Summary The combination of 5-fluorouracil (5-FU) and interferon-alpha (IFN-a) has reported activity in the treatment of advanced colorectal carcinoma. Laboratory studies of IFN-n suggest that this agent may offer theoretical advantages over IFN-a in combination with 5-FU. A total of 27 patients with advanced or recurrent colorectal carcinoma were treated in a non-randomized open phase 11 study with a combination of 5-fluorouracil (750 mg m-1 daily for 5 days as a continuous intravenous (i.v.) infusion followed, from day 15, by i.v. bolus 750 mg m-2 every 7 days) and recombinant interferon-f [r-hlFN-P-1 a; 9 MIU (total dose) by subcutaneous injection from day 1 on every Monday, Wednesday and Friday throughout the treatment period]. Toxicity was less than that seen with this schedule of 5-FU in combination with IFN-a. Among 21 evaluable patients, four objective responses were seen. Recombinant human interferon-beta-1 a in combination with 5-FU is an acceptable regimen in terms of toxicity. However, the study did not demonstrate a superior response rate when compared with previous reports of treatment with 5-FU alone or in combination with IFN-a.Keywords: interferon-beta; 5-fluorouracil; chemotherapy; colorectal carcinoma Five year survival from carcinoma of the colon and rectum is less than 40%. Although 5-fluorouracil has been the mainstay of palliative systemic therapy for advanced colorectal carcinoma for over 30 years, it is not curative in patients with metastatic disease. This agent gives objective responses in less than 25% of patients when used as a single agent, and there has been much interest in the potential for modulating its effects. Several studies indicate that 5-FU and IFN-a act in synergy to inhibit the growth of tumour cell lines in vitro (Wadler and Schwartz, 1990), and in some clinical studies this combination results in response rates (35-62%) higher than that predicted for 5-FU alone (Pazdur et al, 1990; Wadler and Wiernick, 1990;Wadler et al, 1991). However, this activity has not been confirmed in other trials (Corfu-A Study Group, 1995;Hill et al, 1995) Although IFN-,B has only 30% homology with interferon-a (de Grado et al, 1982), it binds with greater affinity to certain subclasses of interferon receptors (Ruzicka et al, 1987). In laboratory studies, r-hIFN-P-la exhibits greater antiproliferative activity than IFN-a against some tumour cell lines (Borden et al, 1982) and, against colorectal carcinoma cell lines, it demonstrates both direct antiproliferative activity and synergy in combination with 5-FU (Wong et al, 1989;Kase et al, 1993 (Lillis et al, 1987;Triozzi et al, 1987). This is similar to the single-agent activity of IFN-a, which has been reported to produce three responses among 66 patients (Kemeny and Younes, 1992). There is, therefore, a significant rationale for the testing of IFN-P in combination with 5-FU in this patient group.A phase II study examining the combination of r-hIFN-p-1a and 5-FU was designed modelled on the 5-FU/IFN-a regimen of Wadler et al (1990). In phase ...