Abstract
BACKGROUND: It is controversial for the use of survival surrogate end points including response rate (RR), disease control rate (DCR), time to progression, and progression-free survival (PFS) in trials of molecular targeted agents. Our aim was to determine the correlations of these surrogates with survival in the treatment of advanced non-small-cell lung cancer (ANSCLC) with epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs), gefitinib and erlotinib. METHODS: Summary data of median survival time (MST) and surrogates from prospective trials of EGFR-TKIs in ANSCLC were identified. Patient- or trial-related characteristics were introduced as covariates. Simple and multivariate linear regression models were fitted for MST and each surrogate, respectively. And the significance of each surrogate as a survival marker was compared by calculating the area under their receiver operating characteristic (ROC) curves. RESULTS: Sixty eligible trials (9,903 patients) were enrolled. RR, DCR, and PFS were all strongly associated with MST. In their simple linear regression models, the coefficient of determination (R(2)) was 0.83 (p < 0.000001), 0.58 (p < 0.0001), and 0.70 (p < 0.0001), respectively. And in their multivariate linear regression models, the standard coefficient was 0.71 (p < 0.001), 0.40 (p < 0.001), and 0.74 (p < 0.001), respectively, while RR and PFS were the preferred survival predictors in the ROC analysis. CONCLUSION: RR or PFS may serve as an appropriate survival surrogate in the clinical trials of EGFR-TKIs for ANSCLC.