Genomic DNA Methylation-Derived Algorithm Enables Accurate Detection of Malignant Prostate Tissues

Area of Science:

• Genomics
• Cancer Research
• Molecular Diagnostics

Background:

• Prostate cancer (PCa) diagnosis relies on biopsies, which have limitations due to tumor heterogeneity and high false-negative rates.
• Molecular markers hold potential for improving PCa detection and overcoming diagnostic challenges.
• Genome-wide DNA methylation analysis is explored for novel PCa diagnostic markers.

Purpose of the Study:

• To identify genome-wide DNA methylation changes in prostate cancer.
• To develop a diagnostic algorithm for PCa screening using identified genomic markers.
• To investigate the role of DNA methylation in PCa development and progression.

Main Methods:

• Genome-wide DNA methylation arrays were used on archival PCa and normal tissues.
• Differentially methylated regions (DMRs) underwent functional and pathway analyses.
• A classification algorithm was trained using logistic regression on specific CpG sites, validated on an external cohort.

Main Results:

• Significant DNA methylation changes, primarily hypermethylation, were observed in approximately 6,000 probes and 600 regions.
• Enrichment analysis revealed associations with cell communication, neurogenesis, proliferation, and tumor suppressor binding sites.
• A classification model using four CpG sites demonstrated 100% accuracy in initial testing and 96% sensitivity/98% specificity in validation, detecting metastatic lesions.

Conclusions:

• Prostate cancer DNA methylation profiles contain driver events potentially linked to tumor suppressor disruption.
• A diagnostic algorithm based on as few as four CpG sites can effectively screen for prostate cancer.
• This methylation-based approach offers a promising alternative or adjunct to current PCa diagnostic methods.