Results of an independent oncology review board of pivotal clinical trials of gemcitabine in non-small cell lung cancer.

Response rates reported in early phase II clinical trials are often not reproduced in subsequent larger or phase III studies. Independent review of claimed partial or complete responders to gemcitabine was undertaken in four pivotal, open-label phase II studies of advanced, non-small cell lung cancer (NSCLC) to provide accurate, consistent, reproducible response rates. Patients were chemonaive and had stage III or IV NSCLC. In three trials, gemcitabine (800 and 1250 mg/m2) was administered once-weekly for 3 weeks followed by a rest week. In the fourth, gemcitabine (90 mg/m2) was given twice-weekly for 3 weeks in every 4 weeks. The primary endpoint was response rate. Of the 374 evaluable patients, 114 (30%) were claimed as responders. Independent review reduced this to 79 (21%). The response range was reduced from 25-35 to 20-23% after validation; 95% confidence intervals did not overlap. Consistent application of response criteria by an independent panel significantly reduced response rates but produced greater consistency and reproducibility. These results confirm that gemcitabine is active against NSCLC. Subsequent larger-scale studies have produced comparable response rates, vindicating the use of independent review. Independent review is recommended for all trials using response rate as a primary endpoint.