BackgroundTwo randomised studies were performed with trimetrexate (TMTX) as a biochemical modulator of 5-fluorouracil (5-FU)/leucovorin (LV) in advanced colorectal cancer (ACC), one in Europe and one in the United States. Both studies were similarly designed to detect a statistically significant difference in progression-free survival (PFS). Overall survival (OS), however, was later adopted as the primary outcome measure for approvability of agents for first-line treatment of ACC. Therefore, an integrated analysis of survival data from the European and USA trials was performed to detect a clinically relevant difference in survival. Patients and methodsThe experimental arm was identical in both studies and consisted of TMTX 110 mg/m2 intravenously (i.v.) followed 24 h later by i.v. LV 200 mg/m2/5-FU 500 mg/m2 plus oral LV rescue. The 5-FU dose in the control arm was 600 mg/m2 in the European study and 500 mg/m2 in the USA study, and the USA study was placebo-controlled. Treatment was given weekly for 6 weeks every 8 weeks. ResultsA total of 746 patients were analysed. Median OS was 13.0 months for 5-FU/LV and 14.6 months for TMTX/5-FU/LV (P = 0.15; Wilcoxon). Median PFS was 4.4 months and 5.4 months, respectively (P = 0.07; Wilcoxon). ConclusionsThe addition of TMTX to a weekly regimen of 5-FU/LV does not improve the outcome for patients with ACC.