1. Introduction Lawrence J. Lesko, Ph.D., FCP Director of the Office of Clinical Pharmacology and Biopharmaceutics Center for Drug Evaluation and Research Food and Drug Administration Clinical Pharmacology Subcommittee (CPSC) of the Advisory Committee for Pharmaceutical Sciences November 14-15, 2005 Rockville, Maryland

2. CPSC Meeting History November 2-3, 2002 April 22-23, 2003 November 17-18, 2003 November 3-4, 2004 November 14-15, 2005

3. Recent CPSC Topics I Quantitative methods –M/S to optimize dosing adjustments and reduce risk in patient subgroups Pharmacogenomics –Label revisions of thiopurines and irinotecan to include genomic data for guiding dosing Evaluation of drug interactions –Labeling and evaluation of enzyme and transporter mechanisms for guidance revision

4. Recent CPSC Topics II Critical path initiatives –End-of-phase 2A (EOP2A) meetings –Framework for biomarker evaluation “Opportunity: ….. Biomarkers to target responders, monitor clinical response and measures of drug effectiveness.

5. This Meeting 1.Pharmacogenomic Data in Product Labels - best way to include PGx data in product labels - evidence for including PGx data in warfarin label 2.Model-Based Drug Development - recap experience with EOP2A meetings - stratification issue using clinical trial simulation 3.Biomarkers and Individualization - update on critical path initiatives - including biomarker data in product labels

6. Drug Labeling: The Legal Basis of Prescribing “If evidence is available to support the safety and effectiveness of the drug only in selected subgroups of the larger population with a disease, the labeling shall describe the evidence and identify specific tests needed for selection and monitoring of patients who need the drug.” - 21 CFR 201.57

7. Label Revisions Are Common Labels among most frequently consulted information sources –Label updates one of the main tools for informing physicians and patients about new risks Original version reflects pre-approval data –Efficacy documented; safety provisional New insights post-approval alter B/R and drive regular label revisions –Particularly important for individualizing therapy Martin-Facklam, Eur J Clin Pharmacol 2004

8. Label Revisions Have Limitations While physicians wish for precise management advice, e.g., specific dose adjustments, evidence may sometimes be descriptive and actions general, e.g., reduce the dose, titrate carefully or monitor more closely Lack of perfect evidence (e.g., specific dose reductions) is not a reason to support inaction

9. Irinotecan: November 3, 2004 Irinotecan R (camptosar) ~ proven 1 st (5-FU and leucovorin) and 2 nd line therapy for metastatic colon/rectal cancer Providers/patients face a clinical predicament ~ what is the optimal dose –Incidence of grade 3-4 neutropenia is 35% –Nearly 70% of patients need dose reduction –Toxicity associated with SN-38 exposure “…causes severe myelosuppression…” “...death due to sepsis following myelosuppression…” “...adjust doses based on neutrophil count…”

10. Problem: Accumulation of SN-38 Exposure dependent on metabolism of SN-38 by UGT1A1 –Wide interpatient variability in UGT1A1 activity –Patients with *28 variant (7 TA repeats) have reduced enzyme activity –Homozygous deficient (7/7 genotype) patients have the greatest risk of neutropenia –Neutropenia matters to patients Original label was silent on UGT information; approved dose not optimized

11. Risk Assessment by Genotype Would an adjunct UGT diagnostic test to identify patients who are 7/7 genotype lead to lower risk of neutropenia vs SOC? Patient GroupPrevalenceRisk of Neutropenia All Patients No Test -----10 in 100 Wild-type 6/6 Genotype 50%0 in 100 Heterozygous-deficient 6/7 Genotype 40%12 in 100 Homozygous-deficient 7/7 Genotype 10%50 in 100 From Innocenti et al in Clin Pharmacol Ther (2004)

12. Camptosar Label Revised and FDA Approved UGT Test

13. Optimizing Warfarin Benefit/Risk with CYP 2C9 Genotypes There has been over 20 label revisions for warfarin since 1954. The most recent revision in September 2005 had to due with interactions with cranberry juice and proton pump inhibitors.

14. Success and Failure of Drug Therapy: Inborn Predisposition or Susceptibility “By nature, men are nearly alike; by practice, they get to be wide apart.” Confucius, Analects Chinese Philosopher 551 BC – 479 BC

15. Warfarin Discovered 60 years ago and one of the most widely prescribed drugs in the world Intended to prevent and treat thromboembolisms –Afib, recurrent stroke, DVT, pulmonary embolism, heart valve prosthesis Multi-source anticoagulant –1, 2, 2.5, 3, 4, 5, 6, 7.5 and 10 mg tablet strengths Significant increase in Rx’s over past 10 years especially in the elderly

16. Trends in Warfarin Use: 1.5-fold Increase (45%) in Last 6 Years

17. Efficacy of Warfarin Prospective clinical trials unequivocally demonstrate effectiveness Mortality risk in untreated patients with AFib is 2.5X greater than in warfarin-treated patients Risk of ischemic stroke in patients with AFib treated with warfarin is reduced by 65% NNT (vs placebo) to prevent one stroke ~ 32 Linkins et al, Ann Intern Med 139:893-900, 2003 Schulman, N Engl J Med 349:675-683, 2003 Eikelboom, Med J Australia 180:549-551, 2004

18. Global Problem of Warfarin AEs ~ 2 million people in the US receiving warfarin; near the top in most surveys of AEs ~ 70,000 patients in Sweden receiving warfarin; it tops the list of drug-induced AEs ~ 600,000 patients in UK receiving warfarin; 6% of patients over 80 years; 10-24 episodes of hemorrhage per 100 patients ~ Account for 3.6% of all drug-induced AEs (4 th ranked drug) but 15.1% of all severe AEs (2 nd to digoxin) over 10 year period Evans, Annals of Pharmaco 39:1161-1168, 2005 Wadelius, The Pharmacogenomics J, 5:262-270, 2005 Pirmohamed (Personal Communication)

19. Safety of Warfarin Major risk is bleeding: frequent and severe 1.2 – 7 major bleeding episodes per 100 patients Responsible for 1 in 10 hospital admissions Relative risk of fatal extracranial bleeds 0 - 4.8% NNH for major bleed ~ 333 Schulman, N Engl J Med 349:675-683, 2003 Pirmohamed, British Med J 329:15-19, 2004 DaSilva, Seminars Vasc Surg 15:256-267, 2002 Eikelboom, Med J Australia 180:549-551, 2004

20. Dosing of Warfarin is Complex Narrow therapeutic index –Small separation between dose-response curves for preventing emboli and excess coagulation Nonlinear dose-response (INR) –Small changes in dose may cause large changes in INR with a time lag Wide range (50x) of doses (2-112 mg/week) to achieve target INR of 2-3 –Patient intrinsic and extrinsic factors

21. PK of Warfarin: Mechanistic Basis of Dosing Problem Large interindividual variability related to S- warfarin metabolism by CYP2C9 (genetics) –*1 (wild type), *2 and *3 (variant alleles) Genotype (N = 188) Prevalence% Enzyme Activity S/R Warfarin (mg/L) Weekly Doses (mg) Clearance/ LBW (ml/min/kg) 2C9 *1/*163%100%0.45 (0.11) 34.1 (19.5) 0.065 (0.025) 2C9 *1/*X31%50-70%0.69 (0.28) 19.0 (10.8) 0.041 (0.021) 2C9 *X/*X6%10%1.43 (0.63) 11.5 (7.2) 0.020 (0.011) Herman et al, The Pharmacogenomics J 4:1-10. 2005

22. Dosing Adjustments Based on Genotype- Specific S-Warfarin Clearance Stefanovic and Samardzija, Clin Chem & Lab Med, 42(1) 2004

23. PD of Warfarin: Mechanistic Basis of Variability in Response INR: measure of intensity of anticoagulation –Dose-plasma levels-INR Plasma warfarin was a strong predictor of changes in INR measurements (p < 0.0001) Accounted for 15.3% of variance in effects of warfarin Wide interindividual variability with stronger correlations at higher INR values Response to given INR is also variable Difficulty in achieving target INR and frequency of AEs shows the limitations of INR White, Clin Pharmacol Ther 58:588-93, 1995

24. Benefit: INR and Stroke Prevention Hirsch, J Amer Coll Cardio 41:1633-1652, 2003

25. Risk: INR and Intracranial Hemorrhage

26. Unequivocal Association Between 2C9 Alleles and Warfarin-Induced Bleeding Higashi, JAMA 287:1690-1698, 2002 Margaglione, Thromb Haemost 84, 775-778, 2000 Ogg, The Lancet 354:1124, 1999 Sanderson, Genetics in Medicine, 7:97-104, 2005

27. Quality of Anticoagulation is Generally Poor Despite INR Monitoring Mean % time patients (n = 600) spend within target INR range was 62%. More time below (25%) than above (13%). Target INR range (n = 100) was achieved on 44% of time. Sub-therapeutic levels (38%) exceeded supra- therapeutic levels (18%) Only 14% (n = 52) of patients met criteria for quality anticoagulation control (>70% time in target INR range) Davis, Annals of Pharmacotherapy, 39:632-636, 2005 Lin, Europ Soc Cardiology, 7:15-20, 2005 Menzin, Annals of Pharmacotherapy, 39:446-451, 2005 Peterson, J Am Coll Cardiol, 41:1445-1451, 2003

28. Two Phases of Warfarin Dosing Induction Phase: When initiating warfarin treatment to achieve target INR (2-3) - daily, bi-weekly, weekly INR - frequent dose adjustments in response to INR - reach INR target in 4-5 days on average - may take 7 – 30 days to reach steady state Maintenance Phase: When target INR (2-3) is achieved - following the induction phase - monthly INR, relatively stable doses - dose adjustments needed based on changes in co-variates

29. Initial Dosing During Induction Phase Initial dose: estimated maintenance dose (5 mg/day) based on patient co-factors Predictors of higher (> 5 mg/day) doses –Indication, e.g., cardiac replacement valves –Co-morbidities, e.g., diabetes –Age < 55 y –Male gender –African-American ethnicity –Vitamin K intake –Weight –Concomitant drugs, e.g., carbamazepine Absher, Annals of Pharmacotherapy 36:1512-1517, 2002 Hillman, Pharmacogenetics 14:539-547, 2004

30. Individualize dosing based on rise in INR INR Monitoring During Induction Phase Initial doses suppress factor VII with little effect on factors II, IX, X INR may appear to reach stable target in 3-5 days Continued dosing inhibits factors with longer t 1/2 (II, IX, X) resulting in over-shooting target INR INR in first 4 days have a 65% success rate in predicting dose Vitamin K Dependent Clotting Factors

31. Schematic of Warfarin Dosing: One Size Fits Few Initial Dose: 35 mg/week Age Gender BSA Concomitant Drugs Co-morbidities 30-35%20-25% INR 23 Increase DOSE DecreaseRepeat INR: Adjust Dose Stable Maintenance Dose INR 23 29 mg/wk 28 mg/wk 24 mg/wk 18 mg/wk 6 mg/wk

32. Clinical Example: Problem with Initial Anticoagulation Rate and INR Monitoring -Elderly woman in nursing home -Sent to ER with lower GI bleed -Dx with femoral v thrombosis -Started warfarin 5 mg/day -After 7 days, INR was 2.5 -Advised to continue for 12 wks -INR of 66, treated, discharged -4 days later, hospitalized -Unexpected rise in INR ~ 7.5 -No changes in drug, diet -No medication errors -Warfarin half-life ~ 10 days CYP2C9 PGx analysis = heterozygote, two variant alleles, 2C9*2/2C9*3

33. Implications of Difficult Induction Phase for Patients with 2C9 Alleles More frequent changes in daily dose Delayed stabilization and hospital discharge Multiple visits to clinic or hospital Additional investigation to seek solution Increased risk of bleeding Peyvandi, Clin Pharmacol Ther 75:198-203, 2004 Aithal, The Lancet, 353: 717:719, 1999 2C9 *2 and *3 unequivocally risk factors consistently across studies; magnitude of risk increase is variable

34. Incorrect dosing, especially during the induction phase, carries a high risk of either severe bleeding (too high) or failure to prevent thromboembolisms (too low). The majority of warfarin-related AEs occur during the first 30 days of therapy and are preventable with optimal dosing. Risks of Warfarin Are Greatest During Induction Phase Schmelzer (Marshfield Clinic), Report to Agency for Healthcare Research and Quality (AHRQ), 2001

35. Frequency of Major Bleeds Following Initiation of Warfarin Dosing Landefeld, Am J Med 87:144-152, 1989

36. Prospective Genotyping of CYP 2C9: Translation of Data to Practice Would knowledge of a patient’s genotype improve warfarin dosing during the induction phase and reduce the incidence of warfarin-related adverse events, i.e., unintentional bleeding (overdosing) and embolisms (underdosing)? Note: Genotyping is not a replacement for other co-factors but as an additional piece of information

37. Incremental Value: Accounting for Interpatient Variability in Dosing Reference2C9 Alleles AloneAll Other Factors* Herman, 200527%10% Wadelius, 200512%18% Hillman, 200420%27% Relative % of Variability in Dose Explained * Age, body weight/BSA, indication, gender, interacting drugs; VKORC 1 SNPs not included

38. Can Genotyping Data in Label Help Anticoagulation During Induction Phase? Identify high risk patients for AE (e.g., 2C9 *X) at risk prior to or during induction phase –No need to delay initial dosing –More conservative dose increases –More frequent INR measurements –Lower target maintenance dose Identify patients likely to require higher maintenance doses (e.g., 2C9 *1/*1) Identify low risk patients in need of anticoagulation –Select warfarin alternatives, e.g., aspirin Investigations of unexpected toxicity or resistance

39. Acknowledgements Dr. Myong-Jin Kim Dr. Felix Frueh Dr. Shiew-Mei Huang Dr. Atik Rahman

40. Review of Genotype Data in Labels Generally and Evidence Related to Warfarin Specifically Next