基于失巢凋亡相关基因并利用 LASSO-cox 方法构建肝细胞癌预后特征

To develop a signature based on anoikis-related genes (ARGs) for predicting the prognosis of patients with hepatocellular carcinoma (HCC), and to elucidate the molecular mechanisms involved. In this study, bioinformatic algorithms were applied to integrate and analyze 777 HCC RNA-seq samples from the cancer genome atlas and international cancer genome consortium repositories.

A prognostic signature was developed via the least absolute shrinkage and selection operator-cox regression method. To evaluate the accuracy of the signature in predicting events, multi-type technical means, such as Kaplan-Meier plots, receiver operating characteristic curve analysis, nomogram construction, and univariate and multivariate Cox regression studies were performed.

We investigated the underlying molecular biological mechanisms and immune mechanisms of the signature using gene set enrichment analysis and the CIBERSORT R package, respectively. Meanwhile, immunohistochemical staining acquired from the human protein atlas was used to confirm the differential expression levels of hub genes involved in the prognostic signature.

We developed an HCC prognostic signature with a collection of 5 ARGs, and the prognostic value was successfully assessed and verified in both the test and validation cohorts. The risk scores calculated by the prognostic signature were proved to be an independent negative prognostic factor for overall survival. A set of nomograms based on risk scores was established and found to be effective in predicting OS.

Further investigation of the underlying molecular biological mechanisms and immune mechanisms indicated that the signature may be relevant to metabolic dysregulation and infiltration of gamma delta T cells in the tumor. The survival prognosis of HCC patients can be predicted by the anoikis-related prognostic signature, and it serves as a valuable reference for individualized HCC therapy.