Use of targeted therapy in patients with advanced non-small cell lung cancer in response to broad genomic profiling.

11155 Background: Broad genomic profiling (BGP) is increasingly performed for patients with advanced non-small cell lung cancer (aNSCLC) to inform the use of targeted therapy (TT). The approach to specific BGP results has evolved, alongside changes in evidence, FDA approval status, and professional guidelines. Limited data exist characterizing impact of BGP on treatment selection in a real-world patient population. Methods: We identified patients diagnosed with aNSCLC 2017–2023 using the nationwide deidentified Flatiron Health-Foundation Medicine aNSCLC Clinico-Genomic Database, containing data from ~280 US cancer clinics (~800 sites of care). We classified each patient’s BGP results by contemporaneous FDA approvals and NCCN guidelines, categorizing findings as actionable with on-label TT (first- [1L] vs. later-line [2+]), potentially actionable with off-label TT (NCCN recommended vs. not recommended), or not actionable. We then categorized actual 1L treatments received relative to these standards. Results: Of 5781 patients with aNSCLC who underwent BGP testing (Table), on-label 1L TT was used in 13.1% of patients, while 3.3% (1.3% guideline recommended + 2.0% not recommended) received off-label 1L TT. Overall, 17.6% had a 1L targetable alteration (74.5% of whom received on-label TT), while 6.0% had a 2L+ on-label targetable alteration (9.7% received TT in 1L, against label + recommendation). In addition, 4.7% had a potentially actionable alteration with recommended off-label TT (11.4% of whom received off-label 1L TT), while an additional 23.3% had potentially actionable alterations without recommended TT (4.3% received off-label TT in 1L, against NCCN recommendation). Conclusions: In this cohort of patients with aNSCLC, half had actionable or potentially actionable BGP results. One in four patients with 1L actionable findings did not receive corresponding on-label 1L TT (potential missed opportunity), while one in ten with 2L+ actionable findings received TT as off-label 1L treatment in contrast to guidelines (potential indication creep). 1L off-label TT use was uncommon and largely guideline discordant. First-line treatments by mutation actionability in patients with advanced non-small cell lung cancer, as recommended (Rec’ed) by contemporaneous clinical guidelines. Mutation Category On-Label TT Rec’edOff-Label TT NOT Rec’ed Off-Label TT Non-Targeted Systemic Therapy Actionable, TT 1 st LineN = 1018 (17.6%) 758 (74.5%) 44 (4.3%) ≤5 (≤0.5%) ≥211 (≥20.7%) Actionable, TT 2 nd LineN = 349 (6.0%) . . 34 (9.7%) 315 (90.3%) Potentially actionable –Off-Label TT Rec’edN = 273 (4.7%) . 31 (11.4%) ≤5 (≤1.8%) ≥237 (≥86.8%) Potentially actionable –Off-Label TT NOT Rec’edN = 1348 (23.3%) . . 58 (4.3%) 1290 (95.7%) Not actionableN = 2793 (48.3%) . . 16 (0.6%) 2777 (99.4%) TOTALN = 5781 758 (13.1%) 75 (1.3%) 115 (2.0%) 4833 (83.6%) Cells ≤5 patients censored; TT = targeted therapy.