1 Tier 1 EDSP: Other Scientifically Relevant Information Barbara Neal Exponent December 13, 2010

2 Endocrine Disruptor Screening Program  Tier 1 Test Orders for 67 chemicals, mostly pesticides  Tier 1 screening assays include testing for estrogen, androgen and thyroid (EAT) system interactions  Specific in vitro binding and or transactivation assays (ER and AR)  Less specific in vitro assays (aromatase inhibition and steroidogenesis)  Relatively specific in vivo assays (uterotrophic for estrogenic activity); Hershberger for androgenic and anti-androgenic activity); frog metamorphosis (anti-thyroid activity)  Apical assays (male and female pubertals); fish reproduction 2

3 Available Data on Pesticides  Pesticides have an extensive GLP mammalian toxicological and ecotoxicological data base developed under FIFRA Part 158 requirements  Multiple studies from this data base provide apical information regarding the potential for interaction with the EAT systems  EPA has been developing specific in vitro information on many of these pesticides under the ToxCast program  Published in vivo and in vitro data of varying quality are also available for many pesticides 3

4 OSRI Other Scientifically Relevant Information  OMB encouraged EPA to consider Other Scientifically Relevant Information (OSRI) in deciding whether any Tier 1 assays could be waived  EPA issued very limited guidance on what would be considered acceptable OSRI  Functionally equivalent data exist for many Tier 1 assays for pesticides 4

5 EPA OSRI Definition “Other scientifically relevant information” is information that informs the determination as to whether the substance may have an effect that is similar to an effect produced by a substance that interacts with the estrogen, androgen, and/or thyroid hormonal systems (e.g., information that identifies substances as having the potential to interact with the estrogen, androgen, and/or thyroid system(s); information demonstrating whether substances have an effect on the functioning of the endocrine system). OSRI may either be functionally equivalent to information obtained from the Tier 1 assays—that is, data from assays that perform the same function as EDSP Tier 1 assays—or may include data that provide information on a potential consequence or effect that could be due to effects on the estrogen, androgen or thyroid systems.” 5

6 Examples of OSRI Submitted  1998 Guideline Reproductive Toxicity : Equivalent to Tier 2 Mammalian Reproduction Study  Pre-1998 Guideline Reproductive Toxicity: Includes multiple endpoints: mating, fertility indices; pup sex ratio; reproductive organ weights and histopathology  Time to mating data provides information on estrous cyclicity  DNT: provides information on VO and PPS  Sub-chronic toxicity: reproductive organ, thyroid adrenal, pituitary organ weights and histopathology  Chronic toxicity/oncogenicity: increases or decreases in endocrine sensitive tumors  Fish Life cycle study: fertility, fecundity; behavior; development 6

7 Part 158 Studies Predictive for Endocrine Effects  Anti-androgens used as weak positive controls identified in multigeneration reproductive toxicity studies  Linuron  Vinclozolin  Owens and Ashby (2002) reviewed the predictive value of the uterotrophic assay by comparison with observations made in the multigenerational reproductive and developmental toxicity assays. They compared no observed effect levels (NOEL) and lowest observed effect levels (LOEL) for uterotrophic and rat multigeneration studies, and found high concordance for NOEL and LOEL values for estrogenicity. 7

8 EPA OSRI Evaluation Process Ongoing and Late  OSRI for 15 chemicals evaluated to date; responses for some are published   Many EPA OSRI responses are more than 6 months late; EPA initial estimate 90 days for evaluation from receipt  EPA indicated challenges: “the diversity of approaches to OSRI, as well as the volume and frequency of the order/DCI responses and short time frame (90 days). Including OSRI from public.”  EPA Data table for OSRI responses incomplete  Some reviews unavailable; no links provided  Some link to incorrect chemicals 8

9 OSRI Evaluation Process at EPA Ongoing and Changing  Appears to be improvisational approach to OSRI evaluation and approved options  Example: Bypass option (move directly to Tier 2) issued for several chemicals, then withdrawn  As noted above, still a limited number of OSRI evaluations to draw conclusions from, but some patterns appear to be emerging 9

10 Patterns in EPA OSRI Evaluations  Appears likely that the 1998 Two-generation study will be accepted as a Tier 2 mammalian study  To date these have been accepted as the basis for waivers from mammalian in vivo testing in at least two instances  In both cases the chemicals were found to show potential interaction with the EAT systems 10

11 OSRI ‘Deficiencies’ Not Addressed by Tier 2 Mammalian Study  Thyroid evaluations faulted because a 5 point subjective scale was not used for histopathologic characterization of thyroid lesions, including follicle height and colloid  Current 2-generation study does not require thyroid evaluations  Comparative thyroid studies do not specify subjective scale for evaluation  Examination of follicle height and colloid are key components of analyses, regardless of qualitative scale 11

12 OSRI ‘Deficiencies’ Not Addressed by Tier 2 Mammalian Study  Benfluralin and malathion: “organ weight data are inadequate because they were obtained from adult animals whereas in the Tier 1 pubertal studies these weights are obtained in pubertal animals.”  1998 Reproductive toxicity guideline requires organ weights and requires histopathological evaluations in weanling and adult animals, not in pubertal animals.  Benfluralin: “LABC and ventral prostate weight not obtained”  Endpoints not required in the 1998 mammalian 2-generation reproductive toxicity study 12

13 Patterns in OSRI Responses—ToxCast Data EPA ToxCast data uniformly rejected, even in context of extensive other supporting information ToxCast assays include in vitro ER and AR binding and transactivation; multiplex assay; thyroid activity; aromatase ToxCast assays used to screen potential endocrine toxicity of oil dispersants used in the Gulf 13

14 Patterns in OSRI Responses—ToxCast Data In most EPA OSRI responses a generic letter from Kavlock is cited providing no detail for the rejection of ToxCast data, however in one recent decision more detail is provided: “The cited ToxCast data may be appropriate for use in priority setting for in vitro assays, however protocol information on individual assays is limited and there is no indication of the sensitivity or specificity of ToxCast, e.g., demonstrated by how weak positive, negative controls and proficiency chemicals would perform in this assay.” Dimethoate OSRI EPA review Nov. 10,

15 Patterns in EPA OSRI Evaluations In vitro studies: several published in vitro assays have been accepted in lieu of Tier 1 screening assays. In these cases there is a strong parallel between the Tier 1 assay methodology and that of the published study Negative in vivo studies: uniformly rejected if any deviations from Tier 1 guidelines, regardless of their quality, consistency of the findings in multiple studies or commitment to perform other Tier 1 assays evaluating closely related endpoints Positive in vivo studies: two accepted as basis for waivers despite poor study quality and/or marked divergence from Tier 1 guidelines 15

16 Patterns in EPA OSRI Evaluations  No WoE process evident for OSRI evaluation  Studies considered individually, without consideration of other studies with the same or closely functionally related endpoints, and  without consideration of any Tier 1 assays the registrant has agreed to perform  Weak non-specific positive responses, or responses in poor quality studies, have been used as indications that an EAT pathway may be affected 16

17 Lack of WoE Evaluation– Specificity and Consistency of Response  Non-specific ‘positive’ responses have been used as indications that an EAT pathway may be affected:  Decreased ovary weights in chronic toxicity/oncogenicity study without evidence of histopathological change  Slight inconsistent decreased fertility in one of two two generation studies  No effects on more specific and/or robust endpoints including :  ER binding,  estrous cyclicity,  time of vaginal opening, and  quantitative ovarian histopathology 17

18 Lack of WoE Evaluation— Consideration of Study Quality  For another compound, a single study from the published literature is used to indicate the material “may have the potential to interact with the thyroid pathway;” and a second very high dose study is used to indicate it “may have the potential to interact with the androgen pathway” regardless of substantial other data not supporting such effects  In combination these published studies are used as the basis to waive a male pubertal assay. There is no indication EPA examined the quality of either study, and neither study corresponds to a male pubertal assay in study design. 18

19 Patterns in EPA OSRI Evaluations  Lack of a WoE approach sets a poor precedent for evaluation of Tier 1 screening results.  OSRI evaluations to date predict any positive finding in an in vivo Tier 1 or OSRI study will lead to a “potential endocrine pathway activity” finding and likely to Tier 2 testing 19