Treatment Patterns Across Lines of Therapy for Advanced Non‐Small Cell Lung Cancer in the United States

ABSTRACT Introduction Treatment strategies for advanced non‐small cell lung cancer (aNSCLC) have evolved, but little is known about real‐world approaches by biomarker groups across different lines of therapy (LOTs), particularly regarding discontinuation rates as a measure of effectiveness and tolerability. Methods Using deidentified patient data from the Flatiron Health electronic health record‐derived database, we constructed a retrospective cohort study of aNSCLC patients diagnosed between January 2016 and June 2020, with follow‐up through August 2023. Patients had either anaplastic lymphoma kinase ( ALK ) rearrangements or epidermal growth factor receptor ( EGFR ) mutations or had programmed cell death‐ligand 1 (PD–L1) biomarker results (PD–L1 < 1%, 1%–49%, or ≥ 50%). For each biomarker group, we summarized treatments received across up to the sixth LOT, calculating the overall discontinuation rate and the rate due to death. Results There were 13,369 patients in our sample (427 patients with ALK rearrangements, 2283 with EGFR mutations, 3663 PD–L1 < 1%, 3281 PD‐L1 1%–49%, and 3715 PD–L1 ≥ 50%). Treatment patterns varied across biomarker groups. The overall discontinuation rate after first‐line therapy was 31.9% for ALK ‐rearrangement, 39.5% for EGFR ‐mutation, and ranged from 50.6% to 57.5% for the PD–L1 groups. The discontinuation rate due to death after first‐line therapy was 8% ( ALK ‐rearrangement), 14.8% ( EGFR ‐mutation), and approximately 21% for the three PD–L1 groups. Patients with gene alterations had lower overall discontinuation rates across second through fifth LOTs than those without. Conclusions Patients with driver alterations had lower overall discontinuation rates in subsequent therapy lines, highlighting the need for effective, tolerable treatments for those without driver alterations. Notably, a substantial proportion of patients with targetable driver alterations did not receive the corresponding TKI, underscoring real‐world gaps in biomarker‐directed treatment implementation.