Medicine & Health
AI Uncovers Key Prostate Cancer Biomarkers for Better Prognosis and Treatment
Original Authors: Tang, Zhang, Tao, Zhang, Xing, Wang, Yan, Gao, Zhang
Prostate cancer remains a significant health challenge globally, marked by its high incidence and complex tumor microenvironment. This complexity, stemming from diverse cell populations, often hinders clear interpretation of gene and biomarker roles in disease progression and immune response modulation. To address this, researchers embarked on a comprehensive investigation, combining bulk and single-cell RNA sequencing data to evaluate the clinical relevance and prognostic potential of prostate cancer-related genes, aiming to achieve a thorough understanding of prognostic characteristics.
Leveraging advanced methodologies, including FindAllMarkers, Dseq2 R package, ssGSEA, and WGCNA across both single-cell and bulk transcriptomes, the team identified 91 genes linked to prognosis within the tumor microenvironment, with 15 specifically associated with biochemical recurrence. The study developed a machine learning approach, integrating 14 algorithms and 162 combinations, to construct a consensus Immune and Prognostic-Related Signature (IPRS). This IPRS underwent systematic validation in training and test cohorts, and multivariate analysis further supported its potential as an independent prognostic marker for prostate cancer progression, revealing significant differences in biological functions, immune infiltration, and genomic mutations across various risk groups.
The clinical applicability of the IPRS was further evaluated in the context of immunotherapy and personalized drug selection. As lead author Tang notes in the paper, "Significantly, the submap method revealed enhanced immunotherapy responsiveness in high-risk patients while highlighting potential pharmacological targets for certain risk subgroups." The team also examined and confirmed both the gene expression associated with IPRS and the expression level and function of B-cell adhesion molecule (BCAM) within prostate cancer cells and tissue. This collection of genes relevant to PCa prognosis and immune characteristics may serve as potential biomarkers with clinical translational value, offering insights for more precise prognostication and personalized therapeutic approaches.