Metabolomics in Prostate Cancer: A Minimally Invasive Method Using Urine after Prostatic Massage to Predict Gleason Grade.

[BACKGROUND] In this study, biomarkers that can predict prostate cancer with a Gleason grade of 8 or higher were explored through nuclear magnetic resonance (NMR).

[METHODS] Patients scheduled for transrectal prostate biopsy were enrolled, and urine samples were collected after prostate massage. Patients with cancer were categorised as having Gleason grades of 6-7 or ≥8. All spectra were acquired using a Bruker Avance III DRX 600 spectrometer. For statistical analysis, univariate and multivariate analyses were conducted using metabolites and clinical variables, and the presence of tumours with Gleason grades of ≥8 was predicted.

[RESULTS] Data were obtained from 107 patients with prostate cancer: 73 (68.2%) with Gleason grades of 6-7 and 34 (31.8%) with Gleason grades of ≥8. A predictive model incorporating the 29 most significant metabolites identified through partial least squares-discriminant analysis was established. Suspicious digital rectal examination (DRE) results were considered. The model predicted a Gleason grade of ≥8, demonstrating an area under the curve of 0.92, sensitivity of 82%, specificity of 92%, positive predictive value of 84% and negative predictive value of 90%. Metabolites associated with amino acid metabolism and glycolysis were prominent in this model.

[CONCLUSIONS] Our study demonstrates that a model combining urinary metabolites with clinical data, specifically DRE findings, can effectively stratify risk in patients with biopsy-confirmed prostate cancer according to Gleason grade. Metabolites linked to glycolysis and amino acid metabolism were particularly relevant. This minimally invasive approach may assist clinical decision-making, although validation in larger multi-centre cohorts is required to confirm its robustness and generalisability.