Observational analysis identifies urinary miRNA profiles for cancer detection and prognosis in diverse cohorts, suggesting promising diagnostics.
Key Points
Machine learning algorithms distinguish urological cancers from healthy samples, achieving AUCs of 0.92 for RCC, 0.92 for PCa, and 0.96 for UC.
A panel of urinary miRNAs showed high discriminatory power for recurrence-free survival with time-dependent AUCs ranging from 0.75 to 0.89 across cancer types.
Prospective analysis included 419 urine samples from renal cell carcinoma, prostate cancer, and urothelial carcinoma, comparing with healthy individuals.
These findings highlight the utility of miRNA signatures as noninvasive diagnostic and prognostic tools for urological malignancies.