Abstract
Immune checkpoint blockade (ICB) therapies that target programmed cell death 1 (PD1) and PD1 ligand 1 (PDL1) have demonstrated promising benefits in lung adenocarcinoma (LUAD), and tumor mutational burden (TMB) is the most robust biomarker associated with the efficacy of PD-1-PD-L1 axis blockade in LUAD, but the assessment of TMB by whole-exome sequencing (WES) is rather expensive and time-consuming. Although targeted panel sequencing has been developed and approved by the US Food and Drug Administration (FDA) to estimate TMB, we found that its predictive accuracy for ICB response was significantly lower than WES in LUAD. Given that previous studies were mainly focusing on genomic variations to explore surrogate biomarkers of TMB, we turned to examine the transcriptome-based correlation with TMB in this study. Combining three immunotherapeutic cohorts with two independent The Cancer Genome Atlas (TCGA) datasets, we revealed that the expression of mutS homolog 2 (MSH2), one of the most crucial genes involved in DNA mismatch repair (MMR) pathway, was the strongest feature associated with increased TMB in multivariate analysis. Furthermore, MSH2 expression also displayed a significantly positive correlation with smoking signature while an inverse association with MMR deficiency (MMRd) signature in LUAD. More importantly, high expression of MSH2 markedly correlated with increased PD-L1 expression and CD8+ T cell infiltration, both suggesting a prominent immunotherapy-responsive microenvironment in LUAD. Notably, detecting MSH2 expression is much easier, faster, and cheaper than TMB in clinical practice. Taken together, this study demonstrates the association of MSH2 expression with TMB and the immune microenvironment in LUAD. MSH2 expression may be developed as a potential surrogate biomarker of TMB to identify ICB responders in LUAD.