TOSCA.

Nearly 4000 patients with high risk stage II (T4, high grade, obstruction/perforation, LVSI, or <12 nodes examined) or stage III colon cancer were randomized to 3 or 6 months of investigator’s choice adjuvant FOLFOX (64%) or CAPOX (36%). The primary endpoint was relapse-free survival (RFS), where a difference of less than 4% was deemed clinically noninferior. While the 3-year and 5-year RFS rates were both within 4%, an unexpectedly (is it really, though?) low number of events resulted in a confidence interval that surpassed the pre-specified limit. Just like in the pooled analysis, 3 months of CAPOX was similar to 6 months, while 3 months of FOLFOX was inferior to 6 months. Completely opposite to the pooled analysis, lower (not higher) risk patients derived the most benefit from longer chemo. What? TBL: TOSCA failed to prove noninferiority of 3 versus 6 months of chemo, with subset analyses contradictory to the pooled analysis. | Sobrero, J Clin Oncol 2018