Cigarette smoking on outcomes with osimertinib in <i>EGFR</i> -mutated non–small cell lung cancer: A real-world analysis.

Abstract

e20680 Background: Osimertinib is a preferred therapy for patients with EGFR-mutated non–small cell lung cancer (NSCLC). However, while the effect of cigarette smoking has been evaluated in clinical trials, there is limited data on real-world populations. This study evaluated whether smoking history influences overall survival (OS) and time to next treatment (TTNT) among patients treated with osimertinib. Methods: We conducted a retrospective cohort study using the TriNetX database, including adults (≥18 years) with EGFR-mutated NSCLC who initiated osimertinib between 2015 and 2024. Patients were stratified by smoking history (current or former smokers vs never smokers). Cohorts were matched 1:1 using propensity scores based on age, sex, race, ethnicity, metastatic status, and comorbidities. Primary outcomes were OS and TTNT. TTNT was defined as time from osimertinib initiation to receipt of any subsequent antineoplastic therapy, with follow-up starting 14 days after osimertinib initiation to avoid misclassification. Kaplan–Meier methods and log-rank testing were used. A sensitivity analysis excluding smoking-related comorbidities was performed to assess the impact of overadjustment. Results: Before matching, 1,748 smokers and 5,249 never smokers were identified; after matching, 1,612 patients remained in each cohort. Mean age after matching was 67.8 years in smokers and 68.1 years in never smokers, with a consistent 60% female distribution. In the primary matched analysis, there was no significant difference in OS between smokers and never smokers (5-year OS 32% vs 32%; median OS 37.7 vs 38.3 months; p = 0.824). TTNT was also similar between groups (99 vs 101 days; p = 0.637). In the sensitivity analysis excluding smoking-related comorbidities, smokers demonstrated inferior OS, with lower 5-year survival and shorter median OS (32% vs 34%; 37.4 vs 41.2 months; p = 0.046), while TTNT remained unchanged (99 vs 102 days; p = 0.545). Conclusions: In a large real-world cohort of EGFR-mutated NSCLC patients treated with osimertinib, smoking history was not associated with differences in OS or TTNT when smoking-related comorbidities were balanced. However, when real-world comorbidity burden was retained, smokers experienced significantly worse survival. These findings suggest that smoking-related comorbidities, rather than smoking history alone, may drive inferior outcomes and should be considered in prognostication and patient counseling.