Urinary creatine riboside for risk stratification of clinically significant prostate cancer.

Abstract

e17149 Background: Distinguishing clinically significant prostate cancer from indolent disease remains a major challenge in prostate cancer diagnosis and contributes to unnecessary biopsy and overtreatment. Prostate-specific antigen (PSA) testing, while widely used, has limited specificity for clinically significant disease and provides limited information for individualized risk assessment. Creatine riboside is a tumor-associated urinary metabolite previously linked to aggressive prostate cancer biology. We evaluated whether urinary creatine riboside could provide clinically interpretable risk stratification for clinically significant prostate cancer using a likelihood-ratio–based framework designed to complement existing diagnostic approaches and inform decision-making beyond PSA alone. Methods: Urinary creatine riboside concentrations were quantified using targeted liquid chromatography–mass spectrometry in men undergoing evaluation for prostate cancer. Clinically significant prostate cancer was defined according to standard pathological criteria. Risk strata were prespecified using percentile thresholds derived from the empirical distribution of urinary creatine riboside concentrations. Diagnostic performance across strata was evaluated using sensitivity, specificity, and positive likelihood ratios (LR+), interpreted according to established clinical benchmarks to define low-risk, intermediate-risk, and high-risk categories. Results: Urinary creatine riboside concentrations were higher among men with clinically significant prostate cancer than among those without. Diagnostic discrimination increased progressively across ascending percentile thresholds. Lower-risk strata were associated with minimal changes in post-test probability (LR+ approximately 1–2), consistent with low-risk classification. Intermediate-risk strata demonstrated moderate enrichment of risk (LR+ approximately 2–5). At higher percentile thresholds, urinary creatine riboside demonstrated strong rule-in performance, with LR+ values exceeding 10 and reaching greater than 30 at the highest thresholds. This likelihood-ratio–based framework enabled clear separation of patients into clinically actionable low-risk, intermediate-risk, and high-risk categories using a single noninvasive biomarker. Conclusions: Urinary creatine riboside enables robust, likelihood-ratio–based risk stratification for clinically significant prostate cancer, providing clinically interpretable decision support beyond PSA. This approach has the potential to reduce unnecessary biopsies while improving identification of aggressive disease and advancing precision prostate cancer diagnostics.