Purpose: To evaluate in patients with advanced colorectal cancer (CRC) three treatment regimens of oral capecitabine in order to select the most appropriate regimen for testing in phase III.
Patients and methods: Three capecitabine schedules were evaluated in a randomized phase II design: arm A, 1,331 mg/m(2)/d bid continuously; arm B, 2,510 mg/m(2)/d bid intermittently (2 weeks on/1 week off); and arm C, 1,657 mg/m(2)/d plus oral leucovorin 60 mg/d bid intermittently (2 weeks on/1 week off).
Results: One hundred nine patients were randomized; 39 patients were assessable for efficacy in arm A, 34 in arm B, and 35 in arm C. Patient characteristics were balanced in the arms. Confirmed tumor responses (partial response [PR] + complete response [CR]) were reported for eight patients with two CRs (21%; 95% confidence interval [CI], 9% to 36%) in arm A, eight patients with one CR (24%; 95% CI, 11% to 41%) in arm B, and eight patients with two CRs (23%; 95% CI, 10% to 40%) in arm C. Median times to progression (TTP) in arms A, B, and C were 127, 230, and 165 days, respectively. Overall, more toxicity was seen with capecitabine plus leucovorin, particularly diarrhea and hand-foot syndrome. There was no grade 3 or 4 marrow toxicity.
Conclusion: Capecitabine offers a new, effective treatment option as an oral single agent in advanced CRC. Promising overall response rates were reported for all three regimens. The addition of leucovorin to the intermittent regimen had no marked effect on tumor response or median TTP. The intermittent single-agent capecitabine schedule is proposed for phase III evaluation, based on considerations of toxicity, dose-intensity, response rate, and TTP.