Identification and validation of a genomic mutation signature as a predictor for immunotherapy in NSCLC

Currently, the benefits of immune checkpoint inhibitor (ICI) therapy prediction via emerging biomarkers have been identified, and the association between genomic mutation signatures (GMS) and immunotherapy benefits has been widely recognized as well. However, the evidence about non-small cell lung cancer (NSCLC) remains limited. We analyzed 310 immunotherapy patients with NSCLC from the Memorial Sloan Kettering Cancer Center (MSKCC) cohort. Lasso Cox regression was used to construct a GMS, and the prognostic value of GMS could be able to verify in the Rizvi cohort (N=240) and Hellmann cohort (N=75). We further conducted immunotherapy-related characteristics analysis in The Cancer Genome Atlas (TCGA) cohort (N=1052). A total of seven genes (ZFHX3, NTRK3, EPHA7, MGA, STK11, EPHA5, TP53) were identified for GMS model construction. Compared with GMS-high patients, patients with GMS-low had longer overall survival (OS; P<0.001) in the MSKCC cohort and progression-free survival (PFS; P<0.001) in the validation cohort. Multivariate Cox analysis revealed that GMS was an independent predictive factor for NSCLC patients in both the MSKCC and validation cohort. Meanwhile, we found that GMS-low patients reflected enhanced antitumor immunity in TCGA cohort. The results indicated that GMS had not only potential predictive value for the benefit of immunotherapy but also may serve as a potential biomarker to guide clinical ICI treatment decisions for NSCLC.