A Urine Exosome Gene Expression Panel Distinguishes between Indolent and Aggressive Prostate Cancers at Biopsy

作者：Indu Kohaar, Yongmei Chen, Sreedatta Banerjee, Talaibek Borbiev, Huai‐Ching Kuo, Amina Ali, Lakshmi Ravindranath, Jacob Kagan, Sudhir Srivastava, Albert Dobi, Isabell A. Sesterhenn, Inger L. Rosner, Jennifer Cullen, Shiv Srivastava, György Petrovics · 发表于：The Journal of Urology · 年份：2020 · DOI：10.1097/ju.0000000000001374 · 被引用次数：38 · 研究领域：Extracellular vesicles in disease、Prostate Cancer Diagnosis and Treatment、Prostate Cancer Treatment and Research

PURPOSE: Prostate cancer is predominantly indolent at diagnosis with a small fraction (15% to 25%) representing aggressive subtype (Gleason score 7-10), which is prone to metastatic progression. It is critical to explore noninvasive assays for the early detection of this aggressive subtype, when it still can be treated effectively. Additionally, there is an emerging need to develop markers that perform equally well across races, as racial differences in the prevalence and mortality of prostate cancer has become evident. MATERIALS AND METHODS: First catch, nondigital rectal examination urine specimens were collected from patients undergoing diagnostic biopsy. Total RNA was extracted from urinary exosomes and a quantitative expression assay protocol using droplet digital polymerase chain reaction was developed for detection of candidate genes in exosomal mRNAs from urine. Clinical performance for the gene expression assay was evaluated to predict high grade cancer (Gleason score 7-10) from low grade cancer (Gleason score 6) and cancer negative cases at biopsy. Assay performance was examined in combination with standard of care to determine improvement in model prediction. RESULTS: ), in combination with standard of care variables, significantly improved the prediction of high grade cancer at diagnosis compared to standard of care variables alone (AUC 0.88 vs 0.80, respectively, p=0.016). Decision curve analysis showed that there is a benefit of adopting the gene panel for detec...