Real-world outcomes of first-line immune checkpoint inhibitors with or without chemotherapy in KRAS G12C altered NSCLC according to PD-L1 status.

Abstract

9136 Background: There is a considerable debate whether treatment approaches should be tailored in advanced KRAS G12C-mutated NSCLC. Increasing evidence suggests that KRAS G12C-mutated NSCLC is associated with genomic heterogeneity that may impact clinical outcomes. Herein, we report a multimodal, real world outcomes analysis of 1L patients with advanced KRAS G12C-mutated NSCLC stratified by treatment type and PD-L1 status. Methods: Deidentified multimodal real-world data (RWD) of 1576 advanced, 1L NSCLC patients across the US were retrospectively analyzed from the Tempus database. Selection criteria included: Tempus xT targeted NGS, known PD-L1 status, absence of EGFR, ALK and ROS1 mutations, and treatment with chemotherapy (CT) + pembrolizumab (P), or P only. Patients were stratified by the presence or absence (wt) of KRAS G12C alterations. Median OS (mOS) was estimated using Kaplan-Meier methods. Subgroup analyses were performed using Cox model stratified by KRAS G12C status, PD-L1 status, and pathogenic alterations of STK11 and KEAP1. Results: In this study, two cohorts were assessed, KRAS G12C (n = 201) and KRAS G12C wt (n = 1375). Characteristics and demographics were balanced between both cohorts. Among TPS < 50 patients treated with CT + P, KRAS G12C (n = 109, 54%) had a mOS of 11.64 months (m) versus 16.81m (HR = 1.32, p = 0.06) for KRAS G12C wt (n = 864, 62.8%). Among TPS < 1 patients treated with CT + P, KRAS G12C (n = 46, 22%) had a mOS of 11.18 m versus 16.44 m (HR = 1.67, p = 0.01) for KRAS G12C wt (n = 489, 35%). Among TPS ≥50 patients treated with P only, KRAS G12C (n = 45, 22.3%) had a mOS of 30.03m compared to 25.03m (HR = 0.94, p = 0.83) in KRAS G12Cwt (n = 236,17.1%). TPS < 50 subgroup was enriched for select co-mutations with KRAS G12C ( STK11, 24.5%; KEAP1, 10.7%) vs. TPS ≥50 ( STK11, 3.8%; KEAP1, 5.1%). When evaluating the impact of co-mutations in the KRAS G12C with TPS < 50, STK11 (n = 30) mOS was 10.59m (HR = 1.01, p = 0.81) and KEAP1 (n = 13) mOS was 7.63m (HR = 2.14, p = 0.03). Conclusions: This largest RWD analysis to date demonstrates that 1L KRAS G12C NSCLC patients with TPS < 1% and < 50% receiving standard CT + P have the shortest survival among all evaluated sub-groups. Novel KRAS G12C targeted combination therapies in development for this patient population may offer the promise of better outcomes. [Table: see text]