Disease control and objective responsive rates in randomized phase II trials evaluating non-first-line chemotherapy for non-small cell lung cancer: a systematic review of 74 trials.

Although objective response rate and disease control rate are commonly used as primary endpoints of lung cancer trials, it remains unclear whether objective response rate and disease control rate correctly reflect the overall survival in a non-small cell lung cancer phase II trial evaluating a non-first-line chemotherapy. Objective response rate might be easily affected by chance because the small number of patients in each trial achieved complete or partial response in the phase II non-first-line setting. This study was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (UMIN000040412). Four databases were searched for eligible trials. A Spearman's rank correlation with hazard ratio of overall survival was calculated each for odds ratio of objective response rate, difference of objective response rate (%), odds ratio of disease control rate, and difference of disease control rate (%). Of 74 eligible trials, 73 reported objective response rate and 68 reported disease control rates. Nine (12%) trials included patients with driver mutation status. Thirteen (18%) and two (3%) RCTs specifically included adenocarcinoma/non-squamous and squamous subtype of non-small cell lung cancer, respectively. The Eastern Cooperative Oncology Group performance status 0-2 (N=41, 55%) and the performance status 0-1 (N=25, 34%) were frequently used performance status criteria. The median number of patients in the two arms was 116 (interquartile range, 82-159). The correlation between trial-level odds ratio of objective response rate and hazard ratio of overall survival was weak (r=-0.29, 95% CI: -0.49 to -0.05, P=0.014). An exploratory subgroup analysis suggested that fewer responders were associated with poorer correlation. Odds ratio of disease control survival (r=-0.53, 95% CI: -0.68 to -0.32, P<0.001) had moderate rank correlations with hazard ratio of overall survival. Instead of objective response rate, disease control rate should be used as the primary endpoint in a randomized phase II trial evaluating non-first-line chemotherapy for non-small cell lung cancer.