Adjusting for differential proportions of second-line treatment in cancer clinical trials. Part II: an application in a clinical trial of unresectable non-small-cell lung cancer.

The CPT-11 trial was conducted to confirm the superiority of CDDP+CPT-11 (CPT-P) arm to CDDP+VDS (VDS-P) arm with regard to survival for patients with previously untreated advanced (stage IIIB/IV) non-small-cell lung cancer. However, the CPT-P arm did not show the survival benefit for patients with stage IIIB. The proportion of patients with stage IIIB treated by active radiation given as second-line treatment (second-line radiotherapy) after initial (first-line) treatment was 45 per cent in the CPT-P arm while 61 per cent in the VDS-P arm. This showed that not proving survival advantage in the CPT-P arm might be due to the benefits of second-line radiotherapy. We apply two causal models appeared in the companion paper to test and estimate the survival differences that would have been observed if all the patients had been received the same second-line radiotherapy. Based on several assumptions, we inferred that the CPT-P arm still would not have showed survival advantage if all the patients had been given identical second-line radiotherapy.