Novel Inflammasome-Based Risk Score for Predicting Survival and Efficacy to Immunotherapy in Early-Stage Non-Small Cell Lung Cancer

Abstract

Immune checkpoint inhibitors (ICI) for early-stage non-small cell lung cancer (NSCLC) have been approved to improve outcomes and reduce recurrence. Biomarkers for patient selection are needed. In this paper, we proposed an inflammasome-based risk score (IRS) system for prognosis and prediction of ICI response for early-stage NSCLC. Cox regression analysis was used to identify significant genes (from 141 core inflammasome genes) for overall survival (OS) in a microarray discovery cohort (n = 467). IRS was established and independently validated by other datasets (n = 1320). We evaluated the inflammasome signaling steps based on five gene sets, which were IL1B-, CASP-1-, IL18-, GSDMD-, and inflammasome-regulated genes. Gene set enrichment analysis, the Kaplan–Meier curve, receiver operator characteristic with area under curve (AUC) analysis, and advanced bioinformatic tools were used to confirm the ability of IRS in prognosis and classification of patients into ICI responders and non-responders. A 30-gene IRS was developed, and it indicated good risk stratification at 10-year OS (AUC = 0.726). Patients were stratified into high- and low-risk groups based on optimal cutoff points, and high-risk IRS had significantly poorer OS and relapse-free survival. In addition, the high-risk group was characterized by an inflamed immunophenotype and higher proportion of ICI responders. Furthermore, expression of SLAMF8 was the key gene in IRS and indicated good correlation with biomarkers associated with immunotherapy. It could serve as a therapeutic target in the clinical setting of immunotherapy.