Small cell lung cancer (SCLC): a randomized trial of cyclophosphamide, adriamycin, vincristine plus etoposide (CAV-E) or teniposide (CAV-T) as induction treatment, followed in complete responders by alpha-interferon or no treatment, as maintenance therapy.

In this phase III, double-random study, we compared CAV-E to CAV-T combination as induction treatment (1st randomization) for SCLC. Subsequently, patients achieving a complete response (CR) were randomized again (2nd randomization) to receive maintenance treatment with α-IFN or no treatment. From June 1990 to June 1992, 75 untreated patients were enrolled in this trial. After stratification according to limited disease (LD) or extensive disease (ED), patients were randomized to receive the following treatment: cyclophosphamide 1000 mg/m 2 , adriamycin 50 mg/m2, vincristine 2 mg, day 1 i. v., plus etoposide (E) 100 mg/m2 (CAV-E: arm-A) or teniposide (T) 60 mg/m2 on day 2, 3, 4 i.v., every 3 weeks (CAV-T: arm-B). LD patients after 3 cycles of treatment received chest radiotherapy and 2 further cycles, whereas ED patients received 5 consecutive cycles. Patients who achieved a CR entered the 2nd randomization receiving a-IFN (3×10 6 I.U., i.m. daily ×9 months) or no treatment. A second-line treatment with carboplatin 300 mg/m2 plus E (if T was initially used) or T (if E was initially used) was also scheduled for patients achieving less than CR to induction treatment. Preliminary results are as follows: 75 patients were randomized, 72 were eligible for survival (arm-A=37 and arm-B=35) and 60 were fully evaluable for response (arm-A =34 and arm-B=26). In patients with LD the overall response rate was 79% (CR 21%) in arm-A vs 92% (CR 50%) in arm-B. In patients with ED, the overall response rate was 80% (CR 33%) in arm-A vs 84% (CR 7%) in arm-B. At a mean observation time of about 1 year (range 1-25 months), median survival of LD patients was 15 months in arm-A and 13 months in arm-B (Chi-square=1.55; p>0.05); in ED patients survival was 10.8 months and 8 months respectively (Chi-square=2.88; p>0.05). Cumulative survival probability was identical (12 months) in all patients of both arms. Toxicity was mainly haematologic and gastrointestinal: WHO grade 3-4 myelosuppression and vomiting were observed in 20% and 11% respectively, of cycles delivered in arm-A, compared to 19% and 8%, respectively, of cycles in arm-B. Two septic deaths occurred with CAV-T, while 1 patient discontinued treatment due to persistent myelosuppression with CAV-E. After the first and second-line treatment 20 patients showed a CR. Five of them refused further treatment, while 15 accepted to enter the 2nd randomization. At the time of analysis, 9 patients were allocated to the a-IFN arm and 6 to the control arm. Although the study is still running, the 2nd randomization accrual appears inadequate to achieve the second objective within a reasonable time