1. 1 Backup Slides

2. 2 Physiology of Anemia

3. 3 J&J Breast Cancer (EPO-ANE-3010) Design Features –Excludes patients who ever had hormonal Rx or who are on anticoagulants –Tumor Assessments and Follow Up are adequate –Routine Assessment for TVEs included –Dosing of epoetin alfa=Procrit label –Stratified Randomization by Prior adjuvant anthracycline Her-2 Neu status DFS Interval between initial diagnosis and metastatic disease

4. 4 BEST (Breast) N93-004 (J&J PMC; SCLC) Henke (Head/Neck) ODAC 2004 Supportive Safety Studies from J&J EPO-ANE-3010 (J&J PMC; Breast) PMC Studies from Amgen 2001-0145 (SCLC) PREPARE (Breast) ARA-03 (Breast) DAHANCA (H/N) GELA (NHL) GBR-7 (H/N) RTOG 9903 (H/N) GER-22 (NSCLC) CAN-20 (NSCLC) CAN-17 (Breast) AGO (Cervical)

5. 5 AGO Neoadjuvant Breast (PREPARE) Randomized, Open Label, Multifactorial 1° and 2° Endpoints: –Q 21 vs Q 14 Dose Intense chemo Influence of Aranesp vs Supportive Care on 1° and 2° endpoints is also a 2° endpoint E=epirubicin, C=cyclophosphamide, T=paclitaxel, M=MTX, F=5FU Breast Cancer >2 cm N=720 EC→T q 21 Dose Dense + Intense E→T→CMF Aranesp Transfusion Support SURGERYSURGERY 1° Objectives RFS, OS 2° Objectives pCR, LN status, in breast recurrence, remission rate Target Hgb 12-13 Study DE2001-0033

6. 6 AGO Neoadjuvant Breast (PREPARE) Trial presented at ODAC 2004 Accrual June 2002 - March 2005 Results not available Limitations –Primary Data not provided to FDA –Open Label –Multifactorial –M0 status confirmed by CXR, Abd US, Bone Scan –Inadequate follow up surveillance for recurrence –Off-label Aranesp dose + dose adjustments –No routine assessment of TVEs Study DE2001-0033

7. 7 WSG Adjuvant Breast (ARA-03) Breast Cancer T1-3 and >3 + LN N=690 of planned 1234 TAC or CEF x 6 Aranesp TAC or CEF x 6 Transfusion Support 1° Endpoint: EFS Target Hgb >14 Randomized, Open Label Trial presented at ODAC 2004 Accrual began January 2004; currently ongoing Study DE 2002-0015 RT+ Hormones 2° Endpoints: OS, Local Relapse TAC=Docetaxel, Doxirubicin, Cyclophosphamide CEF=Cyclophosphamide, Epirubicin, 5-FU

8. 8 WSG Adjuvant Breast (ARA-03) Limitations –Heterogeneous chemotherapy regimens; no attempt to stratify –Open Label –M0 status confirmed by CXR, Abd US, Bone Scan –Inadequate follow up surveillance for recurrence –Off-label Aranesp dose + dose adjustments –No routine assessment of TVEs Study DE 2002-0015

9. 9 GELA DLBCL ( LNH03-6B ) Randomized, Open Label, Multifactorial 1° Endpoint: –Q 21 vs Q 14 Dose Intense chemo Control arm could receive Aranesp DLBCL N=328 of planned 600 R-CHOP Q 14 x 8 + MTX IT x 4 Aranesp Transfusion or Aranesp if Hgb≤9 1° Objective EFS (Q 14 vs Q 21) 2° Objectives RR, DFS, OS, Progression rate, Relapse Rate Target Hgb 13-15 Study FR 2003-2005 R-CHOP Q 21 x 8 + MTX IT x 4

10. 10 GELA DLBCL ( LNH03-6B ) Accrual began December 2003; ongoing Interim results on 130 patients-1 yr OS, 1 yr EFS no difference observed Design Problems –Control arm could receive Aranesp –Multifactorial –Open Label –Response Rate and DFS not clearly defined –Criteria for tumor response or progression not provided –Off-label Aranesp dose + dose adjustments –No routine assessment of TVEs Study FR 2003-2005

11. 11 BEST (Breast) N93-004 (J&J PMC; SCLC) Henke (Head/Neck) ODAC 2004 Supportive Safety Studies from J&J EPO-ANE-3010 (J&J PMC; Breast) PMC Studies from Amgen 2001-0145 (SCLC) PREPARE (Breast) ARA-03 (Breast) DAHANCA (H/N) GELA (NHL) GBR-7 (H/N) RTOG 9903 (H/N) GER-22 (NSCLC) CAN-20 (NSCLC) CAN-17 (Breast) AGO (Cervical)

12. 12 Common Limitations Excessively high Target Hgb Off Label Epoetin alfa dosage and dose adjustments Inadequate radiological assessments to assess for recurrence No routine assessment for TVEs Open Label Primary data not submitted to FDA on trials that have finished enrollment

13. 13 EPO-GBR-7 Head/Neck Head/Neck Cancer (Stage II/III) N=301 Definitive RT Eprex Definitive RT Transfusion Support 1° Endpoint: 2 yr Local DFS 2° Endpoint: 1, 2, 5 yr OS Randomized, Open Label Trial presented at ODAC 2004 Accrual August 1999 – April 2002 Terminated early due to slow accrual (goal = 800 pts) Target Hgb 14.5-15

14. 14 EPO-GBR-7 Head/Neck N=301 EprexControl 12 wk ORR99% Local Relapse in RT field 25%29% Local Relapse outside RT field 13%15% 1 yr OS77.3%79.9%95 % CI (-12.1% to 6.8%), p=0.867 TVE3%1% Result summary from J&J, April 2006 No significant differences for local recurrence in or outside the RT field, 1 yr OS, RR

15. 15 EPO-GBR-7 Head/Neck Limitations –Primary Data not given to FDA –No result for Primary Endpoint, 2 yr local DFS –Assessment method and the frequency of testing for local recurrence was inadequate –Suspected recurrences did not need biopsy –Assessment of Survival was not required for subjects withdrawn from study –Response judged from clinical assessments only –Dose and dose adjustments off-label –Target Hgb excessively high –No routine TVE assessments

16. 16 RTOG 99-03 Head/Neck Head/Neck Cancer (Stage I-IV) N=148 Definitive RT or chemoRT Procrit Definitive RT or chemoRT Transfusion Support 1° Endpoint: 2 yr LRF 2° Endpoint: OS Randomized, Open Label Trial presented at ODAC 2004 Accrual June 2000-October 2003 Terminated early due to DMC trend to lower LRC and OS in Procrit arm (goal = 372 pts) Target Hgb 13.5- 16 M, 12.5-14 F LRF-loco regional failure; LRC-loco regional control

17. 17 RTOG 99-03 Head/Neck Limitations –Primary Data not given to FDA –No result for Primary Endpoint, 2 yr LRF –Dose adjustments off-label –Frequency of testing for local recurrence inadequate –Target Hgb excessively high –No routine TVE assessments Result summary (abstract 2004):

18. 18 EPO-GER-22 NSCLC NSCLC (Stage IIIA/IIIB) N=389 Weekly chemo→RT Eprex Weekly chemo→RT Transfusion Support 1° Endpoint: 2 yr OS 2° Endpoint: Remission Rate, Local Control Randomized, Open Label Trial presented at ODAC 2004 Accrual August 2001-December 2005 Terminated early due to slow accrual (Target = 612 patients) Target Hgb: 12-13

19. 19 EPO-GER-22 NSCLC EprexControlBased on ORR55%47%177/389 pts Median Survival (days) 338+/-49 (95% CI 242-434) 299+/-33 (95% CI 234-364) P=0.61215/389 pts TVE17.7%8.5%P=0.097230/389 pts Limitations –Primary Data not given to FDA –No result for Primary Endpoint, 2 yr OS –Dose and dose adjustments off-label –Target Hgb excessively high –No routine TVE assessments Result summary from J&J, April 2006

20. 20 EPO-CAN-17 Breast Breast Ca (Stage I-IV) N=354 Chemo Eprex Chemo Transfusion Support 1° Endpoint: QOL 2° Endpoint: RR, OS Randomized, Open Label Trial presented at ODAC 2004 Accrual February 2002 – May 2003 Target Hgb: 12-14

21. 21 EPO-CAN-17 Breast N=354 EprexControl ORR (74 pts w/Stg IV dz)36%30%p=0.586 # deaths at 24 months2728P=0.82 Deep Thrombophlebitis6.3%0.6% TVEs (FDA Analysis)7.1%3.6% Limitations Primary Tumor Outcome or Survival data not given to FDA Primary Endpoint: QOL Open Label Systematic tumor restaging in Stg IV dz not required post treatment Length of f/u inadequate Dose adjustments off-label Target Hgb excessively high No routine TVE assessments Results (J&J summary 4/06)

22. 22 AGO/NOGGO Cervical Cervical Ca (Stage Ib-IIb) N=264 Chemo→RT Eprex Chemo→RT Transfusion Support 1° Endpoint: 5 yr RFS 2° Endpoint: OS, TTF Randomized, Open Label Trial presented at ODAC 2004 Accrual January 1999 – March 2001 Target Hgb: 13 SURGERYSURGERY RFS: Relapse Free Survival TTF: Time to Treatment Failure

23. 23AGO/NOGGOLimitations –Primary Data not given to FDA –Open Label –Dose adjustments off-label –Target Hgb excessively high –No routine TVE assessments ESAControl Relapse Rates @ median 105 wks17%25%P=0.074 2 yr RFS70%81%P=0.058 1 yr RFS86%91% Results (J&J summary 4/06) 5 year RFS (primary endpoint), OS, TTF: not reported

24. 24 BEST (Breast) N93-004 (J&J PMC; SCLC) Henke (Head/Neck) ODAC 2004 Supportive Safety Studies from J&J EPO-ANE-3010 (J&J PMC; Breast) PMC Studies from Amgen 2001-0145 (SCLC) PREPARE (Breast) ARA-03 (Breast) DAHANCA (H/N) GELA (NHL) GBR-7 (H/N) RTOG 9903 (H/N) GER-22 (NSCLC) CAN-20 (NSCLC) CAN-17 (Breast) AGO (Cervical) Other Studies Anemia of Cancer Lymphoid Malignancy Non-Myeloid Malignancy Moebus

25. 25 Non-Myeloid Malignancy (2003-0232) Non-Myeloid Ca N=391 Chemo Aranesp Chemo Placebo 1° Endpoint: % transfusion Randomized, Double-Blind, Placebo Controlled Stratified by Tumor Type Accrual February 2004 – October 2004 Primary Data submitted March 2007 Target Hgb: 12-13

26. 26 Non-Myeloid Malignancy (2003-0232) Results (FDA review of primary data) –No significant difference in OS ( HR 0.82 [95% CI 0.43, 1.57]). Limitations –Primary endpoint: % transfusions –Data collection not adequate –No long term follow-up plan –Dosing was off-label in dose (300 µg Q3W) and adjustments –Heterogeneous cancer types –Target Hgb excessively high –No routine TVE assessments

27. 27 BRAVE (Breast) Breast Ca (Stage IV) N=463 Chemo Epoetin beta Chemo Transfusion Support 1° Endpoint: OS 2° Endpoint: PFS Randomized, Open Label Stratified by chemo type and hormonal status Accrual November 2002 – June 2004 Target Hgb: 13-15

28. 28 BRAVE (Breast) Results (summary results) –OS no significant difference (HR 1.07, 95% CI 0.87, 1.33; p=0.522) –PFS no significant difference (HR 1.07, 95% CI 0.89, 1.30; p=0.448) –TVE higher in ESA arm (13% v 6%, RR 2.36, 95% CI 1,23, 4.55; p=0.01) Limitations –Primary data not supplied to FDA –Protocol not submitted to FDA –Target Hgb 13-15 –No routine TVE assessments

29. 29 Moebus (Breast) Multifactorial Accrual January 1999 – March 2001 Breast Ca (Adjuvant Node +) N=1284 Dose Dense/Intense ETC + GCSF Eprex Transfusion Support 1° Objective: EFS (Q14 v Q21) ∆Hgb 2° Objectives: 5 yr OS and DFS (Q14 v Q21), Intramammary recurrence (ESA v control) Target Hgb 12.5-13 Study Standard EC->T

30. 30 Moebus (Breast) Results –5 yr OS and DFS: no significant difference –↑ TVE in ESA arm (3.0% vs 1.7%) Limitations –Primary data not submitted to FDA –Inadequate study endpoints –Multifactorial –Inadequate frequency of radiographic assessments –Dosing adjustments off label –No routine TVE assessments

31. 31 TVE Risks There are increased risks of TVE with the use of ESAs Evidence: –EPO-CAN-15 (SCLC) –PR00-03-006 (Gastric/Rectal cancer) –GOG 191 (Cervical cancer) –Rosenzweig (Breast) –Bohlius meta-analysis JNCI April 2006 –SPINE Study With the exception of EPO-ANE-3010, tumor promotion studies discussed do not routinely assess patients for TVE, & did not specify TVE as prespecified endpoint. –Safety signals for increased risk may be missed

32. 32 EPO-CAN-15 (SCLC) SCLC (Limited Stg) N=106 Chemo Epoetin Chemo Placebo 1° Endpoint: PFS 2° Endpoints: RR, OS, median survival, LC Terminated early due to TVEs TVE OR 7.74 (ESA vs Control) ↑ Mortality (40% vs 19%) in ESA arm Target Hgb: 14- 16

33. 33 PR 00-03-006 (Gastric/Rectal) Gastric/Rectal Ca N=59 ChemoRT→surgery Epoetin ChemoRT →surgery Placebo 1° Endpoint: % transfusion 2° Endpoints: RR, pCR Terminated early due to TVEs TVE OR 3.79 (ESA vs Control) Target Hgb: 14- 15

34. 34 GOG 191 (Cervical) Cervical Ca N=113 ChemoRT Epoetin ChemoRT Transfusion Support 1° Endpoint: PFS 2° Endpoints: OS, LC Terminated early due to TVEs TVE OR 2.65 (ESA vs Control) Target Hgb: 13- 14

35. 35 Rosenzweig (Breast) Breast Ca (Stg 4) N=27 Unclear Epoetin Unclear Transfusion Support 1° Endpoint: Unclear 2° Endpoints: Unclear Terminated early due to TVEs TVE 29% vs 0% (ESA vs Control) Target Hgb: Unclear

36. 36 JNCI Meta Analysis 2006: TVE

37. 37 JNCI Meta Analysis 2006: OS

38. 38 JNCI Meta Analysis 2005 OS: HR 0.81 (95% CI: 0.67 to 0.99) 2006 OS: HR 1.08 (95% CI: 0.99 to 1.18)

39. 39 Final Data/Study Report not received 2001 2002 2003 2004 2005 2006 2007 2008 2009 *=Data never submitted to FDA †=Received limited safety data (no efficacy data) ‡=Received primary data GBR-7 RTOG 9903* GER-22 † CAN-17 † AGO* ‡ 232 ‡ PREPARE ARA-03 GELA 2010 2011 2012 =Last pt accrued =Final Data/Study Report Anticipated

40. 40 Final Data/Study Report not received StudyDate Last Pt accrued Final Data or Study Report Anticipated GBR-7 (H/N) April 20024Q 2007-1Q 2008 RTOG 99-03 (H/N) Nov 20032004* GER-22 (NSCLC) Dec 20051Q-2Q 2008 CAN-17 (Breast) May 20032007 † AGO (Cervical) Mar 20012006* 232 (Non-myeloid) Oct 20042007 ‡ PREPARE (Breast) Mar 2005Q4 2007 ARA-03 (Breast) OngoingQ2 2011 GELA (DLBCL) OngoingQ3 2010 *=Data never submitted to FDA †=Received limited safety data (no efficacy data) ‡=Received primary data

41. 41 Non Inferiority Trials Purpose is to rule out a prespecified margin of inferiority between ESA and control In trials with ESAs in cancer pts specifically designed to look for safety signals, non- inferiority design preferable over superiority design Hazard Ratio 1.0 Worse for Control Worse for ESA Prespecified Margin Non-inferior Inferior

42. 42 N93-004 (PMC; SCLC) ODAC 2004 Epoetin alfa Studies EPO-ANE-3010 (J&J PMC; Breast) Aranesp Studies Studies Initiated by Amgen or J&J Studies Initiated by Amgen or J&J Other Studies Anemia of Cancer (103) Lymphoid Ca (161) Non-Myeloid Ca (232) 2001-0145 (SCLC) GBR-7 (H/N) GER-22 (NSCLC) CAN-17 (Breast)

43. 43 Completed studies w/summary results + no reported safety signal 5 epoetin alfa studies have been completed, and have had summary results submitted to FDA either by a J&J report or by literature publication, and have not reported safety signals Both FDA and J&J have not analyzed data from these trials Study Limitations –3 of the 5 studies have inadequate design to address risk of tumor promotion

44. 44 ODAC 2004 Epoetin alfa Studies Aranesp Studies Completed studies w/summary results + no reported safety signal Completed studies w/summary results + no reported safety signal Other Studies GBR-7 (H/N) RTOG 9903 (H/N) GER-22 (NSCLC) CAN-17 (Breast) AGO (Cervical)

45. 45 Hazards Ratios for Overall Survival by Studies

46. 46 Meta-Analysis results Amgen meta analysis used Peto odds ratio, which: –Is based solely on the # of known deaths and total # of pts in each arm –Does not consider pt follow-up and survival times –Is not interpretable since study patients do not have the same study start day