Abstract 2734: Poly-ligand profiling of plasma samples as a potential predictor of prostate biopsy results

Abstract

Abstract Prostate cancer is the most common cancer in men, excluding skin cancer, and is also one of the leading causes of cancer death. Currently the prostate-specific antigen (PSA) blood test is employed as an initial screening method in men without symptoms, and patients with blood PSA level higher than 4 ng/mL are commonly recommended for prostate biopsy for definitive diagnosis. However, other non-cancerous conditions may also cause elevation in PSA, such as prostatitis, benign prostatic hyperplasia and prostate intraepithelial neoplasia. Among men with PSA levels of 4-10 ng/mL, only about 1 in 4 will show positive (cancer) biopsy, and for those with PSA levels higher than 10 ng/mL, the chance of having prostate cancer increases to over 50%. Moreover, approximately 15% of men with a PSA levels below 4 ng/mL have a positive prostate biopsy. Because of these scenarios, as well as significant risks associated with prostate biopsies, there is a need for diagnostic adjuncts that provide additional data to inform decisions on whether or not to perform biopsies. Here we describe a method using single stranded oligonucleotide (ssODN) libraries to perform Poly-Ligand Profiling on plasma exosomes from men scheduled for biopsy to classify those with or without a subsequent diagnosis of prostate cancer. Through alternating rounds of positive and negative selection in binding assays, we isolated ssODNs that selectively bound to exosomes from prostate cancer patients (PSA level of at least 4 ng/mL; Gleason Score range 7-8, and at least 3 positive cores from at least 10 biopsy cores) and not to benign cases (all negative cores from at least 10 biopsy cores). An independent test set of 29 prostate cancer cases and 28 benign cases was then evaluated in binding assays. Based on 100 runs of 10-fold cross validation with multiple statistic models, an average successful classification rate of 82% was achieved with an area under the ROC curve of 0.89. While further refinement and testing on larger cohorts are needed, these results suggest that Poly-Ligand Profiling may provide an additional approach to stratify men with suspicious PSA levels for prostate biopsies. Citation Format: Xixi Wei, Xiaowei Liu, Radhika Santhanam, Daniel Magee, Anthony Helmstetter, Mark Miglarese, David Spetzler, Xixi Wei. Poly-ligand profiling of plasma samples as a potential predictor of prostate biopsy results [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2734.