This trial was conducted to determine whether high-dose fluorouracil (FU) given as a weekly 24-hour infusion is more active than bolus FU + leucovorin (LV), and whether high-dose infusional FU can be modulated by LV.

A total of 497 patients with previously untreated metastatic colorectal cancer were randomly assigned to receive bolus FU 425 mg/m2 intravenously + LV 20 mg/m2 on days 1 to 5 and repeated on day 28 (FU + LV), or FU 2600 mg/m2 as a 24-hour infusion alone (FU24h) or in combination with 500 mg/m2 LV (FU24h + LV)-all given weekly x6 followed by a 2-week rest period. Survival was the major study end point.

With a median follow-up of more than 3 years, survival did not differ among the treatment groups (median FU + LV, 11.1 months [95% CI, 10.2 to 15.0 months]; FU24h, 13.0 months [95% CI, 10.4 to 15.4 months]; FU24h + LV, 13.7 months [95% CI, 12.0 to 16.4 months]; P =.724). Progression-free survival (PFS) was significantly longer for FU24h + LV (median FU + LV, 4.0 months [95% CI, 3.4 to 4.9]; FU24h, 4.1 months [95% CI, 3.4 to 5.0]; FU24h + LV 5.6 months [95% CI, 4.4 to 6.7]; P =.029). The response rates in the subgroup of patients with measurable disease were 12%, 10%, and 17% for FU + LV, FU24h, and FU24h + LV, respectively (not significant). Occurrence of grade 3 and 4 diarrhea was higher in the FU24h + LV arm (22%) compared with the FU24h (6%) or FU + LV (9%) arms; however, stomatitis (11% in FU + LV v 3% in FU24h v 5% in FU24h + LV arms) and hematologic toxicity were higher in the bolus FU + LV arm. Global quality of life did not differ within the three arms.

Neither FU24h + LV nor FU24h prolong survival, relative to bolus FU + LV. Leucovorin increases PFS if added to FU24h, but increases toxicity.