VICTOR: A phase III placebo-controlled trial of rofecoxib in colorectal cancer patients following surgical resection

Abstract

4038 Background: The cyclo-oxygenase-2 inhibitor, rofecoxib (R) was hypothesised to improve survival in cancer patients who had undergone surgery for colorectal cancer (CRC). This trial recruited from April 2002 until September 2004 when R was withdrawn over concerns about its cardiovascular safety (CVS). This report provides preliminary efficacy results. Methods: Recruited patients had undergone Ro resection of a stage II/III CRC and completion of adjuvant therapy (radiotherapy/chemotherapy/both/neither) less than 12 weeks previously. Excluded patients were those with active peptic ulceration, gastro-intestinal bleeding and those receiving long-term NSAID therapy (except low dose aspirin). 7,000 patients were planned to receive 25mg R daily or an identical placebo (P) for 2 or 5 years, with the goal of detecting a reduction in risk of death - hazard ratio (HR) 0.82. After the trial’s premature closure, a modified protocol of the post- treatment follow-up phase and revised statistical analysis plan permitted the detection of a reduction (HR=0.75) in risk of death with 87% power, type I error 0.05, with one pre-planned event-driven interim analysis. Overall survival (OS) and disease-free survival (DFS) were both measured from randomisation, with DFS defined as the time to recurrence or death from any cause. Results: 1,167 of 1,217 patients randomised to R and 1,160 0f 1,217 randomised to P received treatment with median durations of 7.4 months and 8.2 months respectively. Median follow-up was 3.0 and 3.1 years in the two arms (R vs P), with 177 vs 191 deaths and 291 vs 316 DFS events. This pre-planned Kaplan-Meier and log- rank analysis demonstrated that the R patients had slightly longer OS than the P patients, HR 0.94 (95% CI 0.77–1.16; p=0.57). Similarly DFS was slightly higher in the R patients, HR 0.91 (95% CI 0.78–1.07; p=0.25). 19 patients in each arm died without recurrence of CRC. Conclusions: In this study of short treatment duration treatment with R is unlikely to result in a substantial improvement in OS but a small protective effect against recurrence is suggested. Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Expert Testimony Other Remuneration Merck and Co. Inc Merck & Co Inc