DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
PRELIMINARY DRAFT
Hazard And Dose-Response
Assessment And Characterization
Atrazine
May 22, 2000
U.S. Environmental Protection Agency
Office of Pesticide Programs
Health Effects Division (7509C)
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5 DISCLAIMER
6
7 This document is a preliminary draft: It has not been formally released by the
8 U.S. Environmental Protection Agency and should not at this stage be construed to
9 represent Agency policy, and should not be interpreted as intent to regulate. It is being
10 circulated for comment on its technical accuracy. Mention of trade names or
11 commercial products does not constitute endorsement or recommendation for use.
-------
DRAFT: DO NOT CITE OR QUOTE
1 Table of Contents
2
3 Preface i
4
5 Introduction ii
6
7 List of Acronyms iv
8
9 1. Summary of Effects 2
10 1.1 Effects Attributable to Treatment of Rats and Mice with Atrazine 2
11 1.2 Carcinogenic Effects 3
12 1.2.1 Mammary Carcinomas 3
13 1.2.2 Mammary Fibroadenomas 6
14 1.2.3 Pituitary Tumors 6
15 1.3 Potential Antecedents to Carcinogenicity 8
16 1.3.1 Attenuation of the LH Surge 9
17 1.3.2 Estrous Cycle Disruptions 12
18 1.3.3 Effects on Pituitary Weights 12
19 1.3.4 Histomorphology of Mammary Tissue 13
20 1.4 Mutagenic and Estrogenic Activity 15
21 1.5 Structure Activity Relationships 16
22 1.6 Doses Associated with Effects 16
23 1.7 Chronic, Developmental, and Reproductive Toxicity 16
24 1.7.1 Chronic and Subchronic Toxicity of Atrazine 19
25 1.7.2 Developmental Toxicity of Atrazine 20
26 1.7.3 Reproductive Toxicity of Atrazine 21
27 1.7.4 Special Studies 22
28
29 2. Hazard Characterization And Mode of Action Analysis 26
30 2.1 Human Cancer Studies 26
31 2.2 Carcinogenicity in Female SD Rats 27
32 2.3 Postulated Mode of Carcinogenic Action 28
33 2.3.1 Reproductive Aging in Rats 28
34 2.3.2 Atrazine Effects Relevant to Carcinogenicity 30
35 2.4.1 Key Events 34
36 2.4.2 Correlation of Effects and Dose 36
37 2.4.3 Temporal Association of Effects 39
38 2.4.4 Biological Plausibility and Coherence of the Database 40
39 2.4.5 Other Modes of Action 43
40 2.4.6 Uncertainties and Limitations 46
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.4.7 Preliminary Conclusions on the Postulated Mode of Carcinogenic
2 Action 47
3 2.5 Reproductive and Developmental Toxicity 49
4
5 3. Science Policy Considerations: Human Relevance, Children's Health
6 Concerns, and Dose-Response Analysis 54
7 3.1 Human Relevance 54
8 3.1.1 Potential Neuroendocrine Disruption 54
9 3.1.2 Potential Human Health Consequences Associated with Altered
10 GnRH/Pituitary Function • 56
11 3.1.3 Potential Cancer Risk Associated with Altered GnRH/Pituitary
12 Function 59
13 3.1.4 Breast Cancer 63
14 3.1.5 Cancer Classification 65
15 3.2 Potential Health Effects of Atrazine in Children 66
16 3.2.1 Reproductive/Developmental Hazard 66
17 3.2.2 Cancer Hazard 68
18 3.2.3 Summary of Children's Health Concern 69
19 3.3 Summary of Atrazine Human Hazard Potential 69
20 3.4 Dose-Response Analysis 71
21 3.4.1 Selecting a Point of Departure 72
22 3.4.2 Point of Departure Using LED10 From The Tumor Data 75
23 3.5 Summary and Conclusions on the Proposed OPP Science Policy
24 Positions: Mode of action, Human Relevance, Children's Health
25 Concerns, and Dose-Response Extrapolation 76
26 3.5.1 Postulated Rat Tumor Mode of Action 77
27 3.5.2 Relevance of Rat Mode of Action to Humans and Carcinogenicity
28 Classification 78
29 3.5.3 Children's Hazard 79
30 3.5.4 Dose-Response 80
31 3.6 Other Reviews 80
32
33
34
35
36
-------
DRAFT: DO NOT CITE OR QUOTE
1 List of Tables
2
3 Table 1-1. Carcinogenicity Bioassays with Atrazine: Incidence and Onset of
4 Mammary Adenomas/Carcinomas in Female SD Rats 4
5 Table 1-2. Carcinogenicity Bioassays with Atrazine: Incidence and Onset of
6 Mammary Fibroadenomas in Female SD Rats 5
7 Table 1-3. Carcinogenicity Bioassay: Incidence of Pituitary Adenomas (Thakur,
8 1991a) 8
9 Table 1-4. LH Data (mean + sd) from Animals Repeatedly Bled in the One-Month
10 Study (Morseth, 1996a) (LH values given are in picograms/mL) 10
11 Table 1-5. LH Data (mean ± sd) from Animals Repeatedly Bled in the Six-Month
12 Study (Morseth, 1996b) (doses are in LH values given are in
13 picograms/mL 10
14 Table 1-6. Percentage of Days (± sd) in Estrus for SD Females Following Six-Month
15 Exposure to Atrazine through the Diet (Morseth, 1996b) 12
16 Table 1-7. Effects of Atrazine Treatment on Group Mean Absolute Pituitary Weights
17 (mg ± sd) in Female SD Rats (Thakur, 1991a) 13
18 Table 1-8. NOAELs and LOAELs (mg/kg/day) Associated with Neoplastic Responses
19 of Female SD Rats Treated with Atrazine 17
20 Table 1-9. NOAELs and LOAELs (mg/kg/day) Associated with Non-Neoplastic
21 Responses in Female SD Rats Treated with Atrazine 18
22 Table 1-10. NOAELs/LOAELs (mg/kg/day) for Reproductive and Developmental
23 Effects Following Treatment of Dams or Offspring of Several Rat Strains
24 with Atrazine or its Metabolites 25
25 Table 2-1. LOAELs for Tumor Formation and Non-Neoplastic Effects in Female SD
26 Rats 37
27 Table 2-2. Lowest NOAELs/ LOAELs (mg/kg/day) for Reproductive and
28 Developmental Effects Following Short-term (1-30 Days) Treatment of
29 Rats During Various Stages of the Reproductive Cycle with Atrazine or its
30 Metabolites 53
31 Table 3-2. LED10s in Human Equivalents (And Revised Q*) 76
32 Table 3-3. Other Reviews on the Carcinogenicity of Atrazine 82
33
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 List of Figures
3
4 Figure 1-1. Cumulative Incidence of Pituitary p-Adenomas (Thakur, 1991 a) 7
5 Figure 1-2. Effects of Atrazine Treatment on the LH Surge in Female SD Rats After
6 Six Months of Dosing 11
7 Figure 2-1. Postulated Effects of Atrazine Treatment on the Hypothalamic-Pituitary-
8 Ovarian Axis 33
9 Figure 2-2. Temporal Pattern of Atrazine Effects 41
10 Figure 3-1. Atrazine's Neuroendocrine Mode of Action: Potential Adverse Outcomes
11 71
12 Figure 3-2. Key Endocrine-Related Effects Following Atrazine Treatment of Rats . 72
13
-------
DRAFT: DO NOT CITE OR QUOTE
1 Preface
2
3 Over the last several years there has been increasing concern about the possible
4 human health effects posed by chemicals that may alter the normal function of the
5 endocrine system. Within the scientific community there is much debate and discussion
6 about the extrapolation of animal findings on endocrine disrupters to predict and quantify
7 such potential effects in humans, including children.
8
9 Agency guidance regarding endocrine perturbations in health risk assessment is
10 limited to thyroid follicular cell carcinogens (US EPA, 1998a). Laboratory animal studies
11 available on atrazine indicate that its mode of action in rats involves a perturbation of the
12 neuroendocrine system that results in prolonged exposure to endogenous estrogen and
13 prolactin. This endogenous exposure to estrogen leads to carcinogenic effects on the
14 mammary and pituitary gland. There are also animal data available showing that there is
15 an association between the adverse effects of atrazine on neuroendocrine control of
16 reproductive developmental function. Given the complexity and multiplicity of effects
17 that result from exposure to atrazine, the Office of Pesticide Programs (OPP) is at a
18 point in its assessment of atrazine where external peer review by the FIFRA Scientific
19 Advisory Panel (SAP) would facilitate further development and refinement of the draft
20 health assessment document. Furthermore, very little is understood about the long term
21 consequences that may result from prenatal and early postnatal exposures to
22 neuroendocrine-perturbing chemicals. Thus, presenting the atrazine health assessment
23 to the SAP at this time also allows the OPP an opportunity to obtain comments on the
24 adequacy of the approach taken by OPP to address potential hazard to children.
25
26 The aim of the SAP review is to obtain advice and comment on the draft
27 document on specific science issues, such as: what factors should be considered in
28 evaluating this particular neuroendocrine mode of action?; what are the relevance and
29 implications of this type of perturbation in humans?; what are the key biological events
30 driving the hazard concern; and what are the potential cumulative effects and hazards
31 on the developing brain that could result from the effects of atrazine on the function of
32 the endocrine system? This external scientific peer review is a significant and critical
33 step as the OPP proceeds to develop a sound and scientifically credible health risk
34 assessment on atrazine as part of the mandate under the 1996 Food Quality Protection
35 Act to protect public health and the environment. OPP intends to use the SAP's
36 comments, as well as public comments that are received to further refine this draft
37 document. Thus, the conclusions and analyses presented here within are considered
38 preliminary.
-------
DRAFT: DO NOT CITE OR QUOTE
1 Introduction
2
3 Over 10 years ago, atrazine was found to induce mammary gland tumors in
4 Sprague Dawley female rats (Mayhew, 1986). Shortly afterwards, the Office of
5 Pesticide Programs (OPP) classified atrazine as a possible human carcinogen (Group
6 C) based on "limited evidence for the oncogenicity of the chemical in rats" (Hauswirth,
7 1988a,b). In 1988, OPP asked the FIFRA Scientific Advisory Panel (SAP) to comment
8 on the cancer classification. The SAP agreed with OPP's classification of atrazine as a
9 Group C carcinogen. The 1988 SAP also raised the possibility of a hormonal mode of
10 action underlying atrazine's carcinogenicity (Copley, 1988). Accordingly, OPP
11 encouraged the registrant of atrazine to pursue studies on a potential endocrine
12 mechanism. Since that time, the registrant has completed numerous studies
13 concerning atrazine's potential mode of carcinogenic action to explain the mammary
14 gland tumor response found in female SD rats. The Agency's National Health and
15 Environmental Effects Laboratory has also generated information on atrazine's
16 neuroendocrine effects, as well as its effects on reproductive development in young
17 rats.
18
19 The purpose of this draft document is to update and revise OPP's previous
20 cancer assessment of atrazine by considering new information bearing on it's
21 postulated mode of action. The draft document presents an integrative approach that
22 uses a common neuroendocrine mode of action to evaluate the potential for both
23 cancer and noncancer health effects (especially reproductive and developmental
24 outcomes). This preliminary assessment also addresses how the available mode of
25 action information influences decisions about the human hazard potential including
26 sensitive subpopulations (e.g., children). This draft document is organized into three
27 parts, A, B, and C. Each has its own List of Contents.
28
29 Q Part A summarizes the key conclusions on the cancer and reproductive
30 developmental hazard potential and mode of action, and provides an integrated
31 synthesis and characterization of the main findings:
32
33 » Chapter 1 provides a summary of tumor and other key data supporting the
34 carcinogenicity of atrazine, as well as data on the reproductive
35 developmental effects of atrazine.
-------
DRAFT: DO NOT CITE OR QUOTE
1 > Chapter 2 provides a technical hazard characterization and presents the
2 mode of action analysis. The mode of action analysis is based on a
3 framework described in the Agency's 1999 draft revisions to its guidelines
4 for carcinogen risk assessment (US EPA, 1999a). This framework is used
5 for judging whether the available evidence supports the mode of
6 carcinogenic action in rats postulated for atrazine. This Chapter also
7 discusses the common events in this mode of action which may lead to
8 consequences on reproductive development.
9
10 » Chapter 3 addresses what inferences can be made about the human
11 relevance of the rat based findings on the mode of action conclusions
12 presented in Chapter 2, and discusses whether there is special concern
13 for children. The proposed dose-response extrapolation approach for
14 cancer is also presented.
15
16 Q Part B of the document (Chapters 4-9) presents a detailed carcinogenicity
17 assessment and evaluation of the available epidemiology, toxicology,
18 metabolism, mutagenicity, and mode of action studies on atrazine that are
19 summarized in Chapter 1 of Part A.
20
21 Q Part C of the document (Chapters 10-13) presents an evaluation of special
22 reproductive/developmental studies performed on atrazine, as well as a review of
23 available reproductive epidemiology studies.
in
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Cl C
CL C
DA D
DACT D
DMBA D
ED10 E
re
F-344 F
FSH F
GD G
GnRH G
HOT H
LE L
LED10 Li
a;
LH Li
LOAEL Li
MTD M
NHL ni
NOAEL N
OR 0
OVX 0
PCOS P
PND P
PoD P
PPS P
SD S
PRL P
PIF Pi
Organizations
CARC C
HED Tl
IARC Tl
MARC M
NTP N
SAP Si
List of Acronyms
Confidence Ratio
Corpea Lutea
Dopamine
Diaminochlorotriazine
Dimethylbenzanthracene
Effective Dose - Central estimate on a dose associated with a 10%
response adjusted for background
Fischer-344
Follicle Stimulating Hormone
Gestational Day
Gonadotrophin Releasing Hormone
Highest Dose Tested
Long Evans
Lower Limit on a Effective Dose - 95% Lower confidence limit on a dose
associated with 10% response adjusted for background
Lutenizing Hormone
Lowest Observed Adverse Effect Level
Maximum Tolerated Dose
non-Hodgkins Lymphoma
No Observed Adverse Effect Levels
Odds Ratio
Ovariectomized/Ovariectomy
Polycystic Ovarian Syndrome
Postnatal Day
Point of Departure
Preputial Separation
Sprague-Dawley
Prolactin
Prolactin Inhibiting Factor
Cancer Assessment Review Committee
The Office of Pesticide Program's Health Effects Division
The International Agency for Research on Cancer
Metabolism Assessment Review Committee
National Toxicology Program
Scientific Advisory Panel
IV
-------
DRAFT: DO NOT CITE OR QUOTE
PART A
3
4 Preliminary Hazard and Mode of Action
Characterization
-------
DRAFT: DO NOT CITE OR QUOTE
1 Chapter 1
2
3 1. Summary of Effects
4
5 This Chapter summarizes the data discussed in the hazard assessment portion
6 of this document (Parts B and C). The summary forms the basis for the analysis of the
7 mode of carcinogenic action information presented in Chapter 2 and draft OPP science
8 policy positions on human relevance of the animal tumor findings and the classification
9 of atrazine for human carcinogenic potential developed in Chapter 3. This Chapter also
10 presents a summary of data on the reproductive and developmental toxicity of atrazine.
11
12 1.1 Effects Attributable to Treatment of Rats and Mice with Atrazine
13
14 Treatment of female SD rats with atrazine, but not male SD rats or Fischer
15 344 rats or CD-1 mice of either sex, results in neoplastic responses expressed
16 as an increased incidence and/or an early onset of mammary carcinomas and
17 adenomas, mammary fibroadenomas, and pituitary adenomas. Atrazine
18 treatment of female SD rats also leads to certain non-neoplastic responses
19 which precede and some of which may be antecedents to the neoplastic
20 responses. A prominent effect is an attenuation of the luteinizing hormone (LH)
21 surge that is necessary for normal reproductive cycling and a disruption of the
22 estrous cycle. Effects on mammary tissue, namely markers of estrogen and
23 prolactin (PRL) exposure, include increased incidences or increased severity of
24 alveolar development, acinar development, dilated ducts, increased secretory
25 activity, and galactoceles. Prolactin exposure is more strongly associated with
26 the development of mammary fibroadenomas while estrogen exposure is more
27 supportive of the development of adenomas/carcinomas. Estrogen also
28 stimulates prolactin secreting cells and predisposes them to neoplasia. Data
29 from short-term, high-dose studies suggest that a primary site of action of
30 atrazine is the hypothalamus.
31
32 Results of mutagenicity assays mostly are negative. Assays designed to
33 evaluate direct estrogenic activity of atrazine have failed to attribute exogenous
34 estrogenic activity to atrazine treatment. Treatment with the close structural
35 analogues, simazine and propazine, also lead to the formation of mammary
36 tumors in female SD rats. Treatment of male SD rats or CD-1 mice of either sex
37 with these chemicals does not result in an increased incidence of tumors at any
38 site.
-------
DRAFT: DO NOT CITE OR QUOTE
1 1.2 Carcinogenic Effects
2
3 Evidence from epidemiologic studies is not sufficient to establish whether
4 atrazine may be carcinogenic to humans. Therefore, any inferences as to
5 human carcinogenic potential must be determined from animal studies (see Part
6 B, Chapter 4).
7
8 Table 1-1 summarizes the data on the incidence and onset of mammary
9 adenomas/carcinomas found in carcinogenicity bioassays following
10 administration of atrazine to female SD rats. The data generated on the
11 formation of mammary fibroadenomas in female SD rats treated with atrazine is
12 summarized in Table 1-2. These benign tumors are considered separate from
13 mammary carcinomas because they are of a different cell origin than the tubular
14 and glandular adenomas and carcinomas. Carcinomas arise from
15 undifferentiated terminal end buds and terminal ducts of the mammary gland;
16 fibroadenomas arise from more differentiated structures such as alveolar buds
17 and lobules (Russo and Russo, 1996). In addition to increased incidence/early
18 onset of mammary gland tumors, an early onset is found for pituitary adenomas.
19 Table 1-3 summarizes data regarding the associations between atrazine
20 treatment and the formation of pituitary adenomas.
21
22 1.2.1 Mammary Carcinomas
23
24 Treatment of female SD rats with atrazine leads to an increased
25 incidence of mammary carcinomas and adenomas in one and two year
26 bioassays (Mayhew, 1986; Morseth, 1998; Pettersen and Turnier, 1995).
27 Serial sacrifice data show that atrazine treatment of female SD rats results
28 in an early onset of mammary carcinomas (Thakur, 1991 a; Pettersen and
29 Turnier, 1995). Data on time of onset of mammary carcinomas as
30 determined by palpation also show an early onset of mammary
31 carcinomas (Thakur, 1992a; Morseth, 1998). The lowest dose of atrazine
32 associated with an increased incidence in mammary carcinomas is 3.5
33 mg/kg/day (Mayhew et a/., 1986). The NOAEL in the same study was 0.5
34 mg/kg/day.
35
36
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
5
6
7
8
Table 1-1. Carcinogenicity Bioassays with Atrazine: Incidence and
Onset of Mammary Adenomas/Carcinomas in Female
SDRats
Study
Mayhew et
a/., 1986
Thakur,
1991 a
Thakur,
1992a
Morseth,
1998
Pettersen and
Turnier, 1995
Duration
2-year
2-year serial sacrifice
month
9
12
15
18
24
2-year terminal sacrifice
week of onset*
<52
53-78
79-104
0-104
mean wk. onset
2-year
week of onset*
<52
53-78
79-104
0-104
mean wk. onset
1-year serial sacrifice
month
9
12
(no tumors at 3 & 6
mo.)
Tumor Incidence
Dose (mg/kg/day)
0 0.5 3.5 25 50
15/88**16/67 27/69* 27/68* 45/60**
Dose (mg/kg/day)
0 4.23 26.63
0*** 0 4
0 1 2
2 0 1
524
2 1 0
Dose (mg/kg/day)
0 3.79 23.01
0/14* 3/11 6/18*
8/14 3/11 5/18
6/14 5/11 7/18
17/60 13/59 22/60
78.9 72.5 65.4
Dose (mg/kg/day)
0 1.5 3.1 4.2 24.4
1/11 2/15 0/14 2/10 6/23
5/11 6/15 7/14 6/10 7/23
5/11 6/15 7/14 2/10 10/23
12/80 18/80 20/79 14/80 27/80**
72.6 77.2 78.6 64.4 64.8
Dose (mg/kg/day)
0 0.8 1.7 2.8 4.1 23.9
1/10## 1/11 0/10 0/10 0/10 1/10
1/25 1/24 1/25 2/25 2/24 6/25
9
10
11
12
13
14
15
16
17
18
= p=<0.05; **=p=<0 01; at control=trend, at dose group=pairwise versus control; ***per 10 animals;
#=onset as determined by first palpation of a tumor; ## incidences for adenomas and adenocarcinomas
combined.
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
4
5
6
7
Table 1-2. Carcinogenicity Bioassays with Atrazine: Incidence and Onset
of Mammary Fibroadenomas in Female SD Rats
Study
Mayhew et
a/., 1986
Thakur,
1992a
Thakur,
1991 a
Morseth,
1998
Pettersen
and Turnier,
1995
Duration
2-year
2-year terminal sacrifice
week of onset*
<52
53-78
79-104
0-104
mean wk. onset
2-year serial sacrifice
month
9
12
15
18
24
2-year
week of onset*
<52
53-78
79-104
0-104
mean wk. onset
1-year
month
9
12
(no tumors at 3 & 6 mo.)
Tumor Incidence
Dose (mg/kg/day)
0 0.5 3.5 25 50
20/88 24/65 21/69 21/68 20/89
Dose (mg/kg/day)
0 3.79 23.01
2/35 1/27 3/39
16/35 15/27 18/39
17/35 11/27 18/39
39/60 30/59 41/60
76.4 76.1 72.7
Dose (mg/kg/day)
0 4 23 26 63
0## 0 2
1 0 2
2 5 1
244
334
Dose (mg/kg/day)
0 1.5 3.1 4.2 24.4
0/15 1/18 3/26 1/26 1/22
9/15 11/18 13/26 14/26 9/22
6/15 6/18 10/26 11/26 12/22
16/78 25/79* 34/77**29/78* 25/77*
76.1 72.4 73.7 73.3 76.3
Dose (mg/kg/day)
0 0.8 1.7 2.8 4.1 23.9
1/10 0/10 0/100/10 1/10 1/10
1/25 2/24 2/25 0/25 3/24 3/25
8
9
10
11
12
13
14
15
16
*p=<0.05;**p=<0.01; at control=trend, at dose group=pairwise versus control, #=Time of onset as
determined by first palpation of tumor; ## = Incidence per 10 animals.
-------
DRAFT: DO NOT CITE OR QUOTE
1 In one study, treatment of male and female F344 rats with a high dose of
2 about 38 mg/kg/day of atrazine was reported to lead to an. increased incidence of
3 benign mammary tumors in males (Pinter ef a/., 1990). The finding is difficult to
4 evaluate because, among other shortcomings, no control animals survived to
5 study termination, the study covered a lifetime and at approximately 30 months
6 of age when the study was terminated, background mammary tumor incidence in
7 untreated male rats would be expected to be similar to the incidence reported in
8 the high dose group. Further, a separate study with F344 male and female rats
9 did not show atrazine treatment induced the formation of tumors of any kind
10 (Thakur, 1992b).
11
12 1.2.2 Mammary Fibroadenomas
13
14 With one exception (Morseth, 1998), atrazine treatment has not
15 been shown to lead to a statistically-significant (pairwise comparisons,
16 treatment group versus control) increased incidence of mammary
17 fibroadenomas. The apparent increased incidence in fibroadenomas in
18 the single study may not be treatment related because there is no dose-
19 response trend among treatment groups over a 16-fold increase in doses;
20 the control group incidence is low compared to historical control rates; and
21 the incidences in atrazine treatment groups are within historical control
22 ranges. As illustrated in Figure 1-1, data from one serial sacrifice study
23 (Thakur, 1991a) support an association between atrazine treatment and
24 an early onset of mammary fibroadenomas. However, an early onset of
25 mammary fibroadenomas was not evident in the other serial sacrifice
26 study (Pettersen and Turnier,1995). The Thakur data suggest an early
27 onset of mammary fibroadenomas at the lowest atrazine dose
28 administered, 4.23 mg/kg/day.
29
30 1.2.3 Pituitary Tumors
31
32 There are no increases in the incidences of pituitary tumors at the
33 terminal sacrifice (24 month) in any of the carcinogenicity studies
34 performed with atrazine. Because the background incidence of pituitary
35 tumors is in the range of 80-90% at 24 months of age in SD rats, the lack
36 of an increased incidence in pituitary tumors at terminal sacrifice may not
37 be surprising. However, there is evidence for an earlier onset of pituitary
38 tumors at nine and 12 months in female SD rats treated with atrazine in
6
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
one serial sacrifice study (Thakur 1991 a) but not in a second 12-month
study which included a nine month interim sacrifice .(Pettersen and
Turnier, 1995). Figure 1-1 depicts the dose-response data for the
cumulative incidence of pituitary tumors over time and shows that there is
an apparent early onset of pituitary tumors in the Thakur (1991 a) serial
sacrifice study. The information in Figure 1-1 shows that an early onset of
pituitary tumors can be attributed to atrazine treatment at a dose level of
26.23 mg/kg/day. Neither an early onset nor an increased incidence of
pituitary tumors is evident at an atrazine dose level of 4.23 mg/kg/day.
Table 1-3 provides the incidences of pituitary tumors found at each
sacrifice interval.
Figure 1-1. Cumulative Incidence of Pituitary p-Adenomas (Thakur, 1991 a)
25-
g 20j
c
•
1 i«j
3
5J
• Control
D 4.23 mg/kg/day
^26.63 mg/kg/day
9 12 15 18
Sacrifice Interval (Months)
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Table 1-3. Carcinogenicity Bioassay: Incidence of Pituitary Adenomas (Thakur,
1991a)
Sacrifice Time
(Months)
9*
12
15
18
24
Dose (mg/kg/day)
Control
0"
2
5
9
6
4.23
0
2
3
5
6
26.63
2
6
4
6
2
* = No tumors at one and three months; ** = Incidence/10 Animals
1.3 Potential Antecedents to Carcinogenicity
Chronic atrazine treatment of female SD rats leads to the expression of a
number of non-neoplastic neuroendocrine disruptions and of histomorphologic
effects on mammary and pituitary glands. Neuroendocrine effects include
attenuation of LH surges, disruption of the estrous cycles, and an increase in
pituitary weights. Endocrine associated histomorphologic effects on mammary
tissue include increases in the incidences of acinar/lobular development and
secretory activity and severity of galactoceles, in atrazine treated animals.
8
-------
DRAFT: DO NOT CITE OR QUOTE
1 As discussed in Part B, Chapter 9.3, preliminary data implicate the
2 hypothalamic-pituitary axis as a primary site of atrazine toxicity (Cooper et a/.,
3 1996, 1998; Cooper et a/., 2000). Atrazine appears to affect the catecholamine
4 neurotransmitters in the hypothalamus by decreasing norepinephrine (NE) and
5 increasing dopamine (DA) (Cooper etal., 1998). The decrease in NE results in a
6 decrease in gonadotropin releasing hormone (GnRH), with a corresponding
7 diminution of surges of luteinizing hormone (LH). If serum LH levels do not
8 display a proestrus afternoon surge above a critical level then ovulation does not
9 occur, and the ovarian cycle is disrupted. The inhibition of ovulation following
10 continued atrazine exposure leads to maintenance of a state (prolonged or
11 constant estrus) where ovarian follicles continue to secrete estrogen. Removal
12 of the estrogen stimulus by ovariectomy abolishes the induction of mammary
13 tumors by atrazine treatment.
14
15 1.3.1 Attenuation of the LH Surge
16
17 Table 1-4 is a summary from a one month study on the effects of
18 atrazine treatment on the preovulatory surge of LH in female SD rats while
19 Table 1-5 provides a summary of the LH surge effects following six
20 months treatment. Although LH data were collected at several time
21 periods in addition to those shown, table entries are limited to periods
22 when LH blood levels should be near or at baseline values (1100 hours)
23 and the period when LH blood levels should be near or at the peak surge
24 value (1800 hours). Thus, these time periods are appropriate points for
25 evaluating the fold increase in serum LH compared to baseline values and
26 for ascertaining the effects of atrazine on the preovulatory surge.
27
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 1-4. LH Data (mean ± sd) from Animals Repeatedly Bled in the One-Month
Study (Morseth, 1996a) (LH values given are in picograms/mL)
Dose
mg/kg/day
0
2.5
5.0
40
200
1100 Hours
732 ± 461
11 01 ±652
810 ±519
755 ±389
514 ±503
1800 Hours
2650 ± 2389
3015 ±3220
271 7 ±2542
1450 ±857
812 ±470
Fold Increase*
3.6
2.7
3.3
1.9
1.6
'Increase = 1800 hour values (peak values) divided by the 1100 hour values (baseline values)
Table 1-5. LH Data (mean + sd) from Animals Repeatedly Bled in the Six-Month
Study (Morseth, 1996b) (doses are in LH values given are in
picograms/mL)
Dose
mg/kg/day
0
1.8
365
294
1100 Hours
909+410
1075+621
972+353
1005+482
1800 Hours
3336±3138
3631 ±2732
250011897
8581416
Fold Increase*
3.7
34
2.6
<1.0
'Increase = 1800 hour values (peak values) divided by the 1100 hour values (baseline values)
As shown in Table 1-4, treatment of female SD rats with 200
mg/kg/day of atrazine for one month leads to a pronounced attenuation of
the LH surge while treatment with 40 mg/kg/day suppresses the
preovulatory surge to a lesser degree. Treatment with atrazine over a six
month period (Table 1-5) results in effects at lower doses: an abolishment
of the preovulatory surge at 29.4 mg/kg/day and an attenuation of the LH
surge at 3.65 mg/kg/day. Figure 1-2 presents graphically the LH levels
over the entire sampling period (1100 to 2300 hours) in the six month
study. Atrazine treatment suppresses the LH surge in a time and dose
dependent fashion. In other words, lower doses of atrazine require longer
periods of time to produce an attenuation of the LH surge.
10
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
Figure 1-2. Effects of Atrazine Treatment on the LH Surge in Female SD Rats
After Six Months of Dosing
Re peat bleed LH data
0
1 8mg/kg
3 65 mg/kg
29 4 mg/kg
1100 1400 1600 1800 2000
Biological Time (Hours)
2300
11
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
1.3.2 Estrous Cycle Disruptions
In the normal female SO rat, approximately 20-25% of the days of
the estrous cycle are spent in estrus. Atrazine treatment leads to a
disruption of the normal reproductive cycle as evaluated by vaginal
smears (Table 1-6) (Morseth, 1996b). As early as 13 weeks following
initiation of treatment and continuing throughout the remainder of the six
month study, there is a statistically-significant increase in the percentage
of days spent in estrus (control - 31 %; 29.4 mg/kg/day - 40%). By 21 to
22 weeks of treatment, the effect on the days in estrus is also statistically-
significant in animals treated with 3.65 mg/kg/day atrazine (control - 32%,
3.65 mg/kg/day - 45%).
Table 1-6. Percentage of Days (+ sd) in Estrus for SD Females Following Six-
Month Exposure to Atrazine through the Diet {Morseth, 1996b)
Dose
(mg/kg/day)
0
1.8
365
29.4
9-10
weeks
25 ± 9.4
25 ± 4.8
26 ±10.2
26 ± 9.3
13-14
weeKs
31 ± 22.4
28 ±18.0
31 ±21.1
40 ± 27.6*
17-18
weeks
34 + 24.2
33 ± 24.7
36 ±25.1
45 ± 32.1*
21-22
weeks
32 ± 25.4
41±31.9
45+32.2*
51 ± 34.8"
21-26
weeks
47 ± 32.2
48± 35.5
54 + 35.1
63 ± 37.0*
ps 0.05; ** ps 0.01
1.3.3 Effects on Pituitary Weights
Atrazine treatment of female SD rats leads to an early increase in
pituitary weights by nine months (Table 1-7). Pituitary weights were
increased by 54% over control weights at a dose level of 26.23
mg/kg/day. A less pronounced effect was observed at 4.23 mg/kg/day
(25% increase over control pituitary weights) at nine months but not at
other times).
12
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
Table 1-7. Effects of Atrazine Treatment on Group Mean Absolute Pituitary
Weights (mg ±sd) in Female SD Rats (Thakur, 1991 a)
Dose
(mg/kg/day)
Control
4.23
26.23
3 months
23±4
21. 2 ±30.0 (-8%)*
21 ±8 (-11%)
9 months
24 + 6
30 ±6 (+25%)
37 ±8 (+54%)
12 months
37+20
35 ±26 (-4%)
42 + 15 (+13%)
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
"Values in parenthesis represent percent change relative to control
1.3.4 Histomorphology of Mammary Tissue
Endocrine associated histomorphologic effects on mammary tissue
found following treatment of female SD with atrazine include increases in
the incidence and severity of acinar development, acinar/lobular
development, secretory activity, dilated ducts with secretion, and
galactoceles. Each of these effects are considered to be associated with
exposure of mammary tissue to estrogen and/or prolactin (Part B, Chapter
9).
The incidences and severity of acinar development, which is
primarily associated with estrogen secretion, seemed to be increased at
three and nine months in both the low and high dose groups.
Secretory activity is primarily associated with prolactin exposure.
At nine months, incidences of animals determined to have increased
incidence and severity of secretory activity increased as a function of
increasing atrazine dose-levels.
The development of dilated ducts is primarily influenced by
prolactin secretion. The incidences and severity of dilated ducts (with
secretion) increased markedly at the low and high dose at nine months
and at the high dose at 12 months. There is also a suggestion that the
incidence of lesions of ducts was increased at the high dose at three
months.
13
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 The incidence and severity of galactoceles, primarily a marker of
3 prolactin secretion, were reported to increase at both nine and 12 months
4 in a serial sacrifice study (Thakur, 1991 a). This increase is pronounced in
5 the 26.23 mg/kg/day atrazine treatment group. The response at the 4.23
6 mg/kg/day does not indicate a treatment-related effect.
7
8 An examination of the individual animal data from Thakur (1991 a)
9 is quite useful in demonstrating the relationships between mammary and
10 pituitary tumors, pituitary weights, and histomorphological indications of
11 hormone exposure in the mammary gland. Also evident, when individual
12 animal data from the nine month time point in this study is examined, is
13 the early onset of these parameters. Appendix Tables 27, 28, and 29
14 display these parameters for each individual animal at the nine month
15 time point in this study.
16
17 Early onset of tumors is clear from comparing the control to
18 atrazine-treated animal data displayed in Appendix Tables 27,28 and 29.
19 None of the ten control animals at this time point had a mammary or
20 pituitary tumor while five of ten and two of ten 400 ppm animals had a
21 mammary tumor or pituitary tumor respectively. Early onset of
22 histomorphologic markers of hormone exposure of the mammary gland is
23 also evident. Only one of the ten control animals had a galactocele or
24 had index weighted scores of 3 or greater for secretory activity or dilated
25 ducts with secretion at nine months. At 400 ppm, eight of the ten animals
26 had galactoceles and eight of the ten had weighted index scores of either
27 three or four for secretory activity or dilated ducts with secretion.
28
29
14
-------
DRAFT: DO NOT CITE OR QUOTE
1 The relationship of these parameters to each other is clear when
2 the data from each animal at these time points is examined. For example,
3 the one and only animal which had a galactocele in the control group, also
4 was the only animal with a three or four weighted index score for
5 secretory activity and dilated ducts and also had the heaviest pituitary in
6 the group. A pituitary would be expected to be enlarged due to lactotroph
7 hyperplasia. Lactotroph hyperplasia is associated with increased prolactin
8 secretion; thus, the animal with the heaviest pituitary is secreting the most
9 prolactin and this is why it is the only animal in the group with a
10 galactocele and high scores for markers of prolactin exposure in the
11 mammary gland. Similar examples can be found in the 70 ppm group
12 where the two animals with the heaviest pituitaries both had galactoceles.
13 Two other animals in this group also had a galctocele, but had pituitaries
14 that were close to the average pituitary weight of the group. Though the
15 pituitaries in these two animals did not weigh an exceptional amount, they
16 were the only two animals in this group in which histopathology detected
17 increased focal hyperplasia of the pituitary. Thus, all four animals with
18 galactoceles (a marker of prolactin exposure) had either heavy pituitaries
19 or focal hyperplasia of the pituitary as detected by histopathology.
20
21 1.4 Mutagenic and Estrogenic Activity
22
23 The totality of the evidence from a variety of in vitro and in vivo studies
24 does not support a role for mutagenicity or DMA damaging potential for atrazine.
25 A detailed evaluation of the genotoxicity studies available on atrazine, its
26 metabolites, and structural analogues is provided in Part B, Chapter 6.
27 Additionally, as discussed in Part B, Chapter 7, numerous studies indicate that
28 atrazine does not have exogenous estrogenic activity.
29
30 The mutagenic compound W-Nitrosoatrazine (NNAT) can be formed in
31 vitro when atrazine and nitrite are mixed at an acid pH. Because nitrites and
32 atrazine can be found together in drinking water, concern has been raised about
33 this mutagenic chemical. Although the hypothesis has been advanced that
34 NNAT can be formed in the acid pH found in the stomach, the formation of NNAT
35 in the stomach in vivo has yet to be demonstrated. Further, cancer bioassays in
36 female Swiss mice and female Wistar rats failed to show a carcinogenic
37 response following NNAT exposure.
38
15
-------
DRAFT: DO NOT CITE OR QUOTE
1 1.5 Structure Activity Relationships
2
3 Like atrazine, treatment of female SD rats with simazine and propazine
4 leads to an increased incidence and/or early onset of mammary tumors. Also
5 like atrazine, treatment of male SD rats or CD-1 mice of either sex with simazine
6 or propazine does not lead to an increase in tumor incidences at any site (see
7 Part B, Chapter 8).
8
9 1.6 Doses Associated with Effects
10
11 Tables 1 -8 and 1-9 list NOAELs and LOAELs for the neoplastic and non-
12 neoplastic effects reported to be associated with treatment of female SD rats
13 with atrazine.
14
15 1.7 Chronic, Developmental, and Reproductive Toxicity
16
17 The data summarized in sections 1.1 through 1.6 indicate that primary
18 underlying events that lead to decreases in LH and prolactin release by the
19 pituitary, irregular estrous cycles, and mammary and pituitary tumor formation
20 following treatment of female SD rats with atrazine involve disruption of the
21 hypothalamic mechanisms involved in the regulation (release) of pituitary
22 hormone secretion. The proximal effects of atrazine that lead to these outcomes
23 have been identified as increased dopamine levels and decreases in
24 norepinephrine, and diminished ability to release GnRH from the hypothalamus
25 (Cooper ef a/., 1998). Because reproduction and development are controlled by
26 the neuroendocrine system, there are concerns that atrazine treatment could
27 lead to reproductive or developmental toxicity.
16
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
Table 1 -8. NOAELs and LOAELs (mg/kg/day) Associated with Neoplastic
Responses of Female SD Rats Treated with Atrazine
Response #
Carcinomas
Carcinomas
Carcinomas
Carcinomas
Carcinomas
Carcinomas
Fibro-
adenomas
Pituitary
adenomas
Duration of
Exposure
(Months)
24
24
24
12
12
12
9-15
9-12
Dose in mg/kg/day
(Incidence)
Control NOAEL LOAEL
0 05 3.5
(15/88) (16/67) (27/69)
0 4.2 24.4
(12/80) (14/80) (27/80**)
0 3.79 23.01
(17/60) (13/59) (22/60)
0 4 12 3.9
(1/25) (2/24) 6/25)
0 3.79 23 01
(0/14) (3/11) (6/18*)
0 4.22 4 4
(1/11) (2/10) (6/23)
0 <4.2 4.2
(3/30) (5/30) (5/30)
0 4.23 26.63
(2/20) (2/20) (8/20)
Reference
May hew ef a/.,
1986
Morseth, 1998
Thakur, 1992a
Pettersen &
Turnier,
1995
Thakur, 1991 a
Morseth, 1998
Thakur, 1991 a
Thakur, 1991 a
8
9
10
11
12
13
14
15
16
17
#= mammary unless otherwise specified; *p=<0.05; **p=<0.01; *** when adjusted for
survival
17
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
Table 1 -9. NOAELs and LOAELs (mg/kg/day) Associated with Non-Neoplastic
Responses in Female SD Rats Treated with Atrqzine
Response
Percent days
in estrus
LH-repeat
bleed; fold
increase above
baseline
LH-repeat
bleed fold
increase above
baseline
Mammary
galactoceles
Mammary
secretory
activity1
Mammary
dilated
ducts1
Mammary
acinar
development1
Pituitary
weights
relative to
control2
Duration of
Exposure
(Months)
~5
1
6
9
9
9
3
9
Dose in mg/kg/day
(Response)
0 NOAEL LOAEL
0 1.8 3.65
(32% days) (41% days) (45% days*)
0 5.0 40
(3.6X) (33X) (1.9X)
0 1.8 365
(3 7X) (3.3X) (2.6X)
0(10%) 4.23(40%) 26.23
0(24) <4.23 (28) 4.23 (28)
0(17) <4.23(24) 4.23(24)
0 (23) <4.23 (28) 4.23 (28)
0 <4.23 4.23
(+25%) (+25%)
Reference
Morseth, 1996b
Morseth, 1996a
Morseth, 1996b
Thakur, 1991 a
McConnell,
1995
McConnell,
1995
McConnell,
1995
Thakur, 1991a
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
*p=<0 05; **p=<0.01; when adjusted for survival; 1 - Index Score shown in parenthesis. Each grade was
assigned the following values: absent=0; mimmal=1; mild=2; moderately severe=3; marked=4. The sum
of these values is the index score; 2 - Increase in pituitary weight relative to control shown in
parenthesis.
18
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Standard (EPA Guideline) chronic and subchronic studies conducted with
3 atrazine do not provide insight regarding the potential of atrazine to produce
4 lesions of reproductive organs or tissues that might lead to adverse reproductive
5 or developmental outcomes in male or female animals. Similarly, results of
6 developmental or reproductive toxicity guideline studies with atrazine do not
7 show that the dam or her offspring express effects of atrazine treatment that can
8 be associated with disruption of the hypothalamic-pituitary-ovarian axis.
9 However, results of mode and mechanism of action studies conducted with
10 atrazine in the adult cycling or adult, ovariectomized, estrogen-primed female
11 rats suggest that treatment with atrazine, its structural analogues or metabolites,
12 during other periods of the life cycle would also alter reproductive or
13 developmental function in the dam or offspring. Special studies have been
14 conducted that show that atrazine has reproductive and developmental effects
15 that can be attributed to alterations in endocrine function. Summaries of the
16 guideline and special studies are presented below. Implications of the data
17 summaries presented are discussed in Chapter 2.
18
19 1.7.1 Chronic and Subchronic Toxicity of Atrazine
20
21 There is no clear evidence that chronic or subchronic treatment of
22 rats or dogs with atrazine, its metabolites or structural analogues leads to
23 effects on reproductive organs and tissue with the exception of the
24 carcinogenicity and histomorphologic effects involving mammary tissue
25 discussed previously. The principal effects reported in female SD rats
26 following chronic dietary treatment with high doses of atrazine (50
27 mg/kg/day) include altered hematology and clinical chemistry parameters,
28 retinal degeneration, centrolobular necrosis in the liver, rectus-femoris
29 muscle degeneration, myeloid hyperplasia, transitional epithelial
30 hyperplasia in the bladder and kidney, and extramedullary hematopoiesis
31 (Mayhew et a/., 1986). Other effects observed in this combined
32 chronic/carcinogenicity rat study at the high dose were histopathology
33 findings in male rats consisting of statistically-significant increases in
34 incidences of prostate epithelial hyperplasia and acinar hyperplasia of the
35 mammary gland at the high dose. These effects were observed at the
36 end of the study at which time there was increased survival in the high
37 dose male rats compared with control male rats. Thus, the significance of
38 the effects observed is unclear because the apparent increases may
19
-------
DRAFT: DO NOT CITE OR QUOTE
1 reflect the increased number of animals that survived for 24 months at the
2 high dose compared with the controls.
3
4 When atrazine was fed to dogs for one year, the prominent effect
5 observed was cardiac dysfunction (O'Conner et at, 1987). Chronic
6 effects observed in a 91-week dietary study in mice were limited to
7 hematologic alterations and decreased mean group absolute brain and
8 kidney weights (Hazelette and Green, 1987). The only effect occasionally
9 seen and potentially associated with endocrine alterations following
10 subchconic or chronic treatment with atrazine, its metabolites, or structural
11 analogues is an effect on the weight of the testes in rats and dogs.
12 However, this effect is variable in different studies. Atrazine treatment
13 produced no effects on the testes in a two-year rat bioassay or in a 18-
14 month mouse bioassay. Simazine treatment resulted in a decrease in
15 gonadal weights in males and females in a 90-day rat study. DACT did
16 not produce effects on the gonads when administered to dogs in 90-day
17 orone-year studies or when administered to rats in a 90-day study. G-
18 28279, when administered to dogs for 90-days produced decreased
19 testes weights. On the other hand, treatment with this metabolite led to
20 increased testes weights when administered to rats for 90-days. G-30033
21 treatment led to increased relative testes weights when fed to rats for 90-
22 days but produced no effects on testes weights in a 90-day dog study.
23 The overall conclusion regarding effects on gonadal tissue is that there is
24 no clear pattern of increased or decreased weights.
25
26 1.7.2 Developmental Toxicity of Atrazine
27
28 Results of standard (guideline) rat developmental toxicity studies
29 with atrazine show that effects in maternal animals are confined to
30 increased mortality and decreases in body weight gains and food
31 consumption (Infurana, 1984; Ginkis, 1989). Fetal effects observed in the
32 Infurana study (1984) included incomplete or delayed ossification of skull
33 bones or other sites (NOAEL, 10 mg/kg/day and LOAEL, 70 mg/kg/day).
34 The developmental NOAEL and LOAEL for delayed ossification in the
35 Ginkis study (1989) were 25 and 100 mg/kg/day, respectively.
36 Developmental effects observed in a rabbit developmental toxicity study
37 were reduced litter sizes, increased resorptions, and delayed ossification
38 at maternally toxic doses (appearance of blood in the cage or on the
20
-------
DRAFT: DO NOT CITE OR QUOTE
1 vulva, reduced body weight gain, and reduced food consumption) (Arthur,
2 1984a). The NOAEL and LOAEL (developmental) .in this study were 5
3 mg/kg/day and 75 mg/kg/day, respectively. There are no data that would
4 suggest that the delays in ossification in fetal animals are due to
5 disruption of the hypothalamic-pituitary-ovarian axis by atrazine and the
6 dose-levels for producing the delays in ossification (NOAELs 5-25;
7 LOAELs 70-100 mg/kg/day). Because of the limited histopathology and
8 the lack of measurements of developmental delays (e.g., vaginal opening
9 and preputial separation) in traditional developmental studies, it is not
10 expected that developmental effects of atrazine treatment that are
11 associated with endocrine perturbations would be seen in the guideline rat
12 and rabbit developmental studies.
13
14 1.7.3 Reproductive Toxicity of Atrazine
15
16 The effects on gonadal weights (both increases and decreases)
17 occasionally observed in subchronic and chronic studies with atrazine or
18 its metabolites were seen in multi-generation reproduction studies. In the
19 rat multi-generation studies with atrazine, simazine, and propazine,
20 increases were observed in relative but not absolute testes weights of
21 adult P0 and F1 rats following treatment with atrazine, simazine or
22 propazine at doses ranging from 29 to 50 mg/kg/day (Mainiero et a/.,
23 1987; Epstein et a/., 1991; Jessup, 1979). No effect on testes weights
24 were observed in juvenile pups. The increases in relative testes weights
25 may be due to decreased body weights of the adult animals. As noted
26 from the data on testes weights from subchronic and chronic studies, the
27 significance of this finding is unclear. The multi-generation study results
28 provided no evidence of reproductive or developmental toxicity. However,
29 as in the case of the developmental studies performed with atrazine, the
30 traditional, EPA Guideline studies for reproductive effects do not include
31 observations or measurements that were selected to determine effects
32 related to endocrine imbalances.
33
21
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 1.7.4 Special Studies
3
4 Several special studies have been performed with atrazine with the
5 objective of evaluating the effects of atrazine, or its metabolites, on
6 pregnancy maintenance and postnatal development. Table 1-10 provides
7 a listing of key findings reported in the special studies along with NOAELs
8 and LOAELs for the effects.
9
10 Q Pregnancy Maintenance
11
12 When 0, 50,100, or 200 mg/kg/day of atrazine was
13 administered by gavage to SD, F344, Holtzman, or LE rats during
14 GD 1-8 just prior to the diurnal prolactin surge or just prior to the
15 nocturnal surge of prolactin, a small but significant decline in mean
16 number of implantation sites was seen only in Fischer-344 rats at
17 the two highest doses. Holtzman rats alone showed an increase in
18 postimplantation loss at the two top doses (Cummings ef a/.,
19 submitted). Serum LH levels were significantly decreased in
20 Holtzman, or LE-hooded rats treated with 100 mg/kg, at 200
21 mg/kg/day in F344 rats, but at no dose in SD rats. A decrease in
22 serum progesterone levels was seen only in Holtzman rats treated
23 with 200 mg/kg.
24
25 In a series of experiments assessing the effect of atrazine
26 on pregnancy maintenance in the female rat by Narotsky ef a/.,
27 (submitted, 1999), atrazine was administered by gavage to F344,
28 SD, and LE rats during GD 6-10. The F344 strain was the most
29 sensitive to atrazine's effects on pregnancy maintenance (full-litter
30 resorption); the LE strain was the least sensitive. In F344 rats,
31 surviving litters appeared normal; however, parturition was delayed.
32 In SD rats, full-litter resorptions were also observed, but at higher
33 dose levels: parturition was delayed at the same dose levels as for
34 F344 rats. In contrast, the LE hooded strain showed full-litter
35 resorption at the same dose level as SD rats, but there were no
36 effects on parturition.
37
22
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Q Reproductive and Postnatal Effects
3
4 In a study examining the effect of atrazine on pubertal
5 development, young Wistar rats were treated by gavage with
6 atrazine (12.5, 25, 50,100 or 200 mg/kg/day) during PND 22-41
7 (Laws ef a/.,submitted; Laws etal., 2000). Vaginal opening was
8 significantly delayed (three or four days) by 50 and 100 mg/kg
9 respectively. The 200 mg/kg per day treatment with atrazine for
10 the same period delayed vaginal opening by more than seven days
11 in 18 of 32 females. When vaginal opening did occur, irregular
12 cycles were observed in the 50 and 100 mg/kg dose groups during
13 the ensuing two weeks. Vaginal opening occurred shortly after
14 dosing was stopped in the 200 mg/kg dose group, and these
15 females also demonstrated irregular estrous cycles for the next two
16 weeks. All animals returned to regular estrous cycles by PND 70.
17
18 In a study evaluating the pubertal development in the male,
19 weanling Wistar rats were dosed with atrazine during PND 23-53
20 (Stoker ef a/.,2000a; Stoker et a/., submitted). The significant
21 finding from this study was that atrazine delayed preputial
22 separation. The LOAEL for delay in preputial separation was <12.5
23 mg/kg/day. No consistent effect on serum prolactin and
24 testosterone concentrations was observed, but the serum levels of
25 these two hormones in animals of this age fluctuate widely making
26 significant difference difficult to identify. However, there was a
27 significant dose-related decrease in serum LH on PND 53 (r = -
28 0.92. P < 0.0024).
29
30
23
-------
DRAFT: DO NOT CITE OR QUOTE
1 Other studies have shown that reproductive tissues in the
2 offspring can also be affected if the dam is treated during lactation.
3 Suckling-induced PRL release was measured in Wistar dams
4 treated with atrazine by gavage, twice daily with 0, 6.25,12.5, 25,
5 or 50 mg/kg (the total daily dose was 13, 25, 50 or 100 mg/kg/day)
6 during PND 1 -4 (Cooper ef a/., 2000). Serum PRL in dams was
7 measured on PND 3. A significant rise in serum prolactin release
8 was noted in all control dams within 10 minutes of the initiation of
9 suckling. The 25 and 50 mg/kg/day treatment with atrazine
10 inhibited prolactin release in 40% or 60% of the dams, respectively;
11 the daily dose of 100 mg/kg inhibited this measure in all dams. In
12 this same study, the effect of postnatal atrazine on the incidence
13 and severity of inflammation of the lateral prostate of the offspring
14 was examined in adult males at 90 and 120 days. While no effect
15 was noted at 90 days of age, at 120 days, both the incidence and
16 severity of prostate inflammation was shown to increase in those
17 offspring of atrazine-treated dams (50 or 100 mg/kg/day).
18 Combined treatment of dams with ovine prolactin (oPRL) and
19 atrazine on PND 1 - 4 reduced the incidence of inflammation
20 observed at 120 days, indicating that this increase in inflammation
21 seen after atrazine alone resulted from the suppression of prolactin
22 in the dam. These data demonstrate that atrazine suppresses
23 suckling-induced prolactin release and that this suppression results
24 in lateral prostate inflammation in the offspring. The critical period
25 for this effect is PND1-9. It should be noted that vaginal opening
26 was delayed in the offspring of these dams (Stoker ef a/.,
27 submitted). Whether this effect is also related to changes in
28 prolactin secretion in the dam remains to be determined.
29
30
24
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Table 1-10. NOAELs/LOAELs (mg/kg/day) for Reproductive and Developmental
Effects Following Treatment of Dams or Offspring of Several Rat
Strains with Atrazine or its Metabolites1
Response and exposure
period
Decrease in mean number
of implantation sites- GD 1-
8
Delayed partuntion-GD 6-
10
Full litter resorptions-GD 1-
8 or 6-10
atrazine
DACT
DEAT
DIAT
OHA
Reduction in serum LH-GD
1-8
Decreased prolactin
release- PND 1-4 (dams)
Increased incidence of
prostatitis-PND 1-4
Increased incidence and
severity of prostatitis-PND
1-4
Delayed vaginal opening-
PND 22-41
Delayed preputial
separation-PND 23-53
F344
50/100
50/100
25/50
<67/67
<87/87
>80
<275/275
100/200
N.A.
N.A.
N.A.
N.A.
N.A
SD
>200
50/100
100/200
N.A.
N.A.
N.A.
N.A.
>200
NA.
NA
N.A.
NA
NA
Wistar
NA*
NA.
NA
NA
N.A.
N.A.
N.A
NA
13/25
13/25
25/50
25/50
<13/13
LE
>200
>200
100/200
N.A.
N.A.
N.A
N.A
50/100
N.A.
N.A.
N.A.
NA.
N.A.
Holtz-
man
>200
NA
50/100
NA
NA
NA
NA
50/100
NA
N.A
NA
NA
NA.
ReL
Cummings et
a/., submitted
Narotsky et
al., submitted;
1999
Cummings ef
al., submitted.
Narotsky et
al., submitted;
1999
Cummings ef
al., submitted;
Stoker etal.,
1999
Stoker ef al ,
1999
Stoker et al.,
1999
Laws ef al.,
submitted;
2000
Stoker et al,
submitted;
2000a
'Data are for atrazine unless otherwise noted; * not available
25
-------
DRAFT: DO NOT CITE OR QUOTE
1 Chapter 2
2
3 2. Hazard Characterization And Mode of Action Analysis
4
5 This Chapter presents information characterizing the neoplastic and non-
6 neoptastic effects reported from studies conducted with atrazine and considers them in
7 the context of an analytical framework for evaluating a postulated mode of action as
8 described in the proposed revisions to the guidelines for carcinogen risk assessment
9 (EPA, 1999). The framework is used to judge how well the available data support a
10 mode of action postulated for a carcinogenic agent. This Chapter draws on the
11 information summarized in the preceding Chapter. Complete details on the
12 carcinogenicity and chronic toxicity of atrazine are presented in Part B of this document.
13 This Chapter also evaluates the neuroendocrine effects of atrazine on the development
14 and function of the reproductive system. The details of these studies can be found in
15 PartC.
16
17 2.1 Human Cancer Studies
18
19 Several epidemiologic studies have examined cancers among populations
20 with exposures relevant to the assessment of atrazine, especially among farmers
21 or farm residents (see Part B, Chapter 4 for details). Most are case control
22 studies, although there are ecologic investigations and also a worker mortality
23 study of workers directly employed in the manufacture of triazines. Studies
24 examining the association of triazine exposure with colon cancer, leukemia,
25 multiple myeloma, soft tissue sarcomas, and Hodgkins disease failed to find firm
26 associations. The pooled results of three separate case-referent studies
27 investigating atrazine exposure in the development of non-Hodgkins lymphoma
28 (NHL) concluded that there was essentially no risk of NHL attributable to farm
29 use of atrazine. A mortality study of workers in two triazine manufacturing plants
30 did not find any significant excesses of deaths for any disease category. There
31 were, however, two cases of NHL in plant workers - one of whom was relatively
32 young (31 years). These two cases do not provide evidence of an association
33 between atrazine exposure and NHL, but do indicate that further follow-up of
34 workers in these triazine manufacturing plants is desirable.
35
26
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Associations between triazine exposure and cancer for three hormone-
3 responsive cancers-ovary, breast and prostate cancer has been reported.
4 Although suggestive, these associations should not be considered as conclusive
5 evidence of a correlation between triazine exposure and these tumor types. The
6 studies that showed possible relationships between these tumor types and
7 triazine exposure should be interpreted with caution because of limitations, such
8 as misclassification of subjects, use of surrogate data for exposure, or
9 concurrent exposure to other potentially carcinogenic compounds.
10
11 To summarize, there is suggestive evidence of a possible association of
12 triazine exposure and NHL, prostate, breast and ovarian cancers. This evidence
13 does not show a direct cause and effect relationship between atrazine or triazine
14 exposures and carcinogenicity because of confounding factors and limitations in
15 the available studies. The available evidence emphasize the need for further
16 epidemiologic research into the association of these tumor types with atrazine
17 exposure.
18
19 2.2 Carcinogenicity in Female SD Rats
20
21 There were dose-related increases in the incidence of mammary tumors
22 (adenomas, adenocarcinomas, and carcinosarcomas combined) in female
23 Sprague-Dawley (SD) rats in the seminal carcinogenicity test performed with
24 atrazine (Mayhew et a/., 1986). No dose-related increases in tumor responses
25 were observed in male SD rats. Results of subsequent bioassays, some of
26 which included serial and/or one year sacrifices, confirmed that the predominant
27 response observed following testing of atrazine in female SD rats is an increase
28 in the incidence and/or early onset of mammary adenomas/carcinomas.
29 Although less compelling, there is evidence that there is decreased latency for
30 the formation of mammary fibroadenomas and pituitary adenomas (Thakur,
31 1991 a and 1992a; Petersen and Turnier, 1995) and an increased incidence of
32 mammary fibroadenomas (Morseth, 1998). An increased tumor incidence is not
33 found at any other site in female SD rats, or at any site in male SD rats, or in
34 either sex of Fischer 344 rats and CD-1 mice (Mayhew ef a/., 1986; Hazelette
35 and Green, 1987; Thakur, 1992a,b). Mammary tumors were reported in one
36 study in male Fischer 344 rats that involved lifetime treatment with atrazine
37 (Pinter ef a/., 1990), but the finding is difficult to evaluate in light of the
38 experimental design and shortcomings of the study. Furthermore, this finding is
27
-------
DRAFT: DO NOT CITE OR QUOTE
1 in conflict with the results of a conventional 24-month carcinogenicity study with
2 F344 male rats that showed no increases in mammary tumors (Thakur, 1992b).
3 The closely related structural analogues to atrazine, simazine and propazine,
4 also produce mammary tumors in the female SD rat but no other tumors of any
5 type in the female SO rat and no tumors of any kind in the male SD rat or in CD-
6 1 mice of either sex.
7
8 2.3 Postulated Mode of Carcinogenic Action
g
10 Before presenting the postulated mode of action for atrazine, it is
11 instructive to consider aspects of the normal reproductive biology of the female
12 SD rat and its relevance to tumor formation.
13
14 2.3.1 Reproductive Aging in Rats
15
16 With advancing age, the female Sprague-Dawley, as most strains
17 of rats, normally undergoes a transition from regular ovarian cycles to an
18 acylic pattern of "persistent" or "constant" estrus (Cooper and Walker,
19 1979; Also, see Part B, Chapter 9.1). Typically, this transition occurs prior
20 to one year of age and is related to a disruption in both the timing and
21 amplitude of the preovulatory surge of lutenizing hormone (Cooper et a/.,
22 1980). As a result of this inability to achieve ovulation, the ovaries of the
23 constant estrous female may contain many large follicles (i.e.,
24 polyfollicular ovaries) but no corpora lutea (Huang and Mertes, 1975).
25 These follicles continue to secrete estradiol, while progesterone secretion
26 is minimal (Huang et a/., 1978). This pattern of hormone secretion has
27 been shown to facilitate the development of mammary gland tumors in
28 aging rats and in young females in which a constant estrus has been
29 induced (Nandi etal., 1995; Russo ef a/., 1990; Cutts and Noble. 1964;
30 Meites, 1972). The inability to achieve an ovulatory surge of LH is the
31 result of changes in the ability of the hypothalamus to achieve the proper
32 release of GnRH. Changes in norepinephrine concentration occur prior to
33 the onset of the loss of regular ovarian cyclicity (Wise ef a/., 1997; Wise ef
34 a/., 1999). Conversely, treatment with CNS acting compounds such as
35 the catecholaminergic precursor, L-dopa, will result in a reinitiation of
36 regular cycles (Quadri ef a/., 1973). Similarly, the age at which regular
37 estrous cycles are disrupted can be extended if the female is placed on a
38 diet containing L-tyrosine (i.e., the amino acid precursor of L-dopa).
28
-------
DRAFT: DO NOT CITE OR QUOTE
1 Persistent or constant vaginal estrus, the accompanying pattern of
2 persistent estradiol secretion and no progesterone, also leads to an
3 increase in pituitary weights, development of pituitary hyperplasia, and
4 formation of pituitary adenomas in the aged female rat (Blankenstein et
5 a/., 1984; McConnell, 1989a; Nelson etal., 1980; Meites, 1980;
6 McConnell, 1989b). The majority of the pituitary adenomas seen in the
7 aged female SD have been found to originate from lacotrophs (i.e.,
8 prolactin-secreting cells of the anterior pituitary) (Sandusky et a/., 1988).
9 The increased number of prolactin-secreting cells results in an increased
10 serum level of prolactin and extended or prolonged exposure of mammary
11 tissue to higher than normal levels of prolactin. As indicated above,
12 dietary supplementation with L-dopa and L-tyrosine (precursors to
13 catecholamine synthesis in the central nervous system) delays
14 reproductive aging as evidenced by maintained LH surges, normal
15 reproductive cycling, and delayed onset of mammary gland tumor
16 formation in treated animals compared to controls of the same age. No
17 female Long-Evans (LE) rat developed mammary tumors by 21 months of
18 age when fed a diet supplemented with L-tyrosine compared with a
19 mammary tumor incidence of 67% in control (no supplement) animals
20 (Cooper and Walker, 1979). Restored vaginal cycling is also found when
21 aged female rats are administered L-dopa and L-tyrosine. Ovariectomy
22 also reduces exposure of mammary tissue to estrogen and reduces or
23 eliminates mammary tumor formation.
24
25 In summary, reproductive aging in the female rat appears to result
26 from a disruption of hypothalamic neurotransmitter and neuropeptide
27 (primarily noradrenergic) regulation of GnRH, and subsequently LH
28 secretion. Importantly, the normal age-related disruption of regular
29 cycling can be modified by pharmaceutical treatment or dietary
30 supplementation. Finally, the resultant endorine milieu of enhanced or
31 unopposed estrogen and prolactin secretion, provides an environment
32 that is conducive to the development of mammary gland and pituitary
33 tumors.
34
29
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 2.3.2 Atrazine Effects Relevant to Carcinogenicity
3
4 It is postulated that the carcinogenicity of atrazine is a
5 consequence of the disruption of the normal secretory activity of the
6 hypothalamic-pituitary-ovarian axis. Atrazine exposure adds to the
7 formation of mammary tumors by inducing a sequence of events which
8 intersects, at some point, with the normal reproductive aging pathway.
9 The point of intersection appears to be the attenuation of the proestrous
10 afternoon LH surge. Both in vivo and in vitro experiments demonstrate
11 that atrazine exposure does not directly affect the pituitary (Cooper et al.,
12 2000) and that a decreased ability of the hypothalamus to release GnRH
13 is likely the cause of the attenuated LH surge in the atrazine exposed SD
14 female. Finally, pituitary weight and histomorphologic data in the
15 mammary gland demonstrate that continued estrogen secretion also
16 stimulates prolactin secretion by the pituitary. Again, ongoing secretion of
17 estrogen and prolactin create an endocrinological milieu conducive to
18 mammary gland and pituitary gland cell proliferation and eventual tumor
19 development.
20
21 Females of the F-344 rat strain have a rather low background
22 incidence of mammary tumors. In contrast to SD, LE, and Wistar females,
23 this strain goes through a different pathway for reproductive senescence.
24 F-344 females age through a process termed repeated pseudopregnancy,
25 a condition where there are normal LH surges and ovulation occurs but
26 continued secretion of progesterone by corpus lutea leads to a vaginal
27 cytology indicative of diestrous. Mammary tumors are not induced by
28 atrazine in F344 female rats. It would seem that the differences in
29 reproductive aging between the F-344 and SD strains influence their
30 sensitivity and response to atrazine administration.
31
30
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Figure 2-1 illustrates the postulated mode of action of atrazine in
3 female SD rats on the activity of the hypothalmic-pituitary-ovarian axis and
4 the development of mammary and to some extent pituitary neoplasms.
5 Effects associated with atrazine treatment on the activity of this axis are:
6
7 1. Atrazine exposure affects - either directly or indirectly - the
8 hypothalamus, leading to a decreased secretion of hypothalamic
9 norepinephrine (NE) (Cooper 1998)1.
10
11 2. Hypothalamic NE normally modulates the release of gonadotropin
12 releasing hormone (GnRH) from the hypothalamus. Decreased NE
13 levels result in decreased release of GnRH from the hypothalamus
14 (Cooper, 1998).
15
16 3. GnRH is the hormone responsible for inducing the pituitary gland to
17 release luteinizing hormone (LH). A decreased GnRH level leads
18 to an attenuated LH release (Cooper et a/., 2000, Morseth, 1996a,
19 b).
20
21 4. LH normally provides a signal to the ovaries promoting ovulation.
22 Below some critical level, the decreased serum levels of LH are
23 insufficient to stimulate ovulation.
24
25 5. Estrogen from ovarian follicles normally provides a feed back to the
26 hypothalamus to stimulate a pituitary LH surge which promotes
27 ovulation. Following atrazine exposure, there is insufficient GnRH
28 to stimulate ovulation. Under the tonic secretion of LH and FSH,
29 the ovarian follicles persist and continue to secrete estradiol. In
30 turn, under the continued stimulation of estradiol, the pituitary
31 lactotrophs become hypertrophied and secrete increasing amounts
32 of prolactin.
33
1 Cooper (1998) has also shown that acute atrazine treatment results in an increase in
hypothalamic dopamine which in turn results in a decrease in pituitary prolactin. This
acute effect is not expected to be associated with neoplasia but has potential
reproductive consequences under certain circumstances.
31
-------
DRAFT: DO NOT CITE OR QUOTE
1 6. Estrogen acts on the mammary gland increasing the risk of
2 mammary tumors, especially carcinomas and adenomas.
3
4 7. Prolactin derived from the hyperplastic lactotrophs (prolactin
5 secreting cells) described in step 5 also acts on the mammary
6 gland (in concert with estrogen) to increase the risk of mammary
7 tumors, particularly fibroadenomas.
8
9 8. Tumor formation by atrazine does not appear to involve direct
10 mutagenic effects nor does atrazine act as a direct estrogen
11 agonist.
12
13 2.4 Evaluation of the Postulated Mode of Carcinogenic Action
14
15 In this section, the evidence linking the formation of mammary and
16 pituitary tumors in female SD rats with disruption of biochemical activities in the
17 hypothalamic-pituitary-ovarian axis is examined. These sections also examine
18 the evidence supporting or refuting the postulated mode of action described in
19 Figure 2-1 as the causal mode of action associated with the carcinogenicity of
20 atrazine in female SD rats.
21
22
23
24
32
-------
1
2
3
4
DRAFT: DO NOT CITE OR QUOTE
Figure 2-1. Postulated Effects of Atrazine Treatment on the Hypothalamic-
Pituitary-Ovarian Axis
Hypothaimus
NE
GnRH (-}
2
\/
Atrazine
Pituitary
LH
3
/N
\/
Ovary
Anovulation
PrL
Mammary Gland
\
Increased Risk of
Mammary Tumors
33
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.4.1 Key Events
2
3 Data showing that the hypothalamus appears to be a primary initial
4 site of action for atrazine primarily come from short-term, high dose
5 studies conducted in Long Evans (LE) females2 by the EPA's National
6 Health and Effects Research Laboratory (Cooper et a/., 1998; Cooper,
7 2000). These studies provide evidence that atrazine affects hypothalamic
8 catecholamine levels. A decrease in NE results in a decrease in
9 gonadotropin releasing hormone (GnRH), with a corresponding diminution
10 of pituitary surges of luteinizing hormone (LH). These in vivo observations
11 are further supported by in vitro studies using pheochromocytorna cells.
12 In this cell line, both dopamine and norepinephrine are synthesized
13 constitutively. Das et al. (2000, in press) have shown that catecholamine
14 synthesis is suppressed, in a dose dependent manner, following exposure
15 to atrazine. Evidence for a hypothalamic site of action for the
16 neuroendocrine disrupting effects of atrazine include the following
17 observations:
18
19 Q the pulsatile release of GnRH from the hypothalamus is
20 impaired in the female rat following atrazine exposure
21 (Cooper et al., 1998)
22
23 Q the atrazine-induced suppression of LH secretion can be
24 reversed following treatment with synthetic GnRH (Cooper ef
25 al., 2000)
26
27 Q there is a dramatic increase in the hypothalamic
28 concentration of GnRH following exposure to atrazine
29 demonstrating that release (and not synthesis) of GnRH is
30 impaired (Ford et al., 2000)
2Some rat strains (LE, Wistar and SD included) undergo a similar reproductive aging
process which is characterized by the appearance of persistent (or constant) estrus by
approximately one year of age and under similar neuroendocrine events. Thus, the LE
female rat is considered to be a valid model for evaluating atrazine's mode of action
resulting in mammary tumors in SD females.
34
-------
DRAFT: DO NOT CITE OR QUOTE
1 Q related to these effects on GnRH release, a hypothalamic
2 site of action also appears to be responsible for the inhibition
3 of prolactin release as the atrazine-induced suppression of
4 prolactin secretion is not observed if the pituitary is removed
5 from its normal location (within the sella turcica, beneath the
6 hypothalamus) and placed beneath the kidney capsule
7 (Cooper ef a/. ,2000).
8
9 Suppression of the LH surge in female SD rats is considered to be
10 a necessary precursor for the development of atrazine-induced mammary
11 gland tumors. This is because LH blood levels must reach a sufficient
12 magnitude to induce ovulation and to maintain normal reproductive cycles.
13 When atrazine reduces LH output to the critical point where there is not
14 enough to trigger ovulation, a physiological state results which is
15 characterized by prolonged or persistent estrous. This state leads to
16 continued stimulation of mammary tissue by estrogen. Evidence for an
17 attenuation of the LH surge and an early onset of prolonged and/or
18 persistent estrus is provided in several studies (Morseth 1996a,b; Thakur
19 1991a; Eldridge et a/., 1993a). Removal of the estrogen stimulus by
20 ovariectomy completely abolishes the formation of mammary tumors
21 following chronic administration of atrazine (Morseth, 1998). Estrogen
22 has been strongly implicated in mammary gland cell proliferation and the
23 enhancement of neoplastic transformation in rodents and humans (for
24 review see Russo and Russo, 1996; Nandi, 1996).
25
26 The attenuation of LH surges and disruption of the normal
27 reproductive cycles in female SD and Long-Evans hooded rats treated
28 with atrazine mirrors prominent features of the normal reproductive aging
29 process in these strains. This process features a diminution of LH blood
30 levels, a failure to ovulate, and a state of persistent estrus.
31
35
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Prolonged estrogen secretion resulting from atrazine treatment
3 appears to lead to other consequences. There is evidence that sustained
4 exposure of the pituitary gland to estrogen leads to an increase in pituitary
5 weights, pituitary hyperplasia, development of lesions characteristic of
6 prolactin secretion, and the formation of pituitary adenomas (Thakur,
7 1991a; McConnell, 1995). The sustained secretion of prolactin is believed
8 to play a role in the development of mammary tumors, in general, but a
9 more prominent role in the development of mammary fibroadenomas
10 (Welsch, 1985).
11
12 2.4.2 Correlation of Effects and Dose
13
14 There is a strong association between dose levels of atrazine that
15 lead to an early onset and increased incidences of mammary tumors and
16 doses that produce biochemical perturbations that have been linked to
17 reproductive aging (i.e., suppression of LH surges and prolonged or
18 persistent estrus). Table 2-1, lists the lowest dose (LOAEL) which elicited
19 each of the effects associated with atrazine treatment. Tables 1-8 and 1-
20 9, Chapter 1, may be referred to by the reader for NOAELs and LOAELs
21 of all data on tumor and non-neoplastic effects.
22
23 A dose of 3.5 mg/kg/day and above that leads to an early onset
24 and/or increased incidence of mammary carcinomas in female SD rats
25 also leads to attenuation of LH secretion. Examination of Figure 1-2,
26 Chapter 1, indicates that administration of atrazine at a dose level of 3.65
27 mg/kg/day results in a diminution of the LH surge. This is the same dose
28 that results in estrous cycle perturbations. At a dose level of 29.4
29 mg/kg/day, the LH surge is completely suppressed. If attenuation of the
30 LH surge were indeed a key event in mammary and pituitary tumor
31 formation, then doses that result in an attenuation of the LH surge would
32 be expected to result in an increased incidence or early onset of these
33 tumors. Doses of 4.2 and 24.4 mg/kg/day resulted in an early onset of
34 mammary carcinomas (Morseth, 1998). Doses of 3.79 and 23.01
35 mg/kg/day resulted in an early onset of mammary carcinomas in another
36 study (Thakur, 1992a). The evidence for an early onset of mammary
37 fibroadenomas and pituitary tumors is less strong as these effects were
38 only seen in one study (Thakur, 1991 a).
36
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
5
6
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Table 2-1. LOAELs for Tumor Formation and Non-Neoplastic Effects in Female
SDRats
Effect/Time of Observation
LH-repeat bleed; increase above baseline
(6 months)
Prolonged days in estrus (6 months)
Mammary carcinomas - decreased latency
(12 months)
Mammary carcinomas - increased
incidence (24 months)
Mammary galactoceles
(9 months)
Increased pituitary weights
(9 months)
Pituitary adenomas - decreased latency (9
months)
Mammary fibroadenomas - decreased
latency (15 months)
LOAEL
(mg/kg/day)
3.65
3.65
3.79
3.5
4.23
4.23
26.23
4.23
Reference
Morseth, 1996b
Morseth. 1996b
Thakur, 1992a
May hew eta/., 1986
Thakur, 1991 a
Thakur, 1991 a
Thakur, 1991 a
Thakur. 1991a
There is also a correlation between time spent in estrus and tumor
formation. The data from the 1998 Morseth study, as described in Thakur
(1999), shows that there is a statistically-significant correlation between
percent days spent in estrus during both the one to 46 week and 17 to 26
week time intervals, and an increased mammary carcinomas incidence.
Moreover, examination of the animals in this study, where there was an
especially early tumor onset (prior to 52 weeks), showed that there was
an unusually long period of time spent in estrus. Five of six female SD
rats that developed mammary carcinomas by 52 weeks spent >70 % of
the days in estrus between weeks 17 to 26.
37
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 As discussed previously, estrogen stimulation of the pituitary gland
3 is believed to cause an increase in the secretion of prolactin, a hormone
4 closely associated with the development of mammary tumors, especially
5 fibroadenomas. There is histomorphologic evidence (e.g., acinar
6 development, dilated ducts, and increases in the incidence and severity of
7 galactoceles) of an early onset of increased prolactin secretion at 4.23
8 mg/kg/day (McDonnell, 1995). It is biologically plausible that this early
9 exposure to prolactin may contribute to the early onset of mammary
10 fibroadenomas as seen in Thakur (1991 a). There is also an early onset of
11 increased pituitary weights in this study. Absolute pituitary weights at 4.23
12 mg/kg/day are increased by 25% at nine months. The increase is likely
13 due to the mitogenic effect on pituitary lactotrophs of estrogen derived
14 from unovulated follicles. The larger pituitaries would be expected to
15 secrete increased amounts of prolactin. This is indicated by the early
16 onset of prolactin-dependent histomorphologic parameters and the early
17 onset of mammary fibroadenomas.
18
19 The lowest atrazine dose showing effects on LH, the pituitary
20 gland, and the estrous cycle is somewhere between 3 and 4 mg/kg/day.
21 The LH surge attenuation occurred at 3.65 mg/kg/day, but did not occur at
22 1.8 mg/kg/day. The estrous cycle alterations occurred at 3.1 and 3.65
23 mg/kg/day in two separate studies. In these studies, the estrous cycle
24 alteration did not occur at 1.5 and 1.8 mg/kg/day. There is only one study
25 on serum estradiol levels. Although this study shows an early onset of
26 increased estradiol levels at 4.23 mg/kg/day, a clear dose effect level is
27 uncertain due to variability in the data and the lack of confirmatory data at
28 other timepoints in the same study (e.g., six months). The main factor is
29 that estrogen secretion is prolonged during persistent estrus which results
30 in continuous stimulation of the mammary gland.
31
32
38
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.4.3 Temporal Association of Effects
2
3 Data from chronic studies in female SD rats administered atrazine
4 consistently show that there is an early onset of mammary tumors. This is
5 what would be expected if atrazine accelerated the reproductive aging
6 process. Therefore, it is anticipated that precursor events to mammary
7 tumors would have their onset in atrazine treated females before that of
8 untreated SD females undergoing normal reproductive aging. The
9 temporal pattern of effects found following atrazine treatment are
10 summarized in Figure 2-2.
11
12 In untreated aging female SD rats, prolonged days in estrus begin
13 as early as nine months and shortly thereafter they enter into persistent
14 estrus. Extended days in estrus, a key event associated with the
15 formation of mammary tumors begins earlier in rats treated with atrazine
16 than in controls. An increased number of days in estrus begins as early
17 as 3.5 or 5.5 months in females administered 29.4 mg/kg/day or 3.65
18 mg/kg/day of atrazine, respectively (Morseth, 1996b). These data were
19 confirmed in a separate study which showed that by 3.3 months SD
20 females exposed to 24.4 mg/kg/day were spending approximately 26%
21 more days in estrus than control animals (Thakur, 1999). Dietary
22 administration of 3.65 mg/kg/day of atrazine leads to attenuation of the
23 proestrus afternoon LH surge after as little as six months of atrazine
24 exposure. Thus, exposure to atrazine decreases the onset time of
25 attenuated LH surge and persistent estrus. These effects have been
26 identified as the precursor events in the pathway towards mammary
27 tumors in rats. In keeping with these findings, animals receiving 4.23
28 mg/kg/day (lowest dose tested) manifest an early onset of
29 histomorphologic changes in mammary tissue (e.g., increased incidences
30 and severity of acinar formation, secretory activity, and galactoceles)
31 following six to nine months of treatment of female SD rats with atrazine
32 (McConnell, 1995). These changes are primarily indications of exposure
33 of mammary tissue to prolactin and estrogen. This broad time line
34 illustrates the sequence of events that occur prior to tumor development
35 (as well as the associated effective dose levels for the response).
36
39
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 2.4.4 Biological Plausibility and Coherence of the Database
3
4 The process of normal reproductive senescence in the female SD
5 rat has been implicated in creating a hormonal milieu conducive to
6 mammary tumor and pituitary tumor formation, including:
7
8 Q Attenuation of the pre-ovulatory LH surge;
9
10 Q Increased days in estrus; and
11
12 Q Prolonged exposure to endogenous estrogens and prolactin.
13
14 The events listed above have been well described in the open
15 literature as normal and expected events in the reproductive aging of the
16 female SD rat (Cooper and Walker, 1979; Lu, 1994; Mobbs, 1996; Smith
17 and Conn, 1983; Zuo, 1996). Prolonged exposure to endogenous
18 estrogens has been generally accepted as a major contributor to the high
19 spontaneous mammary and pituitary tumor rates seen in the SD female
20 (Welsch, 1987; Cooper, 1983; Cutts and Noble, 1964). Prolonged
21 exposure of mammary tissue to prolactin, which results from the estrogen-
22 induced pituitary tumors, also has been well established as a contributor
23 to mammary carcinogenesis in the normally aging female SD (Welsch,
24 1970a; Welsch, 1970b; Meites, 1971; Goya et a/., 1990).
40
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
Figure 2-2. Temporal Pattern of Atrazine Effects*
Study Initiation <
15-16 weeks'
21 -22 weeks<
26 weeks <
39 -40 weeks •
52 weeks <
65 weeks •
73 weeks '
75 weeks <
76 weeks •
Increased serumestradiol levels** at 4.25 mg/kg/day (Thakur 1991a)
Increased days in estrus at 3.65 mg/kg/day (Morseth 1996b)
Attenuation of the LH surge at 3.65 mg/kg/day (Morseth, 1996b)
Increased pituitary alteratons andprolactin-associated mammary
gland histology at 4.23 mg/kg/day jfhakur, 1991 a)
Increased incidences of pituitary adenomas at
26.63 mg/kg/day (Thakur, 1991a)
Mean week of onset for mammary adenomas/carcinomas
at 4.2 mg/kg/day (Morseth, 1998)
Mean week of onset for mammaryfibroadenomasin control animals (Morseth, 1998)
Mean week of onset for mammaryfibroadenomas
at 4.2 mg/kg/day (Morseth. 1998)
Mean week of onset for mammary carcinomas in control animals tylorseth. 1998)
Time when effects are first noted was dictated by study desigifOnly one study available
41
-------
DRAFT: DO NOT CITE OR QUOTE
1 The biologic plausibility for the mode of action proposed for
2 atrazine-induced mammary and pituitary carcinogenicity lies in the
3 observation that atrazine exposure has been shown to induce an earlier
4 onset of all three of the events outlined above: attenuation of the pre-
5 ovulatory LH surge; increased days in estrus; and prolonged exposure to
6 endogenous estrogen. Because this sequence of events has been
7 generally accepted as leading to mammary and pituitary carcinogenesis in
8 the normally aging SD female, one can reasonably expect that atrazine
9 administration would lead to the same events, including tumors, only at
10 earlier times than in normally aging females.
11
12 Atrazine dose levels that lead to attenuation of the LH surge also
13 are associated with disruption of the estrous cycle and an early
14 development or increased incidence of mammary and pituitary gland
15 tumors. One study provides histomoiphologic evidence that an early
16 onset of pituitary tumors and mammary fibre-adenomas may be explained
17 by prolonged secretion of estrogen by the anovulatory female rat,
18 stimulation of the pituitary to undergo cell proliferation, and increased
19 prolactin secretion by the estrogen-stimulated pituitary gland. The
20 formation of both mammary carcinomas and mammary fibroadenomas are
21 influenced by prolonged exposure of the mammary gland to follicle -
22 derived estrogen and pituitary-derived prolactin. Thus, in several
23 respects, the effects of atrazine treatment mirror biochemical alterations
24 that have been attributed to the onset of reproductive aging and
25 spontaneous tumor formation in the female SD rat. The mode of action
26 proposed to account for the tumor responses in female SD rats treated
27 with atrazine is biologically plausible because the major key biological and
28 biochemical events shown to be altered by atrazine treatment are the
29 same ones that have been identified as contributors to the formation of
30 mammary and pituitary tumors in aging female SD rats.
31
32
42
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.4.5 Other Modes of Action
2
3 Q Mutagenicity
4
5 Because cancer is the result of a series of genetic defects in
6 genes controlling cell growth, division, and differentiation, an initial
7 and prominent question to be examined is whether atrazine (or an
8 atrazine metabolite) interacts directly with, and mutates DMA. The
9 totality of evidence for atrazine, including data on several
1 o metabolites of atrazine and close structural analogues, does not
11 support a mutagenic potential for atrazine, and indicates that a
12 direct DNA reactive, mutagenic mode of action is unlikely to be an
13 influence on atrazine tumor development. The genetic toxicology
14 database for atrazine shows consistent negative responses in
15 bacterial tests and inconsistent positive responses across other
16 phylogenetic lines (which are typically weak, found at high doses,
17 or cannot be reproduced). No subset of data points clearly
18 establishes a direct DNA reactive mode of action for atrazine
19 associated with the carcinogenicity.
20
21 Q Estrogen Agonistic Action
22
23 The available evidence from in vivo and in vitro studies
24 indicates that atrazine does not bind to the estrogen receptor or
25 possess direct estrogenic activity. Under equilibrium conditions,
26 atrazine does not compete with estradiol for binding to SD rat
27 estrogen receptors (Tennant et a/., 1994b). Atrazine treatment
28 does not induce changes in estrogen-responsive tissues (e.g.,
29 increased uterine weight, Increased uterine cell proliferation,
30 uterine peroxidase activity and uterine progesterone receptors) in
31 ovariectomized SD rats. Atrazine does not affect basal or estradiol
32 induced cell proliferation in a human breast cancer cell line (MCF-
33 7) (Safe et a/., 1995). Atrazine does not have agonist or antagonist
34 action against estradiol induced luciferase activity in MCF-7 cells
35 transfected with a Gal4-regulated human estrogen receptor
36 chimera, thus showing failure to bind to the estrogen receptor
37 (Conner et a/., 1996). Although estrogen binding was found for
38 atrazine (Tennant et a/., 1994b), it was demonstrated only under
43
-------
DRAFT: DO NOT CITE OR QUOTE
1 conditions that favored binding and at very high doses of atrazine
2 relative to those that induced mammary tumors in SD females.
3 DACT (a metabolite of atrazine) and simazine (analogue of
4 atrazine) also do not appear to bind to the estrogen receptor
5 (Tennant et a/., 1994b; Connor ef a/., 1996).
6
7 Q Other Endocrine Imbalances
8
9 Data presented in abstract form indicates that atrazine
10 depresses hypothalamic norepinephrine (NE) levels and increases
11 hypothalamic dopamine (DA) levels (Cooper, 1998). NE levels
12 correlate directly with hypothalamic GnRH release (i.e., increased
13 NE leads to increased GnRH release) while DA levels are inversely
14 related to pituitary prolactin release (i.e., increased DA levels leads
15 to decreased prolactin secretion). Thus, a decrease in
16 hypothalamic GnRH secretion and a decrease in pituitary prolactin
17 secretion might be expected from alterations of these
18 neurotransmitters.
19
20 Chronic exposure to doses of atrazine as low as -4
21 mg/kg/day leads to elevated prolactin secretion, as indicated by
22 histomorphologic markers, presumably because of estrogen-
23 induced pituitary lactroph hyperplasia in the anovluatory female.
24 Prolonged exposure to serum prolactin contributes to mammary
25 gland carcinogenesis in the rat (Welsch, 1985) because of its
26 proliferate effect on the mammary gtand tissue.
27
28 The pituitary does not appear to be a direct target of atrazine
29 toxicity. When pituitary hormone secretion is removed from the
30 influence of CNS hypothalamic factors (r'.e., by placing pituitary
31 grafts under the kidney capsule) in LE females, there is no effect
32 on prolactin release when animals are exposed to a dose of
33 atrazine that suppresses the prolactin surge (Cooper, 2000). The
34 atrazine-induced attenuation of the LH surge can be reversed by
35 intravenous exposure to exogenous GnRH (Cooper, 2000). This
36 implies that the pituitary is functional and the deficit responsible for
37 LH surge attenuation is a hypothalamic insufficiency of GnRH
38 release. In vitro studies provide additional support that effects on
44
-------
DRAFT: DO NOT CITE OR QUOTE
1 the LH and prolactin surges are not due to a direct pituitary
2 response to atrazine exposure (Cooper, 2000). No differences in
3 either LH or prolactin release were found from the pituitaries of
4 untreated females exposed to atrazine in vitro. These three lines of
5 evidence indicate that the effect of atrazine on the LH surge (and
6 the high-dose effect on the prolactin surge) involves a disruption of
7 the GnRH pulse from the hypothalamus, rather than a direct effect
8 on the pituitary.
9
10 There is some evidence that atrazine may enhance
11 estrogenic activity by stimulating aromatase activity. Aromatase is
12 an enzyme that converts androgens to estrogens. Treatment of
13 human adrenocortical cells in vitro with atrazine has been shown to
14 stimulate aromatase activity (Sanderson et at., 2000). Crain et a/.
15 (1997) have also shown that atrazine treatment of male hatchling
16 alligators leads to an increase in aromatase activity. Although an
17 increase in aromatase activity would be consistent with dose-
18 response increases in estradiol and estrone and decreases in
19 testicular testosterone noted in a study that examined the effects of
20 atrazine on pubertal development, the doses that resulted in effects
21 on these hormones were well above doses that led to reproductive
22 developmental effects (Stoker et a/., submitted; 2000a). It is
23 plausible that enhanced aromatase activity may have some
24 influence on the development of mammary tumors in SD female
25 rats but whether or not enhanced aromatase activity is a significant
26 contribution to the carcinogenicity, or other effects, of atrazine
27 remains to be determined.
28
29 No other modes of action, apart from disruption of the
30 hypothalamus-pituitary-ovarian axis, are plausible biochemical
31 processes that could account for the early onset and increased
32 incidence of mammary and pituitary gland tumors in female SD
33 rats.
34
45
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 2.4.6 Uncertainties and Limitations
3
4 Despite numerous studies directed at understanding the mode of
5 action for the carcinogenicity of atrazine, several uncertainties remain. In
6 themselves they do not discount the postulated mode of action. Although
7 the available data show that attenuation of the LH surge and disruption of
8 the estrous cycle occur before mammary tissue and pituitary gland tumor
9 formation, precise dose and time correlations are not available for each of
10 the key events due to differences in study design and dose selection.
11 Serum LH values are highly variable within dose groups, which makes it
12 very difficult to determine accurately biologically relevant doses that are
13 associated with effects.
14
15 There is some concern that the lack of direct effects on the pituitary
16 was established using ectopic pituitaries and using prolactin secretion as
17 a marker of LH secretion. There also is a lack of robust data on blood
18 prolactin measures and serum estradiol measurements. Because
19 prolactin measurements are not available from chronic studies,
20 confirmation of the role of the hormone in the formation of the
21 histomorphologic changes in mammary tissue is not possible.
22 Histomorphologic markers are, however, generally viewed as valid
23 indicators of prolactin secretion.
24
25 Stop-dose studies to demonstrate that induced toxicological
26 processes leading to cancer are reversible are limited but this deficiency
27 is offset, once again, by the lack of effects in ovariectomized female SD
28 rats.
29
30 Finally, the initial interaction between atrazine and the rat brain has
31 not been established, albeit effects on hypothalamic catecholamine
32 neurotransmitter levels have been shown.
33
46
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.4.7 Preliminary Conclusions on the Postulated Mode of
2 Carcinogenic Action
3
4 Support for atrazine's mode of action comes from several
5 lines of evidence.
6
7 Q Atrazine's induced LH and cyclicity effects have been shown
8 in two different laboratories and in two different strains of
9 rats (LE and SD);
10
11 Q A strong correlation has been shown for atrazine induced
12 persistent estrus and induction of mammary tumors;
13
14 Q Generally, there is a strong temporal and dose-response
15 correlation between tumor formation and precursor effects;
16 precise correlations are not possible due to differences in
17 study designs and dose selection;
18
19 Q Although robust data on estrogen and prolactin levels are
20 not available, ovariectomized SD rats treated with atrazine
21 do not develop tumors, thus demonstrating the role of
22 ovarian estrogen in atrazine's mode of action;
23
24 Q A strong correlation was demonstrated between increased
25 pituitary weights and histomorphological markers of prolactin
26 exposure in the mammary gland, thus supporting the role of
27 prolonged estrogen and prolactin exposure in tumor
28 development; and
29
30 Q Despite the lack of precise effective dose levels (LOAELs),
31 data from multiple hormonal and carcinogenicity studies
32 show that no effects of atrazine treatment are observed at a
33 dose level between 0.5 and 1.8 mg/kg/day.
34
47
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 The absence of data on the detailed steps in the hypothalamus,
3 would provide insights regarding the mechanism of action of atrazine.
4 However, as stated in the proposed Guidelines for Carcinogen Risk
5 Assessment (USEPA 1999), mode of action is contrasted with mechanism
6 of action which implies a more detailed molecular description of events
7 than does mode of action. The focus of a mode of action analysis is on a
8 sequence of key events which lead to cancer formation and whether data
9 are sufficient to establish a cause and effect relationship between key
10 events and cancer. This guidance was followed in reaching the
11 conclusion stated below.
12
13 Given the overall strengths, consistency, and specificity of the
14 evidence, it is concluded that it is biologically plausible that treatment of
15 female SD rats with atrazine leads to an increased incidence and/or
16 decreased latency in the formation of mammary adenomas, carcinomas,
17 fibroadenomas, and pituitary adenomas through a mode of action
18 involving disruption of the hypothalamic-pituitary-ovarian axis. Disruption
19 of the axis leads to suppression of LH surges, prolonged days spent in
20 estrus, and exposure of mammary tissue and the pituitary gland to
21 estrogen for a extended period. Exposure of the pituitary gland to
22 estrogen stimulates the secretion of prolactin. Exposure of the mammary
23 tissue to estrogen and prolactin and the pituitary gland to estrogen
24 creates an endogenous endocrine milieu conducive to cell proliferation
25 and tumor formation. The available data do not support a role for direct
26 mutagenic or direct estrogenic activity for effects attributed to atrazine
27 treatment.
28
29
48
-------
DRAFT: DO NOT CITE OR QUOTE
1 2.5 Reproductive and Developmental Toxicity
2
3 The natural progression from prepubertal to postpubertal status is
4 dependent upon the normal function of the hypothalamic-pituitary-gonadal axis.
5 Likewise, many of the same hypothalamic mechanisms controlling pituitary
6 function and the pituitary hormones themselves (especially LH and prolactin)
7 play a key role in pubertal development. For example, it has been shown that an
8 increased turnover rate in hypothalamic GnRH, NE and DA precedes the
9 gonadal development (Matsumoto et a/., 1986; Ojeda, 1986).
10
11 At the time of puberty (e.g., vaginal opening and first ovulation) the CMS
12 and pituitary respond to increased concentrations of estradiol in a positive
13 feedback fashion culminating in the first LH surge (Ojeda and Urbanski, 1994).
14
15 These processes are not specific to the rat. Inhibition of GnRH release in
16 neonatal rhesus monkeys suppresses gonadatrophin secretion and testosterone
17 production; this effect is associated at the time of puberty with compromised
18 testicular growth and testosterone secretion (Plant, 1994). This same author
19 postulated that there is a coupling between a rise in circulating LH and FSH
20 concentrations and the transition into puberty that is a general characteristic of
21 sexual maturation in higher primates. Thus, given that atrazine treatment of rats
22 suppresses GnRH, LH, and prolactin release, there is a concern for potential
23 adverse reproductive and developmental effects of atrazine in maternal animals
24 and their offspring.
25
26 Adverse reproductive or developmental consequences have been
27 identified following treatment of different strains of pregnant rats or neonates with
28 atrazine or its metabolites. As noted in Chapter 1, Section 1.7, this evidence
29 does not come from results of EPA guideline studies but from results of special
30 studies conducted with atrazine or its metabolites. The results of these studies
31 show that atrazine or its metabolites produce effects in pregnant, neonatal, or
32 young adult SD, F344, Wistar, Holtzman, or LE rats that may be associated with
33 disruption of the hypothalamic-pituitary axis. The developmental/reproductive
34 effects observed in these studies include reductions in implantation sites, failure
35 to maintain pregnancy, attenuation of suckling-induced prolactin release and the
36 development of prostatitis, delayed vaginal opening, and delayed preputial
37 separation. Table 2-2 provides a listing of the lowest NOAELs and LOAELs
38 reported for these effects. NOAELs and LOAELS for effects on prolactin and LH
49
-------
DRAFT: DO NOT CITE OR QUOTE
1 release following acute or short-term repeat dosing treatment of rats with
2 atrazine are also provided in order to allow comparisons of
3 developmental/reproductive NOAELs/LOAELs with the NOAEULOAELs that
4 result in disruption of neuroendocrine parameters.
5
6 Treatment of young Wistar rats with atrazine during PND 22-41 delays
7 vaginal opening three or four days and produces irregular cycles (Laws et a/.,
8 submitted; 2000). Treatment of weanling male Wistar rats with atrazine during
9 PND 23-53 leads to delays in preputial separation (Stoker et a/., submitted;
10 2000a). No consistent effects on serum progesterone or LH were observed in
11 this study but variability in hormonal levels in animals of the age studied makes
12 comparisons with control animals difficult.
13
14 In addition to affecting the onset of puberty, the offspring of dams
15 exposed to atrazine have also been found to be affected adversely.
16 Administration of atrazine to dams during PND 1-4 inhibits suckling-induced
17 prolactin release in the dams and leads to lateral prostate inflammation in the
18 offspring. The effect on the prostate is reversible if the offspring are treated with
19 ovine prolactin, which provides evidence that prolactin has a role in the
20 development of prostatitis. Also, the incidence and severity of lateral prostate
21 inflammation correlates with decreases in serum levels of prolactin. The effects
22 on the prostate of offspring of dams treated with atrazine and the delays in
23 pubertal developmental observed when young rats are treated with atrazine are
24 associated with the endocrine imbalances that have been identified as critical
25 events in the neuroendocrine mode of action attributed to the carcinogenic
26 activity of atrazine.
27
28 As stated earlier, atrazine may also affect pregnancy maintenance in the
29 rat. The full-litter resorptions reported following atrazine exposure on GD 6-10
30 (roughly coinciding with the LH-dependent period of pregnancy) are consistent
31 with a neuroendocrine mode of action (Narotsky et at, submitted). Although this
32 effect was observed at maternally toxic doses (as defined by a decrease in body
33 weight), treatment after the LH-dependent period caused a similar degree of
34 maternal toxicity, but had no effect on pregnancy maintenance. Hormone
35 measurements on GD 9 (following treatment on GD 1-8) did not show a
36 consistent pattern across strains for prolactin, estradiol, or progesterone.
37 However, for the Holtzman strain, the only strain of four tested to show full-litter
38 resorptions following treatment on GD 1-8, there were reductions in serum
50
-------
DRAFT: DO NOT CITE OR QUOTE
1 progesterone and LH; although not proof, these data are consistent with an LH-
2 mediated mechanism of pregnancy loss. In contrast to treatment on GD 6-10,
3 exposure on GD 1-8 did not cause full-litter resorption in F344 or SD rats. An
4 explanation for the lack of effect following exposure on GD 1-8, remains unclear,
5 but may be related to the truncated dosing regimen within the LH-dependent
6 period, or examination of the litter on GD 9, possibly prior to the actual time of
7 pregnancy loss.
8
9 Because pubertal development is under neuroendocrine control, it may be
10 expected that administration of atrazine to young rats leads to delays in vaginal
11 opening or preputial separation. The dose levels that led to delays in vaginal
12 opening also produced irregular ovarian cycles in offspring, which supports a role
13 for disruption of neuroendocrine control in young animals treated with atrazine or
14 its metabolites. The reductions in implantation sites and the full-litter absorptions
15 reported following treatment of dams with atrazine during the LH-dependent
16 phase of pregnancy are also consistent with an effect on neuroendocrine control
17 but other modes of action can not be discounted (e.g., general toxicity at high-
18 dose levels).
19
20 There are uncertainties, in particular, regarding the dose-response data
21 on preputial separation (PPS). A statistically-significant effect was reported at a
22 dose-level of 13 mg/kg/day (PPS -42 days in controls and PPS - 44 days at 13
23 mg/kg/day, the lowest dose tested). It should be noted that this dose has a
24 significance of p <; 0.05. The next higher dose of 25 mg/kg/day approached
25 statistical significance but did not achieve significance (i.e., p= 0.07). Statistical-
26 significance (p £ 0.05) was achieved at the next three dose levels (50,100, or
27 150 mg/kg/day). At 200 mg/kg/day there was a statistically-significant effect of
28 delayed preputial separation (-42 days in controls and -45 days in the high-dose
29 rats). There was a significant dose-related decrease in LH; however, no
30 statistically-significant effects were observed for testosterone or prolactin
31 concentrations. The variability in levels of these hormones in young rats should
32 be considered before much weight is placed on these data.
33
34
51
-------
DRAFT: DO NOT CITE OR QUOTE
1 In summary, reproductive and developmental effects in various strains of
2 rats that are associated with atrazine treatment include preimplantation and
3 postimplantation losses, prostatitis in adult male offspring of treated lactating
4 females, delays in vaginal opening and preputial separation, and disruption of
5 the estrous cycle in young females. A reduction in prolactin release in nursing
6 dams is strongly associated with the development of prostatitis in male adult
7 offspring. Decreases in serum LH or prolactin were not observed to occur at
8 dose-levels that led to delays in vaginal opening (50 mg/kg/day) and preputial
9 separation (13 mg/kg/day) in the same study but it is presumed that the
10 variability in levels of these hormones in juvenile animals preclude obtaining
11 definitive data. On the other hand, a separate study using dams showed that a
12 daily dose of ~13 mg/kg/day was sufficient to depress serum levels of prolactin in
13 the lactating dam. To the extent that decreased prolactin levels can serve as a
14 marker for effects on neuroendocrine control, there is a linkage between pubertal
15 development and an effect on the hypothalamic-pituitary axis.
16
52
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Table 2-2. Lowest NOAELs/ LOAELs (mg/kg/day) for Reproductive and
Developmental Effects Following Short-term (1-30 Days) Treatment
of Rats During Various Stages of the Reproductive Cycle with
Atrazine or its Metabolites
Response
Preimplantation loss-nocturnal
dosing only
Postimplantation loss-diumal and
nocturnal dosing
Oams prolactin release decreased
Increased incidence of prostatitis
in offspring
Increased incidence and seventy
of prostatitis in offspring
Delayed vaginal opening
Delayed preputial separation
Attenuation of LH surge
Attenuation of prolactin release
Disruption of estrous cycle
Exposure
Period
GD1-8
GD6-10
PND1-4
PND1-4
PND1-4
PND 22-41
PND 23-53
Adult females -
single dose
3 daily doses
21 daily doses
21 daily doses
30 daily doses
Dams-
GD1-8
Adult females -
single dose
3 daily doses
21 daily doses
21 daily doses
PND 22-41
Rat Strain
F344
Holtzman
Wistar
Wistar
Wistar
Wistar
Wistar
LE
LE
LE
SD
SD
LE&
Holtzman
LE
LE
LE
SD
Wistar
NOAEL/LOAEL
50/100
50/100
13/25
13/25
25/50
25/50
<13/13
200/300
<50/50
<75/75
<75/75
5/40
50/100
200/300 serum
<50/50 pituitary
<75/75 pituitary
<75/75 pituitary
25/50
Reference
Cummings et
a/., submitted
.Cummings ef
a/., submitted
Stoker et a/.,
1999
Stoker era/.,
1999
Stoker ef a/.,
1999
Laws ef a/.,
submitted; 2000
Stoker ef a/.,
submitted;
2000a
Cooper ef a/.,
2000; Morseth.
1996a;
Cummings ef
a/ , submitted
Cooper ef a/.,
2000
Laws ef a/ ,
submitted; 2000
18
19
20
53
-------
DRAFT: DO NOT CITE OR QUOTE
1 Chapter 3
2
3 3. Science Policy Considerations: Human Relevance, Children's Health
4 Concerns, and Dose-Response Analysis
5
6 This Chapter evaluates and characterizes the human relevance of the rat
7 toxicological findings of atrazine and postulated mode of action. This analysis focuses
8 on the question of whether the mode of action found to be operative in rats is also
9 operative in humans and whether any human subpoulations or life stage are apt to
10 qualitatively respond to the mode of action differently than the general population. The
11 key questions and rationales are presented in addressing the issue of human
12 relevance. Also, based on the mode of action understanding, a dose-response
13 extrapolation approach is proposed for atrazine.
14
15 3.1 Human Relevance
16
17 3.1.1 Potential Neuroendocrine Disruption
18
19 As discussed in Chapter 2, there are data supporting an
20 understanding of how atrazine induces tumor development in the rat.
21 Briefly, the mode of carcinogenic action underlying mammary and pituitary
22 gland tumor formation in female SD rats involves a lack of adequate
23 secretion of pituitary LH to stimulate ovulation, the development of
24 persistent estrus, and prolonged stimulation of the mammary and pituitary
25 glands by estrogen and prolactin. These hormones promote cell
26 proliferation and predispose cells to become neoplastic. Other
27 endocrinopathies found in the rat (e.g., delayed puberty, prostatitis) are
28 also associated with the neuroendocrine effects of atrazine on pituitary
29 function (i.e., secretion of LH and/or prolactin).
30
31
54
-------
DRAFT: DO NOT CITE OR QUOTE
1 There is clear evidence (discussed in Chapter 2) that atrazine
2 alters hypothalamic GnRH release in rats. There are some data that
3 show that atrazine diminishes NE in the rat hypothalamus as a initial or
4 early site of action which in turn leads to diminished GnRH release.
5 Atrazine also increases dopamine levels which can result in a diminished
6 pituitary prolactin secretion. Therefore, a key question to address is
7 whether this neuroendocrine mode of action at the level of the
8 hypothalamus may be operative in humans. In both humans and rats,
9 hypothalamic GnRH controls pituitary hormone secretion (e.g., LH,
10 prolactin). The hypothalamic-pituitary axis is involved in the development
11 of the reproductive system, and its maintenance and functioning in
12 adulthood. Additionally, reproductive hormones modulate the function of
13 numerous other metabolic processes (i.e., bone formation, and immune,
14 CMS and cardiovascular functions) (Cooper et al., 1986, Plant, 1994).
15 Given that the primary site of atrazine's effect on GnRH secretion in the
16 rat is at the level of the hypothalamus, it is important to address the
17 questions below:
18
19 Question. Is there evidence in primates including humans of
20 central neural modulation of GnRH secretion by the
21 hypothalamus? Is this central mechanism conserved
22 across species?
23
24 Although GnRH secretion is influenced by a number of factors in
25 primates and humans (such as circulating steroids), and the precise
26 control mechanisms remain to be fully understood, the prevailing view is a
27 central neural control system is involved in governing GnRH release (as
28 reviewed by Marshall and Eagleson, 1999; Plant, 1994). For example,
29 there have been studies in both rats and primates showing that CNS-
30 altering drugs (e.g., opiates) can alter the menstrual cycle or pubertal
31 development (see review by Plant, 1994; Ojeda, 1986). Further, there is
32 evidence that endogenous opioids are involved in GnRH/LH secretion in
33 primates (Ferin and Van de Wiele, 1984), indicating that GnRH neurons
34 are modulated by other hypothalamic neural inputs like in the rat.
35 Therefore, if atrazine affected the hypothalamic GnRH in humans like in
36 the rat, it is plausible to assume that this neuroendocrine mode of action
37 would apply to humans.
38
55
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.1.2 Potential Human Health Consequences Associated with
3 Altered GnRH/Pituitary Function
4
5 Given that the rat neuroendocrine mode of action may be operative
6 in humans, it is important to address:
7
8 Question: What neuroendocrinopathies may result in humans if
9 exposed to atrazine?
10
11 Atrazine interferes with the CMS control of pituitary-ovarian function
12 and leads to irregular cycles and inhibition of ovulation in SD and LE rats.
13 In humans and primates reproductive function/ovarian cycling is also
14 influenced by the hypothalamic GnRH (Goldfien and Monroe, 1997; Plant
15 1994; Nishihara era/., 1992; Terasawa and Nyberg, 1997). Therefore, a
16 potential consequence in humans is disrupted or irregular menstrual
17 cycles which can lead to gynecological problems such as diminished
18 fertility, prolonged menses or excessive bleeding.
19
20 It is important to evaluate what is understood about the role of
21 altered GnRH secretion in human ovulatory disorders. Also, a further
22 evaluation of human anovulatory conditions may give some clues as to
23 potential downstream endocrine effects and other health consequences.
24 Hypothalamic amenorrhea (HA) is one model of disrupted cyclicity. HA is
25 a manifestation of a variety of disorders associated with emotional stress,
26 heavy exercise, self-imposed weight loss and oral contraceptive use and
27 occurs in the absence of pathology in the pituitary and ovaries (Reame et
28 a/., 1985). HA has been found to represent a spectrum of disordered
29 GnRH secretion (presumably low frequency and variable or low amplitude
30 pulses) that can vary over time (Perkins ef a/., 1999). Clinically, persons
31 fail to ovulate, as in atrazine treated SD rats. HA is characterized by
32 normal to moderately low serum estrogen and normal to low serum LH.
33 When serum LH is lowered, the cause appears to be a reduction in
34 hypothalamic GnRH secretion (Perkins ef a/., 1999). These
35 manifestations of HA are similar to those seen with atrazine treated SD
36 and LE rats: decreased hypothalamic GnRH, decreased pituitary LH, and
37 failure to ovulate. These observations suggest that certain of the
38 manifestations may be the same in humans and rats if atrazine affects the
56
-------
DRAFT: DO NOT CITE OR QUOTE
1 hypothalamic neurons in similar ways. In addition to the gynecological
2 problems associated with disrupted ovarian cycling,. HA patients can
3 suffer other health consequences. For example, they can be at an
4 increase risk of osteoporosis later in life given that these women are
5 estrogen deficient, and thus can experience significant losses in bone
6 density. Women who are hypoestrogenic may also suffer from vasomotor
7 symptoms, urogenital atrophy, cardiovascular disease, and possibly
8 diminished cognitive and memory functions (Wren, 1997).
9
10 Polycystic ovary syndrome (PCOS) is another model of
11 anovulation, which occurs in 6% to 8% of premenopausal women
12 (Marshall and Eagleson, 1999). PCOS is often characterized by irregular
13 menstrual cycles or amenorrhea, infertility, obesity, and ovaries that are
14 polycystic with many unovulated follicles in various stages of development
15 and atresia. Hirsutism is associated with PCOS (Schildkraut et a/., 1996;
16 Hershlag and Peterson, 1996). Some of the other manifestations of
17 PCOS are very different from that seen in atrazine treated rats. There
18 commonly is an increase in LH secretion from the pituitary and increased
19 synthesis of androgens (hyperandrogenism) and their conversion to
20 estrogens. This can result in unopposed exposure to estrogen. The
21 mechanism underlying the excess ovarian androgen secretion is unknown
22 but may be multifactorial, and include abnormalities of steroidogenesis,
23 effects of hyperinsulinemia, and abnormal gonadotropin secretion in
24 stimulating ovarian steroidogenesis (Ehrmann etal., 1995; Utiger, 1996;
25 Marshall and Eagleson, 1999). PCOS is not an exact model for
26 evaluating the consequences of atrazine exposure in humans, other than
27 in some cases it is associated with abnormal GnRH secretion (with
28 presumably high frequency-low amplitude pulses), anovulation, and
29 unopposed exposure to estrogen.
30
57
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 As discussed in Chapter 2, atrazine has also, been shown to
3 increase the hypothalamic neurotransmitter, dopamine, which in turn
4 results in a decrease of pituitary prolactin secretion in female rats. In both
5 rats and humans, prolactm is one of the hormones involved in
6 lactogenesis. It is the suckling action of the neonate that stimulates
7 prolactin secretion, and thus the maintenance of milk production.
8 Therefore, in humans, diminished production and secretion of milk could
9 result if atrazine were to affect hypothalamic dopamine and suppress
10 prolactin as in the rat. Given that the initial sucking induced prolactin
11 response is relatively robust, atrazine exposure would not be anticipated
12 to impact the initiation of lactation, but could potentially impact the ability
13 to sustain milk production with continuous exposure.
14
15 Therefore, there is support from the primate literature that
16 atrazine's neuroendocrine mode of action (CMS perturbation of GnRH
17 secretion) may apply to humans. Human ovulatory disorders can be
18 associated with aberrant hypothalamic GnRH pulses. These conditions
19 indicate that altered hypothalamic GnRH secretion can broadly affect an
20 individual's functional status, and thus lead to a variety of clinically
21 important health consequences. These human conditions, HA and
22 PCOS, do not prove but raise the possibility that if atrazine produced
23 effects on hypothalamic GnRH in the human, like that seen in atrazine-
24 treated rats, adverse health effects may ensue. The potential ability of
25 atrazine to affect dopamine and prolactin in humans must also be
26 considered. Below, the potential human cancer risk associated with this
27 neuroendocrine mode of action is discussed.
28
58
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.1.3 Potential Cancer Risk Associated with Altered GnRH/Pituitary
3 Function
4
5 It is standard Agency practice to assume that chemically induced
6 tumors in animals have human relevance unless there are data to the
7 contrary (US EPA 1999a). Target organ concordance is not necessarily a
8 prerequisite for evaluation of the implications or relevance of animal tumor
9 findings for humans. Even if there is a mode of carcinogenic action
10 understanding for the rodent tumor findings, site concordance may or may
11 not be expected. In the case of atrazine, there is an increased incidence
12 and early onset of mammary gland tumors in female SD rats. As
13 discussed below, it does not seem plausible that humans would be at an
14 increase risk for breast cancer given that atrazine would potentially reduce
15 the cumulative number of normal ovarian cycles (i.e., one of the risk
16 factors for humans). In fact, the neuroendocrine mode of action for
17 atrazine raises the possibility of tumor development at other hormone-
18 responsive site.
19
20 In assessing potential human risk, human data are generally
21 preferable over animal data when of good quality, and should be given
22 greater weight in the hazard characterization of an agent. Therefore, an
23 obvious question to address is:
24
25 Question: Are there data in humans to determine the human
26 cancer potential and neuroendocrine mode of action
27 for atrazine?
28
29 As summarized in Chapter 1 (and discussed in detail in Part B-Chapter 4),
30 there is suggestive evidence of a possible association of triazine exposure
31 and cancer occurrence for three hormone-responsive cancers-ovary,
32 breast and prostate cancer. However, these associations should not be
33 considered as conclusive evidence of an association of triazine exposure
34 with these tumor types. There are no human or primate studies that
35 directly examine the potential for atrazine to induce endocrine effects as
36 have been described in the SD or LE rat special studies.
37
38
59
-------
DRAFT: DO NOT CITE OR QUOTE
1 One important aspect of atrazine's postulated mode of
2 carcinogenic action involves components in common with the reproductive
3 aging process in SD female rats. It is well recognized that in SD female
4 rats, as well as in other strains of rats such as LE and Wistar, reproductive
5 aging is due to failure of hypothalamtc-pituitary-gonadal function resulting
6 in the normally aging female spending an increased percentage of days of
7 their ovarian cycle in estrus (i.e., constant estrus) (discussed in detail in
8 Part B-Chapter 9.1). Therefore, an aging female SD rat experiences a
9 dampening of the preovulatory pituitary LH surge which results in
10 prolonged exposure to estrogen. In contrast, the prevailing view for
11 humans is that reproductive aging results from a depletion of follicles from
12 the ovary (i.e., atresia). However, the potential that an age-associated
13 loss of the hypothalamic control of GnRH secretion may contribute to
14 significant changes in menstrual function during the perimenopausal
15 period in women can not be discounted (e.g., Wise et al., 1996; 1999).
16
17 Nevertheless, to the extent that the carcinogenic effects of atrazine
18 in SD rats are intimately tied to an interaction between effects of the
19 chemical and the normal aging process in rats, then there may be
20 questions as to the applicability of the carcinogenic effects to humans.
21
22 Question: Can atrazine lead to cancer through a process not
23 involving reproductive aging; and can the
24 neuroendocrine effects of atrazine alone set up a
25 milieu favorable to the development of cancer in
26 humans?
27
28 In addressing the above questions, it is important to note that of the
29 key events identified in Figure 2-1 based on laboratory in vitro and in vivo
30 data, atrazine's initial site of action appears to be at the level of the
31 hypothalamus (i.e., effects on hypothalamic catecholamine and GnRH
32 levels). As discussed above, CMS control of hypothalamic GnRH is
33 similar in primates and humans, and human conditions of anovulation,
34 which can be associated with aberrant GnRH release, lead to a variety of
35 health consequences. It is important to look at the cancer risk associated
36 with the human ovulatory conditions discussed above.
37
60
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 HA has not been found to be associated with a cancer risk based
3 on epidemiologic studies, although it is clearly associated with other
4 health consequences as discussed above. Because of the prevalence of
5 PCOS in the population, several epidemiologic studies have assessed its
6 role in breast cancer. One showed a significantly increased risk, but only
7 in the postmenopausal period (Coulam et al., 1983); the remaining three
8 failed to show breast cancer increases (Gammon and Thompson, 1990;
9 Anderson eta!., 1997; Pierpoint etal., 1998). Such findings have been
10 interpreted as lending little support for PCOS being a risk factor for breast
11 cancer (Solomon, 1999). A small number of patients, however, have
12 enough estrogen to maintain the endometrium, which has the potential to
13 become hyperplastic over time (Mansfield and Emans, 1989; Schachter
14 and Shoham, 1994); endometrial hyperplasia is a risk factor for
15 endometrial cancer (Rose, 1996). Case reports suggest that PCOS may
16 predispose women to endometrial cancer at an early age, in contrast to
17 this cancer's usual occurrence with advancing age (Jafari et al., 1978;
18 Dahlgren et al., 1991). A statistically-significant increase in relative risk for
19 endometrial cancer was noted among documented PCOS patients
20 (Coulam et al., 1983). Information linking PCOS to ovarian cancer is less
21 well developed. One study of epithelial ovarian cancer showed a
22 statistically-significant increase of persons with PCOS (Schildkraut et al.,
23 1996), while another did not (Coulam et al., 1983).
24
25
61
-------
DRAFT: DO NOT CITE OR QUOTE
1 The human conditions of anovulation only supply inferential
2 information concerning potential cancer risks. These human disease
3 models do not prove a potential cancer risk associated with atrazine's
4 neuroendocrine mode of action. But on the other hand, they do not allow
5 one to discount the possibility that if atrazine produced effects on
6 hypothalamic GnRH like is seen in atrazine treated SD rats, disrupted
7 cyclicity may result in an endocrine environment that may be conducive to
8 tumor development at hormone-responsive sites. Mammary gland site
9 concordance with SD rats should not be expected (as discussed further
10 below), but the mode of action responsible for the rat tumors and
11 information on PCOS raise the possibility of other endocrine sites (i.e.,
12 endometrial and ovarian). Also, conditions of human anovulation
13 disorders suggest that atrazine exposure alone may produce an
14 endocrine imbalance that may be conducive to tumor development.
15 Given that hypothalamic GnRH control of the preovulatory pituitary LH
16 surge is similar in rats and primates, it seems possible that this process
17 could be independent of the reproductive aging pattern as seen in rats.
18
62
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.1.4 Breast Cancer
3
4 Given that atrazine induced mammary gland tumors in SD rats, it is
5 important to evaluate what is understood about endocrine influences for
6 breast cancer.
7
8 Q Estrogen seems to be an important influence in breast
9 cancer development, as indirect indicators of estrogen
10 stimulation are known risk factors for the disease: early age
11 of menarche, late onset of menopause and nulliparity.
12 However, it seems that the cumulative number of regular
13 and not irregular ovarian cycles is the important input into
14 breast cancer development (Henderson et a/., 1988; den
15 Tonkelaar and de Waard, 1996). Consistent with this,
16 regular exercise is associated with reduction in breast
17 cancer risk, possibly by reducing the number of normal
18 ovulatory cycles as is seen in hypothalamic amenorrhea
19 (Bernstein ef a/., 1994).
20
21 Q Prolactin plays a role in mammary gland carcinogenesis in
22 rodents, but its importance in human breast cancer
23 development is not at all established. Prolactin together with
24 estrogen, stimulates the human breast tissue during
25 lactation. Unlike rats where there are significant changes in
26 prolactin levels throughout the ovarian cycle, there is little
27 modification during the human menstrual cycle (Goldfien and
28 Monroe, 1997). Rats and humans do show circadian
29 variations in prolactin. Two prospective studies among
30 postmenopausal women have found increases in breast
31 cancer with elevated prolactin levels, although only one was
32 statistically-significant; retrospective studies of
33 premenopausal women have been variable in their
34 outcomes (Wang ef a/., 1992; Hankinson ef a/., 1999).
35
63
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Many different Pharmaceuticals induce hyperprolactinemia,
3 including the raulwolfia drugs used to treat hypertension, tricyclic
4 antidepressants, antipsychotic phenothiazines and methyldopa, the drug
5 used to treat Parkinson's disease. A number of epidemiologic studies
6 have been conducted with the raulwolfia derivatives. Some have noted
7 no increase in breast cancer risk while others have indicated rather limited
8 increases in postmenopausal women (Shapiro et a/., 1984; Williams et a/.,
9 1978). It has been argued that agents which produce about a 50%
10 increase in prolactin levels may account for the small increase in cancer
11 risk in some of the studies (Ross et a/., 1984). One investigation showed
12 that with dosing for at least 10 years or with initiation of dosing at least 10
13 years prior to diagnosis, significant risk ratios of about four were found
14 (Stanford et a/., 1986). Antidepressants also lead to increases in prolactin
15 levels. The relationship between their use and breast cancer have led to
16 differing outcomes (CoHerchio et a/., 2000; Kelly et a/., 1998; Wallace et
17 a/., 1982). There has been less investigation of other psychiatric drugs
18 that produce hyperprolactinemia and its association with increases in
19 breast cancer risk. An investigation of all 9156 schizophrenic patients in
20 Denmark that had there first hospital admission between 1970-1987
21 showed no indication of increase in breast cancer risk (Mortensen, 1994),
22 in keeping with other studies. More work is needed to probe these
23 relationships.
24
25 Interestingly, mammary gland and breast cancers have receptors
26 for prolactin, and studies show that prolactin mRNA and the hormone
27 itself are synthesized by tumor cells (Clevenger et a!., 1995; Mershon et
28 a/., 1995). It has been hypothesized that the local formation of prolactin
29 may serve autocrine or paracrine functions within the mammary gland
30 (Ben-Jonathan ef a/., 1996; Vondehaar, 1999). These observations
31 reopen the question of the role of prolactin in human breast cancer
32 development. As of yet, the regulation and effects of locally synthesized
33 prolactin on the breast have not been determined.
34
64
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.1.5 Cancer Classification
3
4 In the past, OPP had classified atrazine as a Group C, possible
5 human carcinogen based on an increased incidence of combined
6 mammary carcinomas/adenomas and fibroadenomas in female SD rats, in
7 accordance with the 1986 cancer risk assessment guidelines. Recently,
8 the OPP Cancer Assessment Review Committee (CARC) proposed that
9 atrazine should be classified as a likely human carcinogen in
10 accordance with the draft 1999 revisions to the cancer risk assessment
11 guidelines (i.e., US EPA, 1999a). The basis of this current proposal is as
12 follows:
13
14 Q Consistent findings in female SD rats of an increased
15 incidence and early onset of mammary gland
16 carcinomas/adenomas in several studies, and suggestive
17 evidence of an early onset of pituitary adenomas and
18 mammary fibroadenomas;
19
20 Q Mode of action evidence that indicates hypothalamic
21 disruption of GnRH control of pituitary function by atrazine,
22 and critical reductions in LH and resultant anovulation; and
23
24 Q Similarity in humans and rats for CMS control of pituitary
25 function.
26
27 Therefore, if atrazine affected hypothalamic GnRH as in the rat,
28 this opens the possibility that an endocrine imbalance may result which
29 could lead to several different health consequences including cancer at
30 hormone responsive tissues.
31
65
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.2 Potential Health Effects of Atrazine in Children
3
4 3.2.1 Reproductive/Developmental Hazard
5
6 The data summarized in Chapter 2 indicates that the primary
7 underlying process that leads to mammary and atrazine involves
8 disruption of the hypothalamic pituitary-gonadal-axis. pituitary gland tumor
9 development in female SD rats following treatment with This axis is also
10 involved in reproductive development. Therefore, as summarized in
11 Chapter 1.7, it is not surprising that atrazine treatment also results in
12 adverse reproductive and developmental outcomes in special studies
13 using several different strains of rats (i.e., F344, SD, Wistar, LE,
14 Holtzman). These outcomes include interruption of regular ovarian
15 cycling, decreased suckling induced prolactin release and increased
16 incidence and severity of prostatitis, and delays in vaginal opening and
17 preputial separation.
18
19 Rat and human reproductive development and puberty are under
20 similar hypothalamic-pituitary control, especially LH and prolactin
21 (Matsumoto et a/., 1986, Ojeda, 1986). After the first trimester in humans,
22 fetal LH and FSH are used to complete genital maturation (Hsing, 1997).
23 There is an appreciable release of LH commencing at parturition that
24 extends until four to six months of postnatal life. Thereafter, LH is
25 suppressed until puberty begins. There is a re-awakening of the
26 hypothalamtc-pituitary-gonadal axis at puberty. The exact mechanism
27 underlying this pubertal LH release is unknown. For male sexual
28 development, LH is required to stimulate the Leydig cells for testosterone
29 production, and androgens are responsible for the outward signs of
30 pubertal development. LH and FSH are required to begin ovarian
31 activation, follicle growth, and steroid production in female sexual
32 development. Estrogen secreted from the ovary triggers breast growth
33 and other body changes. Some adolescent patients with delayed puberty
34 display low levels of LH and/or FSH (Styne 1997; Kulin 1996). Therefore,
35 there is concern that if children were exposed to atrazine and if it affected
36 the hypothalamic-pituitary-gonadal axis and the pituitary LH and PRL
37 releases as in rats, there is the potential for delayed puberty or altered
38 pubertal growth in both female and male adolescents. Delayed puberty is
66
-------
DRAFT: DO NOT CITE OR QUOTE
1 not without health consequences. For example, girls with delayed
2 menarche show a higher incidence of scoliosis, stress fractures, and
3 osteopenia than do girls with normal time of menarche (Goldfien and
4 Monroe, 1997). Additionally, abnormal puberty may result in problems
5 manifested later in life (e.g., osteoporosis (Styne, 1997).
6
7 Exposure to atrazine in lactating dams (Wistar rats) suppresses
8 suckling-induced prolactin release which eventually results in
9 hyperprolactinemia and prostatitis in the lateral prostate in young adult
10 offspring. It is reasonable to assume that this suppression of pituitary
11 prolactin secretion in the dam is due to atrazine's effect on hypothalamic
12 catecholamine levels (i.e., dopamine). Prolactin does play a role in the
13 development and maintenance of the human prostate. Critical periods for
14 developmental exposures and the hormonal involvement in the induction
15 of prostatitis remain unknown in humans. In humans, nonbacterial
16 prostatitis of undefined etiology is an important clinical problem that has
17 been associated with infertility (Meares, 1998; Huaijin et a/., 1998). There
18 is a suggestion in the literature that chronic proliferative inflammation in
19 the prostate may be a precursor event to prostatic carcinogenesis (De
20 Marzo et al., 2000; Leav et a/., 1999). It should be acknowledged that the
21 relevance of effects in the rat prostate as a human model has been
22 debated. However, Because the dorsal and lateral prostate of the rat are
23 considered to be the most homologous to the human prostate (Price,
24 1963), the increase in inflammation observed in young male rat offspring
25 should not be discounted.
26
27 In summary, because of the similarity between rats and humans of
28 the influence of hypothalamic GnRH on the growth and morphogenesis of
29 the reproductive system, the concern is raised about the potential health
30 effects due to early life exposure to atrazine, some of which may not be
31 manifested until later in life.
32
67
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.2.2 Cancer Hazard
3
4 As stated in the July 1999 Draft revisions to the EPA's cancer risk
5 assessment guidelines, when information is developed to show a mode of
6 carcinogenic action that is expected to be relevant to adults, an evaluation
7 needs to be made as to whether this mode of action is relevant to
8 children. When there is no cancer information on children perse, a
9 "cogent biological rationale needs to be developed regarding whether the
10 mode of action is applicable to children." In the case of atrazine, although
11 there are no animal data directly evaluating its neoplastic potential from
12 pre- and postnatal exposures perse, there is information indicating that
13 atrazine can affect the hypothalamic-pituitary axis and cyclicity in young
14 animals. So reliance is placed upon both data concerning the
15 neuroendocrine effects in young animals as well as using biological
16 arguments to evaluate children's cancer concern.
17
18 If atrazine were to produce neuroendocrine effects in humans like it
19 does in SD rats, projections can be made as to potential consequences in
20 children, using what is understood about the key events described for its
21 postulated mode of action. Components of the neuroendocrine system
22 develop during fetal life, with varying manifestations at different times. As
23 discussed above, the preovulatory LH surge controlling ovulation does not
24 happen until puberty. Considering the purported mode of atrazine action
25 involving attenuation of the preovulatory LH surge and disruption of
26 ovarian cycling as a critical event, it is reasonable to assume that this
27 mode of action may also be operative in children from puberty onward.
28 Furthermore, the rodent cancer bioassays on atrazine as well as the
29 accompanying LH/cyclicity mode of action studies used young pubertal
30 rats (six to eight weeks of age). Thus, there is a potential cancer concern
31 for children as a result of exposure during puberty and continued over a
32 lifetime. The rat studies on decreased suckling induced prolactin release
33 and increased incidence and severity of prostatitis in male offspring,
34 however, raise the question of whether prepubertal exposure may lead to
35 a potential prostate cancer risk later in adult life. At this time there is no
36 indication of such an outcome, however, conventional cancer testing may
37 not screen for such potential. Further study would be needed to
38 determine whether there is or is not any hazard capability.
-------
DRAFT: DO NOT CITE OR QUOTE
1 3.2.3 Summary of Children's Health Concern
2
3 Rat studies using atrazine treatment in utero or during early life
4 demonstrate a wide spectrum of endocrinopathies (e.g., delayed puberty,
5 disrupted cycling, prostatitis, reproductive organ weight changes,
6 hyperprolactemia) associated with the disruption of the neuroendocrine
7 control of pituitary function. There are numerous studies in the literature
8 indicating that altered neuroendocrine status in children lead to a variety
9 of health outcomes. Furthermore, as discussed previously, CNS-GnRH
10 control of reproductive development is similar in primates and rats. Thus,
11 the rat studies on atrazine raise concern for the susceptibility of the fetus
12 and young child if exposed to atrazine. The consequence in children due
13 to this neuroendocrine mode of action would depend on the
14 developmental stage of exposure and the duration of exposure. For
15 example, prepubertal exposures would most likely result in developmental
16 effects, and postpubertal exposure may result in a variety of health
17 consequences including cancer. There is no direct information on cancer
18 responses following pre- or postnatal exposure.
19
20 3.3 Summary of Atrazine Human Hazard Potential
21
22 As shown in Figure 3-1, atrazine operates via a neuroendocrine mode of
23 action that alters hypothalamic GnRH and pituitary LH and PRL secretions. It is
24 recognized that across species and even among different strains of a species
25 endocrinological interactions can differ significantly (Neumann et a/., 1996).
26 However, atrazine's central neuroendocrine mode of action is likely to be
27 operative in humans given that in both rats and primates a central neural control
28 influences GnRH and pituitary function. The variety of endocrinopathies found in
29 the atrazine treated rats (e.g., mammary and pituitary gland tumors, delayed
30 puberty, disrupted cyclicity, prostatitis in young rats) raise concern about the
31 potential human health consequences that may ensue from this neuroendocrine
32 perturbation, including adverse reproductive and developmental outcomes or
33 delayed acquisition of normal reproductive potentialities. This neuroendocrine
34 mechanism also raises concern for potential cancer risk in humans.
69
-------
DRAFT: DO NOT CITE OR QUOTE
Figure 3-1. Atrazine's Neuroendocrine Mode of
1
2
Hfedts on Bore WbssCererty
CEteopcfcss
Caioer
Physidogy
(BxboineDsorclers)
Drririshsd Fertility
CarticvasolarBfects
OS effects
70
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.4 Dose-Response Analysis
3
4 In 1988, the U.S. EPA presented a dose-response assessment of atrazine
5 (Hauswirth 1988a; US EPA 1988). That assessment used the female SD
6 mammary tumor incidence from the study by Mayhew et a/. (1986) and the
7 linearized multistage (LMS) model to estimate an oral slope factor and a unit risk
8 of 2.22 x 10'1 [mg/kg/day]'1. The current dose-response analysis considers the
9 mode of action data as discussed in Chapter 2. Additionally, the two-step
10 approach to dose response assessment as described in the proposed revisions
11 to U.S. EPA Guidelines for Carcinogen Risk Assessment (US EPA, 1999a) are
12 utilized in this dose-response analysis. This two-step process distinguishes
13 between the observed range of empirical data and the range of extrapolation.
14
15 The weight of evidence does not support mutagenicity nor direct
16 estrogenicity as components of atrazine's mode of carcinogenic action. As
17 discussed in Chapter 2, the weight of evidence supports a conclusion that
18 atrazine acts to cause mammary and pituitary gland tumors in female Sprague-
19 Dawley rats by causing a attenuation of the preovulatory surge of LH which
20 results in anovulation and an endocrine milieu that is conducive to tumor
21 development. The critical event, the attenuation of the LH surge, is consistent
22 with a nonlinear phenomenon in that there is a dose of atrazine that does not
23 affect the LH surge or disrupt cyclicity. Therefore, it is proposed that dose-
24 response assessment should proceed by a margin of exposure analysis.
25
26 An increased incidence and/or early onset of mammary and pituitary
27 gland tumors in the rat is only one endocrinopathy found after atrazine treatment.
28 The reproductive and developmental consequences (e.g., disrupted cyclicity,
29 delayed puberty, prostatitis in male offspring) that are found after atrazine
30 treatment are of equal concern. These reproductive/developmental effects also
31 originate from the effects of atrazine on the hypothalamic control of pituitary
32 function through its interference with hypothalamic catecholamines and GnRH
33 neurotransmitters. Thus, given the commonality in the mode of action, it is
34 recommended that a point of departure for dose-response extrapolation be
35 based on the most sensitive effects associated with atrazine's neuroendocrine
36 mode of action.
37
38
71
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
3.4.1 Selecting a Point of Departure
Margin of exposure analysis begins with selection of a point of
departure (POD) considered to represent the lowest reliable endpoint in
the range of observation, being either tumor incidence, a key endocrine
related effect or data on a proximal event that is an integral part of the
mode of action process. Figure 3-2, provides an overview of the
NOAELs/LOAELs for key endocrine related effects of atrazine from
various studies in rats at different life stages and for different treatment
durations. As discussed in Chapter 1 and Part C, the NOAELs for the
effects of atrazine on pregnancy, pubertal onset and prostatitis are, for the
most part, at or above 25 mg/kg/day. The exceptions are:
Figure 3-2. Key Endocrine-Related Effects
Following Atrazine Treatment of Rats
GD6-10 PND1-4 PND22-41 PND23-53 3 days 21 days 30 days -«months 24 months
V
Timing and Duration of Exposure
Adult Females .LDT = iowest dose Iested
72
-------
DRAFT: DO NOT CITE OR QUOTE
1 Q The NOAEL for delay of preputial separation in Wistar rats is
2 not clear as a significant delay was seen at ~13 mg/kg/day
3 (albeit near statistical significance; p = 0.07), but not at the
4 next highest dose of 25 mg/kg/day; and
5
6 Q The NOAEL for dam decreased prolactin release and
7 resultant prostatitis in male offspring is -13 mg/kg/day.
8
9 Dose-response data from long term repeat dosing studies are
10 lacking for the effects on hypothalamic catecholamines and GnRH, i.e.,
11 atrazine's initial site of action. However, it is the pulsatile GnRH secretion
12 from the hypothalamus that determines the pituitary LH secretion (a
13 critical event in atrazine's mode of action). Therefore, it is assumed that
14 effects on LH secretion are a mirror of effects on GnRH secretion and that
15 data on the serum LH are reasonable surrogate measures of the GNRH
16 secretion. There are more data over different doses and time points for
17 the attenuation of the LH surge. Thus, this is an appropriate POD. As
18 illustrated in Figure 3-2, the NOAELs/LOAELs for the LH data are
19 compared to the other key end points (mammary gland tumors, increased
20 days in estrus, delayed puberty, suppression of suckling-induced prolactin
21 and resultant prostatitis) that result from this neuroendocrine mode of
22 action (also see Tables 1.9,1.10, 2.1, and 2.2).
23
24 Selecting the NOAELs for the attenuation of the LH surge to
25 determine a point of departure rather than from curve-fitting in the
26 observable range of LH data is done here because it is not known over
27 just what level of attenuation of the LH surge is necessary in order to
28 produce clinically relevant effects. As discussed in the draft 1999
29 guidelines for carcinogen risk assessment, "the observed range of data
30 may be represented by a NOAEL/LOAEL procedure when a margin of
31 exposure analysis is chosen as the default procedure for nonlinear dose-
32 response extrapolation" (US EPA,1999a).
33
73
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 As the treatment duration is increased, the dose that is needed to
3 attenuate the preovulatory LH surge decreases. For example, it takes an
4 extremely high dose of a single day of dosing of atrazine to attenuate the
5 LH surge (i.e., NOAEL = 200; LOAEL = 300mg/kg; Cooper et a/., 1999).
6 However, with longer durations of dosing, much lower doses of atrazine
7 can attenuate the LH surge (i.e., NOAEL = -2 mg/kg; LOAEL = ~4 mg/kg
8 after six months of dosing; Morseth, 1996b). As shown in Figure 3-1,
9 several types of reproductive/developmental effects can arise in postnatal
10 rats following a few days of dosing up to several weeks of dosing with
11 atrazine (e.g., delayed puberty, prostatitis, increased days in estrus). As
12 depicted in Table 1-10, NOAELs for these reproductive effects range
13 froml 3 mg/kg/day up to 100 mg/kg/day.
14
15 With respect to effects that result from longer durations, LOAELs
16 for precursor events associated with carcinogenesis (increased days in
17 estrus, attenuation of the LH surge) and tumors consistently ranged
18 between ~3 to 4 mg/kg/day. Likewise, NOAELs for various parameters
19 were ~2 mg/kg/day or higher in all cases except one. In the Mayhew
20 (1986) study, a significant tumor increase was noted at ~4 mg/kg/day but
21 not at the lowest dose tested, 0.5 mg/kg/day. Based on consideration of
22 the all the bioassay studies in SD rats and the repeat dose LH studies, as
23 well as consideration of the dose spread in the Mayhew (1986) bioassay,
24 LOAELs for carcinogenic, LH, and cyclicity effects tended to be
25 approximately 4 mg/kg/day and NOAELs tended to be ~2 mg/kg/day.
26 Clearly, there is a correspondence of doses that lead to tumor formation
27 and doses that produce effects on LH levels and cyclicity. Thus, the point
28 of departure for chronic effects is the dose of 1.8 mg/kg/day which is the
29 NOAEL for attenuation of the proestrus afternoon LH surge in Morseth
30 (1996b).
31
74
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.4.2 Point of Departure Using LED10 From The Tumor Data
3
4 Although data based on the attenuation of the LH surge is the
5 preferred POD, for comparative purposes PODs based on the modeling of
6 tumor data to derived LED10.S are also presented. The most appropriate
7 study to use in selecting a point of departure for tumors is Morseth, 1998.
8 Five bioassays using the SD rat are available which examine tumor
9 incidence and early onset. One of these studies (Petterson and Turnier,
10 1995) is a one year study and is not deemed appropriate for that reason.
11 Another study (Thakur, 1991 a) is not considered because only two dose
12 groups were used and the study employed many serial sacrifices which
13 resulted in a very small "n" value by the later timepoints in the study. A
14 third bioassay (Thakur, 1992a) used only two dose groups. The two
15 remaining studies, Mayhew (1986) and Morseth (1998), which employed
16 four dose groups, both may be considered for use in selecting a point of
17 departure. LED10s for both of these studies are presented in Table 3-3.
18 And ranged from ~2 to 3 mg/kg/day for mammary gland carcinomas and
19 adenomas combined. These values represent equivalent human doses.3
20 The NOAELs/LOAELs for mammary gland tumors are 0.5 and 3.5
21 mg/kg/day; and 4.2 and 24.4 from the Mayhew and Morseth studies,
22 respectively. It should be noted that Morseth (1998) provides time to
23 tumor information and used contemporary criteria for pathological
24 evaluations. Also, Morseth (1998) had accompanying estrus cycling data.
25
26 The NOAEL for the LH surge attenuation and the LED10 for
27 carcinomas and adenomas from Morseth (1998) are 1.8 mg/kg/day (i.e.,
28 0.48 mg/kg/day in human equivalents. Therefore, a POD based on the
29 NOAEL for attenuation of the LH surge is comparable to a POD based on
30 tumor response.
Conversion to human equivalents performed by multiplying the rat dose in mg/kg/day
by 0.266.
75
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
Table 3-2. LED10s in Human Equivalents (And Revised Q/)
Study
Mayhew, 1986
Mayhew, 1986
Morseth, 1998
Morseth, 1998
Morseth, 1998
Mammary Gland Tumors
Combined adenomas, carcinomas, and
adenosarcomas
Fibroadenomas
Combined adenomas and carcinomas
Fibroadenomas
Incidence of combined carcinomas and
adenomas
LED10 (mg/kg/day)
2.1
3.0
1.8
3.5
{Q/M.12X10-1
mg/kg/day)
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Data in this table from US EPA, 1999b and 1999d.
Table 3-2, also provides a revised Q* estimate for comparison
purposes only. Given the mode of action understanding for atrazine, the
nonlinear extrapolation approach is preferred over the linear default
approach. The linear extrapolation is not supported by the mode of action
data.
3.5 Summary and Conclusions on the Proposed OPP Science Policy
Positions: Mode of action, Human Relevance, Children's Health
Concerns, and Dose-Response Extrapolation
Listed below are the proposed science policy conclusions regarding the
postulated mode of carcinogenic action in SD female rats. The relevance of the
rat reproductive/developmental studies and the female SD rat tumor findings
their mode of action to humans, including concerns for children.
Recommendations are also made for the dose-response approach that should
be considered in the cancer risk assessment.
76
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.5.1 Postulated Rat Tumor Mode of Action
3
4 Members of the pesticide program Cancer Assessment Review
5 Committee (CARC) reviewed information on atrazine bearing on the
6 formation of mammary and pituitary tumors in female SD rats. The CARC
7 concluded that the increased incidence and early onset of mammary
8 gland carcinomas and adenomas were well supported by several rat
9 bioassay studies. The evidence for an early onset of mammary
10 fibroadenomas and pituitary adenomas was considered to be suggestive.
11
12 Based on the Mode of Action Framework Analysis presented in
13 Chapter 2, judgments were made on three considerations underpinning
14 the mode of action of these tumors. The Committee agreed that:
15
16 Q Atrazine does not have a significant mutagenic component
17 to its mode of action;
18
19 G Direct atrazine binding to the estrogen receptor is not an
20 influence on tumor development; and
21
22 Q The neuroendocrine mode of action for the mammary and
23 pituitary tumors is "biologically plausible" and is supported
24 overall by the weight of the evidence.
25
26 As discussed in Chapter 2, there are several strengths of the mode
27 of action proposal. For example, atrazine's induced LH and cyclicity
28 effects have been shown in two different laboratories and in two different
29 strains of rats (LE and SD). Furthermore, there is a strong correlation has
30 been shown for atrazine induced persistent estrus and induction of
31 mammary tumors. Generally, there is a strong temporal and dose-
32 response correlation between tumor formation and precursor effects.
33 Ovariectomized SD rats treated with atrazine do not develop tumors, thus
34 demonstrating the role of ovarian estrogen in atrazine's mode of action.
35 Finally, a strong correlation was demonstrated between increased
36 pituitary weights and histomorphological markers of prolactin exposure in
37 the mammary gland, thus supporting the role of prolonged estrogen and
38 prolactin exposure in tumor development. Although significant amounts of
77
-------
DRAFT: DO NOT CITE OR QUOTE
1 data have been developed to demonstrate how atrazine may produce
2 mammary and pituitary tumors in SD rats, there are uncertainties or
3 limitations in the available data base (as discussed in Chapter 2.4.5). It
4 should be emphasized that the uncertainties or limitations in the data in
5 themselves do not discount the postulated mode of action, and that the
6 strengths of the data provides compelling evidence in support of the
7 postulated mode of action. However, the uncertainties/weaknesses in the
8 data should be should be considered in the final risk characterization.
9
10 3.5.2 Relevance of Rat Mode of Action to Humans and
11 Carcinogenicity Classification
12
13 It is proposed that the postulated mode of action is assumed as
14 being relevant to human cancer potential given that a primary initial
15 site of action in rat involves the CNS control of pituitary function. It
16 is EPA science policy that animal tumor responses are presumed to be
17 indicative of human cancer potential unless there is substantive
18 information to the contrary. This default is intended to be public health
1 g protective and departure from this default must have a strong
20 accompanying scientific basis. OPP views the differences between
21 reproductive aging in humans and rats as an insufficient scientific basis to
22 depart from the default. Therefore, if atrazine were to act on the
23 hypothalamus of humans as in the rat and caused CNS alterations which
24 influence endocrine function on physiological processes including ovarian
25 cycling, there is the potential for various adverse health outcomes,
26 including cancer.
27
28 The OPP Cancer Assessment Review Committee proposed that
29 atrazine should be classified as a likely human carcinogen (US EPA,
30 1999a).
31
78
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 3.5.3 Children's Hazard
3
4 Data are available from animal studies on atrazine to assess
5 potential effects in children that may be associated with its
6 neuroendocrine mode of action. Based on the endocrinopathies found
7 in postnatal rats, it is reasonable to assume that children would
8 potentialty be susceptible to atrazine's neuroendocrine mode of
9 action which may lead to a variety of health consequences (See
10 section 3.2). How atrazine's neuroendocrine mode of action is manifested
11 depends on the life stage exposed as well as the duration and level of
12 exposure. Data following prepubertal exposures in rats demonstrate
13 adverse developmental effects including delay in puberty and prostatitis.
14 In reference to the mammary tumors in rats and their mode of action, a
15 cogent biological rationale informs that situation. LH secretion is
16 quiescent until puberty. Therefore, it is not expected that atrazine would
17 pose a cancer hazard following prepubertal exposure. However, starting
18 with exposures at puberty, cancer hazard may be evident. As with adult
19 exposures, certain endocrine responsive sites in the female may be at risk
20 for cancer development.
21
22
79
-------
DRAFT: DO NOT CITE OR QUOTE
1 3.5.4 Dose-Response
2
3 Based on atrazine's mode of carcinogenic action, a nonlinear
4 dose-response extrapolation approach is the preferred approach for
5 quantifying the cancer risk. A cancer hazard in adults resulting from
6 infant and children exposure to atrazine cannot be ruled out. Infants and
7 children, however, would not be expected to demonstrate a unique
8 susceptibility to tumors induced by this mode of action, with the possible
9 exception of an increased postpubertal risk of tumors. In order to assure
10 adequate protection of all susceptible subpopulations (i.e., women and
11 children) for both cancer and noncancer effects for potential exposures
12 throughout their lifetime, it is recommended that the health risk
13 assessment be performed utilizing the most sensitive endooint
14 associated with atrazine's neuroendocrine mode of action. ANOAEL
15 of ~2 mg/kg bw/day, based on attenuation of the LH surge following six
16 months of atrazine treatment, is recommended as the point of departure
17 for the health risk assessment using the MOE approach. For continuous
18 exposures, this NOAEL is viewed as appropriate given atrazine's
19 neuroendocrine mode of action which potentially leads to a variety of
20 health consequences including cancer, and is viewed protective of all
21 populations (including women and children).
22
23 3.6 Other Reviews
24
25 There have been a number of reviews on the carcinogenicity of atrazine
26 by other organizations:
27
28 Q Draft report of the Cornell University Program on Breast Cancer
29 and Environmental Risk Factors in New York State (Snedeker and
30 Clark, 1999);.
31
32 Q The International Agency for Research on Cancer (IARC, 1999);
33
34 Q The National Registration Authority for Agricultural and Veterinary
35 Chemicals of Australia (NRA, 1997);
36
37 Q The United Kingdom - in a report to the European Commission
38 (United Kingdom Pesticide Directorate, 1996);
80
-------
DRAFT: DO NOT CITE OR QUOTE
1
2 Q A report published by a U.S. consulting group under contract to the
3 Triazine Network (a national coalition of grower organizations and
4 individuals) (Cantox, 2000); and
5
6 Q A consensus report of a scientific panel commissioned by Novartis
7 Crop Protection (Consensus Panel, 2000).
8
9 It should be noted that the current EPA draft atrazine assessment has
10 generally reached similar conclusions with the above reviews on several issues
11 concerning the carcinogenicity of atrazine (see Table 3-3). There appears to be
12 consensus that mutagenicity and direct binding to the estrogen receptor do not
13 play a significant role in atrazine's carcinogenic action in SD rats (IARC, 1999;
14 Snedeker and Clark, 1999; Cantox 2000; Consensus Panel 2000; NRA, 1997;
15 United Kingdom Pesticide Directorate, 1996). Further, these reviews have also
16 concluded that an endocrine mode of carcinogenic action in SD rats is
17 biologically plausible and is supported by the evidence. Although there is
18 general agreement about support for a mode of action, there are different views
19 on the role of accelerated reproductive senescence in the SD rat tumor
20 response. For example, the United Kingdom Pesticide Directorate (1996) states
21 that the reproductive aging hypothesis is not adequately proven, but that the
22 tumors do appear to be caused by a "disturbance of endogenous hormone
23 levels." Also, Snedeker and Clark (1999), concluded that there were
24 inconsistencies or lack of data on certain hormonal measures (such prolactin and
25 estradiol) which did not lead support to the premature reproductive aging
26 hypothesis, but "there is evidence that it can affect hormones along the
27 hypothalamic pituitary gonadal axis."
28
29 Although there is general agreement among different organizations, there
30 are differences in the conclusions regarding human relevance and cancer
31 classification. Snedeker and Clark (1999) concluded that atrazine is a "possible
32 breast carcinogen." This document concludes that site concordance should not
33 be assumed and that the potential exists for cancer at other hormone-responsive
34 sites (e.g., endometrium). Several other organizations including IARC (1999)
35 concluded that the mode of carcinogenic action in SD rats in not relevant to
36 humans. EPA/OPP may have had more data on the mode of action than these
37 reviews, particularly on the hypothalamus as a primary site of action (Cooper et
38 a/., 2000; Das et a/., 2000). But more importantly, these analyses considered
81
-------
DRAFT: DO NOT CITE OR QUOTE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
the disruption of hypothalamic control by atrazine in a broader sense leading to
several neuroendocrinopathies (e.g., delayed puberty, prostatitis, mammary
gland tumors) in the rat, rather than focusing on the reproductive aging process
and induction of mammary gland tumors in rats. Unlike these reviews, this
analyses evaluated the neuroendocrine controls of pituitary function in rodents
and primates, including humans, and concluded that there is a potential for
carcinogenic effects independent of reproductive aging, and that primates may
have some aging components in common with rat. Also, the LH response were
not limited to SD rats also found in LE rats. The reproductive effects were also
found in other strains such as Wistar rats.
Table 3-3. Other Reviews on the Carcinogenicity of Atrazine*
EPA/OPP
(This Draft)
Snedeker and
Clark (1999)
IARC (1999)
Cantox (2000)
Consensus Panel
(2000)
NRA (1997)
United Kingdom
Pesticide
Directorate (1996)
Mutagenic
No
u
H
U
U
u
u
Direct
Estrogenicity
No
u
u
u
u
u
Mode of
Carcinogenic
Action
Support
Some Support
Support
Support
Support
Support
Support
Human Cancer
Concern
"Likely human
carcinogen"
"Possible breast
carcinogen"
"Not relevant1
(Group 3)
'Not likely to be
carcinogenic"
"Not relevant"
"Not considered
to be relevant"
"A strong case
for non-
classification"
82
-------
DRAFT: DO NOT CITE OR QUOTE
PartB
Hazard Assessment and Review of
Available Studies
-------
DRAFT: DO NOT CITE OR QUOTE
List of Contents
4 Human Epidemiological Studies 7
4.1 Colon Cancer 7
4.2 Non-Hodgkins Lymphoma 7
4.3 Soft Tissue Sarcoma 8
4.4 Other Hematologic Cancers: Hodgkins Disease, Leukemia, Multiple
Myeloma 9
4.5 Ovarian Cancer 9
4.6 Breast Cancer 10
4.7 Prostate Cancer 10
4.8 Stomach Cancer 10
4.9 Summary 10
4.10 Conclusions 12
5 Chronic Rodent Bioassay Studies 14
5.1 Mayhew etal., 1986 17
5.2 Thakur Studies 19
5.2.1 Serial Sacrifice Protocol (Thakur, 1991a) 19
5.2.2 Terminal Sacrifice Protocol (Thakur, 1992a) 21
5.3 Morseth, 1998 22
5.4 Pettersen and Turnier, 1995 25
5.5 Hazelette and Green, 1987 26
5.6 F-344 Two-Year Bioassays 27
5.6.1 Serial Sacrifice Protocol (Thakur, 1991b) 27
5.6.2 Terminal Sacrifice Protocol (Thakur, 1992b) 28
5.7 Pinter et al., 1990 30
5.8 Pinter et al., 1990 30
5.9 Summary and Discussion of the Two-Year Bioassay Studies 33
6 Genotoxicity Studies 35
6.1 Mutation Studies 36
6.2 Chromosome Aberration Studies 38
6.2.1 In VffroAssays 38
6.2.2 In Vivo Assays 40
6.3 Other Indicators of DNA Damage or Mutagen Exposure 41
6.4 Mutagenicity Studies in Plants 42
6.5 Metabolites of Atrazine 42
6.5.1 Diaminochlortriazine metabolite (DACT - 6-chloro-1,3,5-triazine-
-------
DRAFT: DO NOT CITE OR QUOTE
2,4-diamine; didealkyl atrazine) 43
6.5.2 G-28279 metabolite - 6-chloro-N-ethyl-1,3,5-triazine-2,4-diamine;
deisopropyl atrazine 43
6.5.3 G-30033 metabolite - 6-chloro-N-(1 -Methyl ethyl)-1,3,5-triazine-2,4-
diamine; deethyl atrazine 43
6.5.4 Hydroxy-atrazine, G-34048 44
6.6 Close Structural Analogues: Simazine and Propazine 45
6.6.1 Simazine 46
6.6.2 Propazine 47
6.7 Summary and Discussion of Mutagenicity Data 48
7 Estrogenic Activity 49
7.1 In Vivo Assays 49
7.2 In Vitro Assays 53
7.3 Special Carcinogenicity Bioassay Study (Morseth, 1998) 57
7.4 Noncancer Effects Relevant to Estrogenic Activity 58
7.4.1 Subchronic Dog Studies 59
7.4.2 Subchronic Rat Studies 61
7.4.3 Chronic Dog Studies 62
7.4.4 Multi-Generation Reproduction Studies 63
7.4.5 Rat Developmental Toxicity Studies 65
7.4.6 Rabbit Developmental Toxicity Studies 66
7.5 Overall Conclusions of Estrogenic Activity Data 68
8 Structure Activity Relationship 69
9 Hormonal and Estrus Cyclicity Studies 74
9.1 Rat Reproductive Aging Process 76
9.1.1 Sprague-Dawley 76
9.1.2 Fischer-344 79
9.1.3 Summary of the Reproductive Aging Process in SD and F-344 Rats
80
9.1.4 Strain Differences in Reproductive Aging and Mammary Tumors
83
9.1.5 Studies in Which Premature Aging Was Artificially Delayed or
Induced and How Mammary Tumor Incidences Were Affected . 86
9.1.6 The Correlation Between Increased Days in Estrus and Mammary
Tumors in a Two-Year Bioassay 87
9.1.7 Summary and Discussion of the Hypothesis that Mammary Tumors
-------
DRAFT: DO NOT CITE OR QUOTE
Are Induced by the Reproductive Aging Process 88
9.2 The Hypothesis that Atrazine Exposure Induces an Early Onset of
Attenuated LH Surge, Increased Days In Estrus, and Prolonged Exposure
to Estradiol 89
9.2.1 Time-to-Tumor 90
9.2.2 Conclusions Of The Time-to-Tumor Data 98
9.2.3 Alterations in the Ovary and Vagina 98
9.2.4 Summary And Discussion From The Ovarian Histomorphology and
Estrous Cycle Measurements In F-344 and SD Strains 109
9.2.5 Serum Estradiol and Prolactin Levels 111
9.2.6 Summary And Discussion Of The Hormone Measurements and
Histomorphologic Alterations In F-344 And SD Strains 117
9.2.7 Preovulatory LH Levels 121
9.2.8 Summary And Discussion Of The LH Surge Studies 128
9.3 The Site of Action for Atrazine Attenuation of the LH Surge 130
9.4 The Data Examining the Association Between Atrazine Exposure and An
Attenuated Proestrus Afternoon LH Surge, Increased Days and Estrus
and a Prolonged Exposure to an Elevated Level of Estradiol 132
9.4.1 Atrazine Exposure Results in an Earlier Onset of Increased Days in
Estrus 133
9.4.2 Atrazine Exposure Results in an Earlier Onset of Increased Serum
Estradiol Levels 133
9.4.3 Atrazine Exposure Results in an Earlier Onset of Attenuated LH
Surges 134
9.4.4 Atrazine Exposure Results in an Earlier Tumor Onset 134
9.5 Pituitary Adenomas 135
9.5.1 Onset of Pituitary Alterations Following Atrazine Exposure ... 135
9.5.2 Role of Early Onset of Pituitary Alterations in Mammary
Carcinogenesis 136
9.5.3 Pathogenesis of Pituitary Alterations 137
9.5.4 Summary and Conclusion for Pituitary Alterations 138
-------
DRAFT: DO NOT CITE OR QUOTE
List of Tables
Table 4-1. Odds Ratios (OR), Risk Ratio (RR) or Correlation Coefficient 13
Table 5-1. Summary of Female Mammary Tumor Incidence
in Two- and One-Year Rodent Bioassays Using Atrazine 15
Table 5-2. Summary of Female Pituitary Adenoma Incidence
in Two- and One -Year Rodent Bioassays Using Atrazine 16
Table 5-3. Mammary Tumor Incidence in the
Mayhew Study (as determined by US EPA, 1988) 18
Table 5-4. Mortality in the Mayhew Study (as determined by US EPA, 1988) 18
Table 5-5. Pituitary 3-Adenoma Incidences by
Timepoint in Thakur Serial Sac, 1991a 20
Table 5-6. Female Mammary Gland Tumor Incidences in
the SD Terminal Sacrifice Protocol (Thakur, 1992a)
(calculated using Cox-Tarone and Gehan-Breslow tests) 22
Table 5-7. Mammary Gland Tumor Incidence
in Intact Animals in Morseth, 1998 Study 24
Table 5-8. Number Of Animals With Mammary
Tumors In The Pettersen and Turnier, 1995 Study 26
Table 5-9. Female Mammary Tumor Incidence In F-344 Terminal
Sacrifice Protocol In The Thakur Terminal Sacrifice Study (1992b) 29
Table 5-10. Mammary Tumors In Males in Pinter et a/., 1990 31
Table 5-11. Mammary Gland Fibroadenomas in Male
F-344 Rats by 12 Week Time Periods (Solleveld, 1984) 32
Table 7-1. In vitro and in vivo Hormonal Studies with Atrazine 50
-------
DRAFT: DO NOT CITE OR QUOTE
Table 8-1. Results of Two-Year Bioassays with Alkylamino,
alkoxy, and alkythio-triazine Compounds 70
Table 8-2. Results of Two-Year Bioassays with Chloro-triazine Compounds 70
Table 9-1. Mammary Tumor and Pituitary Adenoma Historical
Control Incidence Data For Sprague-Dawley Females At 24 Months 75
Table 9-2. Relationship of Reproductive Aging and
Mammary Tumor Incidence In Various Rodent Strains 85
Table 9-3. Time to Mammary Tumor in the Female
SD Rat- Thakur (1992a) Terminal Sacrifice Protocol 91
Table 9-4. Time to Mammary Tumor in the Two-year Morseth (1988) Study 92
Table 9-5. Time to Mammary Tumor with Simazine
in SD rats (McCormick et a/., 1988) 93
Table 9-6. Time to Mammary Tumor with Propazine in SD rats (Jessup, 1980a) ... 94
List of Figures
Figure 8-1. Structures of Atrazine and Major Metabolites 71
Figure 8-2. Structure of the Amino-s-Triazine Ring 72
Figure 8-3. Structure of Chlorsulfuron 72
Figure 8-4. Structures of Simazine and Propazine 73
Figure 9-1. Summary of the Reproductive Aging Process in Different Rat Strains .. 82
6
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 4
4 Human Epidemiological Studies
Several epidemiologic studies have examined cancers among populations with
exposures relevant to the assessment of atrazine, especially among farmers or farm
residents (Table 4-1). Most are case control studies, although others include ecologic
investigations and a worker mortality study associated with triazine manufacturing.
4.1 Colon Cancer
Associations between herbicide use by farmers and colon cancer in
Kansas was investigated in a case-control study (Hoar et a/., 1985). Starting
with 57 cases of colon cancer and 948 controls, the odds ratio (OR) for the
subset using triazine herbicides was 1.4 (95% C.I. 0.2-7.9). The sample size in
this group was very small with only two cancer cases with confirmed exposure to
triazines and 43 controls. The study author stated that the data did not support
an association between colon cancer and herbicide exposure
An ecologic study of ecodistricts in Canada compared triazine exposures
and cancer incidences (Van Leeuwen et a/., 1999). Association of triazine
exposure with several cancers, including colon, were examined. Significant
negative associations were found in both sexes (p = 0.041 in females and 0.006
in males).
4.2 Non-Hodgkins Lymphoma
Zahm et a/. (1993b) pooled results of three case-referent studies
conducted in three midwestern states that investigated atrazine exposure in the
development of non-Hodgkins lymphoma (NHL). Starting with 993 males with
NHL and 2918 controls, persons were queried as to their pesticide exposures.
An OR = 1.4 (95% Cl 1.1-1.8) was found for atrazine use and NHL However,
when adjustments were made for use of 2,4-dichloroacetic acid and
organophosphate use, the OR = 1.2 (0.9-1.7). The authors concluded that there
was essentially no risk of NHL attributable to farm use of atrazine.
-------
DRAFT: DO NOT CITE OR QUOTE
NHL was investigated among women who lived or worked on a farm in
eastern Nebraska (Zahm et a/., 1993a). The OR for those who reported that
they lived on a farm where atrazine was used was 1.4 (95% Ci 0.6 - 3.0) with 11
cases and 31 controls. For those women who reported having personally used
atrazine the OR = 2.2 (0.1 - 31.5) with only one case and two controls.
Cyanazine was also investigated and the OR for NHL and those who reported
using cyanazine was 1.3 (0.3 - 4.5) with four cases and 12 controls. The study
author noted that there were too few subjects in any of these analyses to
adequately assess associations.
Correlations were made between pesticide use (1993) and NHL incidence
(1988-1992) among California counties (Mills, 1998). There were negative
correlations for white males and females and Hispanic males; the correlation was
positive for Hispanic females (0.12), but it was not statistically-significant.
An ecologic study of ecodistricts in Canada compared triazine exposures
and cancer incidences (Van Leeuwen et a/., 1999). Association of triazine
exposure with several cancers, including NHL, were examined. No association
was found in females and a negative association was found in males.
A mortality study of workers in two triazine manufacturing plants that was
supplied to EPA (Delzell and Sathiakumar, 1996) did not find any significant
excesses of deaths for any disease category. There were, however, two cases
of NHL in plant workers - one of whom was relatively young (31 years). These
two cases do not provide evidence of an association between atrazine exposure
and NHL, but do indicate that further follow-up of workers in these triazine
manufacturing plants would be helpful.
4.3 Soft Tissue Sarcoma
A population-based case-control study of soft-tissue sarcoma (STS) in
Kansas demonstrated that there was no increased risk among farmers (Hoar et
a/., 1986). The lack of association persisted when years of herbicide use or
frequency of herbicide use were considered. Analyses examining atrazine
specifically were not conducted.
A study previously described above (Mills, 1998) examined associations
between pesticide use and STS. Positive correlations were not noted for either
males or females for atrazine use and STS.
8
-------
DRAFT: DO NOT CITE OR QUOTE
4.4 Other Hematologic Cancers: Hodgkins Disease, Leukemia, Multiple
Myeloma
Hoar etal. (1986) examined associations between herbicide use and
Hodgkins disease (HD). An OR for atrazine exposure and HD was not reported
in the study, but for herbicide use in general the OR was 0.9 (95% Cl 0.5 -1.5).
The study author did not consider herbicide exposure to be associated with HD.
Association of pesticide exposure (including atrazine and cyanazine) to
leukemia was investigated in a population-based case-control study of adult
white men in Iowa and Minnesota (Brown et a/., 1990). The OR in those who
reported mixing, loading and applying atrazine or cyanazine was 1.0 (0.6 -1.5)
for atrazine and 0.9 (0.5 -1.6) for cyanazine.
Mills (1998) also examined associations between pesticide use and
leukemia. Positive correlations were not noted for either Hispanic or white males
or females for atrazine use and leukemia.
Triazine exposure and multiple myeloma in Iowa farmers was
investigated in a case-control study by Burnmeister (1990). The OR for triazine
use and multiple myeloma was 1.29 and was not significant.
4.5 Ovarian Cancer
A case-control study of epithelial ovarian cancer was conducted in woman
between the ages of 20 and 69 who lived in a province in Italy where triazine
herbicides are used in farming (Donna et a/., 1989). A relative risk (RR) of 2.7
(95% Cl 1.0 - 6.9) was found for subjects who reported that they definitely had
been exposed to triazine herbicides; the sample size in this subgroup was seven
cases and seven controls. The authors considered that there was some risk of
ovarian cancer among women who were exposed to triazines.
An ecologic study of ecodistricts in Canada compared triazine exposures
and cancer incidences (Van Leeuwen etal., 1999). Association of triazine
exposure with several cancers, including ovarian, were examined. No
association between atrazine exposure and ovarian cancer was found.
9
-------
DRAFT: DO NOT CITE OR QUOTE
4.6 Breast Cancer
An ecologic study of counties in Kentucky compared measures of triazine
exposures (ground and surface water measurements, and acres of land planted
with com and triazine application rates) with state cancer incidences (Kettles et
at., 1997). For the years 1993-1994, the OR = 1.14 (95% Cl 1.08 -1.19) for
counties with medium triazine exposure, compared with OR = 1.2 (1.13 -1.28)
for high exposure counties. Although only slightly greater than 1.0, these OR's
were still statistically-significant (p<0.0001).
4.7 Prostate Cancer
Pounds of atrazine applied in California counties during the calendar year
1993 were compared with state cancer incidences (Mills, 1998). A statistically-
significant correlation coefficient of 0.67 was obtained for blacks and prostate
cancer. The correlation coefficients for whites, Asians and Hispanics were not
statistically significant.
4.8 Stomach Cancer
An ecologic study of ecodistricts in Canada compared triazine exposures
and cancer incidences (Van Leeuwen et a/., 1999). Association of triazine
exposure with several cancers, including stomach, were examined. A significant
positive association was found in both sexes (p = 0.242 in females and 0.046 in
males).
4.9 Summary
Colon cancer does not appear to be associated with triazine exposure as
suggested by a non-significant OR of 1.4 in a single study (Hoar, 1985). The
sample size in this study for triazine use was, however, very small. An ecologic
study found negative associations between atrazine exposure and colon cancer
in both sexes (Van Leeuwen etal., 1999).
Soft tissue sarcoma does not appear to be associated with atrazine
exposure. The correlation coefficient from Mills, 1998 was not significant and the
OR from Hoar et al. (1986) was 0.9. Hodgkins disease does not appear to
associated with atrazine as indicated by an OR of 0.9 in Hoar et al. (1986).
10
-------
DRAFT: DO NOT CITE OR QUOTE
Two studies examined an association of atrazine to leukemia. One found
an OR of 1.0 (Brown, et a/., 1990) and the other found a correlation coefficient
which was not significant (Mills, 1998). Leukemia does not appear to be
associated with atrazine exposure.
Triazine exposure and multiple myeloma do not appear to be associated.
The OR in a single study was 1.29 (Burnmeister, 1990). This OR was not
significant.
The results in regards to non-Hodgkins lymphoma (NHL) are mixed, but
overall indicative of a lack of association of triazines with NHL. One study found
a significant association with an OR of 2.5 (Hoar, et a/., 1985). When the data
from this study is pooled with data from two other studies, a much lower OR of
1.4 is found (Zahm, 1993b). The Zahm, 1993b study represents the pooled data
from three separate studies. Thus, the sample size is quite large - 130 cases
and 249 controls. The positive Hoar et a/., 1985 study, by comparison, had 14
cases and 43 controls. An OR of 2.2 was found for women who had reported
using atrazine (Zahm, 1993a). The sample size in this study was very small with
only one case and two controls. A fourth study failed to find any positive
correlations for either Hispanics or white males or females for atrazine use and
NHL (Mills, 1998). The low OR in the pooled study with a large sample size,
combined with the lack of positive correlations in Mills, 1998, indicates that
atrazine has not yet been clearly shown to be associated with atrazine. Further
research in this area is desirable though given the positive association seen in
Hoar, et a/., 1985 and the previously described incidence of two cases of NHL in
workers employed at triazine manufacturing plants.
The most clear associations between atrazine and cancer occurs for
ovary, breast and prostate cancer. Interestingly, all three of these cancers are
known to be hormone-responsive. These associations should not be considered
as conclusive evidence of an association of atrazine with these tumor types
though.
11
-------
DRAFT: DO NOT CITE OR QUOTE
Two of the associations (breast and prostate cancer) were found in
ecologic studies. Ecologic studies contain inherent limitations and causal effects
can not be found from ecologic studies. The primary limitation of an ecologic
study is that chemical exposure can not be confirmed. Exposure in
epidemiologic studies can sometimes be uncertain. For example, interview-
based studies rely on a persons memory to determine exposure and a recall bias
may be evident. But in ecologic studies exposure is most uncertain. In ecologic
studies the researcher has no idea at all if the persons who contracted cancer
had any exposure at all to the chemical in question. The researcher only knows
that the person lived in a county in which the chemical was used or lived in a
county which had chemical contamination of the water supplies.
The association with ovarian cancer seen in Donna, etal., 1989, is also
weakened by confounding variables. The most dramatic weakness in this study
is the small sample size in the "defiantly exposed" group. This group consists of
only seven cases and seven controls. Furthermore, close examination of this
group reveals that it may be even smaller. A description of the exposure of the
seven women in the "defiantly exposed" group is included as an appendix to the
study. Examination of the descriptions in the appendix show that three out of the
seven did not actually recall exposures to triazines at all. Rather, these three
noted only that they had worked in fields where herbicides were used, but that
they could not recall the names of the herbicides. The small sample size limits of
the defiantly exposed group weakens the conclusions from this study.
4.10 Conclusions
The results of the human epidemiology studies do not provide clear
evidence of an association between triazines and cancer. Some of the studies,
particularly those in which hormone-responsive cancers such as breast, ovary
and prostate, were examined, are suggestive of a possible association. There is
also suggestive evidence of a possible association of triazine exposure and NHL.
Further epidemiologic research is needed - especially in the area of
hormone-responsive cancers.
12
-------
DRAFT: DO NOT CITE OR QUOTE
Table 4-1. Odds Ratios (OR), Risk Ratio (RR) or Correlation Coefficient
Study
Hoarefa/., 1985
Hoar et a/., 1986
Donna et a/., 1989
Brown et a/., 1990
Burmeister, 1990
Zahmefa/., 1993a
Zahmefa/., 1993b
Kettles eta/, 1997
Mills, 1998
Van Leu wen ef a/.,
1999
Cancer
Colon
Non-Hodgkms
lymphoma1
Ovary
Leukemia
Multiple
myeloma
Non-Hodgkms
lymphoma in
women
Non-Hodgkins
lymphoma in
men
Breast
Prostate2
Stomach3
Risk Measure or Correlation
Triazines- OR = 1 4 95% Cl= 0.2 - 7.9
Tnazmes. OR = 2.5 95% Cl= 1 .2 - 5 4
Tnazines. RR = 2.7 95% Cl= 1 .0 - 6.9
Atrazme. OR = 1.0 95%CI=0.6- 1.5
Cyanazine. OR = 0.9 95%CI=0.5- 1.6
Triazines: OR = 1 .3 95% Cl not given
Use on farm: OR = 1 .4 95% Cl= 0 6 - 3.0
Personal use: OR = 2.2 95% Cl= 0.1 - 31.5
Atrazme. OR = 1 .4 95% Cl= 1 . 1 - 1 .8
Triazine exposure
Medium: OR = 1.14 95% Cl = 1.08 - 1.19
High: OR = 1 .2 95% Cl = 1 . 1 3-1 .28
Statistically-significant correlation = 0.67
between atrazine exposure and prostate
cancer in blacks, but not in whites, Hispanics
and Asians
Female p= +0.242
Male p = +0.046
1Soft tissue sarcoma and Hodgkms disease were also investigated and determined, by
the study author, not to have a significant association with atrazme exposure.
2Leukemia, non-Hodgkms lymphoma, soft tissue, brain and testis cancer were
examined, but a significant correlation was seen only with prostate cancer.
3 Bladder, colon, brain, NHL and ovary cancer were also examined, but either no
association or a negative association was seen in each case.
13
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 5
5 Chronic Rodent Bioassay Studies
The carcinogenicity of atrazine in the female Sprague-Dawley (SD) rat has been
confirmed in several two-year bioassays. These studies show that atrazine exposure
results in an increased incidence and an early onset of mammary tumors in female SD
rats (Mayhew etal., 1986; Thakur, 199131; Thakur, 1992a; Morseth, 1998; Pettersen
and Turnier, 1995). No tumor response is seen in SD male rats, however.
A two-year bioassay in both sexes of the mouse was negative for
carcinogenicity, as were two-year bioassays in male and female F-344 rats (Hazeltte
and Green, 1987; Thakur, 1991b; Thakur, 1992b).
Table 5-1 displays summaries of the mammary tumor incidence and onset in all
the rodent bioassays that have been submitted to the Agency and also a study from the
open literature (Pinter etal., 1990). Additional details concerning these studies can be
found in the discussion that follows. Appendix Table 1 also summarizes in further detail
the results from the studies performed in the SD rat.
Table 5-2 displays summaries of pituitary adenoma incidences in all the rodent
bioassays that have been submitted to the Agency. Pituitary tumor onset is difficult to
determine as pituitary tumors are not palpable as are mammary tumors. However, in a
serial sacrifice study, an early onset of pituitary tumors can be discerned in female SD
rats (Thakur, 1991a). Only female pituitary tumor incidence is displayed in Table 5-2.
1Data from the studies referred to here as Thakur, 1991a, Thakur, 1991b, Thakur
1992a and Thakur 1992b, have been published in the open literature as Wetzel et a/.
1994.
14
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-1. Summary of Female Mammary Tumor Incidence
in Two- and One-Year Rodent Bioassays Using Atrazine
Study
Mayhew et
al .1986
Thakur,
1991a
Thakur,
1992a
Morseth,
1998
Pettersen and
Tumier, 1995
Hazelette and
Green, 1987
Thakur,
1991b
Thakur,
1992b
Pinter et al ,
1990
Species/
Strain
Rat/SD
Rat/SD
Rat/SD
Rat/SD, both
OVXand
intact
Rat/SD
Mousc/CD-1
Rat/F-344
Rat/F-344
Rat/F-344
Duration
2 year
2- year with
serial
sacrifices
2- year
2-year
1-year
91 weeks
2- year with
serial
sacrifices
2-year
Lifetime
Mammary Tumor
Incidence
Statistically-significant increase
in female carcinomas at 3 5
mg/kg/day when adjusted for
survival
A significant positive trend for
fibroadenomas is seen
No statistically-significant
increases in female
fibroadenomas or carcinomas
seen at either 3 79 or 24 01
mg/kg/day
No tumors seen in OVX animals
Carcinoma, and fibroadenoma
incidences at 3 1 mg/kg/day are
increased two-fold over control
values in intact animals
six carcinomas/adenomas and
four fibroadenomas are seen at
the 23.9 mg/kg/day group
compared to one carcinoma and
two fibroadenomas in the control
group.
No increase in any tumor in
either sex with exposures up to
386 mg/kg/day for males and
483 mg/kg/day for females
No increase in any tumor in
either sex with exposures up to
34 mg/kg/day in both
No increase in tumors of any
kind in either sex with exposures
up to 20 mg/kg/day for males
and 26 mg/kg/day for females
Statistically-significant increase
in male benign mammary tumors
Mammary Tumor Onset
Not determined in this study
The percentage of carcinomas
occurring in the first year of the
study was 0 in controls, 33% at 4 23
mg/kg/day, and 50% at 26 23
mg/kg/day
The percentage of carcinomas and
adenomas occurring in the first year
of the study in controls was 0% while
at 3 79 mg/kg and 23.01 mg/kg/day
27 3 and 33.3% of the carcinomas
appeared in the first year of the
study
The mean week of onset for
carcinomas and adenomas in
controls was 72.6 while the mean
week of onset for the 1.5, 31,42
and 24 4 mg/kg/day groups was
77 2, 78.6, 64 4 and 64 8
The increased incidence of tumors at
one year indicates an earlier onset.
Not altered in Atrazine exposed
animals
Not altered in Atrazine exposed
animals
Not altered in Atrazine exposed
animals
Increased survival in dose groups
versus controls resulted in delayed
time of onset
15
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-2. Summary of Female Pituitary Adenoma Incidence
in Two- and One -Year Rodent Bioassays Using Atrazine
Study
Mayhew
era/. .1986
Thakur,
1991a
Thakur,
1992a
Morseth,
1998
Petersen and
Tumier, 1995
Hazelette
and Green,
1987
Thakur,
1991b
Thakur,
1992b
Pinter et a/.,
1990
Species/
Strain
Rat/SD
Rat/SD
Rat/SD
Rat/SD.
both OVX
and intact
Rat/SD
Mouse/CD-1
Rat/F-344
Rat/F-344
Rat/F-344
Duration
2 year
2- year
with serial
sacrifices
2- year
2- year
1-year
91 weeks
2- year
with senal
sacrifices
2-year
Lifetime
Pituitary Adenoma Incidence
by Dose Group (doses in mg/kg/day)
ControN 47/68 (69%); 05 = 41/63 (65%); 3.5 = 49/68 (72%); 25=
47/65 (72%); 50= 35/63 (56%)
Control = 22/70 (31%); 4 23 = 16/70 (23%); 26.23 = 20/70 (29%)
Control = 43/58 (74%); 3.79 = 45/58 (78%); 23.01 = 46/60 (77%)
OVX- Control = 42/80 (53%); 1.5= 39/80 (49%);
3.1 = 35/80 (44%). 4.2 = 42/80 (53%); 24.4 = 41/80 (51%)
Intact- Control = 56/80 (70%); 1.5= 60/80 (75%);
3.1 = 52/80 (65%); 4.2 = 56/80 (70%); 24.4 = 54/80 (68%)
Control 2/55 (4%); 08 = 5/55 (9%); 1.7 =6/55 (11%);
2.8 = 4/55 (7%); 4.1 = 1/55 (2%); 23.9 =5/55 (9%)
Control= 0/60; 1.6 = 0/60; 47.4 = 0/60; 246.9 = 3/60 (5%); 482.7
= 0/60
Control= 9/67 (13%), 0.68 = 6/69 (9%); 4 82 = 7/65 (11%), 14.05
= 5/66 (8%); 34.33 = 5/67 (7%)
Control = 22/60 (37%). 0.49 = 26/60 (43%), 3.43 = 20/58 (34%);
9.87= 19/59 (32%), 20.17 = 13/59 (22%)
Control = 32/41 (78%); 18.75 = 23/43 (53%); 37.5 = 35/50 (70%)
16
-------
DRAFT: DO NOT CITE OR QUOTE
5.1 May hew ef a/., 1986
The initial study that raised concerns about the possible carcinogenic
effects of atrazine exposure was a carcinogenicity study conducted in male and
female Sprague-Dawley rats at dietary dose levels of 0,10, 70, 500 or 1000 ppm
(0, 0.5, 3.5, 25 or 50 mg/kg/day). The Maximum Tolerated Dose (MTD) was
likely exceeded in this study at the 1000 ppm dose in females. Mortality was
significantly increased from 49% mortality at 104 weeks in the controls to 75%
mortality at 104 weeks in females of the 1000 ppm group (see Table 5-4).
Terminal body weight was also significantly decreased in 1000 ppm females in
this study. There was a 27.2% decrease in group mean body weight (p<0.01) in
the 1000 ppm females compared to controls. Male survival in the 1000 ppm
group was significantly increased at 1000 ppm compared to controls, but body
weight was significantly decreased compared to controls -18.7% less than
controls at 104 weeks. Based on the decreased body weight and increased
mortality seen in females, and the decreased body weight seen in males, 1000
ppm is deemed to exceed the MTD of Atrazine in this strain of rats. The
second-highest dose in this study, 500 ppm, likely is very close to the MTD.
Male body weight at this dose is reduced 8.2% compared to controls at 104
weeks while male survival is not significantly altered compared to controls.
Female survival is not significantly altered at this dose compared to controls, but
body weight is reduced by 18.9% (p<0.05) compared to controls at 104 weeks.
Based on the lack of significant effect in males seen at the 500 ppm dose and
the uncertain effect seen in females (lack of a significant increase in mortality
with a significant decrease in body weight) it seems likely that 500 ppm is very
close to the MTD for atrazine in this strain of rat.
The conclusions drawn about the MTD of atrazine are important given that
1000 ppm exceeds the MTD and 500 ppm is assumed to be very close to the
MTD. Thus, subsequent two-year carcinogenicity studies have used a dose of
400 ppm as the high dose to have the high dose be slightly below the MTD.
The mammary tumor incidences seen in this study are reported below in
Table 5-3 and mortality is shown in Table 5-4.
17
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-3. Mammary Tumor Incidence in the
Mayhew Study (as determined by US EPA, 1988)
Tumor Type
adenocarcinomas/
carcmosarcomas
combined
adenomas and
fibroadenomas
combined
Dose (mg/kg/day)
Control
15/88
17%
0.000**
20/88
23%
0446
0.5
16/67
24%
0.39
24/65
37%
0.110
3.5
27/69
39%
0.024*
21/69
30%
0.373
25
27/68
40%
0.019*
21/68
31%
0373
50
45/60
51%
0.000**
20/89
22%
0.468
NOTE: Significance for the trend is indicated at control. Significance of pairwise
comparison vs. controls is noted at dose group.
Incidence values are number of tumor bearing animals over number of animals at
risk
"1p< 0.05; ~p<0.01 as indicated by Peto Prevalence Test
Table 5-4. Mortality in the Mayhew Study (as determined by US EPA. 1988)
Mortality at
terminal
sacrifice
Dose (mg/kg/day)
Control
34/59
49%**
0.5
39/70
56%
3.5
40/70
57%
25
44/70
63%
50
52/69
75%**
NOTE: Significance for the trend is indicated at control. Significance of
pairwise comparison noted at dose group. Statistical test used are cox's or
Generalized Krushkal-Wallis.
"p<0.01
18
-------
DRAFT: DO NOT CITE OR QUOTE
A type of statistical analysis that examines tumor incidence and mortality
is the Peto Prevalence test. The results of this test are displayed in Table 5-3.
The results of this test showed that for the 70, 500 and 1000 ppm groups there
was a statistically-significant (SS) pairwise increase in incidence of mammary
adenocarcinomas and carcinosarcomas combined at 70, 500 and 1000 ppm,
and that there was a dose-related trend for these tumors that was also SS (p<
0.01). The study authors of the Mayhew report also conducted Cox-Tarone and
Gehan-Breslow tests to examine tumor incidence in light of the decrease
mortality in the females. The results from these tests were similar to the results
from the Peto test.
5.2 Thakur Studies
These are four studies - two using the SD strain and two using F-344
strain. These studies consisted of both terminal (all animals sacrificed after
two-years exposure) and serial sacrifice (10 animals per group sacrificed at
varying timepoints) protocols. The studies using the SD strain are discussed
below while the studies with the F-344 strain are discussed in section 5.6
Appendix Table 2 displays summaries of the study design for these studies.
5.2.1 Serial Sacrifice Protocol (Thakur, 1991 a)
Seventy SD female rats (no males were used) were exposed
through the diet to doses of atrazine (97%) at 0, 70 and 400 ppm (0, 4.23
and 26.23 mg/kg/day) for two years. Ten females per dose were
sacrificed at one, three, nine, 12,15,18 and 24 months.
Mortality was increased in a dose-dependent manner. There were
five unscheduled deaths in the control group, six in the 70 ppm group, and
eight at 400 ppm. Using the Gehan-Breslow test there was a statistically-
significant (SS) negative trend for survival (survival decreased as the dose
increased). Another statistical test - the Cox-Tarone test - did not indicate
a significant trend in either direction. A statistically-significant reduction in
body weights were found at several timepoints in the 400 ppm group
compared to controls.
19
-------
DRAFT: DO NOT CITE OR QUOTE
The results of this study is significant in regards to pituitary tumors.
Because pituitary tumors cannot be detected by palpation, a serial
sacrifice study is the most appropriate way to determine onset for pituitary
tumors. Pituitary tumor incidences by timepoint are displayed in Table
5-5, below. The hormonal basis for early onset of pituitary tumors is
discussed in section 9.2.6
Table 5-5. Pituitary (3-Adenoma Incidences by
Timepoint in Thakur Serial Sac, 1991 a
Sacrifice
time (mo.)
1
3
9
12
15
18
24
0-12
0-24
Control
0
0
0
2
5
9
6
2
22
4.23 mg/kg/day
0
0
0
2
3
5
6
2
16
26.63 mg/kg/day
0
0
2
6
4
6
2
8
20
NOTE: Ten animals in each group. Unscheduled sacrifice
animals are included in this table
20
-------
DRAFT: DO NOT CITE OR QUOTE
5.2.2 Terminal Sacrifice Protocol (Thakur, 1992a)
Sixty SD females (no males were used) were exposed through the
diet to doses of atrazine of 0, 70 and 400 ppm (0, 3.79 and 23.01
mg/kg/day) for two years (Appendix Table 2 summarizes the protocol
used for the Thakur studies with both SD and F-344 rats). Mortality was
high in the controls and increased in a dose-related manner. Mortality
was 48% in controls, 58% in the 70 ppm group, and 63% in the 400 ppm
group. Two survival tests were contradictory in determining whether or not
this increase in mortality was significant. Analysis with the
Gehan-Breslow test showed a negative trend in survival with dose
(increased mortality with increasing dose) while a Cox-Tarone test found
that the increases in mortality were not significant. Group mean body
weights were significantly decreased, compared to controls, at the 400
ppm group as early as four weeks and remained significantly decreased
up to, and including, week 76. Body weight gains were significantly
decreased for the period from study initiation to week 76. Both absolute
body weight at week 104 and body weight gain from week five to 104 of
the 400 ppm group, were lower than controls, but not
statistically-significantly. Group mean food consumption in the 400 ppm
group was decreased compared to controls for the first 13 weeks of the
study. After 13 weeks though, there was no significant difference. The
only finding at gross necropsy that may have been related to compound
exposure was an increase in enlarged spleens in the 400 ppm dose
group. The control and 70 ppm groups were observed to have five and
three animals with enlarged spleens, respectively, while the 400 ppm
group had 15. This finding was not observed in the Mayhew study, in the
Thakur terminal sacrifice study, or in any other bioassays that followed the
Thakur series of studies.
Histopathology revealed that mammary and pituitary neoplasms
were a common occurrence. Table 5-6 displays mammary tumor
incidence data by tumor type. There was not a statistically-significant
increase in fibroadenomas or carcinomas at the doses tested, compared
to controls. This is true whether or not mortality is taken into account
through Cox or Gehan-Breslow tests. Table 5-6 displays the p values
calculated by the study authors for mammary tumor incidences.
21
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-6. Female Mammary Gland Tumor Incidences in
the SD Terminal Sacrifice Protocol (Thakur, 1992a)
(calculated using Cox-Tarone and Gehan-Breslow tests)
Fibroadenoma
p value
Cox-Tarone
Gehan- Breslow
Carcinoma
p value
Cox-Tarone
Gehan- Breslow
Dose (mg/kg/day)
Control
39/60
(65%)
17/60
(28%)
3.79
30/59
(51%)
0.9141
0.6401
13/59
(22%)
0.8316
0.7613
23.01
41/60
(68.3%)
0.1070
0.2114
22/60
(33.6%)
0.1590
0.0810
5.3 Morseth, 1998
Atrazine (97.1%) was administered to 800 female Sprague-Dawley rats.
The rats were divided into two groups of 400 each. One group was
ovarectomized (OVX) while the other was left intact. Atrazine was mixed with
the diet at dose levels of 0 (control) 25, 50, 70 and 400 ppm (0,1.5, 3.1,4.2,
24.4 mg/kg/day for intact animals and 0,1.2, 2.5, 3.5, and 20.9 mg/kg/day for
OVX animals) for two years. There were 80 females at each dose level - 20 for a
12-month sacrifice and 60 for a 24-month sacrifice.
The trend for survival was statistically-significantly (SS) decreased in the
dosed groups compared to the controls. Survival was as follows: 43.3% in
controls; 31.7% - 25 ppm; 28.8% - 50 ppm; 31.6% - 70 ppm; 21.7% 400 ppm.
Body weight was SS reduced in the first half of the study in the 400 ppm group
(other groups were not significantly altered), but by the end of the study body
weights were similar to control values.
22
-------
DRAFT: DO NOT CITE OR QUOTE
Neoplastic histopathological findings were mostly limited to the pituitary
and the mammary gland. Neither intact nor OVX dosed animals showed an
increase in pituitary tumors compared to their respective controls, but intact
animals did show a 20-30% greater incidence of pituitary adenomas compared to
OVX animals.
There were few mammary tumors in the interim sacrifice animals, which is
not surprising given that these animals were sacrificed after only one-year.
Excluding the interim sacrifice and looking only at those animals that were
sacrificed at 24 months and those that died prematurely, there was an increase
in mammary tumor incidence at all intact dose groups compared to controls.
Looking at carcinomas alone incidence values are: 18.3%; 36.7%; 33.9%; 20%;
and 41.7% for the control, 25, 50, 70 and 400 ppm dose groups respectively.
Fibroadenomas alone were: 26.6%; 40%; 52.5%; 45%; and 40% for the control,
25, 50, 70 and 400 ppm dose groups respectively.
Table 5-7 below displays a statistical analysis of mammary tumor
incidence in intact animals using a Pete's Prevalence Test. Incidence values
shown below differ from those described in the paragraph above because interim
sacrifice animals are included in the analysis shown in Table 5-8 (USEPA,
1999c). A different survival-adjusted statistical analysis (Cox-Tarone) conducted
by the study author showed significant pairwise increases at 3.1 mg/kg/day for
fibroadenomas compared to control, but did not show significant pairwise
comparison to control for the 1.5 mg/kg/day group.
23
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-7. Mammary Gland Tumor Incidence
in Intact Animals in Morseth, 1998 Study
Fibroadenoma
Carcinoma
Adenoma
Dose (mg/kg/day)
Controls
16/78
(21%)
0.233
12/80
(15%)
0.002"
0/28
1.5
25/79
(32%)
0030'
18/80
(22%)
0.112
0/24
3.1
34/77
(44%)
0.000"
20/79
(25%)
0.067
1/20 (5%)
4.2
29/78
(37%)
0014"
14/80
(18%)
0.395
0/21
24.4
25177
(32%)
0.014'
27/80
(34%)
0.007"
0/15
NOTE: Significance for the trend is indicated at control. Significance of
pain/rise comparison noted at dose group. Incidence values are number of
tumor bearing animals over number of animals at risk.
p<0.05; ~p<0.01
Not a single mammary tumor of any sort was seen in any OVX animal.
The lack of mammary tumors in OVX animals provides evidence indicates that
an intact ovary is mandatory for mammary tumorogenesis in the SD female. The
results found in OVX animals will be discussed more fully in section 7.3.
Bi-weekly estrous cycle measurements were also made in this study. The
results of these measurements are discussed below under section 9.2 Estrous
Cycle.
24
-------
DRAFT: DO NOT CITE OR QUOTE
5.4 Pettersen and Turnier, 1995
This study exposed female SD rats (55 per group) through the diet to
doses of atrazine of 0,15, 30, 50, 70, or 400 ppm (0, 0.8,1.7, 2.8, 4.1, or 23.9
mg/kg/day). This study was a serial sacrifice protocol in which 10 animals in
each group were sacrificed at 3, six, and nine months and the remaining 25
animals were sacrificed at one year following initiation of dosing. Dosing
appeared to be adequate, as body weights in the last 10 months of the study
were reduced 8 to 12% in the 400 ppm animals compared to controls. Body
weight gains over the last 10 months of the study were reduced 11 to 18% in the
400 ppm group compared to controls. There were no differences in survival
among dose groups in this study. No mammary tumors were seen in the three-
and six-month sacrifices. There was a significant positive dose-related trend in
mammary tumors as well as a significant increase in mammary tumors between
control and 400 ppm animals using a pain/vise comparison. Table 5-8 displays
tumor incidences from this study.
25
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-8. Number Of Animals With Mammary
Tumors In The Pettersen and Turnier, 1995 Study
3 and 6
months
9 month
12 month
(Terminal
Sac)
Total
Dose(mg/kg/day)
0
0/20 had
tumors
F=1/10
C=1/25
F=1/25
A=0/25
C/A=1/55
F=2/55
0.8
0/20 had
tumors
C=11
no other
tumors
C=1/24
F=2/24
A=0/24
C/A=2/55
F=2/55
1.7
0/20 had
tumors
0/10 had
tumors
C=0/25
F=2/25
A=1/25
C/A=1/55
F=2/55
2.8
0/20 had
tumors
0/10 had
tumors
C=1/25
F=0/25
A=1/25
C/A=2/55
F=0/55
4.1
0/20 had
tumors
F=1/10
C=1/24
F=3/24
A=1/24
C/A=2/55
F=4/55
23.9
0/20 had
tumors
C=1/10
F=1/10
C=5/25
F=3/25
A=1/25
C/A=7/55
F=4/55
NOTE: Numerator is the number of animals with tumors, denominator is the number of
animals examined
C=adenocarcinoma; F= fibroadenoma; A = adenoma
'This tumor occurred in an animal scheduled to be sacrificed at 12 months but found dead
on study day 218.
5.5 Hazelette and Green, 1987
Atrazine (purity not given) was administered to CD-1 mice through the diet
to 59-60 animals/sex/dose, at dose levels of 0,10,300,1500 and 3000 ppm
(male/female mean daily dose 0/0, 1.4/1.6, 38.4/47.9, 194.0/246.9, 385.7/482.7
mg/kg/day)for 91 weeks. The doses given were adequate as indicated by toxic
effects, such as a decrease in mean body weight gain of both sexes
(23.5%/11%, M/F) and an increase in cardiac thrombi in the females, are seen at
both 1500 and 3000 ppm, while no dose-related toxic effects are seen at 10 and
300 ppm. There was also an increase in mortality (p < 0.05) in 3000 ppm
females, but not males, with only 25% of the females surviving versus 39-43% of
the females surviving in the other female dose groups. At the doses tested,
there was not a treatment-related increase in tumor incidence when compared to
controls.
26
-------
DRAFT: DO NOT CITE OR QUOTE
5.6 F-344 Two-Year Bioassays
5.6.1 Serial Sacrifice Protocol (Thakur, 1991b)
Seventy F-344 rats (females only) per dose were exposed ad
libitum to diet that had been mixed with atrazine (97.1%) to the
appropriate doses of 0 (negative control), 10, 70, 200 and 400 ppm (0,
0.68, 4.82,14.05, 34.33 mg/kg/day). Ten animals per dose group were
sacrificed after approximately one, three, nine, 12,15, and 18 months
exposure to the test article.
There was not an increase in mortality due to compound exposure,
and there was no increased incidence of clinical signs in dosed animals
compared to controls. The doses tested appeared to be sufficiently high
because there was a decreased absolute body weight and body weight
gain in the 400 ppm group compared to the controls. Group mean
absolute body weight in the 400 ppm group compared to controls was
also significantly decreased compared to controls at several time points
though the final mean body weight was not significantly decreased
compared to controls. The final group mean body weight for the 400 ppm
group was 6.6% less than the mean control value. During the course of
the study the 400 ppm animals gained an average of 116.7 gm compared
to the weight gain in the control group of 133.3 gm (14% less than
controls). This difference in body weight gain was statistically-significant
at a p value of 0.05. There was not an increase in mammary tumors or
any other type of tumor at any dose group.
27
-------
DRAFT: DO NOT CITE OR QUOTE
5.6.2 Terminal Sacrifice Protocol (Thakur, 1992b)
Sixty F-344 rats per sex per dose were fed technical grade atrazine
through the diet at doses of 0 (negative control), 10, 70, 200 and 400 ppm
(0, 0.49, 3.43, 9.87 and 20.17 mg/kg/day for males and 0, 0.61, 4.35,
12.71, and 26.18 mg/kg/day for females) for two years. Mortality in either
sex was not affected by treatment. Male control mortality was 30% while
male mortality in the 400 ppm group was 32%. Mortality in the other male
dose groups was slightly lower than controls, ranging from 22 to 25%.
Female mortality was 22% in the controls and 27% in the 400 ppm group.
Female mortality in the other dose groups ranged from 17 to 25%. Body
weights and body weight gains were adversely affected by compound
exposure, especially at the 400 ppm dose in each sex. Mean group body
weights were statistically-significantly reduced versus controls at the four,
13, 24, 52, 76 and 104 week timepoint in the both the male and female
400 ppm group. Percent body weight reductions ranged from 5.1 to 9.3%
in the males and 5.3 to 6.4% in the females. Percent body weight gains
were also significantly decreased in both sexes of the 400 ppm group for
all the time periods examined - 0-4, 0-13, 0-24, 0-52, 0-76 and 0-104
weeks. The range of percent reductions compared to controls was 11.3 to
15.9% in males and 10.7 to 17.4% in females. The reduction in percent
body weight gain, compared to controls, in males for the entire study
(weeks 0-104) was 11.3% and for females it was 11.6%. Mean group
food consumption was significantly decreased (4.8% versus controls) for
the 0-104 week period in 400 ppm males, but was not significantly
decreased in females. There were no findings at gross necropsy that
could be attributed to compound exposure and organ weights were not
altered between control and dosed animals.
28
-------
DRAFT: DO NOT CITE OR QUOTE
The incidence of mammary gland fibroadenomas was increased in
dosed females compared to controls. This increase was not statistically-
significant. Even at its highest dose level the percentage of animals with
fibroadenomas was below the historical level for the laboratory where the
Thakur studies were conducted. There were no increases in mammary
tumors of any type in dosed males versus controls. Two out of the 55
control males examined at histopathology were found to have a mammary
tumor (both were fibroadenomas). Only one out of 54 males in the 10
ppm group and one out of 58 in the 400 ppm group were found to have
mammary tumors (one fibroadenoma and one carcinoma) while not males
in the 70 and 200 ppm group had a mammary tumor of any type.
Therefore, dosing with atrazine did not increase mammary tumor
incidence in F-344 males.
Table 5-9. Female Mammary Tumor Incidence In F-344 Terminal
Sacrifice Protocol In The Thakur Terminal Sacrifice Study (1992b)
Fibroadenoma
Unadjusted
p value
Carcinoma
Unadjusted
p value
Dose (mg/kg/day)
Control
2/60
(3.3%)
0.2514
2/60
(3.3%)
0.4640
0.5
5/60
(8.3%)
0.2198
0/60
(0%)
0.2479
3.4
5/60
(8.3%)
0.2195
2/60
(3.3%)
0.0907
9.9
7/60
(11.7%)
0.0815
3/60
(5%)
0.5000
20.2
6/59
(10.2)
01295
2/59
(3.4%)
0.6843
Historical
Control
Mean = 14.9%
Range= 3-23%
Mean = 3.8%
Range=2-15%
NOTE: Historical control data from Hazelton Labs, 1984
29
-------
DRAFT: DO NOT CITE OR QUOTE
5.7 Pinter etal., 1990
The Pinter et at. study exposed Fischer-344 rats of both sexes to atrazine
(98.9%) that was mixed in the diet at concentrations of 0, 500 and 1000 ppm.
The control groups started with 56 males and 50 females; the 375/500 ppm
group started with 55 males and 53 females; and the 750/1000 ppm group
started with 53 males and 55 females. Unlike most carcinogenicity assays
where surviving animals are sacrificed after approximately 104 weeks on study,
in this study animals were allowed to live out their natural life span, except for
four males and six females that were sacrificed moribund. Table 5-10 displays
the mammary tumor incidence in males in Pinter et a/., 1990.
Table 5-10. Mammary Tumors In Males in Pinter et a/., 19901
Total number of
tumor-bearing males
Total number of
benign tumors
Total number of
malignant tumors
Control
1/48
1/48
0/48
375 ppm
1/51
1/51
0/51
750 ppm
8/53
9/53"
1/53
'Data from Table 2 in Pinter ef a/., 1990.
~p<0.01 using Fisher Exact test comparing high dose group to
low dose group.
5.8 Pinter etal., 1990
The Pinter et a/, study exposed Fischer-344 rats of both sexes to atrazine
(98.9%) that was mixed in the diet at concentrations of 0, 500 and 1000 ppm (the
500 and 1000 ppm doses were reduced to 375 and 500 ppm after eight weeks of
treatment due to toxicity). The control groups started with 56 males and 50
females; the 375/500 ppm group started with 55 males and 53 females; and the
750/1000 ppm group started with 53 males and 55 females. Unlike most
carcinogenicity assays where surviving animals are sacrificed after
approximately 104 weeks on study, in this study animals were allowed to live out
their natural life span, except for four males and six females that were sacrificed
moribund. Table 5-10 displays the mammary tumor incidence in males in Pinter
etal., 1990.
30
-------
DRAFT: DO NOT CITE OR QUOTE
Table 5-10. Mammary Tumors In Males in Pinter ef a/., 19901
Total number of
tumor-beanng males
Total number of
benign tumors
Total number of
malignant tumors
Control
1/48
1/48
0/48
375 ppm
1/51
1/51
0/51
750 ppm
8/53
9/53"
1/53
'Data from Table 2 in Pinter ef a/., 1990.
"p<0.01 using Fisher Exact test comparing high dose group to
low dose group.
Mammary gland tumors in the dosed females in this study were not
altered in incidence compared to controls. Mammary gland tumors in dosed
males were altered. The incidence of benign tumors (adenomas,
fibroadenomas, and fibromas) was one tumor in 48 animals in controls, 1/51 in
the 375 ppm group and 9/53 in the 750 ppm group. The study authors
performed a statistical analysis to determine if this increase in tumors was
significant. The authors found that when a pairwise comparison was done
between the 750 ppm group and the 375 ppm group there was a significant
increase at 750 ppm (p<0.01). Generally dose groups are compared to controls
to determine changes in tumor incidence following dosing. The study authors
choose not to perform a pairwise comparison to controls in this case, however,
because the animals in the control group died much sooner than the animals in
either dose group. The last control male died before study week 120; the last
375 ppm male died between week 120 and 130; and the last 750 ppm male died
between weeks 130 and 140. The differences in survival in this study confound
the results. The study authors state: 'The tumors in the high-dose group
appeared later in time than those in the control or low-dose group." The one
tumor in the control group occurred at week 111 and the sole tumor in the 375
ppm group occurred at 119 weeks. By contrast, the average mean time of tumor
appearance in the 750 ppm group was 121.3 week ± 15.4 weeks. The increase
in tumors seen in the older animals could be due to the exposure to atrazine or
could also be due to the simple fact that these were old animals.
31
-------
DRAFT: DO NOT CITE OR QUOTE
Mammary tumors in F-344 males have been shown to increase in
incidence in untreated males as they age (Solleveld, 1984). Table 5-11 displays
mammary tumor incidences in untreated, aged F-344 males. Table 5-11 shows
that a difference of only a few months can greatly increase the incidence of
mammary tumors as the rats age. Tumor incidences double as the rats age from
98-110 weeks to 111-123 weeks. Incidences double again as the rats age from
111-123 weeks to 124 to 136 weeks.
Table 5-11. Mammary Gland Fibroadenomas in Male
F-344 Rats by 12 Week Time Periods (Solleveld, 1984)
Mammary
Fibroadenoma
98-110
Weeks
3/77
(4%)
111-123
Weeks
13/143
(9%)
124-136
Weeks
27/148
(18%)
>137
Weeks
22/95
(23%)
The study authors seem to realize the possible relationship between the
age of the high-dose males and their high tumor incidences. The authors cite
Solleveld (1984) in an attempt to show that, even given the increased tumor
incidence in aged males, the atrazine exposed males had increased tumor
incidences. Pinter etat., (1990) notes:
'The incidence of benign mammary gland tumors in male
F344 rats was reported to be 2.2% for 110 to 16 weeks; in
the life span studies (more than 116 weeks), however,
13.4% of the male animals had benign mammary gland
tumors [Solleveld, 1984 is cited]. In our study, 16.9% of the
high-dose, males had benign mammary tumors."
The 2.2% the authors of the Pinter et al. study refer to in the above quote
is historical control data from untreated males in several National Toxicology
Program two-year bioassays. The 13.4% is historical control data for mammary
tumors in untreated males greater than 116 weeks in age in life-span studies.
The difference between 2.2% mammary tumor incidence at 116 weeks in the
two-year bioassays and the 13.4% in the life span studies again emphasizes the
dramatic increases in tumors that occur as the animals age beyond
approximately two years of age.
32
-------
DRAFT: DO NOT CITE OR QUOTE
The study authors seem to believe that atrazine exposure is inducing the
additional tumors between 13.4% and 16.9%. However, the study author's Table
2 on page 537 of their publication shows that only eight of 53 (15%, not 16.9%)
males had a mammary tumor of any sort. The origin of the 16.9% value is
unknown. The true difference - according to the data the study authors present
in their Table 2 - is between the 13.4% and 15%. Additionally, there was one
incidence of an adenocarcinoma in the high-dose males. The Pinter et a/.
publication does not include any description of which male had this carcinoma
but if this tumor occurred in an animal that did not also have a benign tumor then
the number of animals with benign tumors drops to seven out of 53 - 13.2% -
almost identical to the 13.4% cited by the study authors from the Solleveld
(1984) paper.
It is concluded that the authors of this study have not made a case that
the increase in male benign mammary tumors is due to atrazine exposure. The
tumors appearing in the high-dose males do not appear to be found at a rate any
higher than what would be expected for F-344 males of a comparable age.
5.9 Summary and Discussion of the Two-Year Bioassay Studies
Increased incidences of mammary fibroadenomas or carcinomas were
seen in three out of four separate two-year bioassay studies using
Sprague-Dawley rats.
Atrazine exposure in a two-year bioassay using CD-1 mice did not result
in increased incidences of mammary tumors in either sex, despite the compound
being given to the mice at doses that resulted in decreases in body weight gain
of 23.5% in females and 11% in males and a significant increase in mortality in
females.
33
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine exposure in two separate two-year bioassays using the F-344
strain of rats also did not result in increased incidences of mammary tumors.
One of the bioassays used rats of both sexes (terminal sacrifice protocol) and did
not see increases in mammary tumors in either sex, while the other study
employed only females (serial sacrifice protocol) without seeing an increased
incidence of mammary tumors. Dosing was adequate in both studies as
indicated by the 14% decrease in female body weight gain the 400 ppm dose
group compared to controls (serial sacrifice protocol) and the 11.3% () and
11.6% (?) reductions in percent body weight gain at 400 ppm compared to
controls in the terminal sacrifice study. Doses of atrazine that produced
equivalent, or even less than, reductions in body weight in the two-year
bioassays with the SD strain reductions) produced mammary tumors. For
example, in Thakur, 1992a (the terminal sacrifice protocol), female body weights
gains were reduced 12 to13% at 104 weeks in the 400 ppm group; in Mayhew et
a/. (1986) female body weight in the 70 ppm group were reduced only 3.25, yet
this was still sufficient to produce mammary tumors.
A study by Pinter et ai. (1990) also showed a lack of carcinogenic effect
for atrazine in F-344 females. This study did find an increase in benign
mammary tumors in F-344 males when high-dose males were compared to low-
dose males. However, this study was a life-span study rather than a two-year
bioassay and the males of the high-dose group (in which the increases in benign
mammary tumors were seen) survived significantly longer than males in the
controls or low-dose group. Examination of the tumor incidence in the high-dose
males from this study suggest that mammary tumor incidences were most likely
comparable to what would be expected for males of this strain and age.
34
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 6
Genotoxicity Studies
An important question to address in the hazard assessment is whether atrazine
also has the potential to be DMA reactive and act as an initiator within the context of the
multistage model of carcinogenesis. Thus, following initiation (i.e., a mammary gland
cell acquires a mutation that results in unregulated proliferation), tumor promotion (i.e.,
clonal expansion of the genetically altered cell) in the mammary tissue would then be
hormonally-mediated. The most desirable data to address this issue would be
information on genetic alterations in the relevant target tissue. This information is not
available for atrazine. There are a large number of studies using standard genotoxicity
assays to evaluate the mutagenic potential of atrazine.
Atrazine has been examined for its ability to induce mutations in microorganisms,
insect, and plants, and to induce chromosomal aberrations in vitro and in vivo in both
mammalian and nonmammalian organisms. Additionally, atrazine has been tested in
other assays using endpoints that are indicative of DNA damage, but are not measures
of mutation perse (e.g., genetic recombination, sister chromatid exchanges, DNA
strand breakage, and unscheduled DNA synthesis).
Although more than 50 studies are available on atrazine, some findings reported
in the literature are presented in insufficient detail for evaluation or results are
inconclusive due to study design problems. Furthermore, the results on atrazine are
inconsistent even within the same test system and genetic endpoint evaluated. Thus,
in evaluating the mutagenic potential it is important to take a weight- of- evidence
approach that considers the overall response patterns or trends for mutation,
chromosomal damage, and other indicators of DNA damage. In looking at the overall
trends in the database, it is important to consider the type of end point evaluated and
the test system/organism used. More emphasis is placed on end points that are direct
measures of mutation and chromosomal aberrations rather than indicators of DNA
damage (e.g., sister chromatid exchanges, DNA strand breaks). Also, results from
mammalian systems are emphasized more than results from assays using
nonmammalian organisms. Likewise, mammalian in vivo data are preferred over data
from in vitro tests.
35
-------
DRAFT: DO NOT CITE OR QUOTE
As summarized in Appendix Table 5 Most of the mutagenicity studies on atrazine
have been reported as negative. The majority of these negative results come from
mutation studies in bacteria. Beyond the bacteria results, the response profile for
atrazine is heterogenous and closer to a split between negatives and positives.
Nevertheless, the response patterns or trends for mutation and chromosomal damage
tend to be more convincing for the number and type of negative responses found after
atrazine treatment than for the positive data, which typically were weak, observed at
high treatment concentrations of atrazine, or were not repeatable. Data on several
metabolites of atrazine and its close structural analogues (propazine and simazine) do
not support a mutagenic potential for these compounds. Therefore, the totality of
evidence does not support a mutagenic potential for atrazine, and indicates; that a direct
DMA reactive/mutagenic mode of action is unlikely to be an influence of atrazine on
mammary gland tumor development (or at any other site). A discussion of the literature
supporting this conclusion follows.
6.1 Mutation Studies
As summarized in Appendix Table 5, most studies on atrazine for
mutation induction are bacteria tests with a few assays in yeast, fungi and in the
fruit fly Drosophila melanogaster. There is no compelling evidence for mutation
induction as a mode of carcinogenic action for atrazine given the consistent
negative responses in bacterial tests, and the inconsistent positive responses
across other phylogenetic lines (where responses tended to be weak, found at
high doses, and/or were not reproducible).
When atrazine was evaluated in the Ames assay (with a variety of
Salmonella typhimurium tester strains) by several different laboratories, it was
consistently negative even when a mammalian liver metabolic activation system
was incorporated (Seller, 1973; Poole and Simmon era/., 1977; Lusby eta!.,
1979; Bartsch era/., 1980; Sumneref a/., 1984; Deparde, 1986; Kappas, 1988;
Mersch-Sundermann era/., 1988; Zeiger ef a/., 1988 1992; Ruiz and Marzin,
1997).
There are no acceptable mutation studies in: Butler and Hoagland, 1989;
Anderson et a/., 1972; Morichetti etal. 1992, mammalian systems. Although
Adler (1980) reported a negative result for a gene mutation test (HPRT assay) in
V79 cells, this paper does not contain sufficient detail to allow an independent
assessment of the finding.
36
-------
DRAFT: DO NOT CITE OR QUOTE
Tests in yeast and fungi have yielded heterogenous results. For example,
mutation induction was reported in Schizosaccharomyces pombe with or without
plant cell activation (Mathias, 1987) and in Aspergillus nidulans only with
mammalian cell activation (Benigni et a/., 1979). When atrazine was evaluated
in Saccharomyces cerevisiae without exogenous activation, a negative result
was reported in one paper (Emnova et a/., 1987), while a weak positive finding
was observed by Morichetti et a/., 1992. The reported positives are mostly found
at high doses of atrazine. Furthermore, gene conversion and mitotic
recombination, which are indicators of DNA damage, were not increased in yeast
and fungi exposed to atrazine (de Bertoldi etal., 1980; Emnova et a/., 1987;
Kappas, 1988), except when plant cell activation was incorporated into the assay
(Plewa and Gentile, 1976). Because these lower eucaryotic assays have
intrinsic rates of positive responses that occur sporadically, this conflicting
database in fungi and yeast must be interpreted carefully in the context of the
weight-of- evidence and results from other organisms.
Microbial systems (Salmonella, E. coli, yeast) have been used as
indicators of mutational damage after atrazine treatment in host-mediated assays
(Adler, 1980; Simmon etal., 1977). The mouse host-mediated assays on
atrazine have yielded mix results, with Salmonella (injected intraperitoneally)
being negative and E. coli and yeast (injected into the mouse testes) as positive.
However, because of the variability in cell recovery, these assays are not viewed
as reliable indicators of mutagenicity.
Some information is available in insects. In general, positive results in
Drosophila were reported for somatic mutation in the spot wing test (Tripathy et
a/., 1993; Torres etal., 1992) or in the sex-linked recessive lethal assay (Tripathy
etal., 1993; Murnik and Nash, 1977) under certain conditions (larval feeding or
at high doses). Murnik and Nash (1977) tested atrazine, simazine and
cyanazine in the Drosophila sex linked recessive lethal assay. The authors
further concluded that, "these triazine herbicides may be weak mutagens," and
that "Much larger experiments are needed to determine with confidence the
mutagenic potential of the herbicides." The results reported by Murnik and Nash
(1977) were considered inconclusive by an expert Gene-Tox panel because of
the inadequate sample size used and possible variability confounding the
interpretation (Lee etal., 1983).
37
-------
DRAFT: DO NOT CITE OR QUOTE
6.2 Chromosome Aberration Studies
Several studies are available for the induction of chromosome aberrations
in mammal systems using both in vitro and in vivo assays. Although the in vitro
results have been conflicting, the majority of available in vivo data indicate that
atrazine is not clastogenic (chromosome breaking), particularly in the bone
marrow or in germ cells of the mouse. The few positive findings found in vitro
tests are likely the result of cellular toxicity or stress and not a direct DNA
mechanism of action. It should be noted that in vitro cytogenetic assays tend to
have a relatively high frequency of sporadic positive responses that are usually
associated with toxicity or other nonmutagenic events (e.g., high osmolality, low
pH)(Brusickefa/., 1998).
6.2.1 In Vitro Assays
Atrazine did not produce chromosomal aberrations in Chinese
hamster cells (Ishidate, 1988). A marginal increase in chromosome
aberrations (less than a doubling in the response over background, and
may be within the variation of background) was reported in human
peripheral blood lymphocytes up to 1.0 A
-------
DRAFT: DO NOT CITE OR QUOTE
The lack of a clear SCE response suggests that the Lioi et al.
positive findings for chromosome aberrations may reflect cellular toxicity
or stress. Lioi et al. state that their positive chromosomal aberration
response for atrazine "...indicated an induction of a pro-oxidant state of
the cells as an initial response to pesticide exposure." given the increase
found for glucose 6-phosphate dehydrogenase activity in exposed cells.
Positive results were reported using flow cytometry methods in Chinese
hamster ovary cells (Rayburn and Biradar, 1995.) This study is flawed
and considered inconclusive because the method of cell lysis and staining
of the nuclei used by the authors may have introduced artifacts. Flow
cytometry analysis (which essentially measures the distribution of DNA
between cells undergoing mitosis) is not as reliable as direct cell analysis
by microscopy for evaluating clastogenicity. Many factors can alter the
flow cytometry results, such as cleanliness of the machine, flow rate, cell
number, air bubbles, incomplete cytolysis, incomplete RNA'ase digestion
and sample preparation, thus making it different to interpret induced
genetic changes by a chemical versus induced artifacts due to study
conduct.
EPA NHEERL conducted an in vitro cytogenetic study on atrazine
(as well as simazine and cyanazine) to resolve the contradictory
cytogenetic findings reported in the literature on human peripheral blood
lymphocytes. No induction of chromosomal aberration or SCE's was
found in human peripheral blood lymphocytes after exposure to atrazine
up to a dose of 50 ^g/mL (Kligerman, et al., 2000a).
39
-------
DRAFT: DO NOT CITE OR QUOTE
6.2.2 In Vivo Assays
Atrazine administered in drinking water at 20 Atg/mL (20 ppm), was
found to be negative in a 30- and 90-day mouse ( B6C3F1 males and
females) study using metaphase analysis to evaluate the incidence of
chromosomal aberrations (Meisner et a/., 1992; Roloff eta/., 1992). It
should be noted that bone marrow cytogenetic evaluations with out
chromosome painting are insensitive for chronic studies. Atrazine tested
negative in a mouse (Tif:MAGf) bone marrow assay evaluating
micronuclei induction (Ceresa, 1988a). The study by Ceresa (1988a)
consisted of two parts. In the first phase, both sexes of mice were dosed
with a single gastric intubation of 2250 mg/kg atrazine in carboxymethyl
cellulose, with animal sacrifices at 16,24 or 48 hours following treatment.
In the second phase of the study, both sexes of mice were treated with a
single dose of atrazine at 562.5,1175 or 2250 mg/kg with bone marrow
cells harvested 24 hours post-treatment
More recently, Gebel etal. (1997) examined a variety of herbicides,
including atrazine in the mouse bone marrow micronuclei assay. In this
study, NMRI mice of both sexes were gavaged with several doses of
atrazine dissolved in corn oil up to 1750 mg/kg, and 48 hours later the
animals were sacrificed. The results from this study showed that atrazine
only induced a small increased in micronuclei in female mice at a dose of
1400 mg/kg that is approximately 80% of the LD5Q. It should be noted that
this is a very high dose of atrazine because at the next higher dose (1750
mg/kg), half the animals died (i.e., it was the LD50). The study authors
state that "(A)trazine... revealed significant aneugenic/clastogenic
activities in the micronucleus test in vivo in female NMRI mice. However,
these results only could be achieved in female animals at doses near to
the maximum tolerated dose. Thus, an in vivo genotoxic potential for...
Atrazine seems questionable." Although Adler (1980) cites a mouse bone
marrow cytogenetic study in which atrazine was given by oral gavage and
was found to be positive for clastogenetic effects at 2000 mg/kg (also a
very high dose), no further details were provided to evaluate the
acceptability of this finding. It should be noted that EPA's NHEERL found
both atrazine and simazine to be negative for micronuclei induction in
mice (Kligerman, etal., 2000b).
40
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine has been evaluated in dominant lethal assays for germ
cell chromosomal damage. In a negative study by Hertner (1993), male
mice were exposed via oral gavage up to 2400 mg/kg bw of atrazine.
Males were mated sequentially with untreated virgin females at different
days to allow evaluation of exposed male gametes at various germ cell
stages of development. There were no significant increases for
resorptions or dead fetuses at any dose. Although Adier (1980) cites a
dominant lethal study in the mouse in which atrazine given by oral gavage
caused an increase in dominant lethal mutations at 1500 and 2000 mg/kg,
the lack of details precludes an independent assessment of the study.
6.3 Other Indicators of DNA Damage or Mutagen Exposure
Atrazine has been negative for the induction of unscheduled DNA
synthesis (UDS) in rat hepatocyte cultures (Hertner, 1992; Puri and Muller,
1984). In the study by Hertner (1992), there was no evidence of UDS when
hepatocytes from adult male Tif:RAIf rats were exposed in vitro to atrazine at
several concentrations up to 1670 //g/mL for 16 to 18 hours (139 ^g/ml_ was a
precipitating concentration). In agreement, atrazine exposure did not induce
UDS in the study by Puri and Muller (1984), which also used primary rat
hepatocytes from adult male Tif:RAIf rats that were exposed to several
concentrations of atrazine for five hours up to 150 ug/mL, where precipitation of
the test article occurred.
Ribas et a/. (1995) used the single cell-gel electrophoresis assay (SCGE,
or the comet assay) to examine DNA strand breakage in human lymphocytes
treated in vitro with several concentrations up to 200 ug/mL of atrazine for four
hours both with and without S9 rat liver activation. In this study, atrazine was
found to cause a marginal increase in alkaline labile sites only in the absence of
mammalian liver S9 activation. The study authors refer to the atrazine results as
a "weak positive" and noted: "The extent of DNA migration showed that only in
cultures treated without S9 fraction there was a slight but significant increase and
this took place only when the concentrations were high (100 and 200 ug/mL)."
The weak positive findings by Ribas et a/, were similar to the weak effect
reported by another laboratory using the DNA alkaline elution assay to detect
DNA strand breakage in stomach, kidney, and liver of Sprague-Dawley female
rats treated with a single oral dose of 875 mg/kg of atrazine or with 350 mg/kg of
atrazine given five or 15 successive days (Pino et a/., 1988). Another study
41
-------
DRAFT: DO NOT CITE OR QUOTE
using an excision repair assay to evaluate atrazine for DNA damage in human
lymphocytes up to a dose of 100 ug/mL, reported negative results without
activation (Surralles et a/., 1995). Although a positive finding was reported in
another comet assay evaluating DNA damage in the tadpole, Rana catesbeiana
(Clements et at., 1997), it is uncertain whether this finding can be attributed to
atrazine because the study was conducted with commercial formulation (Aatrex)
that was of low atrazine purity (43%). It should be noted that Roundup was
reported as positive in this study. The active ingredient in Roundup, glyphosate,
is nonmutagenic when tested in standard genotoxicity assays (Flowers and Kier,
1978; Li, 1983; Shirasu et at., 1976). Atrazine was also found to negative in an
SOS chromotest (Ruiz and Marzin, 1997).
6.4 Mutagenicity Studies in Plants
Plant assays have yielded mix results. The induction of mutations but not
chromosome aberrations have been observed in Zea mays (Morgun et ai,
1982). Conflicting results are reported in Hordeum vulgare for both mutation and
chromosome aberrations (Wuu and Grant, 1966; Stroev, 1968; Muller et ai,
1972). The induction of chromosome aberrations is found when a high enough
dose is evaluated in Vicia faba (Wuu and Grant, 1967). Although the
metabolism in plants is qualitatively similar to that in mammalian, quantitative
differences may exist in certain plant systems that may allow for expression of
mutagenicity.
6.5 Metabolites of Atrazine
Metabolism of the triazine herbicides, atrazine, simazine and propazine, in
mammalian species results primarily in chloro-s-triazine metabolites (Simoneux,
1995). The major pathway for metabolism of the triazine herbicides in plants is
hydroxylation (Simoneux, 1995). The major plant metabolite is hydroxy-atrazine.
Several standard mutagenicity studies on hydroxy-atrazine and a variety of these
chloro-metabolites (Diaminochlortriazine, G-28279, and G-30033) have been
consistently negative for mutagenicity, and thus do not appear to exhibit a
mutagenic potential (summarized in Appendix Table 4).
Though not truly a metabolite of atrazine, the mutagenic potential of N-
nitrosoatrazine is also discussed.
42
-------
DRAFT: DO NOT CITE OR QUOTE
6.5.1 Diaminochlortriazine metabolite (DACT - 6-chloro-1,3,5-
triazine-2,4-diamine; didealkyl atrazine)
DACT (the structure is shown in Figure 8-1) was negative in the
Sa/mone//a/Ames assay when evaluated up to the limit concentration of
5000 ug per plate in tester strains TA 98, TA100, TA1535, and TA1537
with and without metabolic activation from the S9 fraction of Aroclor-
treated rats (Deparde and Karimi, 1987). In an UDS assay using isolated
human fibroblasts, OACT was also negative up to 600 ug/mL (which
exceeded the solubility limit of 400 ug/mL) (Meyer, 1987).
6.5.2 G-28279 metabolite - 6-chloro-N-ethyl-1,3,5-triazine-2,4-
diamine; deisopropyl atrazine
G-28279 (structure is shown in Figure 8-1) was tested up to a
concentration 5000 ug per plate and found to be negative in the
Salmonella/Ames assay when evaluated in tester strains TA 98, TA100,
TA1535, and TA1537 with and without metabolic activation from the S9
fraction of Aroclor-treated rats (Deparde, 1990). G-28279 was also
negative for inducing UDS in exposed hepatocytes from adult male
Tif:RAIf rats when tested up to a cytotoxic dose (800 ug/mL) (Gelnick,
1991 a). G-28279 was tested at the maximum tolerated dose of 480
mg/kg without inducing an increase in micronuclei in bone marrow cells of
exposed adult Tif:MAGf mice of both sexes (Ogorek, 1991 a).
6.5.3 G-30033 metabolite - 6-chloro-N-(1 -Methyl ethyl)-1,3,5-triazine-
2,4-diamine; deethyl atrazine
G-30033 (structure is shown in Figure 8-1) was tested up to a
concentration 5000 ug per plate and found to be negative in the
Sa/mone//a/Ames assay when evaluated in tester strains TA 98, TA100,
TA1535, and TA1537 with and without metabolic activation from the S9
fraction of Aroclor-treated rats (Deparde, 1989). G-30033 was also
negative for inducing UDS in exposed hepatocytes from adult male
Tif:RAIf rats when tested up to a cytotoxic dose (1000 jug/mL) (Gelnick,
1991 b). G-30033 was tested at the maximum tolerated dose of 480
mg/kg without inducing an increase in micronuclei in bone marrow cells of
exposed adult Tif:MAGf mice of both sexes (Ogorek, 1991b).
43
-------
DRAFT: DO NOT CITE OR QUOTE
6.5.4 Hydroxy-atrazine, G-34048
Hydroxy-atrazine (structure is shown in Figure 8-1) was negative
when evaluated at concentrations up to 5000 //g/plate in Salmonella tester
strains TA 98, TA100, TA1535 and TA1537 (Deparde, 1988). Tests were
conducted in the presence and absence of mammalian metabolic
activation S9 fraction of Tif:RAIf rats treated with Aroclor 1254. When
hepatocytes from adult male Tif:RAIf rats were exposed to hydroxy-
atrazine at concentrations up to 1500 ug/mL (precipitation seen at doses
2:12.5 ug/mL), no increase in UDS was found (Hertner, 1988). Hydroxy-
atrazine was negative in a UDS assay in which human fibroblast cells
were exposed in vitro to concentrations up to 1500 ug/mL under
nonactivating conditions only (Meyer, 1988). In vivo, hydroxy-atrazine
was tested up to the limit dose of 5000 mg/kg without inducing an
increase in micronuclei in bone marrow cells of exposed adult Tif:MAGf
mice (Ceresa, 1988c).
6.5.5 W-Nitrosoatrazine
A/-Nitrosoatrazine (NNAT) can be formed in vitro when atrazine and
nitrite are mixed at an acid pH (Wolfe, et a/., 1976). Because nitrites and
atrazine can be found together in drinking water, the hypothesis has been
advanced that NNAT can be formed in the acid pH found in the stomach.
The formation of NNAT in the stomach in vivo has yet to be demonstrated.
The genotoxicity of NNAT has been tested in the Ames assay, V-
79 mutation assay, newt micronucleus test, and a clastogenicity assay
using human lymphocytes.
Results of a modified Ames assay (available only as an abstract)
showed NNAT to cause an increase in revertants in the TA 100 and 98
strains with hamster S9 fraction at 525 ug/plate and 1 mg/plate,
respectively (Weisenberger ef a/., 1987). The study authors considered
these results to be "mildly mutagenic." This abstract also noted that
atrazine was tested and found to be nonmutagenic.
44
-------
DRAFT: DO NOT CITE OR QUOTE
A V79 assay (available only as an abstract) was considered by the
study investigators to produce results that indicated that NNAT was
"strongly mutagenic" (Weisenberger ef a/., 1988). This abstract also
noted that atrazine was tested and found to be nonmutagenic.
A micronucleus assay using peripheral red blood cells (RBC) from
newt larvae resulted in an increase micronuclei at doses of 7.5 and 15
ppm after a 12 day exposure while no increase in micronuclei was seen at
3.75 ppm (Haridon, 1993).
A clastogenicity assay in lymphocytes from normal human
volunteers found that NNAT at doses as low as 0.0001 ug/mL produced
significant elevations of chromosome break frequency and percent of cells
with aberrations. A significant elevation of the mitotic index was not
observed at 0.0001 ug/mL. Mitotic index was significantly increased at
the next highest dose of 0.001 ug/mL (Meisner, et a/., 1993).
6.6 Close Structural Analogues: Simazine and Propazine
As discussed in Chapter 8, atrazine is an s-triazine pesticide and is
closely related to simazine and propazine as 2-chloro-4,6-bis-(alkyamino)-s-
triazines. To further explore the mutagenicity of atrazine, the available
databases on propazine and simazine were also evaluated. The available •
studies are predominantly negative, and thus do not provide convincing evidence
of a mutagenic potential for simazine or propazine. Although there were some
positives reported they tended to be marginal, found at very high concentrations
or were not reproducible.
45
-------
DRAFT: DO NOT CITE OR QUOTE
6.6.1 Simazine
As summarized is in Appendix Table 6 simazine has been
evaluated in bacterial for mutation including various Salmonella tester
strains and found to be negative by several different laboratories even
when metabolic activation was incorporated into the assay (Mersch-
Sundermann et a/., 1988; Simmon et a/., 1977; Jones et a/., 1984; Seller,
1973; Lasinski, Kapeghian, and Green, 1987; Fahrig, 1974). Simazine
was negative for the induction of gene mutation and gene conversion in S.
cerevisiae (Fahrig, 1974; Siebert and Lemperle, 1974; Simmon ef a/.,
1977; Jones et a/., 1984; Emnova et a/., 1987). Like atrazine, conflicting
results were reported in plant and insect studies. Only one mammalian in
vitro gene mutation assay was found for simazine that reported a weak
positive in the presence of metabolic activation (Jones ef a/., 1984). This
study is considered inconclusive by an expert Gene-Tox panel (Mitchell ef
a/., 1997). Simazine is also negative at concentrations up to the solubility
limit in the primary rat hepatocyte UDS assay (Hertner, 1992).
Most in vitro cytogenetic assays on simazine are inconclusive due
to study design problems, but negative results have been reported for the
induction of SCE's and chromosomal aberrations. To resolve the
inconclusive cytogenetic literature, EPA NHEERL recently evaluated
simazine in human peripheral blood lymphocytes with in vitro exposures
up to 37.5 /^g/rnL (a dose at the limit of toxicity and solubility) and found
no induction of either chromosomal aberrations or SCE's (Kligemnan ef
a/., 2000b). In vivo, simazine was tested up to the limit dose of 5000
mg/kg without inducing an increase in micronuclei in the bone marrow of
exposed adult Tif:MAGf mice of both sexes (Ceresa, 19885). EPA's
NHEERL found simazine to be negative when evaluated by the mouse
micronucleus test (Kligerman, ef a/., 2000b).
46
-------
DRAFT: DO NOT CITE OR QUOTE
6.6.2 Propazine
The available genotoxicity studies on propazine are summarized in
Appendix Table 7. Propazine has been negative in bacterial mutagenicity
assays (Kappas, 1988; Shirasu, 1975). Propazine has also been
evaluated in vitro for the induction of gene mutation (at the HPRT locus) in
Chinese hamster lung cells (V79) under both activating and nonactivating
conditions (Ciba-Geigy, 1986). Studies without metabolic activation
exposed V79 cells for 21 hours up to 1000 ug/mL of atrazine. Studies
with metabolic activation exposed V79 cells for five hours up to 2000
ug/mL of atrazine along with the S9 fraction from Arochlor-treated male
rats. In the experiment without S9 activation, a weak positive response
(dose-related) was found at concentrations 400 ug/l. An equivocal
response that was not dose-related was seen in the experiment with S9
activation.
Propazine was negative for UDS when tested up to the solubility
limit in (62.5 ug/mL) exposed hepatocytes from adult, male TiF:RAIf rats
(Puri, 1984). Propazine was reported to be negative for chromosome
aberrations in Chinese hamster cells in vitro (Ishidate, etal., 1981;
Ishidate 1983),and negative for the induction of micronuclei when tested
up to 5000 mg/kg in adult female Chinese hamsters (Ciba-Geigy, 1984).
47
-------
DRAFT: DO NOT CITE OR QUOTE
6.7 Summary and Discussion of Mutagenicity Data
The genetic toxicology database for atrazine shows consistent negative
results for bacterial mutation assays. Beyond the bacterial tests, the database is
heterogeneous and contains conflicting test responses. Reported positive
responses tended to be weak and found at high doses of atrazine. No subset of
data points clearly establishes a direct DMA reactive mode of action for atrazine
associated with the carcinogenicity. Although the DMA damaging potential of
atrazine can not be entirely dismissed in Drosophila and plants, these finding
may be the result of species specific metabolism. Although some positive
findings were reported for clastogenicity in mammalian systems, these
responses were in conflict with other studies using the same assay approach
and may be associated with toxicity. Data on several metabolites of atrazine and
its close structural analogues (propazine and simazine) do not support a
mutagenic potential for these chemicals.
In summary, the totality of evidence does not support a mutagenic
potential for atrazine, and indicates that a direct DMA reactive/mutagenic mode
of action is unlikely to be a component of atrazine-induced mammary gland
neoplasia (or on tumor development at any other site).
N-nitrosoatrazine has been shown to be mutagenic in four different types
of mutagenicity assays. However, the chemical reaction which generates NNAT
has never been demonstrated to occur in vivo, and cancer bioassays in female
Swiss mice and female Wistar rats failed to show a carcinogenic response
following NNAT exposure (Weisenberger,1990 - abstract).
In conclusion, exposure to atrazine is not likely to pose a mutagenic
hazard to humans, especially at lower exposure levels experienced by humans.
48
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 7
7 Estrogenic Activity
Several studies employing both in vitro and in vivo assays are available
concerning the potential of atrazine to act as an estrogen agonist (i.e., mimic the effects
of estradiol exposure). Some of the assays also test the ability of atrazine to function
as a progesterone mimic. The results of these studies together provide evidence that
atrazine does not have direct estrogenic effects. These studies are discussed below
and their results are also summarized in Table 7-1.
7.1 In Vivo Assays
Tennant etal. conducted a study known as the uterotrophic response
assay. This is an accepted procedure for measuring estrogenic activity (Korach
and McLachlan, 1995). As an in vivo assay, it incorporates aspects such as
metabolism, serum binding and pharmacokinetics. Exposure to an estrogenic
agent increases the weight of the uterus by causing cellular proliferation of
endometrial cells, leading to an increase in thickness of the uterine endometrium.
Rats or mice are exposed to the suspected estrogenic agent for three or four
days, sacrificed, and their uterine weight is compared to that of the control
animals. Immature or OVX animals are used to decrease interference from
endogenous estrogens.
In this study, OVX Sprague-Dawley rats were treated for three days with
20,200 or 300 mg/kg of atrazine. An increase in uterine weights over controls
was not found.
Tennant etal. (1994a) also evaluated atrazine for uterine thymidine
incorporation. This test measures an increase in uterine cell proliferation.
Rather than measuring proliferation of uterine cells by weighing the uterus, this
test measures the incorporation of thymidine into cellular DNA prior to cell
division.
In this assay, female Sprague-Dawley rats were fed radiolabeled
thymidine in their diet and then exposed to 1,10, 50,100 or 300 mg/kg/day
atrazine for three days. The atrazine-treated animals had less thymidine
incorporation into uterine cells than did control rats.
49
-------
DRAFT: DO NOT CITE OR QUOTE
Table 7-1. In vitro and in vivo Hormonal Studies with Atrazine
Study Type
In vivo
Uterine bioassay in OVX
SD females1
In vivo
Progesterone receptor
binding assay in OVX
SD females '
In vitro
Uterine Thymidine
Incorporation assay1
In vitro
MCF-7 Cell Proliferation
Assay2
In vitro
Gel-shift assay for PR-
PRE complex2
Dose/Duration
Atrazine and DACT at 0. 20, 200 or 300
mg/kg for three days.
Simazme at 100 and 300 mg/kg/day for
three days
Atrazine at 0, 50 or 300 mg/kg for three
days followed by 5 #g/kg E2
Atrazine at 0. 1.0. 10, 50. 100 or
300 mg/kg for three days followed by 0 15
n9 of E2
Atrazine/simazine at 0 01 . 0 1, 1.0, 10 and
100 MM for 11 days
Atrazine/simazine at 1 \iM
Results
An increase in uterine weiaht indicates estroaenic activity
+ control - estradiol at 2 ug stimulated uterine weight gain;
Atrazine/DACT/Simazine - uterine weights in dose groups were similar to control weights
An increase in the abilitv of orooesterone receotor (PR) to bind its aaonist indicates
estroaenic activity
+ control - estradiol at 2 \ig increased PR binding of agonist;
Atrazine - 300 mg/kg/day decreased the ability of PR to bind a PR agonist;
DACT - 300 mg/kg/day decreased the ability of PR to bind a PR agonist;
Simazine - 300 mg/kg/day decreased the ability of PR to bind a PR agonist
An increase in uterine thymidine incorooration indicates estroaenic activitv
+ control - estradiol at 0 15 ug increased uterine thymidine incorporation;
Atrazine - 300 mg/kg/day decreased uterine thymidine incorporation;
DACT - 300 mg/kg/day decreased uterine thymidine incorporation;
Simazine - 300 mg/kg/day decreased uterine thymidine incorporation
An increase in the proliferation of MCF-7 cells indicates estroaenic activitv
+ control - 1 nM estradiol induced a two-fold increase in cell number;
Atrazine/simazine - no increase in cell numbers was seen at any dose of these
compounds
An increase in the retardation through an electroohoretic gel of a comolex consisting of PR
+ PR agonist + PR DNA binding region, indicates estroQenic activity
+ control - 1 nM estradiol resulted in a large increase in the retardation of the complex
through a gel;
Atrazine/simazine -Movement of the complex through the gel was not altered by these
compounds
50
-------
DRAFT: DO NOT CITE OR QUOTE
In vitro
Luciferase reporter gene
assays in MCF-7 cells2
Atrazine/simazine at 10'9.10'8. 10'7.10e, 10*
M
An increase in luciferase activity indicates estroaenic activity
+ control - estradiol at doses as low as 10'12 M resulted in increases in luciferase activity
Atrazine/simazine - no increase in luciferase activity was seen with either compound at
any dose
In vitro
Estrogen receptor
competitive binding
assay using uteri from
rats that had not
previously been
exposed to tnazines 3
Atrazine/simazine/DACT at 10'10,
10-", 10.-8.10'7 ID"6,10'5, ia4. and 10'3 for
equilibrium conditions
A variety of experiments were also run in
which conditions favored triazine binding to
the estrogen receptor. These were termed
disequilibrium conditions.
Displacement of estrogen from its receptor indicates estroaenic activity
No displacement of estradiol from its receptor was observed at any dose with any of the
triazine compounds under equilibrium conditions
Under conditions that favor triazine binding to the estrogen receptor (disequilibrium
conditions) triazine will displace estrogen from its receptor.
In vitro
Estrogen receptor
competitive binding
assay using uteri from
rats previously exposed
to triazines3
Atrazine/simazine/DACT at 50 and 300
mg/kg/day for two days to OVX SD females
Displacement of estrogen from its receptor indicates estroaenic activity
At 50 mg/kg/day displacement of estrogen from it receptor was not significantly altered
with any of the triazine compounds.
At 300 mg/kg/day there was a significant decrease in estrogen binding with all three
triazine compounds.
In vitro
Yeast assay using
transfected human
estrogen receptor4
Both maximal (1 .Ox 10 9 M. in this system)
and submaximal (2.5 x 10'9 M)
concentrations of atrazine, and the
chloratrazme metabolites alone and in
varying combinations were tested
An increase in 0-aalactosidase activity indicates binding of atrazine to the estrogen receptor
Nether atrazine alone, nor any of the many combinations of atrazine with the atrazine
metabolites caused an increase in P-galactosidase activity.
In vitro
Estrogen receptor
mediated growth in
yeast5
Atrazine and simazine at 10 uM
Growth of the veast indicates estroaenic activity
+ control - Yeast colonies exposed to estradiol at 1nM proliferated
Atrazine/simazine - Yeast colonies exposed to atrazine and simazine did not proliferate.
In vivo
Uterine peroxidase
assay using uteri from
female SD rats
Atrazine and simazine at 50, 150. and 300
mg/kg/day were exposed female SD rat
through gavage for three days
Uterine oeroxldase activity indicates estroaenicitv
+ control- estradiol at 10 ug/day results in a 10-fold increase in uterine peroxidase
activity.
Atrazine/simazine - neither compound at any dose increased uterine peroxidase activity.
'Tennant et a/.. 1994a.; 2Safe ef a/, 1995.;3Tennat et a/., 1994b.; "Graumann et a/.. 1999:5Conner et a/.. 1996.
51
-------
DRAFT: DO NOT CITE OR QUOTE
7.2 In Vitro Assays
Safe et at. (1995) used MCF-7 cells, a human breast cancer cell line that
proliferates best in the presence of estrogen, to evaluate the potential estrogenic
activity of atrazine. In this method, cells are grown in culture medium that has
been charcoal-filtered to ensure that it is free of estrogens. The assay is
conducted by adding the suspected estrogenic compound to the cell's medium
and measuring cell proliferation, usually by counting the cells.
MCF-7 cells exposed to 10 uM atrazine did not show increased
proliferation over control cells in this study.
Tennant et a/. (1994b) evaluated the ability of atrazine, simazine and
diaminochlorotriazine (DACT, a metabolite of both atrazine and simazine) to bind
to the rat uterine estrogen receptor. A uterine cytosol extract was prepared from
female Sprague-Dawley rats. These preparations are expected to be rich in
estrogen receptors. The uterine cytosol was incubated with radiolabeled
estrogen and one of the test chemicals. After an appropriate incubation time,
estrogen bound to its receptor was separated from unbound estrogen. High
levels of unbound estrogen would indicate that one of the test compounds was
competing with estrogen for binding to estrogen receptors.
Neither atrazine, simazine, nor DACT treatment was able to compete with
estrogen for binding to the estrogen receptor. No competitive binding was
apparent under conditions of equilibrium. Only when excessive amounts of
triazines were used (10,000-fold molar excess), was a slight competitive binding
observed. Atrazine, simazine and DACT were not considered, under the
conditions of this study, to effectively compete with estrogen for binding to the
estrogen receptor.
Tennant et a/. (1994b) also used an assay approach similar to the one
described above with the exception that the triazines were administered by oral
gavage, and uterine slices were incubated with radiolabeled estrogen instead of
cytosol extracts. The advantage of this method compared to the test without in
vivo preincubation is that exposure to the triazine compounds is done in vivo.
52
-------
DRAFT: DO NOT CITE OR QUOTE
In this study OVX female Sprague-Dawley rats were exposed to 50 or 300
mg/kg/day of either atrazine, simazine, or DACT for two days. The animals were
sacrificed, uterine slices were prepared, and incubated with radiolabeled
estrogen. Dosing at 300 mg/kg/day statistically reduced estrogen binding by
33% with atrazine, 39% with simazine and 24% with DACT. Dosing at 50
mg/kg/day reduced estrogen binding by 18% with atrazine, 21% with simazine
and 13% with DACT, but values were not statistically-significant. At high doses,
triazine compounds are able to bind to rat uterine estrogen receptors in vivo.
Conner et a/. (1996) used the reporter gene assays to evaluate atrazine.
In these experiments two constructs—a Gal4-HEGO chimeric receptor and a
GAL4-regulated promoter containing five ERE's and the luciferase gene-were
placed into the MCF-7 cell line. The luciferase gene is a reporter gene whose
product can be easily measured because it emits light when cells are exposed to
a compound that binds the estrogen receptor.
Treatment of these MCF-7 cells with estrogen in the nM range produced
large increases in luciferase activity compared to controls. However, treatment
with atrazine or simazine at doses as high as 10 uM failed to produce any
increases in luciferase activity compared to controls. These experiments show
that, in this system, atrazine and simazine fail to bind to the estrogen receptor.
Conner et a/. (1996) evaluated atrazine estrogen-dependent growth yeast
strain. The PL3 yeast strain requires for growth a medium supplemented with
the amino acids histidine and leucine, and the pyrimidine base uracil. If these
yeast cells were transformed so as to contain the human estrogen receptor, then
estrogen could take the place of uracil and allow growth of this strain on histidine
and leucine-containing medium. ER-positive PL3 cells in media supplemented
with histidine, leucine and estrogen will show growth after only one day of culture
and continue to proliferate for at least five days.
ER-positive PL3 cells supplemented with histidine, leucine and either
10 uM atrazine or 10 uM simazine show no proliferation even after five days of
culture. This study demonstrates that, in a yeast cell expressing human ER,
neither atrazine nor simazine have estrogenic effects.
53
-------
DRAFT: DO NOT CITE OR QUOTE
Bradlow etal. (1995) and Safe and McDougal (1998) examined the 16a to
2-hydroxyestrone metabolite ratios after atrazine treatment. Estradiol forms
many metabolites. It has been hypothesized by some investigators that the ratio
of the 16 metabolites to the 2-hydroxyestrone (2-OHE) metabolites may be a
factor in mammary carcinogenesis. It is believed that a high 16a to 2-OHE ratio
may be correlated with mammary carcinogenesis (Telang etal., 1997).
Nevertheless, others question whether this ratio is truly predictive.
Bradlow ef a/. (1995), exposed MCF-7 cells to several different chemicals
including atrazine. MCF-7 cells were placed in medium containing [16a-3H]
estradiol (to determine 16a metabolite levels) and [2-3H] estradiol (to determine
2-OHE metabolite levels) and atrazine at 1* 10"s M. Incubation was for 48 hours
following which measurements of the various metabolites were made. The 16a to
2-OHE ratio in the atrazine exposed cells was approximately seven to nine times
higher than the negative control and three to four times higher than ratios seen
with the known mammary carcinogen dimethyl benzanthracene (DMBA).
Safe and McDougal (1998) exposed MCF-7 cells to several different
chemicals including atrazine. Exposure to atrazine was at 1 xl O'5 M for 48 hours
followed by another 48 hours of atrazine plus [16cc-3H]estradiol or 48 hours of
atrazine plus radiolabeled [2-3H]estradiol. The ratio of 16ametabolites to 2-OHE
metabolites was not increased following exposure to atrazine (in fact it was
somewhat decreased compared to negative controls).
The results with atrazine in this assay have been contradictory; one group
of investigators (Bradlow et a/., 1995) reported an increased ratio, supportive of
potential carcinogenicity, while another group (Safe and McDougal, 1998) could
not confirm these results.
Tran etal. (1996) used a yeast strain transfected with the human estrogen
receptor to evaluate atrazine. A yeast cell line was transfected with the human
estrogen receptor linked to the lacZ gene. This system is similar to that used by
Conner ef at. described above; the major exceptions were that the cells are yeast
rather than MCF-7, and the reporter system is B-galactosidase rather than
luciferase. To examine the role of different domains of the estrogen receptor,
these investigators also transfected a gene encoding for the human estrogen
receptor minus the first 179 amino acids.
54
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine, cyanazine, simazine and the DACT metabolite were used in
concentrations ranging from 207 nM to 2075 nM for atrazine, simazine and
DACT, and up to 10,000 nM for cyanazine. All exposures included estradiol at
either 20 or 0.5 nM (referred to the by the study authors as maximal and
sub-maximal concentrations, respectively). With both the full and the truncated
receptor, no concentration of triazine, in combination with either 0.5 or 20 nM or
estradiol, resulted in P-galactosidase activity greater than what was seen with
estradiol alone.
The experiments done with the complete estrogen receptor plus 0.5 nM
estradiol (but not 20 nM) showed that all four triazines resulted in less
3-galactosidase activity than with estradiol alone. Studies with the truncated
receptor showed that at all triazine doses with both 0.5 and 20 nM estradiol,
3-galactosidase activity was not altered compared to estradiol alone. Under none
of the conditions tested did any of the triazines have estrogenic activity.
Anti-estrogenic activity was seen, but only with the submaximal concentration
and not the maximal concentration of estradiol. Since, the anti-estrogenic
activity was not seen in the experiments done with the truncated receptor, it can
be assumed that the amino-terminus of the estrogen receptor is responsible for
the anti-estrogenic activity.
Experiments by Graumann et a/. (1999) used also employed yeast that
had been transfected with DNA coding for the human estrogen receptor linked to
a P-galactosidase expression system. Only the full receptor was used.
Yeast were exposed to atrazine or atrazine plus the atrazine metabolites
Desethyatrazine (DE; 6-chloro-N-ethyl-1,3,5-triazine-2,4-diamine) and
Desisopropylatrazine (Dl; 6-chloro-N-(1 -Methyl ethyl)-1,3,5-triazine-2,4-diamine)
over a variety of concentrations and combinations. Atrazine alone, atrazine plus
DE, atrazine plus Dl and all three together were used. Like the Iran et a/, study,
both maximal (1.0x 10~9 M, in this system) and submaximal (2.5 x 10'9 M)
estradiol concentrations, along with varying concentrations and combination of
triazines, were used.
In agreement with the data of Iran et al. (1996), no concentration or
combination of triazines at either maximal or submaximal conditions resulted in
an increase in p-galactosidase activity. In contrast to the Iran paper, no
concentration or combination of triazines at either maximal or submaximal
conditions resulted in a decrease in p-galactosidase activity either.
55
-------
DRAFT: DO NOT CITE OR QUOTE
The Tran and Graumann papers, using similar systems, found that none
of the triazine compounds tested possessed intrinsic estrogenic activity. The
results in regards to anti-estrogenic activity are contradictory with one study
(Tran) finding anti-estrogenic activity and one study (Graumann et a/., 1999) not
finding such activity.
A study has been published which examines the upregulation of
aromatase activity by atrazine. Aromatase - also known as cytochrome P450 19
(CYP19) - is the enzyme responsible for the conversion of androgens to
estrogens (specifically, the catalysis of androstenedione to estrone and
testostrone to estradiol). An upregulation of CYP19 activity would be expected
to result in increased serum estrogen levels as an increased amount of
androgens are converted to estrogens. An increase in CYP19 activity has been
observed following atrazine treatment in a human adrenocortical carcinoma cell
line - H295R (Sanderson ef a/., 2000).
7.3 Special Carcinogenicity Bioassay Study (Morseth, 1998)
As discussed in Chapter 5, in a two-year oncogenicity study (Morseth,
1998), groups of estradiol-implanted OVX and intact female SD rats were
exposed to 97.1% atrazine at doses of 0, 25, 50, 70 and 400 ppm. The primary
purpose of this study was to examine the effect of ovariectomy on mammary
tumor development in the SD rat following atrazine exposure. This study also
provides information concerning the potential estrogenic activity of atrazine in
relation to mammary carcinogenesis.
If atrazine induces mammary tumors by acting directly on mammary
epithelium, then OVX animals might be expected to develop mammary tumors
as do intact animals. Removal of the ovaries of an animal should not have
affected the ability of atrazine to affect mammary epithelial cells and commence
carcinogenesis. However, the difference in mammary tumor incidence between
OVX and intact animals was striking. Not a single OVX animal in any dose
group developed a mammary tumor of any type. In contrast, among intact
animals, 38.3% of controls and 68.3% of the 400 ppm group developed
mammary neoplasia.
56
-------
DRAFT: DO NOT CITE OR QUOTE
Overall, the results from this study, showing a complete lack of mammary
tumors in OVX animals, provides further evidence that atrazine does not induce
mammary tumors by binding to and activating mammary tissue estrogen
receptors. This evidence is also consistent with the conclusion that mutagenicity
is not a key component of atrazine's carcinogenicity.
7.4 Noncancer Effects Relevant to Estrogenic Activity
A variety of bioassays using atrazine have been submitted to the Agency.
Many of these assays measure parameters that might be expected to be altered
were atrazine acting as an estrogenic agent. This section examines subchronic
dog and rat studies; chronic dog studies; multi-generation reproduction studies in
the rat; and developmental toxicity studies in the rat and rabbit, for alterations in
parameters that may be indicative of a possible estrogenic effect.
Alterations in the chronic dog and subchronic rat and dog studies that
were considered to possibly indicate an estrogenic effect were: changes in
testes or prostate weights in males; changes in ovarian or uterine weight in
females; histopathology findings in the testes (including seminiferous tubules) or
prostate of males; histopathology findings in the ovaries, uterus, vagina or
mammary gland of females.
Data in the multigeneration studies that were examined included: parental
testes, ovary and uterine weights; parental histopathology findings in the testes
or prostate of males; histopathology findings in the ovaries, uterus, vagina or
mammary gland of female parents; malformations or variations of the testes and
prostate of male offspring and the ovaries, uterus, vagina or mammary gland of
female offspring; and, the ability of male offspring to impregnate females and the
ability of females to become pregnant and deliver healthy pups.
In the developmental toxicity studies, data from visceral examinations
were examined for malformations or variations of the gonads in offspring. In
addition, the ratio of male to female offspring was examined. Maternal organ
weights and histopathology were not determined in these studies (which were
conducted under the Subdivision F guidelines). Thus, maternal organ weight
and histopathology data are not available to analyze for potential estrogenic
effects.
57
-------
DRAFT: DO NOT CITE OR QUOTE
In addition to examining data from the above mentioned atrazine
bioassays, data from bioassays with simazine and the atrazine/simazine
metabolites DACT, G-28279, and G-30033 were also scrutinized for potential
estrogenic effects.
7.4.1 Subchronic Dog Studies
Subchronic studies with atrazine and propazine in the dog are not
available, but subchrcnic dog studies using simazine, G-28279 and
G-30033 are available. Prostate and uterine weights were determined in
the G-28279 and G-30033 studies, but not in the simazine or DACT
studies.
Simazine. Testes weights in male Beagle dogs were decreased
(46% decrease in absolute testes weight and 27% decrease in
testes weight relative to body weight, compared to controls) in male
Beagle dogs exposed to the high dose of 133.6 mg/kg/day of
simazine. Ovary weights in females were not significantly altered
at any dose including the high dose tested of 136 mg/kg/day. No
increases in histopathology findings in either the male testes or
prostate or the female ovaries, uterus, vagina or mammary gland
were observed in dose groups compared to controls (Tai et a/.,
1985a)
58
-------
DRAFT: DO NOT CITE OR QUOTE
G-28279. Testes weights in male Beagle dogs exposed to the two
highest doses tested in this study of 18.9 or 33.4 mg/kg/day were
decreased. Compared to controls absolute testes weight
decreased 31.4% at 33.4 mg/kg/day and testes weight relative to
brain weight was decreased 36.7%. The decreases compared to
controls in the 18.9 mg/kg/day group were 22.85 and 25.3% for
absolute testes weight and testes weight relative to brain weight.
Prostate weight in these two dose groups were also reduced
compared to controls. The percent reductions at the high dose
were 60.2% and 62.6% for absolute weight and prostate weight
relative to brain weight. The reductions in the 18.9 mg/kg/day
group were 51.55 and 51.9% for absolute weight and for prostate
weight relative to brain weight. Ovary or uterus weights in the
females of any dose group were not altered compared to controls.
No increases in histopathology findings in either the male testes or
prostate or the female ovaries, uterus, vagina or mammary gland
were observed in dose groups compared to controls (Thompson et
a/., 1992).
G-3Q033. Male testes and prostate weights were not significantly
altered in Beagle dogs exposed to a high dose 28.85 mg/kg/day.
Females ovary weights were not altered in females exposed to a
high dose of 32.18 mg/kg/day. Both absolute uterine weights and
uterine weight relative to brain weight were significantly decreased
in the high-dose females. However, these changes were deemed
by the study reviewer to likely be secondary to body weight loss
rather than a direct effect of G-30033 exposure. No increases in
histopathology findings in either the male testes or prostate or the
female ovaries, uterus, vagina or mammary gland were observed in
dose groups compared to controls (Rudzki et a/., 1992).
59
-------
DRAFT: DO NOT CITE OR QUOTE
PACT. Male and female Beagle dogs used in this study showed
no significant alteration in testes or ovary weight up to the high
doses of 24.1 and 32.7 mg/kg/day for males and females
respectively. There was no increase in histopathology findings in
the female ovary, uterus, vagina or mammary gland in dose groups
compared to controls. There was, however, an increased
incidence of hypospermia and hypospermatogenesis in all four of
the 24.1 mg/kg/day males, while no animal in any of the other
group, including the controls, displayed these findings (Thompson
era/., 1990).
7.4.2 Subchronic Rat Studies
Subchronic studies with atrazine and propazine in the rat are not
available, but Subchronic rat studies using simazine, G-28279 and
G-30033 are available.
Simazine. Absolute testes weights were decreased (15.4%) in
male SD rats exposed to the high dose of 276 mg/kg/day of
atrazine while testes weights relative to body weight were not
altered compared to controls. Absolute ovary weights in the
females exposed to either 142 or 276 mg/kg/day were reduced by
20% and 40%, respectively, while relative ovary weights were not
altered compared to controls. No increases in histopathology
findings in either the males testes or prostate or the female ovaries,
uterus, vagina or mammary gland were observed in any dose
groups compared to controls (Tai et a/., 1985b).
60
-------
DRAFT: DO NOT CITE OR QUOTE
G-28279. Male SD rats showed increased testes weights relative
to body weight at the high dose tested (34.9 mg/kg/day) in this
study. Relative testes weights were increased 139% compared to
controls, but absolute testes weights at this dose were within 5% of
control values. Ovary weights were not significantly altered at any
dose. No increases in histopathology findings in either the testes
or prostate or the ovaries, uterus, vagina or mammary gland were
observed in any dose groups compared to controls (Schneider,
1992).
G-30033. Testes and ovary weights of the RAIf rats used in this
study were not significantly altered even at the high does of 35.1
for males and 38 mg/kg/day for females. No increases in
histopathology findings in either the males testes or prostate or the
female ovaries, uterus, vagina or mammary gland were observed in
dose groups compared to controls (Gerspach, 1991).
PACT. Sprague-Dawley rats were exposed to doses of DACT of
up to 34.1 mg/kg/day for males and 40.2 mg/kg/day for females.
Testes weights relative to body weight were increased 22% at 34.1
mg/kg/day and absolute testes weights were increased 6.6%
compared to controls. No significant alterations in ovary weights
were noted at any dose. No increases in histopathology findings in
either the male testes or prostate or the female ovaries, uterus,
vagina or mammary gland were observed (Pettersen et a/., 1991).
7.4.3 Chronic Dog Studies
Chronic (12-month) dog studies with atrazine DACT, and simazine
are available. Chronic dog studies with propazine, G-28279 and G-30033
are not available.
Atrazine. The Beagle dogs used in this study showed no
significant alteration in testes or ovary weights up to the high dose
used of 33.65 and 33.8 mg/kg/day for males and females,
respectively. There were no histopathology findings in the testes,
prostate, ovaries, uterus, vagina or mammary gland (O'Connor et
a/., 1987).
61
-------
DRAFT: DO NOT CITE OR QUOTE
Simazine. Doses of up to approximately 43 mg/kg/day did not alter
either absolute or relative testes or ovary weights in beagle dogs
exposed to simazine for one year. No increases in histopathology
findings in either the male testes or prostate or the female ovaries,
uterus, vagina or mammary gland were observed in dose groups
compared to controls.
PACT. The Beagle dogs used in this study showed no significant
alteration in testes or ovary weights up to the high dose used of
24.1 and 32.7 mg/kg/day for males and females, respectively.
There was no increased incidence of histopathology findings in the
female ovary, uterus, vagina or mammary gland in dose groups
compared to controls. There was an increased incidence of
hypospermia and hypospermatogenesis in two of the four 24.1
mg/kg/day males, but not in any other group, including the controls
(Thompson etal., 1990).
7.4.4 Multi-Generation Reproduction Studies
Atrazine. Sprague-Dawley rats were exposed to atrazine at
concentrations up to a high dose of 39 mg/kg/day for males and
42.8 mg/kg/day for females. Testes weight relative to body weight
was significantly increased in both parental generations at the high
dose (11.1% increase in the F0 generation and 11.1 % increase in
the F1 generation). Although slightly decreased compared to
controls, absolute testes weights were within 10% of control values.
Ovary weights were not significantly altered compared to controls.
There was no increased incidence of histopathology findings in the
testes, prostate, uterus, vagina or ovaries (histopathology was not
done on the mammary gland) at any dose in either generation of
parental animals, compared to controls. There were no necropsy
or histopathology findings in the male offspring of atrazine-treated
dams that would indicate excessive maternal exposure to
estrogens (Mainiero etal., 1987).
62
-------
DRAFT: DO NOT CITE OR QUOTE
Simazine. Sprague-Dawley rats were exposed to simazine at
concentrations up to a high dose of 28.89 mg/kg/day for males and
34.96 mg/kg/day for females. Testes weight relative to body weight
was significantly increased in both parental generations at the high
dose (11.5% increase in the F0 generation and 20% increase in the
F, generation). Absolute testes weights were within 10% of control
values for both parental generations at this dose. Ovary weights
relative to body weights were significantly increased in the F1
parental generation only at the high dose. Absolute ovary weights
in both parental generations at all doses, were not significantly
altered compared to controls. There was not an increased
incidence of histopathology findings in the testes, prostate, uterus,
vagina or ovaries (histopathology was not done on the mammary
gland) at any dose in either generation of parental animals,
compared to controls. There were no necropsy or histopathology
findings in the male offspring of atrazine-treated dams that would
indicate excessive maternal exposure to estrogens (Epstein et a/.,
1991).
Prooazine. Sprague-Dawley rats were exposed to doses of up to
50 mg/kg/day of propazine in this three-generation study. F0
paternal testes weights relative to body weight, at the high dose
only, were significantly increased (15.6%) compared to controls.
Absolute testes weights at this dose (and every other dose) in this
generation were within 5% of control values. Testes weight relative
to body weight in the F, generation was not significantly altered
though it was increased 16.6% in the high-dose groups compared
to controls. Absolute testes weight in all dose groups was within
5% of control values. Testes weights relative to body weights were
significantly increased (22%) compared to controls in the high-dose
of the F2 generation parents. Absolute testes weights were within
5% of control values for all other dose groups tests in this
generation. Ovary weights in the first two generations were not
significantly different, either absolutely or relative to body weight,
than control ovary weights. Ovary weights in the F2 generation
parents were significantly decreased (27.6%) at the mid-dose of
5 mg/kg/day. The ovary weights at the other two doses in this
generation were higher than the ovary weights at 5 mg/kg/day and
were within 10% of control values. There was not an increased
63
-------
DRAFT: DO NOT CITE OR QUOTE
incidence of histopathology findings in the testes, prostate, uterus,
vagina or ovaries (histopathology was not done on the mammary
gland) at any dose in any generation of parental animals,
compared to controls. There were no necropsy or histopathology
findings in the male offspring of atrazine-treated dams that would
indicate excessive maternal exposure to estrogens (Jessup, 1979).
7.4.5 Rat Developmental Toxicity Studies
Atrazine. Pregnant Sprague-Dawley rats were exposed to up to
100 mg/kg/day of atrazine from days six to 15 of their pregnancies.
There were no findings, in the visceral examinations, of any
malformations or variations in the gonads at any dose. The sex
ratios in all the dose groups were similar to the control sex ratio
(Ginkis, 1991).
Simazine. Pregnant Sprague-Dawley rats were exposed to up to
600 mg/kg/day of simazine from days six to 15 of their
pregnancies. There were no findings, in the visceral examinations,
of any malformations or variations in the gonads at any dose. The
sex ratios in all the dose groups were similar to the control sex ratio
(Mainieroefa/., 1986).
Prooazine. Pregnant Sprague-Dawley rats were exposed to up to
600 mg/kg/day of propazine from days six to 15 of their
pregnancies. There were no findings, in the visceral examinations,
of any malformations or variations in the gonads at any dose. The
sex ratios in all the dose groups were similar to the control sex ratio
(Fritz, 1976).
PACT. Pregnant Sprague-Dawley rats were exposed to up to 150
mg/kg/day of DACT from days six to 15 of their pregnancies.
There were no findings, in the visceral examinations, of any
malformations or variations in the gonads at any dose. The sex
ratios in all the dose groups were similar to the control sex ratio
(Hummalefa/., 1989).
64
-------
DRAFT: DO NOT CITE OR QUOTE
G-30033. Pregnant Tif:RAI rats were exposed to up to 100
mg/kg/day of G-30033 from days six to 15 of their pregnancies.
There were no findings, in the visceral examinations, of any
malformations or variations in the gonads at any dose. The sex
ratios in all the dose groups were similar to the control sex ratio
(Gerspach, 1991).
G-28279. Pregnant Tif:RAI rats were exposed to up to 100
mg/kg/day of G-28279 from days six to 15 of their pregnancies.
There were no findings, in the visceral examinations, of any
malformations or variations in the gonads at any dose. The sex
ratios in all the dose groups were similar to the control sex ratio
(Marty, 1992).
7.4.6 Rabbit Developmental Toxicity Studies
Atrazine. Pregnant New Zealand White rabbits were exposed to up
to 75 mg/kg/day of atrazine from days seven to 19 of their
pregnancies. There were no findings, in the visceral examinations,
of any malformations or variations in the gonads at any dose. The
sex ratios in all the dose groups were similar to the control sex ratio
(Arthur, 1984a).
Simazine. Pregnant New Zealand White rabbits were exposed to
up to 200 mg/kg/day of simazine from days seven to 19 of their
pregnancies. There were no findings, in the visceral examinations,
of any malformations or variations in the gonads at any dose. The
sex ratios in all the dose groups were similar to the control sex ratio
(Arthur, 1984b).
Propazine. Pregnant New Zealand White rabbits were exposed to
up to 50 mg/kg/day of atrazine from days seven to 19 of their
pregnancies. One male offspring in the mid-dose group of 10
mg/kg/day was found to have absent testes. Otherwise, there were
no findings, in the visceral examinations, of any malformations or
variations in the gonads at any dose. The sex ratios in all the dose
groups were similar to the control sex ratio (Knapp, 1995).
65
-------
DRAFT: DO NOT CITE OR QUOTE
The bioassays described above provide little evidence of an estrogenic
effect. Dosing in utero with estrogenic agents has been associated with findings
such as cryptochidism, hypospadias and anorchism (Daston etal., 1997; Danzo,
1998). Developmental toxicity studies in both the rat and rabbit and
multigeneration reproduction studies in the rat, failed to show an increased
incidence of these, or any other, abnormalities of the gonads. Multigeneration
reproduction studies in the rat did not indicate that male offspring of atrazine-
treated dams had any difficulties impregnating females although decreased
fertility in males exposed to increased levels of estrogen or estrogen mimicking
compounds in utero, has been described (Daston etal., 1997; Danzo, 1998).
Developmental toxicity studies in both the rat and rabbit and multigeneration
reproduction studies in the rat, failed to show any alterations in sex ratios
compared to controls. Ample evidence is available that estrogen exposure can
result in germ cell depletion in the seminiferous tubules within a few weeks of
commencement of estrogen administration (Bianco-Rodriguez and
Martinez-Garcia, 1996).
Subchronic rat and dog studies and chronic dog studies failed to show
any histopathology alterations in the seminiferous tubules or in any part of the
testes, or in the prostate. Excessive exposure to estrogen induces proliferation
of the uterine endometrium. Subchronic rat and dog studies and chronic dog
studies failed to show any histopathology alterations in the uterus, and, in the
studies where uterine weights were determined, no alteration in uterine weights
in response to triazine exposure were noted either. Excessive estrogen
exposure might also be expected to result in alterations in the vagina - such as
vaginal hyperplasia and vaginal wall thickening, or changes in the ovaries - such
as an increase in ovarian stromal cells or enlarged ovaries. Subchronic rat and
dog studies and chronic dog studies failed to show any histopathology
alterations in the vagina or ovaries following exposure to the triazines. Ovarian
weights were also not affected by triazine exposure.
The only potential endocrine alteration that is consistently seen in these
bioassays is an alteration in either relative or absolute testes weight. This
alteration is, however, somewhat variable. In seven rat studies and one dog
study testes weights were increased; in two rat studies weight were decreased;
in three dog studies and one rat study testes weights were not altered. Thus, in
a total of 14 studies, testes weight was increased in eight studies, decreased in
two, and not altered in four. Excessive estrogen stimulation would be expected
to decrease testes weight.
66
-------
DRAFT: DO NOT CITE OR QUOTE
The studies described above do not indicate an estrogenic effect of
atrazine, simazine, propazine, DACT, G-30033, or G-28279. All the parameters
examined for possible estrogenic effects were unaltered save one- testes weight.
The testes weight alterations were variable, and the majority of the time they
were altered, they were increased when one would expect that they would be
decreased.
7.5 Overall Conclusions of Estrogenic Activity Data
Multiple studies that examine the estrogenic activity of atrazine, simazine
and DACT have been reviewed by the Agency. Most, but not all, studies
indicate that these chemicals do not possess estrogenic activity. Utertrophic
response assays, uterine thymidine incorporation assays, MCF-7 cell
proliferation assays, luciferase reporter gene assays in MCF-7 cells, in vivo
progesterone receptor binding assays, uterine peroxidase assays, and estrogen
receptor mediated growth in were all negative for estrogenic or progesterone
activity of the triazine herbicides.
The few that yielded positive results were performed under conditions that
favored an estrogenic effect of atrazine, or were performed at very high dose
levels relative to the doses that induce mammary tumors in SD females.
A study examining atrazine's effects on aromatase activity indicated an
upregulation of aromatase in vitro activity following atrazine treatment. The role
this finding may play in bringing about an increase in serum estrogen in vivo is
unclear as this study was conducted in a cell line and assays such as utertrophic
response assays, and uterine thymidine incorporation assays, conducted in vivo,
were negative.
An examination of the data from several bioassays (chronic, subchronic,
developmental, and reproductive studies) employing atrazine, simazine,
propazine and the major atrazine metabolites did not provide any evidence of an
estrogenic effect resulting from exposure to these compounds.
67
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 8
8 Structure Activity Relationship
Atrazine belongs to a class of compounds known as triazines in reference to the
triazine ring structure that they contain. Several compounds containing triazine rings
are presently registered for use in the U.S. as pesticides. Many of these pesticides (but
not all) will have nitrogen atoms attached to carbon two and four. Compounds with this
moiety are more appropriately referred to as "amino-s-triazines." These triazine-
containing chemicals can be divided into a several classes of compounds:
sulfonylurea-triazine compounds (that do not have the nitrogen at C2 and 4);
alkyamino, alkythio-triazines, alkoxy-triazines and chloro-triazines. Atrazine is a
chlorotriazine. The distinction between the chemical classes lies in the groups attached
to the R1 position of the triazine ring (C6). Chloro-triazines will have a chlorine atom at
the R1 position. The alkyamnio-triazines will have an alkyamnio moiety at R1 and the
alkoxy will have an hydroxyl moiety at R1. Alkythio compounds will have an alkythio
group at R1. Sulfonylurea compounds will have, at R1, a sulfonated urea moiety to
which another structure, frequently a benzene ring, is attached. The structures of
atrazine and several atrazine metabolites are shown in Figure 8-1. The structures of
the amino-s-triazine ring itself and of chlosulfuron, a representative example of a
sulfonylurea compound, are shown in Figures 8-2 and 8-3. The structures of the
chloro-triazine pesticides simazine and propazine are shown in Figure 8-4.
Two-year bioassay studies using female Sprague-Dawley rats have been
performed on several of the triazine compounds. As shown in Tables 8-1 and 8-2,
(Only those studies that used Sprague-Dawley or CD rats, were submitted to EPA as
guideline studies, and have undergone an EPA review are included in these tables)
bioassay results demonstrate that the triazine ring structure, in and of itself, is not
carcinogenic for mammary tumors in the female Sprague-Dawley rat (Spencer, 1991).
Not shown in Tables 8-1 and 8-2 are the results from five two-year bioassays using
sulfonyurea compounds. Four out of the five studies with sulfonyurea compounds were
negative for carcinogenicity. The fact that four out of five sulfonyurea and three out of
four alkyl amino, alkoxy, or alkvthio-triazine compounds failed to induce tumors,
including mammary tumors, in SD rats indicates that simply containing a triazine ring is
not sufficient to render a compound carcinogenic. Chlorine at the R1 position seems to
promote an increased carcinogenic potential. For example, all four of the chlorotriazine
compounds are able to induce mammary tumors in female Sprague-Dawley rats. The
68
-------
DRAFT: DO NOT CITE OR QUOTE
importance of a chlorine in the R1 position is further demonstrated by the lack of
carcinogenicity seen with hydroxyatrazine (2-hydroxy-4-ethyl amino-6-isopropyl amino-
s-triazine) (Chow and Hart, 1995). Hydroxyatrazine is a major plant metabolite of
atrazine that differs from atrazine only in that the chlorine at R1 is replaced by a
hydroxyl-group.
Table 8-1. Results of Two-Year Bioassays with Alkylamino,
alkoxy, and alkythio-triazine Compounds
Chemical
Cyromazine
R1 - NH2
Prometryn
R1- SCH3
Terbutryn
R1-SCH3
Prometon
R1-OCH3
Species/
Strain
Rat-
Sprague-Dawley
Rat-
Sprague-Dawley
Rat-
CD«BR
Rat-
Sprague-Dawley
Results
Negative for oncogenicity in
doses up to 3000 ppm
Negative for oncogenicity in
doses up to 1500 ppm
Positive for female mammary
tumors at 3000 ppm
Negative for oncogenicity in
doses up to 1000 ppm
Reference
Blair, ef a/., 1981
Chau, efa/.,1991
Ciba-Geigy, 1980
O'Conner, era/., 1988
Table 8-2. Results of Two-Year Bioassays with Chloro-triazine Compounds
Chemical
Atrazine
R1-CI
Simazine
R1-CI
Propazine
R1-CI
Cyanazine
R1-CI
Species/
Strain
Rat-
Sprague-Dawley
Rat-
Sprague-Dawley
Rat-
Sprague-Dawley
Rat-
Sprague-Dawley
Results
Positive for female mammary
tumors at 70 ppm
Positive for female mammary
tumors at 100 ppm
Positive for female mammary
tumors at 3 ppm
Positive for female mammary
tumors at 5 ppm
Reference
Mayhew, etal., 1986
McCormick, etal.,
1988
Jessup, 1980a
Bogdanffy, 1990
69
-------
DRAFT: DO NOT CITE OR QUOTE
The lack of carcinogenicrty of the hydroxyatrazine metabolite is further supported
by decisions reached by the HED Metabolism Committee (MARC) which concluded in a
September 29,1995 meeting that: "For atrazine, the residues of concern for cancer
dietary risk are parent and chloro metabolites" (US EPA, 1995).
Figure 8-1. Structures of Atrazine and Major Metabolites
Atrazine G-30033
Cl Cl
x,
CH3CH2N ^N/ NHCH(CH3)2 H2W N NHCH2
G-26279 G-28273
Hydroxyatrazine
OH
N
CH3CH2W N NCH(CH3)2
H H
70
-------
DRAFT: DO NOT CITE OR QUOTE
Figure 8-2. Structure of the Amino-s-Triazine Ring
R1
N
\
R3-N N_R2
Amino-s-Tnazine Ring
Figure 8-3. Structure of Chlorsulfuron
OCH3
Chlorsulfuron
71
-------
DRAFT: DO NOT CITE OR QUOTE
Figure 8-4. Structures of Simazine and Propazine
Propazine
Cl
HN
Of special interest when looking at structural analogues of atrazine is the
functional similarity between those compounds most similar to atrazine (simazine,
propazine and cyanazine) in regards to carcinogenicity. Like atrazine, all three of these
compounds are positive for mammary tumors in the female Sprague-Dawley rat (see
Table 8-2) and all three have been tested in two-year mouse bioassays and have been
found to be negative for carcinogenicity (Hazelette and Green, 1988; Jessup, 1980b;
Gellatly, 1981). Like atrazine, genotoxicity studies with simazine and propazine do not
support a mutagenic potential for these compounds (see Chapter 6 - Structural Analogs
of atrazine of this document). Mutagenicity studies with cyanazine have yielded mixed
results, although the presence of the cyano group in cyanazine confounds comparison
of this compound with the three other chloro-s-triazines relative to mutagenicity.
72
-------
DRAFT: DO NOT CITE OR QUOTE
Chapter 9
9 Hormonal and Estrus Cyclicity Studies
The four previously described two-year bioassays in the SD rat clearly
demonstrated that atrazine may increase mammary tumor incidence or decrease
latency in the SD female. Atrazine exposure may also decrease pituitary adenoma
latency in the SD female, but does not appear to alter pituitary tumor incidence.
Reviews of numerous mutagenicity studies indicates that DNA damage does not
appear to be contributing to these carcinogenic effects. Furthermore, results indicate
that atrazine does not appear to be acting as a xenoestrogen to induce mammary
tumors. The two-year bioassays with the CD-1 mouse and the F-344 rat demonstrated
that there are strain/species differences in the carcinogenic effects of atrazine.
The lack of a mutagenic or exogenous estrogenic effect of atrazine, combined
with the known hormonal dependence of rat mammary tumors, suggests that a
perturbation of an endogenous hormonal mechanism may be responsible for the
increase in mammary tumors and decreased latency of pituitary adenomas seen
following atrazine exposure in the SD rat.
Mammary tumors and pituitary adenomas in the SD female are both very
common occurrences. The historical control data shown in Table 9-1 demonstrates the
high background tumor incidence rate for these tumor types in the female SD rat. The
background tumor incidence rates of mammary tumors in SD males is <2% - much
lower than in females (Lang, 1992; McMartin etal., 1992). Pituitary adenomas, on the
other hand, are very common in the SD male. Spontaneous pituitary adenoma rates
are approximately 60% in SD males (Lang, 1992; McMartin etal., 1992).
73
-------
DRAFT: DO NOT CITE OR QUOTE
Table 9-1. Mammary Tumor and Pituitary Adenoma Historical
Control Incidence Data For Sprague-Dawley Females At 24 Months
Source
Pooled Charles River1 mammary
tissues from 1250 female rats in 19
studies examined
Pooled Ciba-Geigy, Summit, N.J.2
mammary tissues from 585 female
rats in nine studies examined
Mammary
Fibroadenoma
= 31.4%
Range= 13 7-49.0%
= 31.3%
Range= 20-43 3%
Mammary
Carcinoma
= 17.68%
Range= 7.1-
31.4%
= 16.8%
Range= 6.730%
Pituitary
Adenoma
= 72.1%
Range= 31 .4-88.8%
=84.7%
Range= 79.7-90%
'Lang, 1992
2McMartinefa/., 1992
It has been hypothesized that the effect of atrazine exposure is to decrease the
latency of mammary cancer in the SD female rat (Stevens, 1994).
"...it has been hypothesized that the lifetime administration of s-triazines to
female Sprague-Dawley rats produces an endocrine-mediated imbalance,
which causes precocious age-related changes, possibly resulting in the
earlier onset or increased incidence of mammary tumors."
Implicit in this hypothesis is the belief that the high mammary tumor incidences
seen in control SD females are due to the reproductive aging process in that strain. It is
reasonable to assume the same mode of action proposed for mammary tumors
(induction of an early onset of a state resembling reproductive aging and its associated
hormonal imbalance) is also responsible for the early onset of pituitary tumors.
Pituitary tumors in the SD female are also known to be age-related (Blankenstein
etal., 1984; McComb et a/., 1984; Sandusky era/., 1988). Pituitary tumors in the
female rodent are known to be, at least in part, estrogen-dependent, and increased
exposure to estrogens is associated with increased incidences of pituitary adenomas,
hyperplasia and increased pituitary weights (Blankenstein et a/., 1984; McConnell,
1989a; Nelson era/., 1980; Meites, 1981; McConnell, 1989b).
74
-------
DRAFT: DO NOT CITE OR QUOTE
9.1 Rat Reproductive Aging Process
Mammary tumors are recognized as a common and expected occurrence
in female SD rats, and appear to be increased primarily in aging animals. The
age-related nature of the mammary tumors in female SD rats has been
commented on by various authors (Cooper, 1983; Cults and Noble, 1964). The
age-related nature of the mammary tumors in SD females is illustrated by the
mammary tumor incidences in control animals in the two-year oncogenicity
studies described in this document. The great majority (75%) of the mammary
tumors in control rats in the Thakur study (1991 a) occurred when the rats were
greater than one-year old. The other Thakur study (1992a) also had 75% of the
mammary tumors in controls appearing after one year of age. A third two-year
oncogenicity study found that 82% of the mammary tumors in control animals
occurred when the animals were greater than one-year old (Morseth, 1998).
Understanding the hypothesis that the mechanism of reproductive aging
in the female Sprague-Dawley leads to a high incidence of mammary tumors in
females of that strain necessitates understanding the processes of reproductive
aging in SD rats that are believed to lead to increased incidences of mammary
tumors. At this point, a brief review and summary of these processes in the SD
rat are presented. For comparisons sake, the reproductive aging process in the
F-344 rat is also presented.
9.1.1 Sprague-Dawley
9.1.1.1 Alterations at the Ovary And Vagina
For reasons that are not yet completely understood, an
aging female SD rat experiences a dampening of the preovulatory
pituitary gonadotropin (luteinizing hormone, LH) surge (Zou, 1996;
Cooper and Walker, 1979; Huang et a/., 1978). The preovulatory
LH surge is responsible for inducing ovulation and, when the
amplitude of the surge falls below a critical threshold, there is
failure to ovulate. The ovaries of females with subthreshold LH
surges will have reduced numbers of, or no, functional corpea lutea
(CL), an increased number of secondary and antral follicles, and an
increased number of follicles undergoing atresia (Smith and Conn,
1983; Cooper et a/., 1996; Huang and Meites, 1975). The increase
in numbers of unruptured follicles results in prolonged exposure to
75
-------
DRAFT: DO NOT CITE OR QUOTE
moderately elevated levels of serum estrogens as these follicles
continue to secrete estrogen. Eventually, the unovulated follicles
do undergo the process of atresia. Thus, the ovaries of aging SD
rats will display increased numbers of atretic follicles. In the aging
SD rat, each successive wave of follicles that undergoes atresia is
replaced by a new crop that will again sustain the serum estradiol
levels at a moderately elevated level.
Corpea lutea form from follicles that have ovulated and
undergo the process of "lutenization" in which the granulosa cells
of the follicle begin to secrete progesterone instead of estradiol.
Because follicle ovulation is reduced in aging SD females, CL
numbers will be reduced.
The vaginal smears of aging SD rats will consist primarily of
cornified epithelial cells reflecting the tonic level of estradiol in the
serum and the absence of, or low levels of, progesterone. The
presence of vaginal cornification day after day as a result of the
aging process is termed "constant estrus."
These changes typically begin to take place in a normally
aging rat at approximately nine months of age. A typical estrous
cycle in a young SD rat (< nine months in age) is four or five days
in length. The first two days are diestrus, the third is proestrus and
the fourth is estrus. Five-day cycles typically have an extra day of
diestrus. During the aging process, cycles first become irregular
and then the majority of the females transition into constant estrus.
In the final months of the females life, she may become completely
acyclic. Vaginal smears performed on animals at this time will
indicate that the animals are in a state of extended or persistent
diestrus.
The above described pattern should be considered a
general rule only. The exact age at which these changes take
place and their exact order may vary with the individual rat or with
separate colonies of rats (LeFevere and McClintock, 1988).
76
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.1.2 Alterations at the Pituitary
Age-associated pituitary alterations have been
well-described in the SD female. Pituitary weight, pituitary
hyperplasia and pituitary adenomas have all been reported to be
increased in aged female rats that undergo constant estrus as their
primary mode of reproductive aging (SD and Long-Evans, for
example) (Console era/., 1997; McComb etal., 1984). A
proliferation of the cells of the anterior pituitary that secrete
prolactin (i.e., lactotrophs) has been shown to be responsible for
much of the increase in pituitary weight and hyperplasia seen in the
aged female SD. The great majority of the pituitary adenomas
seen in the aging female SD have also been found to have
originated from lactotrophs (Sandusky era/., 1988).
An increase in lacotroph number might be expected to result
in increased serum prolactin levels in the aging rat as these cells
continue to secrete prolactin. Increased serum prolactin levels are
seen in aged SD females, and these increased serum prolactin
levels have been correlated with increased pituitary weight,
increased pituitary hyperplasia, and increased incidence of pituitary
adenomas (Baird et a/., 1990; McComb et a/., 1986; van Putten et
a/., 1988).
The alterations seen at the pituitary are believed to be due
to the prolonged exposure to moderately elevated levels of serum
estrogens that occur following anovulation (Nelson etal., 1980;
Goya et a/., 1990; McConnell, 1989b). Serum estrogen is known to
be mitogenic to pituitary lactotrophs, and the prolonged exposure to
moderately elevated serum estradiol levels likely mediates the
increases in pituitary weight, hyperplasia and adenomas seen in
the aging SD female. Additionally, estrogen appears to damage
tuberoinfundibular neurons in the hypothalmus to inhibit production
of prolactin inhibiting factor (PIF) (Sarkar et al., 1982). PIF inhibits
the production by the pituitary of prolactin. Thus, inhibition of its
production by estrogen would have the effect of increasing
prolactin production by the pituitary.
77
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.2 Fischer-344
9.1.2.1 Alterations at the Ovary and Vagina
Like the young-adult Sprague-Dawley female, the
young-adult female F-344 typically displays regular four- to five-day
estrous cycles. However, unlike the SD female, the ability of the
aging F-344 rat to obtain an ovulatory LH surge is not
compromised. Also, unlike the SD female that develops a pattern
of constant estrus within the first year of life, the regular ovarian
cycles present in the F-344 give way to a pattern of repetitive
pseudopregnancies (Huang et a/., 1978; McConnell, 1989a; Estes,
1982). In this condition, ovulation occurs periodically and the newly
formed CL are maintained for extended periods (10 to16 days).
There are two distinct changes in the pattern of hormone secretion
that are present in the pseudopregnant female - increases in serum
prolactin and serum progesterone. First, there are two, daily
prolactin peaks that occur just before lights on and just before lights
out. In response to the elevated prolactin levels, the ovarian CL
are maintained. The functional CL are responsible for the elevated
levels of progesterone present in these animals. In addition to CL,
the ovaries of an aging F-344 will contain a moderate number of
antral and secondary follicles, and a moderate number of atretic
follicles. Again, these findings are in contrast to the SD female that
will contain few, if any CL, many secondary and antral follicles, and
many atretic follicles. Also, in contrast to the constant estrus
female that is characterized by uninterrupted cornification (typical
of the aging SD rat), the vaginal cytology of the pseudopregnant
female is primarily leukocytic. This vaginal smear pattern is similar
to that seen during the diestrus period of the estrous cycle and this
pattern may be referred to as "persistent diestrus," although the
term "repetitive pseudopregnancy" is also used, and is the more
descriptive term. As was the case with the SD, in the final months
of the F-344 females life, she may become completely acyclic. The
same disclaimer about variability in SD rat reproductive aging
process applies to the F-344. The description above can only be
seen as a generality.
78
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.2.2 Alterations at the Pituitary
The alterations seen in the pituitary of the female SD are not
seen in the aging F-344 rat. The incidence of pituitary adenomas
in female F-344 rats at two-years of age is much less than the
incidence of pituitary adenomas in female SD rats at two-years of
age (Sandusky et a/., 1988). Increases in pituitary weights and
pituitary hyperplasia are also not commonly seen in the aging
F-344 female.
9.1.3 Summary of the Reproductive Aging Process in SD and F-344
Rats
Characteristics of reproductive aging in the female SD have been
throughly studied (Cooper, 1983; Cooper ef a/., 1996; LeFevereand
McClintock, 1988; Meites era/., 1980). The appeal of hypothesizing that
these events may result in mammary carcinogenicity comes from the
prolonged exposure to estrogen and prolactin that results from these
events. As noted above - the increased days spent in estrus in the aging
SD female appears to be related to the attenuated LH surge, which results
in numerous unovulated ovarian follicles constantly producing estrogen.
The prolonged exposure to estrogen acts at the pituitary to stimulate
increased prolactin production.
The role of prolonged or inappropriate exposure to estrogen and
prolactin in mammary carcinogenicity in rats has been well-established
(Nandi era/., 1995; Russo era/., 1990; Cutts and Noble, 1964; Meites,
1972). It is reasonable to hypothesize that the prolonged exposure to
estrogen and prolactin that results from the normal reproductive aging
process in SD females may be responsible for the high levels of mammary
tumors normally seen in this strain. This hypothesis was first advanced as
long ago as 1966 (Durbin et a/., 1966).
79
-------
DRAFT: DO NOT CITE OR QUOTE
The mode of reproductive aging in the F-344 rat is different from
that of the SD female. Reproductive aging in the F-344 does not result in
prolonged estrogen exposure. Prolactin levels are increased in the aging
F-344, as are progesterone levels.
A summary of the different modes of reproductive aging in the SD,
and F-334 is shown in Figure 9-1.
80
-------
DRAFT: DO NOT CITE OR QUOTE
Figure 9-1. Summary of the Reproductive Aging Process in Different Rat Strains
Sprague-Dawley, Long Evans, Wlstar1
F-344
Normal cycle is a four to five days with 25% of the time
spent in estrus, 25% spent in proestrus and 50% spent in
diestrus;
Reproductive aging becomes evident at approximately 12
months;
Reproductive aging is characterized by increased prolactin
surges that leads to maintenance of the corpea lutea;
There is an increase in the days spent in diestrus, and
increased exposure to progesterone.
In very aged animals, acyclicity is common
» Normal cycle is a four to five day cycle with 25% of the time
spent in estrus, 25% spent in proestrus and 50% spent in
diestrus;
» Reproductive aging becomes evident at approximately nine
to 12 months,
» Reproductive aging is characterized by decreased
gonadotropin surges that leads to maintenance of primary,
secondary and antral ovarian follicles;
> An irregular cycling pattern develops followed by an
increase in the days spent in estrus, and prolonged
exposure to estrogen;
» Pituitary alterations such as increase in pituitary weight,
increases in pituitary hyperplasia and pituitary -adenomas
become common as the animal ages;
» Acyclicity develops in the final months of life
*• Normal cycling -»irregular cycles -» prolonged estrus -» acyclicity
(begins around (occurs in the last few
5-6 weeks old) months of life
2.21 months of age)
'There may be temporal differences between and among strains
81
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.4 Strain Differences in Reproductive Aging and Mammary
Tumors
Evidence supporting the hypothesis that the normal reproductive
aging process in the SO female leads to mammary tumors in this strain
can be found by comparing the hormonal environment associated with
reproductive aging and mammary tumor historical control values in the
F-344 to the SD. Direct experimental evidence implicating the
reproductive aging process of the SD is also available and is discussed
below.
9.1.4.1 The Different Hormonal Environments of the SD
and F-344 Rats During Reproductive Aging
The hormonal environment of the normally aging SD female
is one of increased exposure to both estrogen and prolactin. The
hormonal environment of the normally aging F-344 rat is one of
increased exposure to prolactin and progesterone.
Increased or prolonged exposure to estrogen or prolactin is
implicated in mammary carcinogenesis. Female rats that have
been implanted subcutaneously with pellets of estrogen have very
high incidences of mammary tumors and an earlier onset of those
tumors than would be expected (Cooper, 1983; Cutts and Noble,
1964). Animals that have had pituitaries implanted under their
renal capsule2 or have been given lesions in the arcuate
nucleus-median eminence of the hypothalmus3 also have high
incidences and earlier onset of mammary tumors (Welsch et a/.,
1970a and 1970b). In contrast, increased or prolonged exposure
to progesterone does not increase the incidence or decrease the
onset of mammary tumors in the rat (Welsch, 1987).
Implantation of pituitaries under the renal capsule results in high levels of prolactin secretion
from the implanted pituitaries
3Lesions in this area of the brain result in production of high levels of pituitary prolactin.
82
-------
DRAFT: DO NOT CITE OR QUOTE
Increased exposure to either estrogen or prolactin increases
mammary tumor risk in the rat, while increased exposure to
progesterone - in the absence of increased exposure to either
estrogen or prolactin - does not increase mammary tumor risk.
Progesterone can, in fact, be anti-carcinogenic. The ratio of
estrogen to progesterone is an important aspect to consider when
examining mammary carcinogenesis. Progesterone has been
shown to "oppose" estrogen. This means that high levels of
progesterone can counteract high levels of estrogen and reduce
the increased risk of mammary cancer seen with high serum
estrogen levels. The molecular mechanism by which progesterone
opposes estrogen's actions appears to be by down-regulation of
mammary, hypothalamic and pituitary estrogen receptors
(Mauvais-Jarvis etal., 1987; Libertun etal., 1979; Cho etal., 1993;
Brannefa/., 1988).
9.1.4.2 Different Mammary Tumor Historical Control
Values in the SD and the F-344
The Thakur studies described above show the low incidence
of mammary tumors in control F-344 females. Charles River
Laboratories has reported relatively low historical control rates for
female F-344 mammary carcinomas (1.5%) and fibroadenomas
(12%) while control rates for other types of mammary tumors are
even lower (Lang, 1990). A recent report of the spontaneous
neoplasms seen in the National Toxicology Program (NTP) two-
year bioassays confirms a low mammary carcinoma rate for F-344
females (3.1%), but finds that the fibroadenoma incidence rates are
now 41.2% after having increased from 28% in 1985 (Haseman ef
a/., 1998). Hazelton Labs reports the historical control incidence of
F-344 female mammary carcinomas (adenocarcinomas and
carcinomas combined) as 3.8% and the rate of fibroadenomas as
14.9% (Hazelton Labs, 1994). The values for background
mammary tumors in the F-344 are much lower than in the SD (with
the exception of adenocarcinomas in the NTP studies). These
differences in background mammary tumor incidences could be
due the different mechanisms of reproductive aging.
83
-------
DRAFT: DO NOT CITE OR QUOTE
Historical control data from rat strains other than the SD and
F-344 confirms that animals whose predominant mode of
reproductive aging is constant estrus have higher background
mammary tumor incidences than F-344. Table 9-2 displays the
mode of aging in several strains of rats and the historical control
incidence of mammary tumors in each strain.
Table 9-2. Relationship of Reproductive Aging and
Mammary Tumor Incidence In Various Rodent Strains
Strain/
Species
SD/rat
Wistar/rat
Long-Evans
Hooded/rat
F-344/rat
Reproductive Aging
Constant estrus
Constant estrus
Constant estrus
Pseudopregnancy
Spontaneous
Mammary Tumor
Incidence
-30% fibroadenoma
-12% carcinoma
-25% fibroadenoma
-13% carcinoma
-50% with either
fibroadenoma or
carcinoma
-12% fibroadenoma
-2% carcinoma
References
Aging -Estes, 1982
Tumors - Lang, 1992
Aging - Mora ef a/.. 1994a; Mora
efa/.,1994b
Tumors - Walsh and Poteracki,
1994
Aging -Estes, 1982
Tumors - Cooper, 1983
Aging -Estes. 1982
Tumors - Lang, 1990
84
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.5 Studies in Which Premature Aging Was Artificially Delayed or
Induced and How Mammary Tumor Incidences Were Affected
Experimental evidence is available that examines the relationship
between reproductive aging and mammary carcinogenesis.
9.1.5.1 Mammary Tumor Onset Is Delayed in Rats Fed a
Diet Supplemented with L-tyrosine
Supplementing the diet of the female Long-Evans rat with
the amino acid L-tyrosine will delay the onset of constant estrus
(Cooper and Walker, 1979). These studies showed that
approximately 95% of the female Long-Evans rats on regular diets
will be in constant estrus by 13 months of age while 60% of the
females fed L-tyrosine will still be cycling normally at 16 months of
age. Mammary tumor onset in the Long-Evans females fed
L-tyrosine supplemented diets was also delayed. Mammary tumor
incidence in control LE females was 67% by 19 to 21 months of
age. No female fed an L-tyrosine supplemented diet had a
mammary tumor at 19 to 21 months. Even at 25 months of age,
L-tyrosine supplemented animals tumor incidence rates are only
25%.
The implication from these findings is that the delay in the
appearance of mammary tumors is due to the delay in the onset of
constant estrus (i.e., the delay in reproductive aging).
85
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.5.2 Mammary Tumor Onset Is Accelerated in Females
Receiving High Levels of Estrogen Through
Implants
The hormonal state of constant estrus rodents is one of
moderately elevated levels of serum estrogen and low levels of
serum progesterone (Huang et a/., 1978). Ovariectomizing female
rodents and then exposing them to a chronically high level of
estrogen through subcutaneous implants of silastic tubing
containing estrogen will mimic the hormonal state seen in constant
estrus females and these females will develop mammary tumors
much earlier than control animals (Cooper, 1983). The mammary
tumor incidence rate at 12 months for Long-Evans females who
were OVX at four months of age and given implants is 100% while
the control rate is less than 5%.
9.1.5.3 Mammary Tumor Onset Is Accelerated in Females
Whose Reproductive Aging Has Been Accelerated
by Exposure to Constant Light
Rats exposed to constant light (rather than a light cycle such
as 12 hours light/12 hours dark or 14 hours light/10 hours dark)
show an earlier onset of constant estrus. Such rats show an
increased incidence and earlier onset (latency was decreased by
34% compared to animals on light/dark cycles) of mammary
tumors- primarily adenocarcinomas (Molina etal., 1981).
9.1.6 The Correlation Between Increased Days in Estrus and
Mammary Tumors in a Two-Year Bioassay
Estrous cycle evaluations were performed in a two-year bioassay
using SD females (tumor data from this study - Morseth, 1998 - is
discussed in section 5.3 while estrous cycle data is discussed below
under section 9.2.3 - Results of Estrous Cycle Measurements in the SD
rat - Thakur, 1999). These evaluations showed that the percent of total
days spent in estrus was significantly higher in animals with mammary
tumors than in those without. Tumor latency was also reduced in animals
that spent a longer time in estrus.
86
-------
DRAFT: DO NOT CITE OR QUOTE
9.1.7 Summary and Discussion of the Hypothesis that Mammary
Tumors Are Induced by the Reproductive Aging Process
The hypothesis that the mode of reproductive aging in some strains
of rat (including the SD) may contribute to mammary tumor formation has
received attention for many years. The hypothesis is well established in
that:
» Female SD rats undergo a mode of reproductive aging that
results in extended periods of estrus;
> These periods of extended estrus result in prolonged
exposure to moderately elevated serum estrogen derived
from unovulated ovarian follicles;
> Serum estrogen acts at the pituitary to increase prolactin
secretion;
» Increased exposure to estrogen and prolactin increases the
risk of mammary carcinogenicity in rats.
The connection between the mode of reproductive aging and
mammary carcinogenicity can be confirmed by examining rodent strains in
that reproductive aging involves induction of a primarily constant estrus
state. The estrous cycles of the SD, Wistar, and Long-Evans strain of rat
all progress through a state of constant estrus as the animals age and all
these strains have high background incidences of mammary tumors. The
estrous cycle of the aging F-344 strain of rat is primarily one of
pseuctopregnancy and background mammary tumor incidences in the
F-344 rat are low.
87
-------
DRAFT: DO NOT CITE OR QUOTE
Experimental evidence showing the relationship between
reproductive aging and mammary carcinogenicity is provided by studies in
which the reproductive aging process in rats was either delayed or
accelerated resulting in mammary tumor onset being delayed or
accelerated. Delaying the onset of reproductive aging by feeding rats diet
supplemented with L-tyrosine delayed the onset of mammary tumors.
Accelerating the reproductive aging process by creating a hormonal state
that mimics the hormonal state of reproductive aging results in mammary
tumors appearing at a much earlier age. Accelerating the reproductive
aging process by exposure to constant light also accelerates the onset of
mammary tumors.
9.2 The Hypothesis that Atrazine Exposure Induces an Early Onset of
Attenuated LH Surge, Increased Days In Estrus, and Prolonged
Exposure to Estradiol
It is postulated that atrazine induces an early onset or increased incidence
of mammary tumors by inducing an early occurrence of an attenuated LH surge,
increased days in estrus, and prolonged exposure to estrogen and prolactin.
The evidence testing this hypothesis can be found in six areas:
* Examination of the time-to-tumor in atrazine exposed rats;
»• Examination of serum estradiol levels in atrazine exposed rats;
>• Examination of alterations in the ovary (early onset of decreased
CL, increased antral, secondary and atretic follicles) and vagina
(early onset of increased days in estrus) of atrazine exposed rats;
»• Examination of the pre-ovulatory LH surge in atrazine exposed
rats.
* Examination of pituitary alterations in atrazine-exposed rats.
»> Examination of mammary gland alterations that are indicative of
prolactin exposure in atrazine-exposed rats.
88
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.1 Time-to-Tumor
Examination of the onset time of mammary tumors in female SD
rats in the two-year bioassays (excluding the Mayhew study that is not
amenable to this type of analysis) shows that atrazine exposure did
induce an earlier onset of carcinomas (Tables 9-3 and 9-4). An early
onset of fibroadenomas was less evident, but nevertheless, there also
appeared to be an early of onset of this tumor type. Onset time was
determined in these studies by examining the histopathology data and the
clinical observations and correlating the appearance, by palpation, of each
mass that was subsequently confirmed by histopathology to be a
mammary tumor. Analysis of the onset time for mammary tumors in the
two-year bioassays with structurally related s-triazines simazine and
propazine, are also considered (Tables 9-5 and 9-6). Additional data
concerning onset time can be obtained from a one-year study submitted
by the registrant (Pettersen and Turnier, 1995).
89
-------
DRAFT: DO NOT CITE OR QUOTE
Table 9-3. Time to Mammary Tumor in the Female
SD Rat- Thakur (1992a) Terminal Sacrifice Protocol
x wkof
appear
s52wk
p-value
53-78 wk
p-value
79-1 04 wk
p-value
Dose (mg/kg/day) and Tumor Type
0
Fib
76.4
2/35
(5.7%)
0.379
16/35
(45.7%)
0.414
17/35
(48.6%)
0.15
3.79
Fib
76.1
1/27
(3.7%)
0.598
15/27
(55.6%)
0.304
11/27
(40.7%)
0.361
23.01
Fib
727
3/39
(7.7%)
0.552
18/39
(46.2%)
0.578
18/39
(46.2%)
0.510
0
Care
789
0/14
(0%)
0.024'
8/14
(57.1%)
0.099
6/14
(42.9%)
0.423
3.79
Care
72.5
3/11
(27.3%)
0.072
3/11
(27.3%)
0.138
5/11
(45.5%)
0.608
23.01
Care
65.4
6/18
(33.3%)
0.021'
5/18
(27.8%)
0.094
7/18
(38.9%)
0.553
Notes: The dose shown is in mg/kg/day.
Fib= fibroadenomas. Carc= carcinomas and adenomas combined.
psO.05
90
-------
DRAFT: DO NOT CITE OR QUOTE
&ft&
*52wk
p-value
53-78 wk
p-value
79-104 wk
p-value
Table 9-4. Time to Mammary Tumor in the Two-year Morseth (1988) Study
Dose (mg/kg/day) and Tumor Type
0
Fib
76.1
0/15
(0%)
0.549N
9/15
(60.0%)
0.1 04N
6/15
(40.0%)
0.092
1.5
Fib
724
1/18
(5.6%)
0.546
11/18
(61.1%)
0614
6/18
(33 3%)
0.486N
3.1
Fib
73.7
3/26
(11 5%)
0.244
13/26
(50 0%)
0.386N
10/26
(38.5%)
0590N
4.2
Fib
73.3
1/26
(3.8%)
0634
14/26
(53 8%)
0.479N
11/26
(42.3%)
0.575
24.4
Fib
76.3
1/22
(4.5%)
0.595
9/22
(40 9%)
0.211N
12/22
(54.5%)
0.297
0
Care
72.6
1/11
(91%)
0.047'
5/11
(45.5%)
0.110N
5/11
(45.5%)
0.526
1.5
Care
772
2/15
(13.3%)
0.619
6/15
(40.0%)
0548
7/15
(46.7%)
0.632
3.1
Care
78.6
0/14
(0%)
0.440N
7/14
(50.0%)
0570
7/14
(50 0%)
0.570
4.2
Care
64.4
2/10
(20 0%)
0.462
6/10
(60.0%)
0410
2/10
(20.0%)
0.221 N
24.4
Care
64.8
6/23
(26.1%)
0.252
7/23
(30.4%)
0.315
10/23
(43 5%)
0.600
NOTEs. The dose shown is in mg/kg/day.
Fib- fibroadenomas. Carc= carcinomas and adenomas combined
Trend for the dose is shown in the control column. "N" indicates that the trend is negative, otherwise trend is positive.
'psO 05
91
-------
DRAFT: DO NOT CITE OR QUOTE
Table 9-5. Time to Mammary Tumor with Simazine
in SD rats (McCormick et a/., 1988)
wkof
appear
s52wk
p-value
53-78 wk
p-value
79-1 04 wk
p-value
Dose (mg/kg/day) and Tumor Type
0
Fib
86
1/23
(4.3%)
0.029*
6/23
(26.1%)
0.000**
16/23
(69.6%)
0.000**N
0.52
Fib
81.7
1/28
(3.6%)
0.704
7/28
(25.0%)
0.590
20/28
(71.4%)
0.563
5.3
Fib
89.6
0/18
(0%)
0.561
3/18
(16.7%)
0.370
15/18
(83.3%)
0.260
63.1
Fib
68.8
6/42
(14.3%)
0.212
26/32
(61 9%)
0.006"
10/42
(23.8%)
0000-N
0
Care
834
2/18
(11.1%)
0.026*
3/18
(16.6%)
0.013*
13/18
(72.2%)
0.000** N
0.52
Care
80.7
3/17
(17.6%)
0.472
4/17
(23.5%)
0.466
10/17
(58.8%)
0.316 N
5.3
Care
82.1
2/20
(10.0%)
0.656
4/20
(20.0%)
0.563
14/20
(70.0%)
0.583 N
63.1
Care
65.9
13/44
(29.5%)
0.110
18/44
(40.9%)
0.059
13/44
(29.5%)
0.003*" N
NOTEs: The dose shown is in mg/kg/day.
Fib= fibroadenomas. Carc= carcinomas and adenomas combined
Trend for the dose is shown in the control column. "N" indicates that the trend is negative;
otherwise trend is positive.
*ps0.05,~ps0.01.
92
-------
DRAFT: DO NOT CITE OR QUOTE
Table 9-6. Time to Mammary Tumor with Propazine in SD rats (Jessup, 1980a)
wkof
appearance
*52wk
p-value
53-78 wk
p-value
79-1 04 wk
p-value
Dose (mg/kg/day) and tumor type
0
Fib
84
1/21
(4.8%)
0.406
4/21
(19.0%)
0.005**
16/21
(76.2%)
0.01 0*N
0.2
Fib
80.6
0/21
(0%)
0.500N
10/21
(47.6%)
0.050
11/21
(52.4%)
0.099N
6.4
Fib
77.6
1/17
(5.9%)
0701
8/17
(47.1%)
0.067
8/17
(47.1%)
0.065N
68
Fib
75.7
0/19(0%)
0.525N
13/19
(68.4%)
0.002**
6/19
(31.6%)
0.006** N
0
Care
90.2
0/7
(0%)
0.170
2/7
(28.6%)
0.430
5/7
(71.4%)
0.220N
0.2
Care
77.8
0/15
(0%)
1.000
8/15
(53.3%)
0.268
7/15
(46.7%)
0.268N
6.4
Care
85.2
0/13
(0%)
1.000
4/13
(30.8%)
0.664
9/13
(69.2%)
0.664N
68
Care
77.28
2/25
(8.0%)
0.605
11/25
(44.0%)
0.389
12/25
(48.0%)
0.254N
NOTEs: The dose shown is in mg/kg/day.
Fib= fibroadenomas. Carc= carcinomas and adenomas combined.
Trend for the dose is shown in the control column. "N" indicates that the trend is negative; otherwise
trend is positive.
*ps0.05;**ps0.01.
93
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.1.1 Thakur, 1992a
The mean week of fibroadenoma onset in controls in this
study was 76.4 weeks. The mean week of fibroadenoma onset in
the 3.79 mg/kg/day and 23.01 mg/kg/day groups was 76.1 and
72.7, respectively. The mean week of onset for carcinomas was
78.9 in controls while the mean week of onset for carcinomas and
adenomas in the 3.79 mg/kg/day and 23.01 mg/kg/day groups was
72.5 and 65.4.
The percentage of carcinomas and adenomas occurring in
the first year of the study in controls was 0% while at
3.79 mg/kg/day and 23.01 mg/kg/day 27.3 and 33.3% of the
carcinomas appeared in the first year of the study.
9.2.1.2 Thakur, 1991a
The percentage of fibroadenomas occurring in the first year
of the study in the controls was 16.7. At 4.23 mg/kg/day and 26.23
mg/kg/day the percentage was 0 and 20%.
The percentage of carcinomas occurring in the first year of
the study was 0 in controls and 33% at 4.23 mg/kg/day and 50% at
26.23 mg/kg/day.
94
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.1.3 Morseth, 1998
The mean week of fibroadenoma onset in controls in this
study was 76.1 weeks. The mean onset in the 1.5, 3.1,4.2 and
24.4 mg/kg/day groups was 72.4, 73.7, 73.3 and 76.3 weeks,
respectively. The mean week of onset for carcinomas and
adenomas in controls was 72.6 while the mean week of onset for
the 1.5, 3.1,4.2 and 24.4 mg/kg/day groups was 77.2, 78.6, 64.4
and 64.8, respectively.
The percentage of fibre-adenomas in the control group that
occurred in the first year of the study was 0. The percentage of
fibroadenomas that occurred in the first year of the study in the 1.5,
3.1, 4.2 and 24.4 mg/kg/day groups was 5.6,11.5, 3.8 and 4.5,
respectively. The percentage of carcinomas and adenomas that
occurred in the first year of the study was 9.1 and the percentage
that occurred in the 1.5, 3.1,4.2 and 24.4 mg/kg/day groups was
13.3, 0, 20, 26.1, respectively.
There was an increased mammary tumor incidence at the 53
week interim sacrifice, which also indicates and earlier tumor onset.
There were no fibroadenomas at the 53 week sacrifice in the 20
females of the control group, but in the dose groups the
fibroadenoma rates were 1/20, 2/19, 2/20,1/20 in the 1.5, 3.1,4.2
and 24.4 mg/kg/day groups, respectively.
9.2.1.4 Pettersen and Turnier, 1995
The percentage of animals with fibroadenomas in the
controls was 5.9 and the percentage that occurred at the 0.8,1.7,
2.8, 4.1 and 23.9 mg/kg/day dose groups was 5.9, 5.9, 0, 8.8, and
8.8, respectively. The percentage of animals with carcinomas and
adenomas in the control group was 2.9 and the percentage that
occurred at the 0.8,1.7, 2.8, 4.1 and 23.9 mg/kg/day dose groups
was 2.9, 2.9, 5.9, 5.9 and 14.7, respectively.
95
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.1.5 McCormick et a/., 1988 (simazine)
The mean week of fibroadenoma onset in controls in this
study was 86 weeks. The mean onset in the 0.52, 5.3, and 63.1
mg/kg/day groups was 81.7, 89.6 and 68.8 weeks, respectively.
The mean week of onset for carcinomas and adenomas was 83.4
for controls and 80.7, 82.1 and 65.9 for the 0.52, 5.3, and 63.1
mg/kg/day groups, respectively.
The percentage of fibroadenomas occurring in the first year
of the study in controls was 4.3 and the percentage of the 0.52,
5.3, and 63.1 mg/kg/day groups occurring in the first year of the
study was 3.6, 0, and 14.3, respectively. The percentage of
carcinomas occurring in the first year of the study in controls was
11.1 and the percentage occurring in the dose groups was 17.6,10
and 29.5, respectively.
9.2.1.6 Jessup, 1980a (propazine)
The mean week of fibroadenoma onset in controls in this
study was 84 weeks. The mean fibroadenoma onset in the 0.2,6.4
and 68 mg/kg/day groups was 80.6, 77.6 and 75.7 weeks,
respectively. The mean week of carcinoma onset in controls was
90.2 and the mean week of onset for the 0.2, 6.4 and 68 mg/kg/day
groups was 77.8, 85.2 and 77.3, respectively.
The percentage of fibroadenomas occurring in the first year
of the study was 4.8 and the percentage of fibroadenomas
occurring in the first year in the 0.2, 6.4 and 68 mg/kg/day group
was 0, 5.9, and 0, respectively. The percentage of carcinomas that
occurred in the first year of the study in controls was 0 and the
percentage occurring in the dose groups was 0, 0, and eight,
respectively.
96
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.2 Conclusions Of The Time-to-Tumor Data
The data from four two-year and one-year bioassays demonstrate
that there is a decreased time-to-tumor for mammary tumors in the female
SD rat following atrazine exposure. The earlier onset is more evident for
carcinomas and adenomas than for the fibroadenomas. The mean week
of carcinoma onset drops with atrazine exposure in both of the two-year
studies with atrazine for which mean week of tumor onset is applied and
also drops with exposure in the two-year bioassays with simazine and
propazine.
The percentage of carcinomas occurring in the first year of the
study is increased with exposure in three two-year bioassays with atrazine
and is also increased in a one-year bioassay. Both the simazine and
propazine studies showed an increase in percent of carcinomas and
fibroadenomas occurring in the first year in dosed animals compared to
controls.
9.2.3 Alterations in the Ovary and Vagina
9.2.3.1 Sprague-Dawley
Aging SD rats normally undergo alterations in their estrous
cycles. These alterations are described above under the section
9.1 and are summarized in Figure 9.1. In brief, the alterations in a
normally aging SD female rat are an increase in the percentage of
days of the estrous cycle spent in estrus from 25% prior to nine
months old to >40% after nine months old. The increase in days
spent in estrus occurs at the expense of days spent in diestrus and
proestrus.
97
-------
DRAFT: DO NOT CITE OR QUOTE
Eldridge etal. (1993a) examined estrous cycle alterations in
female SD rats in response to atrazine exposure. Care should be
taken when reading this document not to confuse the study that is
referred to as Thakur, 1991 a and the study referred to as Eldridge,
1993a. The Thakur and Eldridge studies are, in fact, the same
studies. The hormone and estrous cycle evaluation are referred to
in this document as Eldridge, 1993a while the animal necropsy and
histopathology portions of this study is referred to as Thakur,
1991 a.
Histomorphologic evaluation of the ovaries and other tissues
(including the vagina and mammary gland) from the
Thakur/Eldridge studies was performed and is referred to as
McConnell, 1995 in this document. Only the histomorphologic
evaluation of the ovaries and the vagina are discussed in this
section of the document. Discussed under section 9.2.5 is the
histomorphic evaluation of the mammary gland of atrazine-exposed
rats.
Therefore, the three citations - Thakur, 1991a; Eldridge,
1993a; and McConnell, 1995 - refer to different analysis performed
on the same group of SD or F-344 rats as part of the same study.
Additionally, estrous cycles were evaluated in one month, six
month and two-year studies - Morseth, 1996a, 1996b, 1998). The
protocol and measurements for estrous cycles in these studies are
described below.
9.2.3.2 Fischer-344
Aging F-344 rats normally undergo alterations in their
estrous cycles. These alterations are described above and are
summarized in Figure 9.1. In brief, a state of pseudopregnancy or
persistent diestrus becomes a common occurrence in F-344 rats as
they progress beyond approximately nine months of age. The
ovaries of animals in persistent diestrus contain CL, a moderate
number of secondary and antral follicles, and moderate numbers of
atretic follicles.
98
-------
DRAFT: DO NOT CITE OR QUOTE
The same study that examined serum hormone levels in
F-344 (Eldridge et a/., 1993b) also examined estrous cycle
alterations in response to atrazine exposure. Doses and sacrifice
schedules for these two studies were previously described.
Histomorphologic evaluation of the ovaries and other tissues
(including the vagina, mammary gland, and pituitary) from the
Thakur/Eldridge studies was performed and is referred to as
McConnell, 1995 in this document.
9.2.3.3 Protocol For Examination of the Ovaries and
Estrous Cycle Measurements
Thakur Studies. Two separate studies were
conducted: one with SD females and one with F-344
females. In both studies, 10 animals per dose group were
sacrificed after approximately one, three, nine, 12,15, and
18 months of exposure to atrazine. Two weeks before
scheduled sacrifice daily vaginal smears were performed.
The smears were examined for the presence of keratinized
epithelium, nucleated epithelium and leucocytes. The
presence of well-defined keratinized cells was taken to be
indicative of estrus. The presence of leucocytes in the
vaginal smears indicated diestrus and the presence of
moderate- to- dense nucleated epithelium and moderate
cornified epithelium was taken to be indicative of proestrus.
Following the 14 days of smear collection animals were
sacrificed at their next proestrus day. Thus, some animals
were sacrificed on their scheduled sacrifice date (on day 15
after vaginal smears were begun) while others were
sacrificed in the week following their scheduled sacrifice
date (on days 16 to 21 following initiation of vaginal smears,
depending on when their next proestrus phase occurred).
Any animals that had not had a proestrus phase by day 21
after initiation of vaginal smears was sacrificed on day 21.
Surviving animals were sacrificed at 24 months, again
following the procedure described above whereby each
animal is sacrificed on a proestrus phase, if possible.
99
-------
DRAFT: DO NOT CITE OR QUOTE
Results of the estrous cycle evaluations were
examined at: the dose groups at each individual timepoint
compared to controls; for trend within dose at each
individual timepoint; and the effect of treatment over time.
These are the same parameters for which the hormone
measurements were analyzed.
The ovaries of all females in these studies (both the
SD and F-344 studies) were examined by standard
histomorphologic techniques for several parameters
including (but not limited to): absence of CL; reduced
number of CL; presence of secondary and antral follicles;
presence of atretic follicles.
The parameters of absence of CL and reduced
number of CL were graded as simply being "present" or
"absent." The parameters of secondary, antral follicles, and
atretic follicles were graded on a scale of zero to five with
zero indicating the parameter was absent and one to five
indicating the parameter was present - the higher the
number the more of the follicles were present.
Morseth. 1996a (one-month study). Vaginal smears
were performed after seven days of treatment and continued
daily for three weeks. The criteria for evaluation and
classification of vaginal smears was similar to that followed
in the Thakur studies.
Morseth. 1996b (six-month study). Vaginal smears
were performed on the first day of treatment and continued
daily for 14 consecutive days every four weeks. Thus, an
animal was smeared in cycles of two weeks smearing
followed by two weeks non-smearing. This cycle continued
throughout the study. The criteria for evaluation and
classification of vaginal smears was similar to that followed
in the Thakur studies.
100
-------
DRAFT: DO NOT CITE OR QUOTE
Morseth. 1998 (two-year study). Estrous cycles
evaluations were performed on all intact females in this
study (80 per dose group). Vaginal smears were performed
for two consecutive weeks every two weeks starting on
study week one. Thus, animals had two weeks of smears
followed by two weeks without smears. The criteria for
evaluation and classification of vaginal smears was similar to
that followed in the Thakur studies. Estrous cycle data for
the first 46 weeks of the study have been analyzed.
9.2.3.4 Results of Ovarian Histomorphology and Estrous
Cycle Measurements in the SD Rat (McConnell,
1995 and Eld ridge, 1993a)
The only study that performed histomorphologic examination
of the ovaries of atrazine-treated SD rats was the Thakur, 1991 a
study. The histomorphologic data are presented in McConnell,
1995 while the vaginal smear data (for determination of phase in
estrous cycle) are presented in Eldridge, 1993a.
Ovarian Histomorpholoaic Examination (McConnell,
1995). An early onset of anovulation is seen as early as
three months following atrazine exposure. After three
months of exposure the number of control animals with an
absence of CL was zero of ten while the number of treated
animals with no CL was one of ten at the 4.23 mg/kg/day
group and two of ten in the 26.23 mg/kg/day group. This
small increase in animals with no CL indicates that incidence
of anovulation was increased in the dose groups. The fact
that no CL were seen at all indicates that ovulation had not
even occurred in the recent past, as even the CL from
previous cycles in which ovulation did occur had regressed
(i.e., become corpus albicans). The number of animals at
three months with reduced numbers of CL was increased
from two in the control group to three at 4.23 mg/kg/day and
three at 26.23 mg/kg/day. Reduced CL number also
indicates anovulation, but in these animals ovulation likely
did occur in the recent past as CL of some type were
present. The antral follicle group mean graded index (antral
101
-------
DRAFT: DO NOT CITE OR QUOTE
follicle score on the above described scale of zero to five
divided by the number of animals in the group) was 1.2 for
control, 2.1 for 4.23 mg/kg/day and 2.2 for 26.23 mg/kg/day.
This increase in antral follicles again indicates anovulation.
Evidence of early onset of anovulation was also seen
at nine months. After nine months of exposure the number
of control animals with an absence of CL was six of ten while
the number of treated animals with no CL was seven of ten
at the 4.23 mg/kg/day group and ten of ten in the 26.23
mg/kg/day group. The antral follicle group mean graded
index was 2.6 for control, 3.1 for 4.23 mg/kg/day and 3.8 for
26.23 mg/kg/day.
By 12 months of exposure the ovarian
histomorphology indicated that nearly all the animals in all
dose groups were not ovulating.
Estrous Cycle Evaluations (Eldridae. 1993a).
Appendix Table 8 displays the results of the estrous cycle
analyses done with the SD rat in Eldridge, 1993a.
The estrous cycle results seen in this study are what
would be expected in the SD strain of rat. The percent days
spent in estrus in control animals increases as the animals
age at the expense of days spent in diestrus, and proestrus.
Linear regression analysis indicated that the decrease in
diestrus and proestrus and the increase in estrus over the
period of one to18 months was statistically- significant at
p<0.01. The % days in both diestrus and proestrus are also
decreased in a statistically-significant manner (p<0.01) in
both dose groups for the period from one to 18 months.
Both dose groups showed a significant (p<0.01) trend
toward an increased % days spent in estrus from months
one to 18. Such an increase was seen in controls, is to be
expected, and did not appear to be altered by exposure to
atrazine.
102
-------
DRAFT: DO NOT CITE OR QUOTE
The percent days of the cycle spent in estrus was
dramatically increased in dosed animals compared to
controls at nine, 12 and 18 months. Females in the 4.23
and 26.23 mg/kg/day groups at nine months spent an
average of 34.3 and 44.8% of their days in estrus,
respectively (p<0.05 at 70 and p<0.01 at 400 ppm). This is
compared to 24% spent in estrus at the controls. The
dose-related trend, as determined using a
Terpstra-Jonckbeere Trend Test, showed a significant
increase in % days spent in estrus (p<0.01). At 12 months
there was also a dose-related trend (p<0.05) but the
increases in dosed groups compared to controls as
determined by ANOVA were not significant at either dose.
The percent days in proestrus is similar in dosed
animals compared to controls for all timepoints. There is a
significant trend (p<0.01) in both dose groups for a decrease
in % days spent in proestrus over time. This decrease was
seen in controls also and is to be expected. Atrazine
exposure did not appear to effect this trend.
The percent days spent in diestrus was significantly
decreased in dose groups compared to controls at nine, 12
and 18 months. Females at nine months spent a mean of
44.8% of the days of the estrous cycle in diestrus. Dosed
females spent 36.2% (p<0.05 compared to control) for the
70 ppm dose and 25.9% (p<0.01 compared to control) of the
days in diestrus. The dose-related trend at nine months was
significant at p<0.01. At 12 months there was also a
dose-related trend that was significant at p<0.01. The 4.23
mg/kg/day group, at 12 months, was decreased 14%
compared to control, but this was not statistically-significant.
The 24.23 mg/kg/day group at 12 months was decreased
28% compared to controls and this was significant (p<0.05).
The trend at 18 months was decreased in a dose-related
manner (p<0.01) while both the 4.23 and 24.23 mg/kg/day
groups compared to controls by ANOVA were also
statistically-significantly decreased (p<0.05, for both).
103
-------
DRAFT: DO NOT CITE OR QUOTE
To summarize the estrous cycle data from this study:
* When analyzed for effect over time it was seen that
animals in both dose groups and the controls
exhibited an increase in % days in estrus and a
decrease in % days in di- and proestrus as the study
progressed. These alterations are to be expected in
an aging SD female rat and exposure to atrazine did
not appear to affect these parameters compared to
controls.
» When analyzed for effect of dose it was seen that
there was a significant increase, compared to
controls, in percent days spent in estrus at nine, 12
and 18 months in both dose groups. The percent
days spent in diestrus significantly decreased in both
dose groups compared to controls.
9.2.3.5 Results of Estrous Cycle Measurements in the SD
Rat (Morseth, 1996a and 1996b)
These studies examined the effect of atrazine on the estrous
cycle and on plasma concentrations of the hormones LH and
prolactin. The hormone measurement data from these studies will
be discussed below. Preovulatory LH levels, and the estrous cycle
evaluations will be discussed here. The studies were of one month
(Morseth, 1996a) and six months (Morseth, 1996b) in duration.
One-Month Study. Female SD rats were exposed to
0, 2.5, 5, 40 and 200 mg/kg/day technical grade atrazine for
28 to 31 days. Ninety females per dose group were used.
Dosing was by gavage once a day, at approximately the
same time each day.
104
-------
DRAFT: DO NOT CITE OR QUOTE
The results of the smears indicated an effect of
atrazine exposure on vaginal cycling. Atrazine exposure
induced a dose-dependent increase in the number of
animals having irregular cycles. The nature of the irregular
cycles was both an increase in estrous cycle blocks (two
consecutive days in estrus) and an increase in diestrus
blocks (four consecutive days in diestrus). The effects of
atrazine on the estrous cycle in this study are most
pronounced at 40 and 200 mg/kg/day with a statistically-
significant increase in females displaying diestrus blocks at
both 40 and 200 mg/kg/day and a statistically-significant
increase in females displaying estrus blocks at 200
mg/kg/day only (p 0.05 for both diestrus and estrus blocks
using pairwise comparison). These data are shown in
Appendix Table 9.
Six-Month Study. Female SD rats were exposed,
through the diet, to 0, 1.8. 3.65 and 29.4 mg/kg/day
technical grade atrazine for six months. Ninety females per
dose group were used.
A statistically-significant increase in percent days in
estrus was evident as early as 3.5 months into the study in
the high dose group, and in the mid-dose group by 5.5
months into the study. The low-dose group never
experienced statistically-significant alterations in their estrus
cycles. These data are shown in Appendix Table 10.
105
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.3.6 Results of Estrous Cycle Measurements in the SD
Rat (Morseth, 1998)
The estrous cycle evaluations performed in this two-year
bioassay demonstrate that atrazine-treated females display
increased days in estrus sooner than control animals. All animals,
irrespective of dose group, spent a normal amount of time in estrus
(approximately 25% of the days spent in estrus) for the first 10
weeks of the study. By the 13 to 14 week measurement period
though, the atrazine-treated animals began to display more days
spent in estrus. These increases were most evident in animals of
the high-dose group while, for most timepoints, the other three
dose groups showed only marginal increases in percent days in
estrus. The differences between the control and dose groups were
most evident at week 25 to 26 where controls spent 53.7% of the
days in estrus compared to 63.8, 59.7, 55.4, and 72.1% of the days
in estrus for 1.5, 3.1, 4.2, and 24.4 mg/kg/day groups. Appendix
Table 11 displays the percent days in estrus by dose group for all
the time periods up to 46 weeks. Animals dosed with atrazine also
showed an increase in the likelihood of having an estrus block of
seven days during one of the three measurement periods in the
17-26 week interval. Again, this effect was most evident at the
high-dose group, but was also seen at the other dose groups.
Appendix Table 12 displays these data.
In addition to examining the amount of time spent in estrus
by atrazine-treated animals compared to control animals, this study
also examined the relationship between amount of time spent in
estrus and mammary tumor incidence. There was a clear
relationship between the amount of time spent in estrus and
mammary tumor onset. These data are shown in Appendix Tables
13 and 14.
106
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.3.7 Results of Ovarian Histomorphology and Estrous
Cycle Measurements in the F-344 Rat (McConnell,
1995)
The only study that performed histomorphologic examination
of the ovaries of atrazine-treated Fischer rats was the Thakur,
1991b study. The histomorphologic evaluation data is presented
as McConnell, 1995 while the vaginal smear data (for
determination of phase in estrous cycle) is presented as Eldridge,
1993b. HED believes that the estrous cycle evaluation (vaginal
smears) reported in Eldridge, 1993b are unreliable. Thus, these
data are not reported. The histomorphology data from McConnell,
1995 are instead used to determine stage of the estrous cycle.
Ovarian Histomorpholoaic Examination in the F-344
(McConnell, 1995). The great majority of F-344 rats in all
groups, both control and dose groups, maintained CL
throughout most of the study. Only at the final, 24-month,
timepoint were there dramatic decreases in CL numbers.
The atrazine-treated animals at this timepoint did not show
decreases in CL numbers any more severe than the control
animals. The reduction in CL numbers at this late timepoint
appears to be a consequence of a natural progression of the
animals from persistent diestrus into acyclicity. All animals
in all dose groups maintained moderate numbers of
secondary, antral and atretic follicles throughout the study -
including the 24-month timepoint.
107
-------
DRAFT: DO NOT CITE OR QUOTE
Histomorpholoav to Determine Stage of the Estrous
Cycle in the F-344 (McConnell, 1995). All animals in all
dose groups appeared to maintain normal cycles through
the first 12 months of the study. At the 15-month timepoint
approximately half the animals in all dose groups (five often
in control; three of ten at 0.68 mg/kg/day; 6 of 11 at 4.82
mg/kg/day; five often at 14.05 mg/kg/day, and five often at
34.33 mg/kg/day) were in a state of extended diestrus. The
state of extended diestrus was indicated by the presence of
increased vaginal mucification - which is indicative of
extended diestrus (McConnell, 1989a). At the 18-month
timepoint approximately 70 to 80% of the animals in all dose
groups displayed increased vaginal mucification. The
persistent diestrus (or pseudopregnancy) that is
characteristic in an aging F-344 was evident by 15 months
and was quite common by 18 months. At 24 months the
incidence of vaginal mucification was still high, but the
previously mentioned reduction in animals with CL indicated
that a progression towards acyclicity was occurring in the
animals in this study.
9.2.4 Summary And Discussion From The Ovarian Histomorphology
and Estrous Cycle Measurements In F-344 and SD Strains
Over the duration of the study, both strains of rat exhibited ovarian
histomorphology and estrous cycles that would be expected for those
strains. The F-344 maintained CL throughout most the 24-month study
and showed an increase in days spent in diestrus (as indicated by vaginal
mucification) at the post-12 month timepoints (McConnell, 1995). The SD
showed decreased numbers (and frequently, a complete absence of) CL,
increases in secondary, antral and atretic follicles, and an increase in
percentage of days in estrus (as indicated by the vaginal smears) as the
study progressed (Eldridge, 1993b; McConnell, 1995).
108
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine exposure in the F-344 did not seem to alter estrous
cycles; atrazine exposure in the SD increased the number of estrus and
diestrus blocks after as little as one month of exposure (Morseth, 1996a).
Three months of exposure increased the percentage of days spent in
estrus and decreased the percentage of days spent in diestrus, but not
significantly so (Thakur, 1991 a). By nine months of atrazine exposure the
percentage of days spent in estrus were significantly increased compared
to controls and the percentage of days spent in diestrus was significantly
decreased. The increase in estrus and diestrus blocks after one month of
atrazine exposure indicates that these females were cycling irregularly.
The increased days spent in estrus at three and nine months indicates
that the atrazine-exposed animals in this study were entering constant
estrus sooner than the control animals.
As noted above, reproductive aging is manifested in the estrous
cycle first as irregular cycling, then as constant estrus, and finally as
acyclicity/persistent diestrus. Atrazine exposure was able to induce
irregular cycling in the SD females after only one month of exposure -
when the animals in this study were about three months of age.
The Thakur, 1991 a; Eldridge, 1993a, and the two shorter duration
Morseth studies indicate that atrazine exposure can alter estrous cyclicity
in the female SD rat. Atrazine exposure did not seem to effect estrous
cyclicity in the F-344 rat (McConnell, 1995).
109
-------
DRAFT: DO NOT CITE OR QUOTE
The estrous cycle data from the Morseth two-year study (described
in Thakur, 1999) confirmed the findings from Eldridge (1993a) and the
shorter duration Morseth studies (Morseth 1996a and 1996b). In this
study atrazine exposure resulted in SD females spending an increased
amount of time in estrus earlier than control animals. The high-dose
group (26.23 mg/kg/day) females in the Thakur study showed a group
mean of 44.8% of the days in estrus at nine months. The control animals
reached this level of days in estrus, but not until 12 months when their
group mean days in estrus was 42.6%. The high-dose (24.4 mg/kg/day)
females in the Morseth, 1998 study had a group mean of 47.8% of the
days in estrus during weeks 17 to 18. The control animals reached this
level of days in estrus, but not until weeks 21 to 22 when their group mean
days in estrus was 45.6%. All four studies together provide strong
evidence that atrazine can disrupt estrous cycles in the SD female and
can lead to an early onset of increased percent days in estrus compared
to control animals.
The Morseth (1998) study examines correlations between days in
estrus and mammary tumor incidence and onset and demonstrates that
increased days in estrus early in the life of an animal decreases the time
to onset of mammary tumors. This Morseth study also demonstrated that
even animals who are not spending increased time in estrus during the
early period of life, will show an increased risk of mammary cancer when
chronically fed atrazine.
9.2.5 Serum Estradiol and Prolactin Levels
Given the postulated mode of action, it is important that one
examine serum estradiol and prolactin levels and confirm that they are, in
fact, altered following exposure to atrazine.
Serum hormone levels of estradiol (as well as progesterone,
prolactin and corticosterone) in atrazine exposed SD female rats were
examined in Eldridge, 1993a. Serum hormone levels of these four
hormones were examined in F-344 females in Eldridge, 1993b.
110
-------
DRAFT: DO NOT CITE OR QUOTE
Serum corticosterone was measured as an indicator of stress while
estradiol, prolactin, and progesterone were measured because of their
roles (estrogen and prolactin as promoters of and progesterone as a
possible inhibitor of) mammary carcinogenesis.
Eld ridge 1993a and 1993b are discussed below. The
histomorphological evaluation of the mammary gland in atrazine-exposed
SD rats is also discussed below (McConnell, 1995)
9.2.5.1 Protocol and Rationale for Hormone Measurement
Doses and sacrifice schedules for these two studies were
previously described. The protocol and measurements for
hormone levels in these studies are described below.
Blood was collected from the trunk of all animals at
scheduled sacrifice. Two weeks prior to each scheduled sacrifice
vaginal smears were performed. Animals were sacrificed during
proestrus, if possible. If, after 21 days of vaginal smears, an
animal was not in proestrus, then this animal was sacrificed,
irrespective of what stage of the estrous cycle the animal was in.
Blood samples were used for the determination of serum estradiol,
progesterone, prolactin, and corticosterone. The hormones
estradiol, progesterone, and prolactin all are important in the
regulation and maintenance of normal reproductive functioning in
the rat and could potentially play roles in mammary tumor
pathogenesis. Corticosterone measurements were taken to test
the hypothesis that dosing the animals with atrazine produces
stress (and thus an elevation of serum corticosterone levels) that
may play a role in mammary gland neoplasia.
111
-------
DRAFT: DO NOT CITE OR QUOTE
Radioimmunoassay techniques were used to measure levels
of each hormone. Standard curves were constructed and sample
values were compared to the standard curve. The results of the
measurements were examined for alterations in:
» each individual timepoint compared to controls;
* trend within dose, and;
* effect of treatment over time.
9.2.5.2 Results of Hormone Measurements - F-344
{Eldridge.era/., 1993b)
The results of the hormone measurements did not reveal
any consistently statistically-significant alterations in serum
hormone levels compared to controls for any of the hormones
tested. There were occasional significant alterations such as
significantly decreased (p< 0.05) progesterone levels in the 4.82
mg/kg/day dose and corticosterone levels in the 34.33 mg/kg/day
group at the 12-month timepoint compared to controls; significant
negative trends in estradiol levels at the 12-month timepoint,
progesterone levels at the 12- and 18-month timepoints, and
corticosterone levels at the 12- and 15-month timepoints, and
prolactin levels displayed a significant positive trend at the three-
month timepoint only. Careful consideration of these alterations
indicated that they did not appear to be related to atrazine
exposure.
There were alterations in serum hormone levels that were
seen in control as well as treated rats. These were: decreases in
estradiol levels in the later half of the study; significantly increasing
progesterone levels from one to18 months. Decreased estradiol
and increased progesterone are expected in rats undergoing a
reproductive aging process involving pseudopregnancy (Huang et
a/., 1978). Exposure to atrazine did not alter the age-related
changes in estradiol or progesterone levels.
112
-------
DRAFT: DO NOT CITE OR QUOTE
An increase in prolactin levels might be expected in an aging
rat undergoing pseudopregnancy. However, consistent increases
in serum prolactin levels were not seen.
9.2.5.3 Results of Hormone Measurements - SD (Eldridge
efa/.,1993a)
Appendix Table 15 displays the results of the hormone
measurements in the SD females.
Serum progesterone and corticosterone levels did not show
any significant dose-related alterations compared to controls.
Serum prolactin levels in dosed groups did not show any significant
dose-related alterations with the exception of the 26.23 mg/kg/day
group that did however, show a negative trend (p<0.01) over the
nine to 18 month period and a positive trend at nine months.
These alterations in serum prolactin likely were related to
compound exposure.
Serum estradiol levels in the control rats in this study
showed a positive trend (levels increased as time increased) over
months one through nine. Exposure to atrazine did not alter this
trend. The 4.23 and 26.23 mg/kg/day dose groups also showed a
significantly positive trend from months one through nine. These
trends are expected as constant estrus would be expected to begin
to set in as these animals approach nine months of age.
Examination of the pairwise comparisons at three months indicates
that treated animals had an early onset of increased serum
estradiol levels compared to controls. At three months control
estradiol levels were 3.5 ng/mL, 70 ppm levels were 11.2, and 400
ppm levels were 16.2 ng/mL. The increase at 70 ppm was
significant at p<0.05, the increase at 400 ppm was significant at
p<0.01 and the trend, determined using a Terpstra-Jonckbeere
Trend Test, was positive at p<0.05. At nine months control and
4.23 mg/kg/day group estradiol levels were similar, but the 26.23
mg/kg/day group compared to controls was increased 44%. At
nine months estradiol levels were elevated compared to control,
but not significantly so.
113
-------
DRAFT: DO NOT CITE OR QUOTE
Prolactin levels were not altered either as a result of atrazine
exposure or as result of aging in this study.
9.2.5.4 Results of the Histomorphologic Evaluation of the
Mammary Gland - SD (McConnell, 1995)
The mammary gland is clearly a hormone-responsive tissue.
Various tissues in the mammary gland contain receptors for the
hormones estrogen, progesterone and prolactin and exposure to
these hormones effects on these tissues. A detailed
histomorphologic analysis of mammary gland (and other) tissues
from the rats in the Thakur, 1991 a study was performed. This
histomorphologic analysis is referred to in this document as
McConnell, 1995.
The mammary glands were examined for these alterations:
» Acinar development - indicative primarily of estradiol
exposure
» Acinar/lobular development- indicative of both
prolactin and progesterone exposure
» Secretory activity - indicative of prolactin, and to a
lesser extent, estrogen and progesterone exposure
» Dilated ducts with secretion - indicative of prolactin,
and to a lesser extent, progesterone exposure
» Galactocele (milk cyst) - indicative of prolactin, and to
a lesser extent, progesterone exposure
The alterations in the tissues of the acinar region are
indicative of estrogen exposure (ductal epithelial hyperplasia and
acinar development). The alterations related to milk production are
primarily prolactin-dependant (secretory activity, dilated ducts with
secretion, galactocele).
114
-------
DRAFT: DO NOT CITE OR QUOTE
Appendix Table 16 displays the results of the
histomorphologic analysis of the mammary glands at months one,
three, nine and 12 in the SD rats from Thakur, 1991 a. Values at
15,18 and 24 months are not shown as these values are similar to
control values. The index weighted scores of several of the above
listed parameters are shown in this table. An index weighted score
assigns a numerical value to the severity of the finding assigned by
the pathologist. The higher the index weighted score, the more
severe was the finding in that group.
Increased prolactin exposure in the rat is associated with
formation of galactoceles (milk cysts). Galactocele incidence and
severity in this study are shown in Appendix Table17. The results
of the histomorphologic analysis are described below:
Acinar Development. An early onset of increased
exposure to estrogen is indicated by examination of the
column of Appendix Table 16 labeled "Acinar Development."
The index scores at the one-month timepoint are slightly
higher in the dose groups than in the control, but a
dose-response relationship was not seen. At the three-
month timepoint the index scores are again increased over
control. The increase is at this timepoint is dose-related
though with the high dose being more severe than the low
dose. At both nine and 12 months the index score again
indicate more severe acinar development in the dose groups
compared to the controls with the increase in severity being
especially obvious at the high dose.
Acinar/Lobular Development. An early onset of this
parameter is evident. The one- and three-month timepoints
have index scores in the dose groups that are similar to the
control index scores. The dose group index scores are
clearly increased compared to controls at nine and 12 month
timepoints though. This indicates that the atrazine-treated
animals were exposed to elevated levels of prolactin at an
earlier timepoint than the control animals.
115
-------
DRAFT: DO NOT CITE OR QUOTE
Secretory Activity. The index scores for secretory
activity also demonstrate an early onset of increased
prolactin exposure in the dose groups compared to the
controls. Index scores at one and three months in the dose
groups were similar to control values, while index scores in
both dose groups were clearly elevated compared to
controls at the nine- and 12-month timepoints.
Dilated Ducts with Secretion. Index scores at one
and three months in the dose groups were similar to control
values. The index scores for the 4.23 mg/kg/day group
compared to controls were only slightly elevated. The index
scores for the 26.23 mg/kg/day group were greatly elevated
compared to controls
Galactocele Incidence and Severity. No galactoceles
were observed in any group at the one- and three-month
timepoints. At nine and 12 months galactoceles in the dose
groups were increased in both number and severity. By 15
months galactocele incidence and severity were similar
between control and dose groups.
9.2.6 Summary And Discussion Of The Hormone Measurements and
Histomorphologic Alterations In F-344 And SD Strains
Serum corticosterone levels were not altered in F-344 or SD female
rats by atrazine exposure. Serum corticosterone levels reflect stress and
the lack of any alteration in these levels indicates that the dosing, and
more importantly, the regular vaginal lavages, were not causing the
animals in these studies undue stress that may have compromised the
results of these studies.
116
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine exposure in the F-344 females did not alter serum
hormone levels of estradiol, progesterone or prolactin. Both control and
dose groups did see decreases in serum estradiol levels for the latter half
of the study and a generally increasing level of progesterone for the one
to 18 month period. These changes in estradiol and progesterone are to
be expected in animals undergoing reproductive aging through
pseudopregnancy. What was not seen, but would be expected in an
animal undergoing reproductive aging through pseudopregnancy, were
increases in serum prolactin levels.
Accurate serum prolactin levels from the rat can be difficult to
obtain because many different factors can cause dramatic alterations in
serum prolactin levels. For example, simple inadvertent stimulation of a
female rats nipples can induce large increases in serum prolactin
(Freeman, 1981). Prolactin levels are also very sensitive to stress. A rat
in pseudopregnancy is especially difficult to obtain accurate serum
prolactin measurements from as these animals will display twice-daily
prolactin surges. Prolactin measurements from these animals will vary
dramatically depending on whether or not a measurement is taken during
a surge, and, if taken during a surge, at what point in the surge. Accurate
prolactin measurements in a young, unmated rat, are easier to obtain as
these animals have relatively static prolactin levels except for have one, or
possibly two, prolactin surges every four days (a proestrus afternoon
surge and, sometimes, a smaller surge on estrus) (Freeman, 1981;
Butcher era/., 1974).
Despite these difficulties, it is still surprising that increased prolactin
levels were not measured in this study in the aging animals.
117
-------
DRAFT: DO NOT CITE OR QUOTE
Atrazine exposure in the SD females did not alter serum
progesterone levels or serum prolactin levels in dose groups compared to
controls. The serum estradiol levels in the SD rat were dramatically
altered by atrazine exposure and deserve special attention. Atrazine
exposure resulted in an early exposure to high levels of estrogen. The
levels of serum estrogen in the dosed groups of 70 and 400 ppm females
at three months were 11.2 ± 12.6 and 16.2± 13 ng/mL compared to only
3.5 ± 6.4 ng/mL in the control females at this time point. While the
standard deviations in these groups are large the increases are three to
over four-fold, and the increases are statistically-significant using a
pairwise comparison. The atrazine exposed groups had higher serum
estrogen levels than controls at the three-month timepoint. There was
also a positive dose-related trend over the first nine months of the study in
estradiol levels. Exposure to such high levels of estrogen this early in the
rats life is not normal. Exposure to these levels of estrogen at nine
months, as can be seen from Appendix Table 15, is normal. The early
exposure to these high estrogen levels may be leading to an earlier onset
of mammary tumors.
As was the case with the F-344, increased prolactin levels would
be expected as a consequence of the normal aging process in the SD; yet
increases in serum prolactin were not seen in this study. Old SD females
with pituitary adenomas can have serum prolactin levels that are
approximately 13 times higher than in young SD females (Sarkar et a/.,
1982). An increase in all the groups (both control and atrazine-treated)
would be expected as the pituitaries increased in size and pituitary
adenomas became common. By the 18- and 24-month timepoints the
majority of the animals in all groups had pituitary adenomas; yet serum
prolactin levels at these two timepoints were similar to control values.
Increases in serum prolactin levels have been seen in numerous studies
in the published literature and are accepted to be a normal part of the
aging process in the SD female and other rats that undergo constant
estrus as the predominant mode of reproductive aging (Console et a/.,
1997; McComb et a/., 1984; Sandusky era/., 1988; Baird era/., 1990;
McComb et a/., 1986; van Putten ef aL, 1988).
118
-------
DRAFT: DO NOT CITE OR QUOTE
The fact that measured serum prolactin levels were not increased
with age in either strain must be considered a weakness in this study,
despite the difficulties of such measurements. This is especially true in
the SD study where very large increases in serum prolactin levels should
have been evident.
Because of the lack of increased serum prolactin (as determined by
direct serum measurements) in the aged animals in Eldridge, 1993a, the
histomorphologic data from this study was examined especially closely for
any signs of increased prolactin exposure.
The incidences and severity of all four parameters that indicate
increased exposure to prolactin (acinar/lobular development, secretory
activity, dilated ducts with secretion, galactoceles) were increased in
atrazine-treated groups compared to controls at the nine and 12-month
timepoints. Index scores for values before nine months and after 12
months were similar among dose groups. The increased index scores at
nine and 12 months in dose groups compared to controls indicates an
early onset of increased prolactin exposure in the dose groups compared
to the control. With time, the normal aging process proceeds in the
control animals and by 15 months the control animals have "equalized" or
"caught up" with the dose groups. Thus, index scores are similar for the
timepoints after 12 months. Index scores are similar for the timepoints
prior to nine months because increases in prolactin require first that
estrogen be increased. The increased estrogen then acts at the pituitary
to induce lactotroph hyperplasia that results in increased prolactin levels
and, finally, increased incidence and severity of these mammary gland
findings. Appendix Table 15 shows increases in serum estradiol at three
months in the dose groups compared to controls while Appendix table 16
shows a dose-related increase in acinar development (primarily an
estrogen-dependent effect) at three months in the dose groups compared
to control groups. The increased serum estrogen levels seen at three
months are affecting the pituitary between three and by nine months the
pituitary lactotrophs are proliferating and producing prolactin such that by
nine months prolactin-dependent alterations at the mammary gland are
evident in the dosed animals.
119
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.7 Preovulatory LH Levels
The effect of aging on the preovulatory luteinizing hormone (LH)
surge, has been briefly discussed previously in this document. As in
humans, ovulation is triggered in rats by a sudden and dramatic increase
(a surge) in serum LH levels. The attenuation of the preovulatory LH
surge in aging female rat strains, including the SD, has been
well-described (Lu et a/., 1979, Cooper et a/., 1980). One- and six-month
studies examining the effect of atrazine exposure on the preovulatory LH
surge are available (Morseth, 1996a; Morseth, 1996b; Minnema, 2000).
These data show that atrazine exposures as short as one month can
dramatically attenuate the pre ovulatory LH surge. Plasma prolactin levels
were also determined in the Morseth, 1996a study and results of this
analysis were reported. Plasma prolactin concentrations have been
shown to undergo a preovulatory surge in rats similar to the LH surge
(Butcher et a/., 1974). However, this document will not go into a detailed
discussion of the effect of atrazine on the plasma prolactin surge simply
because the role of this event in female rat inducing ovulation are not as
well described as a the LH surge.
9.2.7.1 Protocol for LH Surge Measurement - One-Month
Study (Morseth, 1996a)
Female SD rats were exposed to 0, 2.5, 5,40 and 200
mg/kg/day technical grade atrazine for 28 to 31 days. Ninety
females per dose group were used. Dosing was by gavage once a
day, at approximately the same time each day. After 28 to 31 days
of atrazine exposure the animals were OVX. Vaginal smears were
performed from days seven to the day prior to OVX to determine
the animals cycling patterns. Estradiol implants were placed in
each animal seven days following OVX and the animals were
sacrificed three days later. Thus, OVX occurred 10 days prior to
sacrifice and estradiol implants occurred three days prior to
sacrifice. This protocol of OVX followed by estradiol implantation,
followed by sacrifice has been previously used by other
investigators to induce an LH surge in SD female rats (Legan etal.,
1975).
120
-------
DRAFT: DO NOT CITE OR QUOTE
Ovarectomization of the animals followed by implantation of
estradiol implants was done in an attempt to synchronize the
estrous cycles of the animals so that the LH surge in each animal
would occur at approximately the same time. This allows for
comparison of the surge between dose groups and also points the
investigator to a specific time at which the LH surge should be
occurring. With this information the animals may be bled and
serum LH levels measured. Animals were bled for serum hormone
measurements at six different timepoints spread over 12 hours.
The first two time points were 1100 and 1400 in biologic time
(biologic time being time in the light cycle - biologic time 1200 is
mid-point of the light cycle or noon of the light cycle). These first
two timepoints are baseline. The other timepoints for serum
measurement are biologic time 1600,1800, 2000, and 2300. The
peak of the LH surge would be expected to occur in the late
afternoon of the light cycle (around 1800 biologic time). This
equates to the late afternoon of proestrus in a normally cycling rat.
By 2300 biologic time the LH surge would be expected to be over
and LH values should return to baseline levels.
Out of the 90 animals in each group 10 were "repeat bleed."
These animals were bled at each timepoint from the jugular vein for
the first four bleedings, through the ocular venous plexus for the
fifth bleed and the trunk for the last timepoint. The remaining 80
animals in each dose group were sacrificed, and trunk blood was
collected according to the schedule shown in Appendix Table 18.
121
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.7.2 Results of LH Surge Measurements - One-Month
Study
The means and standard deviations for serum LH
measurements from this study are shown in Appendix Table 19.
Non-repeat Bleed. Plasma LH values of the 200
mg/kg/group were significantly decreased at 1600 and 1800
compared to controls at that timepoint (specific p value not
given). There were also non-significant decreases in plasma
LH levels, compared to controls, in the 5 and 40 mg/kg
group (45.4% and 36.8%, respectively). There was not a
great increase in the 200 mg/kg group in the magnitude of
the peak response over its own baseline value. Control
mean baseline (1100 and 1400 hours) values are 998 and
1122 pg/mL compared to a peak value of 5138 pg/mL at
1800: approximately a five-fold increase. The baseline
values for the 200 mg/kg group are similar to controls - 873
and 1099 pg/mL However, the peak value in the 200 mg/kg
group is only 2752 pg/mL: an increase of only 2.5-fold.
Repeat Bleed. Peak values compared to controls in
the 200 mg/kg group were, as in the non-repeat bleed set,
significantly decreased. The peak control value (1800
hours) was 2650 pg/mL while the peak 200 mg/kg value was
812 pg/mL (1800 hours). The 40 mg/kg group was
decreased compared to controls (1450 pg/mL) but not
significantly so. There was little increase in the 200 mg/kg
group in the magnitude of the response over its own
baseline value. Control baseline values were 732 and 786
pg/mL compared to peak values of 2650 pg/mL:
approximately a 3.5-fold increase. Peak values in the 200
mg/kg group were increased only about 45% over baseline
values (812 vs. 514 and 453 pg/mL).
122
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.7.3 Summary of the Plasma LH Measurements from
the One-Month Study
The plasma LH values seen in the control animals in both
the repeat and non-repeat bleed data indicate that one month of
atrazine exposure dramatically attenuates the preovulatory LH
surge in the high-dose group of 200 mg/kg/day. This effect is only
statistically-significant in the high-dose group of 200 mg/kg, but it
can be seen in the 40 mg/kg/day group also.
The results seen in this study are confirmed by a separate
28-day using similar protocols and identical doses (Minnema,
2000). The LH surge, in this study, was statistically significantly
attenuated at both the 40 and 200 mg/kg/day doses.
9.2.7.4 Protocol for LH Surge Measurement - Six-Month
Study (Morseth, 1996b4)
The protocol for this study was very similar to the protocol
for the one-month study. The main differences were:
» duration of atrazine exposure (26 weeks);
» route of exposure (through the diet);
•• and, dose levels (25, 50 and 400 ppm -1.8, 3.65, and
29.44 mg/kg/day).
Other than these differences, the six-month study was
conducted much the same as the one-month study. Animals were
OVX 10 days prior to sacrifice and implanted with estradiol implants
three days prior to sacrifice. Ninety females per group were used
in a sacrifice schedule identical to that used in the one-month study
(shown in Appendix Table 18). Blood collection and plasma
hormone measurements were performed identical to the methods
used in the one-month study.
"Data from the study referred to here as Morseth 1996b has been published in
the open literature as Eldridge etal., 1999.
123
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.7.5 Results of LH Surge Measurements - Six-Month
Study
The means and standard deviations for serum LH
measurements from this study are shown in Appendix Table 20.
Non-repeat Bleed. Plasma LH values of the 400 ppm
group were significantly decreased at 1400,1800, and 2000
hours compared to controls at those timepoints (specific p
value not given). There was not a large increase in the 200
mg/kg group in the magnitude of the response over its own
baseline value (specific p value not given). Control mean
baseline (1100 and 1400 hours) values are 1900 and 2326
pg/mL compared to a peak value of 3458 pg/mL at 1800:
approximately a 1.6 -fold increase. The baseline values for
the 200 mg/kg group are slightly less than controls -1863
and 1420 pg/mL. However, the values at 1600 and 1800 in
the 200 mg/kg group are only 1913 and 1356 pg/mL. The
average of the 1100 and 1440 hour and 1600 and 1800 hour
in the 200 mg/kg groups are essentially the same -1641 and
1634 pg/mL
Appendix Table 20 displays the baseline values, peak LH
values and % increase of peak values over baseline. Examination
of this table shows that at the high-dose group of 29.4 mg/kg/day
there is clearly a decrease in the strength of the LH surge. At this
dose level the surge does not seem to be occurring at all.
124
-------
DRAFT: DO NOT CITE OR QUOTE
Repeat Bleed. Baseline values for plasma LH are
similar among controls and all dose groups. There is a
statistically-significant decrease at 1600,1800, 2000, and
2300 in the 29.4 mg/kg/group compared to controls. The 50
ppm (3.65 mg/kg/day) group had a decrease at 1800 (25%)
compared to controls. Compared to its own baseline values,
the LH values in the 29.4 mg/kg group were not altered.
Values at 1600 and 1800 are actually slightly lower than
baseline values. Appendix Table 20 displays the baseline
values, peak LH values and % increase of peak values over
baseline. Appendix Figure 1 displays a line graph of the
results from the repeat bleed group of this study.
9.2.7.6 Summary of the Plasma LH Measurements from
the Six-Month Study
An attenuation of the LH surge at the high dose of 400 ppm
(29.4 mg/kg/day) is clear. Examination of the data from Appendix
Table 20 shows that plasma LH values for both the repeat and
non-repeat bleeds at this dose remain essentially flat over the six
timepoints.
The other dose groups do not appear to be as affected by
atrazine exposure. The non-repeat bleed data for the dose groups
is very similar to controls. There is a decrease in the strength of
the LH surge at the mid-dose of 50 ppm (3.65 mg/kg/day), but the
magnitude of this decrease is not large and given the variability
inherent in this assay (as indicated by the large standard
deviations), it is difficult to draw firm conclusions based on this
decrease.
125
-------
DRAFT: DO NOT CITE OR QUOTE
9.2.8 Summary And Discussion Of The LH Surge Studies
In the non-repeat bleed set of the one-month study the baseline
values for the 200 mg/kg/day group are similar to controls. However, the
peak value in the 200 mg/kg/day group is increased only 2.5 fold - much
less than the five-fold increase seen in control peak values versus control
baseline values. The results from the repeat bleed set for the one-month
study are even more indicative of an attenuated LH surge. Baseline
values for both control and 200 mg/kg/day groups are, again, similar. The
peak LH values are much less in the 200 mg/kg group compared to the
control though. Peak control LH values in the repeat bleed set were
increased approximately 3.5-fold over baseline. Peak values in the 200
mg/kg group were increased only about 45% over baseline values.
The results seen the six-month study in the 29.4 mg/kg/day group
also indicate an attenuated LH surge. Both the repeat and non-repeat
data sets show plasma LH levels that are flat overtime. Control LH
values at their peak are about 67% above baseline values for the non
repeat bleed data set and about 226% above baseline values in the
repeat bleed set. Peak values in the 29.4 mg/kg group are 17%
(non-repeat bleed) and 20% (repeat bleed) lower than baseline values
though.
126
-------
DRAFT: DO NOT CITE OR QUOTE
A major factor to consider when drawing conclusions from this
study is, as indicated above, the large standard deviations in the data.
Examination of the data shown in Tables 20 and 21 shows that the
standard deviations are quite large and, in fact, sometimes exceed the
means. An extreme amount of variability is to be expected with this type
of data. Variability can be expected between rats in both timing and
magnitude of the LH surge. Although the animals were synchronized by
photoperiod and by OVX, there will still be variability in timing of the surge
among rats in a dose group. It is hoped that all the rats will have their
plasma LH levels be at their peak at 1800 hours when they are sacrificed
and blood is collected. Clearly, however, this will not be the case. Some
rats will experience a peak LH surge prior to 1800 hours and some after.
The magnitude of the LH surge peak will also vary among rats. This is
due largely to the differential rate at which the animals reproductive
systems age. The variability of reproductive aging among female rats in a
particular strain has been well described (Cooper et a/., 1986; Lu et a/.,
1994; LeFevere and McClintock, 1988). Because the LH surge is
attenuated as part of the female SD rats reproductive aging process, the
variability in rate of reproductive aging means that animals of the same
chronological age will have LH surges of varying magnitude.
Even given the variability inherent in the LH measurements in this
type of study, there are some conclusions that can be reached with
confidence. There is little doubt that in the one-month study at 200
mg/kg/day and in the six-month study at 29.4 mg/kg/day, there is an
attenuation of the LH surge. Appendix Figure 1 displays a line graph of
the mean plasma LH levels in the repeat bleed group from the six-month
study displayed without standard deviation or standard error bars. The
decreased LH surge at 3.65 mg/kg/day in the six-month study is less
apparent but may be considered biologically-significant.
127
-------
DRAFT: DO NOT CITE OR QUOTE
9.3 The Site of Action for Atrazine Attenuation of the LH Surge
Experiments have been conducted examining the mechanisms underlying
the attenuation of the LH surge produced by atrazine. The focus of these
experiments has been the target site of action in the brain for atrazine. The LH
for the proestrus afternoon LH surge comes from the anterior pituitary and the
release of LH from the pituitary is controlled by gonadotropin releasing hormone
(GnRH), which is produced in the hypothalamus. Thus, atrazine could be
altering the LH surge by acting at either the hypothalamus (and affecting GnRH -
the signal to release LH) or by acting at the pituitary and directly affecting its
ability to secrete LH.
Experiments examining the effects of atrazine exposure on the
hypothalamus and pituitary were conducted at the Reproductive Toxicology
Division of National Health and Environmental Effects Research Laboratories
(Cooper et a/., 1998, Cooper et a/., 2000, Das et a/., 1999, Das et a/.,
submitted). These studies, through the results of both in vivo and in vitro
experiments, demonstrate that atrazine appears to be attenuating the LH surge
by acting on the hypothalamus, rather than directly affecting the pituitary.
An in vivo experiment using the Long-Evans (LE) strain of rat,
administered GnRH through a cardiac catheter to OVX atrazine-treated females
to see if GnRH exposure could reverse the atrazine-induced attenuation of the
LH surge. Females, in this study, given atrazine only showed an attenuation of
the LH surge - which was expected. Females given atrazine plus 50 ng/rat of
GnRH, did not display an attenuated LH surge. This provides evidence that
atrazine is affecting the ability of the hypothalamus to release GnRH.
An in vitro experiment using perfused anterior pituitaries removed from
untreated female LE, showed that atrazine could not directly affect the ability of
the anterior pituitary to secrete LH. Pituitaries perfused in vitro were able to
produce an LH surge when primed with estradiol. Adding 100 uM atrazine to the
perfusion system did not affect the ability of the pituitaries to produce an LH
surge following estradiol priming.
128
-------
DRAFT: DO NOT CITE OR QUOTE
These two experiments taken together provide evidence that atrazine
inhibits the proestrus afternoon LH surge through an action on the hypothalamus
rather than a direct action on the anterior pituitary. Specifically, it appears that
atrazine may somehow inhibit the hypothalamic secretion of GnRH.
Data from the Cooper lab indicates that a decrease in hypothalamic
norepinephrine levels may be responsible for the reduced capacity of the
hypothalamus to secrete GnRH (Cooper ef a/., 1999a). In these studies,
exposure of LE rats to a three-day exposure of 50,100,200 and 300 mg/mUday
of atrazine resulted in significant depressions of hypothalamic norepinephrine
levels at all dose levels. This study is supported by in vitro studies using
triazines and PC12 cells that showed that exposure through the medium of 50,
100 and 200 uM atrazine resulted in dramatic decreases in norepinephrine
release at all dose levels starting as early as six hours following the start of
exposure and continuing for up to 48 hours following exposure (Das, ef a/.,
1999).
Disruption of GABAergic neurotransmission by atrazine may also play a
role in the decrease in GnRH release seen following atrazine exposure.
Gamma-aminobutyric acid type a receptors (GABAJ are known to play a crucial
role in GnRH release. In vitro, atrazine (and cyanazine) have been shown to
disrupt agonist binding to the GABAA receptor in cortex from male Long-Evans
rats (Schafer, ef a/., 1999). Such disruption could contribute to the decreased
hypothalmic GnRH release seen following atrazine exposure.
129
-------
DRAFT: DO NOT CITE OR QUOTE
9.4 The Data Examining the Association Between Atrazine Exposure and
An Attenuated Proestrus Afternoon LH Surge, Increased Days and
Estrus and a Prolonged Exposure to an Elevated Level of Estradiol
This document presents several studies examining an association
between atrazine exposure and an early onset of alterations in the above
described parameters - estrous cycle, serum estradiol levels; serum LH levels. A
study is also available examining the correlation between estrous cycles and
mammary tumor incidence/onset. The time of tumor onset is also examined in
several studies in an attempt to confirm that normally occurring events
(mammary tumor induction) are, in fact, occurring earlier following atrazine
exposure.
The results indicate that atrazine exposure does appear to result in an
early onset of these parameters. Increased days in estrus, increased serum
estradiol levels, attenuated LH surge, and onset of mammary tumors all occur
earlier in atrazine-treated females than they do in untreated females.
130
-------
DRAFT: DO NOT CITE OR QUOTE
9.4.1 Atrazine Exposure Results in an Earlier Onset of Increased
Days in Estrus
One-month of atrazine exposure at 40 mg/kg/day induced estrus
blocks. Longer term studies revealed that atrazine exposure at lower
levels also has the ability to increase days in estrus if the exposure is of
long enough duration. A dose of 3.65 mg/kg/day in a six-month study was
able to induce an increase in days spent in estrus over controls as early
as 5.5 months into the study. A dose of 4.23 mg/kg/day in a two-year
study (the low-dose tested in this particular study) induced an increased
percentage of days spent in estrus at nine months - an event that is not
seen in control SD females in this study until about 12 months into the
study. A separate two-year study showed that as little as 1.5 mg/kg/day
could induce an increase in percent days spent in estrus. This increase
was marginal and was not statistically-significant though. Furthermore,
the increase in days in estrus seen at this dose was apparently not of the
magnitude to increase a female rats risk of mammary cancer as there was
not an increase in mammary tumor incidence or decrease in mammary
tumor onset at this dose in this study. The next highest dose in the same
study (3.1 mg/kg/day) also resulted in an increase in days in estrus. The
increase in days estrus seen in this study was slightly greater than the
increase seen at 1.5 mg/kg/day, but was still not statistically-significant. It
was apparently enough to cause an increase an animals risk for
mammary cancer as mammary cancer incidences at this dose were
increased about two-fold over concurrent control values.
9.4.2 Atrazine Exposure Results in an Earlier Onset of Increased
Serum Estradiol Levels
Atrazine exposure of only three months resulted in an increase in
serum estradiol levels in SD females. Levels equivalent to those seen at
three months were not seen in control animals in this study until nine
months. Because there was no intermediate timepoint between three and
nine months in this study it is difficult to determine just when the control
animals achieved serum estradiol levels equivalent to those seen in the
dosed animals at three months. Clearly, though, the dosed animals had
higher serum estradiol at three months than the controls did - indicating
that this parameter of reproductive aging was achieved earlier in the
dosed rat than in the control rat.
131
-------
DRAFT: DO NOT CITE OR QUOTE
9.4.3 Atrazine Exposure Results in an Earlier Onset of Attenuated
LH Surges
Atrazine exposure at high levels (40 and 200 mg/kg/day) was able
to attenuate the preovulatory LH surge after only one month of exposure.
Exposure to atrazine at lower levels (29.4 mg/kg/day) significantly
weakened the LH surge after six months exposure. The lowest dose of
atrazine that was able to induce a weakening of the LH surge was 3.65
mg/kg/day, but this weakening was not statistically-significant. It is not
clear exactly when a normally aging animal would be expected to
experience an weakened LH surge, but the studies described here
showed that the control animals did not experience a weakened LH surge
when the concurrently run dosed animals did.
9.4.4 Atrazine Exposure Results in an Earlier Tumor Onset
Tumor onset times were consistently decreased following atrazine
exposure. The decrease in tumor onset times implies that the process of
tumor formation is occurring at an earlier chronological age due to
atrazine exposure. Tumor incidence rate were not always increased
following atrazine exposure- they were not increased in Thakur, 1992a
and only the trend for fibroadenomas was increased in Thakur, 1991 a.
Tumor onset times were decreased in every study in which they were
examined - Thakur, 1991a and 1992a; and Morseth, 1998.
132
-------
DRAFT: DO NOT CITE OR QUOTE
9.5 Pituitary Adenomas
9.5.1 Onset of Pituitary Alterations Following Atrazine Exposure
Table 5-5 provides evidence that the latency period of pituitary
adenomas is decreased following atrazine exposure. Pituitary adenomas
are known to be age-related in the rodent. The time to onset of other
age-related pituitary alterations may also be decreased in atrazine
exposed SD females. As described above under section 9.1, pituitary
weights and incidences of pituitary hyperplasia are also increased in the
untreated female SD rat with age. There is evidence that an early onset
of increased pituitary weights and an early onset of pituitary hyperplasia
may also be occurring in response to atrazine exposure. Appendix Tables
21 and 22 display absolute and relative (to body weight) pituitary weights
from Thakur, 1991 a. Appendix Table 23 displays absolute and relative (to
body) pituitary weight from SD females exposed to atrazine for six months
in Morseth, 1996b.
Three months of exposure did not result in an increase in either
absolute or relative- to-body pituitary weights in the dose groups
compared to controls. By nine months of exposure there was a clear
dose-related increase in both absolute and relative pituitary weights. The
effect was still evident at 12 months, but only at the high dose and the
effect was less severe at the high dose at 12 months compared to nine
months. Pituitary weights at 15,18 and 24 months are comparable in
dose groups compared to controls. The incidence of pituitary focal
hyperplasia (as recorded in the histopathology records of this study) was
marginally increased in dose groups compared to the control at nine
months. There was only one incidence of this histology finding in the
control group (it was assigned the grade "slight" by the examining
pathologist) while there were two incidences at 4.23 mg/kg/day (both
"slight") and two incidences at 26.23 mg/kg/day (one "minimal" and one
"slight"). The early onset of increased pituitary weights following atrazine
exposure seen in Thakur, 1991 a is confirmed in Morseth, 1996b. Both
absolute and relative pituitary weights increases are >20% at 29.4
mg/kg/day after six months of exposure.
133
-------
DRAFT: DO NOT CITE OR QUOTE
9.5.2 Role of Early Onset of Pituitary Alterations in Mammary
Carcinogenesis
The pituitary alterations - adenomas, hyperplasia and increased
pituitary weight - all result in increased serum prolactin levels (Baird ef a/.,
1990; McComb ef a/., 1986; van Putten ef a/., 1988). The association
between increased prolactin exposure and mammary tumors has been
well-described (Meites, 1972; Meites, 1981; Russo ef a/., 1990).
Likewise, an association between pituitary alterations of the types
mentioned above and mammary tumors had been well-described
(Blankenstein ef a/., 1984; McConnell, 1989a; Goya ef a/., 1990). The
studies described in this document also provide evidence of an
association between these pituitary alterations and mammary tumors.
Appendix Table 24 shows that majority of SD females with mammary
tumors also had pituitary adenomas. Females without mammary tumors
also had high incidences of pituitary adenomas, but the incidences were
generally lower than for animals with mammary tumors. Appendix Tables
25a and b show that pituitary weights in females with mammary tumors
were higher than pituitary weights in females without mammary tumors.
The mean absolute pituitary weight in animals with mammary tumors in
Morseth, 1998 was 46% greater than the mean pituitary weight in animals
that did not have mammary tumors. The difference was even more
apparent in the Thakur, 1992a study where the mean absolute pituitary
weight in females with mammary tumors was 94% greater than in those
without mammary tumors.
9.5.3 Pathogenesis of Pituitary Alterations
As has been previously discussed, estrogen is mitogenic to the
pituitary lactotrophs of the rodent. It is therefore biologically plausible that
the same increase in anovulation and accompanying prolonged exposure
to serum estrogens that is believed to contribute to mammary
carcinogenesis following atrazine exposure would also contribute to
pituitary hyperplasia and neoplasia.
If an increase in serum estrogens leads to the afore-mentioned
pituitary alterations then it would be reasonable to expect that increases in
serum estrogen would precede the pituitary alterations. Indeed, increases
in serum estradiol are seen as early as three months following the
134
-------
DRAFT: DO NOT CITE OR QUOTE
initiation of atrazine exposure (Appendix Table 15). Dose-related
increases in acinar development (a histomorphologic alteration highly
dependent on estrogen) are found following three months of exposure
while dose-related increases of galactoceles, secretory activity and other
histomorphologic alterations indicating prolactin exposure are not seen
until nine months of exposure (Appendix Tables 16 and 17).
If estrogens derived from unovulated follicles contribute to pituitary
alterations then OVX animals would be expected to have lower incidences
of these pituitary alterations. The OVX animals in Morseth, 1998 did have
lower incidences of pituitary adenomas, but not as much less as might be
expected. As previously noted, ovariectomy was able to reduce
mammary tumor incidences from about 50% having some sort of
mammary tumor to zero percent having any sort of mammary tumor. The
effect of ovariectomy on the pituitary adenoma rate was much less
pronounced. Ovariectomy dropped the pituitary adenoma rate from about
70% at terminal sacrifice for the intact animals to about 50% at terminal
sacrifice in the OVX animals. Interestingly, though ovariectomy had only
mild impact on pituitary tumor incidences, ovariectomy was able to
dramatically reduce pituitary weights at the end of the study. Absolute
pituitary weights at terminal sacrifice in the OVX animals were only about
a quarter the value of the intact animals. Relative pituitary weights were
only 35% the weight in OVX compared to intact animals. Appendix Table
26 displays these data. The dramatic decrease in pituitary weight in
conjunction with only a mild decrease in pituitary adenoma incidence may
be explained by an earlier onset of the pituitary tumors in the intact
animals compared to the OVX. The data in Appendix Table 26 provides
evidence of an early onset of pituitary adenomas in intact animals versus
OVX when it shows a 6% incidence of pituitary adenomas in the interim
sacrifice OVX animals compared to a 17% incidence in interim sacrifice
intact animals. The data in Appendix Table 26 also shows that OVX
animals had a much reduced incidence of "enlarged" pituitaries compared
to intact animals. The decreased incidence of enlarged pituitaries and the
reduced weight of the pituitaries may indicate that, though many OVX
animals still got pituitary tumors, these tumors occurred later in life and
thus, by the time the animals were sacrificed, had not had as much time to
grow and were thus of a smaller size.
135
-------
DRAFT: DO NOT CITE OR QUOTE
A serial sacrifice study comparing OVX and intact animals would be
useful in determining if pituitary tumor onset is delayed in OVX animals
versus intact animals. While HED is not interested in pituitary tumor onset
in OVX animals perse; the fact that pituitary tumors are not more
dramatically decreased in OVX animals raises some doubts about a mode
of action for atrazine-mediated pituitary tumors that depends on prolonged
exposure to follicular-derived estrogen. Were the same mode of action to
apply to pituitary tumors that applies to mammary tumors then one would
expect pituitary tumors to behave like mammary tumors when the ovaries
are removed. That is, one would expect pituitary tumor rates to drop to
zero, or close to zero, following OVX.
9.5.4 Summary and Conclusion for Pituitary Alterations
There is ample evidence in the open literature that exposure to
follicular-derived estrogen in CE rats leads to an increased incidence of
prolactin-secreting pituitary adenomas and increased pituitary weight or
focal hyperplasia. Being that atrazine exposure seems to result in an
early onset of constant estrus and increased estradiol exposure, one
would expect that these pituitary alterations would also show an early
onset following atrazine exposure. Pituitary weight data from Thakur,
1991 a and Morseth, 1996 both show an early onset of increased pituitary
weight following atrazine exposure. Pituitary adenoma data from Thakur,
1991 a shows that there is an early onset of pituitary adenomas following
atrazine exposure.
Further research is desirable into why ovariectomy does not reduce
pituitary tumor incidence to the same extent as it does mammary tumors
incidence despite there apparently having the mode of action.
136
-------
PartC
Hazard Assessment and Review of
Available Studies
-------
List of Contents
10. Possible Effects of Atrazine Exposure on Testosterone Metabolism 3
11. Epidemiology 4
11.1 IUGR in Iowa Communities 4
11.2 Birth defects in Rural Minnesota 6
11.3 Male Pesticide Exposure and Pregnancy Outcome 7
11.4 Summary and Conclusions 8
12. Background for Pregnancy, Pubertal and Prostatitis Papers 8
12.1 Role of Prolactin and LH in Pregnancy 8
12.2 Role of Prolactin and LH in Pubertal Development 9
12.2.1 Female 9
12.2.2 Male 10
12.3 Role of Prolactin in Prostatitis 11
13. Data 12
13.1 Atrazine Effects on Prolactin 12
13.2 Atrazine Effects on Pregnancy 13
13.2.1 Implantation and Early Pregnancy 13
13.2.2 Pregnancy Maintenance: Strain Comparisons of Sensitivity to
Atrazine-lnduced Pregnancy Loss in Rats 14
13.3 Atrazine Effects on Pubertal Development 15
13.3.1 Female 15
13.3.2 Male 17
13.4 Atrazine Effects on Prostate 19
13.5 Summary/Conclusion 20
-------
Part B of this document described a variety of experiments demonstrating an
effect of atrazine on the pituitary-hypothalamic-ovarian axis of the rodent. These
alterations would be expected to result in the establishment of a hormonal environment
conducive to the development of mammary and pituitary tumors (i.e., prolonged
exposure to endogenous serum estrogens and serum prolactin).These neuroendocrine
perturbations might also be expected to have effects other than tumorgenicity.
The suppression of LH release from the pituitary (described in Part B, section
9.2.7) might be expected to affect pregnancy in the rat. Alterations in LH release from
the anterior pituitary might also be expected to affect pubertal development. Atrazine
exposure has also been shown to alter serum prolactin levels following acute, high-
dose exposure (Cooper etal., 2000; these data are not discussed in Part B).
Alterations of serum prolactin could affect pregnancy and pubertal development as well
as having other potentially adverse effects - specifically an increased risk of prostate
inflammation in adults male rats whose mothers were exposed to atrazine while
nursing.
Data describing atrazine effects on serum prolactin and possible effects of LH
suppression and serum prolactin alterations on pregnancy, pubertal development, and
prostatitis are described below. Most of these data are derived from the labs of Dr.
Ralph Cooper at EPA. Also described is a study by Dr. Barry Zirkin of Johns Hopkins
University examining atrazine effects on pubertal development in male rats.
Understanding the data described below requires some background knowledge of the
role LH and prolactin play in pregnancy and pubertal development in the rodent.
Brief backgrounds of the roles of prolactin and LH in pregnancy and pubertal
development, followed by a discussion of the data showing atrazine's effects on these
parameters, constitute the bulk of this chapter. The remainder of Part C is a brief
discussion of a trio of open literature epidemiology studies that investigate potential
associations of reproductive anomalies with atrazine exposure, and a discussion of a
trio of studies from the lab of Dr. Jasna Kniwald in Yugoslavia that examine possible
effects of atrazine on testosterone metabolism in the male rat.
The discussion of the testosterone metabolism studies immediately follows. A
discussion of the epidemiology studies is next followed by a discussion of atrazine's
effects on pregnancy, puberty and prostatitis.
-------
10. Possible Effects of Atrazine Exposure on Testosterone Metabolism
The steroid hormone testosterone is converted into Sot-reduced metabolites by
tissues in vivo. These Sot-reduced metabolites are primarily 5a-dihydrotestostrone (DHT),
Sa-diol and androstenedione. Tissues such as the prostate, seminal vesicles,
hypothalamus and pituitary contain 5ct-reductase enzymes and are primarily responsible
for reducing testosterone to these compounds. The enzymes responsible for these
conversions are 5ot-reductase, 3a-HSD and 170-HSD. The Sot-reduced metabolites
(particularly DHT) are able to bind the testosterone receptor, and testosterone receptor-
5a-reduced metabolite complexes are believed to be more stable than testosterone-
receptor complexes. The rat prostate also has receptors specific for DHT.
Three published papers have examined the effect of atrazine on testosterone
metabolism. The first of these papers examined the effects of atrazine exposure with
either in utero or in utero plus early postnatal exposure (Kniewald et a/., 1987). Female
Fischer-344 rats were treated by s.c. injection once a day during their entire pregnancy
with 16.6 mg/kg/day of atrazine or deethyatrazine (DET) and sacrificed at PND 28; or
were treated once daily by s.c. injection with the same dose of both compounds during
both pregnancy and lactation and were sacrificed at 21 days postnatal. Activity of
pituitary reductase enzymes, and DHT and estradiol receptor binding sites in the
prostate and uterus were determined in all animals.
A second study exposed young adult male Fischer-344 rats to either 60 or 120
mg/kg/day atrazine or DET by gavage for seven days (Babic-Gojmerac ef a/., 1989).
On the eighth day animals were sacrificed, and the anterior pituitary and hypothalamus
were excised and activity of reductase enzymes responsible for testosterone
metabolism were measured. In vitro studies were also conducted in which the anterior
pituitary and hypothalamus were removed from young adult males and exposed to
atrazine and DET in vitro. The activity of reductase enzymes was then determined.
The third study also used young-adult male Fischer-344 rats and dosed them by
gavage with 120 mg/kg/day of atrazine for seven days (Simic et a/., 1991). On the
eighth day the animals were sacrificed and DHT-receptor complexes in the prostate
were measured.
-------
These studies indicate that testosterone metabolism is not greatly altered if
animals are treated only in utero. Testosterone metabolism can be significantly altered
if male and females are treated in utero and during lactation or if males are treated as
young adults. The studies indicate that atrazine and DET are able to inhibit the
conversion of testosterone to its more active reduced forms. The inhibition of
metabolism of testosterone to its more active forms by atrazine indicates an anti-
androgenic effect of atrazine which may result in adverse consequences.
11. Epidemiology
11.1 IUGR in Iowa Communities
The incidence of intrauterine growth retardation (IUGR) in Iowa was
investigated in an ecologic study (Munger et a/., 1997). The definition of IUGR
can vary considerably (Seeds, 1984). IUGR can, generally speaking, be defined
as low birth weight. Infants that are delivered full-term are usually defined as
displaying IUGR if they weigh less than 2,500 grams at birth. For infants that are
delivered prior to full term, birth weight below the tenth percentile for gestational
age is frequently used to define IUGR.
IUGR was defined in the Munger study as birth weight below the 10th
percentile for gestational age as defined by California standards for non-Hispanic
whites. The Munger study was a an ecologic study in which estimated
exposures to pesticides through drinking water were compared to birth weights in
an attempt to determine associations between the two. Data on levels of
pesticide contamination in drinking water were obtained from the 1986 to 1987
statewide municipal water survey of Iowa which included data from 856
municipal water sources across the state. Birth weight data were obtained from
birth certificate data obtained from the Iowa Department of Public Health.
Factors such as maternal smoking, maternal education, quality of prenatal care,
geographic region and community size were evaluated in an attempt to account
for potentially confounding data.
-------
An association between IUGR and maternal exposure to drinking water
from the Rathburn system was found. The Rathburn system is a community
drinking water system that obtains its water from the Rathburn reservoir in
southern Iowa and serves several communities and counties in southern Iowa.
The 1986 to 1987 drinking water survey found the Rathburn system to have
elevated levels of herbicide contamination with the most notable contaminant
being atrazine. The mean atrazine contaminant level in the Rathbum system 2.2
ug/L compared to a mean atrazine contaminant level of 0.6 ug/L in all other Iowa
surface water suppliers. The rate of IUGR incidence from 1984 to 1990 in
communities served by the Rathburn system was 11.2% compared to 6.4% for
all other surface water suppliers in Iowa.
Regression models of IUGR showed that atrazine had the best fit (i.e., the
best positive association between drinking water contaminant levels and IUGR
incidence) of all the contaminants examined in the study. The study authors
note: "Atrazine had the best fit in the regression models of IUGR, but
independent effects of other herbicides, which are intercorrelated, cannot be
ruled out."
Although useful, the Munger study should be, because it is an ecologic
study, regarded as a preliminary study which needs to be verified by more
detailed epidemiologic studies. The deficiencies of ecologic studies are
discussed in Part B section 4.9 and the study authors of the Munger study state:
"Because of the limitations of the ecologic design of this
study, including aggregate rather than individual measures
of exposure and limited ability to control for confounding
factors related to source of drinking water and risk of IUGR,
a causal relationship between any specific water
contaminant and risk of IUGR cannot be inferred."
6
-------
11.2 Birth defects in Rural Minnesota
A study published in the open literature examined the possibility that
offspring of pesticide appliers may display higher incidences of birth defects
(Garry etal., 1996). This study used information from the Minnesota Department
of Agriculture (MDA) to identify persons in Minnesota who were certified to apply
restricted use pesticides in 1991. This applicator information was linked to birth
data supplied by the Minnesota Center for Health Statistics of the Minnesota
Department of Health to determine birth defect rates among pesticide applicators
in Minnesota. The birth data were also compared to pesticide use data supplied
by MDA to examine associations between birth defects and quantitative pesticide
use for both pesticide applicators and the general population.
Minnesota pesticide applicators did have significantly (p<0.001) more
children born with a birth anomaly than did the general population of Minnesota.
However, when birth data from pesticide applicators from one of three crop
growing regions are compared to birth data from the general population from the
same region, a statistically-significant difference (p<0.02) is seen only in one of
the three regions. Use of any specific pesticide was not implicated in the
increases in birth defects seen in pesticide appliers vs. the general population.
When use of specific pesticides was examined by comparing pesticide
use by county cluster to birth defects in that cluster, the authors noted that use of
chlorophenoxy herbicide and fungicides seemed to have the strongest
association with birth defects. The results for atrazine use did not show a clear
association of atrazine use with birth anomalies. There was a significant
increase in birth defects when counties with atrazine use leveEs of > 100,000 IDS
Al/county cluster are compared to county clusters with < 100,000 Ibs Al, but if
clusters with use levels of >25,000 Ibs Al are compared to clusters with <25,000
Ibs Al/cluster then a significant association is not seen.
-------
11.3 Male Pesticide Exposure and Pregnancy Outcome
To examine the potential effects of paternal pesticide exposure on
pregnancy outcome, farm activity in male fanners in Ontario, Canada was
evaluated for the three months prior to conception through conception (Savitz et
a/., 1997). Male farmers were surveyed about their farming activities for the past
five years to obtain information about their activities during the above-described
time around conception. Pregnancy outcomes were determined by
questionnaires completed by the farm couples. Farm activities by the males
were compared to pregnancy outcome to determine potential relationships of
farm activities (specifically mixing, loading or applying pesticides) to pregnancy
outcomes (specifically the risk of preterm delivery, miscarriage, small for
gestational age [SGA] and sex ratio). If the males had used pesticides in the
time period around conception, then the specific pesticides used were reported.
An increased risk of miscarriage was not associated with atrazine use as
a crop herbicide (adjusted odds ratio [AOR] 1.5) or atrazine use as a yard
herbicide (AOR = 1.2). An increased risk of SGA was not associated with
atrazine or cyanazine use as a crop herbicide (AOR 0.5 and 0.8 respectively) or
with atrazine use as a yard herbicide (AOR= 0.5). Sex ratio data were not
separated out into exposures to specific chemicals but, since the AORs for
proportion of male births for all farm activities involving chemicals ranged from
0.8 to 1.1, it is apparent that there was not an increased risk of alterations in sex
ratio associated with use of any pesticide in males.
There was, however, an increased risk of preterm delivery associated with
atrazine use by males around the time of conception. The AOR for use of
atrazine as a crop herbicide was 2.4 (95% Cl 0.8 to 7.0) and the AOR for use of
atrazine as a yard herbicide was 4.9 (95% Cl 1.6-15).
8
-------
11.4 Summary and Conclusions
An association of exposure to atrazine-contaminated water and IUGR was
seen, but because of the limitations of this study (i.e., an ecologic study) the
study authors conclude that the results are, "a preliminary finding that needs to
be verified by more detailed epidemiologic studies." An association of atrazine
use and birth defects in Minnesota communities was also seen, but the
association was inconsistent. The inconsistencies in the data weaken the
positive finding seen and the study authors do not dwell on the positive finding
noted for atrazine. Male atrazine use around the time of conception failed to
show an association for three of the four reproductive parameters examined
(miscarriage, SGA and sex ratio). A positive association was seen for preterm
delivery though. There is little biologic plausibility in associating a parameter
such as preterm delivery with male chemical exposure. The effect of male
chemical exposures on this endpoint has not been extensively studied and, as
the study authors note: "Maternal characteristics, particularly reproductive and
medical, are most strongly associated with preterm delivery."
The data provided by these three epidemiology studies do not provide
clear evidence of an association between atrazine exposure and reproductive
anomalies.
12. Background for Pregnancy, Pubertal and Prostatitis Papers
12.1 Role of Prolactin and LH in Pregnancy
Progesterone, acting at the uterus, is essential to maintain pregnancies in
mammals. The major source of this vital progesterone in all mammals is the
corpus luteum (CL). All mammals require progesterone throughout pregnancy
from fertilization to parturition.
The sole source of progesterone during pregnancy in the rat is the CL.
During early pregnancy (from implantation to approximately GD seven in the rat)
the CL is maintained by prolactin derived from the anterior pituitary (Terkel,
1988). During mid-gestation (from GD 7 to 10) the CL is maintained by
lutenizing hormone (LH) (Rothchild, 1981; Terkel, 1988). After GD 10, prolactin-
like compounds produced by the placenta (i.e., placental lactogens) function to
maintain the CL throughout the remainder of the pregnancy (Gibori, era/., 1988;
Linznerand Fisher, 1999).
-------
The source of progesterone in pregnancy in humans is the CL for about
the first six weeks of pregnancy. At approximately the sixth week of pregnancy,
humans display a "luteal-placental shift" in which the placenta assumes the
responsibility of progesterone secretion and the CL becomes quiescent (Stouffer,
efa/.,1989).
Thus, all mammals require progesterone throughout pregnancy, but in the
rat the only source of this progesterone is the CL whereas in the humans the CL
is the initial source, and the placenta supplies progesterone for the greater part
of the pregnancy. Maintenance of the CL in the rat is accomplished by prolactin,
LH and placental lactogens. Pituitary-derived prolactin plays the primary role in
maintaining the CL in early pregnancy; LH is primarily responsible for
maintaining the CL in mid-gestation; and, placental lactogens maintain the CL
during late gestation.
12.2 Role of Prolactin and LH in Pubertal Development
12.2.1 Female
Pubertal development in the female rat has been well-
characterized (Ojeda, 1980; Ojeda, 1983). The onset of puberty in the
female is a transitional period that culminates with the initiation of cyclic
surges of luteinizing hormone (LH) from the pituitary that stimulate
ovulation. Vaginal opening generally coincides with the first ovulation and
occurs at 32 or 33 days of age in the female rat. The hormonal changes
which induce the first ovutation are similar in many respects to the
hormonal changes which induce all other ovulations in rodents. The
sequence of hormonal changes preceding the first ovulation is as follows:
1. Serum estradiol levels increase followed by;
2. A dramatic increase (surge) in serum luteinizing hormone
(LH);
3. Serum prolactin levels dramatically increase concomitant
with the LH surge.
10
-------
Exposure to atrazine has been shown to attenuate the proestrus
LH and prolactin surges in Long-Evans and Sprague-Dawley rats. Since
both of these hormones are important for normal pubertal development, it
is reasonable to hypothesize that atrazine may affect the onset of puberty
in the female rodent. Atrazine's attenuation of the proestrus LH surge is
described in detail in Part B sections 9.2.7 and 9.2.8. Atrazine's
attenuation of prolactin release is described below In section 13.1.
In addition, reports that atrazine can reduce hypothalamic
norepinephrine concentrations (Cooper et a/., 1998) and that intravenous
injections of GnRH restore the estrogen-induced secretion of LH in
ovariectomized, atrazine-treated female rats (Cooper et a/., 2000) suggest
that possible effects on neurotransmitters and their regulation of pituitary
hormone synthesis/secretion could also alter the onset of puberty.
Thus, to examine the effects of atrazine on female pubertal
development, a study was conducted using the "Research Protocol for the
Assessment of Pubertal Development and Thyroid Function in Juvenile
Female Rats" (U.S. EPA, 1998b; Goldman ef a/., 2000).
12.2.2 Male
The onset of puberty in the male rat involves a complex interplay of
several hormones including LH, FSH, testosterone and prolactin (Nazian
and Mahesh, 1980; Piacsek and Goodspeed, 1978). It has been shown
that an increased turnover rate in hypothalamic GnRH, NE and DA
precedes the dramatic increase in testosterone (Matsumoto et a/., 1986)
prior to the onset of puberty. LH stimulates testosterone secretion by the
Leydig cells. At the same time, LH secretion varies only slightly as
puberty approaches. However, there is an increased sensitivity of the
testes to LH prior to puberty, due to other hormonal influences, such as
increased prolactin secretion, that facilitate an upregulation of LH
receptors (Kamberi ef a/., 1980; Odell era/., 1973; Vihko era/., 1991). In
contrast, there is a higher threshold for the gonadotropin/gonadal steroid
feedback mechanism in the adult male (Gupta et at., 1975; Nazian and
Mahesh, 1980) as compared to the immature male, making the immature
male more sensitive to the feedback of testosterone. As this feedback
sensitivity decreases, the hypothalamic-pituitary unit becomes more
effective at stimulating testicular development, because there is less
inhibition of gonadotropins by testosterone.
11
-------
Development of the size of the penis and cornification of the
epithelium of the prepuce and preputial separation in immature rats are
regulated by androgens (Marshall, 1966). A decrease in testosterone
during the juvenile period can delay preputial separation (Lyons et a/.,
1942) and reduce the size of the androgen-dependent tissues, such as
the ventral prostate and seminal vesicles. Normally, testosterone levels
rise gradually from PND 20 to 40, and abruptly double by PND 50
(Matsumoto era/., 1986; Monosson et a/., 1999). Atrazine exposure has
been shown to alter LH and prolactin secretion in female rats. An effect
on LH and prolactin secretion in immature male rats, and thus on pubertal
onset, may also be possible.
To examine the effects of atrazine on male pubertal development,
a study was conducted using the "Research Protocol for the Assessment
of Pubertal Development and Thyroid function in Juvenile Male Rats"
(U.S. EPA, 1998b).
12.3 Role of Prolactin in Prostatitis
Hyperprolactinemia prior to puberty in male rats has been shown to lead
to lateral prostate inflammation in young adult rats (Stoker era/., 2000b). One
possible cause of hyperprolactinemia in immature male rats is a deficiency in
milk-derived prolactin. Milk-derived prolactin plays a critical role in the
development of the tuberoinfundibular dopaminergic neurons (TIDA) of the
hypothalamus of a developing rat (Shyr, et a/., 1986). The TIDA neurons
function to inhibit prolactin secretion from the anterior pituitary. Organization and
development of these neurons occurs mainly during the first postnatal week in
the rat (Ojeda and McCann, 1974).
Thus, if developing rats do not receive a sufficient amount of prolactin
from their mothers milk during the first week after birth, the TIDA neurons will not
develop properly and may not be able to sufficiently provide an inhibitory check
to prolactin secretion in the adult animal. The resultant hyperprolactinemia is
associated with development of prostatitis in the adult.
12
-------
13. Data
13.1 Atrazine Effects on Prolactin
Similar to the preovulatory LH surge that is described in Part B section
9.1.1. of this document, rodents also display a preovulatory prolactin surge
(Blank, 1986). Studies demonstrating atrazine-associated effects on this
prolactin surge are presented in Cooper, et a/., 2000 (other data from this
publication relating to the site of action of atrazine-associated LH alterations are
described in Part B section 9.3 of this document).
The prolactin studies described in Cooper et a/., 2000, use adult Long-
Evans (LE) and Sprague-Dawley (SD) female rats which were ovariectomized
(OVX) and given estrogen-containing implants. Three days later they were
dosed by gavage with a single dose of 97.1% atrazine suspended in
carboxymethylcellulose at dose levels of 0, 50,100, 200, or 300 mg/kg (this
protocol of OVX, estradiol implantation, and sacrifice three days later, is an
established model of the LH and prolactin surges and was also used in the LH
surge experiments described in Part B section 9.2.7). Separate groups of LE
and SD females were OVX, implanted with estradiol pellets, and given daily
doses at the same dose levels as the single-exposure animals for the three days
leading up to sacrifice. Lastly, other LE and SD females were OVX, dosed daily
for 21 days at dose levels of 0, 75,150 or 300 mg/kg/day, implanted with
estrogen pellets on day 21 and then sacrificed three days later.
Thus, atrazine exposures consisted of: single exposures; three-day
exposures; and,21-day exposures. Following sacrifice, blood was collected and
serum prolactin was measured.
A single exposure in the SD females resulted in no statistically-significant
differences between controls and any of the dose groups. The LE rats showed
statistically-significant decreases in serum prolactin levels in the high-dose group
of 300 mg/kg only.
Three days of atrazine exposure resulted in statistically-significant
attenuation of the prolactin surge in SD females at 300 mg/kg/day, but not at any
other dose. The prolactin surge in LE rats under this exposure protocol was
significantly attenuated at 100,200 and 300 mg/kg/day and was delayed at 50
mg/kg/day.
13
-------
The 21-day exposure resulted in significantly lower prolactin levels in both
strain at the 150 and 300 mg/kg/day doses, but at 75 mg/kg/day prolactin levels
were not significantly different from control levels.
These data clearly demonstrate that atrazine has the ability to suppress
estrogen-induced prolactin surges. Further data described in this paper
demonstrate that, as was the case for the atrazine-associated LH surge
attenuation, the effect of atrazine on prolactin secretion does not appear to be
due to a direct effect on the pituitary. Rather, the hypothalamus appears to be
the site of action for this effect of atrazine exposure.
13.2 Atrazine Effects on Pregnancy
13.2.1 Implantation and Early Pregnancy
Cummings et a/., 2000 (submitted) examined the effects of atrazine
on implantation and early pregnancy in several strains of rats. Technical
grade atrazine (ATR) of 97.1% purity was administered daily by gavage to
rats during GD 1 to 8 (day 0 = sperm +). Dose levels included 0, 50,100,
and 200 mg/kg/day of ATR. Rats were divided into groups such that half
were dosed at 2 p.m. (just prior to the diurnal prolactin surge of early
pregnancy) and half were dosed at 2 a.m. (just prior to the nocturnal surge
of prolactin). Within each time interval group, four strains of rats were
each tested at each of the four dose levels listed above. Rat strains used
were Holtzman, Fischer-344, Sprague-Dawley, and Long-Evans hooded.
Clinical signs of toxicity consisted of decreased mean body weight at
necropsy in the 200 mg/kg groups. Necropsies were performed on GD 9
of pregnancy. A small but significant decline in mean number of
implantation sites was seen at 100 mg/kg in Fischer-344 rats (nocturnal
dosing interval) and Sprague-Dawley rats (diurnal dosing interval), as well
as at 200 mg/kg in Holtzman rats (nocturnal dosing). Holtzman rats alone
showed both an increase in resorptions and a decrease in serum
progesterone (at 200 mg/kg) as well as a decrease in serum LH at the
same dose. Long-Evans and Fischer-344 rats also exhibited a decrease
in serum LH in the 200 mg/kg dose group.
14
-------
Summary/Conclusion
The LOAEL is 100 mg/kg for the effect on implantation and 200
mg/kg for all other parameters.
The NOEL is 50 mg/kg for the effect on implantation and 100
mg/kg for all other parameters.
13.2.2 Pregnancy Maintenance: Strain Comparisons of Sensitivity to
Atrazine-lnduced Pregnancy Loss in Rats
In a series of developmental toxicity studies (Narotsky et a/., 1999,
Narotsky et a/., submitted) technical grade atrazine (97.1%) was
administered by gavage, in 1% methylcellulose, to F344, Sprague-
Dawley, and Long-Evans hooded rats at 0, 25, 50,100, or 200 mg/kg/day
on GD 6 to 10. This time frame was selected because it coincides with
the LH-dependent period of pregnancy. In preliminary work, the authors
identified this period of pregnancy as the most sensitive to the effects of
atrazine. Using 200 mg/kg atrazine (by gavage), they found full-litter
resorptions in 20 of 30 dams dosed from GD 6 to 10 while the same
treatment was without effect in nine dams dosed from GD 11 to 15.
Based on this information, the following study compared potential strain
differences in response to atrazine. The dams were allowed to deliver
and their litters were examined on PND's one and six. The F344 strain
was the most sensitive to atrazine's effects on pregnancy maintenance;
the Long-Evans strain was the least sensitive. In the F344 rats, maternal
toxicity (weight loss, piloerection) and developmental toxicity (full-litter
resorption, i.e., pregnancy loss) were observed at ;>50 mg/kg. Among
surviving litters, increased prenatal mortality was observed at 200 mg/kg,
and parturition was delayed at 100 mg/kg. In Sprague-Dawley rats,
similar effects were observed, albeit at different dose levels; maternal
weight loss was noted at *25 mg/kg, full-litter resorption was observed
only at 200 mg/kg, and delayed parturition was seen at * 100 mg/kg. In
contrast, the Long-Evans hooded strain showed maternal weight loss at
^100 mg/kg and full-litter resorption at 200 mg/kg, but no effects on
parturition.
15
-------
Similar experiments conducted using the atrazine metabolites
desethylatrazine, desisipropyl atrazine, diaminochlorotriazine and
hydroxyatrazine, demonstrated that these metabolites were of equal or
lesser potency than parent atrazine.
Summary/Conclusion
The maternal and developmental NOAELs and LOAELs for each strain
are tabulated below.
Maternal NOAEL
Maternal LOAEL
Developmental
NOAEL
Developmental
LOAEL
F344
25 mg/kg
50 mg/kg
25 mg/kg
50 mg/kg
Sprague-
Dawley
—
25 mg/kg
50 mg/kg
100 mg/kg
Long -Evans
50 mg/kg
100 mg/kg
100 mg/kg
200 mg/kg
13.3 Atrazine Effects on Pubertal Development
13.3.1 Female
A recently completed study (Laws et a/., 2000, Laws et al.
submitted) evaluated the effects of atrazine on pubertal development in
the female Wistar rat. Atrazine (97.1%) was administered by oral gavage
(in a suspension of 1% methyl cellulose) to 165 female Wistar rats, 15 or
30 rats/dose, at dose levels of 0,12.5, 25, 50,100 or 200 mg/kg/day, from
PND 22 through 41. To evaluate the effects of lower body weight gain
during treatment, a pair-fed group (n=15) was included where the food
intake of each pair-fed rat was dependent upon the amount consumed by
its respective mate in the ATR 200 mg/kg/day group. Half of the rats were
killed on PND 41 and liver, kidney, adrenal, ovary, uterus and pituitary
weights were collected. Estrous cyclicity was evaluated in the remaining
females by monitoring changes in vaginal epithelial cells from vaginal
opening through PND 70.
16
-------
The mean body weights ± SEM for all treatment groups were equal
on PND 22. Body weight on PND 41 was unaltered by 12.5, 25, 50 and
100 mg/kg/day, but was significantly reduced by 11% in the 200
mg/kg/day and 9% in the pair-fed groups. As compared with the control,
the total gain in body weight during the 20-day treatment period was
reduced to 17% and 15% in the ATR 200 mg/kg/day and pair-fed groups,
respectively.
Vaginal opening was significantly delayed 2.3, 3.9 and 7.1 days
following exposure to 50,100 and 200 mg ATR/kg, respectively. In
addition, vaginal opening did not occur in 18/31 females in the highest
ATR dose group by the end of the dosing period. Body weight at the time
of vaginal opening was significantly increased in the 50,100 and 200
mg/kg/day groups as compared with the controls (as would be expected
due to the increase in age at vaginal opening). However, no significant
difference in the age or body weight at the time of vaginal opening was
observed between the control and the pair-fed groups.
Irregular estrous cycles (e.g., increased number of days of
diestrus) were observed between the time of vaginal opening and PND 41
in females exposed to 50 and 100 mg/kg/day, but returned to normal by
the end of the 30-day exposure period. Once dosing was discontinued,
vaginal opening occurred in all females in the 200 ATR group within four
to five days. The estrous cycles in the ATR 200 females were irregular
during the first 15-day interval following vaginal opening, but also returned
to regular four to five day cycles by PND 70.
Summary/Conclusion
Atrazine exposure delayed vaginal opening and altered estrous
cycles in female Wistar rats following oral exposure during PND 22 to 41.
The LOAEL for vaginal opening is 50 mg/kg/day and the NOAEL is 25
mg/kg/day under the conditions of this protocol. The LOAEL for estrous
cycle alteration is 50 mg/kg/day and the NOAEL is 25 mg/kg/day under
the conditions of these assays. The effect on the estrous cycle is
reversible as indicated by the fact that normal estrous cycles resumed in
all females by the end of the 30-day post-exposure period. Data from this
study are consistent with an effect on the central nervous system and
subsequent alterations in hormonal control during pubertal development.
17
-------
13.3.2 Male
Work from two separate laboratories has examined the effect of
atrazine exposure on puberty onset in male rats. Data from the lab of Dr.
Ralph Cooper (et a/., 2000) will be described first followed by data from
the lab of Dr. Barry Zirkin (Trentacoste et a/., 2000).
Stoker et a/., 2000a, and Stoker et a/., submitted undertook a
series of experiments using the weanling male Wistar rat. Animals were
treated from PND day 23 to 53 with atrazine. Atrazine (97.1%) was
administered by daily gavage to male Wistar rats of similar body weight at
doses of 12.5, 25, 50,100,150 and 200 mg/kg/day. Six rats per dose
were used at 12.5, 25, and 150 mg/kg/day and 20 to 24 per dose were
used at 50,100 and 200 mg/kg/day. An additional ten rats were pair-fed
to match the food intake of the 200 mg/kg/day rats. Parameters
measured were: body weights; prostate, seminal vesicle, epididymis and
testes weights; preputial separation (PPS); and serum testosterone,
estradiol, estrone, LH and prolactin levels. Organ weights and hormone
measures were taken on day 53. Body weights and preputial separation
were determined daily fro PND 23 to 53.
Body weights in the 200 mg/kg/day group were significantly
decreased from day 43 to 53 compared to controls. There were no
significant alterations in body weight in any other dose group. Testes
weights (neither absolute nor relative to body weight) were not altered in
any dose group. Absolute epidydimal and seminal vesicle weights were
significantly decreased in the 200 mg/kg/day group and the pair-fed
group. When adjusted for body weight, the seminal vesicles were still
significantly reduced but the epididymis were not. Lateral prostate
weights were not altered by atrazine treatment, but ventral prostate
weights, both absolute and relative to body weight, were significantly
decreased in all dose groups from 50 to 200 mg/kg/day. Serum hormone
levels, for the most part, were not significantly altered by treatment with
atrazine. There was, however, a statistically-significant increase in serum
estrone and estradiol levels at 200 mg/kg/day compared to controls.
18
-------
The major effect of atrazine on the male rats in this study was a
delay in preputial separation. Preputial separation, which occurred on
about day 42 in the controls in this study, was delayed by 2.3,1.7,1.7,1.7
and 3 days in the 12.5, 50,100,150 and 200 mg/kg/day groups. The
pair-fed animals displayed delays of two days compared to controls. A
significant delay was not seen at the 25 mg/kg/day "dose with the mean
day of preputial separation being 43 days.
Trenatcoste, et a/., 2000 dosed male Sprague-Dawely rats by
gavage from PND 22 to 47. Nine to twelve animals per dose level were
dosed at 1, 2.5, 5,10,25, 50,100 or 200 mg/kg/day of atrazine (96.1%).
A separate study termed a "food deprivation study" was also conducted.
In this study animals were dosed at 100 mg/kg/day and the amount of
food consumed on a daily basis was measured. A second group of rats
was vehicle-treated and fed the average daily intake of food consumed by
the atrazine-treated group while a third group was vehicle-treated and fed
ad lib. Parameters measured in both studies were body weights, serum
and intratesticular (interstitial fluid) testosterone levels, serum LH levels,
testes, epididymis, ventral prostate and seminal vesicle weights. Unlike
the above described Stoker et a/., 2000a and Stoker et a/., submitted
study, PPS was not measured in this study.
Body weights were significantly reduced in the 100 and 200
mg/kg/day dose groups compared to controls. Serum and intratesticular
testosterone levels at 100 and 200 mg/kg/day were significantly reduced.
Serum LH reduced 17 and 20% at he 100 and 200 mg/kg/day groups,
respectively. Only the 200 mg/kg/day decrease in LH concentration was
significant. Significant reductions in seminal vesicle and ventral prostate
weight were seen at 100 and 200 mg/kg/day. Testes and epididymis
weights were not significantly altered at any dose. Body weights, organ
weights and hormone levels were not significantly altered at any dose
from one to 50 mg/kg/day.
19
-------
The food deprivation study also showed significant reductions in
body weight, serum and interstitial testosterone, serum LH, ventral
prostate weight, and seminal vesicle weight in rats treated with 100
mg/kg/day atrazine compared to vehicle-treated controls fed ad lib. The
pair-fed, food deprived rats (the vehicle-treated rats fed the average daily
intake of the 100 mg/kg/day rats) also showed significant decreases in
serum and interstitial testosterone, serum LH, ventral prostate weight, and
seminal vesicle weight compared to animals fed ad lib. The reductions in
body weight, serum LH, and ventral prostate and seminal vesicle weights
were almost identical between the 100 mg/kg/day and the pair-fed, food-
deprived rats.
13.4 Atrazine Effects on Prostate
As described above, atrazine has been shown to depress the secretion of
prolactin. Section 3.3. above describes the role of milk-derived prolactin in
development of the TIDA neurons in the neonatal rat hypothalmus, and the
resulting hyperprolactinemia followed by lateral prostatitis that is the
consequence of incomplete development of these neurons. To summarize these
points: without early lactational exposure to PRL, TIOA neuronal growth is
impaired and elevated PRL levels are present in the prepubertal male.
Hyperprolactinemia in the adult male rat has been implicated in the development
of prostatitis.
Thus, early lactational exposure of dams to agents that suppress suckling-
induced PRL release (possibly atrazine) could lead to a disruption in TIDA
development in the suckling male offspring, followed by altered PRL regulation
and subsequent hyperprolactinemia and prostatitis in these male offspring.
To test the hypothesis that atrazine exposure of dams during lactation
could initiate the above-described sequence of events, Cooper era/., 1999,
measured suckling-induced PRL release in Wistar dams treated with atrazine (by
gavage, twice daily on PND 1 to 4 at 0, 6.25,12.5, 25, and 50 mg/kg) or the
dopamine receptor agonist bromocriptine (BROM, s.c., twice daily at 0.052,
0.104, 0.208 and 0.417 mg/kg). BROM is known to suppress PRL release.
Serum PRL was measured on PND 3 using a serial sampling technique and
indwelling cardiac catheters.
20
-------
A significant rise in serum PRL release was noted in all control females
within 10 minutes of the initiation of suckling. Fifty mg/kg ATR inhibited suckling-
induced PRL release in all females, whereas 25 and 12.5 mg/kg ATR inhibited
this measure in some dams and had no discernible effect in others. The 6.25
mg/kg dose of ATR was without effect. BROM also inhibited suckling-induced
PRL release at the two highest doses.
To examine the effect of postnatal ATR and BROM on the incidence and
severity of inflammation (INF) of the lateral prostate of the offspring, adult males
were examined at 90 and 120 days. While no effect was noted at 90 days of
age, at 120 days both the incidence and severity of prostate inflammation was
increased in those offspring of ATR-treated dams (25 and 50 mg/kg). The 12.5
mg/kg ATR and the two highest doses of BROM increased the incidence, but not
severity, of prostatitis. Combined treatment of ovine prolactin (oPRL) and 25 or
50 mg/kg ATR on PND 1 to 4 reduced the incidence of inflammation observed at
120 days, indicating that this increase in INF seen after ATR alone resulted from
the suppression of PRL in the dam. Testing to determine whether or not there is
a critical period for these effects revealed that the critical period for this effect is
PND 1 to 9.
13.5 Summary/Conclusion
These data demonstrate that ATR suppresses suckling-induced PRL
release and that this suppression results in an increase in lateral prostate
inflammation in the offspring and that the critical period for this effect is PND 1 to
9.
13.5.1 Summary
The NOAELs for the above-described effects on pregnancy, pubertal
onset and prostatitis are, for the most part, at or above 25 mg/kg/day. The
exceptions are:
• A NOAEL for maternal effects in the pregnancy maintenance
studies, in the SD rat strain only, was not found and the LOAEL is
25 mg/kg/day;
• The NOAEL for delay of pubertal onset in males is not clear as a
significant delay was seen at 12.5 mg/kg/day, but not at the next
highest dose of 25 mg/kg/day;
• The NOAEL for prostatitis is 12.5 mg/kg/day.
21
-------
DRAFT: DO NOT CITE OR QUOTE
References
-------
DRAFT: DO NOT CITE OR QUOTE
References
Adler, I.D. 1980. A review on the coordinated research effort on comparison of test
systems for the detection of mutagenic efforts. Mut. Res.: 74:77-93.
Anderson, K.J., Leighty, E.G. & Takahashi, M.T. 1972. Evaluation of herbicides for
possible mutagenic properties. J. Agric. Food Chem.: 20: 649-656.
Anderson, K.E. Sellers, T.A., Chen, P-L, Rich S.S., Hong C.P.. Folsom A.R. 1997.
Association of Stein-Leventhal syndrome with the incidence of postmenopausal breast
carcinoma in a large prospective study of women in Iowa. Cancer. 79: 494-499.
Arthur, AT. 1984a. A teratology study of atrazine technical in New Zealand White
rabbits. Safety Evaluation Facility, Ciba-Geigy, Summit, NJ. Project No. 68-84. Ace. no.
254979.
Arthur, AT. 1984b. A teratology study of simazine technical in New Zealand White
rabbits. Safety Evaluation Facility, Ciba-Geigy, Summit, NJ. Project No. 62-83. Ace. no.
252938.
Babic- Gojmerac, T., Kniewald, Z. Kniewald, J. 1989. Testostrone metabolism in
neuroendocrine organs in male rats under atrazine and deethyatrazine influence. J.
Steroid Bichem. 33(1):141-146.
Baird, R.E., Losinski, N.E., Kovacs, K. and Thorner, M.O. 1990. Correlation study of
serum prolactin levels, pituitary weights, number and structure of pituitary lactotrophs in
females Long-Evans rats. Exp. Path.: 40: 95-103.
Bartsch, H., Malaveille, C., Camus, A.-M., Martel-Planche, G., Brun, G., Hautefeuille,
A., Sabadie, N., Barbin, A., Kuroki, T., Drevon, C., Piccoli, C. & Montesano, R. 1980.
Validation and comparative studies on 180 chemicals with S. typhiumurium strains and
V79 Chinese hamster cells in the presence of various metabolizing systems. Mut. Res.:
76:1-50.
Ben-Jonathan, N., Mershon, J.L., Allen, D.L. and Steinmetz, R.W. 1996 Extrapituitary
prolactin: Distritubion, regulation, functions, and clinical aspects. Endocrin. Rev. 17:
639-669.
Ref-2
-------
DRAFT: DO NOT CITE OR QUOTE
Benigni, R., Bignami, M., Camoni, I., Carere, A., Conti, G., lachetta, R., Morpurgo, G. &
Ortali, V.A. 1979. A new in vitro method for testing plant metabolism in mutagenicity
studies. J. Toxicol. Environ. Health: 5: 809-819.
Bernstein, L, Henderson, B.E., Hanisch, R., Sullivan-Halley, J. and Ross, R.K. 1994
Physical exercise and reduced risk of breast cancer in young women. JNCI 86:
1403-1408.
de Bertoldi, M., Griselli, M., Giovannetti, M. & Barale, R. 1980. Mutagenicity of
pesticides evaluated by means of gene conversion in Saccharomyces cerevisiae and in
Aspergillus nidulans. Environ. Mutag.: 2: 359-370.
Blair, M., Kruger, S. and Nair, K. 1981. Two-year chronic and oncogencity study with
CGA-722662 technical and albino rats. International Research and Development
Corporation. IRDC report no. 382-008. IRDC Report no. 382-081. Ace. no. 00103202.
Bianco-Rodriguez, J. and Martinez-Garcia, C. 1996. Further observations on the early
events that contribute to establishing the morphological pattern shown by the oestradiol
suppressed testis. Tissue & Cell. 28(4) 387-399.
Blank, M.S. Ching, M.C. and Dufau, M.L. 1986. Bioactivity of serum and pituitary
protection during rat estrous cycle. Endocrinology 1118 (5)1886-1891.
Blankenstein, M.A., Broerse, J.J., van Zwieten, M.J., and van der Molen, H.J. 1984.
Prolactin concentration in plasma and susceptibility to mammary tumors in female rats
from different strains treated chronically with estradiol-17 3- Breast Cancer Research
and Treatment. 4:137-141.
Bogdanffy, M. 1990. Combined chronic toxicity/Oncogencity - Rat. Du Pont Haskell
Laboratory no. 23-90. MRID 41509902.
Bradlow, H.I., Davis, D.L., Lin, g., Sepkovic, D. and Tiwari, R. 1995. Effects of
pesticides on the ratio of 16a/2-hydroxyestrone: a biologic marker of breast cancer risk.
Environ.. Health Persp.: Vol 103, Supplement 7: 147-150.
Brann, D.W., Rao, I.M. and Mahesh, V.B. 1988. Antagonism of estrogen-induced
prolactin release by progesterone. Bio. Repro. 39:1067-1073.
Ref-3
-------
DRAFT: DO NOT CITE OR QUOTE
Briza, J. 1989. Estimation of mutagenicity and metabolic activation after recurrent
exposures ofNicotiana tabacum L. var. zanthi to 14 pesticides. Biol. Plant: 31:145-
151.
Brown, L.M., Blair, A., Gibson, R., Everett, G.D., Cantor, K.P., Schuman, L.M.,
Burmeister, L.F., Van Lier, S.F., and Dick, F. 1990. Pesticide exposures and other
agricultural risk factors for leukemia among men in Iowa and Minnesota. Cancer Res.
50:6585-6591.
Brusick, D. 1994. An assessment of the genetic toxicity of atrazine: Relevance to
human health and environmental effects. Mut. Res.: 317:133-144.
Brusick D, Albertini R, McRee D, Peterson D, Williams G, Hanawalt P, and J. Preston.
1998. Genotoxicity of radiofrequency radiation. Environ. Molec. Mutagenesis 32:1-16.
Burnmeister, L.F. 1990. Cancer in Iowa farmers: Recent results. Am. J. Ind. Med.
18:295-301.
Butcher, R.L., Collins, W.E., and Fugo, N.W. 1974 . Plasma concentration of LH, FSH,
prolactin, progesterone and estradiol-17B throughout the 4-day estrous cycle of the rat.
Endocrinology. 94:1704.
Butler, M.A. & Hoagland, R.E. 1989. Genotoxicity assessment of atrazine and some
major metabolites in the Ames test. Bull. Environ. Contam. Toxicol.: 43: 797-804
Cantox, 2000. Assesment of the relevance to humans of the mode of action of atrazine
according to the 1999 U.S. EPA revised draft guidelines for carcinogen risk assesment.
Cantox Health Sciences International, Bridgewater, N.J.
Ceresa, C. 1988a. Structural chromosomal aberration test - micronucleus test, mouse.
Genetic Toxicology laboratories, Ciba-Geigy. Laboratory report number: 871546. May
31,1988. Unpublished . MRID 40722301.
Ceresa. C. 1988b. Structural chromosomal aberration test - micronucleus test, mouse.
Ciba-Geigy, Basle, Switzerland. Laboratory report number: 881189. September 15,
1988. Unpublished. MRID 41442901.
Ref-4
-------
DRAFT: DO NOT CITE OR QUOTE
Ceresa. C. 1988c. Structural chromosomal aberration test - micrpnucleus test, mouse.
Ciba-Geigy, Basle, Switzerland. Laboratory report number: 871373. September 15,
1988. MRID 41479401.
Chau, R.F., McCormick, G.C., and Arthur, AT. 1991.104-week Oral
toxicity/Carcinogenicity Study in Rats. Ciba-Geigy Study No. MIN 872225. MRID
419012-01.
Cho, B. N., Sun, Y.H., Yoon, Y.D., Lee, C.C., and Kim, K. 1993. Progesterone inhibits
the estrogen-induced prolactin gene expression in the rat pituitary. Mol. CellEndo.
93:47-52.
Chow, E. and Hart, S.G. 1995. 2-year dietary chroinc toxicity/oncogencity study with G-
34048 in rats. Ciba-Geigy, Farmington, CT. Laboratory study No. F-00125. MRID
43532001.
Ciba-Geigy 1980. Two-year chronic feeding study in rats with terbutryn technical.
International Research and Development Corporation. IRDC report no. 382-008.
Ciba-Geigy, 1984. Nucleous anomaly test. Ciba-Geigy, Basle, Switzerland. Laboratory
report number: 850624. July 20, 1984. MRID 470042031.
Ciba-Geigy, 1986. In vitro test for mutagenic properties of substances in mammalian
cells with and without activation. Ciba-Geigy, Basle, Switzerland. Laboratory report
number: 850624. May 6,1986. MRID 470289044.
Clements, C., Ralph, S. & Petras, M- 1997. Genotoxicity of select herbicides in Rana
catesbeiana tadpoles using the alkaline single-cell gel DNA electrophoresis (comet)
assay. Environ. Mol. Mut: 29: 277-288.
Clevenger, C.V., Chang, W-P., Ngo, W., Pasha, T.L.M., Montone, K.T. and
Tomaszewski, J.E. 1995 Expression of prolactin and prolactin receptor in human
breast carcinoma. Am. J. Pathol. 146: 695- 705.
Conner, K., Chen, I., Liu, H., Berhane, K., Sciarretta, C., Safe, S., and Zacharewski, T.
1996. Failure of chloro-s-triazine-derived compounds to induce estrogen receptor-
mediated responses in vivo and in vitro. Fund. Appl. Tox.: 30:93-101. MRID 4394403.
Ref-5
-------
DRAFT: DO NOT CITE OR QUOTE
Consensus Panel, 2000. Evaluation of a hormonal mechanism for mammary
tunorigenesis of the chlor-s-triazine herbicides: Fourth Consensus Panel Report:
January 13, 2000. In: Background materials submitted to the Science Advisory Panel
for the public meeting to be held January 27-28, 2000, Public Docket OPP-00637.
Novartis Crop Protection, Greensboro, N.C.
Console, G.M., Gomez Dumm, C.L, Brown, O.A., Ferese, C. and Goya, R.G. 1997.
Sexual dimorphism in the age changes of the pituitary lactotrophs in rats. Mech. Aging
and Dev.: 95: 157-166.
Cooper, R.L. and Walker, R.F. 1979. Potential therapeutic consequences of age-
dependent changes in brain physiology. In: CNS aging and its neurophamacology, W.
Meir-Ruge, ed. Karger, Basel, pp. 54-76.
Cooper R.L, Conn, P.M., Walker, R.F. 1980. Characterization of the LH surge in
middle-aged female rats. Biol. Reprod. Oct;23(3):611-5.
Cooper, R.L. 1983. Pharmacological and dietary manipulations of reproductive aging in
the rat. signficance to central nervous system aging. In: Clincal and Experimental
Intervention of the Aging Process, Cooper, R.L. and Walker, R.F., eds. New York:
Marcel Dekker; 27-44.
Cooper, R.L., Goldman, J.M., and Rehnberg, G.L. 1986. Neuroendocrine control of
reproductive function in the aging female rodent. J. Amer. Geron. Soc.: 34:735-751.
Cooper, R.L., Stoker, T.E., Goldman, J.M., Parrish, M.B., and Tyrey, L 1996. Effect of
atrazine on ovarian function in the rat. Reproductive Toxicology. 10 (4):257-264.
Cooper, R.L., Stoker, T.E., McElroy, W.K., and Hein, J. 1998. Atrazine (ATR) disrupts
hypothalamic catecholamines and pituitary function. The Toxicologist. 42:160.
Cooper, R.L., Robinette CL, and Stoker T.E., 1999. Maternal exposure to atrazine
during lactation suppresses suckling-induced prolactin release and results in prostatitis
in the adult offspring. Tox. Sci. 1999 Nov; 52(1):68-79.
Cooper, R.L, Stoker, T.E., Tyrey, L., Goldman, J.M., and McElroy, W.K. 2000. Atrazine
disrupts the hypothalamic control of pituitary-ovarian function. Tox. Sci. 53: 297-307.
Ref-6
-------
DRAFT: DO NOT CITE OR QUOTE
Copley, M.P. 1988. Third peer review of atrazine- reevaluation following the September
7,1988 Scientific Advisory Panel Review. Memorandum from Marion Copley, HED to
Robert Taylor, RD and Jude Andreasen, SRRD.
Cotterchio, M., Kreiger, N. Darlington, G. and Steingart, A. 2000. Antidepressant
medication use and breast cancer risk. Am. J. Epi. 151:951-957.
Couiam, C.B., Annegers, J.F. and Kranz, J.S. 1983. Chronic anovulation sydrome
and associated neoplasia. Obstet. Gynecol. 61: 403-407.
Grain, A.D., Guillette, L.J., Rooney, A.A. and Pickford, D.B. 1997. Alterations in
steroidogenesis in alligators (Alligator mississippiensis) exposed naturally and
experimentally to environmental contaminants. Environ.. Health Persp. 105(5):528-533.
Cummings, A.M., Rhodes, B.E., and Cooper, R.L (Submitted). Effect of atrazine on
implkantation and early pregnancy in four strains of rats.
Cutts, J.H. and Noble, R.L. 1964. Estrone-induced mammary tumors in the rat:
Induction and behavior of tumors. Cancer Res.: 24:1116-1123.
Dahlgren, E., Friberg, L.G., Johansson, S., Lindstrom, B., Oden, A., Samsioe, G. and
Janson, P.O. 1991. Endometrial carcinoma; ovarian dysfunction-a risk factor in
young women. Eur. J. Obstet Gynecol. Reprod. Biol. 41:143-150.
Danzo, B.J. 1998. The effects of environmental hormones on reproduction. Cell. Mol.
LifeSci.. 54:1249-1264.
Das, P.C., McElroy, W.K. and Cooper, R.L. 1999. Cyanazine (CYN) affects
norepinephirine release in pheochromocytoma (PC 12) cells. The Toxicologist. 48: 298.
Das, P.C., McElroy, W.K. and Cooper, R.L. (Submitted). "Differential Modulation of
Catecholamines by Chlorotriazine Herbicides in Pheochromocytoma (PC12) cells in
vitro".
Daston, G.P., Gooch, W., Breslin, W.J., Shuey, D.L., Nikiforov, A. I., Fico, T.A., and
Gorsuch, J.W. 1997. Environmental estrogens and reproductive health: A discussion of
the human and environmental data. Repro. Tox. Vol. 11: 465-481.
Ref-7
-------
DRAFT: DO NOT CITE OR QUOTE
DeMarzo AM, VL Marchi, Jl Epstein, and WG Nelson. 1999. Proliferative inflammation
atrophy of the prostate implications for prostatic carcinogenesis.. Am. J. Path.
155:1985-1992.
den Tonkelaar, I. and de Waard, F. 1996. Regularity of menstrual cycles in women
aged 41-46 in relation to breast cancer risk: results from the DOM-project. Breast
Cancer Res. Treat. 38(3):253-8.
Deparde, E. 1986. Atrazine: Salmonella/mammalian - microsome mutagenicity test.
Ciba-Geigy, Basle, Switzerland. Laboratory report number: 861172. December 5,
1986. MRID 40246601.
Deparde, E. 1988. Gene mutations test Salmonella and Escheria/liver-microsome test.
Ciba-Geigy, Basle, Switzerland . Study No. 871376. February 15,1988. MRID
40722304.
Deparde, E. 1989. Gene mutations test Salmonella and Escheria/liver-microsome test
(G-30033). Ciba-Geigy, Basle, Switzerland . Study No. 891236. December 18,1989.
MRID 43093102.
Deparde, E. 1990. Gene mutations test. Salmonella and Escheria/liver-microsome test
(G-28279). Ciba-Geigy, Basle, Switzerland. Study No. 891243. January 18,1990.
MRID 43049101.
Deparde, E. and Karimi, M. 1987. Salmonella/Mammalian microsome mutagenicity
test. Ciba-Geigy, Basle, Switzerland. Study No. 871372. November 10,1987. MRID
40722302.
Delzell, E and Sathiakumar, N., 1996. A combined analysis of mortality amoung
workers at the Ciba-Geigy Corportation's Mclntosh and St. Garbrial plants - an update.
MRID 44256502.
Donna, A., Crosignani, P., Robutti, F. Betta, P.G., Bocca, R., Mariani, N., Ferrario, F.
Fissi, R., and Berrino, F. 1989. Triazine herbicides and ovarian epithelial neoplasms.
Scan. J. Work Environ. Health. 15:47-53.
Ref-8
-------
DRAFT: DO NOT CITE OR QUOTE
Dunkelberg, H., Fuchs, J., Hengstler, J.G., Klein, E., Oesch, F. and Striider, K. 1994.
Genotoxic effects of the herbicides alachlor, atrazine, pendimethaline, and simazine in
mammalian cells. Bull. Environ. Contam. Tox.: 52:498-504.
Durbin, P.W., Williams, M.H., Jeung, N. and Arnold, J.S. 1966. Development of
spontaneous tumors ver the life span of the female Charles River (Sprague-Dawley) rat:
The influence of ovariectomy, thyroidectomy, and adrenalectomy-ovariectomy. Cancer
Res.: 26 part 1:400-411.
Ehrmann , DA, Barnes, R. and Rosenfield, R.L. 1995. Polycystic ovary syndrome as
a form of functional ovarian hyperandrogenicsm due to dysregulation of androgen
secretion. Endocr. Rev. 16: 322-353.
Eldridge, J., Wetzel, L, Tisdel, M., and Luempert, L.G. 1993a. Determination of
Hormone Levels in Sprague-Dawley Rats Treated with Atrazine Technical: Revised
Supplement to Final Report. Hazleton Washington, Inc. Lab Project Number: 483-278.
MRID 42743902.
Eldridge, J., Wetzel, L, Tisdel, M. and Luempert, L.G. 1993b. Determination of
Hormone Levels in Fischer-344 Rats Treated with Atrazine Technical: Revised
Supplement to Final Report. Hazleton Washington, Inc. Lab Project Number: 483-
279. MRID 42743903.
Eldridge, J.C., Wetzel, L.T., Stevens, J.T., Simpins, J.W. 1999. The mammary tumor
response in triazine-treated female rats: A threshold-mediated interaction with strain
and species-specific reproductive senescence. Steriods 64:672-678.
Emnova, E-.E., Merenyuk, G.V. & Curkan, L.G. 1987. Genetic activity of sim-triazine
herbicides on Saccharomyces cerevisiae. Tsitol. Genet: 21:127-131.
Epstein, D.L., Hazelette, J.R., Yau, E.T. 1991. Simazine Technical: Two-generation
reproductive toxicology study in rats Pharmaceuticals Division, Ciba-Geigy. Laboratory
Report No. 882095. MRID 41803601.
Estes, K.S., Simpkins, J.W. and Kalra, S.P. 1982. Normal and neuronal function and
hyperprolactinemta in old pseudopregant Fischer 344 rats. Neurobio. Aging: 3:247-252.
Ref-9
-------
DRAFT: DO NOT CITE OR QUOTE
Fahrig, R. (1974) Comparative mutagenicity studies with pesticides. IARC Sci Publ,
10,161-181
Ferin M and Vande Wiele R, 1984. Endogenous opioid peptides and the control of the
menstrual cycle. Eur. J. Obstet. Gynecol. Reprod. Biol. 18(5-6):365-73
Flowers, L. and Kier, L.D. 1978. Final report on Salmonella mutagenicity assay of
glyphosate. Monsanto Environmental Health Laboratory, St. Louis MO. Study no. LF-
78-161. June 16, 1978. MRID 00078620.
Ford, M.M., and Eldridge, J.C. 2000. Atrazine disruption of the hypothalamic-pituitary
axis: Consequences for reproductive functions. The Toxicologist 54:260
Freeman, M.E. 1981. Control of the corpus luteum: A model system for toxicology
research. Environ. Health Persp.: 38: 51-54.
Fritz, H. 1976. Reproduction study: G-30028 Tech.: Rat; Seg. II. Pharmaceuticals
Division, Ciba-Geigy. MRID 087879.
Gammon, M.D. and Thompson, W.D. 1991. Polycystic ovaries and the risk of breast
cancer. AmJ.Epi. 134:818-824.
Garry, V.F., Schreinemachers, D., Harkins, M.E., Griffith, J. 1996. Pesticide appliers,
biocides, and birth defects in rural Minnesota. Environ. Health Persp. 104(4):394-399.
Gebel, T., Kevekordes, S., Pav, K., Edenhader, R., Dunkelberg, H. 1997. In vivo
genotoxicity of selected herbicides in the mouse bone-marrow micronucleus test. Arch.
Tox.: 71:193-197.
Gellatly, J.B.M. 1981. A two-year feeding study of Bladex in mice. Shell Toxicology Lab
study no. 1493, Accesion no. 247295-298.
Gelnick, D. 1991 a. Tests for other genotoxic effects- auto-radiographic DMA repair test
on rat hepatocytes (G-28279). Ciba-Geigy, Greensboro, NC.. Study No. 901308. April
12,1991. MRID 43049105.
Ref-10
-------
DRAFT: DO NOT CITE OR QUOTE
Gelnick, D. 1991b. Tests for other genotoxic effects- auto-radiographic DNA repair test
on rat hepatocytes (G-30033). Ciba-Geigy, Greensboro, NC. . Study No. 901310. April
26,1991. MRID 43049106.
Gerspach, R. 1991. 3-month oral toxicity study in rats. Ctba Geigy Limited, Basle,
Switzerland. Study no. 901264. MRID 43013206.
Ghiazza, G., Zavarise, G., Lanero, M. & Ferraro, G. 1984. SCE (sister chromatid
exchanges) induced in chromosomes of human lymphocytes by trifluralin, atrazine and
simazine. Bull. Soc. Hal. Biol. Sper.: 60: 2149-2153.
Gibori, G., Richards, G.S. 1978. Dissociation of two distinct luteotrophic effects of
prolactin: regulation of lutenizing hormone receptor content and progesterone secretion
during pregnancy. Endocrinology, 102:767-74.
Ginkis, M.L. 1989. A teratology (segment II) study in rats. Ciba-Geigy, Summit, N.J.
Study No. 882049. MRID 41065201.
Goldfien, A and Monroe, S.E. 1997. Ovaries. In: Basic and Clincal Endocrinology. Fifth
Ed. Appelton and Lange, Stamford, CT. Greenspan, F.S. and Strewler, G.S. eds. pp.
459.
Goldman, J.M., Laws, S.C., Balchak, S.K., Cooper, R.L and Kavlock, R.J. 2000.
Endocrine-disrupting chemicals: Prepubertal exposures and effects on sexual
maturation and thyroid activity in the female rat. A focus on the EDSTAC
recommendations. Critical Rev. Tox. 30(2): 135-196.
Goya, R.G., Lu, J.K. and Meites, J. 1990. Gonadal function in aging rats and its
relation to pituitary and mammary pathology. Mech. Aging Dev. 56(1):77-88.
Grady, D. and Emster, V.L 1996. In Schottenfeld, D and Fraumeni, J.F., Jr., eds.
Cancer epidemiology and prevention. 2nded. New York: Oxford University, pp. 1058-
1089.
Ref-11
-------
DRAFT: DO NOT CITE OR QUOTE
Graumann, K. Breithofer, A., and Jungbaur, A. 1999. Monitoring of estrogen mimics
by recombinant yeast assay:synergy between natural and synthetic compounds? Sci.
Tot. Environ.: 225:69-79.
Griffiths, AJF 1979. Neurospora prototroph selection system for studying aneuploid
production. Environ. Health Persp., 31: 75-80.
Gupta, D., Zarzycki, J., and Rager, K. 1975. Plasma testosterone and
dihydrotestosterone in male rats during sexual maturation and following orchidectomy
and experimental bilateral cryptorchidism. Steroids 25:33-42.
Hankinson, S.E., Willett, W.C., Michaud, D.S. et al. 1999 Plasma prolactin levels and
subsequent risk of breast cancer in postmenopausal women. JNCI 91: 629-634.
Haridon, J.L., Fernandez, M., Ferrier, V., Bellan, J. 1993. Evaluation of the genotoxicity
of N-nitosoatrazine, N-nitrosodiethanolamine and their precursors in vivo using the newt
micronucleus test. Wat. Res. 5:855-862.
Haseman, J., Hailey, J.R., and Morris, R.W. 1998. Spontaneous neoplasm incidences
in Fischer 344 rats and B6C3F1 mice in two-year carcinogencity studies: a National
Toxicology Program update. Tox .Path.: 26, no. 3:428-441.
Hauswirth, J.W. 1988a. Peer review of atrazine. MEMO from Judith W. Hauswirth, HED
to Robert Taylor/Clare Grubbs, RD.
Hauswirth, J.W. 1988b. Second peer review of atrazine. MEMO from Judith W.
Hauswirth, HED to Robert Taylor/Clare Grubbs, RD.
Hazelette, J. and Green, J. 1987. Oncogenicity Study in Mice: Atrazine Technical.
Ciba-Geigy Corp. Laboratory Study No. 842120. MRID 40431302.
Hazelette, J. and Green, J. 1988. Simazine technical: 95 week oral toxicity/oncogencity
study in mice. Ciba-Geigy Corp. Laboratory Study No. 842121. MRID 40614404
Hazelton Labs 1984. Representative Historical Control Data, Hazelton Labs. March 5,
1984.
Ref-12
-------
DRAFT: DO NOT CITE OR QUOTE
Henderson, B.E., Ross, R. and Bernstein, L. 1988. Estrogens as a cause of human
cancer: The Richard and Hinda Rosenthal Foundation Award Lecture. Cancer Res.
48: 246-253.
Herschlag, A. and Peterson, C.M. 1996. Endocrine Disorders. In: Novak's Gynecology,
12 th Edition. Ed. Johnathan Berek. Willams and Wilkins, Baltimore.
Hertner, T. 1988. Tests for other genotoxic effects - auto-radiographic DMA repair test
on rat hepatocytes. Ciba-Geigy, Basle, Switzerland. Laboratory report number: 871374.
January 22,1988. Unpublished. MRID 40722305
Hertner, T. 1992. Atrazine autoradigraphic DMA repair test on rat hepatocytes test for
other genotoxic effects. Ciba-Geigy Corp, Basle, Switzerland. Laboratory report
number: 911246. April 14,1992. Unpublished. MRID 42547105.
Hertner, T. 1993. Atrazine structural chromosomal aberration test. Dominant lethal test,
mouse 8 weeks. Ciba-Geigy, Basle, Switzerland. Study No. 911247. January 7,1993.
Unpublished. MRID 42637003.
Hoar, S.K., Blair, A., Holmes, F.F., Boysen, C. and Robel, R.J. 1985. Herbicides and
colon cancer. The Lancet. \. 1277-1278.
Hoar, et a/. 1986. Agricultural herbicide use and risk of lymphoma and soft tissue
sarcoma. JAMA. 256:1141-1147.
Huang, H.H. and Meites, J. 1975. Reproductive capacity of aging female rats.
Neuroendo.: 17:289-295.
Huang HH, Steger RW, Bruni J, Meites J. Changes in patterns of sex steroid and
gonadotropin secretion in aging female rats. Endocrinology. 1978; 103:1855-1859.
Huaijin C, Junyan Z, Naiguan C. 1998 Prostatic fluid and sperm examination: 106
cases. Preliminary study on infertility. Acta Urol Belg. 66:19-21.
Hummal, H., Yourneff, M., Ginkis, M., Yau, E.T. 1989. Diaminochlortriazine: A
teratology (Segment II) study in rats. Pharmaceuticals Division, Ciba-Geigy, Summit,
NJ. Study No. MIN 872177. MRID 41392402.
Ref-13
-------
DRAFT: DO NOT CITE OR QUOTE
(ARC, 1999. World Health Organization International Agency for Research on Cancer
(IARC) Monographs on the Evaluation of Carcinogenic Risks to Humans. 73: 59-113.
Infurna, R. N. 1984. A teratology study of atrazine technical in Charles River rats. The
Safty Evaluation Facility, Ciba Corp., Summit, N.J. MRID 40566302.
Ishidate, M. Jr., Sofuni, T., & Yoshikawa, K. (1981) Chromosomal aberration tests in
vitro as a primary screening tool for environmental mutagens and/or carcinogens.
Gann Monograph on Cancer Research, 27, 95-108
Ishidate, M. Jr. (1983) Application of chromosomal aberration tests in vitro to the
primary screening for chemicals with carcinogenic and/or genetic hazards. In:
Short-term Tests for Carcinogenesis: Quo Vadis? Symposium Proceedings,
Montpellier, VT, U.S.A., Excerpta Medica, pp. 58-79
Ishidate, M., Jr 1988. Data Book of Chromosomal Aberration Test In Vitro, rev. Ed.,
New York, Elsevier.
Jafari, K. Javaheri, G. and Ruiz, G. 1978. Endometrial adenocarcinoma and the
Stein-Leventhal syndrome. Obstet Gynecol. 51:97-100.
Jessup, D.C. 1979. Three genration reproduction study in rats. International Reserach
and Devlopment corp, Mattewan, Ml. Reprot No. 382-010. MRID 00041409.
Jessup, D. 1980a. Two year oral chronic toxicity study in rats. IRDC study no. 382-007.
MRID 00041408; Ace. No. 219502401.
Jessup, D. 1980b. Two year carcinogencity study in mice. IRDC study no. 382-004.
MRID 00044335.
Jones, D. C. L, Simmon, V. F., Mortelmans, K. E., Mitchell, A. D., Evans, E. L, Jotz, M.
M., Riccio, E. S., Robinson, D. E., & Kirkland, B. A. (1984) In vitro and in vivo
mutagenicity studies of environmental chemicals. U.S. EPA report EPA-600/1-84-003.
NTIS report # PB84-138973.
Kamberi, I.A., de Vellis, J., Bacleon, E.S., and Inglish, D. 1980. Hormonal patterns of
the hypothalamo-pituitary-gonadal axis in the rat during postnatal development and
sexual maturation. Endokrinologie 75:129-40.
Ref-14
-------
DRAFT: DO NOT CITE OR QUOTE
Kappas, A. 1988. On the mutagenic and recombinogenic activity of certain herbicides
in Salmonella typhimurium and Aspergillus nidulans. Mut. Res.: 204: 615-621.
Kelly, J.P., Rosenberg, L. Rao, R.S., Palmer, J.R. and Shapiro, S. 1998. Use of
antidepressants associated with the occurence of breast cancer? Am. J. Epi.
Supp/.147:S69.
Kettles, MA, Broning, S.R., Prince, T.S., and Horstman, S.W. 1997. Triazine herbicide
exposure and breast cancer incidence: An ecologic study of Kentucky counties.
Environ. Health Persp. 105(11): 1222-1227.
Khudoley, V.V., Mizgireuv, 1. & Pliss, G.B. 1987. The study of mutagenic activity of
carcinogens and other chemical agents with Salmonella typhimurium assays: Testing
of 126 compounds. Arch. Geschwulstforsch.: 57: 453-462.
Kligerman, A. D., Doerr, C. L., Tennant, A. H., and Zucker, R. M. 2000a.
Cytogenetic studies of three triazine herbicides, Mut. Res. (465)1-2: pp. 53-59
Kligerman, A.D., Carolyn L. Doerr, C.L, Tennant, A.M., and Peng, B. 2000b.
Cytogenetic Studies of Three Triazine Herbicides II: In Vivo Studies. Submitted.
Knapp, J.S. 1995. A developmental toxicity study of proapzine technical in rabbits. WIL
Research Laboratories. Study No. WIL-157005. MRID 44153401.
Kniewald, J., Peruzovic, M., Gojmerac, T., Milkovic, K. and Kniewald, Z. 1987. Indirect
influence of s-triazines on rat gonadotropic mechanism at early postnatal period. J.
Steroid Biochem. 27(4-6): 1095-1100.
Korach, K.S. and McLachlan, J.A. 1995. Techniques for the detection of estrogenicity.
Environ.. Health Persp.: 103 (Suppl 7): 5-8.
Kulin, H.E. 1996. Delayed puberty. J. Clin. Endocrinol. Metab. Oct;81(10):3460-4.
Lang. P. 1990. Spontaneous neoplastic lesions in the CDF®[F-344]/CrlBR rat. Charles
River Laboratories.
Ref-15
-------
DRAFT: DO NOT CITE OR QUOTE
Lang, P. 1992. Spontaneous neoplastic lesions in the CD:CD®BR rat. Charles River
Laboratories.
Lasinski, E.R., Kapeghian, J.C. and Green, J.D. 1987. Simazine techincal
Salmonella/mammalian-microsome mutagenicity assay (Ames Assay). Ciba-Geigy,
Greensboro, N.C.. Laboratory report number: 87038. July 8, 1987. MRID:40614406.
Laws, S.C., Ferrel, J.M. and Cooper, R.L. 2000. The effect of atrazine on puberty in
female Wistar rats. The Toxicologist, 54 (1) 366.
Laws, S.C., Ferrell, J.M., Stoker, T.E., Schmid, J. and Cooper, R.L. Submitted. The
Effects of Atrazine on Female Wistar Rats: An Evaluation of the Protocol for
Assessing Pubertal Development and Thyroid Function.
Leav I, FB Merk, KF Lee, M Loda, M Mandoki, JE McNeal, and H Shuk-mei. 1999.
Prolactin receptor expression in the develping human prostate and in hypeplasitic,
dysplastic, and neoplastic lesions. Am. J. Path. 154: 863-870.
Lee KC, GM Rao, FL Barnett, and GH Liang 1974. Further evidence of meiotic
instability induced by atrazine in grain sorghum. Cytologia: 39:697-702.
Lee W.R., Abrahamson, S., Valencia, R. von Halle, E.S., Wurgler, F.E., and Zimmering,
S. 1983. The sex-linked recessive lethal test for mutagenesis in Drosophila
Melanogaster. A report of the US Environmental Protection Agency Gene-Tox
Program. Mut. Res. 123:183-279.
LeFevere, J. and McClintock, M. 1988. Reproductive senescence in female rats: A
longitudinal study of individual differences in estrous cycles and behavior. Bio. Repro.:
38:780-789
Legan, S.J., Coon, G.A., Karsch, F. J. 1975. Role of estrogen as initator of daily LH
surges in the ovariectomized rat. Endocrinology. 96:50.
Levi, F., Randimbison, L., Te, V.C. and La Vecchia, C. 1994 Incidence of breast
cancer in women with fibroadenoma. Int. J. Cancer. 57: 681-683.
Ref-16
-------
DRAFT: DO NOT CITE OR QUOTE
Li, A.P. 1983. In vivo bone marrow cytogenetics study of Glyphosate in Sprague-
Dawley rats. Monsanto Environmental Health Lab, St. Loius, MO. Study no. ML-83-236.
October 20,1983. MRID 00132683.
Liang, G.H. & Liang, Y.T.S. 1972. Effects of atrazine on chromosomal behavior in
sorghum.Can. J. Genet. CytoL: 14:423-427.
Libertun, C., Kaplan, S.E., and DeNicola, A.F. 1979. Progesterone negative feedback
on prolactin secretion: Importance of the brain control and of estradiol.
Neuroendocrinology 28: 64-70.
Linzer, D.I.H., and Fisher, S.J. 1999. The placenta and the prolactin family of
hormones: regulation of the physiology of pregnancy. Mol. Endo. 13(6):837-840.
Ltoi, MB, Scarfi, M.R., Santoro, A., Barbieri, R.Zeni, O., Salvemini, F., Di Berardino, D.
and Ursini, M.V. 1998. Cytogenetic damage and induction of pro-oxidant state in
human lymphocytes exposed in vitro to cliphosate, vinclozoling, atrazine, and DPX-
E9636. Environ.& Molec. Mutagen.: 32:39-46.
Lu, K.H., Hopper, B.R., Vargo, T.M., and Yen, S.S.C. 1979. Chronological changes in
sex steriod, gonadotrpin and prolacin secretion in aging female rats displaying different
reproductive states. Bio. Repro. 21:193-203.
Lu, J.K.H., Anzalone, C.R., and LaPolt, S. 1994. Relation of neuroendocrine function to
reproductive decline during aging in the female rat. Neuobio, Aging: 15(4): 541-544.
Lusby, A.F., Simmons, Z. & McGuire, P.M. 1979. Variation in mutagenicity of s-triazine
compounds tested on four Salmonella strains. Environ.. Mut: 1: 287-290.
Lyons, W.R., Berlin, I., and Friendlander, S. 1942. Cornification of balano-preputial
epithelius in normal rats and in castrated rats treated with testosterone proprionate.
Endocrinology 31:659.
Ma, T., Harris, M.M., Anderson, V.A., Ahmed, I., Mohammad, K., Bare, J.L. & Lin, G.
1984. Tradescantia-micronucleus (Trad-MCN) tests on 140 health-related agents. Mut.
Res.: 138: 157-167.
Ref-17
-------
DRAFT: DO NOT CITE OR QUOTE
Malasanos, T.H. 1997. Sexual development of the fetus and pubertal child. Clin. Obs.
Gyn. 40:153-167.
Mainero, J., Wimbert, K., Wright, J., Infurana, R.N. 1986. Simazine technical; A
teratology study in rats. Pharmaceutical Division, Ciba-Geigy. Laboratory Report No.
B6/110. MRID 40614403.
Mainiero, J., Yourneff, M., Giknis, M.L.A., Yau, E.T. 1987. Atrazine 2-generation
reproduction study. Agricultural Division, Ciba-Geigy, Greenboro, NC. Study no.
852063. MRID 40431303.
Mansfield, M.J. and Emans, S.J. 1989. Anorexia nervosa, athletics, and amenorrhea.
Pediatr. Clin. North Am. 36: 533-549
Marshall (1966) Marshall's Physiology of Reproduction: Longmans, Green and Co.
LTD.
Marshall, J.C. and Eagleson, C.A. 1999. Endocrinology and Metabolism Clinics of
North America Vol. 28 (2) - June pp. 295- 324.
Mauvais-Jarvis, P., Kutten, F. and Gompel, A. 1987. Antiestrogen action of
progesterone in breast tissue. Hormone Res. 28:212-218.
Marty, J.H. 1992. Developmental toxicity (Teratogencity) study in rats with G-28279
technical. Ciba-Geigy Limited, Basel, Switzerland. Stuyd no. 901262. MRID 43013208.
Mathias, M., Gilot-Delhalle, J. & Moutschen, J. 1987. Mutagenicity of atrazine in
Schizosaccharomyces pombe Lindner with and without metabolic activation by maize.
Environ. Exp. Bot: 29: 237-240.
Matsumoto, A.M., Karpas, A.E., South worth, M.B., Dorsa, D.M., and Bremner, W.J.
1986. Evidence for activation of the central nervous system-pituitary mechanism for
gonadotropin secretion at the time of puberty in the male rat. Endocrinology 119:362-9.
Mayhew, DA, Taylor, G.D., Smith, S.H. and Banas, D.A. 1986. Twenty-four month
combined chronic oral toxicity and oncogencity study in rats utilizing atrazine technical.
American Biogenics Corp., Decatur, IL. Lab Study No.: 410-1102. Accession no.
262714-262727. MRID: 00158930.
Ref-18
-------
DRAFT: DO NOT CITE OR QUOTE
McComb, D.J., Kovacs, K., Berl, J. and Zak, F. 1984. Piutitary adenomas in old
Sprague-Dawley rats: A histologic, ultrastructural, and immunocytochemical study.
J/VC/: 73, 5: 1143-1156.
McComb, D.J., Hellman, P., Thorner, M.O., Scott, D., Evans, W.S. and Kovacs, K.
1986. Morphologic effects of bromocriptine on spontaneously occuring pituitary
prolactin-cell hyperplaisa in old Long-Evns rats. Am. J. Path.: 122:7-16.
McConnell, R.F. 1989a. General observations on the effects of sex steriods in rodents
with emphasis on long-term oral contraceptive studies. In; Safety requirements for
contraceptive steriods. Ed. Micheal, P.. Cambridge Union Press, Cambridge, England.
McConnell, R.F. 1989b. Comparative aspects of contraceptive steroids: effects
observed in rats. Tox. Path.; 17(2):385-8
McConnell, R.F. 1995. A histomorphologic reevaluation of the ovaries, uterus, vagina,
mammary gland, and pituitary gland from Sprague-Dawley and Fischer-344 female rats
treated with Atrazine. MRID: 43598622.
McCormick, G.C., Atrhur, AT. and Green, J.D. 1988. Combined Chronic
Toxicity/Oncogencity study in Rats. Ciba-Geigy Corp.. Laboratory study no. 852004.
MRID 40614405.
McMartin, D., Sahota, P.S., Gunson, D.E., Hsu, H., and Spaet, R.H. 1992. Neoplasms
and Related Proliferative Lesions in Control Sprague-Dawley Rats from Carcinogenicity
Studies. Historical Data and Diagnostic considerations.fox. Path .: 20(2), 212-225.
Means, J.C., Plewa, M.J. & Gentile, J.M. 1988. Assessment of the mutagenicity of
fractions from s-triazine-treated Zea mays. Mut. Res.: 197: 325-336.
Meares, E.M., Prostatitis and related disorders. IrrWalsh, P.C. (ed), Campbell's
Urology, 7th ed. Philidelphia,: Saunders; 1998: 615-630.
Meisner, L.F., Belluck, D.A. & Roloff, B.D. 1992. Cytogenetic effects of alachlor and/or
atrazine in vivo and in vitro. Environ. Mol. Mutag.: 19: 77-82.
Meisner, L.F., Roloff, B.D. & Belluck, D.A. 1993. In vitro effects of N-nitrosoatrazine on
chromosome breakage. Arch. Environ. Contam. Toxicol.: 24; 108-112.
Ref-19
-------
DRAFT: DO NOT CITE OR QUOTE
Meites, J. 1972. Relation of prolactin and estrogen to mammary tumorigenesis in the
rat. JNCI; 48:1217-1224.
Meites, J., Steger, R.W., and Huang, H.H. 1980. Relation of neuroendocrine system to
the reproductive decline in aging rats and human subjects. Fed. Proc.: 39:3168-3172.
Merson, J., Sail, W., Mitchner, N. and Ben-Jonathan, N. 1995 Prolactin is a local
growth factor in rat mammary tumors. Endocrinology 136:3619-3623.
Mersch-Sundermann, V., Dickgiesser, N., Hablizel, Y. & Gruber, B. 1988. Examination
of mutagenicity of organic microcontaiminations on the environment. 1. Communication:
The mutagenicity of selected herbicides and insecticides with the Salmonella
microsome-test (Ames test) in consideration of the pathogenetic potence of
contaminated ground- and drinking-water. Zbl. Bakt. Hyg. L Abt. On'g. B: 186: 247-260.
Meyer, A. 1987. Tests for other genotoxic effects. Autoradiographic DMA repair test in
human fibroblasts. Ciba-Geigy, Greenboro, NC.. Study No. 871371. November 20,
1987. MRID 40722303.
Meyer, A. 1988. Hydroxyatrazine: Tests for other genotoxic effects, autoradiographic
DNA repair tests on human fibroblasts. Ciba-Geigy, Basle, Switzerland . Study No.
871375. January 11,1988. Unpublished. MRID 40888101.
Mills, P.K. 1998. Correlation analysis of pesticide use data and cancer incidence rates
in California counties. Arch. Eviron. Health. 53(6): 410-413.
Minnema, D.J. 2000. Comparison of the LH surge in female rats administered atrazine,
simazine, or DACT via oral gavage for one month. Covance Laboratories, Vienna VA.
Study no. 6117-398. MRID 45058701.
Mobbs, C.V. 1996. Neuroendocrinology of aging. In: The Biology of Aging. Eds.
Schneider, E. L. and Rowe, J. W. Academic Press, San Diego, Boston, New York.
Molina, M., Prabhaker, S. and Harbans, J. 1981. Effect of Varying Photoperiods on
Mammary Morphology, DNA Synthesis, and Hormone Profile in Female Rats. JNCI:
Vol. 72 (6) pp: 1411-1416.
Ref-20
-------
DRAFT: DO NOT CITE OR QUOTE
Mitchell AD, Auletta AE, Clive d, Kirby PE, Moore MM. And BC Myhr. 1997 The
L5178Y/tkv+mouse lymphoma specific gene and chromosomal mutation assay: A phase
III report of the U.S. Environmental Protection Agency Gene-Tox Program. Mutation
Research 394:177-303.
Mora, O.A., Canberra, M.M., and Sanchez-Criado, J.E. 1994a. Pheromonal male-
induced diestrus and cyclicity in aging intact and young estrogenized female rats. Bio.
Repro.: 50:603-606.
Mora, O.A., Canberra, M.M., and Sanchez-Criado, J.E. 1994b. Hormonal pattern of the
pheromonal restoration of cyclic activity in aging irregularly cycling and persistent-estrus
female rats. Bio. Repro.: 51:920-925.
Morgun, V.V., Logvinenko, V.F., Merezhinskii, T.G. & Lapina, T.V. 1982. Cytogenetic
and genetic activity of the herbicides atrazine, simazin, prometrin, and linuron. Cytol.
Genet. (USSR): 16:42-45.
Monosson, E., Kelce, W.R., Lambright, C., Ostby, J., and Gray, L.E., Jr. 1999.
Peripubertal exposure to the antiandrogenic fungicide, vinclozolin, delays puberty,
inhibits the development of androgen-dependent tissues, and alters androgen receptor
function in the male rat [In Process Citation]. Toxicol. Ind. Health 15:65-79.
Morichetti, E., Delia Grace, C., Rosellini, D., Bronzetti, G. & Soldani G. 1992. Genetic
and biochemical studies on a commercial preparation of atrazine. Toxicol. Environ.
Chem.: 37:35-41.
Morseth, S. 1996a. Evaluation of the lutenizing hormone (LH) surge in atrazine
exposed female Sprague-Dawley rats. Covance Laboratories Inc. Vienna, VA.
Laboratory study no. CHV 2386-111. January 25, 1996. MRID 43934406.
Morseth, S. 1996b. Evaluation of the lutenizing hormone (LH) surge in atrazine
exposed female Sprague-Dawley rats - 6 month report. Covance Laboratories Inc.
Vienna, VA. Laboratory study no. CHV 2386-111. October 25, 1996. MRID 44152102.
Morseth, S. 1998. Chronic (12-24 month) study in rats with atrazine technical. Covance
Laboratories Inc. Vienna, VA. Study No. 2386-108. April 15,1998. MRID 44544701.
Ref-21
-------
DRAFT: DO NOT CITE OR QUOTE
Mortensen, P.B. 1994. The occurrence of cancer in first admitted schizophrenic
patients. Schizophr. Res. 12:185-194.
Muller, A., Ebert, E., & Gast, A. 1972. Cytogenetic studies with atrazine
(2-chloro-4-ethyl-amino-6- isopropylamino-s-triazine) on plants. Experientia, 28,
704-705
Munger, R., Isacson, P., Hu, S., Burns, T, Hanson, J., Lynch, C.F., Cherryholmes, K.,
Van Dorpe, P., and Hausier, W.J.. 1997. Intrauterine growth retardation in Iowa
communites with herbicide-contaminated drinking water supplies. Environ. Health
Persp. 105(3): 308-314.
Murnik, M.R. and Nash, C.L. 1977. Mutagencity of the triazine herbicides atrazine,
cyanazine, and simazine in Drosophila Melanogastor. J. Tox. Environ. Health: 3:691-
697.
Nandi, S., Guzman, R.C., and Yang, J. 1995. Hormones and mammary
carcinogenesis in mice, rats, and humans: a unifying hypothesis. PNAS: 25;92(9) 3650-
7.
Narotsky, M.G., Guidici, D.L., Best, D.S., Bielmeier, S.R., and Cooper, R.L. 1999.
Strain comparisons of susceptibility to pregnancy loss in rats. The Endocrine Society
Program & Abstracts, 81st Annual Meeting. Endocrine Society Press, Bethesda, MD.
Page 492.
Narotsky, M.G., Best, D.S., Guidici, D.L. and Cooper, R.L. (Submitted) Strain
comparisons of atrazine-induced pregnancy loss in the rat.
Nazan, S.J. and Mahesh. 1980. Hypothalmic, pituitary, testicular, and secondary organ
functions and interactions during the sexual maturation of the male rat. Arch. Andro.
4:283-303.
Nelson, J.F., Felicio, L.S., Sinha, Y.N., and Finch, C.E. 1980. An ovarian role in
spontaneous pituitary tumorigenesis. Gerontologist. 20:171.
Neumann, D.A., Crisp, T.M. and Olin, S.S. 1996. Mammary gland neoplasia. Environ.
Health Perspect. 104:912-914.
Ref-22
-------
DRAFT: DO NOT CITE OR QUOTE
Njagi, G.D.E., Oyolla, F., Hongo-Allego, P. and Goplan, H.N.B. 1980. Mutagenicity
testing of herbicides, fungicides and insecticides, III. Drosophila: rescessive sex-linked
lethals. Bangladesh J. Zoo:. 8(1): 1-4
Nishihara M, Minima H, Kimura F. 1992 . Generating Mechanism of Pulsatile
Gonadotropin Release, pp. 69-91. In, Brain Control of the Reproductive System.
Yokoyama A, ed. CRC Press, Boca Raton.
NRA, 1997.The NRA review of atrazine. Existing Chemical Review Program National
Registration Authority for Agricultural and Vetinary Chemicals of Australia, Canberra,
Australia. Submitted to U.S. EPA as MRID 44597607
O'Conner, D.J. 1988. Combined Chronic Toxicity/Oncogenicity study in Rats. Ciba-
Geigy study no. 852003. MRID 404881-02.
O'Conner, D.J., McCormick, G.C., Green, J.D. 1987. Atrazine technical -53 week
feeding study in dogs. Research Dept. Ciba-Geigy. Study no. 852008. MRID 40431301.
Odell, W.D., Swerdloff, R.S., Jacobs, H.S., and Hescox, M.A. 1973. FSH induction of
sensitivity to LH: one cause of sexual maturation in the male rat. Endocrinology
92:160-5.
Ogorek, B. 1991 a. Structural chromosomal aberration test-micronucleus, mouse (G-
28279). Ciba-Geigy, Greenboro, NC.. Study No. 901307. February 23,1991. MRID
43093103.
Ogorek, B. 1991b. Structural chromosomal aberration test-micronucleus, mouse (G-
30033). Ciba-Geigy, Greensboro, NC.. Study No. 901309. March 25,1991. MRID
43093104
Ojeda and McCann. 1974. Development of dopaminergic and estrogenic control of
prolactin release in the female rat. Endocrinology. 95:1499-1505.
Ojeda, S.R, Advis, J.P., Andrews, W.W. 1980. Neuroendocrine control of the onset of
puberty in the rat. Fed. Proc. 39(7): 2365-2371.
Ojeda, S.R., Aguado, LI., Smith, S. 1983. Neuroendocrine mechanisms controlling the
onset of female puberty: The rat as a model. Neuroendo. 37: 306-313.
Ref-23
-------
DRAFT: DO NOT CITE OR QUOTE
Ojeda, S.R., Urbanski, H.F. and Ahmed, C.E. 1986. The onset of female puberty:
studies in the rat. Recent Prog. Horn. Res. 42: 385-442.
Ojeda, S.R. and Urbanski, H.F. 1994. Puberty in the rat. In: E. Knobil and J.D. Neill
(eds.), The Physiology of Reproduction. 2nd Ed. Raven Press. New York, pp. 363-409.
Perkins, R.B., Hall, J.E., Martin, K.A. 1999. Neuroendocrine abnormalities in
hypothalamic amenorrhea: spectrum, stability, and response to neurotransmitter
modulation. J. Clin. Endocrinol. Metab. 84(6): 1905-11.
Pettersen, J.C., Richter, A.D., Gilles, P.A. 1991. 90-day oral toxicity study in rats. Ciba-
Geigy Environmental Health Center, Farmington, CT. Study No. F-0006. MRID
43013207.
Pettersen, J.C. and Turnier, J.C. 1995. One-year chronic toxicity study with atrazine
technical in rats. Ciba-Geigy Environmental Health Center, Farmington, CT. Lab Study
no. F-00171. MRID 43934402.
Plant, T.M. 1994. Puberty in Primates, in: The Physiology of Reproduction, Second
Edition, Edited by E. Knobil and J.D. Neill, Rave Press, New York, Chapter 42, pages
453-477.
Plewa, MT & Gentile, J.M. 1976. Mutagenicity of atrazine: A maize-microbe bioassay.
Mut. Res.: 38, 287-292.
Plewa, MT 1984. Plant genetic assays and their use in studies on environmental
mutagenesis in developing countries. Basic Life Sci.: 34: 249-268.
Piacsek, B.E. and Goodspeed, M.P. 1978. Maturation of the pituitary-gonadal system in
the male rat. J. Reprod. Pert. 52:29-35.
Pierpont, T., McKeigue, P.M., Isaacs, A.J., Wild, S.H., and Jacobs, H.S. 1998.
Mortality of women with polycystic ovary syndrome at long-term follow-up. J. Clin.
Epidemiol 51: 581-586.
Pino, A., Muara, A. & Grille, P. 1988. DNA damage in stomach, kidney, liver and lung
of rats treated with atrazine. Mut. Res.: 209,145-147.
Ref-24
-------
DRAFT: DO NOT CITE OR QUOTE
Pinter, A., Torek, G., Borzonyi, M., Surjan, A., Csik, M., Kelecsenyi, Z., and Kocsis, Z.
1990. Long-term carcinogencity bioassay of the herbicide atrazine in F-344 rats.
Neoplasms: 37(5): 533- 544
Poole, D. and Simmon, V.F. 1977. Final report: In vitro and in vivo microbiological
assays of six Ciba-Geigy chemicals. Stanford Research Inst. Study no. not applicable.
Dec 29,1977. MRID 00606402.
Price, D. 1963. Comparative aspects of development and structure in the prostate. Nat.
Cancer Inst. Monogr. 12:351-369.
Puri, E. 1984. Autoradiographic DNA repair test on rat hepatocytes. Ciba-Geigy. Basle,
Switzerland. Study No. 831371. May 16,1984. MRID 00150623.
Puri, E. and Muller, D. 1984. Autoradiographic DNA repair test on rat hepatocytes with
G 30027, technical. Ciba-Geigy. Basle, Switzerland. Study No. 8311771. May 16,
1984. MRID 40246601/Ace no. 284052.
Puri, E. and Muller, D. 1984. Autoradiographic DNA repair test on rat hepatocytes with
G 30027, technical. Ciba-Geigy. Basle, Switzerland. Study No. 8311771. May 16,
1984. MRID 40246601/Ace no. 284052.
Quadri SK, Kledzik GS, Meites J. 1973. Effects of L-dopa and methyldopa on growth of
mammary cancers in rats. Proc. Soc. Exp. Biol. Med. 142(3):759-61.
Rayburn, A.L., and Biradar, D.P. 1995. Flow cyotgenic analysis of whole cell
clastogencity of herbicides found in groundwater. Arch. Environ.. Contam.Tox.: 28:13-
18.
Reame, N.E., Sauder, S.E., Case, G. D., Kelch, R.P. and Marshall, J.C. 1985. Pulstile
gonadotropin secretion in women with hypothalamic amenorrhea: Evidence that
reduced frequency of gonadotropin-releasing hormone secretion is the mechanism of
persistent anovulation. J. Clin. Endo. Metab. 61: 851-858.
Ribas, G., Frenzilli, G., Barale, R., and Marcos, R. 1995. Herbicide-induced DNA
damage in human lymphocytes evaluated by the single-cell gel electrophoresis (SCGE)
assay. Mut. Res. 344:41-54.
Ref-25
-------
DRAFT: DO NOT CITE OR QUOTE
Roloff, B.D., Belluck, DA and Meisner, L.F. 1992. Cytogenetic studies of herbicide
interactions in vitro and in vivo using Atrazine and Linuron. Arch.. Environ. Contam. Tox.
22:267-71.
Rose, P.G. 1996. Endometrial carcinoma. N. Engl. J. Med. 335: 640-649.
Ross, R.K., Paganini-Hill, A., Krailo, M.D., Gerkins, V.R., Henderson, B.E. and Pike,
M.C. 1984 Effects of reserpine on prolactin levels and incidence of breast cancer in
postmenopausal women. Cancer Res. 44:3106-3108.
Rothchild, 1.1981. the regulation of the mammalian corpeus luteum. Recent Prog.
Horm. Res. 37:183-298.
Rudzki, M.W., Batastini, G., Atrhur, AT. 1992. Ciba-Geigy Corporation,
Pharmaceuticals Division, Summit, NJ. Study No. MIN 902187. MRID 43013204
Ruiz, M.J. and Marzin, D. 1997. Genotoxicity of six pesticides by Salmonella
mutagenicity test and SOS chromotest. Mut. Res.: 390 (3): 245-255.
Russo, J. and Russo, I. 1996. Experimentally induced mammary tumors in rats. Breast
Can. Res. and Treat: 39:7-20.
Russo, J., Gusterson, B.A., Rogers, A.E., Russo, I.H., Wellings, S.R., and Van Zwieten,
J. 1990. Comparative study of human and rat mammary tumorigenesis. Lab Invest:
62:119-122.
Safe, S., Chen, I., Liu, H. and Zacharewski, T. 1995. Failure of atrazine and simazine
to induce estrogenic responses in MCF-7 human breast cancer cells. Ciba-Geigy,
Greensboro, NC.. Study No. not applicable. March 13, 1995. MRID 43598619.
Safe, S. and Mcdougal, A. 1998. Induction of 16a-/2-hydroxestrone metabolite ratios in
MCF-7 cells by pesticides, carcinogens, and antiestrogens does not predict mammary
carcinogens. Environ.. Health Persp.: 106 (4): 203-206.
Sanderson, J.T., Seinen, W., Giesy, J.P. and van den Berg, M. 2000. 2-chloro-s triazine
herbicides induce aromatase (CYP19) activity in H295R human adrenocortical
carcinoma cell.: A novel mechanism for estrogenicity. Tox. Sci. 54:121-127.
Ref-26
-------
DRAFT: DO NOT CITE OR QUOTE
Sandusky, G.E., Van Pelt, C.S., Todd, G.C. and Wightman, K. 1988. An
immunocytochemical study of pituitary adenomas and focal hype.rplasia in old Sprague-
Dawley and Fisher rats. Tox. Path.: 16: 376-380.
Savitz, D.A., Arbuckle, T. Kaczor, D. and Curtis, K.M. 1997. Male pesticide exposure
and pregnancy outcome. Am. J. Epi. 146(12): 1025-1036.
Sarkar, O.K., Gottschall, P.E., and Meites, J. 1982. Damage to hypothalmic
dopaminergic neurons is associated with development of prolactin-secreting pituitary
tumors. Science; 218: 684-686.
Schacter, M. and Shoham, Z. 1994. Amenorrhea during the reproductive years -is it
safe? Pert. Ster. vol 62, No. 1:1-16.
Schildkraut, J.M., Schwingl, P.J., Bastos, E., Evanhoff, A., Hughes, C. 1996.
Epithelial ovarian cancer risk among women with polycystic ovary syndrome. Obstet.
Gynecol. 88: 554-559.
Schneider, M. 1992. 90-day oral toxicity study in rats. Ciba-Geigy Corporation,
Short/Long-Term Toxicology, Stein, Switzerland. Study no. 901261. MRID 43013205.
Seeds, J.W. 1984. Impaired fetal growth: definition and clincal diagnosis. Obstet.
Gynecol. 64(3): 303-310
Seller, J.P. 1973. A survey on the mutagenicity of various pesticides. Experientia: 29:
622-623.
Schafer, T.J., Ward, T.R., Meachame, C.A. and Cooper, R.L. 1999. Effects of the
chlorotriazine herbicide, cyanazine, on GABAA receptors in cortical tissue from rat brain.
Toxicology 142: 57-68.
Shapiro, S., Parsells, J.L., Rosenberg, L, Kaufman, D.W., Stolley, P.O. and
Schottenfeld, D. 1984. Risk of breast cancer in relation to the use of raulwolfia
alkaloids. Eur. J. Clin. Pharmacol. 26:143-146.
Shirasu, Y. M. Moriya, M., and Ohta, T. 1978. Microbial mutagencity testing on
CP67573 (Glyphosate). Instiute of Environmental Toxicity, Japan. July 20,1978. MRID
00078619.
Ref-27
-------
DRAFT: DO NOT CITE OR QUOTE
Siebert, D. & Lemperle, E. 1974. Genetic effects of herbicides: induction of mitotic
gene conversion in Saccharomyces cerevisiae. Mut. Res., 22,1.11-120
Simmon, V.F., Kauhanen, K. & Tardiff, R.G. 1977. Mutagenic activity of chemicals
identified in drinking water. In: Progress in Genetic Toxicology: Developments in
Toxicology and Environmental Science. Vol.2,. Scott, D., Bridges, B.A. & Sobels,
F.H.,eds. Amsterdam, Elsevier; pp. 249-258.
Simoneaux, B.J. 1995. Nature of atrazine and simazine metabolism in animals and
plants. Ciba-Crop Protection, Greenboro N.C. MRID 43598602.
Simic, B., Kniewald, Z, Davies, J.E., Kniewald, J. 1991. Reversibility of the inhibitory
effect of atrazine and lindane on cytosol 5a- -dihydrotestostrone receptor complex
formation in rat prostate. Bull. Environ. Contam. Toxicol. 46:92-99.
Simpkins, J.W., Eldridge, C. and Wetzel, L.T. 1998. Role of strain-specific reproductive
patterns in the appearance of mammary tumors in atrazine-treated rats. In: Triazine
herbicides risk assessment. (L.G. Ballantine, J.E. McFarland and D.S. Hackett eds.)
American Chemical Society, Washington, DC 399-413.
Shirasu, Y. 1975. Significance of mutagenicity testing on pesticides. Environ. Quality
Safety, 4: 226-231
Shyr, S.W., Crowley, W.R., Grosvenor, C.E. 1986. Effect of neonatl prolactin deficiency
on prepubertal tuberoinfunddibular and tuberohypophyseal dopaminergic activity.
Endocrinology. 119:1217-1221.
Smith, W.A. and Conn, M.P. 1983. Causes and consequenses of altered
gonadatrophin secretion in the aging rat. In: Clincal and Experimental Intervention of
the Aging Process, Cooper, R.L. and Walker, R.F., eds. New York: Marcel Dekker; pp.
27-44.
Snedeker, S.M. and Clark, H. 1999. Critical Evaluation of atrazine's breast cancer risk.
Cornell University Program on Breast Cancer and Environmental risk Factors in New
York State, http://www.cfe.comell.edu/bcerf/>
Solleveld, H.A., Haseman, J.K., and McConnell, E.E. 1984. Natural history of body
weight gain, survival, and neoplasia in the F-344 rat. JNCI: 72 (4): 929-940.
Ref-28
-------
DRAFT: DO NOT CITE OR QUOTE
Solomon, C.G. 1999. The epidemiology of polycystic ovary syndrome: Prevalence
and associated disease risks. Endocrinol. Metabol. Clin. N. Am.. 28: 247-263.
Spencer, H. 1991. Conclusions Drawn From Structure Activity Relationship (SAR)
Analysis of s-Triazine Pesticides and Related compounds From OTS. MEMO: From H.
Spencer and K Baetche, To: J. Andreasen. Dated January 7,1991.
Stevens, J.T. 1994. Hypothesis for mammary tumorigenesis in Sprague-Dawley rats
exposed to certain triazine herbicides. J. Tox. Environ.. Health: 43:139-153. MRID
43598611.
Stanford, J.L., Martin, E.J., Brinton, LA and Hoover, R.N. 1986 Raulwolfia use and
breast cancer: a case-control study. JNCI. 76:817-822.
Stouffer, R.L., Ottobre, J.S., Molskness, T.A. and Zelinski-Wooten, M.B. 1989. The
function and regulation of the primate corpus luteum during the fertile menstrual cycle.
Development of Preimplanatation Embryos and their Environment: 129-142.
Stoker, T.E., Guidici, D.L., McElroy, W.K., Laws, S.C. and Cooper, R.L. 2000a. The
effect of atrazine on puberty in male Wistar rats. The lexicologist, 54 (1) 366
Stoker, T.E., Robbinette, C.L, Cooper, R.L. 2000b. Perinatal exposure to estrogenic
compounds and the subsequent effects on the prostate of the adult rat: Evaluation of
inflammation in the ventral and lateral lobes. In press. Repro. Tox.
Stoker, T.E., Laws, S.C., Guidici, D, and Cooper, R.L. (Submitted) The Effects of
Atrazine on Puberty and Thyroid Function in the Male Wistar Rat: An Evaluation in the
Male Pubertal Protocol.
Stroev, V.S. 1968. Mutagenic effect of herbicides on barley. Sov. Genet: 4,1539-
1542.
Styne, D. 1997. Puberty. In:. Basic and Clinical Endocrinology. 5th ed. Greenspan,
F.S. and Strewler, G.J., eds. Stamford, CT. Appleton & Lange. pp. 521-547
Sumner, D.D., Cassidy, J.E., Szolics, I.M., Marco, G.J., Bakshi, K.S. & Brusick, D.J.
1984. Evaluation of the mutagenic potential of corn (Zea mays L) grown in untreated
and atrazine (AAtrex) treated soil in the field. Drug Chem. Toxicol.: 7, 243-257.
Ref-29
-------
DRAFT: DO NOT CITE OR QUOTE
Surralles, J., Xamena, N., Creus, A. and Marcos, R. 1995. The suitability of the
micronucleus assay in human lymphocytes as a new biomarker of excision repair. Mut.
Res.: 342: 43-59.
Tai, C.N., Breckenridge, C. Green, J.D. 1985a. Subacute oral 13 week toxicity study in
dogs. Ciba-Geigy Corporation, Division of Toxicology/Pathology, Summit, NJ. Study
No.85022. Ace. no. 257692.
Tai, C.N., Breckenridge, C. Green, J.D. 1985b. Simazine techncial: Subacute oral 13
week toxicity study in rats. Ciba-Geigy Corporation, Division of Toxicology/Pathology,
Summit, NJ. Study No.85018. MRID 00143265.
Telang, NT., Katdare, M., Bradlow, H.L., and Osbome, M.P. 1997. Estradiol
metabolism: and endocrine biomarker for modulation of mammary carcinogenesis.
Environ. Health Persp.: 82:6295-6299.
Tennant, M. K.f Hill, D.S., Eldridge, C., Wetzel, L.T., Breckenridge, C.B., and Stevens,
J. 1994a. Possible antiestrogenic properties of chloro-s-triazines in rat uterus. J. Tox.
Environ. Hea/tf743:183-196. MRID 43598617.
Tennant, M. K., Hill, D.S., and Eldridge, C. 1994b. Chloro-s-triazine antagonism of
estrogen action: Limited interaction with estrogen receptor binding. J. Tox. Environ.
Health: 43:197-211. MRID 43598618.
Terasawa E, Nyberg CL. 1997. LHRH Pulse Generation in the Monkey: In Vivo and In
Vitro Studies, pp. 57-70. In, Neural Control of Reproduction: Physiology and Behavior.
Maeda K, Tsukamura H, Yokayama A, eds. Karger, New York
Terkel, J. 1988. Neuroendocrine processes in the establishment of pregnancy and
psuedopregnancy in rats. Psychoneuroendo. 13(1-2): 5-28.
Thakur, A. 1991 a. Determination of Hormone Levels in Sprague-Dawley Rats Treated
with Atrazine Technical. Lab Project Hazleton Washington, Inc . Number:
483-278.MRID 42085001 .
Thakur, A. 1991b. Determination of Hormone levels in Fischer 344 rats treated with
atrazine technical. Hazleton Washington, Inc. Lab Project Number: 483-279. MRID
42146101.
Ref-30
-------
DRAFT: DO NOT CITE OR QUOTE
Thakur, A. 1992a. Oncogenicity Study in Sprague-Dawley Rats with Atrazine
Technical. Hazleton Washington, Inc. Lab Project Number: 483-275. MRID
42204401.
Thakur, A. 1992b. Oncogenicity Study in Fischer 344 Rats with Atrazine Technical.
Hazleton Washington, Inc. Lab Project Number: 483-277. MRID 42227001.
Thakur, A. 1999. Addendum 1 to study no. 2386-108: Chronic (12/24 month) study in
rats with atrazine technical. Unaudited draft interim report of the estrous cycle data
through week 46. Covance Lab study no. 2386-103.
Thompson, S.S., Bastini, G., Arthur, AT. 1990.13/52 week feeding study in dogs.
Ciba-Geigy Corporation, Division of Toxicology/Pathology, Summit, NJ. Study No. MIN
872151. MRID 4392401.
Thompson, S.S., Bastini, G., Arthur, AT. 1992. G-28279:13 week feeding study in
dogs. Ciba-Geigy Corporation, Division of Toxicology/Pathology, Summit, NJ. Study No.
MIN 912021. MRID 43013203.
den Tonkelaar, I and de Waard, F. 1996. Regularity and length of menstrual cycles in
women aged 41-46 in relation to breast cancer risk: Results from the DPOM-project.
Breast Cancer Res. Treat. 38:253-258.
Torres, C., Ribas, G., Xamena, N., Creus, A. & Marcos, R. 1992. Genotoxicity of four
herbicides in the Drosophila wing spot test. Mut. Res.: 280, 291-295.
Tran, D.Q., Kow, K.Y., McLachlan, J.A., Arnold, S.F. 1996. The inhibition of estrgoen
receptor-mediated responses by chloro-s-triazine-derived compounds is dependent on
estradiol concentration in yeast. Bioch. Biophy. Res. Comm. 227:140-146.
Trentacoste, S.V., Friedmann, A.S., Breckenridge, C.B., Zirkin, B.R. (Submitted).
Atrazine effects on testostrone and adrogen-dependent reproductive organs in
peripubertal male rats. Submitted to EPA as MRID 45058702.
Tripathy, N.K., Routray, P.K, Sahu, G.P. & Anandkumar, A. 1993. Atrazine, a triazine
herbicide, is genotoxic in the Drosophila somatic and germ line cells. Biol. Zent. Bl.,
112:312-18.
Ref-31
-------
DRAFT: DO NOT CITE OR QUOTE
United Kingdom Pesticide Directorate, 2000. Atrazine: Report and proposed decison of
the United Kingdom made to the European Commision under Article 7(1) of regulation
3600/92 council directive 91/414/EEC regulation 3600/92. United Kingdom Pesticide
Directorate. Submitted to U.S. EPA as MRID 44315415.
U.S. EPA 1988. Nelson, C.J., Memorandum from C.J. Nelson, Office of Pesticide
Programs Health Effects Division (HED) to Marion Copley, HED. "Atrazine - Updated
qualitative and quantitative risk assesment from a rat 2-year chronic oral
toxicity/oncogencity study." August 23,1988.
U.S. EPA. 1995. Memorandum from John Abbotts, Office of Pesticide Programs
Health Effects Division (HED) to HED Metabolism Committee. "Atrazine (80803),
regregistration case no. 0062, and simazine (80807), reregistration case no. 0070 and
special review. Issues to be presented at the 10/16/95 meeting of the HED Metabolism
Committee." October 10,1995.
U.S. EPA, 1998a. US Environmental Protection Agency, 1998. Assessment of thyroid
follicular cell tumors. Office of Research & Development, Risk Assessment Forum,
EPA/630/R-97/002 March 1998.
U.S. EPA (U.S. Environmental Protection Agency). 1998b. EDSTAC Final Report,
Chap. S.Appendix L, Protocols for Tier I Screening Assays,
http://www.epa.gov/scipol/oscpendo/history/app-1v14.pdf.
U.S. EPA 1999a. Guidelines for carcinogen risk assesment, July 1999 Review DRAFT.
Risk Assesment Forum, Washington, D.C.
http://www.epa.gov/ncea/raf/car2sab/preamble.pdf
U.S. EPA 1999b Brunsman, L.L. Memorandum from Lori Brunsman, Office of
Pesticide Programs Health Effects Division (HED) to Roger Hawks, HED. Atrazine:
EDIO's and LEDIO's Based on Sprague-Dawley Female Rat Dietary Studies.
September 1,1999
U.S. EPA 1999c Brunsman, L.L. Memorandum from Lori Brunsman, Office of
Pesticide Programs Health Effects Division (HED) to Roger Hawks, HED. Atrazine
qualitative risk assesment based on female Sprague-Dawley rat dietary study.
February 9,1999.
Ref-32
-------
DRAFT: DO NOT CITE OR QUOTE
U.S. EPA 1999d Bmnsman, L.L., Memorandum from Lori Brunsman, Office of
Pesticide Programs Health Effects Division (HED) to Roger Hawks, HED. "Atrazine:
Qualitative risk assesment based on female Sprague-Dawley chronic dietary studies
using mg/kg/b.w. 3/4ths cross-species scaling factor." July 7,1999.
Utiger, R.D. 1996. Insulin and the polycystic ovary syndrome. N. Engl. J. Med. 335:
657-658.
Van Leuwen, J.A., Waltner-Toews, D., Abernathy, T., Smit, B. and Shoukri, M. 1999.
Associations between stomach cancer incidence and drinking water contamination with
atrazine and nitrate in Ontario (Canada) agroecosystems, 1987-1991. Inter. J. EpL
28:836-840.
van Putten, L.J.A., van Zwieten, M.J., Mattheij, JAM., and van Kemenade, J.A.M.
1988. Studies on prolactin-secreting cells in aging rats of different strains: I. Alterations
in pituitary histopathology and prolactin levels as related to aging. Mech. Aging Dev.:
42: 75-90.
Vihko, K.K., LaPolt, P.S., Nishimori, K., and Hsueh, A.J. 1991. Stimulatory effects of
recombinant follicle-stimulating hormone on Leydig cell function and spermatogenesis
in immature hypophysectomized rats. Endocrinology 129:1926-32.
Vonderhaar, B.K. 1999 Prolactin involvement in breast cancer. Endocrin. Pel. Cancer
6: 389-404.
Wallace, R.B., Sherman, B.A. and Behn, J.A. 1982. A case control study of breast
cancer and psychotropic drug use. Oncology. 39:279-283.
Walsh, K.M., and Poteracki, J. 1994. Spontaneous neoplasms in control in Wistar rats.
Fun. App. Tox.: 22:65-72.
Wang, D.Y, De Stavola, B.L, Bulbrook, R.D. et al. 1992 Relationship of blood prolactin
levels and the risk of subsequent breast cancer. Int. J. Epidemiol. 21: 214-221.
Welsch, C.W., Jenkins, T.W., and Meites, J. 1970a. Increased incidence of mammary
tumors in the female rat grafted with mulitple piturtaries. Cancer Res.; 30:1024-1029.
Warren, M.P.; Stiehl, A.L 1999. Exercise and female adolescents: effects on the
reproductive and skeletal systems. J. Am. Med .Womens Assoc. Summer;54(3):115-20
Ref-33
-------
DRAFT: DO NOT CITE OR QUOTE
Welsch, C.W., Jenkins, T.W., and Meites, J. 1970a. Increased incidence of mammary
tumors in the female rat grafted with multiple pituitaries. Cancer Res. 30:1024 -1029.
Welsch, C.W., Nagasawa, H., and Meites, J. 1970b. Increased incidence of
spontaneous mammary tumors in female rats with induced hypothalmic lesions. Cancer
Res.; 30: 2310-2313.
Welsch, C.W. 1985. Host factors affecting the growth of carcinogen induced rat
mammary carcinomas: A review and tribute to Charles Brenton Muggins. Cancer Res.
45:3415-3443
Welsch C. W. 1987. Rodent Models to Examine In Vivo Hormonal Regulation of
Mammary Gland Tumorigenesis. In: Cellular and Molecular Biology of Mammary
Cancer. D. Medina, W. Kid well, G. Heppener, and E. Anderson, Eds. Plenum Press,
New York, 1987, p. 163-179
Weisenberger, D.D., Joshi, S.S., Hickman, D.M., Babcock, D.M., Walker, B.A. and
Mirvish, S.S. 1987. N-Nitrosoatrazine (NNAT) synthesis, chemical properties, acute
toxicity, and mutagenicity. Proc. Am. Assoc. Cancer Res. 28:103
Weisenberger, D.D., Joshi, S.S., Hickman, D.M., Babcock, D.M., Walker, B.A. and
Mirvish, S.S. 1988. N-Nitrosoatrazine (NNAT) synthesis, chemical properties, acute
toxicity, and mutagenicity. Proc. Am. Assoc. Cancer Res. 29:106
Weisenberger, D.D., Joshi, S.S., Hickman, D.M., Patil, K.D., Lawson, T.A. and Mirvish,
S.S. 1990. Carinogenesis tests of atrazine and N-nitrosoatrazine - compounds of
special interest to the midwest. Proc. Am. Assoc. Cancer Res. 31:102
Wetzel, L.T., Luempert, L.G., Breckenridge, C.B., Tisdel, M.O., Stevens, J.T., Thakur,
A.K., Exstrom, P.J. and Eldridge, C.J. 1994. Chronic effects of atrazine on estrus and
mammary tumros formation in female Sprague-Dawely and Fischer 344 rats. J. Tox.
Environ. Health 43:169-182.
Williams, R.R., Feinleib, M., Connor, R.J. and Stegens, N.L. 1978 Case-control study
of antihypertensive and diuretic use by women with malignant and benign breast
lesions detected in a mammography screening program. JNCI 61: 327-335.
Ref-34
-------
DRAFT: DO NOT CITE OR QUOTE
Wise, P.M., Kashon, Ml., Krajanak, K.M., Rosewell, K.L., Cai, A., Scarbrough, K.,
Harney, J.P., McShane, T., Lloyd, J.M. and Weiland, N.G. 1997. Aging of the female
reproductive system: a window into brain aging. Recent Prog. Harm. Res. 52:279-
303.
Wise, P.M., Krajnak, K.M., and Kashon, M.L 1996. Menopause: the aging of multiple
pacemakers. Science 273:67-70.
Wise, P.M. 1999. Neuroendocrine modulation of the "menopause": insights into the
aging brain. AJP-Endocrinol. Metab. 277: E965-E970.
Wolfe, N.L., Zepp, R.G., Gordan, J.A., and Fincher, R.C. 1976. N-nitrosoatrazine
formation from atrazine. Bull. Environ. Contam. Toxicol. 15: 342-347.
Wren, 1997. "Progress in the Management of the Menopause", Barry G. Wren, ed.
1997, Parthenon Publishing Group, Inc., New York, 489 pages.
Wuu, K.D. & Grant, W.F. 1966. Morphological and somatic chromosomal aberrations
induced by pesticides in barley (Hordeum vulgare). Can. J. Genet. Cytol.: 8, 481-501.
Wuu, K.D. & Grant, W.F. 1967a. Chromosomal aberrations induced by pesticides in
meiotic cells of barley. Cytologia: 32, 31-41.
Zahm, S.H, Weisenburger, D.D., Saal, R.C, Vaught, J.B., Babbit, P.A. and Blair, A.
1993a. The role of agricultural pesicide use in the development of non-Hodgkin's
lymphoma in women. Arch. Environ. Health. 48 (5) 353-358.
Zahm, S.H, Weisenburger, D.D. , Cantor, K. P., Holmes, F.F., Blair, A. 1993b. Role of
the herbicide atrazine in the development of non-Hodgkin's lymphoma . Scand. J. Work
Environ. Health. 19:108-114.
Zeiger, E., Anderson, B., Haworth, S., Lawlor, T. & Mortelmans, K. 1988. Salmonella
mutagenicity tests: IV. Results from the testing of 300 chemicals. Environ. Mot. Mutag.:
11:1-158.
Ref-35
-------
DRAFT: DO NOT CITE OR QUOTE
Zuo, Z., Mahesh, V., Zamorano, P. and Brann, D. 1996. Decreased gonadotrophin-
relasing hormone neurosecretory response to glutamate agonists in middle-aged
female rats on proestrous afternoon: a possible role in reproductive aging?
Endocrinology. 137:2334-2338.
Ref-36
-------
DRAFT: DO NOT CITE OR QUOTE
Appendices
-------
DRAFT: DO NOT CITE OR QUOTE
Appendix Table 1: Summary of the Atrazine Two Year and One Year Bioassays Using the SD Strain of Rat
Study
Mayhew.
etal.,
1986
Thakur,
1991 a
Thakur,
1992a
Morseth,
19981
Duration
2 year
2- year
with serial
sacrifices
2- year
2 year
Doses Tested
0. 10. 70 , 500 or 1000 ppm
(0, 0.5. 3.5. 25 or 50
mg/kg/day)
0. 70 and 400 ppm
( 0. 4.23 and 26.23
mg/kg/day)
0. 70 and 400 ppm
(0. 3.79 and 23.01
mg/kg/day)
0, 25, 50. 70 and 400 ppm
(0.1.5,3.1.4.2.24.4
mg/kg/day )
Female
Mammary
Fibroadenom
a Incidence
(doses are in
ppm)
0=23%; 10=
37%; 70= 30%;
500= 31%;
1000=22 %
0=1 1.6%; 70=
17.9%;
400= 18.8%
0= 65%; 70=51
%;
400= 68.3%
0=21%; 25=
32%; 50=44 %
70=37 %;
400=32%
Fibroadenoma P
Values Adjusted for
Survival (Trend is
indicated at control)
0=0.446; 10=0.110.
70= 0 373 ;
500=0.373:1000=
0.468
0=0.484 ; 70=0.213;
400=0.0842
0=not meaningful ;
70=0 914; 400=0.1072
0=0.23; 25=0.03;
50=0.00;
70=0.014:400=0.014
Female
Mammary
Carcinom
a
Incidence
(doses are
in ppm)
0= 17%;
10=24%;
70=39 %;
500= 40%;
1000=51 %
0=13%;
70=6%,
400=15.9%
0=28% ;
70=22%;
400=33.6%
0= 15% ;
25=22%
50=25% ;
70=18% ;
400= 34%
Carcinoma P
Values Adjusted for
Survival (Trend is
Indicated at
control)
0=0.00 ; 10=0.39;
70= 0 024 ;
500=0.019:1000=
0.000
0=0092:70=0254;
400=0.61 92
0= not meaningful;
70=0.832;
400=0 1592
0=0.002;25=0.112;
50=0.067; 70=0.395;
400=0.007
A-2
-------
DRAFT: DO NOT CITE OR QUOTE
Study
Pettersen,
and
Turnier,
19953
Duration
1 year
Doses Tested
0.15,30,50, 70, or 400
ppm
(0,0.8, 1.7, 2 8, 4.1, or 23 9
mg/kg/day)
Female
Mammary
Flbroadenom
a Incidence
(doses are in
ppm)
0=5.9%; 15=
5.9% 30= 5.9%;
50=0 %; 70=
11.3%;
400= 1 1 9%
Fibroadenoma P
Values Adjusted for
Survival (Trend is
Indicated at control)
Survival was similar
between groups. Thus, a
survival adjusted
analysis is not
meaingful.
Female
Mammary
Carcinom
a
Incidence
(doses are
in ppm)
0= 2.9%;
15=2.9 %
30=2 9%;
50= 5 9%;
70=2.9%;
400= 17.1%
Carcinoma P
Values Adjusted for
Survival (Trend is
indicated at
control)
Survival was similar
between groups. Thus,
a survival adjusted
analysis is not
meaingful.
1This study employed both ovarectomized and intact animals. However, no ovarectomized animal was found to have a mammary tumor Thus, results shown are
for intact animals only.
2 Based on Cox-Tarone .
3 Incidence rates are based on 9 and 12 month timepoints only
A-3
-------
DRAFT: DO NOT CITE OR QUOTE
Appendix Table 2: The Thakur Study Design
Type of study
Serial
Sacrifice
Terminal
Sacrifice
Estrous cycle
and hormone
evaluations
Sacrifice
at 1,3, 9. 12, 15,
18 and 24 months
after 24 months
serial sacrifices
Strain
one study with
SO
one study with F-
344
one study with
SD
one study with F-
344
one study with
SD
one study with F-
344
Sex
Females only in
both strains
Females only for
SD
both sexes for F-
344
Females only
Doses
0,1 0.70, 200 and 400
ppm for F-344
0, 70 and 400 ppm for
SD
0.10. 70, 200 and 400
ppm for F-344
0, 70 and 400 ppm for
SD
all doses
MRID
42085001 -SD
42146101 - F-344
42204401 - SD
42227001 - F-344
42743903 - F-344
42743902 - SD
A-4
-------
DRAFT: DO NOT CITE OR QUOTE
Appendix Table 3: Tumor Incidence by timepoints in Thakur, 1991a.
Values shown are the number of rats with that type of mammary tumor.
Mammary tumors found in animals which died an unscheduled death are
included in the data for the timepoint which immediately followed the animals
death.
1 month
3 month
9 month
12 month
15 month
18 month
24 month
0-1 2 month
total
0-24 month
total
Control
no mammary tumors of any
type
no mammary tumors of any
type
no mammary tumors of any
type
Fibroadenomas= 1
Carcincmas=0
Fibroadenomas= 2
Carcinornas= 2
Fibroadenomas=2
Carcinomas=5
Fibroadenomas=3
Carcinomas=2
Fibroadenomas= 1
Carcinomas= 0
Fibroadenomas= 8
Carcinomas2 9
4.23 mg/kg/day
no mammary tumors of
any type
no mammary tumors of
any type
no mammary tumors of
any type
Fibroadenomas=0
Carcinomas=1
Fibroadenomas=5
Cardnomas=0
Fibroadenomas=4
Carcinomas=2
Fibroadenomas=3
Carcinomas=1
Fibroadenomas= 0
Carcinomas= 1
Fibroadenomas= 12
Carcinomas^ 4
26.63 mg/kg/day
no mammary tumors of any
type
no mammary tumors of any
type
Fibroadenomas= 2
Carcinomas=4
Fibroadenomas= 2
Carcinomas=2
Fibroadenomas=1
Caranomas=1
Fibroadenomas=4
Carcinomas=4
Fibroadenomas=4
Caranomas=0
Fibroadenomas= 4
Carcinomas^ 6
Fibroadenomas= 13
Carcinomas= 1 1
A-5
-------
DRAFT: DO NOT CITE OR QUOTE
Appendix Table 4: All 20 Guideline studies with Atrazine, Propazine, Simazine or Atrazine mammalian
metabolites that have been submitted and found acceptable by HEP
Chemical
Atrazine
Atrazine
Atrazine
Atrazine
Atrazine
Atrazine
Study Type
Ames Test
Ames Test
Unscheduled DMA
Synthesis
Unscheduled DNA
Synthesis
Micro-nucleus Assay
Dominant Lethal Assay
Endpolnt
Gene Mutation
Gene Mutation
DNA Damage
DNA Damage
Clastogenicity
(Chromosomal
aberrations)
Genotoxicity to
germinal tissue (
effects which
cause embry. or
fetal death)
Cell type/Species
TA 98. 100. 1535. 1537 and
1538
TA98, 100. 1535 and 1537
Isolated rat hepatocytes
strain- Tif:RAIf
Isolated rat hepatocytes
strain- Tif:RAIf
Mouse
Strain-Tif:MAGf
Mouse
Strain-Tif:MAGf
Results
Negative when tested with and w/o
activation up to the limit dose of 5000
ug/plate
Negative when tested with and w/o
activation up to the limit dose of 5000
ug/plate
Negative when tested up to the solubilty
limit
Negative when tested up to the solubilty
limit
Negative up to a dose causing death in
the mouse
Negative when tested in doses which
induced toxicity
MRID
0060642
40246601
00161790/
40246602
42547105
40722301
42637003
A-6
-------
Appendix Table 4 cont.
Chemical
Simazine
Simazine
Simazine
Propazine
Propazine
Propazine
DACT
DACT
G-28279
G-28279
Study Type
Ames Test
Unscheduled DNA
Synthesis
Micronucleus Assay
In vitro mammalian cell
gene mutation assay
Unscheduled DNA
Synthesis
Micronucleus Assay
Ames Test
Unscheduled DNA
Synthesis
Ames Test
Unscheduled DNA
Synthesis
Endpolnt
Gene Mutation
DNA Damage
Clastogenicity
(Chromosomal
aberrations)
Gene Mutation
DNA Damage
Clastogenicity
(Chromosomal
aberrations)
Gene Mutation
DNA Damage
Gene Mutation
DNA Damage
Cell type/Species
TA 98, 100. 1535, 1537, and
1538
Isolated rat hepatocytes
strain- Tif:RAIf
Mouse
strain-Tif:MAGf
V79 cell line -
Chinese Hamster, fibroblast -
like
Isolated rat hepatocytes
strain- Tif:RAIf
Female Chinese Hamsters
TA98, 100, 1535, and 1537
CRL 1521 cell line -human
fibroblast - like
TA98. 100, 1535. and 1537
Isolated rat hepatocytes
strain- Tif:RAIf
Results
Negative both with and w/o activation
when tested up to the solubility limit
Negative when tested up to the solubilty
limit
Negative up to the limit dose of 5000
mg/kg
Positive dose-related response (5-23x
background) w/o activation at 800 and
1000ug/ml
Weak (5x background and non-dose
related) mutagenic response with
activation at 2000 ug/ml
Negative when tested up to the solubilty
limit
Negative up to the limit dose of 5000
mg/kg
Negative with and w/o activation up to
the limit dose of 5000 ug/plate
Negative without activation only when
tested up to solubilty limits
Negative when tested with and w/o
activation up to the limit dose of 5000
ug/plate
Negative up the cytotoxic dose of 800
ug/ml
MRID
40614406
4144902
41442901
0016322
00150623
00150622
40722302
40722303
43049101
43049105
A-7
-------
Appendix Table 4 cont.
Chemical
G-28279
G-30033
G-30033
G-30033
Study Type
Micronucleus Assay
Ames Test
Unscheduled DMA
Synthesis
Micronucleus Assay
Endpoint
Clastogenicity
(Chromosomal
aberrations)
Gene Mutation
DNA Damage
Clastogenicity
(Chromosomal
aberrations)
Cell type/Species
Mouse
strain-Tif:MAGf
TA 98. 100.1535, and 1537
Isolated rat hepatocytes
strain- Tif.RAIf
Mouse
strain-Tif:MAGf
Results
Negative up to the maximum tolerated
dose of 480 mg/kg
Negative when tested with and w/o
activation up to the limit dose of 5000
ug/plate
Negative in doses up to the cytotoxic
dose of 1000 ug/ml
Negative up to the maximum tolerated
dose of 480 mg/kg
MRID
43093103
43093102
43093106
43903104
A-8
-------
AppcnHI* Tahlo «>• Datahaen fnr «ho Ocnnfoyjftity of AtrayJHB*
Test system
Results"
Without With
Exogenous Exogenous
Metabolic Metabolic
Artiuatinn Artiuatinn
Dose"
(LED or HID)
Reference"
MUTATION
Bacteriophage T4, forward mutation
Bacteriophage, reverse mutation
Salmonella typhimunum, forward mutation, 6AGR
Salmonella typhimurium TA100, TA98, TA1535, TA1 537, TA1 538 reverse mutation
Salmonella typhimurium TA100, TA98. TA1 535 reverse mutation
Salmonella typhimurium TA100, TA98 reverse mutation NT
Salmonella typhimurium TA100, TM677 reverse mutation NT
Salmonella rypft/munum TA100, TA98, TA97, TA1535, TA1537, TA1539 reverse
mutation
Salmonella typhimurium TA100, reverse mutation NT
Salmonella typhimurium TA100, TA98, TA97, TA1 02 reverse mutation
Salmonella typhimurium TA100, TA98, TA97, TA1535, TA1537. TA1538 reverse
mutation
Salmonella typhimurium TA100, TA98, TA97 reverse mutation
Salmonella typhimurium TA100, TA98, TA102, TA1535. TA1537 reverse mutation
Salmonella typhimunum TA1 00. TA98, TA1 535, TA1 537 reverse mutation
Salmonella typhimurium TA1 00, TA98 reverse mutation
Salmonella typhimurium TA1530, TA1531, TA1532, TA1534, his G45 reverse mutation
(spot test)
Salmonella typhimunum, (eight unidentified strains) reverse mutation
Salmonella typhimurium, (strains not identified) reverse mutation
Escherichia coli, forward mutation, AMP"
Saccharomyces cerevisiae, reverse mutation (stationary phase cells)
Saccharomyces cerevisiae, reverse mutation (logarithmic phase cells) (+)
Saccharomyces cerevisiae, forward mutation
Schizosaccharomyces pombe, reverse mutation +
Schizosaccharomyces pombe, reverse mutation +
Aspergillus nidulans, forward mutation
Gene mutation, Chinese hamster lung V79 cells, nprt locus
Host-mediated assay, Escherichia coli AmpR in mouse
Drosophila melanogaster, somatic mutation +
Drosophila melanogaster, somatic mutation +
Drosophila melanogaster, sex-linked recessive lethal mutation (+), I6
Drosophila melanogaster, sex-linked recessive lethal mutation
Drosophila melanogaster, sex-linked recessive lethal mutation (+), IE
Drosophila melanogaster, dominant lethal mutation (+) IG,
Drosophila melanogaster, sex-linked recessive lethal mutation
NT
NT
-
-
-
-
_c
-
+=
-
-
NT
-
-
-
NT
NT
-
-
NT
NT
NT
NT
+c
+
_e
+
20 ug/plate
1000 ug/plate
250 ug/ml
5000 ug/plate
100 ug/plate
11 00 ug/plate
30000 ug/plate
1000 ug/plate
NG
1000 ug/plate
1000 ug/plate
2000 ug/plate
1000 ug/plate
5000 ug/plate
1000 ug/plate
NG
NG
NG
430 ug/plate
75600 ug/ml
2160 ug/ml
50 ug/ml
1 7.5 ug/ml
70 ug/ml
2500 ug/ml
2000 ug/ml
100pox1
1000 ug/g feed
200 ug/g feed
100 ug/g feed
2000 ug/g feed
200 ug/g feed
100 ug/g feed
Andersen era/. (1972)
Andersen et al. (1972)
Adler (1980)
Poole & Simmon (1977; DER)
Lusbyefa/ (1979)
Bartschefa/. (1980)
Sumnerera/ (1984)
Kappas(1988)
Means era/ (1988)
Mersch-Sundermann et al (1988)
Zeiger era/, (1988)
Butler &Hoagland (1989)
Ruiz&Marzin (1997)
Deparde(1986; DER)
Morichetti etal. (1992)
Seller (1973)
Andersen et al ( 972)
Adler (1980)
Adler, (1980)
Morichetti et al. (1992)
Morichetti et. al. (1992)
Emnovaera/ (1987)
Mathias (1987)
Mathias (1987)
Benigni etal. (1979)
Adler (1980)
Adler (1980)
Torres etal (1992)
Tripathyefa/. (1993)
Murnik & Nash (1977)
Adler (1960)
Tripathyefa/. (1993)
Murnik & Nash (1977)
10,000 ug/g feed Njagi et al. (1980)
A-9
-------
Table 5 (cont)
Test system
Results8 Dose"
Without With (LED or HID)
Exogenous Exogenous
Metabolic Metabolic
Reference
CHROMOSOME ABERRATIONS-IN VITRO
Chromosomal aberrations, Chinese hamster CHO cells m vitro
Chromosomal aberrations, human lymphocytes in vitro
Chromosomal aberrations, Chinese hamster CHO cells in vitro
Chromosomal aberrations, human lymphocytes in vitro
Chromosomal aberrations, human lymphocytes in vitro
Chromosomal aberrations, human lymphocytes in vitro
.
+ NT
NT
NT
(*) NT
(+) NT
2000 ug/ml
0.1 5 ug/ml
250 ug/ml
50 ug/ml
0.1 ug/ml
1 0 ug/ml
Adler (1980)
Lioi el al (1998)
Isnidate (1988)
Kligerman et al (1 999)
Meisner et al. (1992)
Meisner etal (1993)
CHROMOSOME ABERRATIONS-IN VIVO
Micronucleus formation, Tif.MAGf female mouse bone-marrow cells m vivo
Micronucleus formation, Tif.-MAGf male mouse bone-marrow cells in vivo
Micronucleus formation, NMRI female mouse bone-marrow cells in vivo
Micronucleus formation, NMRI male mouse bone-marrow cells in vivo
Micronucleus formation, female mouse bone-marrow cells m vivo
Chromosome aberrations, mouse bone-marrow cells in vivo
OTHER INDICATORS OF DNA DAMAGE
Eschenchia coli PQ37
Saccharomyes cerevisiae, gene conversion
Saccharomyes cerevisiae, gene conversion
Saccharomyes cerevisiae, gene conversion
Saccharomyes cerevisiae, gene conversion (stationary phase cells)
Saccharomyes cerevisiae, gene conversion (logarithmic phase cells)
Saccharomyes cerevisiae, mitotic recombination
Aspergillus nidulans, gene conversion
Aspergillus mdulans, mitotic recombination
Aspergillus nidulans, mitotic recombination
DNA damage, human lymphocytes in vitro
Unscheduled DNA synthesis, human EUE cells in vitro
Unscheduled DNA synthesis, rat primary hepatocytes m vitro
Unscheduled DNA synthesis, rat primary hepatocytes in vitro
DNA strand breaks, rat stomach, liver and kidney in vivo
DNA strand breaks, rat stomach, liver and kidney in vivo
DNA strand breaks, rat lung in vivo
DNA strand breaks, rat lung in vivo
Rana catesbeiana tadpoles, DNA damage
Sister chromatid exchanges, human lymphocytes in vitro
Sister chromatid exchanges, human lymphocytes in vitro
Sister chromatid exchanges, human lymphocytes in vitro
Sister chromatid exchanges, human lymphocytes in vitro
*c
NT
NT
NT
NT
NT
NT
NT
NT
NT
2250mg/kg po x 1 Ceresa (1988a; DER)
2250mg/kg po x 1 Ceresa (1988a; DER)
1400mg/kg: po x 1 Gebel et al (1997)
1750mg/kg po x 1 Gebel et al (1997)
500mg/kg ip x 2 Kligerman el al (1999)
20 ppm dw Meisner etal (1992)
1000 ug/ml
10 ug/ml
2000 ug/ml
4000 ug/ml
64800 ug/ml
540 ug/ml
50 ug/ml
8000 ug/ml
NG
1000 ug/ml
100 ug/ml
650 ug/ml
139 ug/ml
150 ug/ml
875mg/kg po x
350mg/kg po x
875mg/kg po x
350mg/kg po x
4 8 mg/kg
NG
10 ug/ml
50 ug/ml
0 1 ug/ml
Ruiz & Marzin (1997)
Plewa and Gentile (1976)
Adler (1980)
de Bertoldi et al (1980)
MoricheVi etal. (1992)
Morichetti ef a/. (1992)
Emnovaefa/. (1987)
de Bertoldi et al (1980)
Adler (1980)
Kappas(1988)
Ribase/a/, (1995)
Adler (1980)
Hertner (1992; DER)
Puri&Muller (1984; DER)
1 Pino et al (1988)
15 Pino, el al (1988)
1 Pino el al (1988)
15 Pino et al (1988)
Clements et al (1997)
Ghiazzae/o/ (1984)
Dunkelberg et al (1994)
Kligerman et al (1999)
Lioier al (1998)
A-10
-------
Sister chromatid exchange, Chinese hamster CHO cells in vitro
DNA repair exclusive of unscheduled DMA synthesis, human lymphocytes in vitro
Table 5. cont.
NT
2000 ug/ml
25 ug/ml
Adler(1980)
Surralles et al. (1995)
Test system
Results"
Without
Exogenous
Metabolic
Activation
With
Exogenous
Metabolic
Activation
Dose"
(LED or HID)
Reference
PLANT TESTS
Hordeum vulgare, mutation
Hordeum vulgare. mutation
Zea mays, mutation
Zea mays, mutation
Nicotiana tabacum, mutation
Tradescantia paludosa. micronucleus formation
Hordeum vulgare, chromosomal aberrations
Hordeum Vulgare, chromosomal aberrations
Vicia faba, chromosomal aberrations
Vicia faba, chromosomal aberrations
Sorghum sp., chromosomal aberrations
Sorghum sp., chromosomal aberrations
Sorghum sp., chromosomal aberrations
Nigella damascena, chromosomal aberrations
Nigella damascena, chromosomal aberrations
Zea mays, chromosomal aberrations
ANEUPLOIDV
Aspergillus nidulans
Neurospora crassa,
Drosophila melanogaster
GERM CELL EFFECTS
Dominant lethal effects mouse (all germ cell stages)
Dominant lethal effects mouse spermatids
Sperm morphology, mouse
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
1000 mg/kg Wuu & Grant (1966)
200 mg/kg Stroev (1968)
200 mg/kg Morgun et al. (1982)
NG Plewaefa/ (1984)
NT NG Briza (1999)
200 mg/kg Ma et al, (1984)
500 mg/ kg spray Wuu & Grant (1967 a)
2000 mg/kg Muller et ATI (1972)
400 mg/kg Wu & Grant (1967b)
200 mg/kg Khudoley et al, (1997)
NGd Liang & Liang (1972)
NG Muller efal. (1972)
NG Lee et al. (1974)
320 mg/kg Mathias (1987)
40" mg/kg Mathias (1987)
200 mg/kg Morgun et al. (1992)
2000 ug/ml
NG
100 ug/g feed
Benign! ef a/ (1979)
Griffiths (1979)
Murnik & Nash (1977)
2400mg/kg PO x 1 Hertner (1993; DER)
1500mg/kg po x 1 Adler (1980)
600mg/kg ip x 4 Osterlbh
This Table was adopted and updated from Dearfield et al., 1993
** DER. data entry record-study was submitted by registrant and considered acceptable guideline study after review by EPA's Office of Pesticide Program
8 +. positive; (+) weakly positive; -. negative; IG, determined to be inconclusive finding by the GeneTox Panel; IE, determine to be an inconclusive finding by EPA review;
NT, not tested
b LED, lowest effective dose, HID. highest ineffective does; in vitro test, mg/kg bw/day; NG, not given; po. oral (gavage or gastric intubation); dw, drinking water; d , days;
ip, intraperitoneal
c Tested extracts of atrazine-treated Zea mays
" Commercial pesticide
"Positive with potato microsomes at doses up to 3 mM
' Aneuploidy, chromosome loss or gain is not typically associated with a DNA reactive mutagenic mechanism but usually involves disruption of spindle formation or
chromosomal segreagation
A-11
-------
Table 6. Database for the Gcnotoxicity of Simazine
Test system
Results'
With Without
Exogenous Exogenous
Metabolic Metabolic
Activation Activation
Doseb
_(LED or HID)
Reference
MUTATION
Escherichia coli PQ37, SOS chromotest
Salmonella lyphimurium TAI978/TAI538 and SL525/SL4700 difTerential toxicity
Bacillus sublilis rec strains, difTerential toxicity
Salmonella lyphimurium TAIOO, TA98, TA1535, TAI537, TA1538, reverse mutation
Salmonella lyphimurium TA100, TA98, reverse mutation
Salmonella lyphimurium TA 100, TA 1535, TA 1537, TA 1538, reverse mutation
Salmonella lyphimurium TAIOO, reverse mutation
Salmonella lyphimurium TA 100, TA 102, TA97, reverse mutation
Salmonella lyphimurium TA 1530, TA 1531, TA 1532, TA 1534, G46, reverse mutation
(spot test)
Salmonella lyphimurium, (eight unidentified strains) reverse mutation
Escherichia colt, forward mutation
Escherichia coll WP2 uvr, reverse mutation
Serralia marcescens, reverse mutation
OTHER INDICATIONS OF DNA DAMAGE
Saccharomyces cerevisiae, gene conversion
Saccharomyces cerevisiae, gene conversion
Saccharomyces cerevisiae D3, homozygosis by recombination
Saccharomyces cerevisiae D7, mitotic recombination
Saccharomyces cerevisiae D7, reverse mutation
Saccharomyces cerevisiae D7, gene conversion
Saccharomyces cerevisiae, reverse mutation
Drosophila melanogaster, somatic mutation
Drosophila melanogaster, sex-linked recessive lethal mutation
Drosophila melanogaster, sex-linked recessive lethal mutation
Drosophila melanogaster. sex-linked recessive lethal mutation
Drosophila melanogasler, sex-linked recessive lethal mutation
Drosophila melanogaster, dominant lethal test
Drosophila melanogasler, aneuploidy
Gene mutation, mouse lymphoma L5I78Y cells in vitro, Ik locus in vitro
CHROMOSOME ABERRATION
Chromosomal aberrations, Chinese hamster ovary CHO cells in vitro
Chromosomal aberrations, human lymphocytes in vitro
Micronucleus formation, mouse bone-marrow and peripheral blood cells in vivo
Micronucleus formation, both sexes.mouse bone-marrow in vivo
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
001
37.5
500 po x 2
5000 po x I
NG Mersch-Sundcrmann el al (1988)
2000ug/disc USEPA(I984)
1000 ug/disc Kuroda et al (1992)
NG Simmon el al (1977)
5000ug/plate USEPA(I984)
lOOOug/plate USEPA(1977)
NG Means era/ (1988)
lOOOug/plate Mersch-Sundermann el al (1988)
NG Sieler(l973)
NG Andersen et al (1972)
NG Fahrig(l974)
lOOOug/plate USEPA(I984)
NG Fahng(l974)
NG
I000d
50000
25000
25000
25000
5
2000 ug/g feed
10 ng/fly inj
6 ng/fly inj
6000 ug/g feed
2000 ug/g feed
6000 ug/g feed
6000 ug/g feed
300
Fahrig(l974)
Sicbcrt & Lemperle (1974)
USEPA(I977)
USEPA(I984)
USEPA(I984)
USEPA(I984)
Emnovaetal. (1987)
Tripathye/a/ (1995)
Benes& Sram(1969)
Murnlk&Nash(l977)
Murnik&Nash(l977)
Tripalhyefa/ (1995)
Murmk& Nash (1977)
Murnik& Nash (1977)
Jones etal. (1984)
Biradar & Rayburn (1995)
Kligerman et al (1999)
USEPA(I984)
Ceresa(l988a)
A-12
-------
Table 6 (cent)
Tesl system
Results"
Without
Exogenous
Metabolic
Activation
With
Exogenous
Metabolic
Activation
Doseb
JLED or HID)
Reference
OTHER INDICATORS OF DNA DAMAGE
Sister chromatid exchange, Chinese hamster ovary CHO cells in vitro
Sister chromatid exchange, Chinese hamster lung V79 cells in vitro
Sister chromatid exchange, human lymphocytes in vitro
Sister chromatid exchange, human lymphocytes in vitro
Sister chromatid exchange, human lymphocytes in vitro
Unscheduled DNA synthesis, human lung Wl 38 fibroblasts in vitro
Unscheduled DNA synthesis, rat primary heptocytes
PLANT ASSAYS
Hordeum vulgare, mutation
Hordeum vulgare, mutation
Rizobium meliloti, mutation
Zea mays, chlorophyll mutation
Zea mays, mutation
Fragaria ananassa, mutation
Tradescantia paludosa, micronuclei
Hordeum vulgare, chromosomal aberrations
Hordeum vulgare, chromosomal aberrations
Hordeum vulgare, chromosomal aberrations
Hordeum vulgare, chromosomal aberrations
Viciafaba, chromosomal aberrations
Viciafaba, chromosomal aberrations
Viciafaba, chromosomal aberrations
Allium cepa, chromosomal aberrations
Crepis capillaris. chromosomal aberrations
ANEUPLOIDY TESTS
Neurospara crassa, aneuploidy
NT 1700 USEPA(I984)
NT 2 Kurodaefo/ (1992)
NT 37.5 Kligerman et al (1999)
NT NG Ghiazzae/a/ (1984)
10 Dunkelberg et al (1994)
200 USEPA(I984)
NT Hertner(l992)
NT 1000 Wuu & Grant (1966)
NT 200 Stroev(l968a)
NT 5000 Kaszubiak(l968)
NT 200 Morgunetal. (1982)
NT NG Plewaetal (1984)
NT 2 Malone&Dix(l990)
NT 200 Maetal (1984)
NT 500 Wuu & Grant (1966)
NT 500 spray Wuu & Grant (1967a)
NT 500 Stroev(l968b)
NT 500' Kahlon(l980)
NT 200' Wuu & Grant (1967b)
NT 5 Hakeem &Shehab( 1974)
NT 1000 de Kergommeaux el al (1983)
NT 20 Chubutia & Ugulava (1973)
NT 1000 Voskanyan &• A vakyan (1984)
NT NG Griffiths (1979)
a +, positive; (+), weakly positive, -, negative; NT, not tested
b LCD, lowest effective dose; HID, highest ineffective dose; unless otherwise stated, in-vitro test, ug/mL; in-vivo lest, mg/kg bw/day; NG, not given; mj, injection, po, oral
Tested with extracts of simazine-lreated Zea mays
d Commercial pesticide tested
A-13
-------
Tesl system
Without
Exogenous
Metabolic
Activation
MUTATION
Baclenophage, forward mutation
Bacteriophage, reverse mutation
Salmonella typhimunitm TA100,TA98,TAI535,TAI537,TA1538 reverse mutation
Gene mutation, Chinese hamster lung V79 cells, hpr( locus
Results'
Doseb
Reference**
With (LED or HID)
Exogenous
Metabolic
Activation
NT
NT
(+)
lOOug/plate
2000 ug/plate
5000 ug/plate
400
Andersen el al (1 972)
Andersen el at (1972)
Kappas(l988)
Ciba-Geigy (1986; DER)
CHROMOSOME ABERRATIONS-IN VITRO
Chromosomal aberrations, Chinese hamster CHO cells in vitro
CHROMOSOME ABERRATIONS-IN VIVO
Micronucleus formation, Hamsters in vivo
OTHER INDICATORS OF DNA DAMAGE
Unscheduled DNA synthesis, rat primary hepatocytes -
Aspereillus mdulans, crossing over
NT
62.5
3000 ug/ml
50000 po x I
Puri(l984)
800
Ishidate(l983)
C iba-Geigy ( 1 984 ; DER)
Kappas (1988)
**DER, data entry record-study was submitted by registrant and considered acceptable guideline study after review by EPA's Office of Pesticide Program.
• +, positive; (+), weakly positive, -, negative, 1°, determined to be an inconclusive finding by the GeneTox Panel; 1E, determined to be an inconclusive finding by EPA review; NT, not
tested
b LED, lowest effective dose; HID, highest ineffective dose; in-vivo test, mg/kg bw/day, NG, not given; po, oral (gavage or gastric intubation); dw, drinking water, d, days, ip,
intrapcritoneal
A-14
-------
Appendix Table 8: Data from Eldridge, 1993a. DI=Diestrus; PR=Proestrus; ES=Estrus
mg/kg/day*
0
4.23
26.23
3 months
Dl= 46.1 ±7.1
PR= 29.5 ±4 6
ES= 24 8±7 7
Dl= 42 6±7.4
PR=32.5 ±7.6
ES= 25 2±4.9
Dl= 42.0±6.7
PR= 30.2±6.3
ES= 27.8±7.6
% days in stages of estrous cycle in SD rats
9 months
Dl= 44 8@- ±7.6
PR= 30.9 ±5 7
ES= 24.2@++ ±7.6
Dl= 36.2*±7.9
PR= 29.4±6.5
ES= 34.4"±9.0
Dl= 25.9"±9.1
PR= 29.4±6 2
ES=44.8"*±11.6
12 months
DI=31.1@-±5.3
PR= 26.0±7 9
ES= 42 6@+ ±10 1
Dl= 26.9±8.5
PR= 26.1±8.4
ES=47.2±13.7
Dl= 22.4*±7.5
PR= 24.6±7.3
ES=53.3±11.2
15 months
Dl=36.7±15.1
PR= 19.2±8.2
ES=44.4±122
Dl=328±14.0
PR= 24.6±8.7
ES=42.7±12.6
Dl=280±11.3
PR=22.4±5.1
ES=49.6±12.2
18 months
Dl= 31 .0 @-±5 5
PR= 24.4@-±4 1
ES= 44.9®++ ±5.7
Dl= 23 0*±9.0
PR= 20.1 ±6.4
ES=57.2*±12.5
Dl= 25.5*±17.3
PR= 18.7±7.9
ES= 55.9*±20.7
24 months
Dl=31.6 ±22 2
PR=21.0±5.7
ES=47.8±189
Dl= 37 0±32.8
PR= 13.3±8.1
ES= 50 0±27.3
Dl= 55.0±9.9
PR=21.5±10.6
ES= 24 0±0 0
uose-reiatea trena is statistically significant, in positive direction, at p < 0.05
@++ Dose-related trend is statistically significant, in positive direction, at p~< 0.01
@- Dose-related trend is statistically significant, in negative direction, at p < 0.05
@- Dose-related trend is statistically significant, in negative direction, at p<0.01
* Statistically Significant at p < 0.05
** Statistically significant at p<0.01
Appendix Table 9. Data from Morseth, 1996a Data from One Month study exposing SD females throuqh the diet
Days in Estrus vs. Time
Dose »•
# normally cycling
animals
# animals with diestrus
blocks
# animals with estrus
blocks
0 mg/kg/day
67 (74.4%)
21 (23.3%)
3 (3.3%)
2.5 mg/kg/day
66 (73.3%)
20 (22.2%)
3 (3.3%)
5 mg/kg/day
65 (72.2%)
21 (23 3%)
4 (4.4%)
40 mg/kg/day
50 (55.6%)
36 (40%)
6 (6.6%)
200 mg/kg/day
33 (36.7%)
51 (56 7%)
11 (12.2%)
A-15
-------
Appendix Table 10 : Data from Morseth, S., 1996b. Six month study exposing SD females through the diet
% Days in Estrus or Diestms
Dose (mg/
kg/day)*
0
1.8
3.65
29.4
1-2 weeks
Dl=58±9.2
ES=22±5.2
Dl=57±102
ES= 22 ± 5 6
Dl=56± 10.2
ES= 22 1 5.4
Dl=61 ±11.5
ES= 21 ± 7
5-6 weeks
Dl= 55 ± 8.7
ES= 23 ± 5.1
Dl= 55 ± 7.6
ES= 23 ±4.5
Dl= 53 ± 10 9
ES=25± 10
Dl=55± 10
ES= 24 ± 7.4
9-10 weeks
Dl= 54 ± 7.5
ES=25±9.4
Dl= 54 ± 7.5
ES=25±4.8
Dl= 51 ± 8.7
ES= 26 ± 10.2
Dl=52±10
ES=26±9.3
13-14 weeks
Dl=49± 172
ES= 31 ± 22.4
Dl= 53 ± 15.1
ES=28±18
Dl=49±16
ES=31± 21.1
Dt= 44 ±21. 6
ES=40± 27.6*
17-18 weeks
Dl=47±18.1
ES=34±24.2
Dl=49±19.4
ES= 33 ± 24.7
Dl=47±18.8
ES= 36 ± 25 1
Dl= 41 ± 25 2
ES=45± 32.1*
21-22 weeks
Dl= 51 ± 22 3
ES=32± 254
Dl= 43 ± 24.6
ES=41 ±31.9
Dl= 39 ± 23.9"
ES=45± 32.2*
Dl= 37 ±27.7*"
ES= 51 ± 34.8**
25-26 weeks
Dl=40 ±25.7
ES= 47 ± 32.2
Dl= 42 ±29.6
ES= 48 ± 35 5
Dl= 34 ± 27 3
ES= 54 ± 35.1
Dl= 29 ± 30.2*
ES= 63 t 37.0*
'ps 005; **ps 001
Appendix Table 11: Data from Thakur, 1999
Days in Estrus vs. Time
Dose*
Study Weeks
1-14
17-26
29-38
41-46
Control
26 12 ± 0.71
45 79 ± 2.05
77 22 ±2 13
81 .51 ±2.44
1 .5 mg/kg/day
28 57 ±0.86
50.28 ±2.05
74 63 ± 2 2
75.77 ± 2.73
3.1 mg/kg/day
26.38±0.87
48.67 ±2.25
71.4 ±2 29
70.8 ±2.81
4.2 mg/kg/day
26.45±0.72
48 34 ± 2 1
67.4 ± 2.36
73.41 ± 2 84
24.4 mg/kg/day
28.91 ±0.97
61.3 ± 23
80.75 ± 2 02
83.74 ± 2.24
A-16
-------
Appendix Table 12: Data from Thakur, 1999
Percent of Animals with Estrus Blocks of at least 7 days
Dose »
Study Weeks *
17-18
21-22
25-26
Control
17.5
22.78
30.38
1 .5 mg/kg/day
15
28.75
36.25
3 1 mg/kg/day
21.52
31.65
36.71
4.2 mg/kg/day
17.5
33.74
33.75
24.4 mg/kg/day
2635
50.63
50.63
A-17
-------
Appendix Table 13: Data from Thakur, 1999
Percent Days in Estrus During Weeks 1-46 and
Tumor Response for all Dose Groups Combined
Mean percent days in
estrus
Standard Error
N
No Tumor
50.869
1.27
217
Fibroadenoma
55275*
1.074
128
Carcinoma
60346*
1.596
91
# p=0.0341 compared to animals with no tumor
*p=0 0000 compared to animals with no tumor
Appendix Table 14: Data from Thakur, 1999
Percent Days in Estrus During Weeks 17-26 and
Tumor Response for all Dose Groups Combined
Mean percent days in
estrus
Standard Error
N
No Tumor
48.077
1.864
216
Fibroadenoma
52.223
2489
128
Carcinoma
60.803*
2995
91
*p=0.0003 compared to animals with no tumor
A-18
-------
Appendix Table 15: Data from Eldridge, 1993a. E=Estradiol; PROG=Progesterone; PRL= Prolactin
Serum hormone levels in SD rats (E in pg/ml; PROG and PRL in ng/ml)
mg/kg/day*
0
4.23
26.23
3 months
E= 3.5±B.4
PROG=15.6±79
PRL= N/A
E=11.2±12.6*
PROG= 16.5±10.7
PRL= NW
E=16.2±13*
PROG= 14.3±7 3
PRL= N/A
9 months
E= 22 6*20.6
PROG=11 6±11.9
PRL= 17.8 ±12.44
E=20.7±26.1
PROG= 8.2±6.6
PRL= 24.3*10.4
E=31.2±281
PROG= 7 4±4 1
PRL= 45.8±20**
12 months
E= 131 ±10.6
PROG=4.0±1.5
PRL= 13.2±2.9
E=12.5±21.6
PROG=6.9±11.4
PRL= 1 1 9±6.8
E=11.7±7.5
PROG=32±1.4
PRL=15±36
1 5 months
E= 17.3±12.8
PROG=14.2±19
PRL= 16.1±15.2
E= 18.B±18
PROG=4.1±3
PRL= 11.217 6
E=18.4±7.6
PROG= 20±24.5
PRL= 15111
18 months
E= 3 7 ±3.6
PROG= 19.6 ±29.3
PRL= 20.8 ±8.8
E=16.1±21.6
PROG=11.7*±287
PRL= 17±6.3
E=56±7.1
PROG= 4.4±4.9
PRL=17.5±8.7
24 months
E= 2.1 ±3.3
PROG=28±1.2
PRL=20.3 ±4 9
E= 3.4±5.3
PROG= 13.3±22 1
PRL= 14.2± 6.4
E= 0.9± 0 9
PROG= 3.9±0.6
PRL=135±1.1
& Dose-related trend is statistically significant, in positive direction, at p < 0.05
@ Dose-related trend is statistically significant, in negative direction, at p < 0.05
* Statistically Significant at p < 0 05
" Statistically Significant at p < 0 01
N/A - samples from the 1 and 3 month timepoints were not available for analysis because these samples were inavertatly hydrolyzed.
A-19
-------
Appendix Table 16: Histomorphology Analysis in the SD, McConnell, 1995
Values shown below are Index Weighted Scores at 1, 3, 9 and 12 months into the study.
Index Weighted Score1 at 1, 3, 9 and 12 months
Finding-
Dose*
Control
4.23 mg/kg/day
26 23 mg/kg/day
Acinar Development
(Estrogen)
1=15
3=20
9=28
12=31
1=22
3=24
9=33
12=33
1=21
3=27
9=45
12=41
Acinar/Lobular Development
(Prolactm)
1=9
3=22
9=23
12=25
1=10
3=17
9=28
12=30
1=12
3=16
9=42
12=36
Secretory Activity
(Prolactin)
1=15
3=17
9=24
12=31
1=14
3=14
9=28
12=36
1=12
3=17
9=46
12=39
Dilated Ducts with Secretion
(Protactin)
1=9
3=12
9=17
12=23
1=12
3=11
9=24
12=25
1=9
3=15
9=45
12=41
1 The index weighted score is calculated as such: the severity of the findings, as determined by the examining pathologist, is converted to a numerical value and the
numerical values for each group are summed. The score of absent= 1; minimal =2; mild=3; moderate=4 and marked=5.
A-20
-------
Appendix Table 17: Galactocele Incidence and Severity in the SD Female, Thakur, 1991 a
Finding*
Dose*
Control
4.23 mg/kg/day
26 23 mg/kg/day
One and Three month
None at either limepoint
None at either timepoint
None at either timepoint
Nine Months
1 (slight)1
4 (2-minimal;1 -slight;
1 -moderate)
8 (2-minimal; 1- moderate;
5 moderately severe)
12 Months
5 (4- minimal; 1
moderate)
5 ( 1 -minimal; 3-slight;
1- moderately severe)
10 (5-slight; 4-moderate;
1- moderately severe)
15 Months
7 (2- slight, 1- moderate,
1- moderatly severe; 3-
severe
7 (3 - minimal, 4- slight)
9 (2- minimal; 3- slight; 2-
moderate, 1- moderatly
severe; 1- severe)
1 The scores are: minimal, slight; moderate; moderately severe and marked
Appendix Table 18: Data from Morseth, 1996a and 1996b*
Timepointe for measurement of LH surge In both the one and six month studies
Biologic time
1100
1400
1600
1800
2000
2300
# animals for non-repeat bleed
10
15
15
15
15
10
# animals for repeat bleed
10
10
10
10
10
10
Expected state of serum LH
levels
baseline
baseline
LH surge
LH surge
LH surge
baseline
In a normally cycling rat this is
equivalent to:
proestrus morning
early afternoon proestrus
mid- afternoon proestrus
Late afternoon proestrus
Proestrus evening
Proestrus evening
' There were 90 females in each group: 10 + 15+15 + 15 + 15 + 10 non-repeat bleed animals = 80 animals
plus the 10 repeat bleed animals equals 90 animals per group.
A-21
-------
Appendix Table 19: Data from Morseth, 1996a . Doses are in mg/kg/day. LH values given are in picograms/ml.
LH data from the one-month study
Biologic
Time>
mean and SD
nonrepeat
DiGGu
mean and SD
repeat bleed
1100
0=998 ±614
2.5= 943 ±614
5.0= 1140 ±71 5
40= 121 9 ±467
200= 873 ±656
0= 732 ± 461
25=1101 ±652
5.0= 810 ±51 9
40= 755 ±389
200= 51 4 ±503
1400
0=1 122 ±564
2.5=1171 ±802
5.0= 882 ±926
40= 11 25 ±795
200= 1099 ±863
0= 786 ± 557
2.5= 2222 ±1220
5.0= 1678 ±1602
40= 1037 ±829
200= 453 ±31 3
1600
0=331 5 ±2684
2.5= 20951 ±131 5
5.0= 3099 ±2521
40= 351 8 ±4514
200= 1685 ±2962
0= 1301 ± 1031
2.5= 3029 ±2383
5 0= 4971 ±5047
40= 11 37 ±629
200= 552 ±311
1800
0=51 38 ±4403
2.5= 4489 ±4345
5.0= 2804 ±13
40= 3246 ±1981
200= 2752 ±3137
0= 2650 ± 2389
2.5= 301 5 ±3220
5 0=271 7 ±25
40= 1450 ±857
200= 812 ±470
2000
0=2242 ±1850
2.5= 1118 ±412
5 0=1554 ±14
40= 1740 ±11 57
200= 1853 ±11 38
0= 2606 ± 2076
2.5= 1731 ±1447
5 0=2954 ±351 5
40= 1477 ±1296
200= 11 40 ±328
2300
0= 761 ± 288
2.5= 486 ±138
5.0= 508 ±31 7
40= 689 ±373
200= 11 26 ±81 645
0=1671 ±674
2 5= 1475 ± 456
50= 1431±345
40= 1362 ±329
200= 1080 ±30142
A-22
-------
Appendix Table 20: Data from Morseth, 1996b. Doses are in mg/kg/day. LH values given are in picograms/ml.
LH data from the six-month study
Biologic
Time*
mean and SD
nonrepoat
bleed
mean and SD
repeat bleed
1100
0= 1900 ±775
1 8= 1816 ±543
3.65= 1581 ±791
29.4= 1863 ±786
0=909 ±410
18= 1075 ±621
3.65= 972 ±353
29 4= 1005 ±482
1400
0=2326 ±1082
1.8= 1606 ±926
3.65= 1799 ±933
29.4= 1420 ±622
0=1 136 ±554
1.8= 1468 ±977
3.65= 984 ±466
29.4= 1155 ±620
1600
0= 2669 ±1464
1.8= 2507 ±1008
3.65= 2463 ±1201
29 4= 191 3 ±799
0= 2213 ±2562
1.8= 1603 ±682
3 65= 2277 ±1470
29.4= 850 ±352
1800
0= 3458 ±2310
1.8= 3235 ±2751
3.65= 31 75 ±1685
29.4= 1356 ±760
0=3336 ±31 38
1 8= 3631 ±2732
3.65= 2500 ±1897
29.4= 858 ±416
2000
0= 2327 ±1668
1.8= 2249 ±1498
3.65= 1899 ±752
29.4= 1308 ±477
0= 3388 ±3344
1.8= 2510 ±1138
3.65= 2409 ±1525
29.4= 1042 ±627
2300
0=1 178 ±337
1.8= 1258 ±428
3.65= 1063 ±383
400= 11 29 ±350
0= 1672 ±426
1.8= 1229 ±492
3.65= 1271 ±559
400= 953 ±549
Appendix Table 21: Group Mean Absolute Pituitary weights by Timepoints, Thakur, 1991a
Weights are in mg.
Absolute Pituitary Weights
Dose
(mg/kg/day) *
Control
423
26.23
3 months
s= 23.0
SD= 4.2
x=21.2(-8%)1
SD= 3.0
x=205(-11%)
SD= 8.0
9 months
s=240
SD=64
x= 29.9 (+25%)
SD=6.1
x= 37.0 (+54%)
SD= 7.9
12 months
x=370
SD=19.9
x= 35.4 (-4%)
SD= 26.4
x=41.8(-H3%)
SD= 14.5
1 Values in parenthesis represent percent change relative to control
A-23
-------
Appendix Table 22: Group Mean Pituitary Weights Relative to Body Weight, Thakur, 1991 a
Values represent pituitary weight as a percentage of body weight
Relative Pituitary Weights
Dose
(mg/kg/day) •
Control
4.23
2623
3 months
x= 0.00697
SD= 0.0012
x= 0.00668 (-4%)1
SD= 0.00123
x= 0.00677 (-3%)
SD= 0.00249
9 months
x= 0.00607
SD= 0.00163
x= 0.00765 (+26%)
SD= 0.00187
x= 0.00967 (+59%}
SD= 0.00245
12 months
x= 0.00985
SD= 0.0062
x= 0.00830 (-16%)
SD= 000559
x=0.01239 (+26%)
SD= 0 00572
1 Values in parenthesis represent percent change relative to control
Appendix Table 23: Group Mean Pituitary Weights: Absolute and Relative to Body Weight, Morseth, 1996b
Values represent pituitary weight as a percentage of body weight
Absolute and Relative Pituitary Weights
Dose
(mg/kg/day) »
Absolute
Relative
Control
R=230
SD= 5.0
s= 0.0075
SD=0.0017
1.8
x= 24.0 (+4%)
SD= 5.0
s= 0.0078 (+4%)
SD= 0.0015
3.65
x= 24.0 (+4%)
SD= 5.0
><= 0.0081 (+8%)
SD=0.0015
29.4
x= 28.0 (+22%)
SD= 8.0
x= 0.0096 (+28%)
SD=0003
1 Values in parenthesis represent percent change relative to control
A-24
-------
Appendix Table 24: Association of Pituitary Adenomas/Hyperplasia and Mammary
Tumors
Study »•
Animals Without
Mammary Tumors
Animals With Mammary
Tumors
Thakur, 1992a
86% of the animals had
either pituitary adenoma or
hyperplasia
86% of the animals had
either pituitary adenoma or
hyperplasia
Morseth, 1998
62% of the animals had either
pituitary adenoma or
hyperplasia
93% of the animals had either
pituitary adenoma or
hyperplasia
Appendix Table 25a: Association of Absolute Pituitary Weight with Mammary Tumors
Absolute Pituitary Weight
Study*
Animals Without
Mammary Tumors
Animals With Mammary
Tumors
Thakur, 1992a
x = 0.0742
SD= 0.0823
x = 0.144(94%)1
SD=0109
Morseth, 1998
x = 0.117
SD= 0.076
x = 0.171 (46%)
SD=0.119
1 Value in parenthesis represents percent increase over pituitary weight of animals without tumors
Appendix Table 25b: Association of Relative Pituitary Weight with Mammary Tumors
Relative Pituitary Weight
Study*
Animals Without
Mammary Tumors
Animals With Mammary
Tumors
Thakur, 1992a
K= 0.021
SD= 0.024
• 0.0379 (80%)'
SD= 0.0307
Morseth. 1998
x= 0.0321
SD= 0.029
x= 0.053 (65%)
SD= 0.068
1 Value in parenthesis represents percent increase over pituitary weight of animals without tumors
A-25
-------
Appendix Table 26: Pituitary Tumor and Focal Hyperplasia Incidenes and Piutiatry
Weights In OVX vs Intact Animals, Morseth, 1998
Pituitary Tumor and Enlarged Pituitary Incidence
and Pituitary Weight1
(3-Adenoma (12 months)
(3-Adenoma (24 months)
Absolute Pituitary Wt. in
grams (12 months)
Absolute Pituitary Wt in
grams (24 months)
Relative Pituitary Wt. as a %
of body wt.
(12 months)
Relative Pituitary Wt as a %
of body wt
(24 months)
Enlarged Pituitary
OVX
K= 6%
K= 502%
K= 0.020
K= 0.051
K- 0.0037
K= 0.01 16
x= 34%
Intact
K= 17%
^=69.8%
K= 0.0314
^=0184
• 0.00832
x= 0.0331
x= 86%
1 Values shown represent means for all the OVX or intact animals combined, regardless of dose group. Mean values
between dose groups were very similar.
A-26
-------
Appendix Table 27: Control individual animal data from the nine month timepoint of
the two-year serial sacrifice study Jhakur 1991 a
Control animals at 9 months
Animal*
B94752
B94753
B94754
B94755
B94756
B94757
B94758
B94759
B94760
B94761
Mammary
tumor
none
none
none
none
none
none
none
none
none
none
Pituitary
Alteration
Foe. Hy. -slight
none
none
none
none
none
none
none
none
none
Pituitary wt
(abs. in mg)
26
30
24
27
25
16
19
33
26
14
Galactocele
none
none
none
none
none
none
none
slight
none
none
Acinar/lobular
Development
1
2
-
1
1
2
2
2
2
-
Secretory
Activity
1
2
1
2
1
-
2
3
2
-
Dilated Duct
with
Secretion
-
1
-
1
-
-
1
4
-
-
A-27
-------
Appendix Table 28: Low dose (70 ppm, 4.23 mg/kg/day) individual animal data from
the nine month timepoint of the two-year serial sacrifice study, Thakur 1991 a.
70 ppm animals at 9 months
Animal*
B94822
B94823
B94824
B94825
B94826
B94827
B94828
B94829
B94830
B94831
Mammary
tumor
none
none
none
none
none
none
none
none
none
none
Pituitary
Alteration
none
Foe. Hy -slight
none
none
none
none
none
none
Foe. Hy. -slight
none
Pituitary wt
(abs. in mg)
28
28
41
29
21
27
37
35
29
24
Galactocele
none
minimal
moderate
none
none
none
minimal
none
slight
none
Acinar/lobular
Development
2
1
3
1
2
2
2
2
3
-
Secretory
Activity
2
1
4
2
-
2
2
2
3
-
Dilated Duct
with
Secretion
1
-
4
-
-
2
3
2
2
-
A-28
-------
Appendix Table 29: High dose (400 ppm, 26.23 mg/kg/day) individual animal data
from the nine month timepoint of the two-year serial sacrifice study, Thakur 1991a.
400 ppm animals at 9 months
Animal*
B94892
B94893
B94894
B94895
B94896
B94897
B94898
B94899
B94900
B94901
Mammary
tumor
carcinoma
none
fibroadenoma
none
none
carcinoma
carcinoma
none
none
fibroadenoma
Pituitary
Alteration
none
none
none
none
none
Focal hy.-
mimimal
adenoma
Focal hy.-
slight
none
adenoma
Pituitary wL
(abs. in mg)
36
36
38
23
37
36
55
41
33
35
Galactocele
moderate
moderately
severe
moderately
severe
none
severe
moderately
severe
minimal
moderately
severe
moderately
severe
none
Acinanflobular
Development
2
4
3
4
4
4
2
4
4
1
Secretory
Activity
3
4
4
4
4
4
3
4
4
2
Dilated Duct
with
Secretion
4
4
4
4
4
4
2
4
4
1
A-29
-------
Appendix Figure 1: Mean plasma LH levels from the repeat bleed group of the 6-
month study (Morseth, 1996b)
Re peat bleed LH data
4000
3500
1100
1400 1600 1800
Biological Time (Hours)
2000
2300
A-30
-------
Name
Source Tolerances SF Decision SF List
1 1.12-DODECANEDIOL DIMETHACRYLATE POLYMER
2 1,6-HEXANEDIOLDIMETHYACRYLATE POLYMER
3 12-HYDROXYSTEARIC ACID- POLYETHYLENE GLYCOL
4 2-BUTENEDIOIC ACID (Z)-, POLYMER WITH ETHENOL
5A-(P-NONYLPHENYL)W-HYDROSYPOLY(OXYPROPYLENE
6 A-{P-1,1,3,3-TETRAMETHYLBUTYL)PHENYL} POLY(OXYPROPYLENE
7 A-BUTYL-W- HYDROXYPOLY{OXYPROPYLENE BLOCK POLYMER
8 ACETIC ACID ETHENYL ESTER, POLYMER WITH ETHENOL
9 ACRYLAMIDE POTASSIUM ACRYLATE-ACRYLIC ACID COPOLYMER
10 ACRYLIC ACID-SODIUM ACRYLATE-SODIUM-2-METHYLPROPANE
11 ACRYLIC ACID-STEARYL METHACRYLATE COPOLYMER
12 ACRYLONITRILE-STYRENE-HYDROXYJPROPYL METHACRYLATE
13A-HYDRO-W-HYDROXPOLY(OXYPROPYLENE)
14 A-HYDRO-W-HYDROXYPOLY(OXYETHYLENE)
15 ALKYL (C12-C20) METHACRYLATE- METHACRYLIC ACID COPOLYM
16 CELLULOSE ACETATE
17 CINNAMALDEHYDE
18 ETHYLENE
19 ETHYLENE GLYCOL DIMETHACRYLATE POLYMER
20 ETHYLENE GLYCOL DIMETHYACRYLATE-LAURYL METHACRYLATE
21 GIBBERELLIC ACID
22 HYDROXYETHYL CELLULOSE
23 HYDROXYPROPYL CELLULOSE
24 HYDROXYPROPYL METHYLCELLULOSE
25 LAURYL METHACRYLATE- 1.6- HEXANEDIOL DIMETHACRYLATE
26 MALEIC ACID MONOBUTYL ESTER-VINYL METHYL ETHER COPOL
27 MALEIC ACID MONOISOPROPYL ESTER-VINYL METHYL ETHER
28 MALEIC ANHYDRIDE- METHYL VINYL ETHER. COPOLYMER
29 METHYL METHACRYLATE- 2-SULFOETHYL METHACRYLATE
30 METHYL METHACRYLATE- METHACRYLIC ACID- MONOMETHOXY
31 METHYL VINYL ETHER-MALEIC ACID COPOLYMER
32 METHYLCELLULOSE
33 OCTADECANOIC ACID, 12-HYDROXY-. HOMOPOLYMER OCTADEC
34 PELARGONIC ACID
35 PHOSPHINE RESULTING FROM USE O
36 POLY (VINYL PYRROLIDONE)
37 POLY(OXYETHYLENE/OXYPROPYLENE) MONOALKYL(C6-C10)
38 POLY(OXYPROPYLENE) BLOCK POLYMER WITH POLY
-------
55 SODIUM CARBOXYMETHYLCELLULOSE
56 STEARYL METHACRYLATE-1,5- HEXANEDIOL DIMETHACRYLATE
57 STYRENE-2-ETHYLHEXYL ACRYLATE- GLYCIDYL METHACRYLATE
58 TRICHODERMA VIRIDE SENSU BISBY
59 VINYL ACETATE-ALLYL ACETATE- MONOMETHYL MALEATE
60 VINYL ACETATE-ETHYLENE COPOLYMER
61 VINYL ACETATE-VINYL ALCOHOL ALKYL LACTONE COPOLYMER
62 VINYL ALCOHOL-DISOD1UM ITACONATE COPOLYMER
63 VINYL ALCOHOL-VINYL ACETATE- MONOMETHYL MALEATE
64 VINYL PYRROLIDONE DIMETHYLAMINOETHYLMETHACRYLATE
65 BEAUVARIABASSIANA
66 BENSULFURONMETHYL
67 CARBON DISULFIDE
68 CLOPYRALID (Dichloropyndmeca
69 COPPER CARBONATE
70 COPPER HYDROXIDE
71 COPPER LINOLEATE
72 COPPER OLEATE
73 COPPER OXYCHLORIDE
74 COPPER SULFATE
75 CRYOLITE
76 CYFLUTHRIN
77 DICHLOROPYRIDINECARBOXYLIC ACI
78 FENOXAPROPETHYL
79 FENVALERATE
80 GLYPHOSATE AND ITS METABOLITES
81 Hydroprene
82 LAMBDACYHALOTHRIN
83 TRALOMETHRIN
84 BACILLUS THURINGIENSIS
85 BACILLUS THURINGIENSIS CRYLAB
86 BACILLUS THURINGIENSIS VARIETY
87 BROMACIL
88 BROMOXYNIL
89 BUTRALIN
90 Carbamyl-2,4,5-tnchlorobenzoi
91 COLLETOTRICHUM GLOEOSPORIOIDES
92 DtCHLOBENIL
93 DIPHENYLAMINE
94 HYDRAMETHYLNON
95 METRIBUZIN
96 PARAQUAT DICHLORIDE
97 TERBACIL
98 THIOBENCARB
99 TRICLOPYR
100 ZINC PHOSPHIDE
other
other
other
other
other
other
other
other
other
other
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RD
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
RED
1
1
1
13
1
1
1
1
1
2
16
2
4
39
2
1
1
1
1
36
4
5
4
67
129
1
8
2
2
1
1
2
90
6
38
2
50
3
4
57
116
32
25
34
6
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
none
------- |