English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

We have previously published the results of an infliximab (IFX) population PK model in children and young adults with Crohn’s disease (CD). We found that prediction accuracy was improved with less unexplained variability when the five covariates of drug clearance were included. We also developed a clinical decision support tool (RoadMABTM) to promote physician driven model-informed precision dosing. As IFX has a relatively long half-life, it is currently unknown under what conditions (Bayesian posterior) model performance would be improved. In this study, the aim is to evaluate the precision and bias of three population PK models in a real-world cohort of CD patients. METHODS We investigated CD subjects who were anti-TNF naïve and started IFX induction between 2019-2020 at Cincinnati Children’s Hospital Medical Center. All subjects had a minimum of four IFX doses and a minimum of two IFX concentrations measured with a drug-tolerant assay (Esoterix, LabCorp specialty lab, Calabasas, CA) as a part of their clinical care. We measured the mean predictive error (MPE) and the root mean squared error (RMSE) for describing the bias and precision, respectively, of the prediction of the most recent IFX level (which was excluded from the Bayesian estimation). Predictions were based on the selected PK mode, the available covariates of drug clearance (weight, serum albumin, erythrocyte sedimentation rate, neutrophil CD64 expression and antibody to IFX concentration) and prior IFX levels. Two modes were used to compare model performance. In RoadMABTM, the poly-exponential (POLY-1) mode enables last known covariates (if &amp;lt;182 days) and last known IFX level. The ordinary differential equations (ODE-N) mode enables all covariates within the last 182 days and the last N (-1, -2, -3) IFX levels (Figure 1). RESULTS We identified 105 subjects who met study criteria. The median (IQR) age at first IFX infusion was 14.6 years (11.8-16.8). The median (IQR) starting dose was 8.6 mg/kg (5.4-10) and 47% were female. For all 3 PK models, we found minimal bias and improved precision in the ODE-1 mode (Table1). While there was little difference in the precision (RMSE) when the POLY-1 mode was used, we found an improvement in bias (MPE) in this mode. In a smaller subset (n=36), we also compared model performance to predict an infusion5 or infusion6 level when an infusion4 (week14) level was known. All three models performed equally well in the POLY-1 mode with the Xiong et al. model showing superior precision with ODE-1. Finally, we compared model performance to predict an infusion3 (week6) level from the start of IFX (n=38) based on the covariates of drug clearance alone. CONCLUSION Based on our data, in CD patients, precision dosing for IFX should be largely based on the last known IFX concentration and the most recent covariates of drug clearance.

PREDICTIVE PERFORMANCE OF DIFFERENT POPULATION PHARMACOKINETIC MODELS IN A REAL-WORLD COHORT OF CROHN’S DISEASE PATIENTS