1. ASCO 2008 Annual Meeting Chicago (IL), May 30 – June 3, 2008 FOLFOXIRI (irinotecan, oxaliplatin and infusional 5FU/LV) in combination with bevacizumab (BV) in the first-line treatment of metastatic colorectal cancer (mCRC): a phase II study by the G.O.N.O. group A. Falcone 1,2, G. Masi 2, F. Loupakis 1,2, E. Vasile 1,2, A. Ciarlo 3, D. Cavaciocchi 3, D. Amoroso 4, M. Puglisi 5, E. Fea 6, I. Brunetti 7 1 Department of Oncology, Transplantation and New Technologies in Medicine, University of Pisa, Italy 2 Division of Medical Oncology, Azienda USL 6 - Istituto Toscano Tumori Livorno, Italy, 3 Division of Medical Oncology, Misericordia e Dolce Hospital, Prato, Italy, 4 Division of Medical Oncology, Versilia Hospital, Lido di Camaiore, Italy, 5 Division of Medical Oncology, University “La Sapienza”, Rome, Italy 6 Division of Medical Oncology, S. Croce e Carle Hospital, Cuneo, Italy, 7 Division of Medical Oncology, S. Chiara Hospital, Pisa, Italy Abstract n. 33736

2. ABSTRACT (updated) Background: The GONO-FOLFOXIRI regimen significantly improved response-rate (RR), progression-free survival (PFS), overall survival (OS) and post-CT surgical resection of metastases compared to FOLFIRI in a phase III study. The combination of BV with fluoropyrimidines/oxaliplatin/irinotecan-based doublets is safe and associated with improved efficacy. Methods: This phase II trial evaluates the combination of bevacizumab 5 mg/kg on d1 with the GONO-FOLFOXIRI regimen (irinotecan 165 mg/sqm d1, oxaliplatin 85 mg/sqm d1, l-LV 200 mg/sqm d1 and 5FU 3200 mg/sqm 48-h flat continuous infusion starting on d1) repeated every 2 weeks, as first-line treatment of initially unresectable mCRC pts. Results: Fifty-seven pts have been enrolled. Main pts characteristic are: M/F = 60%/40%, median age (range) = 61 (34-75) years, ECOG-PS 0/1/2 = 69%/26%/5%, primary colon/rectum = 72%/28%, primary on site = 13 pts (23%), sites of disease single/multiple = 56%/44%, liver only mts = 53%. All the patients are assessable for toxicity; the G3-4 maximum observed toxicities were: neutropenia 40% (febrile neutropenia 2%), anemia 4%, diarrhea 11%, nausea 4%, stomatitis 2%, deep venous thrombosis 4% and hypertension 9%; G1 bleeding occurred in 15 pts (26%). No toxic deaths have occurred. Up today 51 pts have been evaluated for response and we observed 6 CR, 32 PR (ORR = 75%) and 13 SD (disease control rate = 100%). So far 8 pts (16%) have undergone to secondary surgery on liver mts and 7 R0 resections have been performed. After a median follow up of 7.7 months, the 10 months actuarial PFS is 73%. Conclusions: The addiction of BV to the GONO-FOLFOXIRI regimen is feasible with manageable toxicities; the characteristic toxicity of BV and FOLFOXIRI occurs with the expected incidence and there were not unexpected adverse events. Data on activity are very promising. The study is still ongoing to complete the follow-up and to better determine the efficacy of the combination. A phase III study comparing FOLFIRI+BV vs FOLFOXIRI+BV has been planned by the GONO group.

3. BACKGROUND The best outcome of mCRC is achieved in pts receiving 5FU, irinotecan, and oxaliplatin in the course of their disease, and the resection-rate of metastases increases with the increase of the response-rate to first-line treatment. A way to expose 100% of pts to all the 3 cytoxic agents and to improve the response-rate is to incorporate them into the first-line therapy. (Grothey JCO 2004, Folprecht Ann Oncol 2005) The “GONO” FOLFOXIRI regimen is the first triple-drug combination demonstrated to be superior to an infusional 5FU containing doublet as FOLFIRI in terms of RR, R0 resections, PFS and OS in metastatic colorectal cancer patients (Falcone JCO 2007) Bevacizumab associated with IFL is feasible and significantly improves response rate, progression free and overall survival compared to IFL alone in the first-line treatment of MCRC pts. (Hurwitz NEJM 2004) Bevacizumab associated with oxaliplatin-based chemotherapy (FOLFOX or XELOX) is feasible and significantly improves efficacy compared to chemotherapy alone in first- and second-line treatment of MCRC pts. (Giantonio JCO 2007, Saltz JCO 2008) Two large phase IV trials (BRiTE, BEAT) evaluating a total of about 4000 MCRC pts show that the combination of Bevacizumab with any first-line chemotherapy is safe and effective

4. OBJECTIVES PRIMARY  Percentage of patients free of progression at 10 months SECONDARY  Response rate  R0 surgery of mets  Progression free survival  Overall survival  Safety profile  Evaluation of potential surrogate markers predictive of bevacizumab activity

5. MAIN SELECTION CRITERIA INCLUSION CRITERIA Histologically confirmed colorectal adenocarcinoma Unresectable and measurable metastatic disease (RECIST criteria) Age > 18 years and ≤ 75 years ECOG PS < 2 if aged < 71 years; ECOG PS = 0 if aged 71-75 years Previous adjuvant CT allowed if more than 12 months have elapsed between the end of adjuvant therapy and first relapse Adequate liver, renal and bone marrow functions Urine dipstick for proteinuria < 2+ EXCLUSION CRITERIA Prior palliative chemotherapy or treatment with bevacizumab Bowel obstruction, inflammatory enteropathy, extensive intestinal resection Presence or history of CNS metastasis Major surgical procedure, open biopsy or significant traumatic injury within 28 days prior to treatment Clinically significant cardiovascular disease, uncontrolled hypertension, thromboembolic or hemorrhagic events within 6 months prior to treatment, bleeding diathesis or coagulopathy

6. STATISTICS In phase II and III studies both the FOLFOXIRI regimen and the combination of bevacizumab with IFL regimens produced a median progression free survival of about 10 months (50% of patients free of progression at 10 months). The combination of FOLFOXIRI with bevacizumab will be considered promising in terms of activity if the percent of patients free of progression at 10 months will increase from 50% to 70%. According to the single-stage design described by Fleming, and selecting the design parameters p0 (percent of patients free of progression at 10 months in null hypothesis) = 0.50, and p1 (percent of patients free of progression at 10 months in alternative hypothesis) = 0.70, and considering also an α and β errors of 0.05 and 0.10, the study required a total of 53 evaluable patients. Treatment will be judged promising if at least 33 out of 53 patients free of progression at 10 months will be observed.

7. STUDY DESIGN FOLFOXIRI plus Bevacizumab was planned for a maximum of 12 cycles, or until progressive disease, or unacceptable toxicities, or patients’ refusal. Maintainance Bevacizumab* was recommended after the first 12 cycles until progressive disease, or unacceptable toxicities or patients’ refusal. Surgical radical resection of residual metastases in responsive patients was highly recommended and its feasibility was evaluated every 2 months. After resection it was recommended to restart FOLFOXIRI plus Bevacizumab (or a different treatment based on previous toxicities) to receive a total of 12 cycles of CT including those administered before surgery. Bevacizumab* was continued for a maximum of 12 months after resection or until tumor progression, unacceptable toxicity, or patient refusal. * On March 2008 the protocol was emended and Bevacizumab + 5FU/LV was recommended as maintainance treatment.

8. FOLFOXIRI + BEVACIZUMAB SCHEDULE So far, 640 total cycles of treatment have been administered: 506 cycles with FOLFOXIRI plus Bevacizumab and 134 cycles of maintanance treatment with Bevacizumab (in 1 patient with Bevacizumab plus 5FU/LV). 5FU flat continuous infusion 3200 mg/sqm L-LV 200 mg/sqm Oxaliplatin 85 mg/sqm 2 hours Repeated every 14 days CPT-11 165 mg/sqm 48 hours Day 1 Day 2 Day 3 1 hour BV 5 mg/Kg 30 min

9. PATIENTS’ CHARACTERISTICS N% Patients57- Age, median (range)61 (34-75)- Sex (M/F)34/2360/40 ECOG PS 0/1/239/15/369/26/5 Primary site (colon/rectum)41/1672/28 Surgery for primary tumor (y/n)44/1377/23 Previous adjuvant CT (y/n)5/529/91 Sites of disease (single/multiple)32/2556/44 Liver only mts3053

10. FOLFOXIRI + BEVACIZUMAB TOXICITY (maximum toxicity per patient; N=57) ToxicityG1 (%)G2 (%)G3 (%)G4 (%) Nausea393040 Vomiting252800 Diarrhea3321110 Stomatitis322120 Neutropenia19 2317 Febrile Neutropenia0002 Thrombocytopenia37000 Anemia493540 Neurotoxicity352800 Hypertension14590 Deep venous thrombosis2040 Bleeding26200 Cardiac ischemia0040 No toxic deaths have so far occurred.

11. ToxicityG1 (%)G2 (%)G3 (%)G4 (%) Nausea13400 Vomiting4400 Diarrhea0000 Stomatitis13400 Neutropenia0040 Febrile Neutropenia0000 Thrombocytopenia17000 Anemia22940 Neurotoxicity94300 Hypertension00130 Deep venous thrombosis0000 Bleeding13000 Cardiac ischemia0000 MAINTAINANCE BEVACIZUMAB ± 5FU/LV TOXICITY (maximum toxicity per patient; N=23) No toxic deaths have so far occurred.

12. RESPONSE RATE (RECIST CRITERIA) Total evaluable patients*N=51 Complete Response (CR)612% Partial Response (PR)3263% Overall Response Rate3875% Stable Disease (SD)1325% Progressive Disease00 Disease Control Rate (CR + PR + SD) 51100% *6 patients too early

13. R0-SURGERY So far 8 pts (16%) with liver only metastases have undergone to secondary surgery on liver mts; 7 R0 and 1 R1 resections have been performed (data still immature) 2 patients have achieved a confirmed pCR No deaths and no relevant morbidities have been observed after surgery 6 patients have received postoperative treatment:  3 with FOLFOXIRI plus Bevacizumab  1 with FOLFOXIRI  1 with FOLFIRI plus Bevacizumab  1 with Bevacizumab alone

14. PROGRESSION-FREE and OVERALL SURVIVAL Data on PFS and OS are promising, but still immature  Median Follow-up is 7.7 months  14 (25%) pts have so far progressed  At 10 months actuarial PFS is 73%  1 (2%) patient have died

15. CONCLUSIONS The addiction of BV to the GONO-FOLFOXIRI regimen is feasible with manageable toxicities The characteristic toxicities of BV and FOLFOXIRI occur with the expected incidence and there were not unexpected adverse events Treatment activity is very promising (RR=75%, Disease Control Rate= 100%, actuarial PFS at 10 months= 73%) Data on surgery of metastases, PFS and OS are still immature, but promising The accrual has been completed; the follow-up of enrolled patients is ongoing to better evaluate the efficacy A phase III study comparing FOLFIRI+BV vs FOLFOXIRI+BV has been planned by the GONO group

16. ACKNOWLEDGEMENTS Livorno Masi G, Loupakis F, Baldi G, Fornaro L, Antonuzzo A, Di Marsico R, Stasi I, Salvatore L, Cupini S, Fontana A, Vasile E, Andreuccetti M, Falcone A. PratoCiarlo A, Cavaciocchi D, Di Leo A. RomaFerraldeschi R, Puglisi M, Cortesi E. VersiliaDonati S, Rondini M, Puccetti C, Amoroso D. CuneoGranetto C, Fea E, Merlano M. Pisa Di Donato S, Brunetti I, Lencioni M, Pfanner E, Petrini I, Ricci S. GenovaSonaglio C, Chiara S.