The study design includes three parallel phase II studies, all with the same rules for patient selection and outcome assessment. In the two studies of cisplatin-based doublets, a dose-finding phase is planned, with subsequent cohorts of 6 patients treated with increasing doses of cisplatin (50-60-70 mg/m²). Following this phase, patients will be assigned to the three studies through a randomization procedure, to reduce selection biases.
Arm A: gemcitabine:
Gemcitabine 1200 mg/m2 intravenously in 2 hours (10 mg/m²/minute) on days 1 \& 8 of each cycle, every 21 days, for a maximum of 6 cycles.
Arm B: cisplatin + vinorelbine:
Vinorelbine 25 mg/m² on days 1 \& 8, followed by Cisplatin (50 or 60 or 70 mg/m²) on day 1 of each cycle, every 21 days, for a maximum of 6 cycles.
Arm C: cisplatin + gemcitabine:
Gemcitabine 1000 mg/m2,, on days 1 \& 8, followed by Cisplatin (50 or 60 or 70 mg/m²) on day 1 of each cycle, every 21 days, for a maximum of 6 cycles.