Urine IRF4/PENK/PXDN Methylation Signatures Enable Machine Learning-Driven Bladder Cancer Detection and Microenvironment Dissection.

IntroductionCystoscopy-based diagnosis and surveillance of bladder cancer (BC) remain challenging. This study presents a urine-based assay that enriches DNA-methylation signals via PCR enrichment and applies a machine-learning model to enable cost-effective early detection.MethodsIn a prospective cohort at hospital (May 2022-November 2023), 155 individuals were enrolled, BC was diagnosed and confirmed by cystoscopy-guided biopsy and histopathology. Targeted next-generation sequencing of urine DNA quantified methylation at 44 CpG sites within IRF4, PENK and PXDN. Supervised classifiers trained on these features distinguished tumor from non-tumor urine samples. Systems analyses around IRF4/PENK/PXDN mapped signaling pathways, protein interaction modules and tumor-microenvironment contexts in BC.ResultsUsing 155 urine samples (68 BC, 87 non-BC) with methylation and transcript expression data in TCGA-BLCA, we found significantly increased methylation of IRF4, PENK, and PXDN in BC (P < 0.0001). Corresponding mRNA levels of IRF4 and PENK were significantly downregulated in tumor tissues, with PXDN showing a declining trend. Methylation of IRF4 and PXDN negatively correlated with their expression (P < 0.0001). A 41-CpG-site random forest classifier targeting IRF4, PENK, and PXND (BladderCando model) demonstrated excellent performance in distinguishing BC from non-BC individuals (AUC = 0.9783, F1 = 0.9773), outperforming urine cytology for low-grade BC detection. Co-expression and enrichment analyses identified DCN as a central hub gene, primarily linked to ECM functions. High expression of PXDN, IRF4, and DCN correlated with upregulated immune checkpoint genes and increased immune cell infiltration. Single-cell sequencing revealed PXDN in fibroblasts and endothelial cells, DCN in fibroblasts, and IRF4 in T cells, with expression patterns in urine mirroring tumor tissue profiles.ConclusionThis study establishes IRF4, PENK and PXDN methylation in urine as robust molecular signatures for BC detection, with machine-learning integration markedly enhancing diagnostic precision. Beyond early surveillance, these epigenetic alterations delineate tumor-microenvironment interactions that may inform future therapeutic strategies.