1. Bevacizumab Beyond Progression ? Axel Grothey Professor of Oncology Mayo Clinic Rochester Axel Grothey Professor of Oncology Mayo Clinic Rochester

2. Continuation of Chemotherapy Beyond Progression FOLFOX  FOLFIRITournigand FOLFIRI  FOLFOXTournigand LV5FU2  FOLFIRIFOCUS LV5FU2  FOLFOXFOCUS Irino  Irino + CetuximabBOND, Saltz FOLFOX  FOLFIRITournigand FOLFIRI  FOLFOXTournigand LV5FU2  FOLFIRIFOCUS LV5FU2  FOLFOXFOCUS Irino  Irino + CetuximabBOND, Saltz

3. Cell membrane VEGF-A VEGF-R1 (Flt-1) Migration Invasion Survival VEGF-R3 (Flt-4) Lymphangio- genesis VEGF-R2 (KDR/Flk-1) Proliferation Survival Permeability PlGF VEGF-B VEGF-C, VEGF-D Functions VEGF Biology

4. VEGF-A VEGF-R1 (Flt-1) Migration Invasion Survival VEGF-R3 (Flt-4) Lymphangio- genesis VEGF-R2 (KDR/Flk-1) Proliferation Survival Permeability PlGF VEGF-B VEGF-C, VEGF-D Functions Large molecule VEGF inhibitors Bevacizumab Ramucirumab Aflibercept (VEGF Trap)

5. Characteristics of Anti-EGFR vs Anti-VEGF Therapy Minimal single agent activity In combination with chemo consistent increase in PFS Decrease in interstitial pressure, better delivery of chemo? “Normalization” of vasculature, better oxygenation? Minimal single agent activity In combination with chemo consistent increase in PFS Decrease in interstitial pressure, better delivery of chemo? “Normalization” of vasculature, better oxygenation? Single agent activity In combination with chemo consistent increase in RR Increased chemo- and radio-sensitivity Resensitization of tumors to chemo (CPT11) Anti-VEGF mAbAnti-EGFR mAb Main target: Tumor cells - genetically instable - Main target: Endothelial cells - genetically stable -

6. Is There a Rationale to Continue Bevacizumab Beyond Progression?

7. Continuation of Bevacizumab Beyond Progression - PRO Mechanism of action targets genetically stable (endothelial) cells Decreased intratumoral interstitial pressure leads to higher concentrations of chemotherapeutic agents Normalization of vasculature and better oxygenation  Cytotoxic effects of all (?) chemotherapeutics, regardless of “line of therapy” enhanced Mechanism of action targets genetically stable (endothelial) cells Decreased intratumoral interstitial pressure leads to higher concentrations of chemotherapeutic agents Normalization of vasculature and better oxygenation  Cytotoxic effects of all (?) chemotherapeutics, regardless of “line of therapy” enhanced

8. Inadequate for tumor growth Dynamic Effects of Anti-VEGF Therapy on Tumor Vasculature Normal Tumor vasculatureDays 2-5: normalized Anti-VEGFR Early effects (days 2-5):  Hypoxia /  Oxygenation Tumor vessel pruning Late effects (day 5): inhibition of blood vessel growth Winkler et al. Cancer Cell. 2004;6:553; Jain. Nat Med. 2003;9:685.

9. Placebo Anti-VEGF mAb *P<0.09 vs placebo. † P<0.05 vs placebo. Wildiers et al. Br J Cancer. 2003;88:1979. Effect of VEGF Inhibition on Vessel Density and Tumoral Chemotherapy Concentration 20 15 10 5 0 Tumor H33342 concentration (100 ng/g) † Tumor irinotecan concentration ( µ g/g) *

10. Continuation of Bevacizumab Beyond Progression - PRO Mechanism of action targets genetically stable (endothelial) cells Decreased intratumoral interstitial pressure leads to higher concentrations of chemotherapeutic agents Normalization of vasculature and better oxygenation  Cytotoxic effects of all (?) chemotherapeutics, regardless of “line of therapy” enhanced Mechanism of action targets genetically stable (endothelial) cells Decreased intratumoral interstitial pressure leads to higher concentrations of chemotherapeutic agents Normalization of vasculature and better oxygenation  Cytotoxic effects of all (?) chemotherapeutics, regardless of “line of therapy” enhanced In experimental models rapid regrowth of blood vessels after withdrawal of VEGF-inhibitors

11. Rapid Regrowth of Tumor Blood Vessels Selective inhibition of VEGFR signaling by AG-028262 in RIP-Tag2 mouse tumors Basement membrane sleeves Mancuso et al. JCI 2006

12. Continuation of Bevacizumab Beyond Progression - CON Potential alternate pathways to activate angiogenesis apart from VEGF Ang-system, FGF, PDGF and others “Co-option” - recruitment of previously established vessels Vascular remodeling, pericyte activation Potential alternate pathways to activate angiogenesis apart from VEGF Ang-system, FGF, PDGF and others “Co-option” - recruitment of previously established vessels Vascular remodeling, pericyte activation

13. The Complex Process of Tumor Angiogenesis

14. Pericytes Proliferate in Tumors Resuming Growth During Chronic VEGF Blockade Huang, J. et al. Mol Cancer Res 2004;2:36-42 Green =  SMA (Pericytes) ControlAntiVEGF PDGF  PDGFR

15. Pericytes Proliferate in Tumors Resuming Growth During Chronic VEGF Blockade Huang, J. et al. Mol Cancer Res 2004;2:36-42

16. Continuation of Bevacizumab Beyond Progression - CON Potential alternate pathways to activate angiogenesis apart from VEGF Ang-system, FGF, PDGF and others “Co-option” - recruitment of previously established vessels Vascular remodeling, pericyte activation Endothelial cells are not necessarily genetically stable Concept of cancer stem cells BEV is not non-toxic (GIP, ATE, HTN, RPLS…) Treatment alternatives exist most of the times BEV is expensive Potential alternate pathways to activate angiogenesis apart from VEGF Ang-system, FGF, PDGF and others “Co-option” - recruitment of previously established vessels Vascular remodeling, pericyte activation Endothelial cells are not necessarily genetically stable Concept of cancer stem cells BEV is not non-toxic (GIP, ATE, HTN, RPLS…) Treatment alternatives exist most of the times BEV is expensive

17. Clinical experience? No prospectively randomized evaluation to date…

18. BBP (n=642) No BBP (n=531) No Post-PD Treatment (n=253) Evaluable patients (n=1953) 1 st Progression (n=1445) BRiTE Registry - Patients with Bevacizumab Beyond Progression (BBP) BRiTE: Total N=1953 1445 pts with 1st PD 932 deaths (1/21/07 cut-off) Median follow-up 19.6 mo Grothey et al. JCO 2008 Physician decision - no randomization

19. BRiTE: Patient Outcome Based on Treatment Post 1st PD BBP (n=642) No BBP (n=531) No Post-PD Treatment (n=253) # of deaths (%) 168 (66%) 306 (58%) 260 (41%) Median OS (mo) 12.619.931.8 1yr OS rate (%) 52.577.387.7 OS after 1st PD (mo) 3.69.519.2 Grothey et al. JCO 2008

20. BRiTE: Continuation of BEV post first progression may increase survival No Treatment (n=253) No BEV post PD (n=531) BEV post PD (n=642) Post-progression Bevacizumab HR=0.48 (0.41-0.57) P<0.001

21. Multivariate Analysis of Pre- and Post- Treatment Variables on Survival Grothey et al. JCO 2008

22. ARIES: Post-progression observation of bevacizumab treatment ARIES* total n=1,548 prospective phase III study primary endpoint: survival beyond progression secondary endpoint: OS, time to first PD, OS, safety ARIES* total n=1,548 prospective phase III study primary endpoint: survival beyond progression secondary endpoint: OS, time to first PD, OS, safety Bevacizumab post-PD (n=406) No post-PD treatment § (n=282) No bevacizumab post-PD (n=336) Physician decision (no randomization) Unresectable mCRC treated with first-line chemotherapy (n=1,548) First progression (n=1,113 ‡ ) First-line chemotherapy + bevacizumab *Non-randomized, observational study ‡ 1,026 patients were alive 2 months after first PD § No treatment ever or bevacizumab and/or chemotherapy ≥2 months after PD Cohn, et al. ASCO 2010

23. ARIES: Potential survival benefit from bevacizumab beyond progression No BEV post-PD * (n=336) BEV post-PD ‡ (n=408) Median OS, months (95% CI) 18.7 (17.5–20.4) 27.5 (25.6–29.0) Median survival beyond first progression, months (95% CI) § 7.5 (6.2–8.7) 14.1 (12.6–16.1) Adjusted HR (95% CI)**1.0 (Reference) 0.52 (0.42–0.63) * Patients alive 2 months post-PD and starting chemotherapy/biologics <2 months post-PD; no bevacizumab ever post-PD ‡ Patients alive 2 months post-PD and starting chemotherapy/biologics + bevacizumab <2 months post-PD §For SBP, t 0 =PD+2 months **Multivariate model adjusted for patient characteristics Cohn, et al. ASCO 2010

24. ARIES*: Does bevacizumab extend survival beyond progression? Survival beyond progression estimate Bevacizumab post-PD (n=408) No bevacizumab post-PD (n=336) 7.514.1 HR=0.52 (95% CI: 0.42–0.63) p<0.001 1.0 0.8 0.6 0.4 0.2 0 051015202530 Months Cohn, et al. ASCO 2010 *Non-randomized, observational study ‡ Post-progression bevacizumab versus no bevacizumab study

25. Limitations of the Analysis Patients were not randomized Actual administration dates for BV and CT not collected; missing BV and CT stop dates Potential bias that patients who survived longer had a greater potential to be treated with BBP – but sensitivity analyses suggest minimal impact of these biases Possibility of unmeasured factors that may have biased these results  Randomized trial needed!

26. AIO 0504 / Roche ML18147 Multinational European Trial Any-OX + BEV Any-IRI + BEV Any-IRIAny-OX Any-OX + BEV RR N = 820 Primary EP: OS Accrual completed May 31, 2010

27. Pertinent Side-Effects of Anti-VEGF Therapy Hypertension Arterial thrombotic/ thromboembolic events (ATEs) Gastrointestinal perforation (GIP) Bleeding Delayed wound healing Proteinuria Hypertension Arterial thrombotic/ thromboembolic events (ATEs) Gastrointestinal perforation (GIP) Bleeding Delayed wound healing Proteinuria

28. Attempt at Classification of AEs Preeclampsia-like syndrome with hypertension, proteinuria, and hypertensive encephalopathy Hypercoagulabilty with increased risk for arterial and - less likely - venous thrombosis and thromboembolic events Anti-angiogenic syndrome with decreased wound healing, risk of gastrointestinal perforation (GIP) and bleeding Preeclampsia-like syndrome with hypertension, proteinuria, and hypertensive encephalopathy Hypercoagulabilty with increased risk for arterial and - less likely - venous thrombosis and thromboembolic events Anti-angiogenic syndrome with decreased wound healing, risk of gastrointestinal perforation (GIP) and bleeding

29. ATE Incidence From Start of BEV Treatment Months from start of BEV Number of events Kozloff et al., Oncologist 2009

30. Incidence of GIP in BRiTE and in BEV Treatment Arms of Phase III Studies in mCRC Study% GIP BRiTE (N=1960) 1.7 AVF2107 IFL+BEV (N=393) 1.5 E3200 FOLFOX+BEV (N=293) 1.7 E3200 BEV mono (N=234) 1.7 Sugrue et al, ASCO, Atlanta, June 2-6, 2006, Kozloff et al., Oncologist 2009 Months from start of BEV Number of events

31. Incidence of BEV-related Safety Events in BRiTE No increase in rates of ATE, bleeding or GI perforation in patients who continued BEV Grothey et al. JCO 2008

32. EFC10262: VELOUR Phase III Trial 2 nd Line FOLFIRI +/- VEGF-TRAP (Aflibercept) Stratification factors: Prior bevacizumab (Y/N) ECOG PS (0 vs 1 vs 2) 1:1 mCRC after failure of an oxaliplatin based regimen R 600 pts Aflibercept 4 mg/kg IV + FOLFIRI q 2 weeks 600 pts Placebo + FOLFIRI q 2 weeks 32 30% of patients had prior BEV PI: Allegra

33. VELOUR: Study Design and Endpoints Multinational, randomized, placebo controlled 28 Countries / 176 Active Sites Primary Endpoint: Overall Survival 90 percent power to detect a 20 % reduction in HR for OS (two-sided log-rank) Secondary Endpoints: Progression free survival Overall response rate Safety Aflibercept pharmacokinetics and immunogenicity Multinational, randomized, placebo controlled 28 Countries / 176 Active Sites Primary Endpoint: Overall Survival 90 percent power to detect a 20 % reduction in HR for OS (two-sided log-rank) Secondary Endpoints: Progression free survival Overall response rate Safety Aflibercept pharmacokinetics and immunogenicity

34. VELOUR: Press Release April 26, 2011 Results to be presented at ESMO GI in Barcelona 2011

35. I4T-MC-JVBBPhase III Trial 2 nd Line FOLFIRI +/- Ramucirumab Stratification factors: Region KRAS status First-line TTP (<>6 mos) 1:1 mCRC after failure FP/oxaliplatin + BEV regimen R 525 pts Ramucirumab IV + FOLFIRI q 2 weeks 525 pts Placebo + FOLFIRI q 2 weeks 35 Primary EP: OS PIs: Tabernero, Grothey

36. Cytokine increase on BEV therapy Kopetz et al., JCO 2010

37. Regorafenib – A Multi-Kinase Inhibitor Cellular Phosphorylation AssaysIC 50 nM VEGFR-2 Phosphorylation, 293 Cells8 TIE2-Receptor Phosphorylation, CHO Cells 31 PDGFR-β Phosphorylation, Aortic SM Cells90 mVEGFR3 Phosphorylation, 293 Cells150 Mutant RET Phosphorylation, Thyroid TT Cells 10 Mutant c-KIT Phosphorylation, GIST 882 Cells20 FGF-10 FGFR MCF-7 Cells 150-300 MAPK ERK-P HCC, HepG2 Cells500 Cell Proliferation Assays IC 50 nM VEGF/HuVEC (2% FCS) BrdU 4 bFGF/ HuVEC (2% FCS) BrdU 120 PDGFBB/Aortic SM (0.1% BSA) BrdU 121 Thyroid TT RET C643W (10% FCS) 33 GIST 882 KIT K642E (10% FCS) 45 Breast, MDA MB 231 (10% FCS) 570 Melanoma, A375 (10% FCS) 900 HCC HepG2 (10% FCS) 560

38. Regorafenib Salvage Therapy Registration Trial Primary endpoint OS:increase OS from 4.5 to 6.0 months; HR = 0.75 Significance level/power: 0.025 (one-sided)/90% Accrual period (months): 26 ( accrual rate 30 pat./month) Study duration (months): 31.5 Total number of events: 582 Total number of patients: 690 Primary endpoint OS:increase OS from 4.5 to 6.0 months; HR = 0.75 Significance level/power: 0.025 (one-sided)/90% Accrual period (months): 26 ( accrual rate 30 pat./month) Study duration (months): 31.5 Total number of events: 582 Total number of patients: 690 Primary endpoint: OS CRC 3rd/4th line Regorafenib 160 mg od 3wks on/1 wk off + BSC Placebo + BSC 2:1 randomization Accrual completed Feb 2011, within 9 mos

39. Conclusions Continuation of BEV beyond progression (BBP) has Preclinical rationale and Support from results of observational cohort studies However… Financial and biological implications of this concept mandate prospective evaluation in randomized phase III trials before BBP can be considered standard of care A pivotal European trial has completed accrual – results are awaited for late 2011 Continuation of BEV beyond progression (BBP) has Preclinical rationale and Support from results of observational cohort studies However… Financial and biological implications of this concept mandate prospective evaluation in randomized phase III trials before BBP can be considered standard of care A pivotal European trial has completed accrual – results are awaited for late 2011

40. How A Cancer Cell Works