Schedule dependency of 21-day oral versus 3-day intravenous etoposide in combination with intravenous cisplatin in extensive-stage small-cell lung cancer: a randomized phase III study of the Cancer and Leukemia Group B.

PURPOSE This was a randomized phase III study to test the schedule dependency of etoposide given as a conventional 3-day intravenous (IV) regimen versus a prolonged 21-day oral regimen for extensive-stage small-cell lung cancer (SCLC). Both regimens contained IV cisplatin. The objectives were to compare survival (primary end point) and to establish response rates, failure-free survival, and toxicity (secondary end points). PATIENTS AND METHODS Patients with untreated measurable or assessable disease and normal organ function were eligible. Randomization was stratified according to performance status 0 versus 1 or 2. Treatment consisted of etoposide 130 mg/m2/d IV for 3 days and cisplatin 25 mg/m2/d IV for 3 days every 21 days for eight courses (schedule 1) versus etoposide 50 mg/m2/d orally for 21 days and cisplatin 33 mg/m2/d IV for 3 days every 28 days for six courses (schedule 2). In 1990, bioavailability of oral etoposide was assumed to be 50%, and the study was designed to deliver the same total doses of etoposide and cisplatin on both regimens over 24 weeks without the use of growth factors. RESULTS Between December 1990 and October 1993, 306 eligible patients were entered. Of these, 69% were male and 66% were > or = 60 years of age; 21% had a performance status of 0, 47% a performance status of 1, and 32% a performance status of 2; 156 were randomized to receive schedule 1 and 150 to receive schedule 2. Overall median survival estimates were 9.5 and 9.9 months (difference not significant) for schedule 1 and schedule 2, respectively. The 95% confidence interval (CI) for overall survival, 8 to 11 months, was the same for both schedules, with 126 and 117 deaths on schedule 1 and 2, respectively. Both schedules also resulted in the same median failure-free survival estimate of 7 months (95% CI, 6 to 8 months on either schedule). Complete and partial responses were observed in 15% and 42% of patients on schedule 1 and 14% and 47% on schedule 2, respectively. The overall maximal hematologic toxicities grade 3 and 4 for leukocytes, neutrophils, platelets, and hemoglobin were, respectively, as follows: schedule 1, 62%, 85%, 32%, and 32%; schedule 2, 83%, 83%, 52%, and 53%. Lethal toxicity due to neutropenia and infection occurred in 4% of patients on schedule 1 and 10% on schedule 2 (difference not statistically significant). CONCLUSION The two schedules of etoposide in combination with cisplatin did not result in differences in treatment outcome with respect to tumor response and survival. However, a significantly greater rate of severe or life-threatening hematologic toxicity was noted on the 21-day oral etoposide treatment schedule.