Comparison of real-world overall survival between atezolizumab- and durvalumab-containing first-line induction and maintenance regimens in extensive-stage small cell lung cancer.

Abstract

8093 Background: Both atezolizumab (atezo) and durvalumab (durva) are preferred first-line (1L) systemic treatment options for extensive-stage small cell lung cancer (ES-SCLC) with similar efficacy in the IMpower133 and CASPIAN phase 3 trials. There is limited evidence comparing their survival outcomes in the real-world (rw) setting. This study described rw overall survival (rwOS) in patients (pts) with ES-SCLC treated in 1L and 1L maintenance (1Lm) settings and also compared rwOS stratified by cohort and treatment with 1L atezo vs durva in combination with platinum-based chemotherapy (chemo) followed by atezo vs durva as 1Lm immunotherapy (IO). Methods: Using the US Flatiron Health Research Database, this retrospective noninferiority study included adults with ES-SCLC across 2 cohorts. The 1L cohort included pts treated with atezo or durva + chemo between Apr 1, 2020, and Sept 30, 2025. The 1Lm cohort included a subset of pts from the 1L cohort who were subsequently treated with atezo or durva monotherapy. The index date was the initiation of atezo or durva in the 1L or 1Lm setting in each cohort. rwOS hazard ratios (HRs) and 95% CIs comparing durva with atezo as reference in the 1L and 1Lm settings were calculated using Cox proportional hazard models, adjusting for prognostic covariates. Based on prior literature, the noninferiority margin was set at HR = 1.25. Results: A total of 2063 and 1122 pts were included in the overall 1L and 1Lm cohorts, respectively. In both cohorts, median age was 68 years, ≥20% of pts had an ECOG PS ≥2, and ≥19% had brain metastases at baseline. The 1L cohort had 1614 (78%) pts on atezo + chemo and 449 (22%) on durva + chemo. The 1Lm cohort had 865 (77%) pts on atezo and 257 (23%) on durva. Pt characteristics were generally similar between those treated with atezo vs durva in 1L or 1Lm settings. In the 1L cohort, median OS (95% CI) from the index date was 8.7 (8.2, 9.3) months for atezo and 8.5 (7.9, 9.5) months for durva. In the 1Lm cohort, median OS (95% CI) from index date (start of maintenance) was 9.3 (8.2, 10.2) and 9.2 (7.5, 10.8) months for atezo and durva. In the 1L cohort, the adjusted HR for OS for durva vs atezo was 0.98 (95% CI: 0.86, 1.10; P = 0.7), indicating no statistically significant difference between the 2 treatments. In the 1Lm cohort, the adjusted HR for OS for durva vs atezo was 1.04 (95% CI: 0.88, 1.22; P = 0.7). In both cohorts, the upper 95% CI was < 1.25, supporting the noninferiority of durva vs atezo. Conclusions: In this rw population, atezo and durva offered comparable rwOS, which was shorter than those reported in phase 3 clinical trials, for pts with ES-SCLC when used with chemo in 1L or as monotherapy in 1Lm.