Abstract 4745: Prostatic fluid-based cfDNA methylation profiling distinguish benign hyperplasia from prostate cancer

Abstract

Abstract Introduction: A challenge in the early diagnosis of prostate cancer (PC) is to distinguish between malignant tumor and benign prostatic hyperplasia (BPH) because of the similar symptoms. Methylation profiles provide a way for discriminating BPH from PC, which could prevent overdiagnosis and overtreatment. Applying methylation biomarkers to liquid biopsy specimens provides higher sensitivity to detect tumors. Methods: Prostatic fluid samples were prospectively collected from 40 patients following digital rectal exam and isolated cell-free DNA to perform whole genome methylation sequencing using PredicineEPIC platform. Combing differentially methylated regions (DMR) with methylation signature identified from massive public data, a panel of methylation patterns was developed. The difference of methylation signal in profiled region were implemented by machine learning to robustly distinguish BPH from prostate cancer. Results: We identified 324 biomarkers from prostatic fluid samples and collected 46 methylation biomarkers reported from previous research. Regions with extreme difference between benign prostatic hyperplasia and prostate cancer were identified as candidate biomarkers. A diagnostic model trained on these 55 biomarkers could distinguish BPH from PC in 10-fold cross-validation (AUC = 0.92). And this model could be applied for discriminating normal tissue from prostate tumor tissue with highly performance (AUC > 0.95). Using this set of regions, we developed an NGS-based panel for detecting methylation alteration of cfDNA from prostate liquid, which inherited the ability of differentiation. Conclusion: These methylation patterns could be developed as the novel biomarkers to support the development of prostatic fluid-based non-invasive diagnostic test and reduce unnecessary biopsies. This panel could be applied for differential diagnosis of benign hyperplasia and prostate cancer. Citation Format: Hang Dong, Haoran Tang, Yue Zhang, Shidong Jia. Prostatic fluid-based cfDNA methylation profiling distinguish benign hyperplasia from prostate cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4745.