Survival advantage for carboplatin substituting cisplatin in combination with vindesine and mitomycin C for stage IIIB and IV squamous-cell bronchogenic carcinoma: a randomized phase III study

This communication represents the definitive report of a randomized phase I II study comparing cisplatin and carboplatin, in combination with vindesine and mitomycin C in stage IIIB and IV squamous-cell bronchogenic carcinoma. A total of 221 patients entered the study and were randomized into two arm s. Of these, 114 patients (109 evaluable for activity) were randomized to a rm A, receiving cisplatin 120 mg/m(2), mitomycin C 8 mg/m(2) and vindesine 3 mg/m(2) per cycle; 107 patients (101 evaluable for activity) were randomi zed to arm B receiving carboplatin 500 mg/m(2) with the same doses of mitom ycin C and vindesine per cycle. Patients with progressive disease (PD) were excluded from the study after the 2nd cycle, and those with stable disease (SID), partial response (PR) and complete response (CR) received six cycle s of chemotherapy (or less in case of early progression). Patients were str atified according to the clinical stage (IIIB vs. IV), performance status ( 0 + 1 vs. 2 + 3) and tumor histological grade (I + II vs. III). In the cisp latin arm two patients (1.9%) achieved a CR, 38 (34.9%) a PR, 45 (41.2%) a SID and 24 (22.0%) had PD; the overall response rate was 40/109 (36.8%). In the carboplatin arm five patients (5.0%) achieved a CR, 31 (30.7%) a PR, 4 0 (39.6%) a SD, and 25 (24.7%) had PD; the overall response rate was 36/101 (35.7%). No statistically significant difference in response rate was pres ent between the two arms, and the response rate was not influenced by perfo rmance status, histological grade or clinical stage. The Kaplan-Meyers curv es displayed a significant advantage both for time to progression (P=0.005) and overall survival (P=0.008) for patients in the carboplatin arm. The ad vantage for patients receiving carboplatin instead of cisplatin appeared ev ident in univariate setting for patients with a good performance status and clinical stage IV, and occurred irrespectively of tumor histological grade ; response duration and survival of responders was identical in the two arm s. Patients achieving a stable disease survived longer in the carboplatin t han in the cisplatin arm (P=0.012). Thus, substitution of cisplatin by carb oplatin in the combination chemotherapy regimen, although more hematologica lly toxic (but less emetogenic) resulted in a similar response rate, but a significantly longer time to progression and overall survival. (C) 2001 Els evier Science Ireland Ltd. All rights reserved.