A novel prognostic signature based on mitochondrial permeability transition-driven necrosis genes for biochemical recurrence prediction in prostate cancer

Background Mitochondrial permeability transition-driven necrosis (MPT-DN) is a therapeutic target and critical driver of prostate adenocarcinoma (PRAD) progression. We investigated MPT-DN-related prognostic features in PRAD. Methods PRAD transcriptomics and MPT-DN-RGs were sourced from public databases. WGCNA, differential expression, Cox regression, and machine learning identified BCR-FS prognostic genes. These genes built a risk model, revealing independent prognostic factors. Patients were stratified into high/low-risk groups. Pathways, immune microenvironment, and drug sensitivities were analyzed between groups. Finally, protein expression was validated in PCa versus normal tissues. Results TREM2, FNDC1, and S100A8 were identified as prognostic genes. The developed risk model demonstrated strong predictive capabilities in BCR-FS, and subsequent analysis confirmed risk score, Gleason, T stage, and prostate specific antigen (PSA) as independent prognostic factors. The majority of the enrichment pathways in the high-risk group (HRG) and low-risk group (LRG) were related to the metabolism. Moreover, it was found that HRG and LRG displayed distinct immune landscapes, with HRG exhibiting immune exclusion and stronger immune evasion capabilities. Lastly, analysis of drug sensitivity showed significant differences for 6 drugs, with all values being lower in the HRG. Conclusion This study identified TREM2, FNDC1, and S100A8 as key MPT-driven necrosis-related genes predicting biochemical recurrence in PRAD. The risk model effectively stratified patients, revealing immune exclusion and drug resistance in high-risk cases, offering prognostic and therapeutic insights.