Paper Details
- Yuki Narita (Department of Clinical Pharmaceutical Sciences, Graduate School of Pharmaceutical Sciences, Kumamoto University / Department of Pharmacy, Kumamoto University Hospital / Faculty of Pharmaceutical Sciences, Sojo University / y-nari@ph.sojo-u.ac.jp)
1) Department of Clinical Pharmaceutical Sciences, Graduate School of Pharmaceutical Sciences, Kumamoto University , 2) Department of Pharmacy, Kumamoto University Hospital , 3) Faculty of Pharmaceutical Sciences, Sojo University , 4) Division of Pharmacology and Therapeutics, Graduate School of Pharmaceutical Sciences, Kumamoto University , 5) Department of Urology, Kamiamakusa General Hospital
Patients undergoing hemodialysis face a significantly elevated risk of cardiovascular disease (CVD) mortality. Uremic toxins, particularly indoxyl sulfate (IS) and p-cresyl sulfate (PCS), accumulate in patients undergoing dialysis and contribute to cardiovascular complications. Although IS and PCS concentrations are often assessed individually, this pilot study developed and validated an integrated scoring system combining these values to improve CVD mortality prediction. This prospective study included 66 patients undergoing hemodialysis at the Hitoyoshi Medical Center between 2008–2018. The IS + PCS score was calculated using Cox proportional hazards analysis. The primary endpoint was CVD death. The IS + PCS score was strongly associated with CVD mortality. IS alone exhibited high sensitivity (91.67%) but low specificity (44.44%), whereas PCS alone showed high specificity (92.59%) but low sensitivity (25.00%). The IS + PCS score achieved balanced sensitivity (83.33%) and specificity (57.41%), with a higher F1 score (0.44) than that of IS alone (0.41) and PCS alone (0.32). Leave-one-out cross-validation confirmed internal validity (mean C-index: 0.645). The IS + PCS score demonstrated potential as a prognostic predictor of CVD mortality in patients undergoing hemodialysis. While the score achieved balanced sensitivity and specificity, with an F1 score exceeding that of the individual markers, time-dependent receiver operating characteristic analysis showed no superiority over IS alone at any time point. These findings suggest that the integrated score combines the complementary strengths of both toxins, but requires validation in larger cohorts to establish clinical superiority.

