AUTHOR=Hu Zixin , Liu Zhening , Zheng Jiabin , Peng Yanmei , Lu Xingyu , Li Jia , Tan Kexin , Cui Huijuan 

TITLE=Microsatellite instability-related prognostic risk score (MSI-pRS) defines a subset of lung squamous cell carcinoma (LUSC) patients with genomic instability and poor clinical outcome

JOURNAL=Frontiers in Genetics

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2023.1061002

DOI=10.3389/fgene.2023.1061002

ISSN=1664-8021

ABSTRACT=Background: Lung squamous cell carcinoma (LUSC) shares less typical onco-drivers and target resistance, whereas a high overall mutation rate and marked genomic complexity. Mismatch repair (MMR) deficiency leads to microsatellite instability (MSI) and act a critical role in the genomic instability. The clinical implication of MSI in LUSC is unknown. In this study, MSI-related prognostic risk score (MSI-pRS) was established by the machine learning and bioinformatics methods. 
Method: MSI status was classified by MMR proteins. Intersection of the differential expression genes (DEGs) and differential methylation probes (DMPs) were classified into functional modules by weighted gene co-expression network analysis (WGCNA). Least absolute shrinkage and selection operator (LASSO) regression and stepwise gene selection were performed for model downscaling. 
Results: Compared with the MSI-low (MSI-L) phenotype, MSI-high (MSI-H) displayed higher genomic instability. 843 genes activated by hypomethylation and 430 genes silenced by hypermethylation in MSI-H tumors were classified into six functional modules. Four genes (CCDC68, LYSMD1, RPS7, CDK20) were used to construct MSI-pRS. Low MSI-pRS was a protective prognostic factor in all cohorts (HR =0.46, 0.47, 0.37; p.val = 7.57e-06, 0.009, 0.021). The model contains tumor stage, age, and MSI-pRS showed well discrimination and calibration. High MSI-pRS was associated with hypermethylation on TP53 promoters, and downregulated cell cycle checkpoint signaling. Tumors with high MSI-pRS were characterized by the inflamed tumor microenvironment (TME), which were more likely to benefit from immunotherapy.
Conclusions: This study established MSI-pRS model with well performation in predicting the prognosis in clinical practice. High MSI-pRS was associated with high genomic instability and shared inflamed TME. The study provided new insights in biomarker exploration of LUSC.