Discover Oncology (Jul 2023)
PTEN-related risk classification models for predicting prognosis and immunotherapy response of hepatocellular carcinoma
Abstract
Abstract Introduction PTEN often mutates in tumors, and its manipulation is suggested to be used in the development of preclinical tools in cancer research. This study aims to explore the biological impact of gene expression related to PTEN mutations and to develop a prognostic classification model based on the heterogeneity of PTEN expression, and to explore its sensitivity as an indicator of prognosis and molecular and biologic features in hepatocellular carcinoma (HCC). Material and methods RNA-seq data and mutation data of the LIHC cohort sample downloaded from The Cancer Genome Atlas (TCGA). The HCC samples were grouped according to the mean expression of PTEN, and the tumor microenvironment (TME) was evaluated by ESTIMATE and ssGSEA. The prognostic classification model related to PTEN were constructed by COX and LASSO regression analysis of differentially expressed genes (DEGs) between PTEN-high and -low expressed group. Results The expression of PTEN was affected by copy number variation (CNV) and negatively correlated with immune score, IFNγ score and immune cell infiltration. 1281 DEGs were detected between PTEN-high and PTEN-low expressed group, 8 of the DEGs were finally filtered for developing a prognosis classification model. This model showed better prognostic value than other clinicopathological parameters, and the prediction accuracy of prognosis and ICB treatment for immunotherapy cohorts was better than that of TIDE model. Conclusions This study demonstrated the effect of CNV on PTEN expression and the negative immune correlation of PTEN, and constructed a classification model related to the expression of PTEN, which was of guiding significance for evaluating prognostic results of HCC patients and ICB treatment response of cancer immunotherapy cohorts.
Keywords
