Abstract
Background: Prostate cancer (PCa) represents a major health burden globally, with over 1.4 million annual diagnoses and almost 400,000 PCa-related deaths. Current approaches to PCa screening and diagnosis are suboptimal, creating challenges for the accurate detection of clinically important PCa. Hence, there is an urgent need for novel, accessible, and clinically deployable biomarkers to optimise PCa care. Circulating tumour DNA (ctDNA) is a minimally invasive, highly specific technology that identifies the presence of a tumour using plasma-derived cell-free DNA (cfDNA). DNA methylation changes are a universal hallmark of cancer that dynamically influence gene expression. We, and others, have shown that ctDNA-based epigenetic alterations are higher in number, potentially enabling earlier tumour detection than genetic mutation analysis. Analysis of the cell-free DNA methylome, therefore, represents an exciting avenue for the development of highly accurate biomarkers for early PCa detection.
Results: In this multicentre, prospective discovery study, we have performed whole-genome scale DNA methylation profiling of matched patient blood and prostate biopsy tissue samples. Men scheduled to undergo prostate biopsy for clinical suspicion of PCa were recruited from several localities across New Zealand, with matched pre-biopsy peripheral blood samples and prostate biopsy tissue collected for molecular profiling. Unbiased, whole-genome scale DNA methylation profiling for both plasma cfDNA and tissue samples was performed using cell-free reduced representation bisulfite sequencing (cfRRBS). Over 200 differentially methylated regions (DMRs) differentiating malignant from benign and healthy samples have been identified from early profiling results of cfDNA samples, demonstrating significant discriminatory power for PCa detection at very early-stage disease. PCa- specific DMRs have also been identified from prostate tissue samples—interestingly, however, the most predictive tissue-derived DMRs cluster distinctly from those detected in cfDNA, even in matched patient samples.
Conclusions: Early profiling results show promise for cfRRBS as a discovery approach for cell-free DNA-based methylation biomarkers, with high prospective utility in the setting of PCa detection. Profiling results will be further refined and correlated to clinical and histopathological features of disease severity to generate a blood-based, targeted methylation panel for early and accurate PCa detection.