors in historical controls provided by Leuschner
(1979). In each of the four sets of historical controls [designated tests
A] - 20/50 (40%), A2 - 24/90 (26.65), A3 - 17/50 (34%), A4 - 32/100
(32%) the incidence of testicular tumors is comparable to that in the high dose
group 16/50 (32%). However, these historical controls were untreated controls
rather than vehicle (acetone) treated controls, and therefore may not be
appropriate for comparision. Moreover, in the low dosage group, while 14
testicular masses were observed macroscopically, only six of those masses were
examined microscopically.. All six masses examined microscopically were
testicular tumors. If the remaining eight masses were examined microscopically
and were also proven-to be testicular tumors, the dose-related trend would no
longer be significant.
46
-------
TABLE \5. INTERSTITIAL-CELL TUMORS T TESTES IN MALE RATS
Dose
Rats wi th
Percent animals
tumors
P-Valuea
with tumors
untreated
controls
22/50
44%
pre-mi x
controls
5/50
12%
10/mg/kg/day
group
12/50
N.S.b
24%
30 mg/kg/day
group
16/50
0.014
32%
aP - Value calculated with Fisher Exact Test (one-tailed).
&N.S. = not significant at P = 0.05.
This study suffers from the following limitations: 1) the maximum tolerated
dose was apparently not used; 2) the observed testicular tumors are often
associated with old-age with variable incidences; 3) testicular masses were
reported in 14/28 of the animals exposed at the low dose (3 mg/kg/day), but only
six of these masses were diagnosed microscopically; and 4) the difference in the
incidences of testicular tumors in the two contol groups makes interpretation of
the significance of the testicular tumor incidence in treated groups uncertain.
In conclusion, the significance of the results concerning the incidence of
testicular tumors is uncertain. In addition, this test cannot be considered a
valid negative study of 2,4,5-T because the highest dose used was less than the
maximum tolerated dose. This reduced the sensitivity of the test for detecting
the possible oncogenic effects of 2,4,5-T.
The tongue, which was a site of increase in tumor incidence in the Kociba
studies was not initially examined microscopically in the Leuschner study.
Therefore, the CAGT requested the histopathologlcal examination*of tongue lesions
47
-------
(cutting the tissues horizontally). Due to some misunderstanding, it appears
that longitudinal sectionals were made (see letter from Wade Richardson to Dr.
Charalingayya Hiremath, Appendix E).
The tongues of the male rats treated with 30 mg 2,4,5-T/kg body weight/day
in the food and those of the untreated male rats (without premix) were
investigated histopathologically after haematoxylin-eosin staining.
Longitudinal sections reaching from the retrolingual region to the tip of the
tongue were prepared. Each of 8 gradual sections of the tongue were
investigated, the mucosal epithelial thickness of the treated and untreated
rats was compared.
The histopathologic examination of the tongue did not reveal any neoplastic
lesions that were observed by Kociba in his 2,4,5-T study at the same dose level
feeding studies (Appendix D). Although the adequacy of the histopathological
examination of the tongues in the Leuschner study is not clear at the present
time, the unusual site of tumor formation in the tongue in the Kociba study and
the apparent lack of reproducabi1ity of this tongue tumor response in the
Leuschner study reduced our judgment of the strength of evidence of the
carcinogenicity of pure 2,4,5-T provided by the Kociba study from substantial to
highly suggestive.
CARCINOGENICITY OF (2,4,5-TRICHL0R0PHEN0XY)PR0PI0NIC ACID (SILVEX)
Innes et al. (Bionetics Laboratories 1968) (Oral) Mouse Study (1969)
Innes et al. (Bionetics Laboratories 1968) (1969), under the sponsorship of
the National Cancer Institute, investigated the carcinogenicity of silvex in two
studies with mice, one oral and the other subcutaneous. In the oral study,
groups of 18 B6C3F1 and 18 B6AKF1 mice of each sex were given the test substance
t
daily in 0.5« gelatin by oral gavage at 46.4 mg/kg body weight beginning at 7
48
-------
days of age and continuing until they reached 28 days of age. At that time, 121
ppm of si 1 vex was administered daily in the diet. This study was carried out
for approximately 18 months. Mice were housed by sex, up to six in a cage, and
were given food and water ad libitum. All animals were observed daily for
clinical signs and weighed weekly. The doses administered were the maximum
tolerated doses, which had been selected from pre-chronic toxicity studies
performed before the initiation of the chronic study. The moribund mice were
killed, necropsied, and selectively examined microscopically, while surviving
animals were killed at approximately 18 months and necropsied. Heart, lungs,
liver, spleen, kidneys, adrenals, stomach, intestines, genital organs, and
tissue masses were placed in formalin. They were later sectioned, stained with
hematoxylin and eosin, and examined microscopically. All but five mice, three
86C3F1 male and two B6AKF1 male or female, survived 18 months. Table 16
identifies the types of tumors and the groups in which they were found.
TABLE 16. TUMORS IN MICE EXPOSED ORALLY TO SILVEX
Type of Tumor
B6C3F1
Mice
B6AKF1
Mice
M
F
M
F
Reticulurn-cell sarcoma, type A
1
1
0
0
Pulmonary adenoma
1
0
1
0
Hepatoma
5
0
0
0
Mammary adenocarcinoma
0
1
0
0
Angioma
1
0
0
0
Gastric papilloma
0
2
0
0
Adrenal cortical adenoma
0
0
0
1
49
-------
There was no significant increase in the incidence of neoplasms in B6C3F1 or
B6AKF1 male or female mice administered si 1 vex orally. However, by current
National Cancer Institute guidelines, there are a number of deficiencies in
these studies: 1) only a single dose level was administered, 2) the number of
animals in the treatment group (18) was too small, and 3) the experiment was
terminated after only about 18 months. Because of these deficiencies, the test
was relatively insensitive for detecting a possible oncogenic effect of silvex
and therefore cannot be considered as significant evidence of the
non-carci-nogenici ty of silvex.
Innes et al. (Bionetics Laboratories 1968) (Subcutaneous) Mouse Study (1969)
In this study, groups of 18 B6C3F1 and 18 B6AKF1 mice of each sex,
approximately 28 days of age, were given a single subcutaneous injection of 215
mg/kg body weight of 2-(2,4,5-trichlorophenoxy)propionic acid (silvex, supplied
by Methson Coleman Bell Co.) suspended in dimethyl sulfoxide (DMSO), and
observed for 18 months. Procedures similar to those described above for the
Bionetics oral study were followed for animal housing, care, observation,
necropsies, selection of tissues, and preparation of histologic slides. All
animals, except two B6C3F1 male mice, survived 18 months. Seven tumors were
diagnosed in the 18 male B6C3F1 mice examined. The tumors were: two
reticulurn-cell sarcomas, type A; two pulmonary adenomas; one hepatoma; and two
hemangiomas. Only one tumor, a gastric papilloma, was diagnosed in one of the
18 female B6C3F1 mice examined. A hepatoma was found in one of the male B6AKF1
mice examined. These incidences were comparable to those seen in the control
animals. There was no significant increase in the incidence of neoplasms in
B6C3F1 or B6AKF1 male or female mice by subcutaneous injection: However, there
50
-------
were a number of deficiencies in this study: 1) only one subcutaneous injection
was given, 2) the number of animals in the treatment group (18) was too small,
and 3) the experiment was terminated after only 18 months. Because of these
deficiencies, the test was relatively insensitive for detecting an oncogenic
effect of silvex.
Dow Chemical Company (Oral) Rat Study, summarized in Mullison (1966) and Gehring
and Betso (1978)
Groups of Wister rats (30 males and 30 females in each group) were fed diets
containing 0.0, 0.03, 0.003, and 0.001% Kurosol®SL (potassium salt of silvex)
for up to 24 months. Administration of the test compound began at 50 days of
age. Animals were sacrificed at 12 and 18 months so that the group sizes at the
end of the 2-year study could not have been more than 21 or 22 per sex; they may
have been even smaller. However, the size of the groups at the end of the study
cannot be exactly determined since no data were provided on the extent to which
animals, other than the ones sacrificed, died before the end of the study.
There was no evidence of a toxic effect or reduced survival in female rats
administered any dose compared to controls. Therefore, it does not appear that
the females were administered the maximum tolerated dose. Since high dose males
exhibited a significant decrease in average body weights, it appears that they
were administered a maximum tolerated dose.
No significant increase in tumors was reported. However, because small
groups of animals were used and the maximum tolerated dose was apparently not
used in the high dose females, this study cannot be considered as significant
evidence of the non-carcinogenicity of silvex in rats.
51
-------
Dow Chemical Company (Oral) Dog Study (1965), summarized in Mullison (1966)
and Gehring and Betso (1978)
Groups of beagle dogs (four males and four females in each group) were fed
diets containing 0.0, 0.056, 0.019, and 0.0056%. Kurosal®SL (potassium salt
of silvex). One male and one female of each group were killed at 12 months.
The remaining animals were killed at the end of the 2-year period.
No incidence of tumors was observed. However, this study does not
constitute a valid cancer study because its duration was far less than the life
expectancy of a beagle dog and the sizes of the animal groups were exceedingly
small. Therefore, this study provides no evidence of the lack of
carcinogenicity of silvex in dogs.
52
-------
CARCINOGENICITY OF TCOD IN RATS AND MICE
Kociba et al. (Oral) Rat Study (1977, 1978)
Although this study was reported in published form in Toxicology and Applied
Pharmacology (1978), a fuller version was submitted in an unpublished report
(Kociba et al., Dow Chemical Company, September 28, 1977).
In this study, groups of 50 Sprague-Dawley rats (Spartan substrain) of each
sex were maintained for up to 2 years on diets providing 0.1, 0.01, or 0.001
ug/kg/day TCDD. Vehicle control groups comprised 86 animals of each sex. The
test was appropriately conducted with the high dose group at a level which
induced signs of tissue toxicity, reduced weight increments in both sexes, and
shortened lifespans in female rats. Clinical tests performed at intervals
during the study monitored organ specific toxicity, particularly of the liver.
Pathologic examinations included histopathologic evaluation of all major tissues
in both the high dose and control animals, but only of selected tissues
identified as possible target organs and suspect tumors in lower dose groups.
This approach is suitable for the identification of a carcinogenic effect, but
does not determine actual tumor incidences in all groups except in those organs
identified as target organs. It, therefore, is adequate to define dose-response
relationships only in these target organs. Tissues examined from most animals
in all dose groups included liver, lungs, kidneys, urinary bladdar, tongue,
brain, testes/ovaries, and prostate/uterus. For these tissues, a quantitative
analysis can be performed using the actual number of tissues examined
histopathologically for animals at risk. For other tissues (excluding skin,
mammary glands, and nasal turbinates/hard palate), actual tumor incidence cannot
be evaluated for the two lower doses. For skin and mammary glands, the number
of animals necropsied is the appropriate denominator to determine incidence,
because detection of these tumors is based on observation of the tumor at
53
-------
necropsy.
A laboratory audit of this study by Spencer and Woodrow, Hazard Evaluation
Division, Office of Pesticide Programs, U.S. Environmental Protection Agency did
not reveal significant new information. Reviewers concluded that the study was
properly conducted, adhering to the accepted procedures (memorandum from Spencer
and Woodrow to Diana Reisa, and Warnick Project Manager for 2,4,5-T).
Based on data reported for food consumption, body weight, and dietary level
of TCDD, the daily doses were reasonably constant for most of the study,
although somewhat below the value expected in most groups during the third
nonth.
High early mortality was observed in all groups in this study but was only
statistically significant in the high dose group. The survival curves show
progressive mortality beginning as early as the 12th month and leading to 50"
mortality by 21 months.* The effects of this early mortality are a reduction
in expected rumor incidence because of a truncated latency period, and a
reduction in sensitivity of the study because of a reduction in number of
animals at risk during the time of expected tumor manifestation. Cumulative
mortality and interva1 mortality rates are given in Tables 111-7 to 10 of
Appendix A (Clement Associates 1979).
The results of this study provide substantial evidence that TCDD is
carcinogenic in rats. TCDD induced a highly statistically significant increase
of both hepatocellular carcinomas and hepatocellular neoplastic nodules in
*In the 0.001 group of males, 44% mortality was as early as 18 months.
The mortality patterns were analyzed by the Whitney-Wilcoxon test and
Kolmogorov-Simonov test. These tests show that mortality was significantly
higher in the high dose females than in controls, and while indications of
increased mortality were found in other groups, they were not part of a
consistent pattern.
54
-------
female rats at doses of 0.1 nd 0.01 ug/kg/day (2200 and 2 0 ppt in the diet,
respectively). The increase of hepatocellular carcinomas alone, in the high
dose females, was also highly significant. In addition, at the highest dose
level, TCDD induced a statistically significant increase in stratified squamous
cell carcinomas of the hard palate and/or nasal turbinates in both males and
females, squamous cell carcinomas of the tongue in males, and keratinizing
squamous cell carcinomas of the-lungs (highly significant) in females (tumor
incidences reported in Tables 17, 18, and 19).
TABLE 17. HEPATOCELLULAR CARCINOMAS AND HEPATOCELLULAR
HYPERPLASTIC NODULES IN FEMALE SPRAGUE-OAWLEY RATS MAINTAINED ON
DIETS CONTAINING TCDD
Dose level
ug/kg/day
Rats with
hepatocellular
hyperplastic
nodules
Rats with
hepatocellul ar
carci nomasa
Total number
of rats wi th
both types
of tumors3
0
8/86 (9%)
1/86 (1%)
9/86 (10%)
0.001
(22 ppt)
3/50 (6%)
0/50 (0%)
3/50 (6%)
0.01
(210 ppt)
18/50 (36%)
2/50 (4%)
18/50 (36%)b
(P = 4.37 x 10"4)
. 0.1
(2200 ppt)
23/48 (48%)
11/48 (23%)
(P- = 5.6 x 10-5)
34/48 (71%)
(P = 9.53 x 10-13)
aP-values calculated using the Fisher Exact Test (one-tailed).
bTwo rats had both hepatocellular carcinomas and hyperplastic nodules.
55
-------
TABLE 18. TUMOR INCIDENCE IN FEMALE RATS FED DIETS CONTAINING TCDD
Dose level
Stratified squamous cell
Keratinizing squamous
ug/kg/day
carcinomas of hard palate
cell carcinomas of
or nasal turbinates
lungs
0
1/54 (2%)
0/86 (0%)
0.001
0/30 (0%)
0/50 (0%)
(22 ppt)
0.01
1/27 (4%)
0/49 (0%)
(210 ppt)
0.1
5/24 (21%)
7/49 {14%)
(2200 ppt)
(P = 0.01)a
(P = 0.0006)2
dP-values calculated
using the Fisher Exact Test (one-tailed).
TABLE 19. TUMOR
INCIDENCE IN MALE RATS FED
DIETS CONTAINING TCDD
Stratified squamous
Hard palate/nasal turbinates
Dose level
cell carcinomas of
stratified squamous cell
ug/kg/day
the tongue
carci noma3
0
0/76 (0%)
0/51 (0%)
0.001
1/49 (2%) N.S.b
1/34 (2%) N.S.b
(22 ppt)
0.01
1/50 (2%) N.S.b
0/27 (0%) N.S.b
(210 ppt)
0.1
3/42 (P = 4.3 x 10-
¦2) 4/30 (13%)(P = 0.016)
(2200 ppt)
dInclude examinations from both original and updated report (2/20/79).
&N.S. = not significant at P = 0.05.
56
-------
Dr. Robert Squire, pathologist at the Johns Hopkins University Medical
School and consultant to the CAG, evluated the histopathological slides from Dow
Chemical Company's 2-year rat feeding studies on
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) by Kociba et al. Dr. Squire and his
associates examined all livers, tongues, hard palates, and nasal turbinates, and
lungs available from TCDD study. His histopathological findings, as well as Dr.
Kociba's histopathological evaluations, are summarized in Tables 20 and 21 and
Appendix B. Although there are some differences between the diagnoses of Kociba
and Squire, the conclusions about the target organ for cancer induction, and the
dose levels at which induction occurred are the same whether Squire's or
Kociba's diagnoses are considered.
57
-------
TABLE 20. nitS. SQUIRE'S AND KOCIBA'S REVIEW OF DOW TCDD ORAL RAT STUDY (8/15/00)
Sprague-Dawley Rats - Spartan Substrain (2 yrs.)
FEMALES
Dose Levels (ug/kg/day)
Tissues and Diagnoses
0 0.001 0.01 0.1
(control)
Lung
Squamous cell
carcinoma
S
0/86
K
0/86
S
0/50
K
0/50
S
0/49
K
0/49
S
8/47
(P = 1.61
K
7/49
x 10-4) (P = i.21 x 10-4)
Nasal turbinate/hard palate
squamous cell
carcinoma 0/54
-------
TABLE 21. DRS. SQUIRE'S AND KOCIBA'S REVIEW OF DOW TCDD ORAL RAT STUDY (8/15/80)
Sprague-Dawley Rats - Spartan Substrain (2 yrs.)
MALES
Tissues and Diagnoses
0
(control)
0.001
Dose Levels
0.01
(ug/kg/day)
0.1
Nasal Turblnates/Hard
Palate squamous cell
carcinomas
S K
0/55 0/51
S K
1/34 1/34
S
0/26
K
0/27
(P =
S
6/30
1.36 x
10"3)
K
4/30?
Tongue
Squamous cell
carcinomas
0/77
2/44
1/49
(P =
3/44
4.60 x
10-2)
3/42
(P = 4.34 x 10-4)
Total - 1 or 2 above 0/77 2/44 1/49 9/44
(each rat had at (P = 6.28 x 10~5)
least one tumor above)
S = Dr. Squire's histopathologic analysis
K = Dr. Koclba's histopathologic analysis
-------
National Cancer Institute (Oral) >t Study (1980a, b)
A cancer bioassay for the possible carcinogenicity of 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD) was tested by the Illinois Institute of Technology in
rats and mice under a contract sponsored by the National Cancer Institute (MCI).
In the rat study, 50 Osborne-Mendel rats of each sex were administered TCDD*
suspended in vehicle of 9:1 corn oil-acetone by gavage 2 days per week for 104
weeks at doses of 0.01, 0.05, or 0.5 mg/kg/wk. Seventy-five rats of each sex
served as vehicle controls. One untreated control group containing 25 rats of
each sex was present in the TCDD treatment room and one untreated control group
containing 25 rats of each sex was present in the vehicle control room. All
surviving rats were killed at 105 to 107 weeks.
In rats, a dose-related depression in mean body weight gain became evident
in the males after week 55 of the bioassay and in the females after week 45.
The results of histopathologic diagnosis of primary tumors caused by the
oral administration of TCDD are presented in Table 22. In male rats an
increased incidence of foilicular-cell adenomas or carcinomas of the thyroid
were dose-related and were statistically significantly higher in the low, mid,
and high dose groups than in the vehicle controls. In addition, a statistically
significant increase in subcutaneous tissue fibromas was found in males of the
high dose group.
"Purity of TCDD was found to be 99.4*; two impurities tentatively identified
as a trichlorodibenzo-p-dioxin and a pentachlorodibenzo-p-dioxin presence of 0.1
to 0.2% hexachlorodibenzo-p-dioxin was detected by gas chromatography and mass
spectometry.
60
-------
TABLE 22. INCIDENCE OF PRIMARY TUMORS IN MALE RATS
ADMINISTERED TCDD BY GAVAGE
ug/kg/week
Type of tumor
Vehicle
Low Dose3
Mid Dosea
High Dose3
control
0.01
0.05
0.5
Subcutaneous tissue
Fibrosarcoma
3/75 (4%)
1/50 (2%)
3/50 (6%)
7/50 (14%)
P = 0.048
Liver
Neoplastic nodule
or hepatocellular
carci noma
0/74 (0%)
0/50 (0%)
0/50 (0%)
3/50 (6%)
Adrenal ¦
Cortical adenoma
6/72 (8%)
9/50 (18%)
12/49 (24%)
9/49 (18%)
Thyroi d
Follicular cell
adenoma
1/69 (1%)
5/48 (10%)
6/50 (16%)
10/50 (20%)
P = 0.042
P = 0.021
P = 0.001
Thyroi d
Follicular cell
adenoma or carcinoma
1/69 (25)
5/48 (10%)
8/50 (16%)
11/50 (22%)
P = 0.042
P = 0.004
P < 0.001
dP-values calculated using the Fisher Exact Test.
In female rats, a statistically significant increase of each of the
following tumors was found in the high dose group: hepatocellular carcinomas
and neoplastic nodules (P = 0.001), subcutaneous tissue fibrosarcomas (P =
0.023), and adrenal cortical adenomas (P = 0.039) as shown in Table 23.
These results confirm the carcinogenic effect observed in the Kociba et al.
(1978) study using Sprague-Dawley (Spartan substrain) rats.
61
-------
TABLE 23. INCIDENCE OF PRIMARY TUMORS IN FEMALE RATS
ADMINISTERED TCDD BY GAVAGE
ug/kg/week
Type of tumor
Vehicle
control
Low dose3
0.01
Mid dose
0.05
High dose3
0.5
Subcutaneous tissue
Fibrosarcoma
0/75 (0%) 2/50 (4%) 3/50 (6%)
4/49 (8%)
P = 0.023
Liver
Neoplastic nodule
5/75 (7%) 1/49 (2%) 3/50 (65)
12/49 (24%)
P = 0.006
Liver
Neoplastic nodule
or hepatocellular
carci noma
Pitui tary
Adenoma
5/75 (7%) 1/49 (2%) 3/50 (6%)
1/66 (25) 5/47 (11%) 2/44 (5%)
P = 0.044
14/49 (29%)
P = 0.001
3/43 (7%)
Adrenal
Cortical adenoma
11/73 (15%) 8/49 (16%) 4/49 (8%)
14/46 (30%)
P = 0.039
aP-values calculated using the Fisher Exact Test.
62
-------
Van Miller et al. (Oral) Rat Study (1977)
Male Sprague-Dawley rats weighing approximately 60 grams each were used.
There were 2 rats in each cage and 10 rats in each group. Rats ingested ground
chow for only 2 weeks. They were then given 2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD) in the following concentrations: 0, 1, 5, 50, 500 parts per trillion
(ppt, 10~12 gram TCDD/gram food); and 1, 5, 50, 500, and 1000 parts per
billion (ppb, 10" 9 gram TCDD/gram food). Rats ingested the diets with TCDD
for 78 weeks, and thereafter were kept on a control diet. Laparotomies were
performed on all surviving rats at the 65th week and biopsies were taken from
all tumors observed. Surviving rats were killed at 95 weeks.
Food intake was significantly lower in rats ingesting 50, 500, or 1000 ppb
TCDD than in the controls, and they lost weight. All of the rats in the dose
groups died between the second and fourth weeks of treatment. The food intake
for rats receiving the other dose levels was similar to that of the controls.
Weight gain was significantly less for rats given 5 ppb TCDO. TCDD intake and
mortality of rats are shown in Table 24.
TABLE 24. TCDD INTAKE AND MORTALITY IN RATS
Dosea
Weekly dose per rat
(ug/kg body weight)
Week of
first death
Number of rats
dead at 95th week
0 ppt
68
6/10 (60%)
1 PPt
0.0003
86
2/10 (20%)
5 ppt
0.001
33
4/10 (40%)
50 ppt
0.01
69
4/10 (40%)
500 ppt
0.1
17
5/10 (50%)
1 ppb
0.4
31
10/10 (100%)
5 ppb
2.0
31
10/10 (100%)
aRats at 50, 500, and 1000 ppb dose levels were all dead within four
weeks.
63
-------
Complete necropsies were done and samples of tissues were taken for
microscopic examination from the control groups and each treatment group
(Laboratory audit* and personal communication with author). Special staining
methods were used as an aid in the diagnosis of neoplasms. Various benign and
malignant tumors were found in each treatment group. No tumors were observed in
the controls (Table 25).
Statistically significant increases of squamous cell tumors of the lungs and
neoplastic nodules of the liver were observed in rats ingesting 5 ppb TCDD
(Table 26). In addition, two animals in the 5 ppb dose group and one animal in
the 1 ppb dose group had liver cholangiocarcinomas, which are rare in
Sprague-Oawley rats. These tumors were not found in any of the lower dose
groups or in the controls. These results provide evidence of a carcinogenic
effect.
The observation of no tumors of any kind in the controls is unusual for
Sprague-Dawley rats. In addition, the reporting of the study was not extensive.
These factors may tend to lessen the reliance which can be placed on the
positive results of this study. However, this study does provide independent
confirmation of the findings of the Kociba study that TCDD causes observable
carcinogenic effects in rats a\. very low doses.
AThe audit of this study brought out the fact that it was intended only to
be a rangefinding study. Therefore, only small numbers of animals were used.
This may have made the study relatively insensitive for detecting carcinogenic
effects at doses lower than 1 ppb.
64
-------
TABLE 25. BENIGN AND MALIGNANT TUMORS IN RATS INGESTING TCDD
Number of
Number of rats
Dosea
Benign
Maiignant
tumors
with tumors
0
0
0
0
0/10 (0%)b
1 ppt
0
0
0
0/10 (0%)
5 ppt
1
5
6C
5/10 (5
-------
TABLE 26. LIVER TUMORS IN RATS INGESTING TCDD
Neoplastic
Choiangi o-
Squamous cell
Dose(ppb)
nodules
carcinomas
tumors of the lungs
0
O
M
o
o
0/10 (0%)
0/10
1
0/10 (0%)
1/10 (10%)
0/10
5
4/10 (20%)
2/10 (20%)a
4/10 (40%)
P = 0.043
P = 0.043
dThe two animals had both neoplastic nodules of the liver and
cholangi ocarci nomas.
66
-------
Toth et al. (Oral) Mouse Study (1979)
This study investigated the carcinogenicity of TCDD in Swiss mice.
Ten-week-old outbred Swiss /H/Riop mice were used. TCDD was administered in a
sunflower oil vehicle by gavage to groups of 45 male mice once a week at doses
of 7.0, 0.7, and 0.007 ug/kg body weight for a year (groups 9, 10, and 11,
respectively, in Table 27). Matched male vehicle controls were administered
sunflower oil once a week. Matched controls to a companion study investigating
the carcinogenicity of (2,4,5-trichlorophenoxy)ethanol (TCPE) contaminated with
low levels of TCDD, were administered carboxymethyl cellulose (the vehicle used
in that study) once a week. Two untreated controls were also maintained.
This study appears to be generally well-conducted. However, the
administration of TCDD over a period of only one year, which is far short of the
life expectancy of the mice used, made the study relatively insensitive.
Animals were followed for their entire lifetimes. Autopsies were performed
after spontaneous death or when the mice were moribund, and all organs were
examined histologically. Sections were stained with hematoxylin and eosin for
light microscopy. Pathological findings were evaluated and analyzed
statistically. The findings-of the TCDD study and the comparison study on TCPE
are given in Table 27 (reproduced from the journal in which this study is
reported).
Analysis of the results of this study focused on the incidence of liver
tumors in the groups treated with TCDD and the incidence of these tumors in the
matched controls (group 12) and in the males in the three other control groups.
Males in groups 3 and 8, the two untreated control groups, had 26% and 33% liver
tumors, respectively (P > 0.20). The carboxymethyl cellulose male controls
(group 7) had 33% (32/96) liver tumors. No significant differences in liver
tumors were observed when males in all four control groups were compared to each
67
-------
TABLE 27. CUMULATIVE DATA ON TUMOUR INCIDENCE
(Taken from Toth et al. 1979)
Group
Treatment
No. of
Animals with tumours of
Effective
tumour
other
TCPE TCDD
Vehicle3
no. of
bearing
1 i ver
lung
lymphomas
organs
Average
(mg/kg)
(ug/kg)
(mg/kg)
Sex
mice
mice
no.
(%).
no.
no.
no.
1i fespan
1
67.0
0.112
50
M
88
69
42b(18)
50
7
16
595
(1.6 ppm)
F
83
61
7
(8)
52
15
25
652
2
70.0
0.007
! 50
98
78
57c(58)
18
11
16
571
(0.1 ppm)
F
96
59
9
(9)
39
15
23
582
3
control
93
63
24
(26)
44
8
17
577
F
84
57
4
(5)
41
23
13
639
4
7.0
0.07
50
93
79
25
(27)
38
18
22
641
(10 ppm)
F
96
60
10
(10)
38
19
19
589
5
7.0
0.0007
50
94
77
23
(24)
50
23
17
660
(0.1 ppm)
F
93
71
8
(9)
42
36
21
500
6
0.7
0.00007
50
97
78
24
(25)
51
20
17
vj J
(0.1 ppm)
F
94
64
5
(5)
38
22
21
566
7
_ _
_ —
50
96
74
32
(33)
44
14
22
615
F
84
55
4
(5)
38
18
17
565
8
control
96
78
32
(33)
38
22
15
651
F
91
57
4
(4)
31
24
19
549
9
7.0
10
M
43
27
13
(30)
11
6
7
424
10
0.7
10
M
44
36
21
(48)
18
12
4
633
11
0.007
10
M
44
39
13
(29)
27
10
6
649
12
- -
10
M
38
27
7
(18)
15
6
7
588
aCarboxymethyl cellulose in groups 1-8, sunflower oil in groups 9-12.
bP < \%
cp < 0.1%
-------
other (P > 0.05). Nevertheless, there was evidence that the incidence of liver
tumors in the control groups was associated with the average lifespan in the
respective groups. The two groups that had less than 600 days average survival
(groups 3 and 12) had the fewest liver tumors (26% and 18%, respectively). On
the other hand, the two groups that had an average survival of greater than 600
days (groups 7 and 8), had 33? liver tumors each. The test for linear trend
(tumors vs. days of average survival) was not quite significant (P = 0.065).
Among the three treatment groups (groups 9, 10, and 11), the middle dose
(0.7 ug/kg) showed the highest incidence of liver tumors (21/44 = 48%). This
incidence was significantly higher than the incidence of liver tumors in either
the sunflower oil controls (P < 0.01) or the pooled controls (all four control
groups combined) (P < 0.025).
The highest dose group (7.0 ug/kg) had an increased incidence of liver
tumors compared to the matched sunflower oil controls (13/43 = 30%) but this
increase was not statistically significant (P = 0.11). The incidence of liver
tumors in the high dose group was comparable to that of the pooled controls.
The highest dose group, however, had a much reduced average survival in
comparison to any of the control groups (only 424 days compared to 577, 588,
615, and 651 days in the four control groups). This poor survival may have
accounted for the lack of a statistically significant increase in liver tumors
in the high dose group. Furthermore, if time-to-tumor data had been available,
it is highly likely that the high dose group would have shown a significant
decrease in time-to-tumor compared to the controls. Therefore, the increase in
liver tumors that was observed in the high dose group in comparison to the
matched control group, although not statistically significant, is considered to
be consistent with an oncogenic effect.
69
-------
In conclusion, the resul ts of this study provide suggestive evidence of an
oncogenic effect.
National Cancer Institute (Oral) Mouse Study (1980a, b)
A cancer bioassay for the possible carcinogenicity of
2,3,7,8-tetrachlorodibenzo-p-dioxin was tested by the Illinois Institute of
Technology in mice under a contract sponsored by the National Cancer Institute
(NCI).
In the mouse study, 50 B6C3F1 mice of each sex were administered TCDD
suspended in a vehicle of 9:1 corn oil-acetone 2 days per week for 104 weeks at
doses of 0.01, 0.05, and 0.5 ug/kg/wk for male mice and 0.04, 0.2, and 2.0
ug/kg/wk for female mice. Seventy-five mice of each sex were used as vehicle
controls. One untreated control group of 25 mice of each sex was present in the
TCDD treatment room. One untreated control group of 25 mice of each sex was
present in the vehicle control room. In mice, the mean body weight gain in the
treated groups was comparable with that of the vehicle control groups. However,
the mean body weight of the treated mice was lower when it was compared with
untreated controls.
The results of the histopathologic diagnosis of primary tumors are presented
in Table 28. The results indicate that, in male mice, TCDD induced a
statistically significant incidence of hepatocellular carcinomas (P = 0.002) and
both hepatocellular carcinomas and neoplastic nodules combined (P = < 0.001) in
male mice of the high dose group.
In female mice, TCDD induced statistically significant increases of
hepatocellular carcinomas (P = 0.014) and both hepatocellular adenomas and
carcinomas (P = 0.002) in the high dose group. In addition, a statistically
»
significant increase in tumor incidences of fibrosarcoma, histiocytic lymphoma,
70
-------
thyroid follicular-cell adenoma, and cortical adenoma or carcinoma were also
observed in the high dose group (Table 29).
The incidence of liver tumors observed in this study confirms the earlier
observation of an increase in liver tumors in the male mouse study performed by
Toth et al. (1979).
TABLE 28. INCIDENCE OF PRIMARY TUMORS IN MALE MICE
ADMINISTERED TCDD BY GAYAGE '
ug/kg/week
Type of tumor
Vehicle
control
Low dose
0.01
Mid dose
0.05
High dose3
0.5
Liver
Hepatocellular
adenoma
7/73 (10%) 3/49 (6%) 5/49 (10%) 10/50 (20%)
Liver
Hepatocellular
carci nomas
8/73 (-115) 9/49 (18%) 8/49 (16%) 17/50 (34%)
P = 0.002
Liver
Hepatocellular
adenoma and
carci nomas
15/73 (21%) 12/49 (24%) 13/49 (27%) 27/50 (54%)
P < 0.001
dP-values calculated using the Fisher Exact Test.
71
-------
TABLE 29. INCIDENCE OF PRIMARY TUMORS IN FEMALE MICE
ADMINISTERED TCDD 3Y GAVAGE
ug/kg/week
Type of tumor
Vehicle
control
Low dose
0.04
Mid dose
0.2
High dose3
2.0
Subcutaneous tissue
Fibrosarcoma 1/74 (1%) 1/50 (2%) 1/48 (2%)
5/47 (11%)
P = 0.032
Hematopoietic
system
Histiocytic lymphoma 9/74 [11%)
4/50 (8%)
4/48 (17%)
14/47 (30")
P = 0.016
Hematopoietic
system
Lymphoma or leukemia 18/74 (24%)
11/50 (22%) 13/48 (27%)
20/47 (43%)
P = 0.029
Hematopoietic
system
All lymphoma
Liver
Hepatocellular
carci nomas
Liver
Hepatocel1ular
adenomas or
carcinomas
Thyroi d
Foilicular-cell
adenoma
18/74 (24%) 12/50 (24%) 13/48 (27%)
1/73 (1%)
3/73 (4%)
0/69 (0%)
2/50 (4%)
6/50 (12%) 6/48 (13%)
3/50 (6%) 1/47 (2%)
20/47 (43%)
P = 0.029
2/48 (4%) 6/47 (13%)
11/47 (23%)
P = 0.002
5/46 (11%)
P = 0.009
aP-values calculated using the Fisher Exact Test.
72
-------
Other Related Studies
Pi tot et al. Prr otion Study in Rats (1980) —
Pi tot et al. (1980) investigated the hypothesis that development of
hepatocellular carcinomas of the liver with chronic administration of TCDD was
the result of the promoting activity of TCDD on cells already initiated by
dietary or other environmental carcinogens. The manuscript of this study has
been submitted to Cancer Research for publication.
In this study, a two-stage model of hepatocarcinogenesis was used.
Twenty-four hours after a partial hepatectomy (to cause cell proliferation),
female Sprague-Dawley rats were divided into seven groups (Table 30). The
animals in groups 1, 5, 6, and 7 received diethylnitrosamine (DEN). The rats in
group 1 were then maintained on a standard laboratory diet for 32 weeks. The
rats in groups 2 and 3 received no DEN, but starting one week after hepatectomy
received biweekly subcutaneous injections of 0.14 or 1.4 ug/kg of TCDD in corn
oil for a period of 28 weeks (TCDD was 98.6« pure and provided by Dow Chemical
Co.). Groups 5 and 6 received DEN, and one week later were initiated on a
regimen of 14 biweekly injections of 0.14 and 1.4 ug/kg of TCDD. The animals in
group 4 received 0.05X sodium phenobarbital in the diet starting one week after
partial hepatectomy for 28 weeks, and the animals in group 5 received DEN and
one week later were also administered 0.05» sodium phenobarbital in the diet for
the duration of the experiment. At the end of the experiment, rats were killed
and sections of the liver were removed and frozen on solid CO2. Serial
sections of the frozen blocks of liver were cut and stained consecutively for
glucose-6-phosphatase (G6Pase), canalicular ATPase,Y-glutamyl transpeptidase
GGTase) with haematoxylin and eosin. The number of enzyme-altered foci were
determined from photographs of histochemically stained sections.
Hepatocarcinomas were diagnosed by standard histopathological criteria.
73
-------
TABLE 30. PROMOTING EFFECT OF 2,3,7,8-TETRACIIL0R0DIGENZ0-P-DI0XIN (TCDD)
ON HEPATOCARCINOGENESIS BY A SINGLE DOSE OF
DIETHYLNITROSAMINE (DEN) AND PARTIAL HEPATECTOMY (Pll)a
Group.
Number
Treatment
Number of enzyme-
altered foci per
cm-* of 1 iver
Percent liver
volume which is
enzyme-altered
foci
Number of
rats wi th
carcinoma
1
PH
+ DEN
(4)
346
+
65
5.0
0
2
PH
+ TCDD (low dose)
(5)
46
+
15
0.1
0
3
PH
+ TCDD (high dose)
(5)
76
+
20
0.1
0
4
PH
+ Phenobarbital
(6)
138
+
40
0.1
0
5
PH
+ DEN + TCDD
(low dose)
(5)
1582
+
300
7.8
OC
6
PH
+ DEN + TCDD
(high dose)
(7)
1280
+
40
35.0
5/7d (P = 0.0075)
7
PH
+ DEN + Phenobarbital
(4)
1510
+
185
5.0
2
subcutaneously) or phenobarbital (0.05% in the diet) administration was begun and continued for 28 weeks at
which time the animals were sacrificed and the livers examined. The low and high doses of TCDD were 0.14
and 1.4 ug/kg/2 weeks, respectively, administered subcutaneously. DEN was given at a dose of 10 mg/kg. See
text for further details.
^The numbers in parentheses denote the number of animals used in each group.
cThree rats showed "neoplastic nodules."
^One rat showed a "neoplastic nodule."
eP-value calculated using the Fisher Exact Test.
-------
The results presented in Table 30 showed that the number of foci with single
enzyme changes, the number of foci with multiple enzyme changes, and the total
liver volume affected, substantially increased with the administration of TCDD.
No carcinomas were detected in four rats treated with DEN only, but five of
seven rats treated biweekly with TCDD at 1.4 ug/kg in addition to DEN had
hepatocellular carcinomas, and six of seven rats had hepatocellular carcinomas
or hepatocellular neoplastic nodules with a statistical significance
(P = 0.0075). Three of five rats treated biweekly with TCDD at 0.14 ug/kg in
addition to DEN had hepatocellular neoplastic nodules (P = 0.083). Rats
receiving only TCDD after partial hepatectomy showed no significant increase in
enzyme-altered foci and no neoplasia.
The results of this study provide evidence that TCDD acts as a potent
promoter in this two-stage model of hepatocarcinogenesis, causing increased
neoplasia and increases in enzyme-altered foci at exceedingly low levels.
National Cancer Institute Skin Painting Study in Mice (1980b) --
This cancer bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) for
possible carcinogenicity was tested by the Illinois Institute of Technology
under a contract sponsored by the National Cancer Institute (NCI) in
Swiss-Webster mice. In this study, groups of 30 male and female Swiss-Webster
mice were used. TCDD in acetone suspension was applied to skin of mice 3 days
per week for 104 weeks. Male mice received 0.001 ug TCDD per application while
the female mice received 0.005 ug TCDD per application.
In another experiment, the same number of animals were pretreated with one
application of 50 ug 7,12-dimethylbenz(a)anthracene (DMBA*) in 0.1 ml acetone
*DMBA obtained from K and K Laboratories (Cleveland, Ohio). Its purity was
not evaluated by NCI but stated by the manufacturer to be at least 95*.
75
-------
one week before TCDD application was litiated. Forty-five mice of ^ach sex
received 0.1 ml acetone three times p - week and 30 animals of each ex were
used as untreated controls; no DMBA control was used.
Mean body weights of male and female groups of mice treated with TCDD or
TCDD following a single application of DMBA were not affected as compared to the
vehicle controls. Mean body weights of treated and vehicle control groups of
females were lower than those of untreated controls. Mean body weights of males
were less than mean body weights of untreated controls.
The results of histopathologic diagnosis are shown in Table 31. The results
show that TCDD induced statistically significant (P < 0.05) increases of
fibrosarcoma in the integumentary systems of female mice treated with TCDD alone
and TCDD following a single initial application of DMBA.
TABLE 31. INCIDENCE OF PRIMARY TUMORS IN MICE ADMINISTERED TCDD
OR TCDD FOLLOWING DMBA 3Y DERMAL APPLICATION
Type of tumors
Vehicle
Control
Dose levels
TCDDa DMBA (50) ug)
pi us TCDDa
Integumentary
system
MALE
0.001 ug x 3/wk 0.001 ug x 3/wk
Fibrosarcoma
3/42 (7%)
6/28 (21%)
P = 0.08
6/30 (20%)
P = 0.10
FEMALE
0.005 ug x 3/wk 0.005 ug x 3/wk
Fibrosarcoma 2/41 (5%) 8/27 (30%) 8/29 (28%)
P = 0.007 P = 0.010
dP-value calculated using the Fisher Exact Test.
76
-------
Berry et al. Skin Painting Study in Mice (1978, 1979) --
Berry et al. (1978) applied TCDD in acetone solution at 0.1 ug/mouse twice
weekly for 30 weeks to the skin of 30 female Charles River CD-I mice after
initiation with a single dermal application of the known skin carcinogen DMBA in
acetone. After 30 weeks of promotion with TCDD, no papillomas were observed on
the DMBA-initiated mice. In the positive controls, DMBA-initiated mice were
treated with 12-0-tetradecanoylphorbol-13-acetate (TPA) for 30 weeks; 92% of
these mice developed tumors.
Berry et al. (1979) also studied the effects of treatment with TCDD and
7,12-dimethylbenz(a)anthracene (DMBA) in a two-stage tumorigenesis bioassay in
mouse skin. In this study, tumors on the shaved skin of female CD-I mice were
initiated by topical application of DMBA and were promoted with TPA.
Pretreatment with TCDD markedly inhibited the initiation of tumors by DMBA. The
effects were greatest when TCDD was applied 3 to 5 days before initiation and
were negligible when it was applied only 5 minutes before initiation. The
inhibition was almost complete (94 to 96#) when a single dose of 1 ug of
TCDD/nouse was applied, but was only slightly less effective (89%) when the dose
was reduced to 10 ug/mouse. The time course of the inhibitory effects was
closely parallel to the time course of induction of arylhydrocarbon hydroxylase
in the skin of the mice. It was also associated with substantial reduction in
the covalent binding of the DMBA metabolite to DNA and RNA, but with no change
in their binding to protein.
The same authors also reported inhibitory effects of TCDD on the initiation
of mouse skin tumors by benz(a)pyrene (BAP), although the effect was not as
large (maximum 65«) with BAP as with DMBA.
77
-------
Cohen Skin Painting Study in Mice (1979) --
Cohen et al. (1979) showed that pretreatment of mice with dermally applied
TCDD resulted in the inhibition of skin tumor induction by subsequent treatment
with DMBA and BAP. The inhibition of skin carcinogenesis by BAP in mice after
pretreatment with TCDD was associated with an increase in covalent binding of
BAP metabolites to DNA, RNA, and protein (in contrast to the results with DMBA,
which showed a reduction in binding to DNA and RNA). However, the BAP
metabolites that were bound to DNA and RNA in mice pretreated with TCDD differed
from those in untreated mice. In particular, pretreatment with TCDD markedly
reduced the formation of the presumptive ultimate carcinogenic metabolite of
BAP, 7,8-diol-9,10-epoxy BAP, and its covalent binding with guanosine in DNA.
Kouri et al. Mouse Study (1978) --
This study was designed as an investigation of the cocarcinogenic activity
of TCDD administered to mice in conjunction with subcutaneous administration of
3-methylcholanthrene (MCA). Two inbred strains of mice, C57Bl/6Cum (abbreviated
B6) and DBA/2Cum (abbreviated D2), were used. These strains are responsive and
nonresponsive, respectively, to the induction of aryl hydrocarbon hydroxylase
(AHH) by MCA.
Groups of mice of both sexes were injected subcutaneously at 4 to 6 weeks of
age with either 150 ug of MCA dissolved in trioctanoin or with trioctanoin
alone. Some groups were also injected with TCDD dissolved in p-dioxane, either
simultaneously with the administration of MCA or 2 days earlier. Two doses of
TCDD (1 ug/kg and 100 ug/kg) were used, and the effects of both intraperitoneal
and subcutaneous injections were investigated. Two sets of experiments,
involving 29 groups of mice, were conducted approximately 1 year apart (Tables
32 and 33).
78
-------
After treatment, the mice were observed for 36 weeks, during which time they
were palpated weekly for the presence of tumors; latency was calculated when the
subcutaneous tumors became 1 cm in diameter. Only tumors characterized
histologically as fibrosarcomas at the site of inoculation were considered. It
is unclear whether or not these were the only tumor types observed. The term
"carcinogenic index" used by the authors was defined as the percentage of tumor
incidence 8 months after treatment divided by the average latency in days
multiplied by 100. No details were given of the number of animals in each group
at the start of each experiment but the numbers dying in the first 28 days and
the numbers at risk (surviving 36 weeks) were tabulated. The results of this
study are shown in Tables 32 and 33.
No subcutaneous tumors were observed in controls or in mice treated with
TCDD alone. In B6 (responsive) mice, the administration of TCDO did not
significantly enhance the induction of tumors by MCA. However, in both
experiments involving D2 (nonresponsive) mice, the administration of TCDD
simultaneously with MCA appeared to enhance the carcinogenic response. The
"carcinogenic index" increased from 1 to 6 in groups treated with MCA alone to
14 in the group treated subcutaneously with TCDD at 1 ug/kg, and 13 to 15 in the
groups treated intraperitoneally with TCDD at 100 ug/kg. The authors concluded
that TCDD acts as a cocarcinogen. They speculated that it may act by local
induction of AHH at the site of inoculation.
A more appropriate statistical analysis would be a comparison of tumor
incidence in TCDO-treated groups with tumor incidence in correspond!'ng MCA-
treated groups within the same experiment. The results of this analysis are
given in Table 34.
From these results, the CAG concluded that the experiment adequately
79
-------
TABLE 32. EFFECTS OF INTRAPERITONEAL ADMINISTRATION OF TCDD ON MCA-INITIATED SUBCUTANEOUS TUMORS
(Kourl et al. 1978)
Treatment
Inbred
No. of mice
No. of mice
No. of
I of mice
Average
Carcino-
strain
dying because
at risk for
mice with
with tumors
latency
genic Index''
-2 days
0 days
of treatment3
tumors"
*tumorsc
(day's)
B6 1.p.
p-d1ox1n
s.c. trloctanoln
1
39
0
0
1.p.
TCDD (100 ug/kg)
s.c. trloctanoln
20
27
0
0
None
s.c. MCA
1
36
29
81
125
65
None
I.p. TCDD (100 ug/kg)
20
30
0
0
None
i.p. TCDD (100 ug/kg)
+ s.c. 30
43
33
71
123
63
MCA
None
i.p. TCDD (1 ug/kg)
4
46
0
0
None
1.p. TCOD (1 ug/kg) +
s.c. 6
27
27
100
132
76
MCA
1 .p.
TCDD (100 ug/kg)
s.c. MCA
20
25
21
84
129
65
1.p.
TCDD (1 ug/kg)
S.C. MCA
6
23
16
70
140
50
D2 1.p.
p-dloxane
s.c. trloctanoln
6
22
0
0
I.p.
TCDD (100 ug/kg)
s.c. trloctanoln
24
25
0
0
None
s.c. MCA
3
34
1
3
217
1
None
I.p. TCDD (100 ug/kg)
30
38
0
0
None
I.p. TCOD (100 ug/kg)
+ s.c. 43
43
10
23
178
13e
MCA
None
I.p. TCDD (1 ug/kg)
5
48
0
0
None
I.p. TCDD (1 ug/kg) +
s.c. 5
34
5
15
199
7
MCA
1 .p.
TCDD (100 ug/kg)
s.c. MCA
20
28
0
0
1.p.
TCDD (1 ug/kg)
s.c. MCA
6
31
0
0
aDuring the first 28 days following treatment
^Defined as the number of mice surviving the 36-week observation period.
cAt the end of the 36-week experiment
^Percentage of Incidence of tumors, divided by the average latency In days, multiplied by 100 (8).
eThis carcinogenic index value lies outside (greater than) the 99% confidence Interval (i.e. P < 0.01) constructed from seven
different studies over the past 5 years during which 150 ug of MCA was given s.c. to D2 mice. These studies Included 295 02 mice, the
mean =50 for all seven studies was a carcinogenic Index of 5.43 = 2.70.
-------
TABLE 33. EFFECT OF INTRAPERITONEAL OR SUBCUTANEOUS ADMINISTRATION OF TCDO GIVEN 2 DAYS BEFORE OR SIMULTANEOUS
WITH SUBCUTANEOUS ADMINISTRATION OF MCA ON TUMORIGENESIS IN D2 MICE
(Kouri et al. 1978)
-2 days
Treatment
0 days
No. of mice
dying because
of treatment
No. of mice
at risk for
tumors
No. of
'* mice with
tumors
% of mice
with tumors
Average
latency
(days)
Carcino-
genic inde)
None
s.c. MCA
0
30
3
10
177
6
l.p. p-dloxane
s.c. MCA
10
40
4
10
194
5
i.p. TCDD (100 ug/kg)
s.c. MCA
35
65
9
14
145
10
None
i.p. p-dloxane x s.c. MCA
5
45
5
11
176
6
None
i.p. TCDD {100 ug/kg) + s.c. MCA
38
62
17
27
183
15a
None
i.p. TCDD (1 ug/kg) + s.c. MCA
22
78
8
10
162
6
None
s.c. p-dioxane + s.c. MCA
2
68
8
12
180
6
None
s.c. TCDD (100 ug/kg)
8
42
0
0
None
s.c. TCDD (100 ug/kg + s.c. MCA
18
82
46
55
145
38a
None
s.c. TCDD (1 ug/kg)
2
48
0
0
None
s.c. TCDD (1 ug/kg + s.c. MCA
2
98
21
21
154
l4a
dThese carcinogenic Index values lie outside the 99% confidence interval.
-------
TABLE 34. INCIDENCE OF TUMORS IN MICE TREATED WITH MCA AND WITH MCA AND TCDD
Experiment
Dose of
TCDD
Route of
administration
Tumor Incidence
TCDD and MCA MCA
P-value^
1
100 ug/kg
Intraperi toneal
10/43
l/34a
P = 0.01
2
100 ug/kg
Intraperitoneal
17/62
5/45
P = 0.03
2
100 ug/kg
Subcutaneous
46/82
5/45
P = 3.0 x 10"7
2
1 ug/kg
Subcutaneous
21/98
5/45
P = 0.1
"Vehicle (p-dioxane) not administered.
bp-value calculated using the Fisher Exact Test (one-tailed).
-------
demonstrated tf-1 enhancement by TCDD of tumor induction when TCDD was
administered s- lultaneously with MCA at the higher dose (100 ug/kg). The
reported results at the lower dose (1 ug/kg) are not statistically significant
unless the reduction in latency is taken into account, which is difficult to do
rigorously. Despite defects in reporting (failure to specify the initial number
of animals in each group and to report tumor incidence by sex), the results
provide convincing evidence that TCDD acts as a cocarcinogen. The failure of
TCDD to induce tumors when administered alone was not unexpected since only a
single dose was administered and the duration of the study was very short (36
weeks).
83
-------
ESTIMATION OF TCDD LEVELS IN 2,4,5-T STUDIES
As discussed above, all of the 2,4,5-T studies which either did not
demonstrate an oncogenic effect or had ambiguous results (i.e., all 2,4,5-T
studies except the one conducted by Kociba), failed to provide significant
evidence of either the oncogenicity or lack of oncogenicity of 2,4,5-T because
of deficiencies in their design. However, it is useful to calculate the highest
doses of TCDD which were administered in these studies as a contaminant of
2,4,5-T and compare these doses to the doses of TCDD which induced an oncogenic
response in the TCDD studies. The calculated doses are based on the reported
contamination levels of the 2,4,5-T used in the studies and the dose of 2,4,5-T
administered. Tables 35 and 36 show that the doses of TCDD administered as a
contaminant in each of the 2,4,5-T studies, with the possible exception of the
Bionetics oral mouse study, were below those doses which produced an observable
oncogenic response in the TCDD studies on the same species. Thus, especially in
view of the deficiencies and insensitivity of the 2,4,5-T studies, it is not
surprising that the TCDD contamination did not induce an observable oncogenic
effect. In the Bionetics* oral mouse study, if it is assumed that the 2,4,5-T
administered was contaminated with 30 ppm TCDD, then the TCDD dose was 0.27
ug/kg/day. This is higher than or equal to the dose which induced an oncogenic
response in the NCI oral mouse study (0.071 and 0.28 ug/kg/day in male and
female mice, respectively). The absence of an oncogenic effect in the Bionetics
study may be explained by several factors. First, the study was much more
insensitive than either the Toth or NCI studies because the group sizes were
much smaller. Second, different strains or substrains of mice were used.
Third, as explained earlier, the TCDD contamination may not have been as high as
30 ppm.
84
-------
TABLE 35. COMPARISON OF DOSE LEVELS OF TCDD IN 2,4,5-1* STUDIES
WITH RESPECT TO THE TCDD STUDY IN MICE WHERE POSITIVE TUMOR
INCIDENCE WAS OBSERVED
Study
Strain of mouse
Route Dose-level
2,4,5-T TCDD
mg/kg/day ug/kg/day
Tumors observeds
(Innes)
Bionetics
hybrid of
7b1/6 and
C3H/AWf (Strain
"A") or "X"
diet
0.27
Muranyi-
Kovacs
NCI
Toth
(Innes)
Bionetics
Fi hybrid of
C57B1/6 and
AKR (Strain "Y"
or "B")
XV11G
C3Hf
diet
0.27
B6C3F1
Male*3
B6C3F1
Femaleb
Swiss male
"A or Y"
"Y or B"
diet
diet
gavage
gavage
gavage
12
12
6.0 x 10~4
6.0 x 10~4
1.42 x 10"3
7.1 x lO-3
7.1 x 10-2
5.7 x 10-3
2.85 x lO-2
0.285
1.0
0.1
0.001
subcutaneous . 215 mg/kg 6.4
(one dose only) (one dose only)
Muranyi-
Kovacs
XVIIG^ subcutaneous 10(4 doses only) 5 x 10~4 (4 doses only)
C3Hf 10(4 doses only) 5 x 10"4 (4 doses only)
dTCDD contaminant in 2,4,5-T
30 ppm--Innes et al. Study (assumed in this analysis, see page 32)
0.05 ppm--Muranyi-Kovacs et al. Study
0.05 ppm--Leuschner et al. (German Study)
0.33 ppb--Dow Chemical Company Study
bCarcinogenic in male and/or female.
85
-------
TABLE 36. COMPARISON OF DOSE LEVELS OF TCDD IN 2,4,5-T STUDIES
WITH RESPECT TO THF. TCDD STUDY IN RATS WHERE POSITIVE TUMOR
INCIDENCE WAS OBSERVED
Study
Strain
Route Dose-level
2,4,5-T TCDD
mg/kg/day ug/kg/day
Tumors observed
Kociba 2,4,5,Ta
Sprague-Dawley
diet
3
1.0 x 10"6
(Spartan)
10
3.3 x 10"6
-
30
9.9 x 10~6
+ (?)
Leuschner
2-4-5-Tb
Sprague-Dawley
ii
3
1.5 x 10~4
-
(SIV50)
10
5 x 10"4
-
30
1.5 x 10~3
•?
Kociba TCDD
Sprague-Dawley
U
1 x 10"3
-
(Spartan)
—
1 x 10"2
+
—
1 x 10"1
+
NCI TCDD
Osborne-Mendel
gavage
--
1.42 x 10-3
+
—
7.1 x 10-3
+
—
7.1 x 10-2
4.
Van Miller TCDDc
Sprague-Dawley
diet
—
5.0 x lO-2
+ (?)
—
2.50 x 10"1
+ (?)
30 mg/kg/day group, compared to matched controls using Kociba's diagnoses;
significant (P < 0.001) when compared to historical controls; significant when using
Squire's diagnoses (P = 0.025). Although no detectable TCDD was present, it is
assumed for this analysis that TCDD is present at the level of detection (0.33 ppb).
^Significance of increase of interstitial cell tumors of testes in the 30
mg/kg/day group is unclear because of great disparity in incidences in two different
control groups and because historical controls untreated with acetone had comparable
incidences of these tumors.
cCertain aspects of this study tend to lessen the reliance wfiich may be placed
on its results.
86
-------
Potency of TCDD
The carcinogenic potency of TCDD is greater than that of aflatoxin B]_,
which is one of the most potent carcinogens known. This conclusion comes from a
comparison of the tumor incidence in male Fischer rats (Wogan et al. 1974),
which were fed 50 ppb of aflatoxin B^, with the incidence of the same tumor
type in female Sprague-Dawley rats (Kociba et al. 1977) fed 0.1 ug/kg/day
(2.2 ppb). The potency of each of these compounds was estimated by calculating
the slope of the linear one-hit model for these compounds. The slope (B) is
calculated according to the following formula:
B = 1 In (1 - Pc)
d TT~n>t)
d = dose inducing carcinogenic effect in the respective studies on TCDD and
afl atoxin.
Pc = tumor incidence in control animals in the respective studies.
= tumor incidence in treated animals in the respective studies at dose d.
This calculation was made on the basis of the lowest dose level at which
TCDD or aflatoxin B]_ caused a significant increase in hepatocellular
carcinomas, the incidence of hepatocellular carcinomas at the respective dose
levels, and the spontaneous incidence of this type of cancer in the control
animals of each study.*
Table 37 shows that TCDD is more potent than aflatoxin by a factor of
0.110/0.032 = 3.45. On this basis, it is estimated that TCDD is a more potent
carcinogen than aflatoxin Bi roughly by a factor of three.
AWogan et al. are not clear on their histologic classification of
preneoplastic lesions. Therefore, only carcinomas were selected for calculating
B.
87
-------
TABLE 37. COMPARISON OF CARCINOGENIC POTENCY
OF TCDD WITH AFLATOXIN Bi
TCDD Aflatoxin Bi
Author Kociba et al. 1977 Wogan et al. 1974
Species Sprague-Dawley rats Fischer rats
Sex Female Male
Tumor incidence in controls (Pc) 1/86 0/18
Dose (d), Tumor incidence in
treated animals (P-t) 2.2 ppb, 11/49 50 ppb, 20/25
Carcinogenic 0.110 (ppb)"1 0.032 (ppb)"1
potency (B.)
SUMMARY OF LABORATORY ANIMAL STUDIES ON 2,4,5-T, SILVEX, AND TCDD
There is highly suggestive evidence that 2,4,5-T is carcinogenic in rats.
The chronic mouse studies on 2,4,5-T suffered from deficiencies in design or
conduct which made them insensitive for detecting an oncogenic response.
Therefore, these studies do not provide significant evidence of either the
carcinogenicity or non-carcinogenicity of 2;4,5-T in mice.
All the chronic animal studies on silvex suffered from deficiencies in
design or conduct which made them insensitive for the purpose of detecting an
oncogenic effect. Therefore, these studies do not provide significant evidence
of either the carcinogenicity or non-carcinogenicity of silvex.
Studies on rats and mice provide substantial evidence that TCDD is
carcinogenic in rats and mice. There is also evidence that TCDD is a potent
liver cancer promoter and is a cocarcinogen. On the basis of the Dow study
(Kociba et al. 1977, 1978), it appears that TCDO is one of the most potent
carcinogens known.
88
-------
The question arises as to whether the carcinogenic action of TCDD by itself
such as exhibited in the Kociba et al. and the N. studies on rats and mice
could be due to the action of TCDD as a carcinogen and/or a promoting agent.
There is evidence that TCDD can be metabolized to a reactive electrophilic
metabolite which could react with DNA and thereby produce genetic damage of the
sort that is associated with the induction of cancer. However, the reactivity
of this metabolite is extremely high with cellular proteins and, to date, the
degree of interaction with DNA that has been demonstrated is low. This may be
peculiar to the tissues that have been examined for this reaction so far but may
not be generally applicable to the reaction of TCDD with DNA in the body.
Furthermore, TCDD has a chemical structure which makes it likely that it could
intercalate into DNA and also act as a genotoxic carcinogen. Promoting agents,
when administered alone, characteristically produce a relatively small increase
in the occurrence of tumors and these tumors are of the sort that occur
spontaneously. This is not characteristic of TCDD, particularly in relation to
its ability to induce squamous carcinomas of the lung and of the hard palate and
nasal turbinates. Squamous carcinomas of the lung are exceedingly uncommon in
the rat in contrast to adenomas of the lung. For these reasons, the CAG
believes that it is prudent, given the present state of knowledge, to regard
TCDD as a complete carcinogen as well as a promoting and cocarcinogenic agent.
89
-------
VI. EPIDEMIOLOGIC STUDIES
SOFT TISSUE SARLUMA
Prompted by clinical observations over a 7-year period of malignant
sarcomas in seven men with previous occupational exposure to phenoxyacetic acid
herbicides (Harden 1977), researchers at the Department of Oncology, University
Hospital, Umea, Sweden initiated epidemiologic studies to test the hypothesis of
an etiologic association. The investigators elected to conduct case-control
studies, a type of epidemiologic research particularly well suited for rare
diseases with long periods of induction (Cole 1979). Cases were defined as male
patients with sarcomas of soft connective tissue, such as smooth muscle
(leiomyosarcoma) and fat (1iposarcoma). The distributions of tumor types in the
two studies are shown in Table 38. Sarcomas of harder connective tissues, such
as bone and cartilage, were excluded because they were not present in the
original series of cases.
TABLE 38. DISTRIBUTION OF TUMOR TYPES IN T^O CASE-CONTROL STUDIES OF
SOFT TISSUE SARCOMA
Percent
of cases
Ti ssue of
Study A
Study B
Diagnosis
origi n
(n=52)
(n=110)
Leiomyosarcoma
Smooth muscle
30
23
Fibrous histiocytoma
Subcutaneous connective
17
25
tissue
Liposarcoma
Fat tissue
14
6
Neurogenic sarcoma
Nerve tissue
10
4
Angiosarcoma
Blood vessels
8
2
Myxosarcoma
Primitive connective tissue
6
8
Fibrosarcoma
Fibrous tissue
4
8
Other sarcomas
11
24
Total
100
100
Sources: Study A, unpublished information supplied by Hardell toEPA.
Study B, Eriksson et al. (1979)
90
-------
Two case-control studies were conducted, the first in northern Sweden
(referred to below as Study A), and the second in the southern part of the
country (Study B). The frequencies of exposure to the substances of primary
interest are shown in Table 39. In the north, occupational exposure to
phenoxyacetic acids took place in both forestry and agricultural work. In the
south, these exposures were predominantly agricultural. The phenoxyacetic acids
to which exposure occurred consisted predominantly of 2,4,5-T and 2,4-0 in both
studies. Exposure to 2,4,5-T in the absence of 2,4-0 was rarely reported in
either study. Exposure to chlorophenols, which contain chlorinated
dibenzodioxin impurities (Levin et al. 1976), occurred mostly in sawmill work
and paper pulp production. Very few persons reported joint exposure to both
phenoxyacetic acids and chlorophenols in these studies.
Of the two phenoxyacetic acids to which exposure predominantly occurred
(2,4,5-T and 2,4-D), only 2,4,5-T is known to be contaminated with TCDD. There
are two published oncogenicity studies on 2,4-0, one in rats (Hanson et al.
1971) and the other in mice (Innes et al. 1969). These studies are inadequate
to assess the carcinogenicity of 2,4-0. In study B, a relative risk of 4.9 (90%
confidence interval 1.6 - 11.1)* was found in relation to exposure to phenoxy,
acid herbicides other than 2,4,5-T (2.,4-D, MCPA, mecoprop, dichloroprop).
Relative risks in relation to the three major categories of exposure arp
shown in Table 40.** Studies A and B indicate a risk of developing soft tissue
*Test-based method of Miettinen (1976); chi-square statistic, no continuity
correction.
**In the analyses considering phenoxyacetic acids only and chlorophenols
only, persons exposed to the other category of substances were excluded. In
study A, the three persons exposed to both chlorophenols and phenoxyacetic acids
were included in all comparisons.
91
-------
TABLE 39. EXPOSURE FREQUENCIES IN TWO CASE-CONTROL STUDIES OF SOFT-TISSUE
SARCOMA
Percent Exposed
Study A
Study B
Cases
Controls
Cases
Controls
Substance(s)
(n=52)
(n=206)
(n=110)
(n=219)
Phenoxyacetic acids only
23.1
6.3
12.7
2.3
Chlorophenols only
11.5
2.4
10.0
3.6
Both
1.9
0.5
0
0
Total
36.5
TT
22.7
T5"
Sources: Study A, Hardell and Sandstrom (1979)
Study B, Eriksson et al. (1979)
TABLE 40. RELATIVE RISKS OF SOFT TISSUE SARCOMA IN RELATION TO EXPOSURE TO
PHENOXYACETIC ACIDS AND CHLOROPHENOLS IN TWO CASE-CONTROL STUDIES
Phenoxyacetic
Phenoxyacetic
Chlorophenols
aci ds
and/or
acids only
only
chlorophenols
Study A Study B
Study A Study B
Study A
Study B
Relati ve
riska
5.3 6.8
6.6 3.3
5.7
4.7
90% Confidence
interval
2.7-10.2 3.1-14.9
2.8-15.6 1.6-7.0
3.2-10.2
2.7-8.3
Unmatched odds ratio
^Test-based method of Miettinen (1976); chi-square statistic, no
continuity correction.
Sources: Study A, Hardell and Sandstrom (1979)
Study B, Eriksson et al. (1979)
92
-------
sarcomas among workers exposed to phenoxyacetic acids only, chlorophenols only,
or phenoxyacetic acids and/or chlorophenols several times higher than among
persons not exposed to these chemicals. In each comparison, the point estimate
of relative risk is high and unlikely to have resulted by chance alone.
Little is known of the etiology of soft tissue sarcoma, so the consideration
of confounding 1n these studies is largely a hypothetical matter. Age, sex, and
place of residence were eliminated as possible confounding factors in the
selection of controls.* Because of the high correlation between exposure to the
substances of interest and employment in agriculture and forestry, a reasonable
hypothesis could be developed that some unknown factor present in these
occupations was responsible for the elevated relative risks.
To test this hypothesis, it is possible to calculate the relative risk in
relation to phenoxyacetic acid exposure in Study B, restricting the analysis to
workers within agriculture and forestry. The result is a relative risk of 6.1
(90% confidence interval 2.4 to 15.4). This finding strongly suggests that a
confounding risk factor for soft tissue sarcoma distributed throughout
agriculture and forestry work was not responsible for the overall increase in
risk found in relation to phenoxyacetic acid exposure.
*Controls were matched individually to cases on the basis of these factors.
Unmatched analyses are presented in Table 40 for the sake of simplicity. The
matched-method relative risks for exposure to phenoxyacetic acids and/or
chlorophenols were 6.2 (90% confidence interval 3.4-11.2) in Study A and 5.1
(90% confidence interval 2.8-9.3) in Study B.
93
-------
Because exposure histories were obtained by means of q -stionnaires and
interviews, the major potential source of bias in these st. iies stems from the
need to rely upon the personal recollection of cases and controls for exposure
histories. The published papers indicate that the researchers paid a great deal
of attention to this potential problem and state that they took all reasonable
precautions to avoid it during the conduct of the study.
In addition, the relative risk calculated by considering the agriculture and
forestry workers who did not report exposure to phenoxyacetic acids or
chlorophenols and comparing them to unexposed persons in other occupations was
0.9 (90% confidence interval 0.3 to 2.4) in Study B. This suggests that a great
deal of recall bias was not present (Axelson 1980).
Of additional interest are the reports of soft tissue sarcoma deaths in two
cohort studies of workers exposed to TCDD. In a study of workers exposed to
TCDD during and after a 1949 trichl orophenol process accident in a ?«1onsanto
Company facility at Nitro, West Virginia, a death from fibrous histiocytoma was
reported (Zack and Suskind 1980). This cause of death was noted by the authors
as a rare event. In a study of Dow Chemical Company workers exposed to TCDD in
a 19c \ "chloracne incident" in a trichlorophenol production area, one of the
three cancer deaths was due to fibrosarcoma (Cook et al. 1980). These two
deaths from this rare form of cancer support the inference from the Swedish
case-control studies that soft-tissue sarcoma may be a hazard of exposure to
TCOD. These and others cohort studies should be continued to determine whether
additional deaths from soft-tissue sarcoma develop.
The associations reported in the two case-control studies of soft tissue
sarcoma are great enough that they are unlikely to have resulted entirely from
random variation, bias, or confounding, even though the possibility cannot be
completely dismissed that bias or confounding was present. These'results are
-------
supported by the occurrence of individual cases of soft-tissue of sarcoma,
usually a relatively rare form of cancer, in two cohort studies of workers
exposed to TCDD and trichlorophenol. Therefore, the studies provide a strong
suggestion that phenoxyacetic acid herbicides, chlorophenols, and/or TCDD are
carcinogenic in humans.
MALIGNANT LYMPHOMA
A separate series of clinical observations at the Department of Oncology in
Umea, Sweden (Harden 1979) led the researchers to conduct a case-control study
of malignant lymphoma in relation to phenoxyacetic acids, chlorophenols, and
other organic compounds (Hardell et al. 1980). Approximately one-third of the
cases in this study were patients with Hodgkin's disease; the remainder of the
lymphomas were non-Hodgkin's forms. MacMahon (1966) and, more recently,
Gutensohn and Cole (1980) have stated that late adult-onset Hodgkin's disease
and the other forms of lymphoma are likely to share similar etiologies.
This study employed essentially the same methods and achieved results
closely comparable to the soft tissue sarcoma studies: fivefold to sixfold
relative risks in relation to phenoxyacetic acids and chlorophenols considered
separately or together. In addition, an elevated relative risk was found in
connection with exposure to organic solvents such as benzene, trichloroethylene,
and styrene. In the published report, the methods and results were incompletely
documented, especially the possibility of confounding by exposure to the organic
solvents. The researchers indicate that an additional report of this study is
in preparation.
Other research has tentatively suggested that lumberjacks may be at
increased risk of lymphoma (Edling and Granstam 1980). In addition, the Zack
l
and Suskind study of workers exposed to TCDD found three deaths from cancers of
SE
-------
the lymphatic and hematopoietic system, against only 0.88 death expected
(relative risk 3.4, 90" confidence interval 0.9-8.8)*
The lymphoma case-control study (Harden et al. 1980) is consistent with the
two soft-tissue sarcoma studies discussed above. On the other hand, the
consistency could also reflect an as yet unidentified methodologic bias in all
these studies. It would be useful for a study using the same methods to be
conducted on a type of cancer not suspected as a hazard of exposure to
phenoxyacetic acids or chlorophenols. If such a study did not show a comparably
elevated relative risk, the likelihood that some methodologic source of bias
produced the results of the soft-tissue sarcoma and lymphoma studies would be
extremely low.
STOMACH CANCER
Studies of two of the oldest cohorts of workers known to have been exposed
to phenoxyacetic acid herbicides and/or TCDD report stomach cancer mortality
rates higher than expected, but the results in each study are based on small
numbers of deaths. In one study (Axelson et al. 1980), 348 Swedish railroad
workers with at least 46 days of herbicide exposure between 1955 and 1972 were
followed through October 1978. The workers were grouped on the basis of their
primary herbicide exposures: those exposed to phenoxyacetic acids (2,4-D and
2,4,5-T) only, to amitrole (aminotriazole) only, and to both types of
herbicides.
In addition to the overall results, the authors were able to provide data
according to the preferred practice of limiting observation to the time
^Obtained by exact Fisher method (Rothman and Boice 1979).
96
-------
following a minimum period of cancer, induction — in this case, 10 years from
first exposure. The results are shown in Table 41. Expected deaths were
derived from Swedish national mortality rates specific for age, sex, and
calendar year.
TABLE 41. STOMACH CANCER MORTALITY IN A GROUP OF SWEDISH RAILROAD WORKERS
EXPOSED TO HERBICIDES, 10 OR MORE YEARS FROM ONSET OF EXPOSURE
Exposure Stomach cancer deaths Relative 90% confidence
category Observed Expected risk interval
Phenoxy acids 2 0.33 6.1 1.1-19.1
Ami trole 0 0.20 —
Amitrole and 1 0.18 5.6 0.3-26.4
phenoxies
Source: Axel son et al. (198U)
The estimate of relative risk of stomach cancer for workers with primary
exposure to phenoxyacetic acids, but not amitrole, is 6.1. Although this
estimate is based on small numbers, the one-tailed Poisson test suggests that it
is not likely to have arisen by chance alone (P = 0.044).
The group of all workers with exposure to the phenoxyacetic acids,
including those who also had amitrole exposure, had a relative risk of stomach
cancer of 5.9 (90% confidence interval 1.6-15.2, three observed stomach cancer
deaths, 0.51 expected).
The other study showing increased stomach cancer mortality is the follow-up
of 75 workers exposed to TCDD during and after a 1953 runaway reaction at a
trichlorophenol manufacturing facility in Ludwigshafen, Federal Republic of
97
-------
Germany (Thiess and Frentzel-Beyme 1977). In this study, three stomach cancer
deaths occurred in contrast to no deaths from stomach cancer in a control group
of 75 men, each matched to cohort members by age and date of entry into
employment and selected at random from a list of over 10,000 persons working in
the same plant. This result indicates an excess of stomach cancer mortality in
the exposed workers.
The researchers also derived expected numbers of deaths for this cohort from
national and regional mortality rates,* yielding results more readily comparable
to those from the Swedish cohort study discussed above. In this analysis, using
1970-1975 rates for the region in which the plant is located and allowing for a
10-year minimum induction period, the researchers found a relative risk of
stomach cancer of 7.5 (90% confidence interval 2.0-19.4, 3 observed stomach
cancer deaths, 0.40 expected). Although strong temporal trends in the regional
stomach cancer mortality rates between 1963 (the start of the follow-up period
in this analysis) and 1970 (the beginning of the period covered by the
comparison rates) would render the expected death figure somewhat inaccurate,
this result confirms the observation of the matched comparison group and are
consistent with the Swedish cohort study.
In summary, the evidence that phenoxyacetic acids and/or TCD0 might increase
the risk of stomach cancer consists of two studies, each of which reports an
excess that is based on only three stomach cancer deaths. Further follow-up of
these and similar cohorts is warranted, but firm conclusions cannot be made on
the basis of the available data.
*An earlier version of the report also included expected deaths calculated
from municipal mortality rates, but these were later found to be inaccurate.
98
-------
OTHER STUDIES
Four additional cohort studies have examined cancer mortality rates in
groups of workers exposed to phenoxyacetic acids and/or TCDD. These are a study
of Dow Chemical Company 2,4,5-T production workers (Ott et al. 1980), a study of
Finnish phenoxyacetic acid herbicide applicators (Riihimaki et al. 1978), and
two studies in which trichlorophenol production workers were exposed to TCDD:
the previously mentioned Nitro study (Zack and Suskind 1980) and study of Dow
Chemical Company employees (Cook et al. 1980).
As noted above, the Mitro study showed a suggestive increase in lymphatic
and hematopoietic cancer mortality. In addition, the Nitro study and the study
by Cook et al. each included a single death from soft-tissue sarcoma.
The CAG has determined that three of these studies as evidenced by the
extremely small numbers of expected cancer deaths in each, have such low
statistical-power that they cannot be taken as strong evidence of the absence of
increased carcinogenic risk in the groups of people studied. In the Mitro
study, 9.04 deaths from all malignant neoplasms and only 0.5 from stomach cancer
were expected. If the researchers had allowed for a minimum period of cancer
induction, these figures would have been even lower. In the study by Ott et
al., only 2.6 deaths from all malignant neoplasms were expected with allowance
for a 10-year minimum induction perod. The study by Cook et al., with only 1.6
expected deaths from all forms of cancer without allowance for a minimum
induction period, had the lowest chance of detecting an effect of all three
studies.
Statistically, the study of Finnish herbicide applicators is inconsistent
with the results of the Swedish and West German cohort studies. Without regard
for induction periods, this study reported 34.5 expected deaths from all
malignant neoplasms. The study, therefore, appears powerful enough to detect
99
-------
relative risks even smaller than those seen in the Swedish and West German
studies. A partial explanation for this apparent inconsistency could lie in the
fact that the Finnish study set the minimum period of herbicide exposure for
membership in the cohort at 10 days (two working weeks) and noted that the
"total length of exposure has, in most cases, been a few weeks only." The
Swedish study of herbicide applicators set the minimum exposure period at 46
days (>1 spraying season).
There are also certain inconsistencies in the data from the Finnish study
which the authors note but find difficult to explain. In particular, no cancer
deaths occurred during the latter part of the study period among Forestry
Authority workers (one of four groups included in the cohort), even though 9.0
deaths were expected. This finding strongly suggests some deficiency in
follow-up or in the source records from which vital status was determined.
In summary, four additional cohort studies of workers exposed to
phenoxyacetic acid herbicides and/or TCDD do not report increased risks of
stomach cancer. Only one of these, however, was statistically powerful enough
to be inconsistent with the two studies that tentatively suggest an increase in
stomach cancer risk. The available report of this study of Finnish herbicide
applicators contains methodologic questions that require clarification.
SUMMARY
Two Swedish case-control studies report highly significant associations
between soft tissue sarcoma and exposure to phenoxyacetic acid herbicides and/or
chlorophenols. These studies provide strongly suggestive evidence for the
carcinogenicity of 2,4,5-T and/or TCDD. The degree of bias or confounding
necessary to produce the highly elevated relative risks in these studies is not
likely to have occurred. Weaker evidence exists from epidemiological studies
100
-------
that exposure to 2,4,5-T and/or TCDD may also increase the risk of malignant
lym^. oma and stomach cancer in hum is. Published studies that have not shown
increases cancer mortality among workers exposed to 2,4,5-T and/or TCDD have low
statistical power and, therefore, do not provide strongly contradictory
evi dence.
101
-------
QUANTITATIVE RISK ASSESSMENT
I. INTRODUCTION
This section assesses the carcinogenic risk posed to humans by 2,4,5-T,
silvex, and the contaminant TCOD as a result of the use of commercial 2,4,5-T
and silvex in the U.S. The highly suggestive evidence of carcinogenicity both
in humans and in animals for 2,4,5-T and the substantial animal evidence for
TCDD dictates that an estimate of risk be made for these chemicals. For silvex
itself there is no direct evidence of carcinogenicity so that the quantitative
estimate of risk will be calculated only for the TCDD contaminant of silvex.
Since TCDD may also act as a promoter or cocarcinogen an additional cancer risk
exists that is dependent upon exposures to other agents. However, since there
is no theoretical basis for quantitatively estimating the risk posed by cancer
promoters or cocarcinogens, this potential is not considered in the quantitative
estimate of risk.
The individual estimated risks are what we consider to be the maximum
plausible primary cancer hazard due to the use of the herbicides 2,4,5-T and
silvex for a given exposure. This opinion is based upon the conservative manner
in which the data used in the extrapolation are selected, the form of the
dose-response model employed, and the conservative assumptions used in
estimating the total years of exposure.
It is well recognized that the risk estimates given are only a rough
measure of the true potential danger, but even so they should be more useful
than no quantified estimate of hazard.
The resulting risk estimates should be viewed as plausible upper bounds
while the plausible lower bound, due to the possibility of highly non-linear
behavior at low doses, is virtually zero. Since all other mathematical models
102
-------
presently in use WfllCfl (.Uinunii iu i.uiiiiiiuii ly ai-i.cpi.cu yt uii,ipi» UI vncmi ju; anu
biology would give risk estimates within this range, we feel that their
employment would not provide any additional useful information.
This risk assessment is based on two main elements: 1) a mathematical model
for extrapolation of animal to human dose-response was developed which can be
utilized to estimate risk given an average lifetime exposure to the herbicides,
and 2) estimates of the lifetime average exposure to various use patterns of the
herbicides were made.
The mathematical model is based on the rationale explained in the
"Carcinogen Assessment Group's Method for Determining the Unit Risk Estimate for
Air Pollutants," July 31, 1980 (Appendix G). All the experimental animal data
for 2,4,5-T and TCOD considered in the employment of the model are fully
explained and the results obtained are given in the next section.
The estimated human exposures from, the use'of these herbicides were supplied
to the CAG by the Hazard Evaluation Division (HED) of the Office of Pesticide
Programs of EPA and is attached as Appendix F. These estimates were used as
given except for the changing of units to mg/kg body wt/day, the appropriate
unit for the mathematical model. All of the qualifications, liabilities,
assumptions, and reservations about the exposure estimate expressed in the HED
document should be kept in mind in evaluating CAG's risk assessment since they
naturally apply to all situations where the exposure estimates are utilized.
Also, quantitative estimates of risk were made for only certain uses and
routes of exposure of commercial 2,4,5-T and silvex. The CAG's analysis is
confined to those situations where HED had sufficient information to generate an
exposure estimate.
Risks are estimated below for exposure to workers in forestry, range and
brush control, rice-weed control, on rights-of-way, and for exposure to the
general population and local populations through the diet by contaminated food.
103
-------
Mo estimates ae made of risk due to general environmental exposure.
II. ESTIMATION OF THE DOSE-RESPONSE MODEL
In order to estimate the slope of the dose-response relationship utilizing
the linearized multistage model, it is necessary for each study to:
1) Obtain the number of animals that have one or more tumors of all types
that are judged to be statistically significantly greater in a test group than
in the control group at the 0.05 level as determined by the Fisher Exact 2x2
Test, or an exact trend test, and are also judged to be biologically meaningful
responses to a carcinogen.
2) Fit the multistage model to these data and test the goodness of fit
utilizing the chi-square statistic. If the fit is rejected at the 0.01 level,
the data are refitted with the highest dose omitted. The process is continued
until an acceptable fit is obtained. For that data set, the largest linear
component that is still consistent with the observed data, which may be viewed
as a 95% upper bound, is used as the slope estimates upon which the risk
estimates are based.
3) These linear components are next translated into human equivalent slope
estimates where exposure to humans is in terms of mg/kg body wt/day by
multiplying the animal slope estimate by the cube root of the ratio of human to
animal weight, where a 70 kg human is assumed.
4) The maximum of these human equivalent slope estimates for all the data
sets is then selected to be used as the slope of the linear or more generally
the one-hit model from which the risk estimates will be made.
In Tables 42-48, all of the tumor data that were judged to be statistically
elevated over control are presented. In Tables 49-58, the data actually used to
fit the models, the value for the goodness of fit test, the fitted
104
-------
parameters of the multistage model, the upper bound linear component, and the
human linear component are all shown for each data set.
In Table 59, the final human slope estimate is given for each data set. The
maximum slope factor for all the data sets are 1.82 x 10~2 (mg/kg/day)~l for
2,4,5-T and 4.25 x 10^ (mg/kg/day)~l for TCDD which are used in the risk
estimation of all subsequent risk.
The slope for TCDD for 2,4,5-T spray applicators may be converted to be used
for exposure given in terms of 2,4,5-T by multiplying the assumed TCDD
contamination rate of 2,4,5-T, 4 x 10*8, by 4.25 x 1Q5, the slope for TCDD,
giving a value of 1.70 x 10"2 (mg/kg/day)"1.
Under these assumptions an estimate of the lifetime probability of cancer
for an applicator due to exposure to a lifetime average exposure of x mg/kg/day
of commercial 2,4,5-T is
p = 1 _ e"^i + B2)x
where 3^ = is the maximum converted human slope for TCDD and B2 = is the
maximum human slope for 2,4,5-T alone, or
p = 1 . e-(0.0170 + 0.0182)x = 1 . e-0.0352x
For applicator exposure to silvex, the risk equation in that case is related
only to the TCDD contaminant
P = 1 - e-Bl x a 1 . e-0.017x
As discussed in detail in the exposure document, the TCDD contaminant of
both 2,4,5-T and silvex is assumed to be present at 40 ppb only for .the sprayer
105
-------
applicator exposure. For dietary exposure due to subsequent contamination of
food, except for rice where silvex is used to predict 2,4,5-T levels, measured
TCDD levels are used for exposure estimates. This is because the environmental
breakdown rates of TCDD and 2,4,5-T are sufficiently different so that the
contamination rate of 2,4,5-T would not be predicted from the observed TCDD
levels.
III. RISKS FOR APPLICATORS
Risks to workers in forestry, range!and, rice levies, and rights-of-way
spraying with 2,4,5-T and silvex have been estimated. As noted above, TCDD has
been assumed to be present as a contaminant in 2,4,5-T and silvex at 40 ppb.
For forestry patterns, 2,4,5-T exposure has been measured in workers by Lavy as
noted by HED. These measurements have also been assumed for other sprayers of
range, rice levies, and rights-of-way. These exposures and associated risks are
shown in Tables 60 and 61.
Worker exposure to silvex has not been measured but silvex has been assumed
to be absorbed into the body in the same manner as 2,4,5-T. Thus, exposure is
also estimated by using the Lavy 2,4,5-T data discussed by HED. It is assumed
that the applicators are exposed to the same number of hours each year that
apply 2,4,5-T as those that apply silvex and the number of workers applying
silvex compaired to 2,4,5-T is the same as the usage ratios presented by HED
(pg. 13 Appendix F) and shown below. As a result risk to applicators applying
silvex would be equivalent to the risks based upon TCDD contaminant which is
shown in Tables 60 and 61.
106
-------
Uses 2,4,5-T:Silvex ratio
Range!and/pasture 10:1
Forestry 100:1
Rice 1000:1
Rights-of-way 10:1
FORESTRY
For forestry sprayers, risks based on measured exposure are shown in Table
60. Lavy gives the exposure as total dose based on the actual clearance of
2,4,5-T from 21 workers. Based on total hours exposed per year and total worker
population exposed and an assumed 40 year working life, a total lifetime
exposure was estimated and lifetime cancer risks have been extrapolated. The
upper limits on these lifetime risks range from 10~4 to 10" ^ with the
highest risk associated with the-aerial mixer-loaders, 2.7 x 10"3. The small
number of workers exposed, however, results in a very small number of cases per
year, even under the assumption of a 40-year working lifetime. Furthermore, the
above analysis does not assume protective clothing.
RANGE AND BRUSH CONTROL
Based on estimated exposure for unprotected range sprayers, Table 61 shows
upper limits on lifetime risks of 10"® to 10~4, with the highest risk of 1.7
x 10"4 to the mixer/loaders. With only 200 of_these estimated, however, the
estimated annual case rate is essentially 0. The risk to each of the 20,000
backpack sprayers is estimated to be 3.5 x 10"®.
RICE-WEED CONTROL
Based on the measured exposure from the forestry workers, adjusted for
application rates of the active ingredient 2,4,5-T, the estimated(lifetime risks
are presented in Table 61. These estimated risks for unprotected workers are
107
-------
in the 10"® to 10~5 range, with the highest risk of 7.8 x 10"^ to the
mixer/loaders. These risks are about one-half to one order of magnitude less
than those for the range sprayers due to the less time exposed.
RIGHTS-OF-WAY BRUSH AND WEED CONTROL
Based on the measured exposure from the forestry worker, adjusted for
application rates of the active ingredient 2,4,5-T, the estimated lifetime
cancer risks for these unprotected workers are presented in Table 61. These
risks are about one-half an order of magnitude greater than the risks for the
forestry workers.' This is mainly because of the higher concentration of 2,4,5-T
used in the application. The upper limit risks for these unprotected workers is
in the TO"4 to 10"^ range.
IV. RISKS DUE TO DIETARY EXPOSURE
Human exposure to TCDD contamination of foods by 2,4,5-T and silvex is
presented. There are four areas where HED has reported TCDD levels: beef fat,
cows' milk, deer and elk fat. HED has also presented an analysis of 2,4,5-T
contamination of milk and rice. The CAG has used these estimates.
BEEF AND MILK
Two monitoring studies for dioxin in beef fat have been discussed by HED,
with emphasis on the Phase One Beef Study of the EPA Dioxin Implementation Plan,
1975. In this study 67 samples of beef were analyzed. Based on a correlation
of those samples analyzed by high resolution mass spectomotry and the
application rate of the herbicide, HED projects a residue of 4.2 ppt TCDD in
beef fat for cattle and cows grazing on land treated that year w.ith 2,4,5-T at a
rate of 1 lb/acre active ingredient. If the land were to be treated with
108
-------
2,4,5-T at a higher rate, up to the legal limit of 4 lb/acre, both the residues
and associated risks would be correspondingly higher.
Based on the 4.2 ppt TCCD contamination level in beef fat and a beef
consumption of approximately 100 1b/person/year, HED estimates that TCDO dietary
intake from beef for the general population is approximately 0.4 pg/day. For the
local population consuming only contaminated beef, dietary intake could be as
high as 31 pg TCDD/person/day assuming a 5-year treatment cycle.
Likewise, for milk contamination, assumption of 4.2 ppt TCDD in fat of
grazing cows would project to as much as 74 pg TCOO/day dietary intake for local
populations or for those consuming only contaminated dairy products.
Measurements of silvex in milk assumed similar for 2,4,5-T, yield exposure
estimates of 7.1 ng/kg/day 2,4,5-T for the local population.
Based on the above exposure estimates Table 62 shows that the upper limit
risk estimates for beef contamination at the above estimated exposures are
1.9 x 10-4 for the local population and 2.4 x 10"6 for the general
population. For the general population this gives an upper limit number of
cases of 7.5/year. For milk and dairy products the upper limit risk estimate
for' estimated exposures is 4.7 x 10~4 for the average consumer of only
contaminated products.
DEER AND ELK
HED has estimated the dietary intake from TCDD contaminated deer and elk
meat to be between 0.14-9.3 pg/kg/meal for deer and 0.05-20.5 pg/kg/meal for
elk. All consumption is assumed to be by the local population of hunters and
their families. The maximum projected risks based on 12 meals per year for life
are 1.3 x 10"4 for deer and 2.9 x 10"5 for elk. These are presented in
Table 63. More or less consumption would lead to corresponding increases or
decreases in risk.
109
-------
RICE
Based on a possible residue of 2,4,5-T on rice of 12 ppb, 10.9 percent crop
treated annually and a food factor of 0.55 percent, HED estimates the average
intake of 2,4,5-T from contaminated rice as 0.011 ug/day for the general
population. On a dose/body weight basis this becomes 0.154 ng/kg/day for a 70
kg person consuming 1.5 kg food/day. HED also estimates that a high consumer of
rice could ingest ten times as much rice and correspondingly, ten times as much
2,4,5-T. These estimates and results are presented in Table 64. the risks are
from 2.8 x 10~9 to 2.8 x 10"^ in the general population to 2.5 x 10"8 to
2.5 x 10"7 in the high consumer group.
IV. SUMMARY
A quantitative assessment has been calculated for the carcinogenic risk
posed to humans by the use of the herbicides 2,4,5-T and silvex. While there is
no evidence for carcinogenicity of silvex, the evidence for 2,4,5-T is highly
suggestive, and that for the contaminant TCDD is substantial. Furthermore,
TCDD is highly carcinogenic to animals.
The assessment of risk from TCDD exposure covers only the herbicide
applicators and dietary exposure to beef, milk, deer, and elk. For unprotected
workers, the upper limits of lifetime risk of induced cancers are in many cases
as high as or in the 10~3 range. For the general population exposed to
contaminated beef, the upper limit of risk for the estimated exposure is
2.4 x 10~6. For local populations consuming only beef which is contaminated
with TCDD, the risk is much greater, as high as 1.9 x 10~4. For local
populations consuming only milk and other dairy products which are contaminated
with TCDD, the risk is 4.7 x 10~4.
110
-------
For contaminated deer and elk meat, risks to the.local population are no greater
than 10~4 for 12 meals a year.
The upper limit of dietary risk associated with estimated exposure to
2,4,5-T in contaminated rice and milk were in the 10"* ? range for a high
consumer eating only contaminated rice or an average consumer drinking only
contaminated milk.
Ill
-------
TABLE 42. DOW (DR. KOCIBA) TCDD ORAL RAT STUDY (1978) WITH DR. R. SQUIRE'S REVIEW
Male Sprague-Dawley Rats - Spartan Substrain (2 yrs.)a
MALES
Tissue and Diagnosis
0
(control)
Dose Levels (ug/kg/day)
0.001 0.01
0.1
Dow (Kociba) Analysis
1. Tongue
Stratified squamous cell
carci noma
0/76
1/49
1/49
3/42
(P = 0.043)
2. Nasal Turblnates/Hard Palate
Squamous cell carcinoma
0/51
1/34
0/27
4/30
(P = 0.016)
Total
0/76
2/49
1/49
7/42
(P = 5.12 x 10-4)
R. Squires Review
1. Tongue
Squamous cell carcinoma
2. Nasal Turb1nates/Uard Palate
Squamous cell carcinoma
0/77
0/55
1/44
1/34
1/49
0/26
3/44
(P = 4.60 x 10-2)
6/30
(P = 1.36 x 10-3)
Total - 1 or 2 above
(each rat had at least
one tumor above)
0/77
2/44
1/49
9/44
(P = 6.28 x 10"5)
aAverage body weight of male rat = 600 grams
-------
TABLE 43. DOW (DR. KOCIBA) TCDD ORAL RAT STUDY (1970) WITH DR. R. SQUIRE'S REVIEW
Female Sprague-Dawley Rats - Spartan Substrain (2 yrs.)a
FEMALES
Tissues and Diagnoses
0
(control)
Dose Levels (ug/kg/day)
0.001 0.01
0.1
Dow (Koclba) Analysis
1. Lung
Keratinizing squamous
cell carcinoma 0/06
Nasal Turblnates/Hard Palate
Stratified squamous cell,
carcinoma (Revised diagnoses
2/19/79) 1/54
Liver
Hepatocellular hyperplastic
nodules/hepatocellul ar
carcinoma 9/06
0/50
0/49
0/30
1/27
3/50
10/50
(2 had both)
(P = 4.37 x 10~4)
7/49
(P = 6.21 x 10-4)
5/24
(P = 9.46 x 10-3)
34/40
(P = 9.53 x 10-13)
Total 1, 2, or 3 above
(each rat had at least
one tumor above) 9/06 3/50 18/50 34/49
(P = 4.37 x 10-4) (p = 2.13 x 10"12)
aAverage body weight of female rat = 450 grams.
(continued on following page)
-------
TABLE 43. (continued)
R. Squire's Review
1. Lung
Squamous cell carcinoma
0/06
0/50
0/49
8/47
(P = 1.61 x 10-4)
2. Nasal Turb1nate/Hard Palate
Squamous cell carcinoma
0/54
0/30
1/27
5/22
(P = 1.43 x 10-3)
3. Liver
Neoplastic nodules/
hepatocellular carcinoma
16/86
0/50
(P =
27/50
2.42 x 10~5)
33/47
(P = 4.92 x 10~9)
Total Combined (1, 2, or 3 above)
(each animal had at
least one tumor above)
16/86
0/50
(P =
27/50
2.42 x 10"5)
34/47
(P = 1.20 x 10"9)
aAverage body weight of female rat = 450 grams
-------
TABLE 44. NCI TCDD (GAVAGE) BIOASSAY (#80-1765)
Osborne-Mendel Rats (2 yrs.) W = 700 g
MALES3
Dose Levels (ug/kg/wk)
Tissues and Diagnoses
vehicle
control
0
low
0.01
medium
0.05
high
0.5
1. Adrenal
Cortical adenoma^
6/72
9/50
(P = 0.093)
N.S.c
12/49
(P = 0.015)
9/49
2. Thyroid
Follicular cell
adenoma carcinoma
1/69
5/48
(P = 0.042)
8/50
(P = 0.004)
11/50
(P = 2.84 x lO-4)
aSubcutaneous combined fibroma or fibrosarcoma - not significant.
^The biological significance of this tumor In old age rats Is questionable, since It Is commonly
observed In control rats and Is associated with the aging process.
CN.S. = Not significant.
-------
TABLE 45. NCI TCUO (GAVAGE) BIOASSAY (#80-1765)
Osborne-Mendel Rats (2 yrs.) W = 450 g
FEMALES
Dose Levels (ug/kg/wk)
Tissues and Diagnoses vehicle
control low medium high
0 0.01 0.05 0.5
1. Liver
Neoplastic nodule or
hepatocellular carcinoma
5/75
1/49
3/50
14/49
(P = 0.001)
2. Adrenala
Cortical adenoma, or
carcinoma
11/73
9/49
5/49
14/46
(P = 0.038)
aThe biological significance of this tumor in old age rats Is questionable, since they are commonly
observed In control rats and associated with aging process.
-------
TABLE 46. NCI TCDO (GAVAGE) BIOASSAY (#80-1765)
B6C3F1 MICE (2 yrs.) W = 48 g
MALES
Dose Levels (u
g/kg/wk)
Tissue and Diagnosis
vehicle
control
0
low
0.01
medium
0.05
high
0.5
Liver
Hepatocellular
adenoma or carcinoma
15/73
12/49
13/49
27/50
(P = 1.31 x 10~4)
-------
TABLE 47. NCI TCDD (GAVAGE) BIOASSAY (#00-1765)
B6C3F1 MICE (2 yrs.)
FEMALES3
Dose Levels (ug/kg/wk)
Tissues and Diagnoses vehicle
control low medium high
0 0.04 0.2 2.0
1. Subcutaneous tissue
Fibrosarcoma
1/74
1/50
1/48
5/47
(P = 0.032)
CD
2. Hematopoietic system
Lymphoma or leukemia
3. Liver
Hepatocellular
adenoma or carcinoma
18/74
12/50
13/48
3/73
6/50
6/48
20/47
(P = 0.028)
11/47
(P = 1.84 x 10~3)
4. Thyroid
Follicular cell
adenoma
0/69
3/50
1/47
5/46
(P = 8.93 x 10"3)
Total 1, 2, 3 or 4 above 22/74 20/50 19/48 31/47
(each mouse had at least (P = 8.94 x 10"^)
one tumor above)
aAverage body weight of female mouse = 40 grams
-------
TABLE 48. DOW (DR. KOCIBA) 2,4,5-T ORAL RAT STUDY (1978)WITH DR. SQUIRE'S REVIEW
Sprague-Dawley Rats - Spartan Substrain
MALES®
Dose Levels (mg/kg/day)
Tissue and Diagnosis
0 3 10 30
(control)
Dow (Dr. Koclba) Analysis
Tongue
Stratified squamous 1/83 1/50 0/46 4/49
cell carcinoma (P = 0.063)
Dr. R. Squire's Review
Tongue
Squamous cell carcinoma 1/83 1/50& 0/46& 5/48
(P = 0.025)
aAvurage weight of male rat - 600 grams
ty)r. Squire examined all slides from the middle and low dose described by Dow (original report) as
exhibiting any lesions, but did not review tongue slides that Dow described as having no lesions. The
incidence numbers for low and medium dose levels in this table represent this combined review Incidence
(I.e., Dow's tongue diagnoses confirmed by Dr. Squire).
-------
TABLE 49. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL OATA BY STUDY AND PATHOLOGIST.
LINEAR PARAMETER (\{, MAXIMIZED TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study Kociba - Dow
Sex-species Male rat
Weight (wa) 600 gm
Tumor sites (one or more)....Tongue - squamous cell carcinomas
Pathologist - Kociba
Exposure level (mg/kg/day) 0
1 x 10"6
1 x I0"5
1
o
X
^•4
+r/na 0/76
1/49
1/49
3/42
+r = number of animals with one or more of the tumors
n = total number of animals examined
aBecause of an error discovered just before press time, the above data cover only the squamous cell
carcinomas of the tongue and omit from the numerator those animals which developed only nasal turbinate and
hard palate tumors. Preliminary calculations have satisfied us, however, that these data would not have
provided us with the maximum slope factor used in the TCDD quantitative risk assessment.
Estimated
multistage parameters
qo
0.01) = 1.68 x 103
qfi = qt (70/wa)V3 = 0.21 x 103, the upper 95'1 limit one-hit slopo factor associated with
hR-n dose response.
-------
ABLE 50. CURVE FIT OF THE MULTISTAGE MODEL PAF ERS TO EXPERIMENTAL DATA BY STUDY AKD PATHOLOG
LINEAR PARAMETER ql, MAXIMIZE iO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study. Dow
Sex-species Male rat
Weight (wa) 600 gm
Tumor sites (one or more)....Nasal turbinates/hard palate - squamous cell carcinoma
Tongue - squamous cell carcinomas
Pathologist - Squire
Exposure Level (mg/kg/day) 0
1 x 10"6 1 x 10-5
1 x 10-4
+r/n 0/77
2/44 1/49
9/44
+r = number of animals with one or more
n = total number of animals examined
of the tumors
Estimated
multistage parameters qo qi
q2 <13 qf
Goodness of fit
X2
When all dose groups
are used 0.015 1.05 x 103
0 109.40 x 109 3.53 x 103
3.90 (d.f.=1)
When the highest dose
group is not used
Above fit is satisfactory
When "the two highest dose
groups are not used
q| the maximum linear component from the model with adequate goodness of fit (P > 0.01) = 3.53 x 10^
^h = ql = 1*73 x 104, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 51. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL DATA BY STUDY AND PATHOLOGIST.
LINEAR PARAMETER q[f MAXIMIZED TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study Dow
Sex-species Female rat
Weight (wa) 450 gm
Tumor sites (one or more)....Liver, lung, hard palate, or nasal tubinates
Pathologist - Kociba
Exposure level (mg/kg/day)
0 1 x 10~6
1 x 10"5
X
0
1
+r/n
9/86 3/50
18/50
34/49
+r = number of animals with
n = total number of animals
one or mure of the tumors
examined
Estimated
multistage parameters
qo
<11 Q2
<13
of
Goodpess of fit
X^
When all dose groups
are used
0.12
1.23 x 104 0
0
8.63 (d.f.=2)
When the highest dose
group is not used
0.09
0 3.5 x 109
4.69 x 104
0.92 (d.f.=1)
When the two highest dose
groups are not used
Above fit is satisfactory
q| the maximum linear component from the model with adequate goodness of fit (P > 0.01) = 4.69 x 104
q* = q* (70/wa)1/3 = 2,52 x 10^, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 52. CURVE FIT OF THE MULTISTAGE MODEL P/> -.TERS TO EXPERIMENTAL DATA BY SJUDY AND PATIIULl' I.
LINEAR PARAMETER ^, MAXIM1 TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study Koclba - Dow
Sex-spec1es Female rat
Weight (wa) 450 gm
Tumor sites (one or more)... .Ijlver, lung, hard palate, or nasal tubinates
Pathologist - Squire
Exposure level (mg/kg/day(
0
1 x 10-6
1 x 10-5
1 x 10-4
+r/n
16/86
8/50
27/50
34/47
+r = number of animals with
n = total number of animals
one or more
examined
of the tumors
Estimated
multistage parameters
%
n m r = i on v rn4
q* = q* (70/wa)^/^ = 4.25 x 10^, the upper 95% limit one-hit slope factor associated with
human aose response.
-------
TABLE 53. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL DATA BY STUDY AND PATHOLOGIST.
LINEAR PARAMETER (\{, MAXIMIZED TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study NCI
Sex-species Male rat
Weight (wa) 700 gm
Tumor sites (one or more)....Thyroid - adenoma or carcinoma
Pathologist - NCI Reviewed
Exposure level (mg/kg/day) 1.43 x 10"6 7.14 x 10"
-6
7.14 x 10-5
+r/n 1/69 5/48 8/50
11/50
+r = number of animals with one or more of the tumors
n = total number of animals examined
Estimated
multistage parameter qg qi q2 <13
If
Goodness of fit
When all dose groups
are used 7.31 x 10"2 2.85 x 103 0 0
5.
.24 x 103
7.13 (d.f.=2)
When the highest dose
group is not used Above fit 1s satisfactory
When,the two highest dose
groups are not used
qj the maximum linear component from the model with adequate goodness of fit (P > 0.01) = 5.24 x 10^
qf = qf (70/wa)^ = 2.43 x 104, the upper 95% limit one-hit slope factor associated with
human aose response.
-------
TABLE 54. CU„VE r,TL?rETHE ™GEqiH°paiL n^jo w ,w» «« -t.
Compound JCDD
Study NCI
Sex-species Female rat
Weight (wa) 450 gm
Tumor sites (one or more)....Liver tumor
Pathologist - NCI Reviewed
Exposure level (mg/kg/day) 0 1.43 x 10~6 7.14 x 10~6 7.14 x 10~5
+r/n 5/75 1/49 3/50 14/49
+r = number of animals with one or more
n = total number of animals examined
of the tumors
Estimated
multldtage parameters qo qj
<12 <13 qf
Goodness of fit
X?
When all dose groups
are used 0.05 0
5.65 x 107 0 6.09 x 103
1.44 (d.f.=2)
When the highest dose
group is not used
Above fit is satisfactory
When the two highest dose
groups are not used
q| the maximum linear component from the model with adequate goodness of fit (P < 0.01) = 6.09 x 103
q/j = q* (70/wa)^/^ = 3.28 x 10^, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 55. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL DATA BY STUDY AND PATHOLOGIST.
LINEAR PARAMETER qp MAXIMIZED TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study NCI
Sex-species Male mice
Weight (wa) 48 gin
Tumor sites (one or more)....Liver
Pathologist - NCI Review
Exposure level (mg/kg/day) 0
1.43 x 10"6
7.14 x 10"6 7.14 x
10"5
+r/n 15/73
12/49
13/49 27/50
+r = number of animals with one or more
n = total number of animals examined
of the tumors
Estimated
multistage parameters qo
-------
ABLE 56. CURVE FIT OF THE MULTISTAGE MODEL PAf TERS TO EXPERIMENTAL DATA BY STUDY A,NO PATHOLO T.
LINEAR PARAMETER ql, MAXIMA TO GIVE UPPER 95% LIMIT qf
Compound TCDD
Study NCI
Sex-species Female mice
Weight (wa) 40 gm
Tumor sites (one or more)....Subcutaneous tlssue-flbrosarcoma, hematopoietic system lymphoma, or leukemia
Llver-hepatocel lular adenoma or carcinoma; Thyroid-follicular cell adenoma
Pathologist - NCI Reviewed
Exposure level (mg/kg/day) 0 5.71 x 10-6 2.86 x 10~5 2.86 x 10~4
+r/n 22/74 20/50 19/48 31/47
+r = number of animals with one or more of the tumors
n = total number of animals examined
Estimated Goodness of fit
multistage parameters qg qi q2 <13 qf X
When all dose groups
are used 0.41 2.38 x 103 0 0 3.78 x 103 1.20 (d.f.=2)
When the highest dose
group 1s not used Above fit Is satisfactory
When the two highest dose
groups are not used
q^ the maximum linear component from the model with adequate goodnes of fit (P < 0.01) = 3.78 x 103
q* = q* (70/wa)^/3 = 4.56 x 10^, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 57. CURVE FIT OF THE MULTISTAGE MODEL PARAMETERS TO EXPERIMENTAL OATA BY STUDY AND PATHOLOGIST.
LINEAR PARAMETER qp MAXIMIZED TO GIVE UPPER 95'* LIMIT qf
Compound. 2,4,5-T
Study Dow
Sex-species Male rat
Weight (wa) 600 gm
Tumor sites (one or more)....Tongue
Pathologist - Kociba
Exposure level (mg/kg/day) 0
3
10 30
+r/n 1/83
1/50
0/46 4/49
+r = number of animals with one or more
n = total number of animals examined
of the tumors
Estimated
multistage parameters qo qi
Q2 93
Goodness of fit
qf X?
When all dose groups
are used 0.01 0
0 2.63 x 10"6
3.38 x 10"3 1.04 (d.f.=2)
When the highest dose
group is not used
Above fit 1s satisfactory
When the two highest dose
groups are not used
q| the maximum linear component from the model with adequate goodness of fit (P < 0.01) = 3.38 x 10"^
q? = qt (70/wa)l/3 = i 65 x 10-2, Upper 95% limit one-hit slope factor associated with
human close response.
-------
ABLE 58. CURVE FIT OF THE MULTISTAGE MODEL PAr TERS TO EXPERIMENTAL DATA DY STUDY AND PA I HOLD' «.
LINEAR PARAMETER qj, MAXIML TO GIVE UPPER 95% LIMIT qj
Compound 2,4,5-T
Study. Dow
Sex-species Male rats
Weight (wa) 600 gm
Tumor sites (one or more)....Tongue
Pathologist - Squire
Exposure level (mg/kg/day) 0 3 10
30
+r/n 1/83 1/50 0/46
5/48
+r = number of animals with one or more of the tumors
n = total number of animals examined
Estimated
multistage parameters qo qj q2 q3 qj
Goodness of fit
X?
When all dose groups
are used 0.01 0 0 3.51 x 10"6 3.72 x 10"3
0.94 (d.f.=2)
When the highest dose
group is not used Above fit is satisfactory
When the two highest dose
groups are not used
q^ the maximum linear component from the model with adequate goodness of fit (P <
0.01) = 3.72 x lO-3
q* = qf (70/wap/3 = 1.82 x 10"2, the upper 95% limit one-hit slope factor associated with
human dose response.
-------
TABLE 59. HUMAN SLOPE ESTIMATES
Compound Species Study Sex Pathologist Human Slope Estimate q£
TCDD
Rat
Dow
Male
Kociba
Squi re
8.21
1.73
X
X
o o
Female
Kociba
Squi re
2.52
4.25
X
X
105
105
NCI
Male
Female
NCI - Reviewed
NCI - Reviewed
2.43
3.28
X
X
104
104
CO
o
Mice
NCI
Male
Female
NCI - Reviewed
NCI - Reviewed
1.33
4.56
X
X
105
104
2,4,5-T
Rat
Dow
Male
Kociba
Squi re
1.65
1.82
X
X
10":
10"'
*Values used in risk analysis
-------
2,4,5-T MEASURED EX 'IREa CALCULATED ON AN HOURLY BASIS
Use pattern
Exposed group
(number
for 2,4,5-T^)
Dose average
mg/kg/hrb
2,4,5-T
(hrs/yr)
Ri skc
mg/kg/day Lifetime
Lifetime 2,4,5-T
2,4,5-T (pure)
Riskd
Lifetime
based on
TCDD
contami nant
Total
Li fetime
risk
commerical
(2,4,5-T)
Average
cases/yre
Total
2,4,5-T plus
silvex
Forestry
1. Aerial
Pilots (73)
0.015(200)
4.6 x 10~3
8.4 x
105
7.8 x
TO"5
1.6 x lO"4
<10"3
Mixer/Loaders
(73-145)
Supervisors (—)
0.062(800)
0.004(800)
7.6 x 10~2
4.9 x 10~3
1.4 x
9.0 x
10"3
10"5
1.3 x
8.4 x
10"3
lO"5
2.7 x 10"3
1.7 x lO"4
0.06
F1aggers (--)
0.003(800)
3.7 x 10"3
6.7 x
10"5
6.3 x
10"5
1.3 x lO"4
—
2. Ground broad-
cast
a. Tractor
mistblower
Mixer/Loaders (180) 0.020(480)
Driver (90) 0.013(240)
1.5 x 10-2
4.8 x 10-3
2.7 x
8.7 x
10"4
10-5
2.5 x
8.2 x
lO"4
lO"5
5.2 x 10-4
1.7 x 10-4
0.001
<10"3
Supervisor (—)
0.006(480)
4.4 x 10~3
8.1 x
10-5
7.5 x
10-5
1.6 x 10-4
—
b. Backpack
sprayer
Applicator (300)
Mixer-supervi sor
0.021(800)
0.003(800)
2.6 x 10-2
3.7 x 10~3
4.7 x
6.7 x
10-4
10-5
4.4 x
6.3 x
10-4
lO'5
9.1 x 10-4
1.3 x 10-4
0.004
dCompared to skin absorption, potential exposure through the lungs was considered negligibly small by Lavy s
measurements.
^Figures from HED (Appendix F). Numbers exposed for silvex given in text.
1 mg/kg/year for 40 years = 40 year x 1 life x 1 year = 1.54 x 10"3 mg/kg/day lifetime.
/I.3 years 365 days
Q2.4.5-T. Slope = 1.82 xK10-2 (mg/kg/day)-l, from Tabie 59.
dTCD6. Slope = 4.25 x 105 (mg/kg/day)"*, from Table 59. Thi
and silvex.
eTotal expected cases 2,4,5-T plus silvex divided by 71.3.
s risk is for the TCDD contaminant of both 2,4,5-T
-------
TABLE 61. LIFETIME PROBABILITY OF INDUCED CANCER FOR 2,4,5-T AND SILVEX APPLICATORS BASED ON FORESTRY
SPRAYER 2,4,5-T MEASURED EXPOSURE AND ON ASSUMED WORKER EXPOSURE FOR
RANGELAND, RICE LEVY AND RIGHTS-OF-WAY SPRAYING3
Use pattern
Exposed group
(number for
2,4,5-T)
Dose average
mg/kg/hr
2,4,5-T
(hrs/yr)
mg/kg/day
Li fetiine
2,4,5-T
Ri sk
Li fetime
2,4,5-T
(pure)
Risk
Li fetiine
based
on TCDD
Contaminant
Total
Li fetiine
risk
commerical
(2,4,5-T)
Average
cases/yr
Total
2,4,5-T plus
Si 1 vex
Range - brush
and weed control
1. Aerial
Pilots (130)
0.008(75)
9.2 x 10"4
1.7 x 10~5
1.6 x lO"5
3.3 x lO"5
210-4
Mixer/loaders
(130)
0.031(100)
4.8 x 10-3
8.7 x 10-5
8.1 x lO"5
1.7 x 10-4
<10-3
F1aggers (800)
0.002(25)
7.7 x 10-5
1.4 x 10-6
1.3 x 10-6
2.7 x 10-6
<10-4
2. Ground
Rice - weed
control (aerial)
Backpack-spot
applicators
(20,000)
0.008(80)
9.8 x 10-4
1.8 x 10-5
1.7 x lO"5
3.5 x lO"5
.010
Pilots (310)
0.008(12)
1.5 x 10-4
2.7 x lO"6
2.5 x lO"6
5.2 x 10-6
<10-4
Mixer/1 oaders
(310)
0.030(48)
2.2 x 10-3
4.0 x lO"5
3.8 x lO"5
7.8 x lO"5
<10-3
Flaggers
(farm labor)
(6500-9500)
0.0021(0.6)
1.9 x lO-6
3.5 x 10~8
3.3 x 10-°
6.8 x 10-8
<10-5
aSee notes on previous table.
(continued on following
-------
TABLE 61. (continued)
Total
Use pattern
Exposed group
Dose average
Risk
Ri sk
LIfetlme
Average
(number for
mg/kg/hr
mg/kg/day
LI fet
me
Li fetlme
risk
cases/yr
2,4,5-T)
2,4,5-T
LIfetlme
2,4,5
-T
based
commerlcal
Total
(hrs/yr)
2,4,5-T
(pure
on TCDO
2,4,5-T
2,4,5-T plus
contaminant
sllvex
Rights-of-way
1. Aerial
Pilots (25)
0.060(400)
3.7 x 10-2
6.7 x
0-4
6.3 x
0-4
1.3 x 10"3
A
o
1
u>
Mixer/loaders
1.5 x 10-1
(25-50)
0.240(400)
2.7 x
o-3
2.5 x
0"3
4.8 x 10-3
0.004
2. Ground
a. Selective
Basal
ApplIcators
1.3 x 10-1
(1380)
0.084(1,000)
2.3 x
0*3
2.2 x
0-3
4.5 x 10-3
0.091
b. Cut
Stump
Applicators (60)
0.053(500)
4.1 x 10-2
7.4 x
0*4
6.9 x
0-4
1.4 x 10-3
0.001
c. Mixed
Handgun
applicators (270)
0.079(660)
8.0 x lO-2
1.5 x
0"3
1.4 x
0"3
2.9 x 10"3
0.005
Brush
Truck/Boom
applicators (180)
0.005(660)
5.1 x 10-3
9.2 x
0~5
8.6 x
0"5
1.8 x 10"4
<10-3
d. Railroad
Crew (of four)
0-4
(110)
0.066(260)
2.6 x 10-2
4.8 x
4.5 x
0-4
9.3 x 10-4
0.002
e. -Electric
Applicators (400)
0.080(660)
8.1 x 10-2
1.5 x
0-3
1.4 x
0-3
2.9 x 10-3
0.017
Power
a See notes on previous tables.
-------
TABLE 62. ESTIMATED DIETARY INTAKE IN PG/KG/DAY OF TCDD* ANO 2,4,5-T FROM
CONTAMINATION OF BEEF AND MILK OF CATTLE AND COWS GRAZED ON 2,4,5-T TREATED
RANGE OR PASTURE LAND. ALSO ESTIMATED LIFETIME CANCER RISK FROM CONTINUOUS
EXPOSURE, BY LOCAL ANO GENERAL POPULATION
Estimated
Beef Fat
Milk and Diary Products
Local Population*
TCDD pg/kg/day
2,4,5-T pg/kg/day
0.44
1.05
<7.14 x 103
Estimated Risk
1.9 x 10~4
4.7 x 10-4
Exposed Population
7,200-12,300
Average Cases/Year
0.02 - .03
General Population
TCDD pg/kg/day
2,4,5-T pg/kg/day
5.7 x 10-3
Estimated Risk
2.4 x 10~5
Exposed Population
220,000,000
Average Cases/Year
7.5
^Exposure estimates from HED divided by 70 kg for proper unit conversion.
134
-------
TABLE 63. ESTIMATED INTAKE OF TCDD FROM CONTAMINATION OF DEER AND ELK MEAT
BY ANIMALS FORAGING ON 2,4,5-T TREATED LAND
ALSO, ESTIMATED LIFETIME CANCER RISKS
Deer Elk
Dietary intake
pg/kg bw/day for one meal 0.14 - 9.3 0.05 - 20.5
Assumed meals/year* 12 12
Equivalent daily dose
pg/kg/bw/day 0.0046 - 0.3058 0.0016 - 0.6740
Estimated risk 2.0xl0-5-l.3xl0~4 6.8xlQ-7-2.9xlO-5
*For higher or lower consumption, the risk will vary proportionately.
135
-------
TABLE 64. ESTIMATED INTAKE OF 2,4,5-T FROM CONTAMINATION OF RICE BY LOCAL
AND GENERAL POPULATION BY AVERAGE AND HIGH COMSUMER
Rice
Local
population
General
population
Amount
Average
Consumer
High
Consumer
Average
Consumer
Hi gh
Consumer
2,4,5-T
ng/kg/day
1.40
14
0.154
1.5
Estimated
risk
2.5 x 10-8
2.5 x 10-7
2.8 x 10-9
2.8 x 10-8
Exposed
population
220,000,000
Average
cases/year
0.009
136
-------
REFERENCES
Allen, J.R., D.A. Barsotti, J.P. Van Miller, L.J. Abrahamson, and J.J. Lalich.
1977. Morphological changes in monkeys consuming a diet containing 500 ppt
of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Food Cosmet. Toxicol.
15:401-410.
Anderson, 0., and J. Styles. 1978. The bacterial mutation test. Br. J. Cancer
37:924-930.
Axel son, 0. 1980. A note on observational bias case-referent studies in
occupational health epidemiology. Scand. J. Work Environ. Health, (in
press).
Axelson, 0., L. Sundell, K. Anderson, C. Edling, C. Hogstedt, and H. Kling.
1980. Herbicide exposure and tumor mortality: an updated epidemiological
investigation on Swedish railroad workers. Scand. J. Work Environ. Health
(in press).
Bauer, H., K.H. Schulz, and V. Spiegelberg. 1961. Occupational intoxication
in the production of chlorinated phenol compounds. Archives for Industrial
Hygiene 18:538-555.
Berry, D.L., J. DiGiovanni, M.R. Jachau, W.M. Bracken, G.L. Gleason, and T.J.
Slaga. 1978. Lack of tumor promoting ability of certain environmental
chemicals in a two-stage mouse skin. Tumorigenesis Assay. Res. Com. in
Chem. Path, and Pharmacol. 20(1):101-108.
Berry, D.L., T.J. Slaga, T. DiGiovanni, and M.R. Jachau. 1979. Studies with
chlorinated dibenzo-p-dioxins, polybrominated biphenyls and polychlorinated
biphenyls in a two-stage sustem of mouse skin tumorigenesis: potent
anti-carcinogenic effects. Ann. N.Y. Acad. Sci. 320:405-414.
Bionetics Research Laboratories, Inc. 1968. Evaluation of carcinogenic,
teratogenic, and mutagenic activities of selected pesticides and industrial
chemicals. I. Carcinogenic study. Prepared for the National Cancer
Institute. (Innes et al. 1979 publication is abstracted from this report).
Bleiberg, J., M. Wallen, R. Brodkin, and I. Applebaum. 1964. Industrially
acquired porphyria. Archives for Dermatology 89:793-797.
Carter, C.D., R.D. Kimbrough, J.A. Liddle, R.E. Cline, M.M. Zack, W.R. Barthel,
R.E. Koehler, and P.E. Philips. 1975. Tetrachlorodibenzodioxin: an
accidental poisoning episode in horse arena:. Science 188:738-740.
Clement Associates. May 15, 1979. Exposure, toxicity, and risk assessment of
2,4,5-T/TCDD. Prepared for EPA under contract no. 68-01-5095.
Cohen, G.M., N.M. Bracken, R.P. Iyer, D.L. Berry, J.K. Selkirn, and J.J. Slaga.
1979. Anti-carcinogenic effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on
benzo(a)pyrene and 7,12-dimethylbenz(a)anthracene tumor initiation and its
relationship to DNA binding. Cancer Res. 39:4027-4033.
137
-------
Cole, P. 1979. The evolving case-control study. J. Chron. Dis. 32:15-27.
Cook, R.R. J.C Townsend, M.G. Ott, and L.G. >ilverstein. 1980. Mortality
experience of employees exposed to 2,3,7.i-tetrachlorodibenzo-p-dioxin
(TCDD). J. Occup. Med. (in press).
Department of Health, Education, and Welfare Subcommittee on Environmental
Mutagenesis. October 12, 1976. Meeting report.
Dow Chemical Company, Biochemical Research Laboratory. April 8, 1965. Results
of a two year dietary feeding study in dogs of Kurosal®SL (potassium salt
of si 1 vex). Submitted to the U.S. EPA, 1980.-
Dow Chemical Company. 1971. Internal report submitted to the EPA's Advisory
Committee.
Ercegovich, C.D., and K.A. Rashid. 1977. Mutagenesis induced in mutant strains
of Salmonella typhimurium by pesticides. 174th American Chemical Society
National Meeting. Division of Pesticide Chemistry. 43:
Eriksson, M., L. Harden, N.O. Berg, T. Moller, and 0. Alexson. 1979.
Case-control study on malignant mesenchymal tumors of the soft tissue and
exposure to chemical substances. Lakartidningen 76:3872-3875 (translation).
Fujita, K., H. Fujita, and Z. Funazaki. 1975. Chromospheric abnormalities
brought about by the use of 2,4,5-T. J. Jpn. Assoc. Rural Med. 24:77-79.
Gehring, P.J., C.G. Kramer, B.A. Schwetz, J.Q. Rose, and V.K. Rowe. 1973.
The fate of 2,4,5-trichlorophenoxy acetic acid (2,4,5-T) following oral
administration to man. Toxicol. Appl. Pharmacol. 26:352.
Gehring, P.J. and J.E. Betso. 1978. Phenoxy acids: effects and fate in
mammals in chlorinated phenoxy acids and their dioxins. Ecological Bulletin
No. 27.
Goldman, P.J. 1972. Extremely severe acute chloracne due to trichlorophenol
decomposition products. Industrial Medicine, Social Medicine, Industrial
Hygiene 7:12-18.
Goldstein, J.A., J.D. McKinney, G.W. Lucier, P. Hickman, H. Bergman, and J.A.
Moore. 1976. Toxicological assessment of hexachlorobiphenyl isomers and
2,3,7,8-tetrachlorodibenzofurans in chicks. II. Effects on drug metabolism
and porphyrin accumulation. Toxic Appl. Pharmac. 36:81-92.
Goldstein, J.A., P. Hickman, H. 3ergman, and J.G. Vos. 1973. Hepatic porphyria
induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin in the mouse. Research
Communications in Chem. Path, and Pharmacol. 6(3):919-928.
Goldstein, J.A., P. Hickman, H. Bergman, J.D. McKinney, and M.P. Walker. 1977.
Separation of pure polychlorinated biphenyl isomers into two types of
inducers on the basis of induction of cytochrome P-450 or P-448. Chemical
Biological Interaction 17:69-87.
138
-------
Goodman, D.G., 1980. Histopathologic tables concerning The Dow Chemical Company
2,4,5-T and TCDD studies.
Grant, W. 1979. The genotoxic effects of 2,4,5-T. Mutat. Res. 65:83-119.
Green, S. 1975. Cytogenetic evaluation of several dioxins in the rat.
14th Annual Meeting Society of Toxicology. March 9-13, 1975.
Green, S., F. Moreland, and C. Sheu. May 1977. Cytogenetic effect of
2,3,7,8-tetrachlorodibenzo-p-dioxin on rat bone marrow cells. FDA By-Lines.
6:292-294.
Guenthner, T.M., J.M. Fysh, and D.W. Nebert. 1979.
2,3,7,8-tetrachlorodibenzo-p-dioxin: Covalent binding of reactive metabolic
intermediates principally to protein in vitro. Pharmacology 19:12-22.
Gupta, B.N., J.G. Vos, J.A. Moore, J.G. Zinkl, and B.C. Bullock. 1973.
Pathological effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory
animals. Environ. Health Perspect. 5:125-140.
Gutensohn, N. and P. Cole. 1980. Epidemiology of Hodgkin's disease. Sem.
Oncol. 7:92-102.
Hansen, W.H., M.L. Quaife, R.T. Habermann, and O.G. Fitzhugh. 1971. Chronic
toxicity of 2,4-dichlorophenoxy acetic acid in rats. Toxicol. Appl.
Pharmacol. 20:122-129.
Hardell, L. 1977. Malignant mesenchymal tumors and exposure to phenoxyacids: a
clinical observation. Lakartidningen 74:2753-2754 (translation).
Hardell, L. 1979. Malignant lymphoma of histiocytic type and exposure to
phenoxyacetic acids or chlorophenols. Lancet 1:55-56.
Hardell, L., and A. Sandstrom. 1979. Case-control study: soft-tissue sarcomas
and exposure to phenoxyacetic acids or chlorophenols. Br. J. Cancer
39:711-717.
Hardell, L., M. Eriksson, and P. Lenner. 1980. Malignant lymphoma and exposure
to chemical substances, especially organic solvents, chlorophenols, and
phenoxy acids. Lakartidningen 77:208-210 (translation).
Hay, A. 1976. Seveso: the aftermath. Nature 263:538-540.
Hook, G.E., J.R. Haseman, and G.W. Lucier. 1975.
2,3,7,8-tetrachlorodibenzo-p-dioxin induced changes in the hydroxylation of
biphenyl by rat 1iver microsomes. Pharmacol. 24:335-340.
Hussain, S., L. Ehrenberg, G. Lofroth, and T. Gejvall. 1972. Mutagenic
effects of TCDD on bacterial systems. Ambio. 1:32-33.
139
-------
Innes, J.R.M., B.M. lllland, M.G. Valerio, L. Petrocelli, L. Fishbein, E.R. Hart,
A.J. Pallota, R.R. Bates, H.L. Falk, J.J. Gart, M. Klein, I. Mitchell, and
J. Peters. 1969. Bioassay of pesticides and industrial chemicals for
tumorgenicity in mice: A preliminary note. J. Nat'l. Cancer Inst.
42:1101-1114.
Jenssen, D., and L. Renberg. 1976. Distribution and cytogenetic test of 2,4-D
and 2,4,5-T pbenoxyacetic acids in mouse blood tissues. Chem. Biol.
Interactions 14:291-299.
Jones, G., and J.B. Greig. 1975. Pathological changes in the liver of mice
given 2,3,7,8-tetrachlorodibenzo-p-dioxin experimentally. Basel.
1(11) :1315-1317.
Khera, K., and J. Ruddick. 1973. Polychlorodibenzo-p-dioxins: perinatal
effects and the dominant lethal test in Wistar rats. Advances in Chemistry
Series 120:70-84.
Kilian, D., M. Benge, R. Johnston, and E. Whorton, Jr. 1975. Cytogenetic
studies of personnel who manufacture 2,4,5-T. Mew York Academy of Sciences
workshop on occupational monitoring and genetic hazards. March 28-29, 1975.
Kimbrough, R.D. 1974. The toxicity of polychlorinated polycyclic compounds and
related chemicals. CRC Crit. Rev. Toxicol. 2:445-498.
Kimmig, T. and K.H. Schulz. 1957. Chloriete aromatische zyklische AEter als
Ursache der sogenanten Chlorakne. Kurze Originale Mitteilungen. 11:337-338.
Kociba, R.J., P.A. Keller, C.N. Park, and P.J. Gehring. 1975.
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): Results of a 13-week oral
toxicity study in rats. Toxicol. Appl. Pharmacol. 34(2):001-002.
Kociba, R.J., D.G. Keyes, J.E. Beyer, R.M. Carreon, C.E. Wade, D.A. Dittenber,
R.P. Kalnins, L.E. Frauson, C.N. Park, S.D. Bernard, R.A. Hummel, and C.G.
Humiston. 1978. Results of a two-year chronic toxicity and oncogenicity
study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol. Appl.
Pharamacol. 46:279-303. Report submitted to EPA September 28, 1977.
Kociba, R.J., D.G. Keyes, R.W. Lisowe, R.P. Kalnins, D.D. Dittenber, C.E. Wade,
S.J. Gorzinski, N.H. Mahle, and B.A. Schwetz. 1979. Results of a two-year
chronic toxicity and oncogenic study of rats ingesting diets containing
2,4,5-trichlorophenoxy acetic acid (2,4,5-T). Report submitted to EPA
September 27, 1978. (Published 1979. Food Cosmet. Toxicol. 17:205-221).
Kouri, R.E., R.A. Salerno, and C.E. Whitenire. 1973. Relationship between
levels of aryl hydrocarbon hydroylase inducibility and sensitivity to
chemically induced subcutaneous sarcomas in various strains of mice. J.
Nat'l. Cancer Inst. 50:363-368.
Kouri, R.E., and H. Ratrie III. 1974. Aryl hydrocarbon hydroxylase induction
in human lymphocyte cultures by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Life
Science 5:1585-1595.
140
-------
Kouri, R.E. 1976. Relationship between levels of aryl hydrocarbon hydroxylase
activity and susceptability to 3-roethylcholanthrene and benzoCa]
pyrene-induced cancers in inbred strains of mice. C rcinogenesis 1:139-151.
Kouri, R.E., T.H. Rude, R. Joglekar, P.M. Dansette, O.M. Jerina, S.A. Atlas,
I.S. Owens, and D.W. Nebert. 1978. 2,3,7,8-tetrachlorodibenzo-p-dioxin as
a cocarcinogen causing 3-methylcholanthrene-initiated subcutaneous tumors in
mice genetically "non-responsive" at Ah locus. Cancer Res. 38:2777-2783.
Leuschner, F., A. Lueschner, F. Hubscher, W. Dontenwill, and P.Y. Rogulja.
Apr. 9, 1979. Chronic oral toxicity of 2,4,5-T. Batch No. 503, Control No.
153574 b. called for short - 2,4,5-T in Sprague-Dawley (SIV 50) rats.
Laboratorium Fur Pharmakologie und Toxikologie.
Levin, J.O., C.F. Rappe, and C.A. Nilsson. 1976. Use of chlorophenols or
fungicides in sawmills. Scand. j. Work Environ. Health 2:71.
MacMahon, B. 1966. Epidemiology of Hodgkin's disease. Cancer Res.
26:1189-1200.
Magnussen, J., C. Ramel, and A. Eriksson. 1977. Mutagenic effects of
chlorinated phenoxy acetic acids in Drosophila melanogaster. Hereditas
87:121-123.
Majumdar, S., and J. Golia. 1974. Mutation test of 2,4,5-trichlorophenoxy
acetic acid on Drosophila melanogaster. Can. J. Genet. Cytol. 16:465-466.
Majumdar, W., and R. Hall. 1973. Cytogenetic effects of 2,4,5-T on in vivo
bone marrow cells of mongolian gerbils. Journal of Heredity 64:213-216.
May, G. 1973. Chloracne from the accidental production of tetrachloro-
dibenzodioxin. Br. J. Ind. Med. 30:276-283.
Miettinen, 0. 1976. Estimability and estimation in case-referent studies.
Amer. J. Epidemiol. 103:226.
Mullison, W.R. 1966. Some toxicological aspects of si 1 vex. Presentation at
Southern Weed Conference, Jacksonville, Florida. Jan. 18, 1966.
Muranyi-Kovacs, I., G.R. Rudali, and J. Imbert. 1976. Bioassay of
2,4,5-trichlorophenoxy acetic acid for carcinogenicity in mice. Br. J.
Cancer 33:626-633.
Muranyi-Kovacs, I., G. Rudali, and J. Imbert. 1977. Study on the
carcinogenicity of 2,4,5-T in mice (meeting abstracts). Fourth Meeting of
the European Association for Cancer Research held at Universite de Lyon,
September 13-15, 1977. European Association for Cancer Research, Lyon,
France.
National Cancer Institute. 1980a. Bioassay of 2,3,7,8-tetrachlorodibenzo-p-
dioxin (gavage study). DHHS publication no. (NIH) 80-1765. Carcinogenesis
testing program, National Cancer Institute, National Institutes of Health,
3ethesda, M0.
141
-------
National Cancer Institute. 1980b. Bioassay of 2,3,7,8-tetrachlorodibenzo-p-
dioxin for possible carcinogenicity. Technical report no. 201. DHHS
publication no. (NIH) 80-1757, U.S. Department of Health and Human Services,
Public Health Service, National Institutes of Health, Bethesda, MD.
National Institute of Occupational Safety and Health (NIOSH). 1976. Registry
of Toxic Effects of Chemical Substances.
Nebert, D., S. Thorgeirsson, and J. Felton. 1976. Genetic differences in
mutagenesis, carcinogenesis, and drug toxicity. In: F. de Serres, J. Fouts,
J. Bend, and R. Phil pot (eds.), In Vitro metabolic activation in
mutagenesis testing. Elsevier/North Holland Biomedical Press, Amsterdam,
pp. 105-124.
Oliver, R.M. 1975. Toxic effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
laboratory workers. Br. J. of Ind. Med. 32:49-53.
Ott, M.G., B.B Holder, and R.D. Olson. 1980. A mortality analysis of employees
engaged in the manufacture of 2,4,5-trichlorophenoxy acetic acid. J. Occup.
Med. 22:47-50.
Piper, W.N., J.Q. Rose, M.L. Leng, and P.J. Gehring. 1973. The fate of
2,4,5-trichlorophenoxy acetic acid following oral administration to rats and
dogs. Env. Health Perspect. 5:241-244.
Pi tot, H.C., T. Goldsworthy, and H. Poland. 1980. Promotion by
2,3,7,8-tetrachlorodibenzo-p-dioxin of hepatocarcinogenesis from
diethylnitrosamine. (Manuscript being submitted to Cancer Research 1980).
Poland, A., D. Smith, G. Metter, and P. Possick. 1971. A healthy survey of
workers in a 2,4-D and 2,4,5-T plant. Arch. Environ. Health. 22:316-327.
Poland, A., and E. Glover. 1973. Chlroinated dibenzo-p-dioxins: Potent
inducers of -amino levulinic acid synthetase and aryl hydrocarbon
hydroxylase. Molecular Pharmacology. 9:736-747.
Poland, A., and E. Glover. 1974. Comparison of 2,3,7,8-tetrachlorodibenzo-p-
dioxin, a potent inducer of aryl hydrocaron hydroylase, with
3-methylcholanthrene. Mol. Pharmacol. 10:349-359.
Poland, A., and E. Glover. 1976a. Stereospecific high affinity binding of
2,3,7,8-tetrachlorodibenzo-p-dioxin by hepatic cytosol. J. Biol. Chem.
51:4936-4946.
Poland, A., and E. Glover. 1976b. 3,4,3',4'-tetrachloroazoxybenzene and
azobenzene: potent inducer of aryl hydrocarbon hydroxylase. Science
194:527-630.
Poland, A., and E. Glover. 1979. An estimate of the maximum in vivo covalent
binding of 2,3,7,8-tetrachlorodibenzo-p-dioxin to rat liver protein,
ribosomal RNA and DNA. CancerRes. 39:3341-3344.
142
-------
Rasmuson, B., and H. 3vahlin. 1978. Mutagenicity tests of 2,4-dichlorophenoxy
acetic acid and ,4,5-trichlorophenoxy acetic acid in genetically stable
and unstable strains of Drosophila melanogaster. Ecol. Bull. (Stockholm)
27:190-192.
Reggiani, G. 1977. Medical problems raised by the TCDD contamination in
Seveso, Italy. Presented at the 5th International Conference on
Occupational Health in the Chemical Industry. Sept. 5-10, 1977.
Renner, H.W. 1979. Monitoring of genetic environmental risk with new
mutagenicity tests. Ecotoxicology and environmental safety 3:122-125.
Riihimaki, Y., S. Asp, A.M. Seppalainen, and S. Hernberg. 1978.
Symptomatology, morbidity, and mortality of experience of chlorinated
phenoxy acid herbicide (2,4-B; 2,4,5-T) sprayers in Finland. A clinical and
epidemiological study. Working paper for an IARC working group meeting on
coordination of epidemiological studies on the long-term hazards of
chlorinated dibenzodioxins and chlorinated dibenzofurans, Lyon, France.
Jan. 10-11, 1978.
Rose, J.Q., T.H. Ransey, R.A. Wentzler, R.A. Hummel, and P.J. Gehring. 1976.
The fate of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) following single or
repeated oral doses to rats. Toxicol. Appl. Pharmacol. 36:209-226.
Rothman, K.J. and J.D. Boice. 1979. Epidemiologic analysis with a programmable
calculator. Bethesda, Maryland: National Institute of Health, NIH
Publication No. 79-1649.
Rowe, V.K., and T.A. Hymas. 1954. Summary of toxicological information on
2,4-D and 2,4,5-T type herbicides and an evaluation of the hazards to
livestock associated with their use. Am. J. Vet. Research 132:2165-219.
Schwetz, B.A., J.M. Norris, G.L. Sparscho, V.K. Rowe, P.J. Gehring, J.L.
Emerson, and C.G. Gerrbig. 1973. Toxicity of chlorinated
dibenzo-p-dioxins. Environ. Health Perspect. 5:87-98.
Seiler, J. 1973. A survey on the mutagenicity of various pesticides.
Experientia 29:622-623.
Thiess, A.M., and R. Frentzel-Beyme. 1977. Mortality study of persons exposed
to dioxin following an accident which occurred in the BASF on 13 November
1955. Proceedings of MEDICHEM Congress V, San Francisco, Sept. 5, 1977. (in
press).
Thigpen, J.E., R.E. Faitt, E.E. McConnell, and J.A. Moore. 1975. Increased
susceptibility to bacterial infections as a sequelae of exposure to
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Infection and Immunity
12:19-1324.
Toth, K., S. Somfai-Relle, J. Sugar, J. Bence. 1979. Carcinogenicity testing
of herbicide 2,4,5-trichlorophenoxy ethanol containing dioxin and of pure
dioxin in Swiss mice.
143
-------
Tuchmann-Duplesis, H. Nov. 26, 1977. Embryo problems posed by the Seveso
accident. Le Concours Medical. Mo. 44.
Van Miller, J.P., J.J. Lalich, and J.R. Allen. 1977. Incidence of neoplasms in
rats exposed to low levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Chemosphere 6(9):537-544.
Vogel, E., and J.L.R. Chandler. 1974. Mutagenicity testing of cyclamate and
some pesticides in Drosophila melanogaster. Experientia 30(6):521-623.
Vos, J.G., J.A. Moore, and J.G. Zinkl. 1973. Effect of 2,3,7,8-tetrachloro-
dibenzo-p-rdioxin on the immune system of laboratory animals. Environ.
Health Perspec. 5:149-162.
Vos, J.G., J.A. Moore, and J.G. Zinkl. 1974. Effect of 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD) on C57BL/6 mice. Toxicol. Appl. Pharmacol. 229-241.
Wogen, G., S. Palialunga, and P. Newburne. 1974. Carcinogenic effects of low
dietary levels of aflatoxin B]_ in rats. Food Cosmet. Toxicol. 12:281.
Yefimenko, L.P. 1974. Data for assessing the gonadotropic and mutagenic effect
of the herbicide butyl ether 2,4,5-T. Gigiena Truda Professional Nye
Zabolevan. 18:24-27.
Zack, J.A., and R.R. Suskind. 1980. The mortality experience of workers
exposed to tetrachlorodibenzo-p-dioxin in a trichlorophenol process
accident. J. Occup. Med. 22:11-14.
Zetterberg, G. 1978. Genetic effects of phenoxy acids on microorganisms. In:
C. Ramel (ed.) Chlorinated phenoxy acids and their dioxins, mode of action,
health risks, and environmental effects.
144
-------
APPENDIX A
TABLE 111-7.
CUMULATIVE MORTALITY
(KOCIBA ET AL. 1977)
OF MALE
RATS
ug/kg/day TCDD
Time (end of 30-day period) N=
Controls
(86)
0.1
(50)
0.01
(50)
0.001
(50)
1-7
0.0
0.0
0.0
2.0
8
0.0
2.0
0.0
2.0
9
0.0
4.0
0.0
2.0
10
0.0
4.0
0.0
2.0
11
2.3
4.0
0.0
2.0
12
5.8
8.0
0.0
2.0
13
7.0
12.0
0.0
2.0
14
10.5
18.0
4.0
4.0
15
12.8
18.0
14.0
14.0
15
16.3
20.0
22.0
14.0
17
18.6
28.0
28.0
24.0
18
24.4
34.0
34.0
44.0*
19
31.4
44.0
46.0
50.0
20
41.9
46.0
54.0
56.0
21
48.8
62.0
68.0
60.0
22
58.1
74.0*
76.0*
68.0
23
69.8
78.0
84.0
74.0
24
77.9
84.0
88.0
76.0
25
82.6
90.0
92.0
78.0
*Interval of greatest difference, 0, in cumulative mortality curves of
controls and treatment group. None of the differences were statistically
significant (Kolmogorov-Smirnov test, P > 0.05).
A-l
-------
TABLE 111-8.
CUMULATIVE MORTALITY OF FEMALE
(K0CIBA ET AL. 1977)
RATS
Controls
Time (end of 30-day period) N= (86)
0.1
(50)
ug/kg/day TCDD
0.01
(50)
0.001
(50)
LO
1
o
0.0
0.0
0.0
0.0
6-8
1.2
0.0
0.0
0.0
9
1.2
2.0
0.0
0.0
10
1.2
4.0
2.0
0.0
11
1.2
8.0
2.0
0.0
12
1.2
16.0
4.0
4.0
13
3.5
20.0
4.0
4.0
14
3.5
26.0
8.0
6.0
15
7.0
28.0
12.0
10.0
16
12.8
32.0
18.0
12.0
17
15.1
38.0
18.0
18.0
18
18.6
44.0
20.0
22.0
19
25.6
56.0*
30.0
34.0*
20
34.9
60.0
36.0
36.0
21
40.7
66.0
46.0*
44.0
22
58.1
82.0-
60.0
52.0
23
64.0
86.0
66.0
58.0
24
70.9
88.0
72.0
66.0
25
70.9
92.0
72.0
68.0
^Interval of greatest difference, D, in cumulative mortality curves of
controls and treatment group. The mortality curve for the rats fed 0.1
ug/kg/day differed significantly from that for controls (D = 30.4, P < 0.01,
Kolmogorov-Smirnov test). The other two groups did not differ significantly
from controls (P > 0.05).
A-2
-------
TABLE 111-9. MALES: INTERVAL MORTALITY RATES
Days
Control
0.1
ug/kg/day
0.01 ug/kg/day
0.001
ug/kg/day
d/1
rate
d/1
rate
d/1
rate
d/1
rate
40-30
0/86
0.000
0/50
0.000
0/50
0.000
1/50
0.020
31-210
0/86
0.000
0/50
0.000
0/50
0.000
0/49
0.000
211-240
0/86
0.000
1/50
0.020
0/50
0.000
0/49
0.000
241-270
0/86
0.000
1/49
0.020
0/50
0.000
0/49
0.000
271-300
0/86
0.000
0/48
0.000
0/50
0.000
0/49
0.000
301-330
2/86
. 0.023
0/48
0.000
0/50
0.000
0/49
0.000
331-360
3/84
0.036
2/48
0.042
0/50
0.000
0/49
0.000
391-420
3/80
0.038
3/44
0.068
2/50
0.040
1/49
0.020
421-450
2/77
0.026
0/41
0.000
5/48
0.104
5/48
0.104
451-480
3/75
0.040
1/41
0.024
4/43
0.093
0/43
0.000
481-510
2/72
0.028
4/40
0.100
3/39
0.077
5/43
0.116
(continued on following page)
-------
TABLE 111-9. (continued)
Days
Control
0.1 ug/kg/day
0.01 ug/kg/day
0.001 ug/kg/day
d/1
rate
d/1
rate
d/1
rate
d/1
rate
511-540
5/70
0.071
3/36
0.083
3/36
0.083
10/38
0.263
541-570
6/65
0.092
5/33
0.152
6/33
0.182
3/28
0.107
571-600
9/59
0.153
1/28
0.036
4/27
0.148
3/25
0.120
601-630
6/50
0.120
8/27
0.296
7/23
0.304
2/22
0.091
631-660
8/44
0.182
6/19
0.316
4/16
0.250
4/20
0.200
661-690
10/36
0.278
2/13
0.154
4/12
0.333
3/16
0.108
691-720
7/26
0.269
3/11
0.273
2/8
0.250
1/13
0.077
721-726
4/19
0.211
3/8
0.375
2/6
0.333
1/12
0.083
Terminal
Kill 15
5
4
11
corrected for continuity
for combined interval:
421-510
7/77
vs
5/41(x2 =
0.04, n.s.)
12/48(x2 = 4.63,
P < 0.05)
10/48(x2
= 2.54, n
• s.)
451-540
10/72
v s
8/41(x2 =
0.37, n.s.)
10/43(x2 = 1.27,
n.s.)
15/4 3(x2
= 6.37, P
< 0.025)
481-570
13/72
vs
12/40(x2 =
1.48, n.s.)
12/39(x2 = 1.67,
n.s.)
18/43(x2
= 6.59, P
< 0.025)
511-600
20/70
vs
0/36(x2 =
0.03, n.s.)
13/36(x2 = 0.32,
n.s.)
16/38(x2
= 1/47, n
.s.)
-------
TABLE 111-10. FEMALES: INTERVAL MORTALITY RATES
Days
Control
0.1
ug/kg/day
0.01 ug/kg/day
0.001 ug/kg/day
d/1
rate
d/1
rate
d/1
rate
d/1
rate
0-150
0/86
0.000
0/50
0.000
0/50
0.000
0/50
0.000
151-180
1/86
0.012
0/50
0.000
0/50
0.000
0/50
0.000
181-240
0/85
0.000
0/50
0.000
0/50
0.000
0/50
0.000
241-270
0/85
0.000
1/50
0.020
0/50
0.000
0/50
0.000
,271-300
0/85
0.000
1/49
0.020
1/50
0.020
0/50
0.000
'301-330
0/85
0.000
2/48
0.042
0/49
0.000
0/50
0.000
331-360
0/85
0.000
4/46
0.087
1/49
0.020
2/50
0.040
361-390
2/85
0.024
2/42
0.048
0/48
0.000
0/48
0.000
391-420
0/83
0.000
3/40
0.075
2/48
0.042
1/48
0.0?1
421-450
3/83
0.036
1/37
0.027
2/46
0.044
2/47
0.043
451-480
5/80
0.063
2/36
0.056
3/44
0.068
1/45
0.022
481-510
2/75
0.027
3/34
0.088
0/41
0.000
3/44
0.068
511-540
3/73
0.041
3/31
0.097
1/41
0.024
2/41
0.049
(continued on following page)
-------
TABLE 111-10. (continued)
Days
Control
0.1
ug/kg/day
0.01 ug/kg/day
0.001 ug/kg/day
d/1
rate
d/1
rate
d/1
rate
d/1
rate
541-570
6/70
0.086
6/28
0.214
5/40
0.125
6/39
0.154
571-600
8/64
0.125
2/22
0.091
3/35
0.086
1/33
0.030
601-630
5/56
0.089
3/20
0.150
5/32
0.156
4/32
0.125
631-660
15/51
0.294
8/17
0.471
7/27
0.259
4/28
0.143
661-690
5/36
0.139
2/9
0.222
3/20
0.150
3/24
0.125
691-720
6/31
0.194
1/7
0.143
3/17
0.177
4/21
0.191
721-726
0/25
0.000
2/6
0.333
0/14
0.000
1/17
0.059
Terminal Kill
25
4
14
16
corrected
for continuity
for combinned interval:
421-510
10/83
vs
6/371x2 =
0.11,
n.s.)
5/46(x2
= 0.0,
n.s.
) 6/47(x2
= 0.0, n.
s.)
451-540
10/80
vs
8/36(x2 =
1.13,
n.s.)
4/44(x2
= 0.8.
n.s.
) 6/45( x2
= 0.01, n
.s.)
481-570
11/75
vs
12/34(x2 =
4.80,
P < 0.05) 6/41(x2
= 0.0,
n.s.
) ll/44(x2
= 1.34,
n.s.)
510-600
17/73
vs
11/31(x2 =
1.08,
n.s.)
9/41 (x2
= 0.0,
n.s.
) 9/41(x2
= 0.0, n.
s.)
-------
APPENDIX B
PATHOLOGIC EVALUATIONS OF SELECTED TISSUES FROM
THE DOW CHEMICAL TCDD k 2,^,5-T RAT STUDIES
Submitted, to
Cancer Assessment Group
The Environmental Protection Agency
Washington, DC 20460
August 15i 1980
hy
Robert A. Squire Associates, Inc.
1515 La3elle Avenue
Ruxton, Maryland. 2120^
B-l
-------
Hcvlood 8/29/80
DOW TCDD CHRONIC OXICITY STUDY IN RATS
TUMOR INCH ,NCE SUMMARY TA3LS
Daily Dose« 0 MCG/fcC .001 MCG/KG .010 MCG/KG 100 MCG/XG
Soxi M F M F M F MF
PRIMARY LUNG NEOPLASMS* (85) (86) (50) (50) (50) (^9) (**8) (1*7)
Bronchoalveolar Adenoma 2 1
Squamous Cell Carcinoma 1 8
Bronchoalveolar Adenocarcinoma 1
PRIMARY NASAL TURBINATE/*
HARD PALATE NEOPLASMS (55) (54) (34) (30) (26) (27) (30) (22)
Squamous Cell Carcinoma 1 1 6 5
Otherx Odontoma, Tooth 1
PRIMARY TONGUE NEOPLASMS* (77) (76) (44) (4) (49) (8) (44) (41)
Squamous Cell Carcinoma 11 1 3 2
Fibrosarcoma 1
PRIMARY LIVER NEOPLASMS* (85) (86) (50) (50) (50) (50) (50) 0+7)
Neoplastic Nodule 4 16 8 2 26 2 23
Hepatocellular Carcinoma 2 1 10
* Number of Animals with Tissue examined Microscopically
B-2
-------
DOW 2.^,5-T CHRONIC TOXICITY STUDY IN MALE RATS
TUMOR INCIDENCE SUMMARY TABLE
CONTROL LEVEL HIGH DOSE LEVEL
INTEGUMENTARY SYSTEM
Skin/Subcutis:
Fibroma 5/86 3/50
Carcinoma 1/86 1/50
Lipsarcoma 1/50
Malignant Fibrous Histiocytoma 2/86
Calcifying Epithelioma 1/86
Squamous Cell Papilloma 2/86 1/50
Squamous Cell Carcinoma 1/86
HEMATOPOIETIC SYSTEM
Lymph node;
Carcinoma, metastatic l/50
Lymphoma 1/50
Malignant Schwannoma, metastatic 1/86
C-ceH Carcinoma, metastatic 1/86
Thymus:
Malignant Schwannoma, metastatic 1/51
Spleenj
Lymphoma 1/86
Multi sites:
Lymphoma 2/86
CIRCULATORY SYSTEM
Heart:
Endocardial Sarcoma 1/50
8-3
-------
ixevisea c/cy/GQ
100
DOW 2.^,5-T CHRONIC TOXICITY STUDY IN HALE RATS
TUMOR INCIDENCE SUMMARY TA3L5
CONTROL LEVEL
HIGH DOSE LE7E
DIGESTIVE SYSTEM
Liverj
Neoplastic Nodule
Hepatocellular Carcinoma
Pancreas«
Acinar Adenoma
Acinar Carcinoma
Islet Adenoma
Islet Carcinoma
Intestines(small)i
Lymphoma
Adenocarcinoma
Sarcoma
Intestines(large):
Lymphoma
Tongue:
Squamous Cell Carcinoma
Salivary Gland
Carcinoma
2/86
3/86
23/86
1/86
10/86
1/86
2/86
2/86
1/83
1/80
1/50
1/50
13/50
5/50
1/50
2/50
1/50
5A8
URINARY SYSTEM
Kidney:
Adenocarcinoma
Tubular Carcinoma
Urinary Bladder
Transitional Cell Papilloma
1/86
1/86
1/50
B-4
-------
DOW 2A.5-T CHRONIC TOXICITY STUDY IN MAUS RATS
TUMOR INCIDENCE SUMMARY TABLE
CONTROL LEVEL .HIGH DCSE LEVEL
ENDOCRINE SYSTEM
Pituitary
Chromophobe Adenoma 15/80 9/^9
Chromophobe Carcinoma 7/80 2/^9
Adrenal:
Pheochromocytoma 37/8^ 19/^9
Cortical Adenoma 8/8^ 7/^9
Cortical Carcinoma 1/84-
Can gli oneur oma i/^9
Thyroids
C-cell Adenoma V"85 6/^7
C-cell Carcinoma 2/85
Parathyroid:
Chief Cell Adenoma lA3
REPRODUCTIVE SYSTEM
Testes:
Interstitial Cell Tumor 2/86
Mammary Gland:
Adenocarcinoma 1/50
Fibroadenoma 1/86 1/50
NERVOUS SYSTEM
Brain:
Astrocytoma 1/86 1/50
Granular Cell Tumor 1/50
Cranial Nerve:
Schwannoma 1/86
B-5
-------
Revised 8/29/80
DOT 2,4,5-T CHRONIC TOXICITY STUDY IN MALE RATS
TUMOR INCIDENCE SUMMARY TABLE
.CONTROL LEVEL HIGH DOSE LEVEL
SPECIAL SENSES
EAR i
Zymbals' Glandi
Sebaceous Carcinoma 1/86
Squamous Cell Carcinoma 3/86 1/50
EYE;
Squamous Cell Carcinoma 1/^6
MUSCULOSKELETAL SYSTEM
30NE:
Rib?
Chondroma 1/86
30DY CAVITIES
Mesentery:
Lipoma 1/86
Mediastinum;
Malignant Schwannoma 1/86
B-6
-------
ROBERT SQUIRE ASSOCIATES. INC.
1515 Lcrielle Avenue
Rux:on, //.orylcnd 21204
(301) 821-0054
' C
•August 25, i960
Br. Bernard Habsman
Cancer Ascecsrsnt Group
Cfilco of Health and Znvironnental
Assessnsr.t
U.S. Environmental Protection Agency
Washington, DC ZCA60
Dear Dr. rL^/i^r-ana
As per cur asreeasnt, we exarnlned tissues frcn only the
control ana high dcse aninsla from the Low 2,^,5-T two year
rat study. Sines finding the ons additional carcinoma in the
tortus of tha high dcse nale, however, I did exssir.a tongues
fron all _zles in all dose groups in which there were any
pathologic alterations reported "by Dow pathologists. Ky
finding agreed >."ith th.eea of Dow pathologists in that I
fcuad no additional ntoplasr3 anong th3 slides exa:±ned*
Sincerely,
A
/¦>
¦(
&
Kobert A. Squirts, D.Y.A.,
cci Richard 2cso#
RAS/ek
-------
APPENDIX C
LAHORATORIUM FUR PHARMAKOLOGI E UNI) TOX1 KOLOG1 H
PROFESSOR OH. P. LEUSC11NI !l
COPY
D-2104 Hamburg 92, January 17th, 1980
Mr.J.Guy Gwynne
Consul
Amerikanisches Generalkonsulat
Handelsabteilung
Alsterufer 27
D-2000 Hamburg 36
Dear Mr. Gwynne,
today I am allowed to answer to the questions which arose
with the telex from EPA, referring to 'The Chronic Oral Toxicity
of 2,4,5-T, batch No. 403, control No. 1535746 - called for short
'2,4,5-T' - in Sprague-Dawley (S1V 50) Rats with special attention
to Carcinogenic Properties' as follows:
A)
A2)
2,4,5-T (untreated rat:;)
fibroma (thorax) 1 female
fibroma (abdomen) 2 males
1 fi-male
fibroma (uterus) none
fibroma (mamma) none
fibroma (limb) none-
interstitial cell
tumour = testes 22 animals
2,4,5-T (acetonc-crc.K od
rats)
none
none
1
1 female
1 male
6 an i nulls
A1 - A4)
A1)
Historical (untreated control rats, no further i-xporictu-c
with acetone-treated animals; all historical studies 2 to
3 years before examinations with 2,4,5-T)
adenofibroma
(mamma)
interstitial cell
tumour (testes)
fibroma (limb)
interstitial cell
tumour
6 of 50 females
20 of 50 animals
3 males and 1 females of each 90 animals
24 of 90 animals
- 2 -
C-l
AN SCH R 1 FT: F R AN COP ER STR. 66b . D-2104 HAMDURG 92 (NEUGRADEN) . TELEPO N : (040) 701 JO 21 - 23 . 794 25 2S
EXPRESSGUTSTATJOr-': 1IAMDUP. C-HAHOURG
k A M If vriM Tri. M A M B U P.c C R S P A a K A S S P ini.7. 200
-------
- 2 -
A3)
A4)
fibroma (limb)
fibroma (ovary)
interstitial cell
tumour (testes)
fibroma (ovary)
fibroma (mamma)
fibroma (abdomen)
fibroma (limb)
fibroma (head
region)
adenofibroma
(mamma)
interstitial cell
tumour (testes)
1 male and 1 female of 50 nnininls c.-h-Ii
1 of 50 animals
17 of 50 animals
1 of 100 animals
1 of 100 females
3 males and 1 female of 100 animals c;u:li
2 of 100 males
1 of 100 males
6 of 100 females
32 of 100 animals
Altogether tumour rates (tumour-bearing animals) for the studies
mentioned:
A)
A1)
A2)
A3)
A4)
80% males
66% males
77% males
64% males
69% males
80% females
60% females
67% females
66% females
71% females
Approximate age of the animals at the begin of the scudies and
duration:
A) born February, 1976
Al) born November/December, 1974
A2) born November/December, 1974
A3) born March/April, 1975
A4) born February/March, 1975
After 5 to 7 lactation weeks or quarantine start of treatment; dura-
tion of study 130 to 132 weeks (30 to 30.5 months).
B) From Group (III) 3 mg 2,4,5-T/kg b.w. only the prematurely decease.;/
killed animals were examined histologically. Fibromas were found m
each 1 male and female (limb and abdomen respectively), interstitial
cell tumours of testes in 5 rats.
C) After chronic examinations with Sprague-Dawlcv^ rat.-. , performed by Liu-
sponsor (personal information) the highest tested close-level (30
was the even subtoxic one. The dos.i^c was fixed by the sponsor.
- 3 -
C-2
-------
J
D) The tongue was examined n-ie roscop u.\->. it y cogetlier with larynx and
pharynx. These investigar ums did :n
-------
APPENDIX D
LAB ORATORIUM FUR PHARMAKOLO GIE UND T OXIKOLOGIE
PROFESSOR OR. P. 12USCHNE3
HISTOPAXHOLOGICAL EXAMINATIONS IN THE TONGUE
Appendix Co
'Chronic oral Toxicity of 2,4,5-T, batch no. 503,
control no. 153574 b - called "2,4,5-T" - in
Sprague-Dawley(SIV 50) rats'
(date of final report: April 9th, 1979)
- with special attention to carcinogenic properties
Senior Pathologist:
Prof.Dr.med.W.Dontenwill
August 6th, 1980
0-1
• ».urnDj» stj. lib ¦ D.2104 HAMBURG 92 (NEUGRA3EN) TE L2 FO N l (0401 701 30 21 • 23 • 796 23 23
-------
1. REPORT ON HISTOPATHOLOGICAL EXAMINATIONS OF THE TONGUE
1.1. General informations:
The test compound 2,4,5-T was examined in 360 Sprague-Dawley
(SIV 50) rats for neoplastigenic properties over 130 weeks at
oral administration. Each 120 further animals served as con-
trols with acetone as premix to the food or without any premix
to the food (see final report April 9th, 1979).
1.2. Conduct of this additional study:
On. August 1st, 1980 the Environmental Protection Agency asked
as a first step for additional histopathological examinations
in-the tc.-.gue of the male rats, treated with Che highest 2,4,5-T
dose-level of 30 mg/kg b.w./day_and of the untreated male con-
trol animals (without any premix). Therefore longitudinal sec-
tions should be prepared and investigated. On 6ch of August,1980
a further advice of the EPA asked for cross sections of the or-
gan mentioned but at this date already longitudinal sections
were taken. By this fact cross sections of the tongue in the
males at 30 mg 2,4,5-T/kg b.w. and the untreated control males
could not be performed.
1.3. Method:
The tongues of the male rats treated with 30 mg 2,4,5-T/kg b.w./
day in the food and those or the untreated male rats (without
premix) were investigated histopathologically afte. haematoxy-
lin-eosin staining. Therefore longitudinal sections reaching
from the retrolingual region to the tip of the tongue were pre-
pared. Each 8 gradual sections of the tongue were investigated.
The mucosal epithelial thickness of the treated and untreated
rats was compared.
1.4. Findings:
The histopathological investigations in the tongue of rats
showed a localised chronic mucosal inflammation with round-
cell infiltration and proliferation of the connective tissue,
whereby the epithelium showed above the inflammation a mode-
rate acanthosis in the male rat no. 37, treated with 30 mg
2,4,5-T/kg b.w./day in the food. The male animal no.'44 at this
dose-level had a severe phlegmonous inflammation of the toneue's
musculature with small mucosal epithelial ulcers.
D-2
-------
2
Apart from these tvo findings no changes could be seen. The
variation of Che epithelial thickness was, as normal, more
marked at the basis of the tongue. A semiquantitative compari-
son did not show signs for demonstrated hyperplasia. No dys-
plasia, papilloma or carcinoma were found.
0-3
-------
- 3 -
TA3LZ Histopathological Investigations in Che Tongue of Rats
Animal No. Findings
Group (I) Control
- males -
1-50 no pathological findings
Group (V) 30 mg 2,4,5-T/kg
- males -
1-36 no pathological findings
37 localised chronic inflammation of the mucosa with
round-cell infiltration and proliferation of the
connective tissue, the epithelium showed above the
inflammation a moderate acanthosis
38 - 43 no pathological findings
44 severe phlegmonous inflammation of the musculature
with mucosal epithelial ulcers (small)
45 - 50 no pathological findings
0-4
-------
LABORATORIUM FUR PHARMAKOLOGIE UND TOXIKOLOGIH
PROP8SSOR SR. 7. LEUSCHNER
QUALITY ASSURANCE STATEMENT
Based on a quality assurance review, ic was concluded chat chis
report accurately reflects the data for the
'Histopachological Examination in the Tongue'
Appendix to: Chronic oral Toxicity of 2,4,5-T,
batch no. 503, control, no. 153574 b - called
"2,4,5-T" - in Sprague-Dawley(SIV 50) rats
(date of final report: April 9th, 1979)
- with special attention to carcinogenic properties
Approved and
submitted by:
August 8th, 1980
Franz Hiibscher
Director of QAU
Date
0-5
PRANCOPER STR. «6b . D - 2104 H AM 8 U R C 91 (N E U G R AB £ N) • TELSPONi (040) 701 SO 11 • 23 • 799 23 ZS
-------
\
mj
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
OFFICE OF
RESEARCH AND DEVELOPMENT
SUBJECT: Clarification of Telephone Convservation with Dr. Leuschner
FROM: Wade Richardson/¦ ^*4®*-
Office of Health and Environmental Assessment (RD-689)
TO: Charalingayya Hiremath, Ph.D.
Carcinogen Assessment Group (RD-689)
In early August, at CAG's request, I made an overseas telephone call to
Dr. Leuschner in Germany and asked if he would be willing to cut histological
sections of the tongues from male rats in his two year chronic toxicity study
on 2,4,5-T. I first indicated that the Agency preferred that horizontal
sections be cut. However, when Dr. Leuschner expressed preference to cut
longitudinal sections, I indicated to him that I would again discuss with the
appropriate people in the Agency how they felt the sections should be cut and
then call him back to confirm the nature of the Agency's request. Due to some
misunderstanding, it appears that longitudinal sections had already been cut
by the time I called Dr. Leuschner back confirming the Agency's wish that
horizontal sections be cut.
E-l
-------
WKtNUiX f
UNITED 5TATE5 ENVIRONMENTAL PROTECTION AGENCY
3ATE September 12, 1980
£ct Exposure Assessment for 2,4,5-T, Silvex and TCDd
--cm Acting Chief, Environmental Fate Branch, HED
TO Elizabeth Anderson
Carcinogen Assessment Group (RD-683)
Attached is the Exposure Assessment for 2,4,5-T, silvex and TCDD.
. ' 'C
David J. Severn, Ph.D.
cc: P. E. McGrath
EPa Form 1320-6 (R»V. 3-76)
F-l
-------
QUANTITATIVE ASSESSMENT OF EXPOSURE TO 2-, 4,5-T, SILVEX AND TCDD
September 12, 1980
F-2
-------
QUANTITATIVE ASSESSMENT OF EXPOSURE TO 2,4,5-T, SILVEX AND TCDD
INTRODUCTION
As part of its risk-benefit balancing procedures, the
Agency generally attempts to estimate potential human exposure to
pesticides in quantitative terms. The ultimate objective of these
assessments is to develop numerical estimates of the amount of
exposure that certain segments of the population may experience
as a result of pesticide use. These exposure data are combined
with toxicity information to generate an overall risk assessment.
The risk assessments are then used to predict potential health
effects based on the toxicologic effects of the pesticide in
question.
This document provides some quantitative estimates of exposure
to 2,4,5-7, silvex, and' TCDD for use in the cancellation hearings.
These estimates are based as far as possible on observed residue
levels in the environment. However, while these estimates are
expressed as numerical values, they are in fact much less precise
than their numerical nature would imply. This is because the
available data are meager, because conditions (spray techniques,
weather, etc.) are so variable, and because many assumptions have
to be utilized in order to arrive at the estimates. This intro-
duction describes some of the reservations which apply to the
numerical estimates presented in this assessment, and comments on
the limitations on the use and interpretations of this information.
F-3
-------
-ii-
General
Agency exposure assessments, including this analysis for
2,4,5-T, silvex, and TCDD, are based where possible on actual
field data. In the present case, the data upon which this
exposure assessment is based include data on chemical residues
in soil, food and other environmental materials, on actual field
exposure data for applicators, and on the data on transport and
fate of these chemicals in the environment.
In addition, information on pesticide use practices and
extent of use is necessary to arrive at reasonable estimates of
exposure. This information includes the crops or sites which may
be treated, the rates and methods of application, and information
on the other activities during their subsequent application. This
information is used to develop estimates of the number of people
potentially exposed to the chemicals by oral, dermal and inhalation
routes as a result of specific use practices.
The information available for use in this exposure assessment,
is variable as to its completeness, quality, and reliability. In
general, the greatest confidence can be placed on the field exposure
and residue data, even though it is incomplete in many ways. The
information relating to use practices is somewhat less certain.
Agency scientists started with information from the pesticide
label to determine application rates and crops or sites likely to
be treated. Estimates relating to the extent of sites or crops
F-4
-------
-iii-
treated and other indicators of the probable extent of contam-
ination ara subject to many uncertainties. In particular, the
numerical values for the populations at risk are highly uncertain.
This is because information on population demographics, whether
or not related to pesticide use, is not well developed.
The uncertainties described above are common, in varying
degrees, to all exposure assessments, including these assess-
ments for 2,4,5-T, silvex and TCDD. In sum, although Agency
scientists have a high degree of confidence about much of the
empirical data which form the basis for this analysis, they are
far less confident about other information. The quantitative
exposure estimates for the populations at risk are limited by
these uncertainties.
Exposure Analysis
The starting point for exposure assessment for pesticides
is descriptive information on pesticide release and distribution
to the different environmental corroartments such as air, water,
soil, and animal and plant tissues during application. In
addition, 2,4,5-T and silvex are known to move from sites of
application to non-target areas under some conditions of
application.
This qualitative information on potential sources of human
exposure is supported by analytical chemical data showing that
residues of these chemicals are present subsequent to application,
F-5
-------
-iv-
both a application sites and at non-target sites. Such chemical
residue information provides the initial numerical base for quanti-
tative estimates of possible human exposure. For example, unlike
many pesticides with relatively short half-lives and relatively
rapid disappearance from the environment, 2,4,5-T and silvex may
persist in the environment for several months after application;
TCDD may remain for several months or years. Therefore, special
concern is raised about 2,4,5-T, silvex and TCDD because they may
remain in the environment in significant concentrations for
several months or years after their application.
However, despite the availability of some useful information,
there are gaps in our knowledge. For example, although large
amounts of 2,4,5-T and silvex are used each year, comprehensive
monitoring information on 2,4,5-T, silvex, and TCDD residues in
the environment is, for the most part, unavailable.^/ This
paucity of residue information limits the Agency's ability to
make quantitative exposure estimates to only some routes of
exposure and only for certain uses.
V The paucity of monitoring data on TCDD is due largely to
the only recent development of analytical methodologies with
sufficient sensitivity to measure the extremely low levels of
TCDD which are of biological concern, to the limited number of
facilities with these analytical capabilities, and to the high
cost of analyzing samples at these levels. For 2,4,5-T and
silvex, the problem of insufficient monitoring information
appears to be largely due to a lack of comprehensive monitoring
programs, or inappropriate sampling.
F-6
-------
-V-
Even when some data are available for one kind of application,
there may be uncertainty as to whether those data are applicable
to other applications which may occur under different conditions.
For example, residue data collected during springtime application
in the Pacific Northwest may not properly describe the amount
and distribution of chemicals under different environmental
conditions at a different time of the year. Often, the only data
available are data derived from laboratory studies, with little
or no field data to verify that the laboratory data accurately
describe the residue levels which might be present under field
conditions.
Further, each of the several different human exposure
pathways provides a different kind of exposure potential. Even
when some empirical residue data on a given route of exposure
are available, there are often uncertainties concerning the
generalization of those data to other routes of exposure. These
uncertainties are a particular concern when estimating exposure
to chemicals such as TCDD which appear to pose risks at very low
levels of exposure.
In attempting to generalize to "average" or "typical" use
patterns, the Agency has encountered a wide variety of practices,
which were very difficult to address. An example is the appli-
cation rate to be used when rangeland vegetation is spot treated.
Despite the fact that the USDA-EPA States Report (Ref. 2) notes a
F-7
-------
-vi-
2 lb/A maximal application rate on grazing lands, it was found
that other rates have been used and are permitted by the label.
Also, despite "typical" 5-15 year recommended intervals between
herbicide spray applications, instances of successive annual
treatments have been substantiated, and may, in fact, be more a
common practice than the USDA Report assumes.
A very difficult aspect of quantitating risk is specifically
identifying and quantitating populations at risk. The Agency
has found, for example, that deer and elk from 2,4,5-T treated
forested areas may contain TCDD residues in their fat at readily
measured levels. Also, it is known that some people include
deer and elk in their diets. But, the proportion of deer and
elk taken by hunters annually that are actually contaminated,
the level of contamination, and the numbers of people who
consume given amounts of contaminated meat is not known.
To extrapolate from the available information to potential
human exposure (and subsequently to risk assessments), assump-
tions based on the observed residue data, information about use
practices, and "typical" consumption patterns are made. These
assumptions may either over- or under-estimate actual risk.
This can be confirmed only by the acquisition of additional data.
Nevertheless, the Agency has developed some numerical values,
however uncertain, to permit the quantitative estimation of risk
for the cancellation proceedings.
F-8
-------
-vii-
The exposures which have been quantified in this document
are as follows;**/
1) Occupational exposure to 2,4,5-T, silvex, and TCDD.
2) Dietary exposure of the general population and local
populations to TCDD residues in beef and local populations to
TCDD residues in dairy products resulting from the use of
2,4,5-T and silvex on rangeland and pasture.
3) Dietary exposure of local populations to TCDD residues
in deer and elk resulting from the forestry use of 2,4,5-T and
silvex.
4) Dietary exposure of the general population and local
population to silvex residues in rice, apples, pears, prunes,
and sugar (from sugarcane) resulting from the use of silvex on
these food products.
5) Dietary exposure of the general population and local
populations to 2,4,5-T and/or silvex residues in rice resulting
from the use of 2,4,5-T and silvex on rice.
Finally, the available data relating to some uses of 2,4,5-T
and silvex are inadequate even to begin assessing potential
human exposure. For some situations, no monitoring information is
known to the Agency, and in other situations the available data
**/ The Agency is still evaluating and generating monitoring
data which were not utilized in these quantitative assessments.
The Agency may utilize these data as they are developed.
F-9
-------
-viii-
are too incomplete or too uncertain to provide the basis for even
a simple estimate of exposure. It is emphasized that the incom-
pleteness of data and the consequent lack of an exposure analysis
mean only that suitable data were not available, not that these
pathways are biologically insignificant.
F-10
-------
-1 -
ESTIMATION OF OCCUPATIONAL EXPOSURE TO 2,4,5-T, SILVEX, AND TOP
Intreduction
This analysis provides a quantitative hunan exposure V estimate for
2,4,5-T, silvex, and dioxin in terms of absorption by the body of these
chemicals under normal agricultural working conditions.
Human exposure estimates are made on the basis of chemical analyses of
dermal and inhaled concentrations of the chemical or chemicals, and if
the information is available, on the basis of the amount of chemical (s)
or their metabolites excreted by the body (e.g. in the urine). **/
In the case of the pesticides and contaminant under consideration, there
are experimental data available on the occupational exposure to pesticide
applicators and farrtwsrkers applying 2,4,5-T under actual use conditions.
These data consist of dermal, inhalation, and urinary concentrations of
2,4,5-T obtained from the field application of 2,4,5-T in forestry and
rice***. Exposures to 2,4,5,-T from other uses and to silvex and TCDD for
all uses were estimated by extrapolation and will be discussed below.
* The term "exposure", as used in this paper, refers to the amount of
chemical absorbed by the body.
** During the past four years, since the initiation of the HPAR process,
the Hazard Evaluation Division has estimated occupational exposures
to many pesticides. In seme cases data on dermal and inhalation
exposure were available for these estimates. In other cases, these
data had not been generated, necessitating extrapolations fran infor-
mation on other pesticides (with similar application techniques) for
purposes of the exposure estimate.
*** Experimental data of the type required for this analysis were found
only for 2,4,5-T. Consequently, exposure to silvex and TCDD was calcu-
lated on the basis of extrapolations fran the 2,4,5-T data as explained
in the text.
F—11
-------
- 2 -
Duration of exposure to specified occupati .:al croups asjz the r.urrer cf
individuals crmprising these groups are critical eissercs in risk assess-
ment. These parameters were estirated frtxn use data fron reference 2
and are summarized in the Appendix (pace 43, et seq.) Occupational exposure
to 2,4,5-T, silvex, and TGDD are estimated for the follcwing uses:
* forestry
* rice
* range and pasture
" rights-of-way
It should be noted that because of infOEraticn gaps, it was necessary to
make a number of assumptions and extrapolations in estimating applicator
exposure to 2,4,5-T, silvex, and TOD. As a result, cur estimates are
subject to a considerable degree cf uncertainty.
Estimation cf Qcsusatlcna1 Zxocsure to 2,4,5-?
v*a are aware cf three studies cn the exposure of applicators to 2,4,5-T
which provide experimental data to be used for exposure assessment. The
rest detailed of these studies is one conducted by Lavy cn forest appli-
cators (?jsf. 14, 15). The data from this study has been analyzed using
a pharmacokinetic mccel in a report by Kansey at al. (Her. 19). Lavy
also conducted a sanewiat abbreviated study of voters applying 2,4,5-T
to rice and forests (3ef. 16). The third study yielding useful exposure
inronnaticn is one by KolmodinrSedman et al. (Ref. 13) in >«iich t>o
professional tractor crews consisting of two persons each were monitored
for 2,4,5-T during and after t>o applications of 2,4,5-T to forests.
F—12
-------
Two ether studies report ad in the literature 2/ provided confirmatory
information cn 2,4,5-T absorption by hunans.-
The information enabling us to estimate the absorption of 2,4,5-7 by cccu-
paticnally exposed individuals is contained in the field study conducted
by Lavy cn rarest'/ applicators (Ser3.14,15). The study was designed to
measure 2,4,5-T exposure to pesticide workers applying this pesticide
in the forest by three different methods:
4 aerial (helicopter)
* ground application by tractor-driven mist blower
" ground application by backpack sprayers
Twenty-one individuals (including two females) participated in this study.
The subjects were encaged in norsal pesticide application activities (e.g.
piloting a. helicopter; driving a tractor and handling pesticide application
equipment; mixing pesticides by dilution, etc.) A oauiiercial product con-
taining 2,4,5-T Ssteron®, was applied at day "0" at a rate of 2 lbs a.a./A*
* Shafik et al. (Ref.24) report an average of 2.4 mg 2,4,5-T/l of urine
in o spray operators engaged in 2,4,5-T application. Mo spray history or
total excretion is given, so it is impossible to calculate total ex-
posure from this experiment. As a matter of feet, the purpose of the
reported study was to develop analytical methodology rather than measure
exposure.
Sispson et al. (Ref.25), in a very brief suimary paper, reported urinary
levels of 2,4,5-T in pesticide applicators handling this herbicide rang-
ing ±ran 0.160 mg/1 to 1.740 mg/1. These incanplete results make it
impossible to calculate total body burden from 2,4,5-T exposure.
* a.e. = acid equivalent.
F—13
-------
4
for tractor-driven rnist 'clover arc. helicopter applications and 1.6 lis./A
in the backpack study. urinalyses for 2,4,5-T (acid) were performed daily
for 7 days including 1 satiple prior to exposure. On the 7th day, the
herbicide application was repeated by the same individuals, and urine
saroles were analyzed as before. Dermal absorption was measured by the
use of cellulcse-backad cauze patches -which were placed according to
directions given by Wolfe, et al. (Hef.31).
Typical attire of individuals participating in the study was long trousers,
shirt (long or short sleeves), cloth sneakers, and leather or field 'coots.
Temperatures during the experiment ranged froit a low of 13°C to a high
of 26 'C. Wind speeds on 5 days of application were recorded at 0 nph while
the wind speed ranged frati 0 -3 irrh on three other days. The experiments
were carried out in South Central Arkansas near Sat Springs, Hairpton,
and Hew >Sarrticello. The terrain there is less hilly than other areas
•where 2,4,5-T and silvex are used, such as that in 'western Washington
and Oregon. It is conceivable that different terrain and -weather
conditions .nay change the exposure pattern of the cccupaticnally exposed
population. Sswever, we krov of no experinsital vork that has been
carried out to investigate these variations. Complete experimental de-
tails ray be found in the Project Carpletion Report (Ref .14) and in the
published japer (J?ef.l5).
According to Pamsey et al. (ref.19), "the total amount of 2,4,5-T excreted
in the urine following exposure represents a miniraun estimate of the amount
F-14
-------
.. .absorbed, sines urinary excretion nay net be complete at termination
or the experiment. ncwever, calculation cf the absorbed cose of 2,4,5-T
based en pharmacokinetic analysis... is not dependent cn total excretion
and can# therefore, provide a acre realistic estimate of the absorbed
dose." rfemsey et al. have chosen maximun estimated deses of 2,4,5-T
obtained from three different kinetic equations (3ef.l9, p. 20).
>fe have used Casey's adjusted data 'cased on Lavy1 s study (iters.14,15) in
estimating .occupational exposure. Results for forestry application of
2,4,5-T are tabulated in the last col una of Table 1, giving the averaos
experimental dose expressed as rag/kg bocy weight/hour. Fran Tables 2-A
and 3-A it rsay be seen that sane individual values varied widely. For
example, the ranges for pilots were 0.C05 - 0.024 rng/kg/hcur and backpack
applicators, 0.009 - 0.036 mg/kgAcur.
Lavy (?efs .14,15) provides experimental data only for forestry uses of
2,4,5-T. Therefore, exposure estimates for uses on rice, rangeiand,
pasture, and rights-of-way were calculated by caiparing application rates,
occupations# and application techniques with the corresponding figures in
forestry use, assuming that exposure '*culd be directly proportional to the
application rate. It was further assumed that the difference in applica-
tion rate was the only variable factor viiich would result in differences
of applicator exposure for each type of occupational group, ctor ex2irole,
the rate used for aerial application of 2,4,5-T in range and pasture is
F—15
-------
- 6 -
1 lb/A (weightad average) and the corresponding rate in Sorest is 2.0
lbs/A (average). Thus, the exposure values for different occupational
groups for ranee and pasture use is estimated by multiplying the experi-
mental value (forestry use) by one-half.*
In order to convert unit exposure values to dose/person/hour, the figure
in the last colonn of Table 1 nay be multiplied by the estimated average
becy 'weight of a male worker, namely TO kg. Table 1 also provides data
on the estimated annual hours of exposure to each occupational croup of
workers and estimated nusfcer of <*crkers in each occupational category.
These nuabers -were derived from the total acreage** treated, found in
Reference 2. The methodologies for arriving at these estimates are
fully explained in the Appendix.
In the Lavy study (Refs.14,15), dermal and inhalation exposures by field
personnel were measured. In addition, urinary 2,4,5-T and ether urine
* Confirmation that absorption, as measured by urinary excretion, is
directly proportional to dese applied has been recently shewn by Franklin,
et al. in a study involving the insecticide azinophesnethyl and orchard
workers (seen to be published) (C.A., Franklin, R.A. Fenske, R. Greenhalgh,
L. Mathieu, H.V. Denlsy, J.T., Laffingwell, and R.C. Spear, A Comparison
of Direct and Indirect :*iethcds of Estimating Dermal Exposure to Guthion
in Orchard vsorkers. Accepted for publication in J. Toxicol. Env.
Health)•
*"* Reference 2 apparently dees net separate 2,4,5-T anr. silvex treatment
for range and pastures, although this is not explicitly stated. Since
under recent usage pattern, silvex represents only ID% (Ref. 35) of the
canbined use of 2,4,5-T and silvex, we feel that our estimates of annual
hours of exposure and nunber of workers in each exposed occupational
group are indeed representative of 2,4,5-T treatment alone without
correcting for the snail percentage of silvex.
F—16
-------
TrsELZ I
Estimated Zxscsure ox Pesti.ci.de Acplicators and ?arrvcrkers to 2,4,5-T
Use Pattern
Exccsed Grcun
Estimated Average
Application No.Exposed Exposure- Exposure^
Rat2^ (li/A) Persons^ Chrs/yr) (.Tg/ko/hr)
RDRESTSf
1.
Aerial
Pilots
2
73
200
0.015
Mixer Readers
2
73-145
800
0.062
FX aggers
2
_ 3
800
0.003
Supervisors
2
3
800
0.004
2.
Grcund ^rcadcast
a. Tractor
Mixer/Loader
2
90-13D
430
0.020
Mistblcwer
Tractor/operator/<*crker 2
90
240
0.013
Supervisor
2
3
430
0.006
b. Backpack
Applicators
1.6
300
800
0.021
Sprayer
Mixer/Supervisor
1.5
3
300
0.005
-WN3Z
MID ?ASTUTE
0.00 S4
1.
Aerial
Pilots
1.0
130
75
Mixer/Loaders
1.0
130-260
100
0.0314
71aggers
1.0
sdo
25
0.0024
2.
Ground Backpack
Applicators
0.6
20,000
80
0.0084
mcs
Aerial
Pilcts
1.0
307
12
0.008"*
Mixer/leader
1.0
307
48
0 .030 4
71aggers
1.0
6 330-9500
0.6
0 .0024
5IGHTS-0F-WPY
1.
Aerial
Pilots
3.0
25
400
0.C604
Mixer/Loaders
8.0
25-50
400
0.2404
2.
Ground
0.0844
a. Selective
Applicators Chard)
6.4
1380
1000
b. Cut Stuns
Applicators Chard)
4.0
60
SO
0.0 S34
c. Mixed 3rush
Arolicators (hand)
6.0
270
660
0.0794
Truck boon Applicators 0.8
178
660
0.0054
A. ftiilTiad
Crsw of ?our
5.(ava)
114
264
0.0664
e. Electric
O.Offl4
Pwer
Applicators (hand)
6.(avg)
400
660
1. See Table 1-A.
2. Reference 19. ' Calculated dose levels; received by c?A on February 14, 1979:
% ISP [30,000/26]; See also Table 2-A for raw data.
3. ( ) indicates that the nurber of individuals cannot be estimated.
4. These values were extrapolated as explained in the text.
F-17
-------
ccrrcnents were analyzed. 3v lavy' 3 calculations, very peer correlation
existed between derrnal exposure to 2,4,5-T, as measured by 2,4,5-T
analyses of the body patches, and the amounts excreted in the urine.*
One explanation for the lack of correlation micht be the fact that the
dermal exposure patches were net always placed in areas of highest
potential exposure, e.g. the hands of mixer-loaders. Thus, the exposure
derived iron dermal patches flight be expectad to be too low, and,
consequently, urinary excretion values would be more realistic.
In the second Lavy 2,4,5-T-exposure study (Sfif.lS), only dermal and no
urinary analyses for 2,4,5-7 were performed. Hcwever, only results fran
urinary excretion experiments were utilized by us for exposure estimates
5ar the following reasons:
1. The pharmacokinetic behavior of 2,4,5-T has been described in
manuals, including nan.
2. Analysis of 2,4,5-T in the urine is a more direct measurement a
2,4,5-T absorption than the use of dermal patches.
Thus, in cur exposure estimates for 2,4,5-T we nave utilized exclusively
urinary excretion data derived frcra Lavy's field study (?e£s.l4,15), tra
posed by pharmacokinetic calculations by 52irsey, et al. (Ref .19).
While we have relied heavily on Lavy's field studies and the pharmaco-
kinetic derivations by Hartsey, et al., based on the same studies, it is
* Exposure through inhalation was roich lower than that from dermal
contact and, therefore, was not included by Lavy in the correlation
test.
F—18
-------
- 9 -
prudent to review these experimental studies and kinetic derivations in
•greater detail. Coring the cross examination testimony of Dr. Niscert,
several experimental deficiencies in the Lavy studies (Sefs.14,15) were
discussed and included apparently incanplete or variable urine collect-
ion and failure to correct urine volumes according to creatinine levels.
T^.e Agency is presently engaged in an independent analysis of the pharma-
cokinetic treatment of Lavy' s field data. After this review has teen
coapletad, the exposure estimates may have to be revised appropriately.
KSEMODJN-fiEDtftN STCIEY
Pecerrtly, another study fran Sweden cn the exposure of tvo tractor crews
to 2.4,5-T has cone to cur attention (Hef .13) • The study consisted cf
the surveillance cf tvo <*ork crews of 2 individuals each. They applied a
mixture of pheiscxy herbicides in a Sorest for cue «ork week and 2-4 hrs/
cay spraying usirn a Gullvik* Forest Tractor equipped with a fan
sprayer. Blood and urine samples were analyzed before application of
the herbicide, cnce cr twice during the application period, and at. 12, 24,
and 36 hours after the last application. Urine samples '*ere net taken
at regular intervals during the study, rreking it less reliable for the
estimation of total exposure than Lavy* 3 study (Befs.14,15). Lavy showed
that even a 6 day period is insufficient for complete elimination of 2,4,5-T
Sran the body. Thus, it is quite certain that Kolrodin's results are cn
* The sake of the Swedish tractor is mentioned because the difference in
exposure between Swedish and U.S. '*crkars may be due to equipment differences.
F—19
-------
- 10 -
the lew sida, sines the list urine sample \%as taken enly 1.5 days after the
last application cf 2,4,5-T. bfeverthelsss, >
-------
The exposure by Crew II in rQHsodin's scucv appears to be 3 to 6 tiaes
higher than that, cf Crew I. The reaacn for this nay pc-ssiily he explained
by the different ucrking axditicris during pesticide application by
Crews I and II. Crsw I changed -*ork clothes each evening and their tractor
had a partially protected seat. Cn the other hand, the mixer/>as "row leader." (A per sen
irarks the row to direct tractor-driver). "When the tractor
turned, he could get spray liquid an his body. Tractor driver
could also receive spray on his body, since tractor had a
canpletely open seat.
* Reference 13.
** Based on 1.5 L urine/day? see T&ble 2 for tabulations.
*** Average 3x5 = 15 hrs/veek spray time.
F-21
-------
- 12 -
Table 4 is a cnrcariscn of ti*.e results fran T2blas 1 anc 3
T£ble 4
Ccmpariscn of Lavy and RolTEden-^eaian Studies
Lavy Study (Refs.14,15) KqIhtV-'t Study (Hef.13)
1 AccLlc. T I ' Appilc. I
Av. Dose Rats Av. Dose (mg/kg/hr) Sate
Occupation (nc/kn/hr) (lbs/A) Crew I Crew H (lbs/A)
y±xer/loader 0.020 2 0.01 0.03 0.66
(ground)
Tracker Driver 0.013 2 0.01 0.06 0.66
3y nultiplying the exposure values obtained by KaLraodin by a factor of 3
(to adjust far the lower applicarJ.cn rata in Rslrodin' a study), the tractor
driver of Grew IX would appear to have a significantly higher exposure (by
a factor of approximately 14) than the corresponding U.S. '^crkers in the
Lavy studies.
If the conditions of described by iNOlrrocin ara typical of these encount-
ered in the Lhited States, it nay be prudent to perform a quantitative
risk assessnerrt using the higher exposure figures.
EXPOSURE TO SILVEX AND TCDD
We could find no reports, either published or unpublished) on the exposure
of '*crkers in the field to silvex or TODD. Therefore, in order to estirrata
occupational exposure to these chenicals, have assumed the following:
1. Silvex exposure is the same as 2,4,5-T exposure, wherever and
•whenever the use pattam for silvex and 2,4,5-T are similar or
identical. We believe that the chemical behavior of silvex and
2,4,5-T is sufficiently similar to justify this assumption.
F-22
-------
- 13 -
2. We are not aware of any induration regarding the rata of cerral
absorption by ran of TCDD relative to 2,4,5-T. In the absence of
this infcrraticn, we are assuming fcr the purpose of estimating
exposure that TCDD and 2,4,5-7 are absorbed at the same rate.*
3. TCDD exposure resulting fran 2,4,5-T application may be estimated
by applying concentration factors obtained by direct analysis of
2,4,5-T fbmolaticris. Lavy reported that TCDD vas present in
the Zstercn® product used in his study (Sefs. 14,15) at a level
of 0.04 pnn (4 x 10-3). Manufacturer's voluntary specifications
of current 2,4,5-T production TCDD concentrations of 0.1 pan
or less.** Thus, TCDD exposure may be estimated by multiplying
2,4,5-T exposure Ssr each applicator group by a factor ranging
from 4 x 10"8 to 1 x 10-7.***
4. Estiirates for rmrber of exposed individuals and annual hours cf
exposure due to silvex use can be trade by using conversion
factors based on ratios of 2,4,5-T treated acres to silvex treated
acres for different uses as shewn in Table 5; these ratios range
fran 1/10 to 1/1000.
* Another assumption is that the concentration of TCDD relative to
2,4,5-T does net diange fran the tine it is formulated until it is
deposited on the skin"of the occupationsUy exposed personnel.
** There are acme mamfacturers ^ho claim that their 2,4,5-T products
contain 0.02 pun or even less dioxin.
*** Since the concentrations of TCDD in 2,4,5-T and silvex are approx-
imately the same, the same factors may be used in estimating ex-
posure to TCDD resulting fran silvex applicatiais. The same nurcber
of persons exposed to 2,4,5-T or silvex are, therefore, assumed to be
exposed to TCDD. Moreover, the annual hours of exposure of a person
to 2,4,5-T and/or silvex are assured to be the same as his annual
hours of exposure to TCDD.
F-23
-------
- 14
Table 5
Comparison of Relative Pates of Usage of 2,4,5-? and Silvex
Uses
2,4,5-T:Silvex Ratio
Hangeiand/pasture®
Forestry (Ref.2)
Rics£
Rights-of-way53
10:1
100:1
1000:1
accx. 10:1
a. Reference 35.
b. Reference 17.
EXPOSURE ESTIMATE - DOSSED CSS GF 2.4,5-T AND SILVEX
The exposure estimates sinner i zed in Table 1 are based on recent pre-
suspension use volume data for 2,4,5-T and silvex. Tor all registered
uses, only a relatively Iw percentage of all potential acreage is actually
treated with these two herbicides. If the acreage treated were to
increase, the total nuxtber of exposure hoars * vculd increase proportionately.
It is extramely unlikely that one hundred percent of the acreage which cculd
"be treated annually with 2,4,5-T or silvex consistent with the labeling wruld in
feet be treated. ** Scwever, because the increase in annual exposure hours
resulting front such -raximn possible use provides an upper limit on the total
nmifcer of annual exposure hours, we are estimating the increase in total rxsrber
of exposure boiurs which wculd result £rcm such maxiraum possible use.
Of the approximately one billion acres of pasture and rare eland in the
U.S., enly 0.33% is treated with either 2,4,5-T or silvex- If all pasture
and rangeland were treatsd annually,** the total annual exposure hours for
*/ Total nuiroer of exposure hours is defined as the product of total
number of <*crkers in a particular occupational croup tines the annual
niaticer of hcurs per worker for this use.
**/ In fact, only 26% of total rangeland and pasture land has undesirable
plants susceptible to treatment by 2,4,5-T or silvex. (Ref. 17)
F-24
-------
-15-
each type of applicator wcul d increase by a factor cf 300 ever cur estimate
of total nzrfcer cf anmal exposure bears estimated to cccur at the tine of
suspension.
s-tTTi-nar projections for increase in total rxmber of exposure hours to
either 2,4,5-T, si!vex, or TCDD might be nade if the extent of use of
2,4,5-T or siivex approached the maximum possible market for caimsrcial
forest land (factor = 500), rice land (factor of 10), or rights-of-way
(factor = 2CO) (ref. 17).
S1W3X CF OCCUPATIONAL EXPOSURE
Based on the Lavy study, which measured 2,4,5-T levels in the urine of
applicators who applied 2,4,5-T, as veil as on a pharmacokinetic analysis
by Sansey of these experimental data, we have estiirated applicator exposure
to 2,4,5-T, siivex and TCDD resulting frcn a asrber cf uses cf 2,4,5-T
and siivex. These estimates are provided in Table 1.
Because of several factors, the exposure estimates made in this document
are subject to considerable uncertainty. Sane of the more important factors
are:
1. It is possible that the degree of care to avoid exposure which
was exercised by the applicators in the Lavy study may net be typical
of that used in routine 2,4,5-T or siivex applications.
2. The applications in the Lavy study were conducted under essentially
windless conditions and on relatively level terrain. At higher
wind velocities or different terrain (roll-. ; hills or mountains)
exposure rates may be quite different
3. In estimating TCDD exposure, it wqs necessary to extrapolate
from cfeta - - 2,4,5-T exposure. In so doirn, it was assumed that
TCDD was i. ?rbed by the body with an efficiency equal to that
of 2,4,5-T. In feet, TCDD may be absorbed at rates considerably
different than those of 2,4,5-T.
F-25
-------
- 16 -
4. The Lavy study may have had certain experimental deficiencies,
including incomplete or variable urine collections.
The Swedish study (ref.13) indicated that under certain conditions, applicator
exposure, at least with respect to tractor drivers, may be considerably
higher than that estimated frcm data generated in the Lavy study. Correct-
ing for differences in application rates, the exposure rate of one of the
tractor—drivers in the Swedish study 'was about 14 times higher than
the exposure rate measured in his American counterpart (0.18 vs. 0.013
mg/kg/hr). Thus, if U.S. field conditions were comparable to those
encountered in the Swedish study, it might be prudent to estimate risk
on the basis of higher levels of exposure than those found in the one
U.S. study.
F-26
-------
17 -
ESTIMATES CP rXMAN IXPCSUHE TO SEE? AND MUX
CarEaMZSiaSSD WITH TC~D
3ACKGOJND
The estimates of human exposure to TCDD fran contaminated beef and milk
•which are developed in this document are based on a twc-cart study (here-
after called chase one and phase two, respectively) initiated under the
E±aein Isplenerrtation Plan in 1975- These studies -v^ere designed to deter-
mine possible residues of TCDD in the fat and liver3 of cattle grazing on
range land treated with 2,4,5-T (ref.26).
Anirals fren selected fams in Missouri, Kansas, Texas and Cklahcma were
taken to carmercial slaughter houses, where samples of fat and liver -*»ere
collected. Alcng with historical information, these samples were forward-
ed to the Toxicant Analysis Canter, at 3ay St. Louis, Mississippi, fcr
extraction, cleanup, and encoding, preparatory to chemical analysis for
tetrachlorcdibenzo-p-o sanples in Ndvenber/December, 1975, fran cattle grazing on forage
treated with 2,4,5-T in May, 1974 and May,. 1975, respectively. In both
parts of the study, the application rates varied fran farm to fam, rang-
ing fran 1/2 to 4 lb 2,4,5-T active ingredient/A (3 lb/A maxi.um applic-
ation rate in phase two). In addition, the percentage of acreage actu-
ally treated varied fran 20% to 100%.
^riciltural practices appear to have been about the same as- those in
use today. Herbicide (2,4,5-T) was aerially applied (with occasional
ground spot-treatnant) to control undesirable vegetation cn grazing
F-27
-------
- 18 -
land. Cattle frcni the selected farms were routinely sent directly to
slaughter (ref.26), rather than to supplenentary feed lots. This practice
proved useful for the subsequent TCDD analyses, since it avoided dilution
of TCDD residues by the deposition of additional interstitial fat. Addit-
ional farms were selected which had no kncwn recent treatment of chlorin-
ated phenoxy herbicides. Adipose and liver samples from animals collected
from this group of farms served as experimental controls.
The decision to search for TCDD in adipose tissues was based on the
presumption .that TCDD, as a lipophilic chemical, would be preferentially
distributed to lipid-rich tissues and could most easily be detected
there. Data on residues confirmed in such tissues could then be used to
estimate levels in other tissues (such as milk or meat) which consist
only partially of fat (5% and 15%, respectively). In fact, this may be
the only technique available to estimate such residues, since TCDD
residues in tissues such as these might well be below the limit of detec-
tion using current technology, and thus cannot be directly measured.
Because the beef studies were conducted as "field" monitoring studies,
they could not be controlled as well as "laboratory" experiments. Con-
sequently, a number of uncertainties must be noted, about which different
assumptions could yield alternative interpretations of the data.
Itoss (ref.22) observed that virtually all of the positive sanples in phase
one (3 lb/A group) came from several adjacent Missouri farms. This is
not surprising, especially in light of the high application rates used
on these farms (3 lbs/A).
F-28
-------
- 19 -
There is also the possibility that the dioxin residues in these fat sanples
might not "be representative of the residues in all cattle allowed to graze
on 2,4,5-T-treated land. Since this study contains the most reliable
field data currently available, however, it is assumed that these residues
are representative of the residues which would result frcm typical
2,4,5-T-use on range land in the United States. Further, it is reasonable
to extend the conclusions regarding 2,4,5-T use to the use of silvex on
pasture land, since the use practices for the tv*o herbicides are very
similar, and both contain ccnparable amounts of TCDD.
toother uncertainty concerns the amount of treated vegetation actually
ingested by the exposed cattle. Since the percentage of 2,4,5-T-treated
grazing lands varied widely frcm farm to farm (frcm 20% to 100%), cattle
might have had the opportunity of ingesting differing percentages of
both treated and untreated vegetation, depending upon the grazing acreage
in which they were allowed to feed. Since the exact situation on each
farm is unkncwn, it is assumed that 100% of the diet of these cattle
consisted of contaminated vegetation, that is, cattle fed selectively
on the treated areas, rather than grazed indiscriminately, and consumed
no supplementary (uncontaminated) feed or forage. This assumption was
made because there appears to be a better correlation between average
application rate and average residue levels when it is assimed that
animals grazed solely on treated vegetation, rather than on both treated
and untreated vegetation.
It is- therefore assumed that the dietary intake of forage in the cattle
from this study consisted of only treated forage. If these cattle.actually
ingested significant quantities of forage frcm untreated areas, or supple-
mented their diets with uncontaminated feed or grain, then it is highly
F-29
-------
-20-
probable that any cattle which might have actually consumed ;xclusively
contaminated vegetation wsuld have had even higher TCDD residues in
their adipose tissues than those measured in this study.
SUMMARY OF RESULTS
Collaborating analytical laboratories included an EPA contractor (North-
rup Laboratory), Wright State University, the Dew Chemical Company and
Harvard University. In phase two of the study, only Wright State Univer-
sity has analyzed the samples thus far. A second laboratory is scheduled
to perform confirmatory analyses.
The results of phase one which were generated by high resolution gas
chranatography/mass spectroscopy are summarized in the Table A-4. Resi-
dues were detected in 7 of 11 samples, ranging from not detected
(ND), (limits of detection ranging from 3 to 24 ppt) to 89 ppt in the 3
lb/A group. The percentage of samples with detectable residues appeared
to decrease with decreasing application rates. The 1/2 lb/A group yielded
no sanples with detectable residues, with limits of detection frcm 2 to
10 ppt.
The 18 control samples (reported in Table 1 of EPA Exhibit 564) were
analyzed by high resolution GC/MS. Sixteen were unequivocally negative
at limits of detection frcm 3 to 23 ppt TCDD. One control was analyzed
five times, and yielded one positive value of 20 ppt (Limit of detection
of 12) and four ND's (limits of detection of 5-14 ppt). The remaining
control '/as analyzed 15 times, and yielded 4 positive values of 3,6,19
and 61 ppt (with limits of detection of 3-10 ppt) and 11 ND's (limits of
detection of 1-20 ppt): Since virtually all analyses were negative, no
F-30
-------
- 21 -
corrections were made to the data summarized in Tables A-4 or A-5.
The preliminary results of phase two are summarized in the Table A-5.
However, these data have been included for canparison only and will
not be incorporated into the dietary estimate because only tv*o samples
were taken frcrn animals grazing on land treated at the highest applica-
tion rate (3 lb./acre). Residues of TCDD found in the adipose tissues
of these cattle ranged fran ND(limits of detection ranging from 7 to 14
ppt) to 34 ppt in the 2 lb/A group, but were all nondetected in the 3/4
lb/A group (with limits of detection of 7-14 ppt). Although of a preli-
minary nature, these results are of the same order of magnitude as those
found in phase one.
ASSIGNED RESIDUE VALUES
Since many of the positive samples tended to occur at levels just above
the limit of detection of current methodology (especially in the cattle
from farms treated at the lower application rates), it is likely that the
samples reported as containing no detectable TCDD actually contained TCDD
residues, at or below the level of detection. Therefore, sane assumptions
were made in order to deal with these kinds of results.
Residues were detected in a majority of the samples in the 3 lb/A group.
This strongly suggests that the ND samples of this set may have contained
residues at, or very close to the limit of detection.
Average residue values were estimated fran the results in Table A-4 by
averaging the test results for each sample, as follows:
a. Only samples vAiich satisfy the criteria used by the Dioxin Moni-
toring Program (Table A-7) have been included in the calculations.
F-31
-------
- 22 -
b. Eata '*hich were reported as ncn-detected (ND) in the 3 lb/A group
have been assigned res due "values" equal to the liniit of detect-
icn (see discussion above).
c. Sainples which vaere reported as having no detectable residues in
the lower application-rate groups have been assigned values equal
to the ratio of the application rate in that group to the 3 lb/A
rate, multiplied by the specified level of detection for that
sample* •
d. When a sample had both positive and ND results, all values were
averaged according to assumptions a. through c., above.
e. Dae to the limited rxnrfcer (2) of samples firan the 4 lb/A group
these data were not included in the dietary exposure estimate.
For the purposes of estimating dietary exposure to residues of TCED in
contaminated beef, it. seens reasonable to assign real values to samples
reported as ND, rather than considering than as zero. Hie justification
for this choice is based on the following reasoning:
A group of highly qualified dioxin analytical chemists established higher
critical standards for TCDD analysis than these applied to many routine
chemical analyses. It is therefore quite possible that sore of the sam-
ples which were classified as ND, might actually have been reported as
positive -were less stringent criteria used.
Further, seven of eleven samples from the 3 lb/A group shewed detectable
ICED residues by the established criteria. This suggests that the three
samples which had no detectable residues did, in fact, have residues
which were close to, but slightly below, the limit of detection. Thus it
* Non-detected (ND) samples are reported as if the TCDD residues were
actually measured (see appendix, Table 3 for modified data) according
to the following scheme:
a. 4 lb/A (net included in this evaluation due to stall sample size)
b. 3 lb/A treatment group, at the limit of detection reported.
c. 2 lb/A treatment group, at 2/3 the Unit of detection reported.
d. 1 lb/A treatment group, at 1/3 the limit of detection reported.
e. 1/2 lb/A treatment group, at 1/6 the limit of detection reported.
F-32
-------
- 23
seems reasonable to assign values equal to the limit of detection to the
"non-detected" samples in this group.
Using the average residue values (which include the assigned positive
values for "ND" test results) we find a strong correlation between the
rate of applied 2,4,5-T (dosage) and the TCDD residues found in the beef
fat. These data are summarized in Table 6. A similar correlation has
been observed by Jensen, et al. (ref .10) in a study where cattle were fed
forage which had been contaminated with various amounts of 2,4,5-T (con-
taining unspecified, but presumably the same, concentration of TCDD). The
observed level of TCDD residues in the adipose tissue appeared to be
directly proportional to the added 2,4,5-T in the daily diet. Based on
Jensen's observations, it seems reasonable to expect that the level of
TCDD in adipose tissues resulting fran ingestion of forage contaminated
with 2,4,5-T or silvex (and consequently TCDD) \*ould be directly proport-
ional to the rate of application of 2,4,5-T or silvex to that forage.
Therefore, it seems reasonable to assign residue values (to samples
which did not have detectable TCDD residues) in seme proportion to
the amount of 2,4,5-T or silvex used on the forage fed to the cattle.
The sensitivity of the method for each particular sanple must also be
taken into account. Since about 70% of the samples frcm the 3 lb/A rate
showed measurable residues, all ND samples were reported as positive at
the level of sensitivity. Samples frcm fields treated at lower rates
were scaled down proportionally (see footnote on page 22).
Finally, Young (ref.32), Zweig (ref.33), and others have observed that
the development of increasingly sensitive methods of analysis have
permitted detection of residues at continually lcwer levels, where few
F-33
-------
- 24 -
residues had been detected by earlier techniques. It may thus be improper
to simply assign a value of "zero" to these ncn-detectsd residues, since
such an assizer.ion wculd terse to underest±nate residues -which .-night be
present, but undetected. Hie phase cne data shown in Table A-4, there-
fore, have been adjusted accordingly (Table A-6) and averaged and are
sunstarized in Tkble 6.
Table 6
Summary cf Adjusted TCDD Residues in Phase Cne Adipose Tissue
%uce of Overall
Application Residues* Mean**
Pate (lb/A) (per: TCDD) (pet TCDD)
3 6.0-46.5 17.1
2 2.0-37.7 10.9
1 2.0-10.0 4.2
1/2 0.3- 1.7 1.2
* Vialues include adjusted UD samples and are averaged for each animal.
** See T^ble A—6.
The adiprse samples of the phase two 2 lb/A group contained sufficient
positive results to compare them with these of phase cne. When the
phase two data are handled in the same way as those in chase cne, the
carputed mean of 8.63 ppt TCDD compares well with the result in the same
group fran phase cne (10.9 ppt TCDD)
PCS5I3LE MEGLANIS4 rCR TCDD IMSSTICN
A possible mechanisn of ingestion of TCDD fran 2,4,5-T- or sil vex-created
range land or pastures may be suggested by a review of several studies.
Mayland et.al. (ref.IS), measured the amount of soil ingested by cattle
grazing cn rang eland. Hs feund that cattle typically ingested up to cne
kilogram of soil per day, possibly fran cropping vegetation close to
the soil surface, tfcder a variety of conditions, TCDD persists at
F-34
-------
measurable levels for long periods of time ("half-life of 1 year or
longer), at or near the soil surface, as shown by Young (ref.32), and
Kearney (ref.ll) and others. These observations suggest that roots
(sub-thatch) and upper layers of soil in range land and pastures treated
with either 2,4,5-T or silvex may constitute a significant reservoir
for the TCEO consumed by grazing animals. Thus TCOD residues, either in
soil or on vegetation, may account for residues observed to occur in
beef animals grazing on 2,4,5-T -treated range land and pasture.
DIETARY INTAKE OF CONTAMINATED BEEF
The reported usage of 2,4,5-T on range land and pasture (ref.2) varies
between 1/4 and 2 lb/A, depending on the area of the country, the
target vegetation, and other parameters. Rangeland uses of 2,4,5-T are
sunmarized in Table 7. In phase one of the beef study, application of
2,4,5-T on sane of the farms studied exceeded these rates (up to 4 lb/A).
This raises the possibility that sane grazing land is treated at levels
considerably higher than the levels reported in Reference 2.
Table 7
Suirmary of 2,4,5-^T-Treated Rangeland*
Method of Target Application Acres Treated
Application Vegetation Rate (lb/A) Per Year
Aerial Mesquite/shinnery oak 1 137,000
Aerial Mesquite/shinnery oak 1/2 330,000
Aerial Mesquite/shinnery oak 1/4 400,000
Aerial Oak Savannah 2 541,000
'Ground Mesquite 1/2 75,000
Ground Oak Savannah 2 60,000
Total Rangeland Treated Annually 1,713,000
* Data from Tables 17 and 18, reference 2.
Using the data from Table 7, the weighted mean application rate was
calculated and found to be 1 lb/A. This represents an "average" use
F-35
-------
- 26 -
of 2,4,5-T on range land. From Table 6, it is seen that a typical TCDD
residue level of about 4.2 ppt in contaminated beef adipose tissues is
estimated to result from a 1-lb/A treatment of range land with 2,4,5-T
or silvex.
DIETARY ESTIMATE
The initial dietary estimate will deal with a group of persons whose
dietary beef intake consists solely of contaminated beef. The maximum
number of persons in this group can be estimated frcm the data provided
by Lee (ref.17), as follows:
a. Total U.S. Beef Production in 1979 was 21.5 billion lbs.
b. Total U.S. population in 1979 was approximately 220 million people.
c. Therefore, per capita beef consumption in 1979 would be about...
21,500,000,000 lbs/vear = 100 lbs/person/year*.
220,000,000 persons
d. Total beef produced on range and aerially-treated pastures treated
with 2,4,5-T and silvex in any year has been estimated to be
from 30 to 137 million pounds dressed weight.**
* Clearly this "average" figure does not take into account persons who
consume no beef whatever, or persons who normally eat larger-than-
average amounts of beef.
** Of the total contaminated beef produced on range and aerially-treated
pastureland, the majority of the beef, by far, canes frcm rangeland.
Specifically, total beef produced on range areas treated with 2,4,5-T
and silvex in one year is estimated to be 73 to 120 million pounds (ref.17).
Total beef produced on pasture aerially treated with 2,4,5-T or silvex
in a year is estimated to be frcm 7 to 17 million pounds.
Approximately 900,000 acres of pasture receive sane 2,4,5-T or silvex
treatment frcm ground application. Since these acres are expected to
receive primarily spot treatments, no estimates could be made of the
amount of beef which would be contaminated by grazing on these areas.
Thus, contaminated beef frcm ground-treated pasture land was not
factored into the exposure analysis. However the amount of such
beef is expected to be a minor portion of all contaminated beef coming
frcm treated grazing lands.
Since approximately 9 times as much grazing land is treated with
2,4,5-T as with silvex, it is estimated that about 90% of the beef
contaminated with TCDD cones frcm 2,4,5-T treated grazing land.
F-36
-------
-21 -
e. The oercentage of heme slaughter beef is estimated -,o be about
0.9%*.
f. Therefore, total beef consumed frcm home slaughter, raised on
tr .ted land is...
80-137 million lbs. x 0 .009 = 720,000 to 1,230,000 lbs.
g. Since about 720,000 to 1,230,000 pounds of contaminated beef could
be consumed at an average rate of 100 lbs/person/year, it is
estimated that between 7,200 and 12,300 persons might consume
only contaminated beef (containing 4.2 ppt TCDD in the adipose
t i ssues).
Beef, consumed at 100 lbs/person/year is equivalent to 124 grams/person/
day* (approximately 1/3 pound). Assuming beef to contain about 15%
(Ref. 18b) fat, a typical daily intake vrould be about 19 grams of conta-
minated fat. Based on 4.2 ppt of TCDD residues in beef adipose tissue
resulting frcm the application of 1-lb/A 2,4,5-T to rangeland, an average
intake of 80 pa TCDD/person/day would be predicted, assuning all beef to
be contaminated. This number represents the dietary intake by a population
whose total beef intake was contaminated (heme slaughter). Exposure to
local populations would be expected to be proportionally higher, if
higher rates of application were used (labels permit treatment up to
4 lb/acre).
The average intake of TCDD by local populations consuming TCDD-
contaminated beef would be expected to be about 80 pg/person/day during
the first year following application of 2,4,5-T or silvex to grazing
lands at 1 lb/A. Reference 2 reports retreatment no more frequently than
once every 5 years. Since it is known that TCDD declines in soil with
a half-life of at least one year (Ref. 11, 32) cattle could reasonably
* Based on data provided by Schmitt (ref.23), dietary intake of beef,
liver and veal would be about 112 grams/day, which agrees well with
Lee's data (ref.17), which is based on more recant information.
F-37
-------
- 27a -
be expected to ingest decreasing quantities of TCDD fran one herbicide
application to the next.
It seems reasonable to assume that residues of TCDD in adipose tissues
will be directly proportional to the amount ingested. Thus, the decline
of tissue TCEO-levels could roughly parallel the decline of soil residues.
If we assune a "typical" half-life of 1 year for TCDD in soil, the
average intake of TCDD during each of the 5 years between applications
may be estimated (Table 7a).
Table 7a
Estimated TCDD Intake
Intra Anolication Period
TCDD Intake
Year (PG /Person/Dav)
1st 90
2nd 40
3rd 20
4th 10
5th 5
6th 80**
* Assumes on application of 2,4,5-T or silvex at 1 lb/A, at a frequency
of every five years.
** Reapplication.
It should be etphasized that seme farmers may choose to treat especially
stubborn (herbicide-resistant) weeds in successive years, or at rates
higher than 1 lb/A. Herbicide use in the state of Missouri in the beef
(phase one) study, for example, was 3 lb/A, with reapplication in tv*o
successive years. Thus, seme local populations may become exposed to
TCDD levels significantly higher than those estimated earlier for "average"
1 lb/A applications at "5 year" intervals.
F-38
-------
- 27b -
The following is an estimate >f the dietary intake by the U.S. peculation
at large of TOD frati contaminated beef. As shewn under "d" above, the
estimated volume of beef from animals grazing on 2,4,5-T or silvex-treated
areas ranges frcm 80 to 137 million pounds dressed weight. The total U.S.
production of beef is estimated to be 21.4 billion pounds. Thus, the
total amount of contaminated beef produced in any one year is estimated
to range from 0.4 to 0.6% of the total U.S. beef production*. The
dietary exposure of the general population to TCDD frcm contaminated
beef, therefore, is estimated to range frcm 0.3 to 0.5 pg TCDD/day.
It should be rioted that only a very small percentage of grazing land is
treated annually with 2,4,5-T or silvex. If the use of these herbicides
were to increase, residues in grazing cattle might reasonably be expected
to increase proportionately.**
INTAKE OF TCDD FFOM CTOTAMTNATED MILK
We have no information on whether or not it is valid to estimate possible
residues of TCDD in the milk of dairy cattle, extrapolated frcm the TCDD
residues in the adipose tissues of beef cattle. It is unclear whether
* These estimates are based on the amount of beef cattle produced
on grazing land treated with 2,4,5-T or silvex during one
calendar year. However, if the assumption that cattle acquire TCDD
residues by ingestion of contaminated soil is correct, then the real
possibility exists that cattle could continually ingest quantities
of TCDD over many years. Thus, the total amount of contaminated beef
produced annually might be considerably higher than these figures.
This is especially true in light of the very long half life of TODD
in soil and lew soil mobility which would tend to ensure continued
dosing of grazing cattle for a number of years follaving herbicide
application.
** If 2,4,5-T or silvex were to be used on all grazing land, to the max-
imum extent permitted by the label, (which is highly unlikely) intake
of TCDD could be expected to increase to 60 - 100 pg/day (200 x 0.3
to 200 x 0.5 pg TODD/day).
F-39
-------
- 28 -
differences in agricultural practices on pasture (as opposed to range-
land) , such as the presuspensicn label-directed post-application animal
reentry interval (for seme silvex products and 6 weeks for 2,4,5-T)
would lead to significantly reduced ingestion of TCDD by the grazing
cattle.
As noted earlier in the discussion of possible mechanises of ingestion
of TODD from 2,4,5-T- or silvex-treated range land or pastures, TCDD
seems to persist in the root area (subthatch) and/or upper layers of
soil for long periods of time. The persistence would be far in excess
of the cost-application intervals mandated by the labels. Therefore, it
seems reasonable to assume that dairy cattle, feeding on 2,4,5-T- or
silvex-treated pasture, might accumulate residues of TCDD in their
tissues very nearly as high as those in beef cattle grazing on similarly
treated range land.
We have no information on vvhich to base an estirate of the number of
persons in the general population which might consume TCDD-contaminated
dairy products. We can only speculate about the level of dietary intake,
giving two hypothetical cases.
a. A person whose dietary intake of dairy products consists solely
of contaminated dairy products frcm farms on which cows consume
vegetation treated with either 2,4,5-T or silvex, at 1-lb/A.
b. The general population, if all rangeland and pasture were to be
treated with either 2,4,5-T or silvex (to the maximum extent per-
mitted by the label) at 1-lb/A .
Assuming that dairy cows have residues canparable to those in beef cattle,
and assigning milk to contain about 4% fat, then milk frcm contaminated
F-40
-------
— 4 7 —
caftle could be expected to contain about 0.17 ppt TCDD*. If the typical
dietary intake of dairy products** consists entirely of TCDD-contaminated
milk and milk products (containing about 43 grams of fat), then the level
of TCDD \*ould then be 190 pg TCDD/day from these dairy products. Exposure
to local populations would be expected to be proportionally higher, if
higher application rates were used***.
DISCUSSION AND 03NCLUSICNS
Assuming recent usage patterns for 2,4,5-T and silvex, the general popu-
lation would be expected to consume approximately 0.5 pg TCDD/day frcm
contaminated beef. Local populations (i.e. heme slaughterers) whose
dietary consumption of beef consists of only contaminated beef are estim-
ated to consume 80 pg TCDD/day, on the average. Although difficult
to identify, there may be local populations whose dietary consumption
of milk and dairy products consists only of contaminated milk and dairy
products. This group is estimated to consune up to about 200 pg TCDD/day.
There might, theoretically, be local populations consuming only contamin-
ated beef and only contaminated milk and dairy products. They are estim-
ated to consume about 300 pg TCDD/day. Levels of 300 pg TCDD/day might
be reached for the general population if all range land and/or all past-
ures were treated with 2,4,5-T or silvex. However, this scenario is
highly unlikely.
* 4.2 ppt TCDD (Table A) x 0.04 = 0.17 ppt TCDD
** Schmitt (ref.23) estimates the daily intake of Milk and Dairy Products
to be about 550 grams, equivalent to about 43 gm of fat. See Table
5-A for carputation.
*** The label permits application of 2,4,5-T at rates up to 4 lbj/A.
F-41
-------
- 30 -
Sane assumptions have been made in this analysis which might tend to
raise or lover the dietary exposure to dioxin from consumption of TCDD-
contaminated beef. For example, in the survey of TCDD residues in beef fat
(the phase one study) the levels of TCDD were found to correlate well
with the rates at which 2,4,5-T was applied to the grazing land. In
order to arrive at this correlation, it was necessary to make the assump-
tion that the animals had consumed only contaminated feed. If, hcwever,
this assumption is incorrect, and the animals had grazed on both trsated
and untreated land, then the diet airy intake of TCDD would be expected to
be even higher than we have estimated.
The evaluation of TODD exposure, resulting fran ingestion of contaminated
beef and dairy products, did not take into account t\so other factors,
neither of which vould be expected to affect the exposure estimates by
an order of magnitude. These factors could conceivably cancel out one
another.
The first factor is the persistence of TCDD in soil, frcm one year to
the next (with a half-life of one year or more). The exposure estimates
for the general population were based on the amount of beef produced on
ranee and pasture treated with herbicide within a single calendar year.
Since it has been shewn that cattle may ingest considerable amounts of
soil and/or subthatch material, and that TCDD has a half-life in soil of
one year or longer, it is possible that cattle could continue to ingest
quantities of TCDD many years after treatment of range land or pasture
vegetation with 2,4,5-T or silvex. Thus, the total amount of beef, milk
and dairy products which may be contaminated with TCDD could conceivably
exceed the levels estimated in this analysis. By not taking this possibility
F-42
-------
into account, one would tend to underestimate the exposure to the general
copulation.
Another factor which should be noted is the ccznman practice*
of fattening calves and yearlings in feeding lots prior to slaughter.
Ingestion of presumably uncontaminated forage and/or grain might tend
to dilute residues of TODD in the adipose tissues. The exact pharma-
cokinetic mechanisns which apply here are unknown. Since none of the
animals in this study were sent to feed lots, their residues were not
diluted by this subsequent feeding. Not taking this factor into account
would tend to overestimate the exposure.
* We are aware of the fact that a significant nimber of beef cattle
avoid the feedlots and are sent directly to slaughter. Therefore,
dioxin in the meat of these animals would not become diluted by
addition of non-contaminated fat. An example of this practice is
a local product, Giant Lean. We do not have any data on hand indicating
the percentage of beef cattle which are in this category.
F-43
-------
- 32 -
DISTAIg 5CPCSUBS TO 2,4,5-T SILVEX
Tliis section of the dccunent represents the analysis cr the possible
intake of 2,4,5-T and silvex by the general U.S. population fran residues
of 2,4,5-T and silvex in food. In estimating dietary exposure to silvex
in treated crocs, a range of exposures has been provided, where appropriate.
Actual residue data, if available, have been used to estimate realistic
dietary exposure levels. Since silvex residues may legally occur up to
the tolerance (or interim tolerance) level* in certain feeds, it has
also been assuned (for purposes of making a conservative estimate),
that silvex may be present at this higher level. Where adequate residue
data axe lacking, exposure estimates were made on the assuroticn that
residues might be present at tol'erance levels, but no range of possible
residue levels could be given.
The only food crop on which 2,4,5-T is used directly is rice. Since no
tolerance for 2,4,5-T residues cn rice has been established, the tolerance
level could rjot be utilized in estimating exposure. As explained ceicw, an
estimate of 2,4,5-T residues in rice has been based cn extrapolation fron
silvex data.
Ranges of exposure of the general population, to 2,4,5-T and silvex residues
in feed are estimated by taking into account the percentage of each food
crop sprayed annually with these herbicides in recent years and the feed
factor (FSf. 23) which quantifies the percentage of the daily diet renre-
* The food tolerance represents the upper legally permissible residue
level of a pesticide and/cr metabolites rsnaining in or cn the crop at
harvest time. The only final tolerance for silvex is the one established
on pears (post-harvest) at 0.05 pun (40 CFR 180.340). There are interim
tolerances for silvex of 0.1 pnn in sugar cane, plans (prunes), apples,
and rice (40 CFR ISO.319). There are no feed tolerances for 2,4,5-T.
F-44
-------
sented by this particular item of food. The food factor is ba rf on the
average food intake of 1.5 kg per day by an 18-year old U.S. ir. ie.
If the percentage of food crops sprayed were to increase, the exposure of
the general population to 2,4,5-T and si1vex residues in these crops would
increase proportionately. For purposes of setting an upper limit, estimates
of potential exposure have also been made for the hypothetical situation
in which 2,4,5-T and silvex are used to the permissible maximum acreage
on food crop, consistent with the pesticide labeling. Although it seems
unlikely that 2,4,5-T and silvex would be used to the maximum extent
permissible, unforeseeable factors could markedly change current usage
patterns so that at least an intermediate exposure might occur.
Exposure to residues of silvex and 2,4,5-T in secondary sources (meat,
milk, and eggs) may occur as a result of livestock feeding on treated
grasslands and rice by-products such as hay, straw, and hulls and poultry
feeding on rice by-products. In addition, exposure to silvex and 2,4,5-T
residues in fish may occur as a result of run-off fir cm rice fields treated
with these herbicides. A quantitative estimate of exposure to 2,4,5-T
and silvex residues in milk and other dairy products has been made for
special situations. Although a quantitative evaluation of the exposure
to silvex and 2,4,5-T residues via other secondary sources cannot be
made at this time, a qualitative discussion follavs in a later section.
SILVEX RESIDUES IN THE HUMAN DIET
The results of the dietary analysis for silvex are given in Tables 8 and
9. Table 8 gives a range for the dietary intake by the general population
estimated frcm residues actually found on the treated crops (where known),
F-45
-------
- 34 -
and frcm tolerance levels. The percent canmodity annually treated (pre-
suspension rate) and the food factors were used in calculating total
dietary burden.
TABLE 8
Estimated Dietarv Exoosure of Silvex
Crop
Possible
Residues
(ppb)
Percent
Crcp
treated
FOod
Factor
(%)
Rate of
Ingestion^
(ug/day)
Dietary
Exposure^
(ng/kg bw/day)
Rice
125 - ]006
0.101'7
0.552
0 .001-0 .008
0.004 - 0.011
Sugar
1006
2.6-4.6
3.642
0.141-0 .251
2.028 - 3.588
Plums
1006
5.4l
0.132
0.011
0.150
Apples
429- 1006
11.01
2.542
0.176-0.419
2.515 - 5.987
Totcil: 4.7 - 9.7 ng/kg/day
1 Ref .17.
2 Ref .23.
3 Based on 1.5kg - daily diet
4 Based on average weight of 70kg per individual
5 In reference 7, 21 samples were at or belcw 10 ppb, the limit of
detection. The Agency made the conservative estimate that residues
could be present up to 10 ppb, and therefore, substituted ID ppb for these 21
samples which in turn were averaged with the 6 positive estimates of
10, 20, 10, 30 and 30 pb. The average (n=27) is 11.85+ 0.36 ppb;
the average was rounded off to 12 ppb.
6 40 CFR 180.319.
7 There are approximately 3000 acres of rice treated with silvex in a
total of 2,979,000 acres (ref. 7). The % of crop treated is:
3000/2,979,000 x 100 = 0.10%
8 Ref.34; 15.3 to 28.6% dariestically grewn sugar cane is treated with
silvex; 2.6-4.6% domestically consimed sugar has been treated;
sugar beets are not treated with silvex.
9 See discussion in text on p. 36.
F-46
-------
Table 9 provides a range far the dietary intake by the ceneral peculation
in the "hypothetical situation of rraxinal trea truant of the crops consistent
with the labeling. This situation, although highly unlikely, gives an
estimated rnaxiarri level of dietary exposure from presently registered
uses of siivex.
TABLE 9
MBCIMJM ESTIMATED DITTAKf 3CCSUEE TO SILVBC
Possible^ Percent? Fcod^- Sate of Dietary
Croo
Residues
(odd)
Crop
Treated
Factor
(%)
Ingestion
(uc/day)
Exposure
(nc/l
-------
- 36 -
Drrrafsr exposure frcm klcz
Residing of silvex have been defected in rlcs treated with silvex grown
in Arkansas, Louisiana, and Texas at levels ranging fmr. 10 to 30 ppb,
with an average of 12 ppb (Ref. 7). The calculations are fully explained
in footnote 5 of Table 8. As seen in Tables 3 and 9, the possible silvex
residues are expressed as a range fran 12 (avg) to 100 ppb, the latter
value being the interna tolerance (0 .1 pcm).
Certain ethnic croups eat nacre rice daily than the general pecula-
tion. Consideration must, therefore, be given to possibly higher dietary
intake of silvex iron treated rice by these ethnic groups. If wa assizes
that these groups (whose nunber3 cannct be easily estimated, but which
could be quite large) substitute rice for potatoes in their daily diet,
the feed factor could increase from 0.55% (rice food factor) to 5.5%
(potato food factor) 'which represents a ten-fold increase in the potential
dietary intake of silvex fran rice.
DUTTAKf EXPOSURE FPCM SUGAR C3NS
The use of silvex is recaroended for sugar cane but net sugar beets.
Zycadlo (Ref.34) estimated that 2.6-4.6% of all darestically consumed
sugar has been treated with 3ilvex. Based cn cur review of current EPA
files, it dees not appear that silvex residues on sugar have been analyzed.
Therefore, it is assuned that residues may be present at the maximum
permissible level, i.e. the interim tolerance of 0.1 pcm (100 pcb). The
percentage of the crop treated (Table S) represents the percentage of
sugar annually produced in the U.S. in recant years which is treated and
censured. The corresponding value in Table 9 is the percentage of
U.S. treated sugar consumed if 1D0% of the U.S. sugar cane were to be
treated with, silvex.
F-48
-------
Digram exposure r?cM pubs
Table 9 reflects the feet that cniy seme plana (Italian prvxes) are
treated 'with silvex, accounting for the fact that the rnaximua treatable
crop is only 12% (the percent of total U.S. pluti production consisting
of Italian plans). Based on our review of current SPA files it dees net
appear that analyses of si!vex residues on plans or prunes have 'seer,
performed. We, therefore, assvme that residues may be present at the
interim tolerance of 0.1 pern.
DIETARY E7TAKE rPOM PEARS
Silvex is applied to Anjou pears trees after harvest. Therefore,
any residues of silvex appear in the foliating years crop. The Agency
has no record of silvex analyses on pears. Based on the ccst-harvest
use pattern, we do not believe that a strong possibility exists for
silvex residues to occur in pears and have, therefore, excluded pears
from the dietary exposure estimate.
DIZTAKf EXPOSURE FROM APPLES
Vie are aware of a study dealing with treatment of apples with silvex
(Ref. 6) In this study, Mclntcsh apples were treated cn the tree 'with
a 20 ppm solution of silvex (according to label instructions) and were
analyzed for silvex residues at different daily intervals up to harvest
time, after 2 weeks storage, and 4 months1 storage (Ref.5). The following
results were obtained.
Silvex Silvex Residues
Residues* After Storage for...
At Harvest 2 weeks 4 norths
Unwashed apples 32 ppb 42 ppb 35 ppb
Washed apples 27 ppb 26 ppb 16 ppb
* 14 days after last application
F-49
-------
- 33 -
Via have iaken the range between the avera.ce residue (42 cpb) found in
unwashed ancles (stored fsr tv
-------
In order to translate these data to possible silvex residues in milk frcm
cows grazing on treated pastures, a study by Bjerke, et al. (Ref. 4) proved
helpful.
Bjerke, et al. (Ref. 4) shewed that feeding milk cows 1000 pan of silvex
in their daily feed resulted in an average of 100 ppb residues of silvex
in the milk at steady state.
If we assume, therefore, that the environmental fate of silvex and 2,4,5-T
are similar, we can use the data of 3ovey and Baur (Ref. 5) to estimate
(by interpolation) the amount of 2,4,5-T, and, therefore, silvex residues,
which would remain on treated grass 1 week after the last application
(There is a 1 week restriction of dairy animals entering silvex-treated
pastures). This value of 50 ppm of silvex in feed, is equivalent to about
5 ppb (0 .005 ppm) of silvex residues in milk, based on an extrapolation
of experimental data (Ref. 4). This extrapolated value is belw the
sensitivity of the method (0.05 ppm). The average male ingests about 500
g of milk and dairy products (ref.23) per day, expressed as of fluid
mil}*. At 5 ppb in the milk, therefore, a person consulting only milk
frcm dairy animals grazing on pastures recently treated with silvex
would ingest 2.5 ug of silvex daily.
2 ,4,5-T DIETAFY EXPOSURE
There are potentially two major sources of dietary intake of 2,4,5-T
from food:
1) the direct application of 2,4,5-T to rice
2) indirect exposure frcm meat, milk, poultry, and eggs derived frcm
chicken and livestock fed on contaminated feed.
Beef and dairy cattle may graze on rangeland and pasture that has
been treated with 2,4,5-T. This possibility is exemplified by the obser-
F-51
-------
- 40 -
vations f jti beef studies phases I and I , discussed in the previous
section, which showed that TCDD residues were detected in beef
cattle from 2,4,5-T-treated fields. However, TCDD and 2,4,5-T have
different metabolic pathways, and no extrapolation frcm TCDD to 2,4,5-T
residue can be attempted without further experimentation. We are not
aware of any specific studies of 2,4,5-T residues in rice or
meat or milk products, so that our dietary analysis will be done by
extrapolation frcm other data on similar compounds, e.g. silvex.
An alternate method of estimating 2,4,5-T exposure, discussed above
under silvex dietary exposure, is suggested by the studies of Bovey and
Baur (ref.5) who shewed that residues of 1.6 ppm 2,4,5-T remained in
grass 6 weeks after treatment with 2,4,5-T. That same period is also the
restricted interval for dairy cows reentering areas treated with 2,4,5-T.
Based on studies of Bjerke et al. (Ref. 4), residues of 1.6 pan in the
feed are estimated by extrapolation to yield residues that are below
1 ppb in the milk. Based on this calculated value of <1 ppb, we
estimate that the daily exposure to those individuals whose 500g/day-
fluid milk might be solely derived frcm contaminated milk, is less than
0.5 ug of 2,4,5-T per day.
Frcrn the work of Devine (Ref.7), it was shown that silvex residues
frcm silvex-treatment of rice actually do occur, and are at an average
level of 12 ppb. Since the use patterns for 2,4,5-T and silvex are
quite similar, and since the two canpounds are closely related chemically,
and since they both have similar environmental characteristics (such as
F-52
-------
similar half-lives), we nay estimate the following dietary exposure to
2,4,5-T for the general population fir an the silvex data on contaminated
rice:
Possible residue: 12 ppb
Percent crop annually treated: 10.9% (Ref.17)
Food Factor: 0.55(Ref.23)
Estimated Rate of Ingestion: 0.011 ug/day/person
Therefore, the estimated dietary exposure, based on recent usage patterns
would be 0.154 ng/kg/day, based on 70 kg body weight.
If the hypothetical, but highly unlikely, situation case may be considered,
in which all rice is treated with 2,4,5-T, the dietary exposure of the
general population would increase to 1.40 ng/kg/day. This might also
represent the exposure for certain limited populations which might
eat contaminated rice exclusively.
We might also consider the possibility that certain ethnic groups could
eat up to 10 times as much rice as the general population and might,
therefore, be exposed to between 1.5 and 14 ng/kg/day, a ten-fold
increase in exposure.
F-53
-------
- 41a -
DIETARY EXPOSURE TO TCDD FROM DEER AND ELK
The EPA Deer and Elk study was begun on/or about October 22, 1977 (Ref.28)
with the collection of perirenal fat sanples from animals in both v&sh-
ington State and Oregon. The rationale for this study was that, since
extensive use of 2,4,5-T on forests (reportedly on more than 250,000
acres in Oregon alone) was occurring in these states, the possibility of
contaminated game needed to be evaluated. Animals were, therefore, sam-
pled frciYi forested areas of these two states where 2,4,5-T was currently
in use. Table 10 surrmarizes the sampling program.
Table 10
Sumnary of Deer and Elk Sairpling
Collection
Dates
Number of
Samples (.M/F)
Description
10/22/77
10/23/77
10/31/77
11/21/77
12 (3/9) Elk Olynpic Peninsula of Washington on
several trips.
9 (7/2) Deer Taken between Aberdeen, Washington
and the Pacific coast
11/D5/77 10 (7/3) Deer Tillanook area of Oregon fran
hunters' catches.
12/33/77 15 (0/15) Elk Single herd between Coos Bay and
Roseburg, Oregon.
F-54
-------
- 42 -
The program in the Northwest was coordinated by Michael 'Watson, a toxicol-
ogist with EPA's Region X office. Dr. Watson enlisted the assistance of
Mr. Reade Brown (Chief, Game Management, Washington Department of Game,
Olympia, Washington) and Mr. Jerry MacLeod (Biologist, Oregon Department
of Fish and Wildlife, Portland, Oregon) who supervised the sample collect-
ion and quality assurance (Refs.29,30)
Dr. Watson provided the appropriate sampling protocol to be used; in
addition, he supplied all necessary equipment (which had been rigorously
cleaned in the laboratory to avoid precontamination with dioxins), so
that the deer and elk adipose tissues could be reliably sampled. Complete
capture records were required for each sample.
Following their collection, the adipose tissue samples were frozen within
24 hours, shipped to Dr. Watson under refrigeration and held in deep
freeze for approximately one year (until 11A4/78). At that time they
were shipped to the EPA Toxicant Analysis Center, in Bay St. Louis,
F—55
-------
- 43 -
Mississippi, for cleanup, extraction and encoding. Extracts, indirj-ing
appropriate blanks and quality central checks were subsequently forwarded
(on 9/06/79) to both the EPA laboratory at Research Triangle Park, Scrth
Carolina and to Wright Stats University, Dayton, Ohio.
RESULTS AND DISCUSSION
Table 11 sunmarizes the deer and elk data. The results of the analyses
of the Washington deer fat were inconclusive, "cur of 3 results require
reanalysis of the samples because of low recoveries, ana one rasult was
ND at a limit of detection of 2 opt. An additional sample '-as net ran at
all, due to small amount of adipose taken.
The results of the analyses of the Oregon ceer fat samples averaged
10.5, 12 and ND(4)ppt. Assuming residues to be present at the limit of
detection for the ND data*, the mean residues for this set '-culd be
about 3.3 pot. T^o of the six samples require reanalysis due to lew
recoveries.
* The rationale for assuning residues to be present at the limit of
detection in samples with reported >TD data, 'while net as defensible as
in the case of the beef data, nevertheless seers raasonable for the
following reasons:
(1) There were a number of confirmed positive results fron deer
and elk taken fran the same area of the respective states.
(2) We have no evidence which <~ould iaad us to conclude that the
NO results reflect no dioxin at all in these samples.
(3) We believe it to be prudent to resolve uncertainties in
data en the side of public safety.
F-56
-------
Table 11
Suanarv of Deer and 213c Dataa
Resort ed*3 TCTD - sec
Animal TJCJ RT? WSJ
Heecrted'0 TC%3 - set
Animal TAC r RTP WSJ
deer T/iArD-i SD
deer "WA-Q-4
deer /iArD-3 7^
SD(2)C
^ j deer CR-O-i 3D (4)
MA I deer CR-Q-5 12 31d
XZ3- I deer CR-Q-6 7 14
elk NA-2-2 9
elk Wfc-2-4 21
elk WA-S-5 12
elk WArS-7 lir£
elk Wfc-3-8 54
elk OR-2-7 24 29
elk OR-S-Q 4 MD(10)
elk OR-2-9 5 ND(8)
elk OR-2-U ND(2) ND(S-)
ND Not Detscted(see Table A-7 for DI? Criteria)
NA Sot Analyzed due to limited amount of sanple.
a. 3er *1.
b. Corrected for recovery losses
c. Parenthetic values are limits of derecticn for the analysis.
d. Sacoveries beicw 50%. Saroles to be rerun.
TAC =» Toxicant Analysis Canter (EPA Lab." in 3ay St. Lcuis, MS)
XT? = EPA Lab at Research Triangle Park, N.C.
ViSU 3 Tfcight State University, Dayton, OH
The results of the analyses of the Washington elk indicated mich higher
residues of TOD in the fat, with average values of 9, 12, 21 and 61 ppt.
The simple mean for this group of samples was 26 pot. Of the ten results,
three samples require reanalysis due to lew recoveries, and one sanple
was net run due to limited size. The high values were confirmed by both
analytical laboratories (21 & 21 ppt, and 54 & 68 pot).
The results of the analyses of the Oregon Elk shewed residues of TGD in
3 of 4 adipose samples, with average values of 5, 7, 7 and 26.5 ppt
TCDD. The mean for this group of saroles V uculd be 21 opt. Of the
2/ See footnote on pace 42.
F-57
-------
- 45 -
3 results, four were ND at good limits of detection. The high result was
confirmed by both participants (24 & 29 pot).
DISIRKf ESTIMATE
The following assumptions were used in the dietary estimate.
a. Tcr the local peculations which might consune the meat cf wild
game, deer and/ar elk neat are substituted partially or --holly
fbr beef in the diet, on an individual basis.
b. The Sat content of deer meat may range fran 4% - 17%, depending
on seasonal variations .V
c. The fat content in elk meat is essentially the sane as that of
deer meat.**/
Residues of TCED ranged fran MD (2-13) to 31 pet in contaminated deer fat
and froa !ffl(0.S-25) to 63 pot in contaminated elk fat. TOD intake fran
contaminated deer or elk is sunnarized in Table 12. This is a worst case
situation in that meat consuned consists of only contaminated deer or
elk meat. We do not have sufficient information to estimate the nunber
of persons in this group.
37 Deer store fat during the spring and surmer, for use over the winter
In a study by Medin and Andersen, (Pef. 19a), Colorado Mule Deer were
analyzed for fat content and found to contain 5-17% (males) and 7-12%
(females), the range reflecting seasonal variation. '-5cwever, Dr. J.
Pennington, a nutrition expert with the US Food and Drvo Administration
reports (Page 1S6), that raw venison contains about 4% fat. riDwever,
Dr. Pennington reported that consurption of venison was rmch higher per
capita (6-3cz. portions) than typical beef consumption (4 oz). Thus,
ccnsunnticn of deer fat contaminated 'with TCZD, might be comparable to
consumption of beef 'with twice the fat content for purposes of estimatin
dietary exposure to TCX.
~*/ We could find rso specific information on the amount of fat in elk
meat. We have no reason to believe that elk contain more or less fat
than deer. Therefore, the ass'oncticn that deer and elk meat contain
comparable percentages of fat seems reasonable.
F-58
-------
Table 12
Dietary Intake of TdD Fran Contaminated Deer cr Elk
TCDD in TCDD in Dietary Intake**/
Animal Fat (pot) Meat (pot)* (pc/perscn/ day) cc/kc bw/cay***/
DEZl ilD(2-L3) - 31 0.08 - 5.27 9.9 - 650 0.14 - 9.3
ELK SD(0.8-25)— 68 0.03 - 11.56 3.7 - 1430 0.05 - 20.5
*/ Assures 4% - 17% fet, depending cn season. Gcrtcuted range
is the lowest percentage fiat multiplied by latest limit of
detection to the highest percent felt irultiplied by the
hi chest detected residues. Thus 2:0.04 =0.08; 3133 .17 ~
5.27; 0.8*3.04 = 0.03; 68:0.17 = 11.56
*"*/ Assumes deer and elk meat is consulted at the same rate as
beef is consigned (124 gns/person/day.).
***/ Assunes a 70 kg perscn
Thus, a perscn consuming contaminated deer meat once a month (cr for a
period of 12 days following the hunting season), for example, could
possibly incest from 1.7 to 111 pg 2,3,7,8-TCn3/kg-5W/ year. Similarly, a
person consulting contaminated elk meat could, at that rate, ingest front
O.o to 246 pg 2,3,7,8—TCD/kg 3W/year.
An informal survey of ten persons was taken during June, 1980 (Par.9)
to determine typical consumption of deer and elk meat. The 10 people
contacted resided in Oregon, and reported having deer and/or elk meat
on hand. One person ccnsuned venison 4 times a -week until all meat cn
hand was gene? six people consulted venison or elk meat about once a
week; the ether three persons ccnsuned venison or elk about once every
two weeks, until the meat was gone. Typical consumption of this group of
people seemed to be about once a week. It is not knovn whether any other
persons were contacted 'who did net have game on hand, or whether this
group of persons were selected because it was suspected that they were
likely to have came on hand.
F-59
-------
- 47 -
Individuals eating contaminated deer cr elk :neat ores a week (or
at a comparable rats) vculd have a dietary intake of TCDD ranging
from 7.3 to 435 pg TCDAg 3W/year fran contaminated deer meat, and froa
2.6 to 10^D pg TODAg HW/year fran contaminated elk meat, assuning a
year's supply of meat v»as available.
If leas meat were available, estimates vculd be correspondingly lower.
Thus, if enly a six month's supply were cn hand, dietary intake might
be one half of the estimated rate.
In addition, if, in fact, individual consunpticn of game is higher than
beef ccnsursxicn for a particular meal then intake of TCDD from this
source .-nay be higher than is reflected in this estimate.
F-60
-------
— *«o —
REFERENCES
1 Accession t9. Strrnerv cf Deer and Elk Studv. 1/23/83. EPA Exhibit
So. 199.
2. Ancnvmcus. 1979. The 3iclccic and Economic Assessment or 2,4,5-T. A
decern of the USDA-States-5?A 2,4,5-T SPAR Assessment Team, Feb. 15,
1979. Giapter 5, 1-212.
3. 3aur, J.R., R.W. Bcvey and J.D. anith. Herbicide Concentrations in
Live Oak Treated with Mixtures of Picloram and 2,4,5-T. Weed Science.
17(4). 567-5*73. October, 1969.
4. Bjerke, 2. L., J. C. 53ecnan, P. W. Miller, and J. H. Wetters (1972).
Residue Study of Thencxy Herbicides in Milk and Cream, J. Aaric.
Food Giem. 20 , 963-967.
5. Bcvey, R.w. and J.R. Baur. Persistence of 2,4,5-T in Grasslands of
Texas. Bull. Eav. Contain. Toxicol. 8(4). 229-233. 1972.
6. Cochrane, W.?., Greenhalgh, R., and Lconev, U.S. (1976). Canadian
J. of Plant Sci., 207-210.
7. Devine, J. M. (1973), Report frexn the Syracuse University Research
Carp., "Silvex Residues in Rough Rice and and Straw", Life Sciences
Division, Pesticide Analysis Laboratory, pg. 11-17.
3. Durham, W.F.and H. R. Vfolfe, 1962. Measurements of the Exposure ox
Workers to Pesticides. Bull. WHD, 26, 75-91.
9. Green, G. Letter to G. • Strei3inger dated June 12, 1980.
10. Jensen, D.J, R.A. Hummel, N.H. Mahle, C.W. Kocher and H.S. iiigcins.
A Residue Study cn 3eef Cattle Consuming 2,3,7,3-Tetrachlorodibenzo-p-
Dicxin (TCED). July 19, 1978, Unpublished. (2A Exhibit No.159).
11. Kearney, Phillip C., Ecwin A. Vfcolson, and Charles P. Ellington, jr.
Persistence and Merabolisn of Cilorodioxins in Soils. Zrrv. Sci. Tech.
6(12). Noveroer, 1972. (EPA Exhibit No. 149)
12. No reference.
13. Kolm3din-He<±nan, 3., K Erne, M. HaJcansson, and A. Engcvist 1979.
Vetensxanlig Skriftserie, 17, 26 pp.
14. Lavy, T.L. 1979. Project Canpletion Report to National Forest Pro-
ducts Association. Measurement of 2,4,5-T Exposure cf Forest "-fcrkers.
F-61
-------
- 49 -
15. Lavy, T.L. J. Scott Shepard, and J.D. Matice, 1980. Exposure Measure
merits of Applicators Spraying 2,4,5-T Acid in the Forest. J. Agric.
and Pood Chem. 28, 626-630.
16. Lavy, T.L., J.S. Shefhard, and D.C. Bouchard, 1980. Field Worker
Exposure arid Helicopter Spray Pattern of 2,4,5-T. Bull. Environm.
Contain. Toxicol. 24 , 90 -96 .
17. Lee, R. E. Econanic Analysis Branch. Memorandum to Gunter Zweig, HED
August 21, 1980.
18. Mayland, H.F., G.E. Shevaraker and R.C. Bull. Soil Ingestion by Cattle
Grazing Crested Wheatgrass. J. Range Manag. 30(4), July 1977 264-265.
18a. Medin, D.E. and A.E. Anderson. Modelling the dynamics of a Colorado
Male Deer Population. Wildlife Monographs 68. July, 1979 (Table 17).
18b. Pennington, J.A. Division of Nutrition, Bureau of Foods, FDA,
Washington, D.C. Personal Ccmmunication on 8/25/80.
19. Ramsey, J.C. , T.L. Lavy, and W.H. Braun, 1979. Exposure of Forest
Workers to 2,4,5-T. Calculated Dose Levels. Unpublished Report.
20. No Reference.
21. No Reference.
22. Foss, Ralph. Critique of Segments of Letters Which Relate to the Ade-
quacy of Analytical Methods for TCDD in Environmental Sairples; in
Final Environmental Statement - Vegetation Management with Herbicides.
U.S.D.A. May 11, 1977. pp H-20-25.
23. Schmitt, R.D. Update of Food Factor Tables. Memorandum to Acting
Chief, Toxicology Branch. May 1, 1978
24. Shafik, M.T., Sullivan, H., and Enos, H.F., 1971. A method for the
determination of lav levels of exposure to 2,4-D and 2,4,5-T. Internat.
& Environm. Anal. Chen. 1: 23-33.
25. Simpson, G.R., Higgins, U., Channan, J., Bermingham, S., 1978. Exposure
of council and forestry workers to 2,4,5-T. Med. J. Austr. 2(x):
536-7.
26. Upholt, W. M. Direct testimony (EPA Exhibit No.139)
2 7. USEPA Hearing Record of 2,4,5-T and Silvex. The Dew Chemical Co.,
et.al., Exhibit No. 494, Exh. Ident. 6729, pp. 6975 ff.
F-62
-------
- 50 -
28. Tfexscn, M. Sunrery: ?a-cicnaie and St'jdy Design for Prcccsed TCZD
Analysis ox Pegion X Six and Deer ?cLipcse Tissue Sarales. 9/13/7Q.
29. Watscn, M. Latter to Mr. Seede Brown dared 10A 7/77
30. Warsen, M. Letters to Mr. Jerry .MacLeod dated 10 AO and 10 A7/77.
31. Vfalis, K.3.., J.?., Amstrong, D.C. Staiif, S.W. Caner, and w.?. Durham,
1975. Exposure of Apple Thinners to Parathicn Pesidues Arch, Envir.
Contain. and Toxicol. 3_, 257-267.
32. Young, Alvin L., Cross Examination Testimony, STFSA. Docket r 415 et.al.
10131-10133. Wednesday, July 23, 1980.
33. 3weig, G. Direct Testimony (Z?A Exhibit No.203)
34. Zyradio, L. 3FSD. Memo to G. Zweig on 3A3/93
F-63
-------
- a -
APPENDIX
ISTT.M3TICN OF 2CPCSZD ?CFJLATICN AND OJPATICN C-F HXPCSjTE TO 2,4,5-T
In order to astinata the annual exposure of 2,4,5-? to a population and to
assess the resultant risk, the following information rrust be taken into
dCCCUnt:
1. Number of individuals within a defined occupational group.
2. Duration of daily and annual exposure for these specified
occupational groups.
These nurbers are for the most part net available as hard data for
2,4,5-T and siivex uses, and consequently an estimation technique has
been utilized. Annual exposure and the nurber of applicators for several
types of uses have been estimated 'cy considering total acreage treated,
area treated/unit tisse, the tiding of application, and by making certain
assumptions abcut duration of exposure*. The basis of data for "total
acreage", "area treated/ unit time", "timing of application", and other
use information is contained in Reference 2, (hereafter referred to as
the Keccrt). The specific uses fcr which these estimations were done
are the following:
* Forestry Pights-of-Way
aerial aerial
ground broadcast ground
backpack mist blower selective basal
backpack spray cut stvrp
mixed brush
* Range and Pasture handgun
aerial truck broau
backpack spray railroad tracks
electric power lines
* Rice
aerial
The results of these estimations are sumnarized in Table A-l.
* In sane cases we have considered that the hours exposed were equivalent
to the hours «crked, as is axslair.ed in the text.
F-64
-------
-52-
TAIUJ5 A-l
2,4,5-T - h'ot imatJon of llie Hiuiil>er of Iho lixjiotiecl Population Ami Duration of lixpouure^ )
_ 2 3 4 5 6 1 0 9 10
EXPOSED ANNUA!.
T0TPA1» TREATIift THfc'Al'MI'.Nr POIU- OAflY LXPOSIJHIi
ACKI.W1I3 AntKttUS DONATION RA'112 daya/yr I.ATION EXPOtUJIUi (hru/yr/
USE PAI'l'liW EXPOSED GIOIP (1000'a) (A/hr) (hra/day) (Ih.al/A) (avg) (no.) (bra.) person)
LOUIES riW
1. Atrial
Pi lota
Ml xer/lx>adars
076
076
60
60
2
2
i .5-3
1.5-3
100
100
73
73-145(2)
2
0
200
ft)0
2. Ground
Broadcast
Boom Tr. Opr.
Mixer/foadera
140
140
6.5(avg)
6.5(av
-------
TAIilJE A-l (continued)
2,4, 5-T - Estimation of the Number of the Exposed Population And Duration of Exposure^ )
2 3 4 5 6 7 0 9 KT
liXPOSI'J) ANNUAL
TOTAL TM\ATW) TRlSA'iWJiNI' POPU- DAT LY IXPOUJRK
ACKKttlK ACRIWili DURATION RATI3 duya/yr IAT10N EXPOSURE (hru/yr/
USE PATI'EKN EXPOSED GROUP (LQOQ'a) (A/hr) (hra/day) (lb.ai/A) (avg) (no.) (lira.) person)
RIGHTS OF MAY
1. Aerial Pilots 206 20<6) 4 0 100 2 5 4 400
Mixer/Loadera 206 20 <6) 4 0 100 25-50 4 400
2. Grouiri
Selective
Ituaal
Applicatora
235
0.5
6
6.4
17)
13) The 20 A/hr figure iu Laken fran Table 35, cliapter 5 of reference 2, alxive.
-------
It should be net ad that we are more certain about our estimate of the
total nurber or axpcsure-hcurs for each specified use and use pattern
than ve are about the exact amber of individuals in each group and the
nunber ox hours >crked by each individual.
Since for each occupational group...
total * exscsure hrs3, » t of workers3 x averace t hrs '*crked or exposed0
even if Co) and (c) were in error, they v-ould vary inversely and (a) would
not change appreciably.
SPECIFIC EXPLANATIONS OF TABL2 A-i
Colaan 3 - Total Acreace
This nunber is taken frtia tables or the text of Part 3 of the Report. Fcr
axairnle, the first figure under aerial forest, 876,000 A, is Sound in
Table 12, p. 5-95 of the report.
Collars 4 and 5 - Acreace Treat ed/tJnit Tina - Duration of Treatment
These nisrbers are usually found in the text or in the "Calculation 3uxary"
of the Report. This is an estimated averace based on the descriptive
portion of the Report or the Calculation Surinary Table. For exarole, on
p. 5-92 of the Report it is stated that it may take 10-30 minutes to
treat 30 acres by helicopter. As stated in Calculation 5 urinary Mo. 1,
one sits of up to 13D acres usually 1-3 hours to treat 'with herbicide.
3ased on this specific information we have chosen 60A/hcur as the acreace
treated cer unit ting and 2hrs/dav as the duration of traatrgnt.
Colunn 6 , Acolicaticn Rates
Application rates are found in the text of the Report or in Calculation
Sunnary tables. When a range is given (e.g., 1.5-3 lb/A) the approximate
F-67
-------
- 55 -
weights^ cr arithmetic average, e.g. 2 lbs/A, is used far further calcula-
tions .
Collar. 7
The davs-cer-year figures are an indication of the r.urtcer of cays 2,4,5-T
might be applied for a certain use throughout the year. Sir.cs mcst
applications are rnade by professional applicators, it i3 assuned that
the same crew <*cuid be applying 2,4,5-T1 in different areas. According
to the infcrraticn on page 5-90 of the report, 2,4,5-7 is applied during
3 periods of the year: Jeb-March; iMay-Jone: July-Sept:. Assigning inclement
'-weather conditions (rain, wind), we estirate, therefore, that the aerial
crew for forestry use "will be applying 2,4,5-T for ICO cays/year.
Column 9 - Number of Ir.djvicr.als of E:ccsed Peculation
This nunfcer is calculated as fbllcws:
= Scray craws = Total ac="3a?e (Col ^.3 ) —
A/hr (Col.4) x hrs/cav (Col.-) x cays/yr (Cci.7)
?or ax2role Scrav crsvs = 8/6,000 73
* * 50 X 2 X ICO
If, according to the 3fipcrt, a spray crew consists of one pilot and 1 to 2
mixer-loaders, then (as in the example afceve) there >culd be 73 pilots
and 73-146 mixer-leaders. A risk assessrent might be performed cn the lew
and high range of the exposed peculation. (Tor one use, aerial- application
of 2,4,5-T to rice, the nunber of pilots is fairly well ccnfirrjed to be
around 300.)
Column 9 - Daily Exposure
*fe assune pilots to have a short exposure. In most cases only t\*o hours
of flight time are permissible during an average <*ork day. Thus,, pilots
F-68
-------
- 56 -
usually ara lis-id as being expcsed for 2 hours/work day. The mixer-lead-
ers in aerial application ara encaged in the leading and :r.ixing cf pesti-
cides during the actual application pericd (2 hcurs) bur ara assured to
be working on other tasks throughout the workday (6-3 hcurs) without a
change of clothes. Thus, we believe that the vorkers will be exposed to .
2,4,5-T during the entire '«ork day by contact through the skin fran wet,
pesticide-ccnraniinated, v»crk clothes.
Colunn 10
Annual Exposure =¦ I^ys/vr. (Col.7) x daily exposure (Col.9).
SPECIFIC DATA P0I3T5 AND AS5J>PTICNS*
Forestry - Air Acclicaticn
Total Acraace - 376,5D0A (Table 12, p. 5-95).
Acraace Traatad - ISOA/day: usually 1-3 hours (Calc. Suimary tfo.l).
Acolicaticn Hate - 1.5-3 lbs/A (Calculation Summary So. 1)
Davs oer year - 100 days (Table 10, p. 5-90)
e.g. Pacific Coast (pine release): Fab-~iarch,
May-June and July-Sect
Daily Exposure - As discussed previously, the assumption is trade that
the pilots are expcsed 2 hours/day based on acrual
flight time and change clothes at the ccmpleticn of
the flight. Cn the other hand, the mixer-loaders
ara assuaed to rsmain in the field engaged in orher
tasks, wearing contaninared apparel during the
norial 'working day of 3 hours. Therafbra, exposure
is estimated at 2 hrs/dav for pilots and 3 hrs/day
for mixer-loaders.
Forestry-:* round Broadcast (Tractor-applied)
Tc-^1 acreage: 140,COO A (Table 12, cp. 5—ICQ ).
Anslicatlcn Rata: 2-3 lbs/A (Table 14).
Acreage Treared: 5-3 A/hour (p. 5—99).
* All other points ara fsund in Table 1.
F-69
-------
Joras-trv-3 round Srcaccisz (Tractor-applied - Continued)
Dur3.-ti.CR of Treatment: 4 hours/cay (p. 5-39).
Days /year: approximately 60 *crkinc days (Mid-April- Mic-Juiy);
Dally Ijcsosure: We are assuming that the mixer leaders -crk a full
'*crk day beyend the 4—hours of application, and are,
therefore, exposed to the pesticide during the errcir
vork day of 3 hours.
Forestry - 3acksack Mist-3lower
Total Acreages 24,COOA (Table 15, p.S-L03)
Hate of Acslicaticn: 2 lbs/A (p. 5-L01)
Davs/vear: May to July - approximately 3 months or 60 working days
Treated Acraaca: 3-5A/day (p. 5-LG1) or approximately 0.5A/hcur
Forestry - 3acksack Scraver:
Tcta.l Acraaces 125,0G0A (Table 15, p. 5-L06)
Accllcaticr. Hare; 2 lbs/A (p. 5-104)
Treated Acrsace: 3-5A/cay, 4 hrs application: 3 hrs '-^crk day
Dav/fear: May-July: 1-Tcv-Maroh; approximate 100 davs/yr.
Rarceand Pastures
Aerial Acolioaticn
Total Acreace: 1,578,X0A (Table 17)
Treated acreace: 100-300A/hr. (p. 5-111) av. 200A/hr.
Acallcaticn Rate: 0.5 - 2 lbs/A (Calculation 3unnary Mo. 2):
weighted average - 1.0 lb/A
Duration of treatment: 6 hours/day (p. 5-L11)
(3 hours AM and 3 hours ?M)
F-70
-------
• 58
Aerial Acglicatica (continued)
Davs/vear: Piicrts and Mixer/Leaders: 1 -4 vrics., 10 cavs (avq)
(p. 5-111)
Flaeperson: about 3 cavs (assumes 4C00A fam at
1200A/day
Daily exscsura: It is assigned that the pilots chance clothes
after each flight period, naking a total of 6
hours exposure. The other -*orker3 are assuned
to retain the same -*crx clothes during an 3-hr
\«oriciay, resulting in 3 hours of exposure.
Z.-ccsec Peculation: .issurting the average ranch to be of 40C0A
size and 2 flag oerscns per ranch, it is estin»-
ated that (1,600^000: 4000) :c 2 = 330 flag
persons will be erolcyed. Other peculations
•were estinated by the calculation shown on p. 52.
Ranee and Pasture
3acksack Scraver:
Total Acreace: 1,060,000a (excluding mescuita, table 13,
Acr33.ee: 3-5A/day (p. 5-113)
Duration cf Treatraarrt: chrs/cay
Rata; 0.5-2 lbs/A (Table 13); weighted average: 0.3 lb/A
The best available infsrnaticn is that 97% rice treatment is by air
(Report, p. 5-i42).
Total acreace: 292,CCOA (p. 5-144)
Treated Acreace: 46A/35min or approximately 30A/hour (p. 5-143)
Duration of Treatment:
Calculated 2 hours/day and o days/year for pilots and loatrnen.
Calculated 0.5 hrs/year for flagperscns.
F-71
-------
- 59 -
Rica (continued)
ELxirrsed Peculation: 307 pilots 'p. 5-149). "or each pilot there is
cne Icacran. Sines the nurber of pilots is 'cscwn,
the yearly exoosura hours '^ra calculated oy the
equation on p. 52. There are roughly 5330 rice
farrvers with an average cf 1.5 ilacpersons being
supplied by each fanner: thus, the nurber cf
flacperscns ranee from 5 5C0-9500 (p. 5-148).
Richts-cf-Wav
i. Aerial
Total acreace; 206,256 A (Table 19)
Treated acreace: 5-15 A/hr. (p. 5-120)
20 A/hr. (p. 5-191); this nunber is used
Typical Work gay: 4 hrs (p. 5-120)
Work Crew: 2D-75, consisting cf pilot
mechanic/ service
mix-truck driver = mixer leader
Duration: 22 weeks or about 100 days
2. Ground
a. Selective ^sal (p. 5-136, calculation
suimary No. 7)
Total Acreace: 235,000 A
Trsstsd Ac^ssLce;
0.5 A/hr or 3 A/day
for 6 hour '*crk cay
Duration cf treatment
34 weeks or approximately 170 days
Scs. cf esecsed Pcculatiai
vaork ~rgw consists of
driver-mixer
2 spray men
Mo of <*crk crews = 235 ,000
170 x 3 = 4c0
Total ncs. of persons = 460 x 3 = 1383
F-72
-------
60
b. Cvrt Stuns (calculation Sumary S)
Tcral Acreace: 9,901 A
Dcsace: 3.2 lb/A - 4.5 lb /A
Average: 4 lb /A
Duration of trsa-teenc
34.7 weeks or 1a3 days / year
Acolicaticn tin*?; 6 hrs / day
Application rate; 0.3 A/hr
(based on esrirata)
No. of workers earcsed:
10,000
3 x 173 =* 20 work crews
Crews made up of 2 spravraen
1 truck driver-mixer
Total =» 50 persons
(Sumary Table 3 lists 76 exposed personnel; this mist include 1 supervisor,
who is net included in cur estirnat.es. We also-assune
that all persons are exposed during entire S hour *^ork day)
c. Mixed 3rush - I-Iandcun (Calculation Sunnary 9)
Total Acraace: 29,400 A
Treated Acrsacs; 0,5 A/hr
6 hour day
3 A/ day
Duration of Annual treatment: 110 days
Sxscsed Peculation: 89 work crews consisting of 4 persons
Total: 356 persons
(Note: There is an error in Calculation Sornery 9? should
toe 39 work crews instead of 39, as written.)
F-73
-------
c. Mi-Ted Hr-Jish (Roadside)
Truck - crocm applicators
(calculation Sumsary Mo. 10)
Total Acreace; 53,447 A.
Duration cf treatment:
22 weeks cr 110 days
Mca. of esrcsed vacrkers:
Driver, mixer leader
Scravman
90 wcrk crews
eared Acreaca; Given 90 work crswg (p. 3-139) and
110 days x 6 hT3 and 50 ,000 A to ce
treated, the calculated trea-ctent rat
is approximately 5 A/day or 1 A/hour.
d. Railroad Crg* (Calculation Sunrary 11)
Total Acreace; 100,000 A
Treated Acreace: 10 A/hr
Duration cf treatment
4 hrs / day
13 weeks / year
260 hrs / year
Total ccculaticn
38 crews x 4= 112 persona
e. Electric Power (Calculation Sunrary 12)
Total acreace: 44,000 A
Treated Acreace: 0.5 A/hr
F-74
-------
- 52 -
Durati.cn:
6 hrs / cay
110 days / year
Total ncs. of serscns:
Driver / mixer-loader
2 spraynen
Ncs. Of crews:
44 ,QC0 =» 133
330
Total ncs. individuals =¦ 400
F-75
-------
-63-
TAnUC A-2
ISat Invited Occupational P.x|xi3ure to2,'1,5-jr (Refs.14,12)
TOrAI.
MTN.
PERSON
AVRIW1E
AVRIVY1R
TIMR
AMJUN1"
(Mi)
mi/iq;
AM3UWP
AM*INI* (MJ/kG)
W0RKRR
WORKRR
WORKRR
CXPOSRI)
AnsoRnRi)
Antt)RURn
Ansomircn
An.«*)Rni-3")
NO.
ACriVI'IY
WEIGUI* (Kq)
A* U*
A
D
A
n
prr mxm
1
Mixer, Oackjiack Sup* r
72.6
im 173
1.040
1.221
.014
.017
.016
.005
2
Rackpnck Sprayer
68.1
M M
9:016
5.0D4
.132
.005
.109
.036
3
•1 II
49.2
II M
4.505
3.173
.092
.063
.070
.026
4
• I M
95.3
M (1
4.206
4.003
.044
.042
.043
.014
5
II ••
52.2
M M
l.rno
0.969
.034
.010
.026
.009
6
II
II
65.0
II
II
2.752
.042
.042
.0 14
7
II
•1
74.9
II
•i
7.312
3.270
.090
.044
.071
.024
Q
Tractor Sup'r
95.3
245
200
3.077
1.169
.032
.012
.022
.006
9
Tractor Driver
04.0
II
II
3.077
3.000
.046
.037
.042
.012
10
II
II
106.7
II
II
4.369
5.766
.041
.054
.048
.014
11
Ml xer
79.5
11
II
6.010
4.240
.005
.053
.070
.020
12
Mlcrofbil Pilot
95.3
55
117
0.193
1.151
.002
.012
.007
.005
13
il
Mixer
109
II
II
10.130
0.002
.093
.001
.007
.061
14
II
Sup* r
134
II
II
0.541
0.455
.006
.005
.006
.004
15
II
Flagman
61.3
II
II
0.469
.000
.000
.006
16
*1
II
74.9
II
II
0.241
.102
.003
.001
.002
.001
17
Raindrop Pilot
72.6
115
116
3.323
3.560
.046
.049
.040
.024
in
II
Mixer
06.3
II
It
7.431
13.501
.006
.156
.121
.063
19
II
Sup* r
01.7
II
II
0.077
0.153
.011
.002
.007
.004
20
II
Flacjnan
06.3
•1
II
0.275
0.112
.003
.002
.003
.002
21
il
95.3
ii
It
0.251
0.345
.003
.001
.004
.002
A and 11 represent tv*o exposures.
-------
64 -
TAELZ
A-3
Esxinstad Cccusaticn?1
Exocsur
s to 2,4,5-T
USE
PATTEST
2CCSED WORKER
; y j." y
WOFKSR
NUMBER
S/G. AMCUOT
ABSOREED
(rsc/kc/hr)
C-RCU?
AVERSGES
(irc/xc/hr)
1
1 AERIAL
1
| it
|
Pilot - Micrefbil
Pilct - Raindrop
1
I 12
1
1 17
1
.CO 5
.024
1 1
1 1
1 .015 I
1 1
1
t 11
1
1 "
1
Mixer - Microfbil
" - Rairairop
1 13
1
1 18
1
.061
.063
1 1
! 1
1 .062 I
! 1
1 1
)
1 ¦»
1
t 11
{
Sup'r - Microfeil
- Sairdrcp
1 14
1
I 19
l
.004
.004
1 1
1 1
1 .004 |
1 !
1 1
1
! 11
I
Flacman - Microfbil
1. IS
1
.006
1 1
1 1
1 1
1
1 "
I
1 11
j
i* U
ti II
1 16
1
1 20
I
.001
.002
1 1
1 1
1 .003 1
1 1
1 1
I
1 11
»
il ii
1 21
1
.002
! !
1 1
1 ' 1
1
1 GHXSD
1
Mixer/Leader - Tractor
1
1 11
1
.020
1 .020 1
I |
1
I "
1
i "
j
Driver - Tractor
if ^ «•
1 ' ID
1
1 9
1
.014
.012
1 1
1 i
1 .013 1
1 1
j. .. j
1
i (>
i
Sup'r -
I
i 3
1
.006
1 .006 1
1 |
1
i "
i
Applicator - Backpack
I
I 7
1
.024
1 1
1 1
1 |
1
i 11
i
(« it
1
i 6
1
.014
1 I
1 1
1 I
I
! "
i
I "
I
il il
il it
l
1 5
1
1 4
I
.009
.014
1 1
1 1
1 .021 |
i i
1
1 11
i
il il
1
1 3
1
.026
1 1
1 1
1 i
I
1 !•
i
ti H
1
1 2
i
.036
I !
i i
i j
1
1 "
i
Mi xer/Supervisor
1
1 1
1
.005
1 .005 1
1 1
F—77
-------
- 65
Table A—1*
TCDP Keaidueat Fran Adipose Tissue - Phage One Deef
Application Sanple** ppt 2,3,7,0-,rCDO*** Application Hanjde** ppt 2,3,7,0-TCDO** *
Rate (lb/A) Nuiiber (limit of detection) Hate (lb/A) Ntaiilier (limit of detection^
4
NZZDIO
6(6), 8(0), N0(6)
2
KZfl)93
NO(3)
4
NZ2019
N0(12)
2
KZ7095
NO(7), NO(0)
2
KZVD90
NO (13)
3
IJZ2002
NO (6)
2
KZa)99
NO (6)
3
IJZ7005
10(9), N0(5), NO(12.5)
2
KZZl 00
N0(3), N0(5)
3
UZ2006
21 (9), 23(5), 29(13), N0(L3), Nl)(20)
3
UZ7007
N0(10), ND(24)
I
NZV030
10(10)
3
UZ2D10
12(3), 16(6), 39(0), 45(13), 40(5), 54(5),
I
NZV06L
5(1), 5(2), 7(6), NO(7), NO(lO)
63(13), 66(13), 75(3), 09(7), Nl)(4)
1
NZV026
NO (6)
3
UZ20U
10(7), N»(3), NO(4), Ml)(1 2)
1
NZ2059
N0(9)
3
(JZZ013
N0(0), ND(16)
I
NZ/f)60
N0(14)
3
UZVM4
NO (21)
I
NZVU69
NO (9)
3
UZ/IH5
14(5), 20(0), NO (3 ), NT)(20)
I
NZT070
NO (6)
3
UZ3U6
0(6), 9(9), 14(10), NO(4), N0(13)
3
UZ2D17
19(12), 20(0), 20(0), 24(20), 29(10)
1/2
NZ2D20
NO (2)
1/2
XZ2072
NO(9)
2
KZ2047
9(5), 22(20), NO(27)
1/2
XZ5H73
NO (6 )
2
KZZ101
22(14), NO(00)
1/2
XZ7075
NO (6 )
2
KZZl 02
30(10), NO(20)
1/2
XZVD/0
NO(IO)
2
KZ7036
NO (6)
1/2
XZ3100
NO(9)
2
KZtf)3n
NO (17)
1/2
XZ7D02
NO (7)
2
KZ7D40
N0(10)
1/2
XZ7T104
Nl")(7)
2
KZtf)43
NO (20)
1/2
X7.M07
NO(0)
2
KZZ049
NI">(13)
1/2
xzvnoo
NO (6 )
¦ lligli resolution MS analyses only.
* Mills table Is modified from Table I of the testimony of (5. Strelslncjei: (I5PA Exhibit No.564).
Itejiorted values have been confirmed from the original worksheets.
** Flacli entry represents a different animal.
*** Analyses wore |ierCbmied on identical extracts of the same nairple.
-------
Table A-5
TC2D Residues fra
a Adircse Tissues
- Phase T\»o Heef
Arolicaticn
Sannle
npc. 2,3,7,3-TCDD
3ate (lb /A)
Nvscer
(Lisdt of Deracticn)
3
BAII-4
iro(io)
3
HAH-5
7(7), 3(7)
2
3AII-9
7(7), 11(7)
2
HAH-i
1TDCL0)
2
2AII-2
sDao)
2
HAH-S
SD(8)
2
SAII-9
SD(10)
2
HAH-12
13(10), 13(10)
2
HAII-16
® (3)
2
HAH-17
HD(7)
2
3AII-13
SD(7)
2
BAH-20
31(3), 34(3)
2
2AII-21
ro(io)
2
HAH-22
SD(10)
2
HAII-*6
SD (3)
2
HAH-47
SD(10)
3/4
HAH-34
®(10)
3/4
HAH-3 5
3D (10)
3/4
HAII-35R
®(10)
3/4
HAH-36
MD(10)
3/4
HAII-36R
mdo.o)
1/2
HAH-LO
iJD(14)
1/2
HAH-11
ND(10)
1/2
HAH-14
.TO (8)
1/2
HAII-23
SD(8)
1/2
HAH-26
ND(7)
1/2
HAH-27
9(7), 10(7)
1/2
HAH-2 3
3D (8)
1/2
HAH-31
MD(7)
F-79
-------
- o / -
Table A-S
Adjusted pjaaicie Data -
ii
TCTD - Phase
Bear
lb/A
Sample*
Adjusted**
Sanple* Adjusted**
timber
Residues (rot)
lb/A
dumber Residues (cct)
3
UZ2D02
6.0
1
SZZC30
10
3
UZ2005
9.2
2TZ2361
4.5
3
UZ2D06
21.2
1
NTZZ026
2.0
3
UZ2C07
17.0
1
XZ&59
3.0
3
UZ2D10
46.5
1
5TZZ260
4.7
3
UZ2Q11
9.3
1
MZ2269
3.0
3
UZ2313
12.0
i
MZ307D
2.0
3
UZ3314
21.0
mean =
4.17 cct
j
UZ3315
14.3
3
UZ3316
9.6
3
UZ2D17
22.4
1/2
JTZ3320
0.33
••nean
= 17.1 opt
1/2
X2X72
1.50
1/2
XZ2D73
1.00
1/2
XZ2373
1.00
2
XZ2D47
16.3
1/2
XZZD78
1.67
2
KZZLOl
37.7
1/2
XZZDS0
1.50
2
KZZL02
21.7
1/2
XZ2382
1.17
2
KZ2336
4.0
1/2
XZ2DS4
1.17
2
KZ3338
11.3
1/2
X22287
1.33
2
KZ2D40
6.7
1/2
XZZD88
1.00
2
:
-------
- 53 -
Table A-7
Criteria Used sv the Dicxin Monitoring
to Confim rc:D Residues "
1. Camllary colann GC/HFMS retention time of reference standard
2,3,7,3-TCDD.
2. Co-Injection of sample fortified with ^^Cl-TC^D and 2,3,7,3-TwD
standard.
3. Correct molecular ion chlorine isotope ratio (m/e 320 and m/e 322).
4. Capillary colann GC/HWS which give simultaneous multiple ion mon-
itoring response (m/e 320, m/e 322 and m/e 323) for TC2D.
5. M/e 320 and m/e 322 'MS response greater than 2.5 x noise level.
6. Psccveries of added TCD most be between 30 and 120%
F-81
-------
- 69 -
Table A-8
ESTIMATED DIETARY INTAKE OF FAT FROM MILK AND DAI5S" PRODUCTS
Milk or Dairv Product
art/person/day*
% fat**
qm fat/day
Fresh Fluid Milk
383.31
4.
15.3
Processed Milk(l)
58.10
8.
4.6
Cream (2)
5.35
28.
1.5
Frozen Milk Desserts(3)
53.14
5.
2.7
Cheese (4)
2i.a
29.
6.3
Butter
7.93
81.
6.4
Other
25.58
25.(5)
6.4
Total Dietary Fat Intake = 43 gms fat/person/day.
* Reference 23.
** Pesticide Analytical Manual, U.S. Food and Drug Administration.
(1) Includes condensed- and evaporated milks.
(2) Includes table- and Whipped creams.
(3) Iced Milk
(4) Includes cream-, spread- and swiss cheeses.
(5) Figure given is a mean of the other reported fat levels.
F-82
-------
APPENDIX G
THE CARCINOGEN ASSESSMENT GROUP'S
METHOD FOR DETERMINING THE UNIT RISC ESTIMATE
PARTICIPATING MEMBERS
Elizabeth L. Anderson, Ph.D.
Larry Anderson, Ph.D.
Dolph Arnicar, 8.A.
Steven 3ayard, Ph.D.
David L. 3ayliss, M.S.
Chao W. Chen, Ph.D.
John R. Fowle III, Ph.D.
Bernard Haberman, D.Y.M., M.S.
Charalingayya Hiremath, Ph.D.
Chang S. Lao, Ph.D.
Robert McGauaHy, Ph.D.
Jeffrey Roser.jiatt, B.S.
Dharm V. Singh, D.Y.M., Ph.D.
Todd W. Thorslund, Sc.D.
FOR AIR POLLUTANTS
f3t*Roy E. Albert, M.D
Roy E. Albert,
Chai man
July 31, 1980
G-l
-------
METHOO FOR DETERMINING THE UNIT RISK ESTIMATE
FOR AIR POLLUTANTS
INTRODUCTION
The unit risk estimate for an air pollutant is defined as the lifetime
cancer risk occurring in a hypothetical population in which all individuals are
exposed continuously from birth throughout their lifetimes to a concentration of
1 ug/m^ of the agent in the air which they breathe. This calculation is done
to estimate in quantitative terms the effectiveness (potency) of the agent as a
carcinogen. Unit risk estimates are used for two purposes: 1) to compare the
carcinogenic potency of several agents with each other, and 2) to give a crude
indication of the population risk which might be associated with air exposure to
these agents, if the actual exposures are known. These two uses have different
1 imitations.
In order to use these estimates intelligently, the nature of the source data
used and the assumptions necessary to derive the estimates must be clearly
understood. This appendix discusses the general approach and the assumptions
common to most unit risk estimates. The credence one can ascribe to each risk
estimate depends heavily on the quality of the studies on which the estimate is
based and the relevance of these studies to the evaluation of air exposure in
humans.
PROCEDURES FOR DETERMINATION OF UNIT RISK
The data used for the quantitative estimate is one or both of two types: I)
lifetime animal studies, and 2) human studies where excess cancer risk has been
associated with exposure to the agent. In animal studies it is assumed, unless
evidence exists to the contrary, that if a carcinogenic response occurs at the
dose levels used in the study, then responses will also occur at all lower coses
G-2
-------
with an incidence determined by the extrapola" on model discussed below.
A. Choice of Model
There is no really solid scientific basis for any mathematical extrapolation
model which relates carcinogen exposure to cancer risks at the extremely low
concentrations that must be dealt with in evaluating environmental hazards. For
practical reasons such low levels of risk cannot be measured directly either by
animal experiments or by epidemiologic studies* We must, therefore, depend on
our current understanding of the mechanisms of carcinogens 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 agents which cause cancer
also cause irreversible damage to DNA. This position is reflected by the fact
that a very large proportion of agents which cause cancer are also mutagenic.
There is reason to expect the quanta! type of biological response that is
characteristic of mutagenesis is associated with a linear non-threshold
dose-response relationship. Indeed, there is substantial evidence from
mutagenesis studies with both ionizing radiation and a wide variety of chemicals
that this type of dose-response model is the appropriate one to use. This is
particularly true at the lower end of the dose-response curve; at higher doses,
there can be an upward curvature probably reflecting the effects of multistage
processes on the mutagenic response. The linear non-threshold dose-response
relationship is also consistent with the relatively few epidemiological studies
of cancer responses to specific agents that contain enough information to make
the evaluation possible (e.g., radiation induced leukemia, breast and thyroid
cancer, skin cancer induced by arsenic in drinking water, liver cancer induced
by aflatoxin in the diet). There is also some evidence from animal
experiments that is consistent with the linear non-thresnold model (e.g., liver
G-3
-------
tumors induced in mice by 2-acetyl ami nofl uorene in the large scale EDoi study
at the National Center for Toxicological Research and the initiation stage of
the two-stage carcinogenesis model in rat liver and mouse skin).
Because it has the best, albeit limited, scientific basis of any of the
current mathematical extrapolation models, the linear non-threshold model has
been adopted as the primary basis for risk extrapolation to low levels of the
dose-response relationship. 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 it
could be smaller.
The mathematical formulation chosen to describe the linear, non-threshold
dose-response relationship at low doses is the improved multistage model
developed by Dr. K. Crump. This model employs enough arbitrary constants to be
able to fit almost any monotonically increasing dose-response data and it
incorporates a procedure for estimating the largest possible linear slope (in
the 95* confidence limit sense) at low extrapolated doses that is consistent
with the data at all dose levels of the experiment.
B. Description of the Extrapolation Model
Let P(d) represent the lifetime risk (probability) of-cancer at dose d. The
multistage model has the form
P(d) = 1 - exp [-(qo + qid + q2 0, i = 0, 1, 2, .... k
G-4
-------
Equivalent^,
A(d) = 1 - exp C-(qid + q2^2 + ... + )3
where
A(d) = P(d) - P(o),
1 - P(o)
1s the extra risk over background rate at dose d.
The point estimate of the coefficients q-j, 1 = 0, 1, 2, k and
consequently the extra risk function A(d) at any given dose d is calculated by
maximizing the likelihood function of the data.
The point estimate and the 95% upper confidence limit of the extra risk A(d)
are calculated by using the computer program GLOBAL 79 developed by Crump and
Watson (1979). The calculation proceeds as follows: Let Lo be the maximum
value of the log-1iklihood function. The 955 upper confidence limit for the
extra risk A(d) has the form
Au(d) 3 1 - exp [-(qi*d +• q£d^ + ... + q^d^)]
where q| is calculated by increasing q^ to a value q^* such that when
the log-1iklihood is remaximized subject to this fixed value q]* for the
linear coefficient, the resulting maximum value of the log-likelihood Lj
satisfies the equation
2 (Lo - Li) = 2.70554
where 2.70554 is the cumulative 90% point of the chi-square distribution with
one degree of freedom, which corresponds to a 95" upper limit (one-sided). The
G-5
-------
quantities $2, Q3, • qk are the maximum likelihood estimates of the
other coefficients given q^ equal to qj*. This approach of computing the
upper confidence limit for the extra risk A(d) is an improvement on the Crump et
al. (1977) model. At low doses, the exponent g(d) = q^*d + ^d^ + ...
+^dk is dominated by the linear term qi*d and hence
Au(d) 3 1 - exp(- q^*d) a q^*d
Therefore, the upper confidence limit for the extra risk A(d) at low doses is
always linear. This is conceptually consistent with the linear non-threshold
concept discussed earlier. The slope qx* is a measure of the potency of the
chemical in inducing cancer at low doses.
In fitting the dose-response model, the number of terms in the polynomial
g(d), is chosen equal to (h-1) where h is the number of dose groups in the
experiment including the control group.
Whenever the multistage model does not fit the data sufficiently well, data
at the highest dose is deleted and the model is refitted to the rest of the
data. This is continued until an acceptable fit to the data is obtained. To
determine whether or not a fit is acceptable, the chi-square statistic
X2 = [ (X,- - NiPO2
z Mi (1 - Pi)
i=l
is calculated where M-j is the number of animals in the itf1 dose group,
is the number of animals in the ith dose group with a tumor response, Pi is
the probability of a response in the i^ dose group estimated by fitting the
multistage model to the data, and h is the number of remaining grauos. The fit
is determined to be unacceptable whenever is larger than the cunulative 99»
G-6
-------
point of the chi-square distribution with f degree of freedom, where f equals
the number of dose groups minus the number of non-zero multistage coefficients.
SELECTION AND FORM OF DATA USED TO ESTIMATE PARAMETERS IN THE EXTRAPOLATION
MODEL
For some chemicals, several studies in different animal species, strains,
and sexes each run at several doses and different routes of exposure are
available. A choice must be made of which of the data sets from several studies
to use in the model. It is also necessary to correct for metabolism differences
between species and absorption factors via different routes of administration.
The procedures used in evaluating these data are consistent with the approach of
making a maximum-!iicely risk estimate. They are listed below.
1. The tumor incidence data are separated according to organ sites or tumor
types. The set of data (I.e., dose and tumor incidence) used in the model is
the set where the incidence is statistically significantly higher than the
control for at least one test dose level and/or where the tumor incidence rate
shows a statistically significant trend with respect to dose level. The :ta
set which gives the highest estimate of the lifetime carcinogenic risk qj*
is selected in most cases. However, efforts are made to exclude data sets which
produce spuriously high risk estimates because of a small number of animals.
That is, if two sets of data show a similar dose-response relationship and one
has a very small sample size, the set of data which has larger sample size is
selected for calculating the carcinogenic potency.
2. If there are two or more data sets of comparable size which are identical
with respect to species, strain, sex, and tumor sites, the geometric mean of
the exponent g(d), estimated from each of these data sets and evaluated at a
specific dose d, is used for risk assessment. The geometric mean of numoers
G-7
-------
A^, A2, ..., A,,, is defined as
(Aj x Aj x ... x A^.)!/1"
3. If two or Tiors significant tumor sites are observed in the same study and
if the data are available to us, the number of animals with at least one of the
specific tumor sites under consideration is used as incidence data in the model.
4. Following the suggestion of Mantel and Schneider-man (1977) we assume
that mg/surface area/day is an equivalent dose between species. Since to a
close approximation the surface area is proportional to the 2/3 rds power of the
weight as would be the case for a perfect sphere, the exposure in mg/day per 2/3
rds power of the weight is also considered to be an equivalent exposure. In an
animal experiment this equivalent dose is computed in the following manner.
Let
Le * duration of experiment
le = duration of exposure
m » average dose per day in mg during administration of the agent
(i.e., duri ng le)
W * average weight of the experimental animal
G-8
-------
Then, the lifetime average exposure is
Often exposures are not given in units of mg/day and it becomes necessary to
convert the given exposures into mg/day. For example in most feeding studies
exposure is in terms of ppm in the diet. In this case the exposure in mg/day is
m = ppm x F x r
where ppm is parts per million in the diet of the carcinogenic agent and F is
the weight of the food consumed per day in kgms and r is the absorption
fraction. In the absence of any data to the contrary r is assumed to be equal
to one. For a uniform diet the weight of the food consumed is proportional to
the calories required which in turn is proportional to the surface area or
2/3rds power of the weight, so that
m a-ppm x W2/3 x r or
m n ppm
7*73-
As a result, ppm in the diet is often assumed to be an equivalent exposure
between species. However, we feel that this is not justified since the
calories/!cg of food is very different in the diet of man compared to laboratory
animals primarily due to moisture content differences. Instead we use an
empirically derived food factor f = F/Vi which is the fraction of a species boay
G-9
-------
weight that is consumed per day as food. We use the rates given below.
Species W f
Man 70 0.02S
Rat 0.35 0.05
Mice 0.03 0.13
Thus, when the exposure is given as a certain dietary concentration in ppm the
exposure i n mg/w2/3 i s
_m ppm x F . ppm x fxW . m x f x wl/3
r x W2/3 i?73
When exposure is given in terms of mg/kg/day » m/Wr = s the conversion is
simply
-T7T- = s X «1/3
ryZ/3
When exposure is via inhalation, the calculation of dose can be considered
for two cases where 1) the carcinogenic agent is either a completely water
soluble gas or an aerosal and is absorbed proportionally to the amount of air
breathed in, and 2) where the carcinogen is a poorly water soluble gas which
reaches an equilibrium between the air breathed and the body compartments.
After equilibrium is reached, the rate of absorption of these agents is expected
to be proportional to the metabolic rate, which in turn is proportional to the
rate of oxygen consumption,- which in turn is a function of surface area.
G-10
-------
Case 1
Agents that are in the form of particulate matter or virtually completely
absorbed gases such as SO2 can reasonably be expected to be absorbed
proportional to the breathing rate. In this case the exposure in mg/day may be
expressed as
m = I x x r
where
I = inhalation rate per day in m3
v = mg/nt3 of the agent in air
r = the absorption, fraction
The inhalation rates, I, for various species can be calculated from the
observations (FASEB 1974) that 25 gm mice breathe 34.5 liters/day and 113 gm
rats breathe 105 liters/day. For mice and rats of other weights, W (in
kilograms), the surface area proportionality can be used to find breathing rates
in rn^/day as follows:
For mice, I = 0.0345 (W/0.025)2/3 m3/day
For rats, I = 0.105 (W/0.113)2/3 nP/day
For humans, the values of 20 u»2/dayt is adopted as a standard breathing rata
(ICRP 1977).
The equivalent exposure in ma/y2/3 for-these agents can be derived frcm
the air intake data in way analogous to the food intake data.
'From "Recommendation of the International Commission on Radiolcaical
Protection", page 9, the average breathing rate is 10"7 cm- per 3 hour work
^ay and 2 x 107 cm3 in 24 hours.
G—11
-------
The empirical factors for the air intake per kg per day, i = I/V< based upon
the previous stated relationships are tabulated below
Species W i = I/W
Han ~7U— 0.23
Rat 0.35 0.54
Mice 0.03 1.3
Therefore, for particulates or completely absorbed gases, the equivalent
exposure in mg/w2/3 is
m , Ivr _ iWvr _ ,,,1/3
wZ7T ^7T "w273 u
In the absence of experimental information or a sound theoretical argument to
the contrary, the fraction absorbed, r, is assumed to be the same for all
species.
Case 2
The dose in mg/day of partially soluble vapors is proportional to the O2
consumption which in turn is proportional to W^/3 and also proportional to the
solubility of the gas in body fluids, which can be expressed as an absorption
coefficiennt r for the gas. Therefore, expressing the O2 consumption as O2
= kw2/3 where k is a cconstant independent of species, it.follows that
m
= 1c W2/3 x v x r
or d 3 JB_ = kvr
As with Case 1, in the absense of experimental information or a sound
theoretical argument to the contrary, the absorption fraction, r, is assumed t:
be the same for all species. Therefore, for these substances a certain
G—12
-------
concentration in ppm or ug/m^ in experimental animals is equivalent to the
same concentration in humans. This is supported by the observation that the
minimum alveolar concentration that is necessary to produce a given "stage" of
anesthesia is similar in man and animals (Dripps, et al. 19/5). When the
animals were exposed via the oral route and human exposure is via inhalation or
vice-versa, the assumption is made, unless there is pharmacokenetic evidence to
the contrary, that absorption is equal by either exposure route.
5. If the duration of experiment (Le) is less than the natural lifespan of
the test animal (L), the slope qi* or more generally the exponent g(d) is
increased by multiplying a factor (L/Le)3. We assume that if the average
dose, 0, is continued, the age specific rate of cancer will continue to increase
as a constant function of the background rate. The age specific rates for
humans increases at least by the 2nd power of the age and often by a
considerably higher power as demonstrated by Doll (1971). Thus, we would expect
the cumulative tumor rate to increase by at least the 3rd power of age. Using
this fact we assume that the slope q^* or more generally the exponent g(d),
would also increase by at least the 3rd power of age. As a result, if the slope
qi* Cor g(d>] is calculated at age Le, we would expect that if the
experiment had been continued for the full lifespan, L, at the given average
exposure, the slope qj* [or g(d)] would have been increased by at least
(L/le)3.
This adjustment is conceptually consistent to the proportional hazard model
proposed by Cox (1972) and the time-to-tumor model considered by Crump et al.
(1979) where the probability of cancer at age t and dose d is given by
P(d,t) = 1 - exoC-f(t) x g(d)l
G-13
-------
This more refined approach would be used in the calculations of unit risk wi* n
the data are available.
CALCULATION OF OF THE UNIT RISK
The risk associate with d mg/kg2/3/day as noted previously is
Ay(d) = l-expC-(qi* d + q2 d2 + ... + q^d^)]
A "unit risk" in units X is simply the risk corresponding to an exposure of X =
1. To estimate this value we simply find the number of mg/kg2/3/day
corresponding to one unit of X and substitute this value into the above
relationship. Thus, for example if X is in units of ug/m3 in the air we have
that for
case (1) d = 0.29 x 7Q1/3 x 10~3 = 1.195 x 10"3
and for case (2) d = 1 when ug/m3 is unit used to compute parameters in
animal experiment.
If exposures are given in terms of ppm in air we may simply use the fact
that
1 ppm = 1.2 x molecular weight (gas) mg/m3
molecular weignt lair)
Note, an equivalent method of calculating unit risk would be to use mg/kg for
the animal exposures and then increase the jth polynomial coefficient by an
amount
(Wh/Wa)j/3 j = 1,2,... ,k
and use mg/kg equivalents for the unit risk values.
6-14
-------
ESTIMATION OF UNIT ore;* RASED ON HUMAN DATA
If human epidemiology studies and sufficiently valid exposure information
are available for the compound, they are always used in some way. If they show
a carcinogenic effect, the data are analyzed to give an estimate of the linear
dependence of cancer rates on lifetime average dose, which is equivalent to the
factor Bh» If they show no carcinogenic effect when positive animal evidence
is available, then it is assumed that a risk does exist but it is smaller than
could have been observed in the epidemiology study, and an upper limit of the
cancer incidence is calculated assuming hypothetically that the true incidence
is just below the level of detection in the cohort studied, which is determined
largely by the cohort size. Whenever possible, human data are used in
preference to animal bioassay data.
In human studies, the response is measured, in terms of the relative risk of
the exposed cohort of individuals compared to the control group. In the
analysis of this data it is assumed that the excess risk, or relative risk minus
one, R(X^) -1,1s proportional to the lifetime average exposure, X^, and
that it is the same for all ages. It follows that the lifetime risk in the
general population exposed to a lifetime average concentration X2, PU2), is
equal to CR(X^) - l^/X^ multiplied by the lifetime risk at that site in
the general population. The unit risk estimate is the value of P when X2 is 1
ug/rn^. Except for an unusually well documented human study, the confidence
limit for the excess risk P is not calculated, due to the difficulty of
accounting for the uncertainty inherited in the data (exposure and cancer
response).
G-I5
-------
INTERPRETATION OF UNIT RISK
The unit risk estimate is a rough indication of the relative potency of a
given agent compared with other carcinogens. The comparative potency of
different agents is more reliable when the comparison is based on studies in the
same test species, strain, and sex and by the same route of exposure, preferably
by inhalation.
For several reasons the unit risk estimate is oniy an approximate indication
of the absolute risk in populations exposed to known air concentrations. First,
there are important species differences in uptake, metabolism, and organ
distribution of carcinogens, as well as species differences in target site
susceptibility, immunological responses, hormone function, dietary factors, and
disease. Secondly, the concept of equivalent doses for humans compared to
animals on a mg/surface area basis is virtually without experimental
verification regarding carcinogenic response. Finally, human populations are
variable with respect to genetic constitution and diet, living environment,
activity patterns, and other cultural factors.
6-16
-------
REFERENCES
Albert, R.E.„et aT. 1977. Rationale developed by the Environmental Protection.
Agency for the assessment of carcinogenic risks. J. Mat!. Cancer Inst.
58:1537-1541.
Cordle, F.f P. Corneliussen, C. Jellinek, B. Hackley, R. Lehman, J.
McLaughlin, R. Rhoden, and R. Shapiro. 1978. Human exposure to
polychlorinated biphenyls and polybrominated biphenyls. Environ. Health
Perspect. 24:157-172.
Cox, C.R. 1972. Regression model and life tables. J. Roy. Stat. Soc. B
34:187-220.
Crump, K.S., H.A. Guess, and L.L. Deal. 1977. Confidence intervals and test of
hypotheses concerning dose-response relations inferred from animal
carcinogenicity data. Biometrics 33:437-451.
Crump, K.S. 1979. Dose-response problems in carcinogenisis. Biometrics
35:157-157.
Crump, K.S., W.W. Watson. 1979. GLOBAL 79. A fortran program to extrapolate
dichotomous animal carcinogenicity data to low dose. Mat!. Inst, of
Environ. Health Science. Contract No. l-ES-2123.
Crump, K.S. 1980. An improved procedure for low-dose carcinogenic risk
assessment from animal data. J. of Environ. Pathology and Toxicology (in
preparation).
Doll, R. 1971., Weibull distribution of cancer. Implications for models of
carcinogenesis, J. Roy. Statistical Soc. A 13:133-166.
Oripps, Robert D., J.E. Eckenhoff, and L.D. Yandam. 1977. Introduction to
anesthesia, the principles of safe practice. 5th Ed. W.3. Saunders Company,
Phil. Pa. pp. 121-123.
FASEB. 1974. Biological Data Books, 2nd ed. Yol. III. Edited by Philip L.
Altman and Dorothy S. Dittmen. Federation of American'Societies for
Experimental Biology. Bethesda, MD. Library of Congress No. 72-87738.
Guess, H., et al. 1977. Uncertainty estimates for low dose rate extrapolations
of animal carcinogenicity data. Cancer Res. 37:3475-3483.
Interagency Regulatory Liaison Group. 1979. Scientific bases for identifying
potential carcinogens and estimating their risks. Feb. 6, 1979.
International Commission on Radiological Protection. 1977. Recommendation of
the International Commission on Radiological Protection, Pub. No. 25,
adopted Jan. 17, 1977. Pergammon Press, Oxford, England.
Mantel, M., and M.A. Schneideman. 1975. Estimating "Safe Levels, A Hazardous
Undertaking. Cancer Res. 35:1379-1236.
G—17
-------
MCI. 1976. Guidelines fo carcinogen bioassay in small rodents. National
Cancer Institute Carcinoaenesis Technical Report Series, No. 1 February
1976.
U.S. EPA. 1975. Interim procedures and impact assessments of suspected
carcinogens. Federal Register 41 -.21^02, May 25, 1976.
G-18
------- |