Meta-analysis of surrogate endpoints for overall survival in extensive-stage small-cell lung cancer

Overall response rate (ORR), disease control rate (DCR), and progression-free survival (PFS) are traditionally used as surrogates for overall survival (OS). However, their validity as reliable predictors of OS in small-cell lung cancer (SCLC) remains unclear.

MEDLINE, Embase, Cochrane, and ClinicalTrials.gov were systematically searched between 2014 and 2024. Study selection followed PRISMA guidelines. Data on ORR, DCR, PFS, and OS were manually extracted from phase III randomized trials. Weighted linear regression and Pearson correlation analyses, with logarithmic transformation of odds ratios and hazard ratios, were used to assess associations between endpoints. Correlation coefficients were computed to evaluate the correlation between surrogate endpoints and OS. All tests were two-sided, with statistical significance set at P < 0.05.

The analysis comprised 23 randomized phase III trials, including 10 340 patients. In first-line studies, a strong correlation was observed between PFS and OS (r = 0.77, P < 0.001). In first-line immunotherapy studies, PFS also correlated significantly with OS (r = 0.80, P = 0.006). In contrast, ORR and DCR did not correlate with OS in either first- or second-line settings. In the second-line setting, DCR correlated with PFS (r = -0.84, P = 0.01) but not with OS. These findings were consistent across both the difference model and the odds ratio model.

PFS demonstrates strong clinical value as a surrogate endpoint for OS in the first-line SCLC setting, including with chemoimmunotherapy, and may serve as a reliable endpoint for future phase III trials. In contrast, ORR and DCR do not reliably predict long-term outcomes, underscoring the need to prioritize PFS over these measures in trial design and interpretation.