The aim of this work is to estimate the area‐under the blood concentration curve of tacrolimus (TAC) following b.i.d. or q.d. dosing in organ transplant patients, using Xgboost machine learning (ML) models. A total of 4,997 and 1,452 TAC interdose area under the curves (AUCs) from patients on b.i.d. and q.d. TAC, sent to our Immunosuppressant Bayesian Dose Adjustment expert system (http://www.pharmaco.chu-limoges.fr/) for AUC estimation and dose recommendation based on TAC concentrations measured at least at 3 sampling times (predose, ~ 1 and 3 hours after dosing) were used to develop 4 ML models based on 2 or 3 concentrations. For each model, data splitting was performed to obtain a training set (75%) and a test set (25%). The Xgboost models in the training set with the lowest root mean square error (RMSE) in a 10‐fold cross‐validation experiment were evaluated in the test set and in 6 independent full‐pharmacokinetic (PK) datasets from renal, liver, and heart transplant patients. ML models based on two or three concentrations, differences between these concentrations, relative deviations from theoretical times of sampling, and four covariates (dose, type of transplantation, age, and time between transplantation and sampling) yielded excellent AUC estimation performance in the test datasets (relative bias < 5% and relative RMSE < 10%) and better performance than maximum a posteriori Bayesian estimation in the six independent full‐PK datasets. The Xgboost ML models described allow accurate estimation of TAC interdose AUC and can be used for routine TAC exposure estimation and dose adjustment. They will soon be implemented in a dedicated web interface.

Study HighlightsWHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC? Mycophenolic acid (MPA) area under the curve from 0 to 12 hours (AUC 0-12 h ) for individual dose adjustment is difficult to estimate in routine patient care. WHAT QUESTION DID THIS STUDY ADDRESS?  We investigated whether machine learning (ML) models could estimate MPA AUC 0-12 h using a limited number of blood concentrations, as well as or even better than deterministic pharmacokinetic (PK) models with Bayesian estimation. WHAT DOES THIS STUDY ADD TO OUR KNOW-LEDGE? We developed and validated in kidney or heart transplants extreme gradient boosting (Xgboost R package) ML models allowing accurate estimation of MPA AUC 0-12 h based on three blood concentrations with better performances than that of the PK approach previously used. HOW MIGHT THIS CHANGE CLINICAL PHARMA-COLOGY OR TRANSLATIONAL SCIENCE?  These models will be soon implemented in an expert system made available to the transplant community through a dedicated website.

No abstract