Randomized phase III trial of bleomycin, vindesine, mitomycin-C, and cisplatin (BEMP) versus cisplatin (P) in disseminated squamous-cell carcinoma of the uterine cervix: an EORTC Gynecological Cancer Cooperative Group study

PURPOSE: Three previous mitomycin-cisplatin-based chemotherapy trials conducted within the EORTC Gynecological Cancer Cooperative Group (GCCG) in patients with disseminated squamous-cell carcinoma of the uterine cervix (SCCUC) suggested that with such regimens a higher overall response rate and a higher complete response rate could be obtained compared to what might have been expected from cisplatin alone. In that respect the combination of bleomycin, vindesine (Eldesine), mitomycin C and cisplatin (BEMP) was the most promising. In the present study BEMP has been compared with the best single agent, cisplatin (P) in the expectation that improved response rates might translate into a better survival. PATIENTS AND METHODS: Eligible patients were those with SCCUC and disseminated measurable disease outside previously irradiated areas, aged < or = 75 years, with a WHO performance status < or = 2 and adequate bone marrow, renal, hepatic and pulmonary function, who gave consent according to regulations followed in individual institutions. Patients were randomized to BEMP: E 3 mg/m2 day 1, P 50 mg/m2 day 1, B 15 mg (24-hour infusion) day 2-4 and M 8 mg/m2 (at alternate cycles), or P 50 mg/m2. The first four cycles were given every 3 weeks (induction phase). Subsequent cycles were given every four weeks (maintenance phase), during which B was deleted from BEMP (MEP). Patients failing on P could be treated with BEM. Of the 287 patients entered, 235 were eligible and 201 evaluable for response. RESULTS: BEMP induced a significantly higher response rate than P (42% vs. 25%, P = 0.006). There was no difference in complete response rate (11% vs. 7%). BEMP was significantly more toxic than P (+/- BEM), both with respect to hematologic and nonhematologic toxicities. After a median follow-up of 6.1 years, survival curves were not significantly different. Median progression-free survival and overall survival were 5.3 and 10.1 months with BEMP and 4.5 and 9.3 months with P (+/- BEM), respectively. In a multivariate analysis of prognostic factors for survival, a lower age (P = 0.003), a lower performance status (P = 0.0001) and a short (