Three Pre-Rule Studies of Chlorpyrifos: Nolan et al. (1982) Honeycutt & DeGeare (1993)

1. Three Pre-Rule Studies of Chlorpyrifos: Nolan et al. (1982) Honeycutt & DeGeare (1993) Kisicki et al. (1999) Human Studies Review Board June 24, 2009 1

2. Sequence of Presentations • Introduction and Context • Anna Lowit, Ph.D. • Science Assessments • Nolan and Kisicki Studies • John Doherty, Ph.D., DABT • Honeycutt & Degeare Study • Wade Britton, MPH • Ethics Assessments • John Carley

3. Chlorpyrifos Introduction and Context Anna Lowit, Ph.D. Senior Scientist Health Effects Division Office of Pesticide Programs 3

4. Introduction • Chlorpyrifos: • Organophosphate pesticide which was first registered in 1965 • In June 2000: • The technical registrants entered into an agreement with the Agency to eliminate and phase out nearly all uses that result in residential exposures. • Human health risk assessment developed for the Interim Registration Eligibility Decision (IRED) relied on adult ChE data from rodents & dogs • Human studies were not used to inform point of departure or uncertainty factors • Honeycutt study used in the worker exposure assessment

5. Introduction • Current regulatory activities leading to new risk assessment: • Registration review: 15-year review cycle under FIFRA for registered pesticides • Update human health & ecological risk assessments • Petition by Natural Resources Defense Council (NRDC) and Pesticide Action Network, North America (PANNA) to revoke all tolerances and cancel all registered uses

6. Introduction • Draft Science Issue Paper reviewed by the FIFRA SAP in 2008 • Review the new science from animal & humans under the context of human health risk assessment • Focus on effects in pregnant women, fetuses, and juveniles as these groups are thought to be more susceptible to chlorpyrifos • Age-dependant metabolism • Epidemiology studies in mothers & children • Rodent studies evaluating non-cholinergic toxicities (i.e., behavior, learning, biochemical responses) • AChE studies in pregnant rats, fetuses, post-natal pups

7. Human Experimental Studies 7

8. Proposed Uses of the Human Studies • For “Bounding” Analyses • Comparing blood & urine data from the human experimental studies with data from animal studies & human epidemiology studies • Comparing levels of AChE/ChE inhibition in humans and animals • To Develop & Refine PBPK models • Physiologically-based pharmacokinetic models • Current models include Nolan et al data • Kisicki et al or Honeycutt & DeGeare not used in current model parameterization 8

9. EPA does NOT propose to use data from the human experimental studies for a point of departure (PoD) or to directly inform the inter-species uncertainty factor • Animal studies provide high-quality dose-response data for ChEI across many doses & multiple life stages • Human studies lack dose-response information • Nolan et al used only one dose level for each route of administration • Kisicki et al showed ChEI in only one subject • Honeycutt & DeGeare was not designed for dose response • Human studies do not address non-cholinergic toxicities • Animal data indicate susceptibility of the developing nervous system to chlorpyrifos • Epidemiology studies in children generally support the animal studies 9

10. Using Human Biomarker Data:Bounding Estimates Rodent Administered Dose, Known & Controlled Human Deliberate Dosing Administered Dose, Known & Controlled Human Epidemiology Exposure, Limited & Uncertain Blood &/or Urine Measures Blood &/or Urine Measures Blood &/or Urine Measures 10

11. Physiologically-Based Pharmacokinetic (PBPK) Models • Represent the anatomy & physiology of the rodent/human • Provide simulations of biological processes such as absorption, distribution, metabolism & elimination • Widely recognized as the “gold standard” in human health risk assessment • Particularly helpful in extrapolations: • Route to route • Inter-species • Across dose range 11

12. SAP Response • SAP was generally supportive of EPA’s preliminary conclusions, and identified areas for revision & additional analysis • “Overall, the Panel agreed that the human deliberate dosing studies contain scientifically useful information for risk assessment, but not for directly establishing PoD or uncertainty factors.” • “The Panel appreciated the Agency’s scientific analysis to compare the blood levels in the deliberate dosing and epidemiological studies, and considered it critically important to maximally use the information from these studies . . . as a basis to “bound” the reference doses/concentrations . . .” • “The Panel encouraged the Agency to consider the use of a PBPK model to widen the application of these bounding data for current or potential human exposures and for the final reference dose or reference concentrations.” 12

13. Human Experimental Studies • Nolan et al (1984) • Historically used to interpret biomonitoring studies (e.g., NHANES & worker) • Provides estimate of dermal absorption • Used in current PBPK models for inter-species scaling

14. Kisicki et al (1999) Lack of plasma ChE is a critical omission from study design decreases its utility Possible reduced absorption of chlorpyrifos from capsule Used in 2002 Timchalk et al PBPK paper For purposes of model evaluation Not used in current parameterization Human Experimental Studies

15. Human Experimental Studies • Honeycutt & DeGeare • Will be used in combination with other available worker biomonitoring studies to evaluate range of urinary TCP concentrations for workers

16. Human Experimental Studies 16

17. The Nolan et al. (1982) and Kisicki et al. (1999) Chlorpyrifos Single Dose Studies in Human Volunteers: Science Assessment • John Doherty, PhD, DABT • Health Effects Division • Office of Pesticide Programs

18. Scope of Presentation The reliability of the analytical data for chlorpyrifos and TCP* and the assessment for ChE and its inhibition with some emphasis on individual variability will be presented. *TCP = 3,5,6-trichloro-2-pyridinol – the principal metabolite of chlorpyrifos.

19. Study Information - Basic Protocols

20. Study InformationChE activity and analysis for chlorpyrifos and TCP

21. Results –Analysis for ChlorpyrifosLevels are near the LOQ- Reliability fair.

22. Results: Analysis for TCPRanges show individual variation

23. Results – Plasma ChE Inhibition and TCPNolan Oral Study • Basal values 0.87±0.09 to 1.42±0.17 are reasonable. • All six subjects maximum 71 to 89% inhibition (good agreement) but time to peak varies (i.e. 6-24 hours). • Maximum blood TCP (715 to 1430 ng/mL) usually before maximum inhibition. • Approximately 700 – 800 ng/mL TCP in blood needed for about 57 to 63% inhibition for two subjects. But 996 ng/mL associated with only 30% inhibition in another subject. • Correlation of urine (µg/hour) TCP with blood (µg/mL) and with inhibition confounded because of units and times of collection and ChE assessment.

24. RBC AChE inhibition in the Kisicki Study. • Basal values (i.e. ♀ group means) of 8576±556 to 9165±709 are reasonable. • 1.7 to 5.6 ng/mL chlorpyrifos - no inhibition. • Blood TCP up to 1300 ng/mL not associated with inhibition. • Urine TCP up to 15,323 ng/mL in one subject did not show inhibition. • Only one subject displayed RBC AChE inhibition. • This subject had highest gastro-intestinal absorption. • Inhibition starts to peak before chlorpyrifos and TCP in the blood and urine peak (next slide).

25. Kisicki Study : RBC AChE inhibition – correlation with chlorpyrifos in blood (ng/mL) and TCP in blood and urine).

26. Chlorpyrifos and ChE inhibition Chlorpyrifos ↓ [Chlorpyrifos Oxon] (Rapid irreversible inhibition of ChE/AChE) ↓ TCP (Also from other pathways)

27. Blood and urine levels of TCP in females dosed with 2 mg/kg.

28. Dermal dosing (Nolan study only) • Borderline inhibition effect on plasma ChE in 3 of 5 subjects, maximum 26% decrease. RBC AChE not inhibited. • Blood TCP at 122 ng/mL with 21% decrease but 36 ng/mL with 26% decrease. No correlation. • Recovery in urine as TCP: 5 mg/kg (five subjects): 1.02±0.57% 0.5 mg/kg (one subject): 2.6%.

29. Some Applications of the Nolan and/or Kisicki Studies • Supports Agency use of low dermal absorption factor (Nolan). • Demonstrates BuChE is more sensitive than RBC AChE in humans (Nolan and Kisicki). • May support PBPK models. • May support “bounding” • Demonstrates variability of humans to absorb chlorpyrifos from the g-i tract.

30. Summary - Strengths • Technical Assessment for ChE/AChE should be reliable in both studies (agree with literature and reasonable CV). • Technical analysis for TCP in blood and urine should be reliable (temporal response in both studies and reasonable dose response in Kisicki).

31. Summary – Limitations: Both Studies • Analytical methods (nano range) are much less sensitive than the epi studies (pico range)*. • Variability in TCP analysis – humans are not equal*. *Provides challenge for interpreting epi studies. • Comparison between Nolan and Kisicki studies confounded because of tablet vs. capsule dosing. • Chlorpyrifos present near LOQ and in some 0 time samples (Nolan). Reliability only fair.

32. Summary – Limitations: Nolan Study • Only one dose resulting in ~70-89% inhibition. • Does not establish NOEL. • Difficult to establish minimal levels of TCP associated with inhibition. • TCP is in units/hour, epi and Kisicki report units/mL. Do not easily compare.

33. Summary –Limitations: Kisicki Study • Does not include plasma ChE assessment. • Only one subject with RBC AChE inhibition limits usefulness.

34. Honeycutt & DeGeare (1993) Science Assessment Wade Britton, MPH Health Effects Division Office of Pesticide Programs

35. Study Information • Agricultural postapplication workers monitored during pruning and picking activities in California citrus • Chlorpyrifos (Lorsban 4E) applied once at each of 3 study locations (5-6 lb ai/acre) • Study conducted between 1991/1992

36. Sampling Strategy • 15 individuals monitored • Actual workers and typical durations • Picking • 5 individuals (5 at 1 site) • Exposure occurred 43 days after application • Pruning • 10 individuals (5 at each of 2 sites) • Exposure occurred 2 days after application

37. Multi-faceted Approach • Biological Monitoring* • Urine collected for 4 days after exposure • Blood sampled 1 day after exposure • Pre-exposure samples collected for each • Passive Dosimetry • Dermal • Inhalation • Leaf surface residues

38. Biological Monitoring - Analysis • Blood analyzed for plasma and red blood cell (RBC) cholinesterase (ChE) levels • Urine analyzed for TCP and creatinine • TCP used to calculate chlorpyrifos body burden • Creatinine used to evaluate completeness of sample collection

39. Blood ChE - Activity

40. Urine Measurements • TCP corrected for pre-study levels • Creatinine corrected based upon literature standard (1.8 g/24 hours)

41. Study Strengths • Monitored both urine and blood (plasma and RBC ChE) in all workers • Actual workers monitored while performing activities in production fields

42. Study Limitations • Not statistically designed to define the relationship between TCP and ChE • TCP exposure can occur from many sources • Dosimetry possibly limited absorption of chlorpyrifos • Potential to underestimate TCP & blood ChE activity

43. Conclusions • Represents the best source for occupational worker chlorpyrifos biological monitoring • Provides urine measures and blood plasma and RBC ChE in the same individuals • Actual workers, activities and duration

44. Ethics Assessments of Three Pre-Rule Studies of Chlorpyrifos John M. Carley Human Research Ethics Review Officer Office of Pesticide Programs 44

45. Nolan, et al. (1982) Nolan, R.; Rick, D.; Freshour, N.; and Saunders, J. (1982) Chlorpyrifos: Pharmacokinetics in Human Volunteers Following Single Oral and Dermal Doses. Unpublished study prepared by the Dow Chemical Company under Protocol HEB-DR-0043-4946-4. 28 p. (MRID 124144) Dow AgroSciences (2009) Supplemental Documentation of Ethical Conduct of Nolan et al. Study. E-mail correspondence April 29 through May 8, 2009, between Kenneth Racke and Tom Myers, with attachments. 22 p. 45

46. Value to Society • Defines absorption, distribution, and elimination of oral and dermal doses of chlorpyrifos • Contributes to weight of evidence linking animal data and human epidemiological data 46

47. Subject Selection • Subject Selection • Subjects were all salaried Dow employees, recruited through in-house advertisements • 6 healthy adult males, screened by a physician not otherwise involved in the research • Women of child-bearing age excluded by IRB • Nature of endpoints and measures ruled out subject bias in reporting 47

48. Risks and Benefits • Risks • Doses based on earlier studies and pilot pre-test, with adequate margins of safety • Expected effects • Inhibition of plasma ChE but not of RBC ChE • No clinical signs • Effects reversible—followed until full return to baseline • Benefits • No benefits to subjects 48

49. Ethics Oversight • Approved by Dow Human Health Research Review Committee • Approved by University of Michigan Committee to Review Grants for Clinical Research and Investigation Involving Human Beings • Approvals documented; gaps typical for research from this period 49

50. Informed Consent • Subjects were given a copy of protocol to review • Subjects were briefed on • Study objectives • Chlorpyrifos properties • Pilot phase results, and study procedures • Benefits, including free meals • Confidential handling of data • Voluntary participation and freedom to withdraw • Subjects signed consent forms reporting that they’d read the protocol and been briefed on the research • Subjects were not paid 50