Interferon-alpha does not improve the antineoplastic efficacy of high-dose infusional 5-fluorouracil plus folinic acid in advanced colorectal cancer: First results of a randomized multicenter study by the Association of Medical Oncology of the German Cancer Society (AIO)

Abstract

High-dose 5-FU given weekly as a 24-h infusion in combination with folinic acid (FA) has been associated with low toxicity and a high response rate. Interferon-alpha (IFN) either alone or in combination with FA has also improved treatment results by modulating 5-FU activity. We therefore initiated a randomized multicenter trial comparing the ability of FA or IFN to modulate infusional 5-FU. The statistical design using a sequential analysis allows us to report on the comparison of 5-FU/FA vs. 5-FU/FA/IFN while randomization of patients into 5-FU/FA vs. 5 FU/IFN continues. Chemotherapy-naive patients with advanced progressive colorectal cancer and measurable metastatic lesions were randomized to receive 5-FU 2600 mg/m2 i.v. as a 24-h infusion, combined with either FA 500 mg/m2 as a 2-h infusion (A), or IFN 3 x 10(6) U s.c. 3 x/week (B), or the combination of FA plus IFN as in arms A and B (C). Treatment arms were repeated weekly for 6 weeks followed by a 2-week rest period. These 8-week cycles were administered until tumor progression. Because of the occurrence of 2 toxic deaths among the first 17 patients treated in arm C, 5-FU was reduced to 2000 mg/m2 for all patients in arm C. Sequential analysis according to Whitehead for objective response was planned with alpha = 0.05/3 and a power of 80% (beta = 0.2) to detect a difference of > or = 25% (delta = 0.25) or equivalence of response rates. For pairwise comparison of treatment arms a minimum of 30 patients per arm and a maximum of 90 patients per arm were expected in case of equivalence or difference. An interim analysis was performed after the first 93 of 149 randomized patients were evaluable for response and toxicity (A 31 pts, B 33 pts, C 29 pts). Despite the 5-FU dose reduction in arm C, 28% of patients experienced grade 3/4 toxicity (CTC) including diarrhea, mucositis and handfoot syndrome compared to 16% in arm A and 12% in arm B (not significant). No treatment related toxic death occurred in arms A or B, but 3 patients (10%) in arm C died of diarrhea and septicemia. Among patients treated with 5-FU/FA objective tumor response occurred in 12/31 patients (39%) (21%-56%, 95% confidence interval) (3 CR, 9 PR), no change in 13/31 (42%) and PD in 6/31 (19%) patients. Eleven of 29 patients (38%) (20%-56%, 95% confidence interval) receiving 5-FU/FA/IFN achieved complete (3 patients) or partial (8 patients) remissions, 10/29 patients (34%) had stable disease and 8/29 patients (28%) tumor progression. According to the sequential analysis the rates of objective responses observed in patients treated with 5-FU/FA or 5-FU/FA/IFN were equivalent. This interim analysis allows the conclusion that infusional 5-FU plus FA/IFN is no more active than infusional 5-FU plus FA alone. However, 5-FU/FA/IFN despite 5-FU dose reduction was associated with unacceptably high toxicity, including 10% deaths. Therefore, further investigation of this regimen is not justified. The study is continued with the comparison of 5-FU/FA vs. 5-FU/IFN.