British journal of clinical pharmacology · 2026
Accelerating tacrolimus model-informed precision dosing in kidney transplant recipients: Model evaluation and refinement strategies.
СтатьяСистематический обзорНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Umpiérrez Martín, Maldonado Cecilia, Guevara Natalia, Cardenas-Rodriguez Magdalena, Menéndez Clara, Seija Mariana, Santiago José, Noboa Oscar, Ibarra Manuel
Аннотация
AIM: To guide the efficient implementation of Model-Informed Precision Dosing (MIPD) by comparing the predictive performance of published tacrolimus population pharmacokinetic (popPK) models with an in-house model in Uruguayan kidney transplant recipients, using comprehensive external evaluation methodology.
METHODS: Twenty-five adult tacrolimus popPK models were identified through systematic review and externally evaluated using retrospective data from 39 patients (396 steady-state concentrations). In parallel, an in-house model was developed with prospective and retrospective data from 49 patients (727 concentrations). Predictive performance was assessed across different data integration levels: a priori predictions and sequential Bayesian forecasting with progressive incorporation of individual observations.
RESULTS: All published models and the in-house model showed poor a priori predictive performance, with even the best performing models yielding MAPE values near 50% and fewer than 30% of predictions within 30% of observed concentrations. Bayesian forecasting progressively improved bias, accuracy and dispersion of prediction errors. Initial incorporation of three trough concentrations produced limited benefit, but substantial improvements were observed upon integration of a pharmacokinetic curve composed of at least 3 post-dose concentrations, at which point several published models and the in-house model achieved high precision, with mean bias below 10% and MAPE below 30%. Two published models exhibited comparable performance to the in-house model, which achieved marginally better results as Bayesian forecasting incorporated successive individual observations, without providing a meaningful overall advantage.
CONCLUSION: Published tacrolimus popPK models can achieve clinically reliable predictions when combined with Bayesian updating, performing comparably to in-house developed models. These findings support a resource-efficient strategy for implementing tacrolimus MIPD without the need for local model development, provided that a pharmacokinetic curve is obtained early post-transplantation.
METHODS: Twenty-five adult tacrolimus popPK models were identified through systematic review and externally evaluated using retrospective data from 39 patients (396 steady-state concentrations). In parallel, an in-house model was developed with prospective and retrospective data from 49 patients (727 concentrations). Predictive performance was assessed across different data integration levels: a priori predictions and sequential Bayesian forecasting with progressive incorporation of individual observations.
RESULTS: All published models and the in-house model showed poor a priori predictive performance, with even the best performing models yielding MAPE values near 50% and fewer than 30% of predictions within 30% of observed concentrations. Bayesian forecasting progressively improved bias, accuracy and dispersion of prediction errors. Initial incorporation of three trough concentrations produced limited benefit, but substantial improvements were observed upon integration of a pharmacokinetic curve composed of at least 3 post-dose concentrations, at which point several published models and the in-house model achieved high precision, with mean bias below 10% and MAPE below 30%. Two published models exhibited comparable performance to the in-house model, which achieved marginally better results as Bayesian forecasting incorporated successive individual observations, without providing a meaningful overall advantage.
CONCLUSION: Published tacrolimus popPK models can achieve clinically reliable predictions when combined with Bayesian updating, performing comparably to in-house developed models. These findings support a resource-efficient strategy for implementing tacrolimus MIPD without the need for local model development, provided that a pharmacokinetic curve is obtained early post-transplantation.