Comprehensive analysis of cholesterol metabolism-related genes in prostate cancer: integrated analysis of single-cell and bulk RNA sequencing.

BACKGROUND

Cholesterol metabolism plays a significant role in cancer progression, including prostate adenocarcinoma (PRAD), making it a promising target for therapeutic intervention. This study aimed to construct and validate a cholesterol metabolism gene (CMG)-related prognostic signature to predict prognosis in PRAD patients, while exploring its biological, clinical, and therapeutic implications.

METHODS

CMGs were retrieved through comprehensive searches in public databases. Prognostic CMGs were determined using univariate Cox regression analysis on The Cancer Genome Atlas (TCGA) PRAD dataset. Patients were classified into subgroups using consensus clustering. Functional enrichment and Gene Set Enrichment Analysis (GSEA) were applied to explore the potential pathways. Importantly, a prognostic signature based on CMGs was constructed using the least absolute shrinkage and selection operator (LASSO) method, with performance evaluated through Kaplan-Meier (KM) analyses and receiver operating characteristic (ROC) curves. The model was validated in three external cohorts, and its clinical relevance was assessed through nomogram construction and drug sensitivity analysis. Immune landscape analysis was also performed to evaluate the PRAD immune microenvironment. Single-cell RNA sequencing analysis was conducted using Seurat package.

RESULTS

18 CMGs were identified to establish the prognostic signature. The risk score derived from this signature demonstrated robust prognostic performance in survival analysis and was significantly associated with key clinical variables, including N-stage, T-stage, and Gleason Score. The risk score of CMG signature was recognized as an independent prognostic parameter, and a nomogram was created to estimate 1-, 3-, and 5-year prognosis in PRAD patients. Additionally, the analysis of drug sensitivity identified variations in responses to commonly used drugs (such as camptothecin, CDK9 inhibitors, docetaxel, mitoxantrone, paclitaxel, and sepantronium bromide) between the two risk groups. Furthermore, immune landscape and single-cell sequencing analyses indicated that biological pathways were significantly correlated with the risk score.

CONCLUSIONS

The CMG-based prognostic model effectively predicts prognosis in PRAD patients and is linked to distinct biological pathways, immune landscapes, and drug sensitivities. This signature has the robust potential to guide personalized therapy and improve prognosis in PRAD.