Retrospective analysis of the effect of CAPOX and mFOLFOX6 dose intensity on survival in colorectal patients in the adjuvant setting.

ABSTRACT

BACKGROUND: Despite lack of a true comparative study, the folfox (5-fluorouracil-leucovorin-oxaliplatin) and capox (capecitabine-oxaliplatin) regimens are believed to be similar in their efficacy and tolerability in the treatment of stage iii colorectal cancer. However, that belief has been disputed, because real-life data suggest that the capox regimen is more toxic, leading to more frequent reductions in the delivered dose intensity-thus raising questions about the effect of dose intensity on clinical outcomes. METHODS: A retrospective data review for two Canadian institutions, the Segal Cancer Centre and the Tom Baker Cancer Centre, considered patients diagnosed with stage iii colorectal cancer during 2006-2013. Primary endpoints were dose intensity and toxicity, with a secondary endpoint of disease-free survival. RESULTS: The study enrolled 180 eligible patients (80 at the Segal Cancer Centre, 100 at the Tom Baker Cancer Centre). Of those 180 patients, 75 received capox, and 105 received mfolfox6. In the capox group, a significant dose reduction was identified for capecitabine compared with 5-fluorouracil in mfolfox6 group (p = 0.0014). Similarly, a significant dose reduction was observed for oxaliplatin in mfolfox6 compared with oxaliplatin in capox (p = 0.0001). Compared with the patients receiving capox, those receiving mfolfox6 were twice as likely to experience a treatment delay of more than 1 cycle-length (p = 0.03855). Toxicity was more frequent in patients receiving mfolfox6 (nausea 30% vs. 18%; diarrhea 47% vs. 24%; peripheral sensory neuropathy 32% vs. 3%). At a median follow-up of 40 months, preliminary data showed no difference in disease-free survival (p = 0.598). Pooled data from both institutions were also separately analyzed, and no significant differences were found. CONCLUSIONS: Our results support the use of capox despite a lack of head-to-head randomized trial data.