Chemotherapy for advanced NSCLC. Will meta-analysis provide the answer?

Meta-analyses of published data show a modest but significant survival or response rate benefit for chemotherapy plus supportive care over best supportive care alone in non-small cell lung cancer (NSCLC), but these findings may be biased in favor of positive results because they are literature based. A meta-analysis of individual patient data from 11 published and unpublished randomized studies shows a significant benefit for cisplatin-based chemotherapy over supportive care, a small (not significant) benefit for vinca alkaloids and etoposide, and no benefit for alkylating agents. The benefit is independent of age, sex, stage, histologic type, or performance status. Single large trials would avoid the pitfalls of meta-analyses, but these have so far proved difficult to organize. Physicians need to be convinced that even a small increase in median survival in NSCLC as a result of chemotherapy may represent a considerable increase in life expectancy for individual patients.