Prediction of Vancomycin Area Under the Curve With Trough Concentrations Only: Performance Evaluation of Pediatric Population Pharmacokinetic Models

Background. In pediatric patients, vancomycin plays a pivotal role in combating infections, necessitating precise therapeutic drug monitoring to ensure efficacy and safety. The adoption of model-informed precision dosing (MIPD) has demonstrated potential in optimizing dosing strategies based on the 24-hour area under the concentration-time curve (AUC(24h)). However, the predictive performance of population pharmacokinetic models by using only trough concentrations to estimate AUC(24h) has not been evaluated. Methods. The predictive performance of 23 vancomycin population pharmacokinetic models was retrospectively evaluated in 2 cohorts: cohort A, 21 patients with postnatal age <50 days; cohort B, 124 patients with postnatal age >= 50 days. Multiple scenarios were investigated by using peak and trough concentration, peak concentration solely, trough concentration solely, or covariate information (a priori). The median AUC(24h) per patient across all models was used as the true AUC(24h). Results. For both cohorts, the relative root mean square error (rRMSE) for the AUC(24h) precision based on trough concentrations was similar to the rRMSE based on a peak and trough sample. For cohort A, the models by Chen, Colin, and Mehrotra showed the best trough-based performance with the lowest relative bias (range, -3.3% to -2.6%) and rRMSE (range, 6.8%-7.3%). For cohort B, the models from Alsultan and Lv illustrated the lowest relative bias (range, 1.75% to -5.4%) and rRMSE (range, 16.6%-15.1%). Conclusions. This study illustrates that trough concentration-based AUC(24h) estimation is a feasible approach in vancomycin MIPD. These findings endorse the selected models for advanced MIPD vancomycin therapy in pediatrics, although further investigation into clinical outcomes is recommended.