Integrated analysis of multiple transcriptomic approaches and machine learning integration algorithms reveals high...

Despite the availability of established surgical and chemotherapy options, the treatment of bladder cancer (BCa) patients remains challenging. While immunotherapy has emerged as a promising approach, its benefits are limited to a subset of patients. The exploration of additional targets to enhance the efficacy of immunotherapy is a valuable research direction.

High endothelial venules (HEV) ssGSEA analysis was conducted using BEST. Through the utilization of R packages Limma, Seurat, SingleR, and Harmony, analyses were performed on spatial transcriptomics, bulk RNA-sequencing (bulk RNA-seq), and single-cell RNA sequencing (scRNA-seq) data, yielding HEV-related genes (HEV.RGs). Molecular subtyping analysis based on HEV.RGs was conducted using R package MOVICS, and various machine learning-integrated algorithm was employed to construct prognostic model. LDLRAD3 was validated through subcutaneous tumor formation in mice, HEV induction, Western blot, and qPCR.

A correlation between higher HEV levels and improved immune response and prognosis was revealed by HEV ssGSEA analysis in BCa patients receiving immunotherapy. HEV.RGs were identified in subsequent transcriptomic analyses. Based on these genes, BCa patients were stratified into two molecular clusters with distinct survival and immune infiltration patterns using various clustering-integrated algorithm. Prognostic model was developed using multiple machine learning-integrated algorithm. Low LDLRAD3 expression may promote HEV generation, leading to enhanced immunotherapy efficacy, as suggested by bulk RNA-seq, scRNA-seq analyses, and experimental validation of LDLRAD3.

HEV served as a predictive factor for immune response and prognosis in BCa patients receiving immunotherapy. LDLRAD3 represented a potential target for HEV induction and enhancing the efficacy of immunotherapy.