Scholarlink

Background <p>The incidence of pediatric inflammatory bowel disease (IBD) rises yearly. Infliximab, a cornerstone of treatment, often has limited efficacy owing to high pharmacokinetic variability. Therapeutic drug monitoring and model-based precision dosing (MIPD) can improve optimal drug exposure.</p> Objective <p>This study aimed to evaluate the predictive performance of available pediatric infliximab population pharmacokinetic (popPK) models implemented in specialized software during the induction phase of IBD treatment.</p> Methods <p>This retrospective observational cohort study included patients from January 2021 to December 2024. Plasma trough concentrations of infliximab were measured using an enzyme immunoassay. Pharmacokinetic simulations were performed with TDMx software, applying eight different pediatric popPK models. Metrics used to evaluate predictive performance included root mean squared error (RMSE), mean squared error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), coefficient of determination, and relative bias.</p> Results <p>A total of 29 patients were included (Crohn’s disease, <i>n</i>&#xa0;=&#xa0;24; ulcerative colitis, <i>n</i>&#xa0;=&#xa0;5; 52% female). Median (range) age was 12 (1–18) years and weight was 32 (7–62) kg. During the induction phase, patients received a median infliximab dose of 6.1 (4.6–10.2) mg/kg per infusion. The model by Dubinsky showed the best performance, including MSE, RMSE, MAE, and R<sup>2</sup>, followed by the models of Wojciechowski and Xiong.</p> Conclusions and Relevance <p>Pharmacotherapeutic assistance using MIPD software for infliximab precision dosing during induction in pediatric IBD is feasible and sufficiently accurate for clinical practice. In our population, the Dubinsky model was optimal for integration into dosing systems, balancing predictive performance with reliable PK representation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluation of Predictive Pharmacokinetic Models to Optimize Infliximab Therapy in Pediatric Inflammatory Bowel Disease

  • 期刊论文

  • 发表日期：

    2025年11月3日

查看全文
  • Paulo Caceres Guido,
  • Guillermo Federico Taboada,
  • Gisela Gruber,
  • Franco García,
  • Laura Pérez,
  • Fernanda Quinteros,
  • David Fabbrini,
  • Mónica Contreras

摘要

Background

The incidence of pediatric inflammatory bowel disease (IBD) rises yearly. Infliximab, a cornerstone of treatment, often has limited efficacy owing to high pharmacokinetic variability. Therapeutic drug monitoring and model-based precision dosing (MIPD) can improve optimal drug exposure.

Objective

This study aimed to evaluate the predictive performance of available pediatric infliximab population pharmacokinetic (popPK) models implemented in specialized software during the induction phase of IBD treatment.

Methods

This retrospective observational cohort study included patients from January 2021 to December 2024. Plasma trough concentrations of infliximab were measured using an enzyme immunoassay. Pharmacokinetic simulations were performed with TDMx software, applying eight different pediatric popPK models. Metrics used to evaluate predictive performance included root mean squared error (RMSE), mean squared error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), coefficient of determination, and relative bias.

Results

A total of 29 patients were included (Crohn’s disease, n = 24; ulcerative colitis, n = 5; 52% female). Median (range) age was 12 (1–18) years and weight was 32 (7–62) kg. During the induction phase, patients received a median infliximab dose of 6.1 (4.6–10.2) mg/kg per infusion. The model by Dubinsky showed the best performance, including MSE, RMSE, MAE, and R2, followed by the models of Wojciechowski and Xiong.

Conclusions and Relevance

Pharmacotherapeutic assistance using MIPD software for infliximab precision dosing during induction in pediatric IBD is feasible and sufficiently accurate for clinical practice. In our population, the Dubinsky model was optimal for integration into dosing systems, balancing predictive performance with reliable PK representation.