SEQUENTIAL METHOTREXATE, 5-FLUOROURACIL (5-FU), AND HIGH-DOSE LEUCOVORIN VERSUS 5-FU AND HIGH-DOSE LEUCOVORIN VERSUS 5-FU ALONE FOR ADVANCED COLORECTAL-CANCER - A MULTIINSTITUTIONAL RANDOMIZED TRIAL

Background. The primary objective of this study was to compare the single-biochemical modulation of 5-fluorouracil (5-FU) and leucovorin with that of the double-biochemical modulation of methotrexate and leucovorin. Because of the Martin et al. study in which an experimental model showed similar effects of 5-FU at maximum tolerated doses to the modulation with leucovorin at standard doses, a third treatment arm of 5-FU alone was also studied. Methods. A randomized trial was performed using a 500-mg/m(2) intravenous (i.v,) 1-hour infusion of methotrexate, and 12 hours later, a 600-mg/m(2) i.v. bolus of 5-FU plus a 200-mg/m(2) i.v. 1-hour infusion of leucovorin (MFL) every 2 weeks versus 5-FU plus leucovorin at an equal dose and schedule (FL), versus a 1200-mg/m(2) i.v dose of 5-FU every 2 weeks. Of 186 patients included in the study, 178 were evaluable. Results. In a preliminary analysis with 94 evaluable patients, two significant statistical differences were shown. First, the toxicity rate of the 5-FU-alone (F) treatment arm was higher than that of the other arms (MFL vs. F, P = 0.0002; FL vs. F, P = 0.00001). Second, the median survival was worse in the F treatment arm with a rate of 12.6 months for the MFL and FL arms and 7.5 months for the F arm (P < 0.05). Considering these results, the F treatment arm was discontinued. The final results included 70 evaluable patients for MFL and 74 patients for FL. No difference was found in the distribution of prognostic factors. The response rates were 25.7% for MFL (95% CI, 16-37.5) and 14.8% for FL (95% CI, 7.6-25), (P = 0.1). The median survival was 14.3 months for patients treated with MFL and 12.3 months for those treated with FL. The hematologic toxicity was mild, with no grade 3/4 leukopenia in either treatment arm. The major nonhematologic toxicity in the MFL and FL treatment arms was ocular; nongrade 3/4 diarrhea also was observed. Conclusions. The results of MFL double-biochemical modulation failed to show a significant statistical difference from that of single-biochemical modulation for this dose and schedule.