1. DEFINITIVE CHEMORADIOTHERAPY (CRT) WITH FOLFOX 4 OR 5FU-CISPLATIN AS FIRST-LINE TREATMENT OF PATIENTS (PTS) WITH INOPERABLE ESOPHAGEAL CANCER (IEC) : FINAL RESULTS OF A RANDOMIZED PHASE II STUDY T. Conroy (1), Y. Yataghene (2), PL. Etienne (3), P. Michel (4), H. Senellart (5), JL. Raoul (6), L. Mineur (7), M. Rives (8), X. Mirabel (9), A. Adenis (9). (1) Centre Alexis Vautrin, Nancy ; (2) sanofi-aventis france, Paris ; (3) Clinique Armoricaine, St-Brieuc ; (4) University Hospital, Rouen; (5) Centre Rene Gauducheau, Nantes ; (6) Centre Eugene Marquis Rennes ; (7) Clinique Sainte Catherine, Avignon ; (8) Centre Claudius Regaud, Toulouse; (9) Centre Oscar Lambret, Lille; France.

2. Abstract  4532 Background: The RTOG regimen of concurrent CRT with 5FU/CDDP represents the standard treatment of pts with IEC (Al-Sarraf JCO, 1997). As new combinations are required to improve survival, we launched a multicenter randomized phase II study to assess the feasibility and endoscopic complete response rate (ECRR) of combination CRT with Folfox 4 (arm A) or 5FU/CDDP (arm B) in pts with IEC. Methods: Radiotherapy (RT) was in both arms 50 Gy (2Gy/fraction) 5 days/ week for 5 weeks. In Arm A, pts received 6 biweekly cycles (cy): Oxaliplatin 85mg/m2 d1 and Leucovorin 200mg/m2 followed by 5-FU 400mg/m2 bolus and 600mg/m2 22h continuous infusion (C.I.) d1-2 ; the first 3 cy were delivered during RT, the 3 other after. Arm B: 4 cy were delivered: CDDP 75mg/m2 d1 followed by 5FU 1000 mg/m2/d C.I. d1-4, the first two cy during RT and 2 other after RT. To be evaluable pts must have completed the concurrent CRT. Response evaluation (esophagoscopy + CT-scan) was done on week 15; ECRR was reviewed by an independent experts committee. Results: 97 pts (A/B= 53/44) were included and 95 treated: 19% adenocarcinoma, 81% squamous cell carcinoma; median age 59 [39-81] and PS 0-1: 98%. Respectively in arms A/B, 89/91% of pts completed the concurrent CRT and 75/70% the full treatment. Main NCI-CTC grade (G) 3-4 toxicities (% per pt) were in arms A/B: neutropenia 19/19, febrile neutropenia 12/5, neutropenic infection 2/7, esophagitis 8/14. Neurotoxicity: no G 3 occurred, in arm A/B, 13%/ 2% of pts had a G 2 and 46%/13% a G 1. 47/40 pts (90%) were evaluable for response. ECRR was 44.7% and 30.0% in arm A and B respectively, leading in all patients (ITT population) to median TTP of 15.0  9.9-18.9] / 9.5  7.6- ongoing  months (mos), median Event Free Survival of 11.6 [8.6-17.2]/ 7.8 [5.1-9.9] mos and median overall survival of 22.7  16.7-ongoing  / 14.7  9.2-17.2  mos. Conclusion: The new concurrent CRT with Folfox4 is a well tolerated convenient combination with promising efficacy associated to very good survival results. Final data will be presented at the meeting. The study continues as a phase III trial.

3. Background (1) In inoperable esophageal cancer, concurrent chemoradiotherapy using 5FU-Cisplatin is the standard of care (Al-Sarraf M et al. J. Clin. Oncol 1997; 15: 277-284). However, with 5FU-Cisplatin regimen: –20% of patients experienced major toxicities –40 % of patients could not complete full sequence of treatment Oxaliplatin has shown encouraging activity in tumors that respond to cisplatin or carboplatin Oxaliplatin demonstrates activity superior to that of cisplatin, having a broader spectrum of action than the parent compound and shows less toxicity Oxaliplatin, in combination with 5-fluorouracil and leucovorin is active on gastric adenocarcinoma and should be active in esophageal adenocarcinoma

4. Background (2) According to a phase I study, Oxaliplatin associated with 5- fluorouracil and leucovorin with concomitant radiotherapy as first-line treatment of inoperable advanced esophageal cancer is safe and active: –MTD was reached at oxaliplatin dose level of 100mg/m² with asthenia and diarrhea as dose-limiting toxicities –Recommended doses were those of Folfox4 regimen with oxaliplatin 85 mg/m² (Giovannini M et al. J Clin. Oncol. 2004; 22 (14S) : 324 s; Abstract 4044) 81% complete responses and 38% pathological complete responses were observed in 38 patients treated with continuous 5-FU, oxaliplatin 85 mg/m² biweekly and 50.4 Gy radiotherapy (Khushalani NI et al. J Clin. Oncol. 2002; 20: 2844- 50)

5. Objectives Primary Objectives : -To assess the feasibility (completion of full planned treatment) 1 of FOLFOX 4 regimen or 5-FU/Cisplatin regimen with concomitant radiotherapy as first-line treatment of inoperable advanced esophageal cancer. -To assess endoscopic complete response 2 rate in both arms. Secondary Objective : - To assess the toxicity profile of each arm using NCI-CTC Version 3.0 1 Complete planned treatment is: CRT and all cycles of chemotherapy 2 Endoscopic complete response is: complete disappearance of any tumor, esophageal ulceration or stenosis, with no new lesion (all endoscopic pictures have to be available) and no CT-scan progression. Endoscopic response was assessed by an Independent Data Monitoring Committee (IDMC) on the population evaluable for response, i.e. those who received concomitant chemoradiotherapy (with or without further chemotherapy)

6. Methods (1) Study design Multicenter (n=21) randomized phase II trial Randomization between 2 arms, stratified by blocks according to centre and histological type (adenocarcinoma or adenosquamous versus squamous cell carcinoma) Statistical considerations The required number of patients for this phase II study was determined according to a Fleming single stage Phase II procedure [Machin D et al. Sample size tables for clinical studies. Second edition. Blackwell Science, Oxford, 1997]. The design parameters assessed the treatment to be insufficiently active if the endoscopic complete response rate was less than 30% and sufficiently active if the objective response rate was at least 50%. A total of 44 evaluable patients per arm needed to be included in the study, assuming an alpha error rate of 0.05 (one-sided) and a beta error rate of 0.15. The 95% confidence limits were provided for response rates using the exact method. Time related parameters were analysed using the Kaplan- Meier method [Kaplan-Meier 1958].

7. Methods (2) At the end of the first part of the trial (randomised phase II) on 88 patients (44 in each arm), the decision to continue the study will be taken by the coordinator, the sponsor and the statistician of the study, with the help of the Independent Data Monitoring Committee (IDMC). Three conditions should be fulfilled to decide to continue the trial : -Good inclusion rates (88 patients in less than 18 months) -Completion of full treatment over 60 % in the experimental arm (Folfox) -Endoscopic complete response rate of Folfox arm equal or superior to the endoscopic response rate of 5FU/Cisplatin arm.

8. Methods (3) Main inclusion criteria Patients with histologically proven adenocarcinoma, squamous cell or adenosquamous carcinoma of the esophagus Patients unfit for surgery or locally advanced esophageal carcinoma (disease status : any T, N0 or N1, M0 or M1a) No prior treatment for esophageal cancer (surgery, chemo- or radiotherapy) Age  18 years and ECOG performance status  2 Adequate bone marrow reserve, normal renal and liver functions Patients receiving sufficient calorific intake (> 1000 Kcal/m²/day) No tracheo-esophageal fistula or invasion of the tracheo-bronchial tree No arteritis, angor, or myocardial infarction (< 6 months)

9. Treatment (1) Radiotherapy : identical in both arms Rx > 6 MV, 3 to 4 beams, 50 Gy (at intersection of all fields), 2 Gy per fraction, 5 fractions per week, all fields used every day, maximum dose to spinal cord 40 Gy. In case of lymph nodes irradiation, total dose 50 Gy. Target volume : GTV = visible tumor CTV = no automatic expansion but limits correspond to mediastinal interfaces (vertebral wall, aortic wall, lung, cardiac wall, vessels walls…) PTV = CTV + 1 cm in all directions

10. Treatment (2) Chemotherapy in Folfox arm : six bi-weekly cycles of FOLFOX 4, the first 3 cycles starting on D1, D15 and D29 concomitant with 5 weeks’ radiotherapy; On day 1, Oxaliplatin 85mg/m², on D1 and D2: leucovorin 200mg/m², 5-FU bolus 400mg/m²/d + 600mg/m²/d continuous IV infusion (1) Giovannini M et al.; J Clin. Oncol. 2004; 22 (14S) : 324 s; Abstract 4044 Arm Folfox 4-RT (50Gy) (1) CRT d29-30 RT d31-33 d36-40 RT Cycle 3 d22-26 RTCRT d15-16 RT d17-19 Cycle 2 CRT d1-2 RT d3-5 d8-12 RT Cycle 1 CT d71-72 Cycle 6 CT d57-58 Cycle 5 CT d43-44 Cycle 4 Tumor assessment W 1W 3W 5 W 7W 9W 11 W 15

11. Treatment (3) Chemotherapy in 5FU-Cisplatin arm : two cycles of 5-FU/Cisplatin on week 1 and 5 of radiotherapy and two cycles of chemotherapy with 5-FU / Cisplatin on weeks 8 and 11; On D1, Cisplatin 75 mg/m² and from day 1 to 4, 5-FU 1000 mg/m²/day (2) Herskovic A et al, N Engl J Med 1992;326:1593-8 Arm 5FU-Cisplatin-RT (50Gy) (2) d34-49 RTCRT d29-32 RT d33 Cycle 2 d22-26 RT d15-19 CRT d1-4 RT d5 d8-12 RT Cycle 1 CT d71-74 Cycle 4 CT d50-53 Cycle 3 W 1 Tumor assessment W 8 W 11 W 15 W 5

12. Results – Patients characteristics (1) FOLFOX regimen n = 53 No. of patients (%) 5FU-Cisplatin regimen n = 44 No. of patients (%) Sex ratio M / F45 (85%) / 8 (15%)38 (86%) / 6 (14%) Median age (years) [range] 59 [39-81]58 [41-80] ECOG PS 0 PS 1 PS 2 36 (68%) 17 (32%) - 24 (54.5%) 18 (41%) 2 (4.5%) Weight loss within 3 months (before study entry) Grade 1 Grade 2 16 (30 %) 7 (13 %) 11 (25 %) 14 (32 %) Inclusion period from october 2004 to december 2005 (14 months)

13. Results – Patients characteristics (2) FOLFOX regimen n = 53 No. of patients (%) 5FU-Cisplatin regimen n = 44 No. of patients (%) Histology Adenocarcinoma Squamous cell carcinoma 11 (21%) 42 (79%) 6 (14%) 38 (86%) Stage M1a4 (7%)3 (7%) TNM Stage I II III IV A - 19 (36%) 30 (57%) 4 (7%) - 17 (39%) 24 (54%) 3 (7%)

14. Results - Feasibility FOLFOX regimen n = 53 No. of patients (%) 5FU-Cisplatin regimen n = 44 No. of patients (%) No treatment (a) Complete radiotherapy (b) Planned concomitant CT (c) => First sequence of treatment 1 (2%) 51 (96%) 47 (89%) 1 (2%) 42 (93%) 41 (93%) Complete planned treatment (d) 40 (75%)31 (70%) (a) one for myocardic ischemia (Folfox arm) and one for physical health degradation before 1 st cycle (5FU-Cisplatin arm) (b) Defined as at least 45 Gy (c) 3 cycles in Folfox arm and 2 cycles in 5FU-Cisplatin arm (d) 1st sequence of RCT and complete sequence of chemotherapy (CT) alone

15. Safety FOLFOX regimen n = 52 No. of patients (%) 5FU-Cisplatin regimen n = 43 No. of patients (%) Main Grade 3-4 hematological toxicities Neutropenia Febrile neutropenia Neutropenic infection Thrombopenia 13 (25%) 4 ( 8%) 3 ( 6%) 5 (10%) 9 (21%) 2 (5%) 5 (12%) Main Grade 3-4 non-hematological toxicities Dysphagia Oesophagitis Asthenia / Fatigue 9 (17%) 4 (8%) 6 (12%) 7 (16%) 6 (14%) Peripheral neurotoxicity Grade 1 Grade 2 Grade 3 29 (56%) 8 (15%) - 3 (7%) 1 (2%) - Toxicity assessed using NCI-CTC Version 3.0

16. Efficacy Results (1) – Endoscopic Complete Response Rate (IDMC assessment) FOLFOX regimen n = 47 No. of patients (%) 5FU-Cisplatin regimen n = 40 No. of patients (%) Complete Response21 (45%) 95% exact CI : 30% to 60% 12 (30%) 95% exact CI : 17% to 46% No Complete Response18 (38%)22 (55%) Not Evaluable (Stenosis)2 (4%)2 (5%) Not Assessed * 6 (13%) 4 (10%) * Patients withdrawn from the study for AE before evaluation : in Folfox arm, 6 patients, including 4 deaths not-related to treatment; in 5FU-Cisplatin arm, 4 patients, including 3 deaths (2 treatment-related and 1 not-related to treatment). For each arm, a tumor assessment was performed during the week 15 and then every 6 months until progression.

17. Efficacy results (2) FOLFOX regimen n = 53 No. of patients (%) 5FU-Cisplatin regimen n = 44 No. of patients (%) Median [Confidence Interval 95%] Event-Free Survival (EFS) 14.3 mth [8.6-16.1]7.8 mth [5.3-12.1] Time to Progression (TTP)15.2 mth [9.9-18.9] 9.4 mth [7.2- ongoing ] Overall Survival (OS)22.7 mth [7.1 - ongoing ] 14.9 mth [9.6-19.0 ]

18. Efficacy Results (3) - Time to Progression Folfox arm 15.2 mth [9.9-18.9] 5-FU / Cisplatin arm 9.4 mth [7.2-ongoing] 5-FU / Cisplatin armFolfox arm

19. Efficacy Results (4) – Overall Survival Folfox arm5-FU / Cisplatin arm Folfox arm 22.7 mth [7.1-ongoing] 5-FU / Cisplatin arm 14.9 mth [9.6-19.0]

20. Conclusion Definitive radiochemotherapy with FOLFOX 4 is feasible with manageable toxicity as first-line treatment of patients with inoperable esophageal cancer. Definitive radiochemotherapy with FOLFOX 4 provides a high complete endoscopic response rate, a promising event-free and overall survival. Based on these results, this study will continue as a phase III trial.

21. Acknowledgments To all principal investigators of active centers Pr. CONROYCentre Alexis Vautrin, NANCYDr. ARTRUClinique Saint-Jean, LYON Dr. BOIGEInstitut Gustave Roussy, VILLEJUIFPr. MICHELCHU Ch. Nicolle, ROUEN Dr. FRANCOISCentre Antoine Lacassagne, NICEPr. BOUCHECHU R. Debré, REIMS Pr. YCHOUCentre Val d’Aurelle, MONTPELLIERDr. MINEURInstitut Ste Catherine, AVIGNON Pr. ADENISCentre Oscar Lambret, LILLEDr. SMITHCHU St André, BORDEAUX Dr. VAUTRONInstitut Bergonié, BORDEAUXDr. CARRIECentre Léon Bérard, LYON Pr. DOUILLARDCentre René Gauducheau, NANTESPr. TUBIANACHU, LIMOGES Dr. ETIENNEClinique Armoricaine, ST BRIEUCDr. CAROLI-BOSCCHU de l’Archet, NICE Pr. RAOULCentre Eugène Marquis, RENNESPr. HOUSSETHEGP, PARIS Dr. JACOBCentre François Baclesse, CAENDr. ROULLETCHU La Milétrie, POITIERS Dr. RIVESInstitut Claudius Regaud, TOULOUSE To IDMC’s Members Pr. BLEIBERG and Dr. MARTIN Oncologists ; Pr. ESCHWEGE and Pr. LABAT; Radiation Oncologists; Pr. PIEDBOIS statistician Oncologist.