1. R.G. Amado Analysis of KRAS Mutations in Patients With Metastatic Colorectal Cancer Receiving Panitumumab Monotherapy Abstract 0007: Wild-type KRAS is Required for Panitumumab Efficacy in Patients With Metastatic Colorectal Cancer: Results From a Randomized, Controlled Trial Abstract 0007: Wild-type KRAS is Required for Panitumumab Efficacy in Patients With Metastatic Colorectal Cancer: Results From a Randomized, Controlled Trial Rafael G. Amado, Michael Wolf, Dan Freeman, Marc Peeters, Eric Van Cutsem, Salvatore Siena, Sid Suggs, Scott Patterson, David Chang Rafael G. Amado, Michael Wolf, Dan Freeman, Marc Peeters, Eric Van Cutsem, Salvatore Siena, Sid Suggs, Scott Patterson, David Chang Abstract 3014: Association of Somatic KRAS Gene Mutations and Clinical Outcome in Patients With Metastatic Colorectal Cancer Receiving Panitumumab Monotherapy Abstract 3014: Association of Somatic KRAS Gene Mutations and Clinical Outcome in Patients With Metastatic Colorectal Cancer Receiving Panitumumab Monotherapy Dan Freeman, Todd Juan, Neal J. Meropol, J. Randolph Hecht, Jordan Berlin, Eric Van Cutsem, Maureen Reiner, Robert Radinsky, Rafael G. Amado, Marc Peeters Dan Freeman, Todd Juan, Neal J. Meropol, J. Randolph Hecht, Jordan Berlin, Eric Van Cutsem, Maureen Reiner, Robert Radinsky, Rafael G. Amado, Marc Peeters

2. R.G. Amado The KRAS Oncogene The KRAS gene encodes the human cellular homolog of the transforming gene Kirsten rat sarcoma-2 virusThe KRAS gene encodes the human cellular homolog of the transforming gene Kirsten rat sarcoma-2 virus KRAS is a self-inactivating signal transducerKRAS is a self-inactivating signal transducer –It cycles from GDP bound (“off” state) to GTP bound (“on” state) in response to receptor activation –This response is transient due to the intrinsic GTPase activity KRAS oncogenes harbor activating mutations yielding proteins with reduced GTPase activityKRAS oncogenes harbor activating mutations yielding proteins with reduced GTPase activity These activating KRAS mutations are among the most common oncogenic alterations in cancer 1These activating KRAS mutations are among the most common oncogenic alterations in cancer 1 1 Malumbres and Barbacid. Nat Rev Cancer. 2003;3:459-465.

3. R.G. Amado EGFr Signal Transduction Akt SOS FOS Myc P13K FKHR mTOR PTEN MEK 1/2 MAPK BAD GSK-3 Shc Grb-2 Ras Raf Jun p27 Cyclin D-1 Ligand Ligand SignalAdapters and Enzymes SignalCascade EGFR dimer Transcription Factors STAT

4. R.G. Amado WT, wild type; MT, mutant; cmab, cetuximab; CT, chemotherapy; pmab, panitumumab Objective Response N (%) Reference Treatment (panitumumab or cetuximab) No of patients (WT:MT) MTWT A. Liévre, et al. (AACR Proceedings, 2007) cmab ± CT76 (49:27)0 (0)24 (49) S. Benvenuti, et al. (Cancer Res, 2007) pmab or cmab or cmab + CT48 (32:16)1 (6)10 (31) W. De Roock, et al. (ASCO Proceedings, 2007) cmab or cmab + irinotecan113 (67:46)0 (0)27 (40) D. Finocchiaro, et al. (ASCO Proceedings, 2007) cmab ± CT81 (49:32)2 (6)13 (26) F. Di Fiore, et al. (Br J Cancer, 2007) cmab + CT59 (43:16)0 (0)12 (28) S. Khambata-Ford, et al. (J Clin Oncol, 2007) cmab80 (50:30)0 (0)5 (10) Single-Arm Studies Support the Hypothesis for KRAS as a Biomarker for EGFr Inhibitors

5. R.G. Amado KRAS Analysis of Single-Arm, Panitumumab Monotherapy Studies Patient samples from 3 Amgen panitumumab monotherapy, single-arm, phase 2 trials in metastatic colorectal cancer were obtained under an ancillary biomarker protocolPatient samples from 3 Amgen panitumumab monotherapy, single-arm, phase 2 trials in metastatic colorectal cancer were obtained under an ancillary biomarker protocol The majority of patient samples were archived tumor samples from the primary resectionThe majority of patient samples were archived tumor samples from the primary resection KRAS mutational status was determined using cloning and sequencing of DNA isolated from paraffin-embedded tumor samplesKRAS mutational status was determined using cloning and sequencing of DNA isolated from paraffin-embedded tumor samples KRAS mutational status was correlated with clinical outcomes including response, progression-free survival, and overall survivalKRAS mutational status was correlated with clinical outcomes including response, progression-free survival, and overall survival

6. R.G. Amado Objective Tumor Response All PatientsMutant KRASWild-type KRAS Total (N = 62) Total (N = 24) Total (N = 38) PR, n (%) 4 (6.5)04 (11) SD, n (%) 25 (40)5 (21)20 (53) PD, n (%) 33 (53)19 (79)14 (37) PR + SD, n (%) 29 (47)5 (21)24 (63) PR partial response; SD, stable disease; PD, disease progression

7. R.G. Amado Progression-Free Survival for Panitumumab- Treated Patients by KRAS Status _ Median (95% CI) in Weeks Mutant: 7.4 (7.1–8.0) -- Wild-type: 16.2 (8.3–23.7) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Proportion with PFS 5550454035302520151050 24 38 Patients at Risk: Mutant Wild-type 24 38 6 25 6 4 17 3 14 2 13 2 7633 Weeks From Enrollment

8. R.G. Amado Randomization stratification ECOG score: 0-1 vs. 2 Geographic region: Western EU vs. Central & Eastern EU vs. Rest of World 1:1 PanitumumabPD Follow-up 6.0 mg/kg Q2W + BSC BSCPD Follow-up RANDOMIZERANDOMIZE Optional Panitumumab Crossover Study Hypothesis: The treatment effect of panitumumab monotherapy is larger in patients with wild-type KRAS compared to patients with mutant KRAS KRAS Analysis of a Phase 3, Randomized, Controlled Trial Comparing Panitumumab vs Best Supportive Care (BSC) in Colorectal Cancer Van Cutsem, Peeters et al. JCO. 2007;25:1658-1664.

9. R.G. Amado Objectives and Analysis Methodology Primary Objective To assess if the effect of panitumumab on progression- free survival (PFS) was significantly greater in patients with wild-type KRAS compared to patients with mutant KRAS Secondary Objectives To assess whether panitumumab improves PFS compared with BSC alone in patients with wild-type KRAS To assess whether panitumumab improves OS compared with BSC alone in patients with wild-type KRAS Compare PFS in wild-type KRAS subset Test for a PFS effect among all randomized patients at a 5% level Test for quantitative PFS effect interaction, i.e., wild-type effect > mutant p ≤ 0.05p > 0.05 p ≤ 0.05 p > 0.05 Compare OR & OS in wild-type KRAS subset STOP

10. R.G. Amado Assay Used to Detect KRAS Mutational Status DNA was isolated from fixed tumor samplesDNA was isolated from fixed tumor samples Mutant KRAS was detected using a KRAS mutation kit (DxS Ltd, Manchester, UK) that used allele-specific, real-time PCRMutant KRAS was detected using a KRAS mutation kit (DxS Ltd, Manchester, UK) that used allele-specific, real-time PCR –The kit can detect approximately 1% of mutant DNA in a background of wild-type genomic DNA –The test identifies 7 somatic mutations in codons 12 and 13 Gly 12 AspGly 12 Asp Gly 12 AlaGly 12 Ala Gly 12 ValGly 12 Val Gly 12 SerGly 12 Ser Gly 12 Arg Gly 12 Cys Gly 13 Asp

11. R.G. Amado Results: Prevalence of Mutant KRAS Panitumumab + BSC BSC aloneTotal Patients randomized, n231232463 KRAS not tested, n (%)11 (5) 7 (3)18 (4) KRAS tests failed, n (%)12 (5)6 (3)18 (4) Patients included in KRAS analysis, n (%) 208 (90)219 (94)427 (92) Wild-type KRAS, n (%)124 (60)119 (54)243 (57) Mutant KRAS, n (%)84 (40)100 (46)184 (43) BSC, best supportive care

12. R.G. Amado Distribution of KRAS Mutations Panitumumab + BSC BSC alone Total Mutation Mutationn%n%n% 12 Ala 12 Ala89.577.0158.2 12 Asp 12 Asp3440.53636.07038.0 12 Arg 12 Arg00.033.031.6 12 Val 12 Val1517.92525.04021.7 12 Cys 12 Cys78.377.0147.6 12 Ser 12 Ser56.099.0147.6 13 Asp 13 Asp1517.91414.02915.8

13. R.G. Amado Baseline Disease Characteristics KRAS Evaluable Group KRAS Evaluable Mutant KRAS Wild-type KRAS Total (N = 427) Total (N = 184) Total (N = 243) Sex, n (%) Men270 (63)111 (60)159 (65) Baseline age, years Median (min, max)62.0 (27, 83) 63.0 (29, 82) Primary diagnosis, n (%) Colon cancer286 (67)118 (64)168 (69) Months since primary diagnosis Median25.124.125.1 ECOG performance status, n (%) 0-1366 (86)155 (84)211 (87) ≥ 261 (14)29 (16)32 (13) Prior adjuvant chemotherapy, n (%) Yes149 (35)67 (36)82 (34) Cells with EGFr membrane staining, n (%) 1% to < 10%103 (24)43 (23)60 (25) 10% to 100%322 (75)140 (76)182 (75) ECOG, Eastern Cooperative Oncology Group; EGFr, epidermal growth factor receptor

14. R.G. Amado KRAS Evaluable Patients: PFS by Treatment 0246810121416182022242628303234 Weeks 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 363840424446485052 1971881781067971645550494937292524191515151299766 20016814275423425231916141410101010986654443 208 219 Proportion with PFS Pmab + BSC BSC Alone Patients at Risk 191/208 (92)8.015.4 209/219 (95)7.3 9.6 Pmab + BSC BSC Alone Events/N (%) Median In Weeks Mean In Weeks HR = 0.59 (95% CI: 0.48–0.72)

15. R.G. Amado Mutant KRAS Subgroup: PFS by Treatment 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Proportion with PFS Pmab + BSC BSC Alone 0246810121416182022242628303234 Weeks 363840424446485052 Patients at Risk 787672261086555544442222222111 91776137221910986554444443332222 84 100 76/84 (90)7.49.9 95/100 (95)7.3 10.2 Pmab + BSC BSC Alone Events/N (%) Median In Weeks Mean In Weeks HR = 0.99 (95% CI: 0.73–1.36)

16. R.G. Amado Wild-type KRAS Subgroup: PFS by Treatment p < 0.0001 for quantitative-interaction test comparing PFS log-HR (pmab/BSC) between KRAS groups

17. R.G. Amado 0.17-0.47 Wild-Type KRAS Subgroup: Treatment Effect on PFS by Demographic/Disease Status HR, hazard ratio; Pmab, panitumumab; BSC, best supportive care Adjusted for randomization factors (ECOG score, geographic region) All Randomized Male Female Age: < 65 Age: 65+ Primary: Colon Primary: Rectal ECOG: 0–1 ECOG: 2–3 Prior regimens: 3+ Met. Sites: 1–2 Met. Sites: 3–5 EGFr: 1–< 10% EGFr: 10–35% EGFr: > 35% EGFr: 1+ EGFr: 2+ EGFr: 3+ Prior regimens: 3 Prior regimens: 2 243 159 84 141 102 168 75 211 32 100 172 69 60 101 81 69 127 47 90 142 0.45 0.42 0.46 0.42 0.47 0.36 0.47 0.35 0.27 0.42 0.52 0.30 0.49 0.34 0.33 0.41 0.37 0.28 0.54 0.34-0.59 0.30-0.59 0.29-0.73 0.29-0.60 0.31-0.73 0.34-0.65 0.21-0.61 0.35-0.62 0.15-0.82 0.17-0.44 0.30-0.59 0.30-0.89 0.16-0.56 0.31-0.75 0.20-0.58 0.18-0.63 0.28-0.60 0.18-0.75 0.38-0.76 0.10.20.30.40.50.60.70.80.91.01.11.2 Hazard Ratio (Pmab/BSC) Favors: PmabBSCFactorsNHR95% CI

18. R.G. Amado Objective Tumor Response (Central Radiology) KRAS All Evaluable n (%) Mutant n (%) Wild-type n (%) Response Pmab (N = 208) BSC (N = 219) Pmab (N = 84) BSC (N = 100) Pmab (N = 124) BSC (N = 119) CR0 (0) PR21 (10)0 (0) 21 (17)0 (0) SD52 (25)22 (10)10 (12)8 (8)42 (34)14 (12) PD104 (50)149 (68)59 (70)60 (60)45 (36)89 (75) CR, PR, SD73 (35)22 (10)10 (12)8 (8)63 (51)14 (12) Pmab, panitumumab; BSC, best supportive care; CR, complete response; PR partial response; SD, stable disease; PD, disease progression

19. R.G. Amado Objective Tumor Response Extension Study 20030194 (Investigator assessment) KRAS All Evaluable n (%) Mutant n (%) Wild-type n (%) ResponsePmab (N = 168) Pmab (N = 77) Pmab (N = 91) CR1 (1)0 (0)1 (1) PR19 (11)0 (0)19 (21) SD55 (33)20 (26)35 (38) PD60 (36)37 (48)23 (25) CR, PR20 (12)0 (0)20 (22) CR, PR, SD75 (45)20 (26)55 (60) Pmab, panitumumab; CR, complete response; PR partial response; SD, stable disease; PD, disease progression

20. R.G. Amado Maximum Percent Decrease in Target Lesions Final Analysis, KRAS Evaluable Group 160 140 120 100 80 60 40 20 % Change -20 -40 -60 -80 0 PR (0%)SD (12%)PD (70%) Mutant Patient Pmab + BSC 160 140 120 100 80 60 40 20 % Change -20 -40 -60 -80 0 PR (17%)SD (34%)PD (36%) Wild-Type Patient 160 140 120 100 80 60 40 20 % Change -20 -40 -60 -80 0 PR (0%)SD (8%)PD (60%) Patient BSC Alone 160 140 120 100 80 60 40 20 % Change -20 -40 -60 -80 0 PR (0%)SD (12%)PD (75%) Patient

21. R.G. Amado Overall Survival by Treatment and KRAS 124925835189321 1198254281810510 8451211063310 100553018113000 Pmab–Wild-type BSC–Wild-type Pmab–Mutant BSC–Mutant 0246810121416182022242628303234 Months 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Survival Probability Pmab–Wild-type 107 / 124 (86) 110 / 119 (92) 8.1 7.6 Events/N (%) Median In Months BSC–Wild-type Wild-type vs. Mutant (treatment arms combined) Pmab–Mutant BSC–Mutant 79 / 84 (94) 95 / 100 (95) 4.9 4.4 HR = 0.67 (95% CI: 0.55–0.82) (adjusted for treatment and randomization factors; ECOG, region)

22. R.G. Amado Overview of Safety From KRAS Analysis ( Panitumumab-Treated Patients) KRAS Evaluable Mutant KRAS Wild-type KRAS Exposure: Mean number of infusions per patient7.94.910.0 Any AE, %100 Grade 3 or 4 AE, %372844 Treatment-related grade 3 AE, %201225 Any grade integument-related AE, %929093 Grade 3201325 Grade 4< 110 Any grade diarrhea, %221924 Grade 3212 Any grade infections, %473555 Grade 3758 Hypomagnesemia as an AE, %102 Infusion reactions, %472 AE leading to withdrawal, %657 AE, Adverse Event

23. R.G. Amado Conclusions The efficacy of panitumumab monotherapy in metastatic CRC seems confined to patients with wild-type KRASThe efficacy of panitumumab monotherapy in metastatic CRC seems confined to patients with wild-type KRAS KRAS genotyping of tumors should be strongly considered in patients with metastatic CRC being treated with panitumumab monotherapyKRAS genotyping of tumors should be strongly considered in patients with metastatic CRC being treated with panitumumab monotherapy Ongoing studies in 1 st and 2 nd lines will prospectively examine the effect of panitumumab in combination with chemotherapy in patients with wild-type and mutant KRAS tumorsOngoing studies in 1 st and 2 nd lines will prospectively examine the effect of panitumumab in combination with chemotherapy in patients with wild-type and mutant KRAS tumors Future studies should investigate the role of KRAS mutational status in the adjuvant setting in colon cancer and in other indicationsFuture studies should investigate the role of KRAS mutational status in the adjuvant setting in colon cancer and in other indications

24. R.G. Amado Acknowledgments We wish to thank the patients and their families for study participationWe wish to thank the patients and their families for study participation We also thank all investigators and study personnelWe also thank all investigators and study personnel Disclosures This study was sponsored by Amgen Inc. R.G. Amado, M. Wolf, D. Freeman, S. Suggs, S. Patterson, D. Chang, T. Juan, M. Reiner, and R. Radinsky are Amgen employees and own Amgen Inc. stock. M. Peeters, Amgen advisory board and Amgen honoraria; E. Van Cutsem, Amgen advisory board and Amgen honoraria; N.J. Meropol, consultant for Amgen, Genentech, and Imclone; J. Berlin, advisory boards for Amgen, BMS, and Imclone.