Genomic landscape and its correlations with tumor mutational burden, PD-L1 expression, and immune cells infiltration...

To depict the genomic landscape of Chinese early-stage lung squamous cell carcinoma (LUSC) and investigate its correlation with tumor mutation burden (TMB), PD-L1 expression, and immune infiltrates.

Whole-exome sequencing was performed on 189 surgically resected LUSC. TMB was defined as the sum of nonsynonymous single nucleotide and indel variants. CD8⁺ tumor-infiltrating lymphocyte (TIL) density and PD-L1 expression were evaluated by immunohistochemistry. Six immune infiltrates were estimated using an online database.

The median TMB was 9.43 mutations per megabase. Positive PD-L1 expression and CD8⁺ TILs density were identified in 24.3% and 78.8%. PIK3CA amplification was associated with significantly higher TMB (P = 0.036). Frequent genetic alterations had no impact on PD-L1 expression but PIK3CA amplification and KEAP1 mutation were independently associated with significantly lower CD8⁺ TIL density (P < 0.001, P = 0.005, respectively). Low TMB and high CD8⁺ TIL density were independently associated with longer disease-free survival (DFS) while none of them could individually predict the overall survival (OS). Combination of TMB and PD-L1 expression or TMB and CD8⁺ TIL density could stratify total populations into two groups with distinct prognosis. Classifying tumor-immune microenvironment based on PD-L1 expression and CD8⁺ TIL density showed discrepant genomic alterations but similar TMB, clinical features, and OS. Notably, patients with different smoking status had distinct prognostic factors.

The combination of TMB, PD-L1 expression, immune infiltrates, and smoking status showed the feasibility to subgroup stratification in Chinese patients with early-stage LUSC, which might be helpful for future design of personalized immunotherapy trials in LUSC.