1. Evolving Safety Issues Associated With Erythropoietin Products Oncology Drugs Advisory CommitteeMay 4, 2004 Harvey Luksenburg, M.D. Division of Therapeutic Biological Oncology Products,ODE 6/OND/CDER

2. FDA Team Harvey Luksenburg, M.D., Clinical Kaushik Shastri, M.D., Clinical Ellis Unger, M.D., Clinical Andrea Weir, Ph.D., Pharm/Tox Clare Gnecco, Ph.D., Statistics Mark Rothmann, Ph.D., Statistics Ruth Wager, Ph.D., Product Emily Shacter, Ph.D., Product

3. Background • Two large randomized studies in cancer patients on chemotherapy ± EPO have shown • Shorter OS • Shorter PFS •  incidence of thrombotic/cardiovascular events in the groups assigned to EPO

4. Background • The EPO products used in these two studies are not licensed in U.S. • Used a treatment strategy to achieve hemoglobin  12 gm/dL [higher than recommended in the labeling for U.S. licensed products]

5. Background • Clinical trials for U.S. licensed EPO products were not designed to assess impact on • Response rate (except N93-004) • TTP/PFS • OS

6. Goals (1) • Safety issues observed with EPREX and NeoRecormon may also apply to U.S. licensed products • Review results of trials with EPREX and NeoRecormon (non-U.S. licensed) regarding safety concerns

7. Goals (2) • Review data available regarding safety from trials of EPOGEN / Procrit and Aranesp (U.S. licensed) • To come to agreement on design of future studies regarding safety

8. Safety Issues With Recombinant Erythropoietins • Thrombotic and cardiovascular adverse events • Tumor progression • Survival

9. Recombinant EPO Products (U.S. Licensed) • Epoetin alfa manufactured by Amgen • EPOGEN (marketed by Amgen Inc.) • Procrit (marketed by Ortho BioTech L.P.) • Darbepoetin alfa (Aranesp; Amgen, Inc.)

10. Recombinant EPO Products(Not Licensed in U.S.) • Epoetin alfa (EPREX, Ortho Biologics) • Epoetin beta (NeoRecormon; Hoffmann-La Roche)

11. EPOs as a Class • FDA considers all of these products members of the same product class • These evolving safety issues are assumed to apply to all products unless adequate and well-controlled trials demonstrate otherwise

12. Differences • Epoetin alfa and beta have same amino acid sequence but differ in glycosylation • Aranesp differs in amino acid sequence (5) and degree of glycosylation

13. Similarities • All exert principal clinical effect by binding to the EPO receptor • All have similar pharmacodynamic effects when used at recommended dosages • Similar toxicity profile across products (except PRCA)

14. Target Hemoglobin: Background • Labels for EPOGEN/Procrit and Aranesp have dosage guidelines based on safety data from registration studies performed in patients with chronic renal failure (CRF)

15. Recommended Hgb/Hct • EPOGEN/Procrit • “Suggested Target Hct. Range: 30% to 36%” • Aranesp • “The dose should be adjusted for each patient to achieve and maintain a target hemoglobin not to exceed 12 g/dL”

16. Dosing Recommendations • For rapid increase in Hgb greater than 1.0 g/dL (or 4 points in Hct) in any 2 week period dose should be reduced • Hold if Hgb > 13 g/dL, until Hgb falls < 12 g/dL, and restart dose at 25% belowprevious dose.

17. Thrombotic and Cardiovascular Adverse Events

19. Thrombotic Events with EPO • Licensing studies for EPOGEN / Procrit and Aranesp demonstrated baseline risk of thrombotic & cardiovascular adverse events at labeled target Hgb (10-12 g/dL)

20. “Normal Hematocrit Study” (Besarab et al, NEJM 1998) • 1,233 patients with chronic renal failure on dialysis with clinical evidence of CHF or ischemic heart disease • EPOGEN at baseline in order to achieve/maintain Hct 27-33%

21. “Normal Hematocrit Study” • Both arms received EPOGEN but randomized to different treatment strategies • Achieve Hct of 42% (±3) (“normal Hct”) • Maintain Hct of 30% (±3) (“low Hct”)

22. “Normal Hematocrit Study”

23. Probability of Death/Non-Fatal MI

25. Breast Cancer Erythropoietin Trial (BEST) (EPREX [Epoetin alfa]) • Randomized, DB, Placebo controlled • 939 women w/ metastatic breast cancer • 1st line therapy • +/- EPREX for 12 mos • Therapy started Hgb < 13 g/dL

26. BEST Trial • Primary objective to demonstrate superior survival at 12 months. • Target Hgb 12-14 g/dL • Stopped by IDMC based on first 4 mos safety data

27. BEST Trial At 4 months •  incidence of fatal thrombotic and cardiovascular events EPREX 2.3% Placebo 0.4%

28. Henke et al (Lancet 2003) (NeoRecormon [Epoetin beta]) • Randomized, DB, PC • 351 patients w/ head and neck cancer • Concurrent radiotherapy • Anemic < 12 g/dL women < 13 g/dL men

29. Henke et al • Primary Objective To demonstrate superior locoregional PFS • Target Hgb ≤14 g/dL women ≤15 g/dL men

30. Henke et al • Hypertension, hemorrhage, venous thrombosis, pulmonary embolism, CVA Epoetin beta 11% Placebo 5% • Patients who died of “cardiac disorders” Epoetin beta 5.5% Placebo 3%

32. Registration Studies-Procrit • Pooled analyses of six multicenter, randomized, DB, placebo-controlled studies • 131 patients – Various primary cancers • Platinum (n=59) & non-platinum (n=72) • Anemic • Primary endpoint: Proportion patients tranfused

33. Registration Studies-Procrit • Incidence of thrombotic & cardiovascular events (pooled data) • Placebo 12% (8/68) • Procrit 3% (2/63)

34. Study N93-004 (Procrit [Epoetin alfa]) • Post-marketing commitment • Randomized, DB, PC, non-inferiority study • Intended to enroll 400 patients with SCLC • First line therapy • Baseline Hgb < 14 g/dL

35. Study N93-004 • Primary Endpoint: Objective RR after 3 cycles chemo to rule out decrement of 15% in ORR with Procrit • No target Hgb, Procrit dose reduced for Hgb > 16 g/dL • Study terminated for poor accrual at 224 patients

36. Study N93-004 Incidence of Thrombotic Vascular Events • Procrit 22% (24/109) • Placebo 23% (27/115) • Includes “chest pain”, MI, angina, stroke, cardiac arrest, peripheral ischemia, cardiovascular disorder, thrombosis, DVT, PE, phlebitis

37. Thrombotic / Vascular Events

38. Aranesp 980297 • Randomized, DB, PC • 320 pts with lung cancer • Platinum-containing chemotherapy • Anemic (< 11 g/dL)

39. Aranesp 980297 • Primary endpoint: Proportion transfused between week 5 and week 12 or end of treatment period (EOTP) • Aranesp held for Hgb > 14 g/dL women & Hgb > 15 g/dL men

40. Aranesp 980297 Thrombotic events • Aranesp 5% (7/155) • Placebo 3% (5/159)

41. Thrombotic/Vascular Events Studies w/ Signal

42. Thrombotic/Vascular Events Studies w/o Signal

43. Higher Target Hemoglobin • In September 2003 three placebo controlled clinical trials in oncology patients in which one arm received EPO to target a higher hemoglobin were terminated because of unexpected rates of thrombotic events in the EPO arm

44. Higher Target Hemoglobin

45. Tumor progression

46. Preclinical Studies • EPO receptors (EPO-R) are present on some tumor cell lines & tumor vasculature (endothelial cells)

47. Preclinical Studies EPO is reported to • Inhibit apoptosis • Stimulate angiogenesis; endothelial cell growth, migration, and proliferation • Reduce survival in tumor xenograft models

48. Tumor Stimulation • Studies supporting approval of Procrit and Aranesp for the treatment of anemia in cancer patients on chemotherapy not designed to assess impact on • Tumor response • Tumor progression • Survival

50. Breast Cancer Erythropoietin Trial (BEST) (EPREX [Epoetin alfa]) • Randomized, DB, Placebo controlled • 939 women w/ metastatic breast cancer • 1st line therapy • +/- 12 mo EPREX • Therapy started Hgb < 13 g/dL