Optimal Sampling Strategy and Threshold of Serum Vancomycin Concentration in Elderly Japanese Patients Undergoing High-Flux Hemodialysis.

Background: The optimal sampling points and thresholds for initial serum vancomycin (VCM) concentrations have not been determined in hemodialysis (HD) patients. To clarify this, multiple blood tests were performed, and the correlations between VCM concentrations at several sampling points and the area under the concentration-time curve for 24 hours (AUC24h) were analyzed.

Methods: A single-center, prospective observational study was conducted. Patients with end-stage renal failure who received VCM treatment while undergoing chronic maintenance HD were enrolled in this study. HD was performed using a high-flux membrane as the dialyzer. After VCM administration, 7 points were sampled between the first and second HD. The AUC24h after the end of the first HD (AUC0-24) and that before the end of the second HD (AUC24-48) were calculated using the linear trapezoidal method. Correlation analysis and simple regression analysis between AUC24h and serum concentrations were performed at each sampling point.

Results: Nine patients were evaluated. Strong correlations were found between AUC24-48 and serum concentrations at 24 hours after the initiation of VCM treatment following the first HD (C24h, R = 0.983 and P < 0.001), between AUC0-24 and C24h (R = 0.967 and P < 0.001), and between AUC24-48 and serum concentration just before the second HD (Cpre(HD2), R = 0.965 and P < 0.001). Regression equations with high coefficients of determination (R2 > 0.9) were obtained, and a C24h of ≥18.0 mg/L and a Cpre(HD2) of ≥16.5 mg/L were required to achieve an AUC24-48 value of ≥400 mg·h/L. In addition, a C24h of ≤23.3 mg/L was estimated to satisfy the AUC0-24 range of ≤600 mg·h/L.

Conclusions: C24h and Cpre(HD2) are optimal sampling points for predicting VCM-AUC24h in HD patients.