Multi‐model averaging improves the performance of model‐guided infliximab dosing in patients with inflammatory bowel diseases

Abstract

Infliximab dosage de-escalation without prior knowledge of drug concentrations may put patients at risk for underexposure and trigger the loss of response. A single-model approach for model-informed precision dosing during infliximab maintenance therapy has proven its clinical benefit in patients with inflammatory bowel diseases. We evaluated the predictive performances of two multimodel approaches, a model selection algorithm and a model averaging algorithm, using 18 published population pharmacokinetic models of infliximab for guiding dosage de-escalation. Data of 54 patients with Crohn's disease and ulcerative colitis who underwent infliximab dosage de-escalation after an earlier escalation were used. A priori prediction (based solely on covariate data) and maximum a posteriori prediction (based on covariate data and trough concentrations) were compared using accuracy and precision metrics and the classification accuracy at the trough concentration target of 5.0 mg/L. A priori prediction was inaccurate and imprecise, with the lowest classification accuracies irrespective of the approach (median 59%, interquartile range 59%-63%). Using the maximum a posteriori prediction, the model averaging algorithm had systematically better predictive performance than the model selection algorithm or the single-model approach with any model, regardless of the number of concentration data. Only a single trough concentration (preferably at the point of care) sufficed for accurate and precise prediction. Predictive performance of both single-and multi-model approaches was robust to the lack of covariate data. Model averaging using four models demonstrated similar predictive performance with a five-fold shorter computation time. This model averaging algorithm was implemented in the TDMx software tool to guide infliximab dosage de-escalation in the forthcoming prospective MODIFI study (NCT04982172).

Best Practice for Therapeutic Drug Monitoring of Infliximab: Position Statement from the International Association of Therapeutic Drug Monitoring and Clinical Toxicology

“…MIPD of infliximab based solely on covariates (a priori prediction) has been shown to be biased and imprecise. 169,170 By incorporating 1 or more measured drug concentrations, the precision and accuracy of MIPD can be significantly improved. [169][170][171] This improvement is expected because covariates generally explain only a small part of the total interindividual variability (IIV, up to 6% for clearance), whereas Bayesian forecasting based on drug concentrations can identify the remaining, often high, "unexplained" IIV (median of 32.7%, interquartile range 28.0%-36.0% on clearance).…”

“…169,170 By incorporating 1 or more measured drug concentrations, the precision and accuracy of MIPD can be significantly improved. [169][170][171] This improvement is expected because covariates generally explain only a small part of the total interindividual variability (IIV, up to 6% for clearance), whereas Bayesian forecasting based on drug concentrations can identify the remaining, often high, "unexplained" IIV (median of 32.7%, interquartile range 28.0%-36.0% on clearance). 172,173 Second, there are significant efforts underway to enhance the methodological components of MIPD, focusing on methods for model selection 171,174,175 and the estimation of PK parameters.…”

“…Finally, the predictive ability of MIPD is typically described by criteria like classification accuracy (eg, the ability to predict the next TC to be higher or lower than a specific threshold) or the difference between predicted and measured values (eg, root mean square error or relative bias). 169,171,175 However, there is a lack of consensus on how to evaluate the success of MIPD, and further investigation into the impact of different evaluation criteria on choosing a suboptimal model is needed.…”

“…Among the published population PK models of infliximab, ATI status, serum albumin, and body weight were the most frequently identified covariates on infliximab clearance, with body weight being the most commonly identified covariate on volumes of distribution. 169 However, the clinical relevance of these statistically significant covariates to therapeutic target attainment remains undetermined. MIPD of infliximab based solely on covariates (a priori prediction) has been shown to be biased and imprecise.…”

“…The findings demonstrated more reliable Bayesian forecasting compared with using a single-model approach. In a recent retrospective study by Kantasiripitak et al, 169 the predictive performance of MIPD of infliximab based on the MAA/MSA algorithm was evaluated in adult patients with IBD undergoing infliximab dose de-escalation. The results indicated that an MAA resulted in the most accurate and precise a posteriori prediction compared with the MSA and a single-model approach.…”