Natural Killer Cell-Associated Radiogenomics Subtyping of Hepatocellular Carcinoma Based on CD2 Expression and Enhanced CT-Derived Radiomics Signatures

Abstract

Abstract Background Chimeric antigen receptor (CAR)-natural killer (NK) cell therapy has shown variable efficacy in patients with hepatocellular carcinoma (HCC). The present study sought to identify NK cell-related prognostic biomarkers, and to develop a non-invasive radiomics signature in patients with HCC. Methods Transcriptomic data from five independent cohorts, consisting of 734 HCC patients, in Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases were analyzed using the Microenvironment Cell Populations-counter (MCP-counter) algorithm. NK cell-related prognostic biomarkers were identified using weighted gene co-expression network analysis (WGCNA) and least absolute shrinkage and selection operator (LASSO)-Cox regression analyses. Radiomics models related to NK cell-related prognostic biomarkers were established using radiomics feature extraction and screening of preoperative enhanced CT images of 168 patients in two datasets from The Cancer Imaging Archive (TCIA) database. HCC radiogenomics subtypes were proposed based on genetic biomarkers and radiomics models. Results CD2 expression was an independent NK cell-associated prognostic biomarker in HCC patients, being associated with improved overall, progression-free, and recurrence-free survival, and correlating with NK cell-associated pathways and biological processes in HCC. A reliable non-invasive prognostic radiomics model for HCC was established using enhanced CT images. The combination of CD2 expression and radioscore identified potential HCC radiogenomics subtypes, thus explaining the prognostic differences caused by heterogeneity based on genomics and NK cell-associated biological processes. Conclusions HCC can be classified into radiogenomic subtypes, explaining in part the complex heterogeneity of HCC. These findings may facilitate the development of personalized treatment strategies using CAR-NK cells in patients with HCC.