Search

Aim: Six months FOLFOX4 or XELOX is the standard adjuvant chemotherapy in radically resected stage III colon cancer, however a shorter duration of therapy could be equally efficacious. Methods: This is an Italian, multicenter, non-inferiority phase III trial randomizing patients (pts) with stage III and high risk stage II colon cancer to receive 3 (arm 3) vs 6 (arm 6) months of FOLFOX4 or XELOX. Primary end-point is relapse free survival (RFS), and the arm 3 will be considered non-inferior if the superior margin of the 95% CI of the HR is Results: From June 2007 to March 2013, 3,759 patients were accrued and 3,738 (1860 in arm 3 and 1878 in arm 6) included in the intent-to-treat (ITT) population, since 21 pts were excluded due to major violations. The proportion of FOLFOX4 vs XELOX was 64% vs 36% in both arms. Among the ITT population, demographic factors (including gender and age) were balanced in the two arms. As expected, compliance was higher in arm 3 than in arm 6: treatment completion rate without any modification of planned dose and timing was 35% in arm 3 and 12% in arm 6, while approximately 55% pts in both arms completed treatment with modifications of dose and/or timing, resulting in 90% vs 67% pts receiving all the planned cycles. Therapy was permanently interrupted due to toxicity in 5% of pts in arm 3 and 17% in arm 6. The overall incidence of grade 3-4 toxicity (including grade 2 oxaliplatin neurotoxicity) events was 37% in arm 3 and 60% in arm 6. Grade 2+ neuropathy was more frequent in arm 6 than in arm 3: 32 % vs 9%, respectively. With a median follow-up of 30 months, 476 pts relapsed, 195 died and 520 relapsed and/or died. Conclusions: TOSCA is the first trial completing the accrual among the international initiative of comparing 3 vs 6 months of adjuvant treatment in colon cancer (IDEA). In general, reported toxicity was low, but, as expected, significantly higher in the 6-month arm compared to that of the 3-month arm; accordingly, compliance was lower in the standard 6-month arm. Data for efficacy analysis are premature. Disclosure: All authors have declared no conflicts of interest.