Purpose: We compared the activity of vinorelbine (VRB) and vindesine (VDS) in a randomized crossover study in patients with previously untreated stages IIIB or IV non-small-cell lung cancer (NSCLC).
Patients and methods: Two hundred four patients were assessable for response and toxicity. VRB was administered at a dose of 25 mg/m2 weekly and VDS at a dose of 3 mg/m2 weekly. Patients who failed to respond after 4 cycles of initial monotherapy were switched to a combination chemotherapy (VRB-->VDS + cisplatin (P) or VDS-->VRB + P).
Results: Objective response was observed in 31.1% of patients in the VRB arm versus 8.9% of those in the VDS arm (P = 0.0002). The median duration of response to VRB was 18.5+ weeks (range, 7.9 to 107.5+ weeks) compared with 11.7+ weeks (range, 6.0 to 35.0+ weeks) for VDS. Of the 69 patients who failed to respond to initial monotherapy, 33 in the VRB group who subsequently received VDS + P did not respond and 13 (26.5%) of 49 initially on VDS who received subsequent VRB + P responded. The rates of grades 3 and 4 leukopenia were similar in the two monotherapy arms (VRB, 55.3% vs. VDS, 48.5%). However, grade 3 anemia was more frequent in the patients on VRB than in those on VDS. The incidence of peripheral neurotoxicity was significantly higher with VDS than with VRB (P = 0.002), but VRB induced a slightly higher rate of local cutaneous reaction than VDS (P = 0.012). With the combination of cisplatin and these vinca alkaloids, peripheral neurotoxicity was less frequent in the VRB group than in the VDS group.
Conclusion: Our results demonstrate that VRB yields a higher response rate than VDS in stage IIIB or IV NSCLC, with the same extent of toxicity in terms of leukocytopenia. The peripheral neurotoxic effects were also milder with VRB than with VDS. In second-line chemotherapy, there was a notable difference in response between the VRB + P and VDS + P regimens.