Differentiation of high risk prostate cancer with a facile urinary exosome detection workflow.

Clear differentiation of high-grade and clinically insignificant prostate cancer (PCa) is critical for clinical decision-making. Here, we developed a proprietary urinary exosome isolation approach (EVLatch) and established a facile diagnostic workflow. We discovered that EEF1A1 levels, abundantly expressed on urinary exosomes, positively correlate to urinary exosome counts irrespective of source and collection time and demonstrated that EEF1A1 enables in-assay quantification of urinary exosomes. Importantly, a prostate cancer urinary EVLatch-based artificial intelligence diagnostics (PURE-AID) classification system utilizing PCA3, HOXC6, and DLX1 as targets with SPDEF for reference and EEF1A1 for quality checking, trained on 271 patients, achieved an area under the receiver operating characteristic curve (AUROC) of 0.76 in the test set of 351 patients. Combination of PURE-AID with prostate-specific antigen (PSA) and age increases AUROC to 0.80 and reduces 54.3% of unnecessary biopsies with 86.8% sensitivity. Our study provides a new classification system for differentiating high-grade PCa in a workflow- and patient-friendly manner.