Interim efficacy analysis of the French Intergroup R98 trial comparing 5FU-leucovorin alone or with irinotecan (CPT-11) in resected stages II-III rectal cancers

Abstract

4015 Background: The R98 trial tests the hypothesis that irinotecan (CPT-11) combined with 5FU/LV is superior to 5FU/LV alone to prevent recurrence or death in resected stages II-III rectal cancers. Methods: 600 pts were planned to be randomized between 5FU/LV (control arm) or 5FU/LV + CPT-11. Recruitment was slow, with only 357 pts included in the study between 03/1999 and 12/2005 (178 in control arm and 179 in CPT-11) at which time the IDMC recommended closure of accrual. The primary endpoint was disease free survival (DFS), relevant events being death (whatever its cause), local or distant recurrence, and any new primary cancer. The trial was stratified by choice of control arm: either a Mayo-Clinic regimen (A: LV 20 mg/m2, 5FU 425 mg/m2 bolus days 1–5, repeated at d29, d57, d92, d127 and d162) or a LV5FU2 regimen (A’: LV 200 mg/m2 over 2-hour, 5FU 400 mg/m2 bolus and 600 mg/m2 22-hour infusion, d1–2, q2w for 12 cycles), depending on centre decision. The experimental arm (B) consisted of LV5FU2 + CPT-11 180 mg/m2 d1 of every cycle. Results: Safety analysis was reported last year (P. Piedbois et al. Proc. ASCO 2008). We present here a first interim analysis of efficacy on half the numbers of events planned per protocol. Median follow-up is 55 months. Pretreatment characteristics were well balanced between the groups; median age 62 years, stage II in 32% of pts. 68% of pts received preoperative RT, and 80% had sphincter conservation. There was a trend in favor of 5FU/LV + CPT-11 in terms of DFS (77 vs 62 events in control and CPT-11 arms, hazard ratio=0.75, p=0.089) with similar effect in the Mayo-Clinic stratum (31 vs 25 events, hazard ratio=0.77) and in the LV5FU2 stratum (46 vs 37 events, hazard ratio=0.74). Conclusions: This interim efficacy analysis shows a trend in favor of 5FU/LV + CPT-11. If performed when at least 220 events are observed (projected in 2010), the next efficacy analysis would have a conditional power of about 60% to detect a true benefit of 5FU/LV + CPT-11 if the treatment effect remained the same as observed today. Study supported by Pfizer, France. [Table: see text]