Drug delivery analysis of the Canadian multicenter trial in non-small-cell lung cancer.

Abstract

PURPOSE The Canadian multicenter trial for advanced non-small-cell lung cancer (NSCLC) reported survival benefit for chemotherapy when best supportive care was compared with vindesine-cisplatin (VP) and the combination of cyclophosphamide, doxorubicin, and cisplatin (CAP). We examined received drug delivery to document dose-intensity (DI) and total dose of drugs given to various groups in this patient population. PATIENTS AND METHODS Plots of cumulatively received chemotherapy against time were used to evaluate drug delivery by regimen, major prognostic factors, and response status. RESULTS Individual CAP patients show a narrow range of received DI, with the median similar to protocol. Drug delivery analysis exposed a wide range of received DI for both drugs in the more intensive VP regimen, and the median received DI was below protocol. The median received DI for cisplatin was still higher for VP than CAP, but only during the first 8 weeks of protocol treatment (20 v 10 mg/m2/wk); thereafter, the ongoing received cisplatin DI was the same (10 mg/m2/wk). The median received DI for cisplatin in each regimen was not influenced by stage, performance status, prior weight loss, sex, or response status. VP-treated patients received a higher total dose of cisplatin than CAP patients (median, 255 mg/m2 v 112.5 mg/m2; P < .0001). Median cisplatin total dose was similar for patients with a chemotherapy response or stable disease and threefold greater than for patients with progressive disease for both regimens. Although patients with chemotherapy response and stable disease had similar survival outcomes for both CAP and VP, the VP regimen had a higher proportion of patients without progressive disease (P = .004), which resulted in an overall survival advantage (P = .01). CONCLUSION The major prognostic factors for advanced NSCLC do not exert their influence on outcome by affecting deliverable chemotherapy DI. Regimen and treatment response determined total dose. Because stable disease patients usually outnumber responding patients in advanced NSCLC trials, controlled studies should be performed that allow assessment of the impact of total received dose on outcome according to response status.