THURSDAY

October 8, 1992





Part IV

Environmental Protection Agency





Amitrole; Preliminary Determination To Terminate Special Review


(This reprint was prepared from the electronic file that accompanied the original signed documents transmitted to the Office of the Federal Register.  This file was certified to be a true copy of the original.)

(This document appeared at 57 FR 46448-46456.)




Federal Register / Vol. 57, No. 196 / Thursday, October 8, 1992 / Notices
ENVIRONMENTAL PROTECTION AGENCY
[OPP - 30000/38A; FRL 4164 - 1]


Amitrole; Preliminary Determination To Terminate Special Review

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed decision to terminate Special Review.




SUMMARY: This Notice sets forth EPA's preliminary determination regarding the continued registration of pesticide products containing amitrole and sets forth the Agency's assessment of the risks and benefits associated with the pesticidal uses of amitrole. On May 15, 1984, the Agency issued a Notice of Special Review of pesticide products containing amitrole based on carcinogenic concerns (49 FR 20546). This Notice proposes to terminate the amitrole Special Review based on the Agency's determination that the benefits of use outweigh the risks.
DATES: Written comments on this Notice must be received on or before November 9, 1992.

ADDRESS: Submit three copies of written comments, bearing the document control number ``OPP - 30000/38A; FRL 4164 - 1'' by mail to: Public Docket and Freedom of Information Section, Field Operations Division (H7506C), Office of Pesticide Programs, Environmental Protection Agency, 401 M St., SW., Washington, DC 20460. In person, bring comments to: Rm. 1132, Crystal Mall #2, 1921 Jefferson Davis Highway, Arlington, VA.
    Information submitted in any comment concerning this Notice may be claimed confidential by marking any part or all of that information as ``Confidential Business Information'' (CBI). Information so marked will not be disclosed except in accordance with procedures set forth in 40 CFR part 2. A copy of the comment that does not contain CBI must be submitted for inclusion in the public record. Information not marked CBI may be publicly disclosed by EPA without prior notice to the submitter. All non-CBI written comments and the correspondence index will be available for public inspection and copying in Rm. 1132 at the Virginia address given above, from 8 a.m. to 4 p.m., Monday through Friday, excluding legal holidays.

FOR FURTHER INFORMATION CONTACT: Philip J. Poli, Review Manager, Special Review Branch, Special Review and Reregistration Division (H7508W), Office of Pesticide Programs, Environmental Protection Agency, 401 M St., SW., Washington, DC 20460. Office location and telephone number: Third floor, Westfield Bldg., 2800 Jefferson Davis Highway, Arlington, VA. (703) 308 - 8038.
SUPPLEMENTARY INFORMATION:
Electronic Availability: This document is available as an electronic file on the Federal Bulletin Board at 9 a.m. on the date of publication in the Federal Register. By modem dial 202-512-1387 or call 202-512-1530 for disks or paper copies. This file is available in Postscript, Wordperfect 5.1 and ASCII.
    This document presents the basis for the Agency's proposed decision to terminate the Special Review of amitrole.

I. Introduction

A. Summary

    Amitrole is the common name for 3-amino-1,2,4-triazole. It is most commonly sold under the trade names AMIZOL~ and Amitrol T~, and is formulated both as a wettable powder and a liquid concentrate.
    Amitrole was first registered as an herbicide under the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) in 1956. It is a systemic broad spectrum, post-emergence herbicide that is used to control many annual and perennial broadleaf weeds, grasses and woody species. Amitrole is classified by EPA as a Restricted Use pesticide [Ref. 1]. Section 3(d) of FIFRA specifies that only certified applicators trained for and familiar with pesticide use, or persons under their direct supervision, can use amitrole containing products. The current registered uses of amitrole are limited to non-crop, commercial sites. All amitrole food uses were canceled by the EPA in 1971 because of the carcinogenic potential from dietary exposure.
    Currently, amitrole is registered for use on land surrounding commercial, industrial and farm premises, on rights-of-way, public utilities, and nurseries. Under the current use pattern, the principal pathway for human exposure is by the dermal route resulting from mixing, loading and applying the pesticide.
    EPA's Special Review was initiated to address the use of amitrole on non-crop sites and by homeowners, and examined the carcinogenic risk to mixers, loaders and applicators. Since the time the Special Review was initiated, the registrant has taken voluntarily actions which have reduced worker exposure to amitrole. These actions include deletion of high exposure application methods such as knapsack sprayers, adoption of a ``no-glug'' container design for the liquid formulation to reduce splashing while pouring, addition of protective clothing requirements to labels, and packaging of the wettable powder formulation in water soluble packets. Additionally, the registrant has canceled all homeowner products.
    EPA has completed its risk/benefit analysis of amitrole and has determined that the benefits from continued use of amitrole outweigh the risks. Accordingly, the Agency is proposing to terminate the Special Review.

B. Legal Background

    In order to obtain a registration for a pesticide under FIFRA, an applicant must demonstrate that the pesticide satisfies the statutory standard for registration. The standard requires, among other things, that the pesticide will not cause ``unreasonable adverse effects on the environment'' [FIFRA section 3(c)(5)]. The term ``unreasonable adverse effects on the environment'' means ``any unreasonable risk to man or the environment, taking into account the economic, social, and environmental costs and benefits of the use of any pesticide'' [FIFRA section 2(bb)]. This standard requires a finding that the benefits of each use of the pesticide outweigh the risks of such use, when the pesticide is used in compliance with the terms and conditions of registration and in accordance with commonly recognized practices.
    The burden of proving that a pesticide satisfies the statutory standard is on the proponents of registration and continues as long as the registration remains in effect. Under FIFRA section 6, the Administrator may cancel the registration of a pesticide or require modification of the terms and conditions of a registration if he determines that the pesticide product causes unreasonable adverse effects to man or the environment. EPA created the Special Review process to facilitate the identification of pesticide uses which may not satisfy the statutory standard for registration and to provide a public procedure to gather and evaluate information about the risks and benefits of these uses.
    A Special Review may be initiated if a pesticide meets or exceeds the risk criteria set out in the regulations at 40 CFR part 154. EPA announces that a Special Review is initiated by publishing a Position Document (PD) in the Federal Register. After a PD is issued, registrants and other interested persons are invited to review the data upon which the review is based and to submit data and information to rebut EPA's conclusions by showing that EPA's initial determination was in error, or by showing that use of the pesticide is not likely to result in unreasonable adverse effects on human health or the environment. In addition to submitting rebuttal evidence, commenters may submit relevant information to aid in the determination of whether the economic, social and environmental benefits of the use of the pesticide outweigh the risks. After reviewing the comments received and other relevant materials obtained during the Special Review process, EPA makes a decision on the future status of registrations of the pesticide.
    The Special Review process may be concluded in various ways depending upon the outcome of EPA's risk/benefit assessment. If EPA concludes that all of its risk concerns have been adequately rebutted, the pesticide registration will be maintained unchanged. If, however, all risk concerns are not rebutted, EPA will proceed to a full risk/benefit assessment. In determining whether the use of a pesticide poses risks which are greater than the benefits, EPA considers possible changes to the terms and conditions of registration which can reduce risks to the level where the benefits outweigh the risks, and it may require that such changes be made in the terms and conditions of the registration. Alternatively, EPA may determine that no changes in the terms and conditions of a registration will adequately assure that use of the pesticide will not cause any unreasonable adverse effects. If EPA makes such a determination, it may seek cancellation, suspension, or change in classification of the pesticide's registration. This determination would be set forth in a Notice of Final Determination issued in accordance with 40 CFR 154.33.
    Issuance of this Notice means that the Agency has assessed the potential adverse effects associated with the uses of amitrole and has preliminarily determined that the benefits override the risks.

C. Regulatory Background

    The Registration Standard for amitrole was published on March 30, 1984. It required submission of product chemistry, environmental fate, toxicology and ecological effects data.
    On May 15, 1984, the EPA issued a notice to initiate a Special Review based on carcinogenic concerns to mixers, loaders and applicators for registrations of products containing amitrole (49 FR 20546). This document, also referred to as Position Document 1 or PD 1, detailed the basis for the Agency's decision to initiate a Special Review. At that time, the Agency determined that all uses, including the homeowner use, would be the subject of the Special Review for amitrole. Subsequently, all products designated for homeowner use were canceled [Ref. 2], and the ensuing Special Review focused on the carcinogenic risk to mixers, loaders and applicators. The Agency had reviewed data concerning the potential adverse effects associated with uses of amitrole which indicated that amitrole induces thyroid and pituitary tumors in the rat, plus liver and thyroid tumors in mice, and had determined that pesticide products containing amitrole met or exceeded the risk criterion in 40 CFR 162.11(a)(3)(ii)(A), (1984 volume). That section required that a Special Review shall be initiated if a pesticide ``induces oncogenic effects in experimental mammalian species or in man as a result of oral, inhalation or dermal exposure.'' That criterion is mirrored by the current provision of 40 CFR 154.7(a)(2), (1991 volume), which sets forth the similar criterion for initiation of a Special Review by EPA.

II. Summary of Toxicological Concerns and Agency Evaluation ofComments

    The Special Review of amitrole was initiated in 1984 because of data indicating that amitrole induces thyroid, pituitary and liver tumors in laboratory animals. In addition, the Agency required further information to be submitted regarding amitrole's other potential effects [Ref. 1]. This section summarizes the Agency's review of studies for all effects of concern and includes discussion of new information and public comments that have been received since publication of the PD 1.

A. Carcinogenicity

    In the PD 1, the Agency indicated its concern about the carcinogenic effects of amitrole. In making its current determination, EPA reviewed nine long-term carcinogenicity studies conducted with amitrole on rats, mice or hamsters. One of the rat studies is a chronic inhalation study; the rest of the studies (4 rat, 3 mouse and 1 hamster) are oral feeding, gavage or drinking water studies [Refs. 3 through 12]. None of these studies individually satisfied EPA testing guidelines; some had major deficiencies (e.g., target doses grossly exceeded [Ref. 3]; problems with the histological examination and presentation of data [Ref. 4]). The rat inhalation study and one of the rat oral studies were classified as invalid. In addition, one perinatal carcinogenicity study in mice could not be evaluated. The chronic feeding study in hamsters was not addressed by the Agency because the hamster was determined to be the least sensitive species. The rest of the studies are classified as supplementary; that is, the studies are scientifically valid, but do not satisfy all Agency guideline requirements. These five studies are discussed below, and do indicate that amitrole is likely to be a carcinogen, inducing thyroid and pituitary tumors in rats, and thyroid and liver tumors in mice. They are considered part of the weight-of-evidence determination for the carcinogenic potential of amitrole. Following these studies are discussions of other pertinent information which support the carcinogenic potential of amitrole.
    1. Rat studies-- a. Keller; Hazleton, 1959. Amitrole was administered in the diet at 0, 10, 50 or 100 ppm to Charworth Farm rats and was associated with numerical increases, in both sexes at the terminal sacrifice, in the incidence of thyroid adenoma at 50 ppm and 100 ppm and in combined thyroid adenoma/carcinoma at 100 ppm. There were statistically significant positive trends for thyroid adenomas and for combined adenoma/
carcinoma in both sexes. There were also statistically significant positive trends in thyroid hyperplasia in both sexes at 68 weeks, which were not seen at 104 weeks.
    b. Johnson; Food and Drug Research, 1981. Amitrole was administered in the diet in five ``pulsed'' dose groups (A, 0 - 0; B, 5 - 100; C, 1 - 20; D, 3 - 60; E, 10 - 200 ppm) to Fischer 344 rats. In treatment groups B, D and E, statistically significant increases in the incidence of thyroid follicular cell adenoma and combined adenoma/carcinoma were reported in both sexes. There were also statistically significant positive trends for adenoma and carcinoma, both individually and combined, in both sexes. In both sexes, there were both statistically significant increases in the incidence of follicular cell hyperplasia in all treatment groups, and also statistically significant positive trends. Increases in thyroid organ weights were observed for groups B and E, both sexes. Thyroid hormone T3 was elevated throughout the study for all treatment groups compared to controls, while T4 values were variable.
    c. Steinhoff, 1979, 1983. Amitrole was administered in the diet at 0, 1, 10 or 100 ppm to Wistar rats and was associated with statistically significant increases in the incidence of thyroid tumors at 100 ppm in males and females when compared to the controls, as well as a statistically significant positive trend in both sexes. There was a numerical increase in the incidence of pituitary tumors in both sexes in all treatment groups, which was statistically significant in both sexes at 100 ppm, with a statistically significant positive trend for females. Percentage accumulations of radioiodine were increased in both sexes at 100 ppm at ``the majority of the test times'', which was thought to be due to increases in physiologically active thyroid tissue; proportional plasma iodine remained fairly constant throughout the study.
    2. Mouse studies-- a. Innes, 1969. Amitrole was administered by stomach tube at 6,667 ppm and by dietary feeding at 2,192 ppm. In this study, amitrole was used as a positive control for screening 120 compounds to C57 mice. Results indicated carcinoma of the thyroid in 64 of 72 treated animals and hepatomas in 67 of 72 animals.
    b. Steinhoff, 1979, 1983. Amitrole was administered in the diet to NMRI mice at 0, 1, 10 or 100 ppm and was associated with statistically significant positive trends for hepatocellular carcinoma and combined liver carcinoma/adenoma in females. Thyroid weights were increased throughout the study in both sexes at 100 ppm.
    3. Mutagenicity. Although the published literature shows that amitrole is negative in a majority of mutagenicity assays [Ref. 13], there is some evidence that amitrole may have some genotoxic activity, as well as transformation activity in mammalian tests. Two sister chromatid exchange assays were reported positive; there were mixed results for DNA damage and unscheduled DNA synthesis, and all in vitro cell transformation assays were positive.
    In addition, several studies suggest that amitrole has mutagenic activity when assayed with metabolic activation systems other than the usual liver preparations (e.g. ``S9'' mixes) [Ref. 14]. Also, Daston et. al. [Ref. 15] found that mutations were induced by amitrole in the mouse lymphoma assay supplemented with a prostaglandin H synthase activation system. Further, Krause and Eling [Ref. 16] have shown PHS- and lactoperoxidase-mediated binding of labelled amitrole to protein and nucleic acid (tRNA), as well as protein binding catalyzed by microsomal thyroid peroxidases; this demonstrates that amitrole can be activated to a reactive species.
    Therefore, the data suggest that evidence of amitrole's possible genotoxicity may play a role in its carcinogenic potential. However, this is not an established conclusion, and the exact role genotoxicity has for amitrole-induced thyroid tumors is unclear at this time.
    4. Mechanism of tumor formation (Threshold Concept). Amitrole is a potent antithyroid agent in laboratory animals, causing thyroid tumors in rats and mice. A plausible theory for amitrole-induced thyroid tumor development is supported by the threshold concept, discussed below. While the Agency recognizes that this concept provides a possible mechanism for thyroid tumor evolution, it has not been proven.
    Experimental evidence indicates that thyroid tumors in rats can only occur as a result of exceeding a threshold (i.e., tumors appear only when hormone concentrations are altered above or below a specific level, identified as the threshold, for an extended period of time). To fully understand the threshold concept, it is necessary to provide a brief explanation of the physiological relationship between the pituitary gland and thyroid gland. Decreased levels of thyroid hormones disturb the physiological equilibrium which causes the anterior pituitary gland to secrete thyroid stimulating hormone (TSH). In turn, TSH causes thyroid hormone levels to rise back to normal levels. The anterior pituitary then decreases production of TSH which causes a subsequent decrease in the production of thyroid hormone. If there is a disruption of this feedback mechanism resulting in decreased thyroid hormone levels that cannot be counterbalanced by increased TSH production, the result is the continuous but futile stimulation of the thyroid by TSH. This constant stimulation by TSH over an extended period of time changes the morphology of the thyroid gland and, at some critical point, results in tumor formation. This is the proposed mechanism for tumor formation involving a threshold. It is unclear what role the genotoxicity of amitrole may play in this process.
    Amitrole may interfere with the synthesis of thyroid hormones by inhibiting iodide peroxidase, which results in positive feedback to the pituitary (described above) for as long as amitrole exposure is sufficient to maintain decreased thyroid levels. It has been argued, therefore, that unless exposure levels of amitrole are high enough to sufficiently inhibit iodide peroxidase for an extended period of time, causing the appropriate hormone levels to exceed a specific concentration, thyroid tumors should not be induced. In other words, doses of amitrole below the threshold would be inadequate to cause this exaggerated positive feedback.
    5. Structure-Activity. Amitrole is a heterocyclic aromatic amine, with a structure somewhat similar to a few triazoles that have been shown to induce liver tumors in mice. Other heterocyclic aromatic amines, for instance, some bicyclo- and tricyclo- heterocyclic aromatic amines, have been shown to be very potent carcinogens and mutagens. Also, some homocyclic aromatic amines (those with double rings attached by a simple ether-like bridge), show a correlation between anti-thyroid activity (inhibition of thyroid peroxidase) and thyroid carcinogenesis. Because the structure of amitrole is remotely related to those of other tumor-inducing aromatic amines, it has been determined that there is a weak structure activity relationship.
    6. Classification of carcinogenic potential. Based on the weight-of-evidence provided collectively by the five studies discussed in this document, amitrole is considered to be a probable human carcinogen. Evidence is considered to be sufficient if there is an increased incidence of tumors: (a) In multiple species or strains of test animals; or (b) in multiple experiments, for example, with different dose levels or routes of administration; or (c) to an unusual degree in a single experiment with regard to high incidence, unusual site or type of tumor, or early age of onset. Although it was generally agreed within the Agency that none of the individual studies were good. However, it was evident that malignant and benign tumors were observed in both sexes of multiple strains of the rat (thyroid) [Refs. 4, 5 and 8] and in two strains of mice (liver) [Refs. 7 and 10]. Other data supporting the conclusion of likely carcinogenic potential include the limited evidence of genotoxicity for amitrole, the structure-activity correlations based on the triazoles (mouse liver tumors), and the slight relationship to heterocyclic aromatic amines.
    7. Potency factor (Q1*). For the purposes of risk characterization, the low dose extrapolation multi-stage model using the thyroid tumor data in the rat was chosen. None of the mouse liver data were amenable to quantification and, furthermore, the rat seemed to be the more sensitive species as tumors occurred at lower doses. Therefore, the amitrole Q1*, which is the geometric mean of estimates computed separately for male and female rats, was determined to be 1.13 (mg/kg/day)-1 [Ref. 17].

B. Developmental and Reproductive Effects

    Studies on rabbits, rats, and mice were conducted to assess the potential for amitrole to induce developmental effects [Refs. 18, 19, 20 and 21]. In all three species, amitrole is a developmental toxicant at dose levels that are maternally toxic. The rabbit is the most sensitive species. In a gavage study in rabbits, dose levels of 0, 4, 40 and 400 mg/kg/day were tested. Developmental and maternal toxicity were observed at 40 mg/kg/day and higher. Maternal effects included loss of body weight, reduced gravid uterine weight, increased relative liver weight (400 mg/kg/day) and increased number of abortions and total litter resorptions, fewer viable implants per litter, blood on the paperboard and yellow/brown fluid in the amniotic sacs (40 and 400 mg/kg/day). The No Observed Effect Level (NOEL) for maternal toxicity was 4.0 mg/kg/day. Developmental effects included decreased fetal body weights and cleft palate (400 mg/kg/day) and increased numbers of early and late resorptions, malformed fore- and hindlimbs, dome-shaped head, hydrocephaly, lateral scoliosis, curved nasal bones, displaced thalamus and poorly ossified bones in skull, trunk and extremities (40 and 400 mg/kg/day). The developmental NOEL was 4.0 mg/kg/day.
    A dermal developmental study also was conducted in rabbits. Again, in this study, amitrole is a developmental toxicant at levels which are maternally toxic. Rabbits were dosed over 10 percent of the body surface with 1,000, 1,500, or 2,000 mg/kg/day of amitrole during days 7 through 19 of gestation. The NOEL for maternal toxicity was 1,000 mg/kg/day and the Lowest Observed Effect Level (LOEL) was 1,500 mg/kg/day based on decreased body weights on gestation day 20, decreased food consumption from gestation days 10 through 20 and ascites. The NOEL for developmental toxicity was 1,000 mg/kg/day and the LOEL was 1,500 mg/kg/day based on decreased gravid uterine weights, decreased fetal bodyweights, increased number of total resorptions and increased skeletal anomalies including unossified hyoid (skull), unossified pubis, absent and/or unossified talus, and left carotid arising from innominate artery.
    A 2 - generation rat reproduction study was not required by the Agency because, based on current use patterns of amitrole, no reproductive risk to workers was expected. However, the Agency has received a 2 - generation reproduction study (determined to be supplementary) [Ref. 22], in which Sherman rats were fed 0, 25 or 50 mg/kg amitrole in the diet for 55 days and then mated. Decreases in body weights and food consumption, enlarged thyroids and reduced liver and kidney weights were observed in the parents of both treated groups. The number of offspring, mean body weights and survival were reduced in the offspring of both treated groups. No malformations were reported. Because this study does not meet Agency guidelines, a determination on the validity of the noted reproductive effects could not be made.
    The Agency has received a range-finding reproduction study [Ref. 23] in which rats were tested at dose levels of 40, 100, 300 and 800 ppm in the diet. There was maternal toxicity at all dose levels (decreased maternal body weight gains from gestation days 0 - 7 at 100 ppm and above, and goiter and increased thyroid-to-body weight ratios at 40 ppm and above). The mean number of pups was significantly lower at dose levels of 100 ppm and above, and decreases in T4 were seen in F1 males at 40 ppm, and in both sexes at 100 ppm and above. TSH was increased at 100 ppm and at 300 ppm. Increased thyroid weights, hyperplastic goiter and follicular hyperplasia also were observed in the offspring.

C. Public Comments and Agency Responses To The Position Document 1

    A number of comments relating to the toxicity of amitrole were received in response to the PD 1. A summary of those comments and the Agency's responses follow.
    1. Comment. Amitrole is a secondary, not a primary, thyroid and pituitary carcinogen in laboratory animals.
    Response. The Agency believes that while the data in the rat are suggestive of a disruption in the thyroid-pituitary status, the data are neither clear and complete nor consistent. Therefore, the Agency treats amitrole as a primary carcinogen for the thyroid and pituitary. The Agency also believes that amitrole is a primary carcinogen for the liver.
    2. Comment. It is inappropriate to extrapolate thyroid and pituitary carcinogenic effects in laboratory animals to humans.
    Response. The Agency takes the position that humans are at least as sensitive as laboratory animals unless data are developed to establish and quantify any differences in sensitivity.
    3. Comment. Exposure to amitrole must be of sufficient magnitude and duration to elicit thyroid, pituitary, and liver carcinogenic effects.
    Response. The Agency believes that exposure to low doses of amitrole can elicit thyroid and pituitary carcinogenic effects. Thyroid tumors were observed in long-term rat studies at doses as low as 60 ppm in the diet. There are some data that indicate that amitrole may induce an increase in liver tumors in mice at low dose levels as well. The Agency agrees that exposure to amitrole must be prolonged to elicit thyroid and pituitary carcinogenic effects. The Agency believes that extended and continuing exposure is necessary to initiate the liver carcinogenic process.
    4. Comment. Amitrole has no mutagenic potential.
    Response. The Agency believes that while a large number of mutagenicity studies are negative (many of them from studies in bacteria), there are a number of studies with positive results (see Hill et al., 1989). In addition, several studies suggest that amitrole has mutagenic activity when assayed with metabolic activation systems other than the usual liver preparations; for example, ``S9'' mixes. (See Mutagenicity, Unit II. A. 3. of this notice) Overall, the genotoxicity evidence for amitrole from available tests is not entirely negative and there are indications of genotoxic, as well as transformation, activity in mammalian tests.
    5. Comment. The appropriate model to estimate thyroid and pituitary carcinogenic risk is a non-linear model, not the multi-stage model.
    Response. The Agency believes that while the data in the rat were suggestive of a disruption in the thyroid-pituitary status, the data were neither clear and complete nor consistent, and did not support the use of the threshold model. Therefore, for the purpose of risk characterization, a low dose extrapolation multi-stage model has been used for quantification of human risk.
    6. Comment. The Agency inappropriately used modeling to estimate risk from exposure to amitrole.
    Response. The Agency believes that the application of modeling techniques to estimate risk is appropriate for thyroid and pituitary carcinogenic effects. The Agency believes that the data are insufficient to support the use of a NOEL/uncertainty factor methodology for estimation of risk.

III. Occupational and Residential Exposure and Risk and Agency Evaluation of Comments

A. Position Document 1

    In the May 1984 Notice of Special Review (PD 1), the Agency concluded that, except for uses associated with homeowner products, the carcinogenic risk associated with all use patterns and application techniques of amitrole may result in unreasonable adverse effects. The Agency's risk analysis was based on exposure estimates obtained from surrogate studies employing other pesticides with uses and application techniques similar to amitrole, and one exposure study which utilized amitrole.
    When conducting the risk assessment, the Agency assumed that all workers were unprotected; that is, they wore only cotton work clothes, short-sleeved shirts but no hat, gloves or respirator. Fifteen percent of the body surface was assumed to be uncovered. Applicator exposure was calculated from surrogate data where amitrole exposure was determined to be a linear function of the pounds of amitrole active ingredient expected to be used vs. the pounds of applied surrogate active ingredient. When no data were available, applicator exposure was calculated as a function of the duration of application and the concentration of the active ingredient in the spray. Mixer/loader exposures were estimated from surrogate data and were assumed to be proportional to the amount of amitrole active ingredient vs. the amount of surrogate active ingredient. Highway tractor (or truck) applicators and forest helicopter applicators were assumed to be tended by mixer/loader personnel. Except for applicators using home-use pressurized aerosols, all other applicators were assumed to be involved in mixing/
loading operations. The Agency's assumptions were conservative and may have overestimated actual exposure.
    The Agency estimated the exposure for each application technique and each site type on which amitrole was registered. These sites included: rights-of-way, marshes and drainage ditches, ornamentals, and land surrounding commercial, industrial, agricultural, domestic, and recreational premises. The typical exposure was based on average field worker exposure, while minimum and maximum values were based on the highest and lowest exposures observed in the data. Dermal exposure, especially to the hands, constituted virtually all of the total amitrole exposure. Exposure estimates ranged from 1 x 10−1 mg/kg/day for certain industrial uses to 3 x 10−7 mg/kg/day for homeowner uses. The use of protective clothing (coveralls, gloves, hats and boots) was expected to reduce the dermal exposure.
    At the time of the PD 1, there were no data available to estimate the dermal penetration of amitrole. Since dermal exposure was the greatest single source of exposure to workers, this was an important parameter in assessing exposure and therefore risk. Because of the lack of data, the Agency calculated the risk to workers using two assumptions. First, EPA assumed that 100 percent of amitrole would be absorbed, a worst case assumption. Risk estimates under this assumption ranged from 10−6 to 10−5 for homeowners to 10−2 to 10−1 for utility power wagon applicators, industry power wagon applicators, industry knapsack/hand-carry applicators, railroad tanktrain mixer/loaders and highway tractor/truck mixer/loaders. Second, EPA also assumed 0.1 percent dermal absorption, based on the chemical properties of amitrole. Based on this assumption, the risk estimates ranged from 10−9 to 10−8 for homeowners to 10−4 to 10−3 for utility power wagon mixer/loader/applicators, and industry power wagon mixer/loader/applicators.

B. Current Exposure and Risk Estimates

    1. Label, packaging, and use changes. Rhone-Poulenc, the sole U.S. registrant of amitrole, has voluntarily canceled all homeowner products (55 FR 41763, October 15, 1990), and has voluntarily taken actions which have reduced worker exposure to amitrole. These actions include deleting the high exposure application methods, such as knapsack sprayers, which had previously resulted in the highest risk. Rhone-Poulenc also adopted a ``no-glug'' container to reduce splashing while pouring, added protective clothing requirements to product labels and canceled all uses except on rights-of-way, public utilities, nurseries, and land surrounding commercial, industrial and farm premises. The current risk assessment for amitrole is for workers mixing, loading and applying amitrole to highway rights-of-way, the use with the greatest exposure and, therefore, risk.
    2. Current exposure estimates and assumptions. The current assessment is based on worker exposure studies conducted for bromoxynil [Ref. 17], another herbicide. The Agency used the bromoxynil worker exposure study as a surrogate for the liquid formulations of amitrole because: (1) The container design for amitrole was amended so it is similar to that of bromoxynil; (2) the label restrictions for protective clothing and engineering controls for amitrole are comparable in effectiveness to those for bromoxynil; (3) the application equipment used to apply amitrole to rights-of-way is similar to that used for bromoxynil; and (4) the ground boom equipment used for bromoxynil agricultural sites would likely result in similar or greater exposure as compared to equipment used for amitrole on highway rights-of-way.
    In general, the exposure assessment assumed that workers wore clean cotton (or cloth) coveralls over long sleeve shirts and long pants in addition to boots or sturdy footwear. For workers handling concentrated product (mixer/loaders) or when repairing and cleaning the equipment used with this product, chemical resistant gloves were also assumed to be used. These protective clothing requirements are on the current labels.
    Usage data utilized for this assessment indicate that the typical usage per person on highway rights-of-way is 100 lbs active ingredient (ai) per day and 1,000 lbs ai/year. The typical application rate is assumed to be 2.5 lbs ai/acre and 40 acres are treated per day. A typical worker is assumed to weigh 70 kg.
    Based on these assumptions, the total yearly exposure to liquid formulations for a 70 kg worker is 2.2 x 10−2 mg/kg/year. The average daily exposure is 5.9 x 10−5 mg/kg/day. The Agency believes that the exposure from the water soluble packets will be far less than the exposure incurred by mixer/loaders from the liquid formulation. Exposures of applicators to mixtures from either liquid or wettable powder formulations are assumed to be the same. Therefore, the exposure estimate of 5.9 x 10−5 mg/kg/day for the liquid is judged to be the highest exposure estimate for all amitrole formulations.
    3. Current risk estimates-- a. Carcinogenic risk. The availability of the bromoxynil exposure studies and the dermal penetration study has allowed EPA to refine its risk estimates. In the PD 1, in the absence of a valid study to demonstrate otherwise, dermal penetration was assumed to be 100 percent. A dermal penetration study on amitrole was subsequently received and reviewed, and was used in the current risk assessment. Dermal penetration in this study was shown to be 0.1 percent [Ref. 24]. Additionally, EPA estimates that the unit cancer risk (Q1*) is 1.13 (mg/kg/day)−1.
    When calculating the Lifetime Average Daily Dose (LADD), EPA assumed a worker life span of 70 years with an amitrole exposure period of over 35 years. Using these values, and an average daily exposure estimate of 5.9 x 10−5 mg/kg/day, the LADD was determined to be 2.9 x 10−5 mg/kg/day.
    The excess lifetime cancer risk to workers involved in mixing, loading, and application of the liquid formulations of amitrole for highway rights-of-way uses is estimated to be 3.3 x 10−5 [Ref. 17].
    b. Other risks. Three developmental studies, two oral and one dermal, are available for amitrole. Because the primary route of exposure is dermal, the most appropriate study to assess potential developmental effects to workers exposed to amitrole is the dermal study. Using the NOEL of 1,000 mg/kg/day from this study to calculate margins of exposure, EPA determined that workers would not be at risk for developmental effects from exposure to amitrole.

C. Public Comments and Agency Responses to the Position Document 1

    Comments relating to exposure to amitrole were received in response to the PD 1. A summary of those comments and the Agency's responses follow.
    1. Comment. The worker exposure estimates used for the Registration Standard and the Position Document 1 in 1984 are unrealistic and represent a worst case exposure scenario.
    Response. The Agency acknowledges that, in the absence of more specific available data, its 1984 worker exposure estimates were conservative and represented a worst case scenario. Both the Registration Standard and the PD 1 specifically state that calculations may overestimate exposure. Since 1984, the Agency has received better data that allow a more accurate estimate of exposure and risk. The Agency originally assumed 100 percent dermal penetration; actual data generated in 1985 by the registrant showed the dermal penetration to be 0.1 percent. Additionally, exposure data from a bromoxynil study were used as a surrogate for amitrole, which allowed the Agency to more accurately estimate exposure to mixers/loaders/
applicators.
    2. Comment. The amitrole exposure study [Ref. 25] demonstrated that hands received only 6 percent of the total dermal exposure. Lower legs, chest and thighs received 94 percent of the total exposure. Respiratory exposure was insignificant (less than 0.1 percent of the total exposure).
    Response. The Agency disagrees with this and estimates hand exposure at approximately 96 percent. The estimate in Baugher (1982) was based on an application time of 2.36 hours with applicators using latex gloves. The actual application time was 7 hours with applicators wearing cotton gloves that absorb and retain moisture, with latex gloves worn underneath. The Agency noted that the exposure estimate was based on the residue on the outside of the latex gloves. The forearms, chest, and inhalation account for the balance of exposure. The Agency concurs that inhalation represents less than 0.1 percent of the total exposure.
    3. Comment. Normal exposure will not alter thyroid function of workers.
    Response. The Agency concurs that typical mixing/loading/
application practices of certified applicators coupled with protective clothing and the exposure reduction measures will reduce exposure to levels which will not likely alter worker thyroid function.

 IV. Summary of Benefits and Evaluation of Alternatives

 A. Importance of Amitrole

    Benefits of amitrole include its relatively low cost, broad spectrum control of newly emerged or established broadleafs and its miscibility with other low cost, broad spectrum residual soil active chemicals. Amitrole controls newly emerged or established broadleafs because it is a contact herbicide that kills growing vegetation. Amitrole is mixed with residual herbicides because its short 2 - 4 week half-life precludes effectiveness against later-germinating weeds. Amitrole provides nonselective weed control when used alone or in combination with longer lasting herbicides on highway berms, guard rails, around sign posts, railroad beds, and similar areas. Highway rights-of-way sometimes require total vegetation control which is generally achieved through use of nonselective herbicides such as amitrole in tank mix combination with a soil residual herbicide.

B. Usage of Amitrole

    Amitrole is imported, not produced in the United States. Rhone-Poulenc is the only U.S. importer of amitrole. Domestic usage has been falling throughout the 1980's and 1990's, partially due to the 1984 classification of amitrole as a Restricted Use pesticide. The decline in use of amitrole may also be due to the recent registration of other herbicides with a broader spectrum of weed control. The EPA estimated that annual usage of amitrole in 1984 was between 500,000 and 800,000 pounds but, by 1989, had decreased to between 50,000 and 100,000 pounds of active ingredient. Total annual usage of amitrole declined even further in 1990 to between 40,000 and 60,000 pounds active ingredient. It is estimated that 80 to 90 percent of amitrole use is for highway rights-of-way. The remaining usage is divided among many minor uses which include landscape management, industrial areas and recreational areas. In the late 1970's, amitrole was used widely on railroad, highway and utility rights-of-way, but currently, its major use is along highways [Ref. 26].
    According to the registrant, voluntary cancellation in 1991 of its California registration was due to its estimates that the projected sales volume of amitrole did not justify expenditures needed to comply with California data requirements [Ref. 27]. The registrant estimated at the time of cancellation that 90 percent of the pesticide's market share existed in that state. Concurrently, Rhone-Poulenc requested from the Agency, and was granted, a 2 - year existing stocks provision to allow products currently in channels of trade to be distributed and sold in California until May 1993. It is expected that the voluntary cancellation in California will accelerate the current declining trend in usage of amitrole.

C. Alternatives Assessment

    Amitrole rights-of-way alternatives are divided into two classes: chemical and mechanical control.
    1. Chemical control. The major alternatives to amitrole are glyphosate, sulfometuron-methyl, diuron, imazapyr, and hexazinone [Ref. 28]. EPA has classified glyphosate as an E carcinogen; that is, evidence indicates that glyphosate is not carcinogenic in humans. Sulfometuron-methyl has not been assigned a carcinogenic classification, but may pose developmental risks. Many of the alternatives to amitrole have outstanding data requirements, thus a satisfactory comparison of the risks for all of the alternatives has not be completed by the Agency. However, glyphosate, at present, appears to be less toxic to humans than amitrole or its alternatives.
    Alternatives to amitrole are generally more expensive, and the loss of amitrole from the market could result in increased cost of weed control. Amitrole alternatives provide equivalent, or in some cases, better control of certain weeds with the exception of poison ivy control, for which amitrole is the most efficacious herbicide. If the most likely alternatives, glyphosate and sulfometuron-methyl, were substituted for amitrole, there would be a chemical cost increase ranging from $12.02 to $40.34 per acre treated [Ref 27].
    An estimated 30,000 to 50,000 acres of rights-of-way in the United States are currently treated with amitrole. Total chemical cost of maintaining rights-of-way in the United States is estimated to be $224 million per year [Ref. 29]. If glyphosate or sulfometuron-methyl were substituted for amitrole, the aggregated chemical cost increase may range from $600,000 for glyphosate to a chemical cost increase of $2 million for sulfometuron-methyl. Thus, although the other chemical alternatives offer similar control (except for poison ivy), the cost per acre is significantly higher.
    2. Mechanical control. Controlling brush on highway rights-of-way by manual or mechanical cutting and controlled burning are high cost alternatives, especially at sites along guardrails and steep slopes, the precise areas of herbicide use.

V. Risk/Benefit Analysis

A. Summary of Risk

    EPA has evaluated the risk posed by amitrole to workers mixing, loading and applying the pesticide to highway rights-of-way. Although amitrole is registered for several additional sites, this use pattern is the one that poses the greatest exposure, and therefore risk. EPA has estimated the excess lifetime cancer risk from this exposure to amitrole to be 3.3 x 10−5.

B. Summary of Benefits

    Amitrole is a relatively inexpensive and efficacious broad-spectrum herbicide. If amitrole were unavailable, growers would have to use more expensive alternatives (e.g. glyphosate, sulfometuron-methyl) as substitutes, with costs increasing from $600,000 to $2 million/year. Also, many of these alternatives may not provide as effective control of certain weeds as amitrole.

C. Cost-Effectiveness

    A cost-effectiveness analysis was developed by the Agency comparing amitrole and its two main alternatives, glyphosate and sulfometuron-methyl [Ref. 30]. The cost-effectiveness analysis reflected two scenarios that took under consideration three possible occupational exposure levels. The first scenario gave a point estimate for 250 applicators (an estimate provided by the registrant as the total annual number of workers exposed, before cancellation in California) and the second scenario utilized 100 applicators (an EPA estimate of the total number of exposed workers, before cancellation in California). The analysis also considered three scenarios for length of exposure: 1 year, 5 years and 35 years. Typically, 1 year of exposure represents college students who are exposed to amitrole only during summer work programs. Five years represents the applicator who finds different employment after several years. Thirty-five years depicts the applicator who is employed in the same profession throughout a lifetime. The exposure figure of 35 years was used for calculating the present risk assessment. This resulted in an estimated risk of 3.3 x 10−5. However, the Agency believes that its assumption that worker exposure occurs for 35 years is conservative and tends to overestimate typical exposure. A more likely length of exposure for a typical worker is considered to be 5 years; the risk estimate for amitrole based on 5 years of exposure is calculated to be 5 x 10−6.
    For all scenarios, the upper end of the treated acreage was used in defining the low and high cost per cancer case avoided, if amitrole was canceled and replaced by one of the two principal alternatives. The cost-effectiveness estimate under the scenario of 250 applicators for a 35 year exposure period would range from $73 million to $244 million per cancer case avoided. The risk resulting from 5 years of exposure, the more likely scenario, would increase these costs associated with avoiding an extra cancer case.

D. Conclusions

    Based on its risk and benefits assessment, the Agency has concluded that the benefits provided from the use of amitrole outweigh the risks. The cost-effectiveness analysis supports this conclusion by demonstrating the extremely high costs of avoiding a cancer case that would be incurred with cancellation of amitrole.

VI. Agency's Decision Regarding Special Review

    Because EPA has concluded that the risks of amitrole are outweighed by the benefits of continued use, EPA proposes that the Special Review based on potential carcinogenic risk of amitrole to workers be concluded. Use of bromoxynil data as a surrogate for amitrole data and the data on dermal penetration have enabled the Agency to refine the amitrole risk assessment to be more realistic than the worst-case assessment conducted for the PD 1. The label modifications and risk reduction measures taken by the registrant, Rhone-Poulenc, have significantly reduced worker exposure to amitrole. Consequently, there is a corresponding reduction in the risks posed by the remaining uses of amitrole. Even given the differences in risk between amitrole and its most likely alternatives, EPA believes that the 3.3 x 10−5 risk associated with 35 years of exposure to amitrole is outweighed by the benefits of use. The Agency believes that amitrole's remaining uses pose no significant threat to workers or the general public, and that its significantly lower costs and higher efficacy rates as compared to its alternatives merit retention of the remaining uses. However, because of the positive carcinogenicity studies, the Agency will continue to require that amitrole remain a Restricted Use pesticide, that the cancer warning statement remain in place, that the current application method remain limited to boom sprayers and that present protective clothing requirements remain on labelling.

VII. Public Comment Opportunity

    During the 30 - day comment period, specific comments are solicited on the preliminary determination set forth in this Notice. The Agency will review and consider any comments received during the official comment period before issuing the final determination to conclude the Special Review of amitrole. Interested persons are invited to submit written comments on this proposal to conclude the Special Review of pesticide products which contain amitrole. All comments and information should be submitted in triplicate by [Insert date 30 days after date of publication in the Federal Register] to the address given in this Notice under the ADDRESS section. The comments and information must bear the document control number, ``OPP - 30000/38A; FRL 4164 - 1.'' All written comments filed pursuant to this notice, except ``CBI'', will be available for public inspection in Rm. 1132, Crystal Mall #2, 1921 Jefferson Davis Highway, Arlington, VA. 22202, from 9 a.m. to 4 p.m., Monday through Friday, excluding legal holidays.

VIII. References

    (1) U.S. Environmental Protection Agency (USEPA). Amitrole Pesticide Registration Standard and Guidance Document. March 30, 1984.
    (2) U. S. Environmental Protection Agency (USEPA). Federal Register Notice (55 FR 41763). October 15, 1990.
    (3) Cox, G.E., Re, T.A. Inhalation Toxicity Study with 3-Amino-1,2,4-Triazole (Amitrole) in Adult Fischer 344 Rats. (Unpublished study prepared by Food and Drug Research Laboratories, Inc. submitted by Union Carbide). MRID #GS009502. (1978).
    (4) Keller, J.G. Final Report: Chronic Feeding-Rats. (Unpublished study received Dec. 20, 1958 under PP0210; prepared by Hazleton Laboratories, submitted by American Cyanamid Co. New York, N.Y.; CDL:090236-F). MRID #0082176. (1959).
    (5) Johnson, W.D., Becci, P.J., Parent, R.A. Lifetime Feeding Study of Amitrole in Fischer 344 Rats. Laboratory No. 5651. (Unpublished study prepared by Food and Drug Research Laboratories, Inc.; submitted by Union Carbide Agricultural Products Co.). MRID #GS009504. (1981).
    (6) Steinhoff, D., Weber, H., Mohr, U., Boehme, K. ``Evaluation of Amitrole (Aminotriazole) for Potential Carcinogenicity in Orally Dosed Rats, Mice, and Golden Hamsters.'' Toxicology and Applied Pharmacology. 69:161 - 169. (1983).
    (7) Steinhoff, D., Boehme, K. Amitrole Cancerogenesis Test with Oral Administration to Mice. Prepared by Bayer AG, submitted by Union Carbide Agricultural Products Co. (1978).
    (8) Steinhoff, D., Boehme, K. Amitrole Cancerogenesis Test with Oral Administration to Rats. Prepared by Bayer AG, submitted by Union Carbide Agricultural Products Co. (1979).
    (9) Steinhoff, D., Boehme, K., Lorke, D., et. al. Cancerogenesis Study on Orally Dosed Golden Hamsters. Prepared by Bayer AG, submitted by Union Carbide Agricultural Products Co. (1978).
    (10) Innes, J.R. et. al. ``Bioassay of Pesticides and Industrial Chemicals for Tumorigenicity in Mice: A Preliminary Note.'' Journal of the National Cancer Institute. (1969).
    (11) Napalkov, N.P. Blastomogenic Effect of 3-Amino-1,2,4-Triazole. Gig. Tr. Prof. Zabol. 6(6):48 - 51. (1962).
    (12) Vesselinovitch, S.D. Perinatal Carcinogenesis. Biol. Res. Pregnancy Perinatology. (1983).
    (13) Hill, R.N., et. al. Thyroid Follicular Cell Carcinogenesis. Fundamental and Applied Toxicology. 12: 629 - 697. (1989).
    (14) Tsutsui, T., et. al. Amitrole-induced cell transformation and gene mutations in Syrian hamster embryo cells in culture. Mutation Research. 140:205 - 207. (1984).
    (15) Daston, D., et. al. Peroxidase-mediated metabolic activation of amitrole. Environ. Mutagen. 8 (Suppl 6):21 - 22. (1986).
    (16) Krause, R. and Eling, T. Macromolecular binding of the thyroid carcinogen 3-amino-1,2,4-triazole (amitrole) catalyzed by prostaglandin H synthase, lactoperoxidase and thyroid peroxidase.Carcinogenesis. 8:659 - 664. (1987).
    (17) Schlosser, A. Memorandum to Philip Poli (USEPA). Assessment of Risk to Workers from Exposure to Amitrole Used on Highway Rights-Of-Way. April 24, 1991.
    (18) Percutaneous Teratology (Developmental Toxicity via Dermal Route of Administration). Study No. HLA 6224 - 107. Unpublished report prepared by Hazleton Laboratories America submitted by Rhone-Poulenc Ag Company. (1988).
    (19) Tjalve, H.; Fetal Uptake and Embryogenetic Effects of Aminotriazole in Mice: Archives of Toxicology. 33:41 - 48. (1974).
    (20) Tyl, R.W.; Bushy Run Research Center, Teratogenicity evaluation of Aminotriazole Technical Administered by Gavage to CD Rats. (1986).
    (21) Tyl, R.W.; Bushy Run Research Center, Teratogenicity evaluation of Aminotriazole Technical Administered by Gavage to New Zealand White Rabbits. (1986).
    (22) Gaines, T.B., Kimbrough, R.D., Linder, R.E. ``The Toxicity of Amitrole in the Rat.'' Toxicology and Applied Pharmacology. 26:118 - 129. (1973).
    (23) Amitrole Report to USEPA. MRID #418909 - 00, 418909 - 01.
    (24) Puhl, R.J., Tisdel, M. Dermal Absorption of C14 Amitrole in Male Rats. Study No. 61589 - 105. (Unpublished study prepared by Hazleton Laboratories America submitted by Union Carbide). MRID #151651. April 19, 1985.
    (25) Baugher, D.G., Bookbinder, M.G., Blundell, M.S. Medical Monitoring and Assessment of Exposure of Workers Applying Amitrole to Utility Rights-of-Way. (Unpublished study received Nov. 18, 1982 prepared by Dynamac Corporation; submitted by Union Carbide Agricultural Products Company, Inc., Research Triangle Park, NC). MRID #GS009517. (1982).
    (26) Jennings, Allen. Memorandum to Rick Tinsworth (USEPA). Preliminary Benefit Analysis of Amitrole. January 18, 1990.
    (27) Griffin, Karen. Memorandum to Special Review Team (USEPA). Summary of 1991 Amitrole Use, Usage, and General Benefits. July 1991.
    (28) Esworthy, Robert. Memorandum to Philip Poli (USEPA). Amitrole Alternatives. January 3, 1992.
    (29) Saulmon, J.G. Memorandum to Philip Poli (USEPA). EPA Estimate of Range of Dollar Costs of Maintaining Rights-of-Way in the USA. December 3, 1991.
    (30) Griffin, Karen. Memorandum to Philip Poli (USEPA). Cost Effectiveness/Benefit Analysis for Amitrole. February 6, 1992.

    All but the published references concerning this Preliminary Determination on amitrole are available for inspection from 8 a.m. to 4 p.m., Monday through Friday, except legal holidays, in the Public Response and Program Resources Branch, Field Operations Division (H7506C), Office of Pesticide Programs, Environmental Protection Agency, Room 1128, Crystal Mall #2, 1921 Jefferson Davis Highway, Arlington, VA 22202.

    Dated: September 30, 1992.

Victor J. Kimm,
Acting Assistant Administrator for Prevention, Pesticides and Toxic Substances.

[FR Doc. 92 - 24555 Filed 10 - 7 - 92; 8:45 am]
BILLING CODE 6560 - 50 - F