1. Hecht ASCO GI 2008 An updated analysis of safety and efficacy of oxaliplatin (Ox)/bevacizumab (bev) +/- panitumumab (pmab) for 1 st ‑ line treatment (tx) of metastatic colorectal cancer (mCRC) from a randomized, controlled trial (PACCE) J. Randolph Hecht, 1 Edith Mitchell, 2 Tarek Chidiac, 3 Carroll Scroggin, 4 Christopher Hagenstad, 5 David Spigel, 6 John Marshall, 7 Allen Cohn, 8 Seta Shahin, 9 Thomas Griffin 9 1 UCLA School of Medicine, Los Angeles, CA; 2 Kimmel Cancer Center of Thomas Jefferson University, Philadelphia, PA; 3 Mid Ohio Oncology/Hematology, Inc, Columbus, OH; 4 NEA Clinic, Jonesboro, AR; 5 Suburban Hematology-Oncology Associates, Lawrenceville, GA; 6 Sarah Cannon Research Institute, Nashville, TN; 7 Georgetown University Hospital, Washington, DC; 8 Rocky Mountain Cancer Centers, Denver, CO; 9 Amgen Inc., Thousand Oaks, CA This study was funded by Amgen Inc.

2. Introduction  Panitumumab, a fully human monoclonal antibody that targets the epidermal growth factor receptor (EGFr), is approved in the US as monotherapy for the treatment of refractory mCRC  Preclinical and preliminary clinical studies have suggested that combining EGFr, VEGF inhibitors, and chemotherapy may improve efficacy over single-agent treatments 1,2  PACCE was a US, community-based study designed to compare the efficacy and safety of bevacizumab and chemotherapy +/- panitumumab for first-line treatment of mCRC  This interim analysis describes updated data available as of May 31, 2007 on safety and efficacy on the Ox-based cohort from PACCE (trial discontinued only panitumumab therapy in March 2007 after a planned interim analysis of ~231 PFS events in the Ox-CT cohort) 1 Shaheen RM et al. Brit J Cancer 2001;85:584–589 2 Saltz LB et al. J Clin Onc 2007;25:4557–4561.

3. PACCE Study Schema Randomized, Open-Label, Controlled Phase 3b Study Stratification Factors: ECOG score, prior adjuvant treatment, disease site, Ox doses/Iri regimen, number of metastatic organs Panitumumab 6 mg/kg Q2W + Ox-CT + Bevacizumab Ox-based CT (eg, FOLFOX) N = 800 Investigator choice Iri-based CT (eg, FOLFIRI) N = 200 Investigator choice Ox-CT + Bevacizumab Panitumumab 6 mg/kg Q2W + Iri-CT + Bevacizumab Iri-CT + Bevacizumab RANDOMIZERANDOMIZE 1:1 SCREENINGSCREENING Tumor Assessments: Q12W until disease progression or intolerability

4. Key Eligibility Criteria  Age  18 years old  Measurable mCRC per modified RECIST  ECOG status 0 or 1  Adequate hematologic, renal, and hepatic function  No prior chemotherapy or biologic therapy for mCRC  No adjuvant chemotherapy within 6 months of randomization  No major surgery within 28 days of randomization or elective and/or planned major surgical procedures during the study  No clinically significant cardiovascular disease within 1 year prior to randomization  No EGFr testing required

5. Investigator Choice – Ox-CT Regimens RegimenOxaliplatinLeucovorin (LV)5-Fluorouracil (5-FU) FOLFOX 485 mg/m 2 IV over 2 hrs on Day 1 per cycle 200 mg/m 2 IV over 2 hrs Days 1 and 2 400 mg/m 2 IV bolus over 2-4 min Day 1, followed by 600 mg/m 2 continuous IV over 22 hrs on Days 1 and 2 FOLFOX 5100 mg/m 2 IV over 2 hrs on Day 1 per cycle 200 mg/m 2 IV over 2hrs Days 1 and 2 400 mg/m 2 IV bolus over 2-4 min Day 1, followed by 600mg/m 2 continuous IV over 22 hrs on Days 1 and 2 FOLFOX 6100 mg/m 2 IV over 2 hrs on Day 1 per cycle 400 mg/m 2 IV over 2 hrs Day 1 400mg/m 2 IV bolus over 2-4 min Day 1, followed by 2400- 3000mg/m 2 continuous IV over 46 hrs Modified FOLFOX 6 85 mg/m 2 IV over 2 hrs on Day 1 per cycle 400 mg/m 2 IV over 2 hrs Day 1 400 mg/m 2 IV bolus over 2-4 min Day 1, followed by 2400- 3000 mg/m 2 continuous IV over 46 hrs FOLFOX 7130 mg/m 2 IV over 2 hrs on Day 1 per cycle 400 mg/m 2 IV over 2 hrs Day 1 2400mg/m 2 continuous IV over 46 hrs Other Ox Q2W TBD by physician

6.  Primary endpoint: –Progression-free survival (PFS) by central review in the Ox-CT stratum a  Secondary endpoints included: –Overall response rate (ORR), time to treatment failure (TTF), overall survival (OS), safety profile  Design Characteristics –To detect a 30% improvement in median PFS in the panitumumab plus bevacizumab/Ox-CT treatment group vs bevacizumab/Ox-CT alone 80% power and  = 0.05 (2-sided) for 462 PFS events (disease progression or death) Planned interim analysis at ~ 231 Ox-CT PFS events Study Endpoints and Design Characteristics a Powered for oxaliplatin stratum only; descriptive for irinotecan stratum

7. Independent DMC Analyses 25 pts Safety 75 pts 150 pts SafetySafety RR Unplanned Safety Efficacy Safety RR Safety Efficacy 500 pts 800 pts ~231 PFS events DMC recommended continuation without protocol modification Mar 2005 Panitumumab dosing was discontinued on Mar 22, 2007 DMC recommended changes to informed consent 1 st pt randomized DMC = data monitoring committee

8. Treatment: Ox-CT Cohort Treatment: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab+ bev/Ox-CT (N = 413) bev/Ox-CT (N = 410) Patients randomized, n (%)413 (100)410 (100) Patients who received any first-line treatment, n (%) 406 (98)398 (97) Median follow-up, weeks Range 53 1-114 55 2-106 Chemotherapy regimen received on week 1, n (%) a FOLFOX497 (24)102 (26) FOLFOX52 ( < 1)1 ( < 1) FOLFOX666 (16) 71 (18) Modified FOLFOX6189 (47) FOLFOX77 ( 2)4 ( 1) bFOL5 ( 1) Other Ox-based Q2W CT b 40 (10)26 (7) a Of those patients who received any first-line treatment; b Allowed per protocol; capecitabine was not allowed

9. Demographics and Baseline Characteristics (Interim Analysis, May 2007 Data Cutoff) pmab + bev / Ox-CT (N = 413) bev / Ox-CT (N = 410) Male, %5658 Race, % White Black Hispanic Other 83 8 6 2 80 10 6 3 Median age, years (range)61 (28–88)62 (22–89) Age  65, % 3943 Baseline ECOG, % 0 1 61 39 58 42 Prior adjuvant therapy for mCRC, % 19 Metastatic organs, % 1 > 1 49 a 50 49 51 a One patient had no metastatic organs

10. Summary of Adverse Events: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab + bev/Ox-CT (N = 407) bev / Ox-CT (N = 397) Any adverse event (AE), %100 Grade 35253 Grade 42819 Grade 5 a 63 Pmab-related grade 51N/A Any serious AE, %6038 Panitumumab treatment-related serious AE, % 19N/A Ended panitumumab treatment due to AE, % 27N/A Based on the safety analysis set (all patients who received ≥ 1 dose of treatment) Graded per NCI CTCAE v 3.0; n/a, not applicable a As reported by investigator – does not include disease progression (ie, neoplasms)

11. Grade 3 or 4 Adverse Events of Interest: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab + bev / Ox-CT (N = 407) bev / Ox-CT (N = 397) Adverse eventGr 3 %Gr 4 %Gr 3 %Gr 4 % Skin toxicity a 35110 Diarrhea222121 Infections b 16282 Dehydration1525< 1 Neutropenia1410177 Nausea1105< 1 Hypokalemia8231 Hypomagnesemia3100 Neuropathy3<170 Paronychia1000 Deep venous thrombosis7080 Pulmonary embolism c 0604 MedDRA v 9.0 preferred terms; graded per NCI CTCAE v 3.0 a Skin toxicity included terms from the skin and subcutaneous and infections system organ class(SOC) b Grade 5 infections occurred in 3 (1%) of pmab + bev/Ox-CT pts and 3 (1%) of bev/Ox-CT pts c Grade 5 pulmonary embolism occurred in 3 (1%) of pmab + bev/Ox-CT pts

12. Deaths (All Causes): Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab + bev/Ox-CT (N = 407) bev / Ox-CT (N = 397) Deaths (all causes), n (%) a 143 (35)107 (27) All-cause deaths within 60 days of first dose 10 (2)6 (2) All-cause deaths within 30 days of last dose of first- line treatment 33 (8)14 (4) a Deaths at any time including long-term follow-up (post-study treatment). Based on the safety analysis set

13. Overall Tumor Response: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) Central Review a Local Review a Tumor Response, n (%) pmab + bev / Ox-CT (N = 413) bev / Ox-CT (N = 410) pmab + bev / Ox-CT (N = 413) bev / Ox-CT (N = 410) Best ORR187 (45)189 (46)210 (51)203 (50) Complete response 0 (0)2 (< 1)22 (5)27 (7) Partial response187 (45)189 (46)188 (46)176 (43) Stable disease121 (29)138 (34)114 (28)134 (33) Progressive disease b 28 (7)17 (4)29 (7)21 (5) Not done or unevaluable c 78 (19)64 (16)60 (15)52 (13) a CT scans performed Q12W; responses did not require confirmation b Central review unable to evaluate clinical disease progression (ie, non-radiographic PD); central review unable to accurately evaluate PD after surgical resections c Included missing and unreadable scans

14. Maximum Percent Change From Baseline in Sum of Longest Diameters by Central Review: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) Panitumumab + bevacizumab / Ox-CT bevacizumab / Ox-CT PR (n = 189) SD (n = 133) PD (n= 16) Unknown (n = 4) Best Overall Response PR (n = 187) SD (n = 120) PD (n= 25) Unknown (n = 2)

15. Progression-Free Survival: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) # PFS events (%) Median (95%CI), mos 599.6 (8.8-10.9) 5211.1 (10.3-11.9) pmab+bev/Ox-CT bev/Ox-CT HR = 1.27 (95% CI: 1.05–1.53) * # PFS events (%) Median (95% CI), mos 709.6 (8.8-10.7) 6411.0 (10.2-11.8) HR = 1.27 (95% CI: 1.07–1.50)* Central Review a Local Review b a Censoring based on last available scan read centrally Q12W b Censoring based on last day of patient contact or visit Q2W *Descriptive only

16. a Descriptive only. Statistical significance is limited by the lack of a prespecified significance boundary. Overall Survival: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab+bev/Ox-CT bev/Ox-CT Death events (%) Median (95%CI), mos 143 (35)19.4 (18.4- 20.8) 108 (26)NE HR = 1.43 (95% CI: 1.11–1.83) a Months Panitumumab N Censored N 413 0 334 34 205 84 73 96 1 55 No Panitumumab N Censored N 410 0 347 44 221 75 71 120 3 61 % Surviving

17. Time to Treatment Failure (Interim Analysis, May 2007 Data Cutoff) Panitumumab N Censored N No Panitumumab N Censored N 413247701910 0112570 410279731240 00141864 Months HR = 1.08 (95% CI: 0.93–1.25) Pmab+bev/Ox-CT5.7 (5.5-6.0) Bev/Ox-CT5.9 (5.7-6.2) Median (95% CI), mos

18. Treatment Exposure: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) Dose Delays % Patients Dose Reductions % Patients Relative Dose Intensity (RDI) % pmab + bev/Ox (N=413) bev/Ox (N=410) pmab + bev/Ox (N=413) bev/Ox (N=410) pmab + bev/Ox (N=413) bev/Ox (N=410) Panitumumab64N/A33N/A86N/A Bevacizumab6054339192 Oxaliplatin574928278387 5-FU56 a 49 a 24 a 17 a 82 b 86 b % Patients Inf 5-FU/Ox/bev ≥ 85% N/A 3342 a Bolus 5-FU b Infusional 5-FU N/A=not applicable

19. Reasons for First-Line Treatment Discontinuation: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab + bev / Ox-CT (413) bev / Ox-CT (410) Patients who discontinued first-line treatment, n (%) 396 (96)380 (93) Progressive events, n (%) a 137 (35)111 (29) Disease progression115 (29)100 (26) Deaths22 (6)11 (3) Non-progressive events, n (%) a 259 (65)269 (71) Adverse events89 (22)90 (24) Protocol violation14 (4)7 (2) Consent withdrawn/ refused treatment63 (16)78 (21) Other b 93 (23)94 (25) a Of those who discontinued first-line treatment b Other included end of first-line treatment without progression, requirement for alternative therapy, administrative, lost to follow-up, and other

20. Rates of Metastases Intervention For Curative Intent: Ox-CT Cohort (Interim Analysis, May 2007 Data Cutoff) pmab + bev / Ox-CT (N = 413) bev / Ox-CT (N = 410) Resection or radiofrequency ablation of metastases (i.e., liver, lung, other), n (%) 40 (10)26 (6)

21. Summary  Based on this updated interim analysis of the PACCE study, panitumumab + bevacizumab/Ox-CT was associated with shorter PFS time and increased toxicity, indicating that this combination has an unfavorable benefit-to-risk profile in unselected patients with mCRC  Lower-dose intensity was observed in the panitumumab + bevacizumab/Ox-CT arm  Most patients withdrew from the study for non-progressive events (65% in the panitumumab + bevacizumab/Ox-CT arm, 71% in the bevacizumab/Ox-CT arm)  Further data collection and analyses are ongoing, including subset analyses based on biomarkers  Phase 3 studies are currently ongoing to investigate panitumumab with chemotherapy in 1 st - and 2 nd -line mCRC –Study 20050181 investigates FOLFIRI +/- panitumumab as 2 nd -line in mCRC (Peeters et al., Abstract #335, Poster #A56) –PRIME/Study 20050203 investigates FOLFOX +/- panitumumab in first-line mCRC (Douillard et al., Abstract #443, Poster #A63)

22. ACKNOWLEDGEMENTS  Patients who participated in this study and their families  All investigators, co-investigators, and study staffs at 194 sites across the US  The Amgen study team