Mismatch repair genes as prognosis biomarkers for hepatocellular carcinoma

Abstract

Abstract Introduction: Great progress was made in early diagnosis and in the treatment of hepatocellular carcinoma (HCC) in most countries, but the mortality rate is still very high. The outcome for HCC patients is influenced not just by the traits of the tumor, but also by its cause, liver functionality, and individual variations, leading to past models not yielding optimal outcomes. Mismatch repair is highly related to the prognosis and progression of liver cancer. However, the prediction model of liver cancer progression by mismatch repair pathway-related genes has not been established. Methods: In this study, mismatch repair pathway-related genes were screened from the TCGA and ICGC databases. We employed both univariate analysis and lasso Cox regression analysis to pinpoint eight genes and formulate a risk score. The model's clinical utility was subsequently confirmed through Cox regression analyses. Results: We chose eight genes (YBX1, PSMD14, NOP58, RUVBL1, HMMR, KPNA2, BSG, and IRAK1) from the set of mismatch repair genes and utilized them to create a prognostic risk factor, which was subsequently validated by using TCGA database. The results indicated a big difference in prognosis between risk groups, categorized based on median risk coefficient. Additionally, we employed a nomogram to predict overall survival. Furthermore, when we conducted functional enrichment analysis, it revealed a connection between the high-risk group and cell cycle process and DNA replication synthesis. Further analysis also suggested that differences in prognosis between various risk groups could be attributed to an immunosuppressed tumor microenvironment. Discussion: The prognostic model composed of 8 mismatch repair pathway-related genes has potential application value and good predictive performance. The related genes may be biomarkers for HCC treatment, which can provide new strategy in guiding the clinical prediction of prognosis.