Dual Biomarker Combining DNA Damage Repair Gene Mutations and PD-L1 Expression for Immune Checkpoint Inhibitors in Non-small Cell Lung Cancer.

Abstract

Immune checkpoint inhibitors (ICIs) improve long-term survival in advanced non-small cell lung cancer (NSCLC) and require robust predictive biomarkers for the selection of responders. This study investigated the optimal implementation of DNA damage repair (DDR) gene mutations to predict response to ICIs in real-world NSCLC patients. We retrospectively reviewed 55 advanced NSCLC patients who had undergone targeted high-throughput sequencing and received ICIs. Patients with two or more DDR gene mutations were defined as DDR2 positive. The patients' median age was 68 (range=44-82) years, and 48 (87.3%) were men. Seventeen patients (30.9%) showed ≥50% high programmed death-ligand 1 (PD-L1) expression. Ten patients (18.2%) received an ICI-chemotherapy combination as first-line therapy, and 38 (69.1%) received ICI monotherapy as more than second-line therapy. Fourteen patients (25.5%) were DDR2-positive. The objective response rate of patients with DDR2-positive or PD-L1 ≥50% was 45.5%, and that of patients with DDR2-negative and PD-L1 <50% was 11.1% (p=0.007). In the PD-L1 low expression subgroup (<50%), patients with DDR2-positive had improved progression-free survival (PFS) and overall survival (OS) after ICIs compared to those with DDR2-negative (PFS: 5.8 vs. 1.9 months, p=0.026, OS: 14.4 vs. 7.2 months, p=0.078). DDR2-positive patients or those with PD-L1 ≥50% (24, 43.6%) had statistically significant improvement in PFS and OS after ICIs compared to DDR2-negative and those with PD-L1 <50% (PFS: 4.4 vs. 1.9 months, p=0.006, OS: 11.6 vs. 7.2 months, p=0.037). A dual biomarker combining DDR gene mutations and PD-L1 expression improves the prediction of response to ICIs in advanced NSCLC.