APOBEC mutational signatures to predict immune checkpoint inhibitor benefit in TMB-low metastatic NSCLC independent of PD-L1 status.

Abstract

3130 Background: Most metastatic non-small cell lung cancer (NSCLC) patients are tumor mutational burden (TMB)-low (<10 mut/Mb) and lack validated biomarkers to guide immune checkpoint inhibitor (ICI) use beyond PD-L1 expression. APOBEC mutagenesis, characterized by distinct mutational signatures, has been linked to ICI response in urothelial and head and neck cancers – but primarily in TMB-high tumors. Whether APOBEC activity predicts ICI benefit specifically in TMB-low NSCLC, independent of PD-L1 status, remains unclear. Methods: We analyzed a discovery cohort of 857 metastatic TMB-low NSCLC patients treated with ICI from the MSK-CHORD cohort. To distinguish predictive from prognostic effects, we examined 1,328 ICI-untreated TMB-low metastatic NSCLC patients as controls. Findings were validated in an independent cohort of 82 TMB-low, ICI-treated stage IV NSCLC patients using whole exome sequencing (WES). APOBEC positivity was determined using MESiCA, a validated machine-learning algorithm designed to detect mutational signatures from targeted gene panels with low mutation counts (≥4 mutations). Primary endpoints were overall survival (OS) and disease control rate (DCR) in the discovery cohort, and progression-free survival (PFS) and objective response rate (ORR) in the validation cohort. Multivariate Cox regression adjusted for PD-L1 status and age. Results: In the discovery cohort, APOBEC-positive (n=52, 6.1%) and APOBEC-negative (n=805) patients were balanced for sex (44% vs 46% male), histology (83% vs 79% adenocarcinoma), and PD-L1 status (40% vs 40%; all p>0.05); APOBEC-positive patients were younger (median 64 vs 69 years, p=0.002). The APOBEC-positive group demonstrated significantly improved OS (median 33.7 vs 17.4 months; HR=0.60; 95% CI, 0.42–0.85; p=0.004). The survival benefit was absent in ICI-untreated patients (HR=0.98; 95% CI, 0.76–1.26; p=0.85). Formal interaction testing confirmed that the APOBEC benefit was specific to ICI-treated patients (interaction HR=0.62; p=0.032). APOBEC-positive tumors demonstrated higher DCR at 180 days (21.4% vs 7.6%; OR=3.32; p=0.006). In multivariate analysis adjusting for PD-L1 and age, APOBEC remained independently predictive (HR=0.57; 95% CI, 0.40–0.81; p=0.002). In the validation cohort, APOBEC independently predicted superior PFS (HR=0.23; 95% CI, 0.05–0.94; p=0.020) after adjusting for PD-L1, with numerically higher ORR (55.6% vs 25.4%, p=0.068). Conclusions: APOBEC mutational signatures represent a robust, independently validated predictive biomarker for ICI response in TMB-low metastatic NSCLC. Unlike prior studies linking APOBEC predominantly to TMB-high tumors, we demonstrate clinical utility specifically in the TMB-low population, where predictive biomarkers are most needed. This biomarker could expand immunotherapy access for NSCLC patients.