A Hybrid Model Associating Population Pharmacokinetics with Machine Learning: A Case Study with Iohexol Clearance Estimation.
Journal
Clinical pharmacokinetics
ISSN: 1179-1926
Titre abrégé: Clin Pharmacokinet
Pays: Switzerland
ID NLM: 7606849
Informations de publication
Date de publication:
08 2022
08 2022
Historique:
accepted:
05
05
2022
pubmed:
1
6
2022
medline:
6
8
2022
entrez:
31
5
2022
Statut:
ppublish
Résumé
Maximum a posteriori Bayesian estimation (MAP-BE) based on a limited sampling strategy and a population pharmacokinetic model is frequently used to estimate pharmacokinetic parameters in individuals, however with some uncertainty (bias). Recent works have shown that the performance in individual estimation or pharmacokinetic parameters can be improved by combining population pharmacokinetic and machine learning algorithms. The objective of this work was to investigate the use of a hybrid machine learning/population pharmacokinetic approach to improve individual iohexol clearance estimation. The reference iohexol clearance values were derived from 500 simulated profiles (samples collected between 0.1 and 24.7 h) using a population pharmacokinetic model we recently developed in Monolix and obtained using all the concentration timepoints available. Xgboost and glmnet algorithms able to predict the error of MAP-BE clearance estimates based on a limited sampling strategy (0.1 h, 1 h, and 9 h) versus reference values were developed in a training subset (75%) and were evaluated in a testing subset (25%) and in 36 real patients. The MAP-BE limited sampling strategy estimated clearance was corrected by the machine learning predicted error leading to a decrease in root mean squared error by 29% and 24%, and in the percentage of profiles with the mean prediction error out of the ± 20% bias by 60% and 40% in the external validation dataset for the glmnet and Xgboost machine learning algorithms, respectively. These results were attributable to a decrease in the eta-shrinkage (shrinkage for a MAP-BE limited sampling strategy = 32.4%, glmnet = 18.2%, and Xgboost = 19.4% in the external dataset). In conclusion, this hybrid algorithm represents a significant improvement in comparison to MAP-BE estimation alone.
Sections du résumé
BACKGROUND
Maximum a posteriori Bayesian estimation (MAP-BE) based on a limited sampling strategy and a population pharmacokinetic model is frequently used to estimate pharmacokinetic parameters in individuals, however with some uncertainty (bias). Recent works have shown that the performance in individual estimation or pharmacokinetic parameters can be improved by combining population pharmacokinetic and machine learning algorithms.
OBJECTIVE
The objective of this work was to investigate the use of a hybrid machine learning/population pharmacokinetic approach to improve individual iohexol clearance estimation.
METHODS
The reference iohexol clearance values were derived from 500 simulated profiles (samples collected between 0.1 and 24.7 h) using a population pharmacokinetic model we recently developed in Monolix and obtained using all the concentration timepoints available. Xgboost and glmnet algorithms able to predict the error of MAP-BE clearance estimates based on a limited sampling strategy (0.1 h, 1 h, and 9 h) versus reference values were developed in a training subset (75%) and were evaluated in a testing subset (25%) and in 36 real patients.
RESULTS
The MAP-BE limited sampling strategy estimated clearance was corrected by the machine learning predicted error leading to a decrease in root mean squared error by 29% and 24%, and in the percentage of profiles with the mean prediction error out of the ± 20% bias by 60% and 40% in the external validation dataset for the glmnet and Xgboost machine learning algorithms, respectively. These results were attributable to a decrease in the eta-shrinkage (shrinkage for a MAP-BE limited sampling strategy = 32.4%, glmnet = 18.2%, and Xgboost = 19.4% in the external dataset).
CONCLUSIONS
In conclusion, this hybrid algorithm represents a significant improvement in comparison to MAP-BE estimation alone.
Identifiants
pubmed: 35641861
doi: 10.1007/s40262-022-01138-x
pii: 10.1007/s40262-022-01138-x
doi:
Substances chimiques
Iohexol
4419T9MX03
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
1157-1165Informations de copyright
© 2022. The Author(s), under exclusive licence to Springer Nature Switzerland AG.
Références
Leger F, Debord J, Le Meur Y, Rousseau A, Buchler M, Lachatre G, et al. Maximum a posteriori Bayesian estimation of oral cyclosporin pharmacokinetics in patients with stable renal transplants. Clin Pharmacokinet. 2002;41:71–80.
doi: 10.2165/00003088-200241010-00006
Marquet P, Destère A, Monchaud C, Rérolle J-P, Buchler M, Mazouz H, et al. Clinical pharmacokinetics and Bayesian estimators for the individual dose adjustment of a generic formulation of tacrolimus in adult kidney transplant recipients. Clin Pharmacokinet. 2021;60:611–22.
doi: 10.1007/s40262-020-00959-y
Guo W, Yu Z, Gao Y, Lan X, Zang Y, Yu P, et al. A machine learning model to predict risperidone active moiety concentration based on initial therapeutic drug monitoring. Front Psychiatry. 2021;12: 711868.
doi: 10.3389/fpsyt.2021.711868
Woillard J, Labriffe M, Debord J, Marquet P. Tacrolimus exposure prediction using machine learning. Clin Pharmacol Ther. 2021;110:361–9.
doi: 10.1002/cpt.2123
Sibieude E, Khandelwal A, Girard P, Hesthaven JS, Terranova N. Population pharmacokinetic model selection assisted by machine learning. J Pharmacokinet Pharmacodyn. 2022;49:257–70.
doi: 10.1007/s10928-021-09793-6
Koch G, Pfister M, Daunhawer I, Wilbaux M, Wellmann S, Vogt JE. Pharmacometrics and machine learning partner to advance clinical data analysis. Clin Pharmacol Ther. 2020;107:926–33.
doi: 10.1002/cpt.1774
Tang B-H, Guan Z, Allegaert K, Wu Y-E, Manolis E, Leroux S, et al. Drug clearance in neonates: a combination of population pharmacokinetic modelling and machine learning approaches to improve individual prediction. Clin Pharmacokinet. 2021;60:1435–48.
doi: 10.1007/s40262-021-01033-x
Hughes JH, Keizer RJ. A hybrid machine learning/pharmacokinetic approach outperforms maximum a posteriori Bayesian estimation by selectively flattening model priors. CPT Pharmacometrics Syst Pharmacol. 2021;10:1150–60.
doi: 10.1002/psp4.12684
Baklouti S, Gandia P, Concordet D. “De-shrinking” EBEs: the solution for Bayesian therapeutic drug monitoring. Clin Pharmacokinet. 2022;61:749–57.
doi: 10.1007/s40262-021-01105-y
Åsberg A, Bjerre A, Almaas R, Luis-Lima S, Robertsen I, Salvador CL, et al. Measured GFR by utilizing population pharmacokinetic methods to determine iohexol clearance. Kidney Int Rep. 2020;5:189–98.
doi: 10.1016/j.ekir.2019.11.012
Benz-de Bretagne I, Le Guellec C, Halimi JM, Gatault P, Barbet C, Alnajjar A, et al. New sampling strategy using a Bayesian approach to assess iohexol clearance in kidney transplant recipients. Ther Drug Monit. 2012;34:289–97.
doi: 10.1097/FTD.0b013e31824a6534
Taubert M, Ebert N, Martus P, van der Giet M, Fuhr U, Schaeffner E. Using a three-compartment model improves the estimation of iohexol clearance to assess glomerular filtration rate. Sci Rep. 2018;8:17723.
doi: 10.1038/s41598-018-35989-x
Taubert M, Schaeffner E, Martus P, van der Giet M, Fuhr U, Lösment A, et al. Advancement of pharmacokinetic models of iohexol in patients aged 70 years or older with impaired kidney function. Sci Rep. 2021;11:22656.
doi: 10.1038/s41598-021-01892-1
Salmon Gandonnière C, Helms J, Le Tilly O, Benz-de Bretagne I, Bretagnol A, Bodet-Contentin L, et al. Glomerular hyper- and hypofiltration during acute circulatory failure: iohexol-based gold-standard descriptive study. Crit Care Med. 2019;47:e623–9.
doi: 10.1097/CCM.0000000000003804
Woillard J-B, Salmon Gandonnière C, Destere A, Ehrmann S, Merdji H, Mathonnet A, et al. A machine learning approach to estimate the glomerular filtration rate in intensive care unit patients based on plasma iohexol concentrations and covariates. Clin Pharmacokinet. 2021;60:223–33.
doi: 10.1007/s40262-020-00927-6
Destere A, Gandonnière CS, Åsberg A, Loustaud-Ratti V, Carrier P, Ehrmann S, et al. A single Bayesian estimator for iohexol clearance estimation in ICU, liver failure and renal transplant patients. Br J Clin Pharmacol. 2021;88:15197.
SimulX version 2020R1. Antony: Lixoft SAS; 2019. http://lixoft.com/products/monolix/ . Accessed 18 May 2022.
Kuhn M, Wickham H. Tidymodels: a collection of packages for modeling and machine learning using tidyverse principles. 2020. https://www.tidymodels.org . Accessed 18 May 2022.
Chen T, Guestrin C. XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco (CA): ACM; 2016: 785–94. https://doi.org/10.1145/2939672.2939785 . Accessed 24 Dec 2021.
Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33:1–22.
doi: 10.18637/jss.v033.i01
Monolix version 2018R1. Antony: Lixoft SAS; 2018. http://lixoft.com/products/monolix/ . Accessed 18 May 2022.
Woillard J, Labriffe M, Debord J, Marquet P. Mycophenolic acid exposure prediction using machine learning. Clin Pharmacol Ther. 2021;110:370–9.
doi: 10.1002/cpt.2216