A Machine Learning Algorithm to Predict the Starting Dose of Daptomycin.


Journal

Clinical pharmacokinetics
ISSN: 1179-1926
Titre abrégé: Clin Pharmacokinet
Pays: Switzerland
ID NLM: 7606849

Informations de publication

Date de publication:
31 Jul 2024
Historique:
accepted: 16 07 2024
medline: 1 8 2024
pubmed: 1 8 2024
entrez: 31 7 2024
Statut: aheadofprint

Résumé

The dosage of daptomycin is usually based on body weight. However, it has been shown that this approach yields too high an exposure in obese patients. Pharmacokinetic and pharmacodynamic indexes (PK/PD) have been proposed for daptomycin's antibacterial effect (AUC/CMI >666) and toxicity (C0 > 24.3 mg/L). We previously developed machine learning (ML) algorithms to predict starting doses based on Monte Carlo simulations. We propose a new way to perform probability of target attainment based on an ML algorithm to predict the daptomycin starting dose. The Dvorchik model of daptomycin was implemented in the mrgsolve R package and 4950 pharmacokinetic profiles were simulated with doses ranging from 4 to 12 mg/kg. We trained and benchmarked four machine learning algorithms and selected the best to iteratively search for the optimal dose of daptomycin maximizing the event (AUC/CMI > 666 and C0 < 24.3 mg/L). The ML algorithm was evaluated in simulations and an external database of real patients in comparison with population pharmacokinetics. The performance of the Xgboost algorithms developed to predict the event (ROC AUC) in the training and test set were 0.762 and 0.761, respectively. The most important prediction variables were dose, creatinine clearance, body weight and sex. In the external database of real patients, the starting dose administered based on the ML algorithm significantly improved the target attainment by 7.9% (p-value = 0.02929) in comparison with the dose administered based on body weight. The developed algorithm improved the target attainment for daptomycin in comparison with weight-based dosing. We built a Shiny app to calculate the optimal starting dose.

Sections du résumé

BACKGROUND AND OBJECTIVE OBJECTIVE
The dosage of daptomycin is usually based on body weight. However, it has been shown that this approach yields too high an exposure in obese patients. Pharmacokinetic and pharmacodynamic indexes (PK/PD) have been proposed for daptomycin's antibacterial effect (AUC/CMI >666) and toxicity (C0 > 24.3 mg/L). We previously developed machine learning (ML) algorithms to predict starting doses based on Monte Carlo simulations. We propose a new way to perform probability of target attainment based on an ML algorithm to predict the daptomycin starting dose.
METHODS METHODS
The Dvorchik model of daptomycin was implemented in the mrgsolve R package and 4950 pharmacokinetic profiles were simulated with doses ranging from 4 to 12 mg/kg. We trained and benchmarked four machine learning algorithms and selected the best to iteratively search for the optimal dose of daptomycin maximizing the event (AUC/CMI > 666 and C0 < 24.3 mg/L). The ML algorithm was evaluated in simulations and an external database of real patients in comparison with population pharmacokinetics.
RESULTS RESULTS
The performance of the Xgboost algorithms developed to predict the event (ROC AUC) in the training and test set were 0.762 and 0.761, respectively. The most important prediction variables were dose, creatinine clearance, body weight and sex. In the external database of real patients, the starting dose administered based on the ML algorithm significantly improved the target attainment by 7.9% (p-value = 0.02929) in comparison with the dose administered based on body weight.
CONCLUSION CONCLUSIONS
The developed algorithm improved the target attainment for daptomycin in comparison with weight-based dosing. We built a Shiny app to calculate the optimal starting dose.

Identifiants

pubmed: 39085523
doi: 10.1007/s40262-024-01405-z
pii: 10.1007/s40262-024-01405-z
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Subventions

Organisme : Agence Nationale de la Recherche
ID : ANR-22-PESN-0017

Informations de copyright

© 2024. The Author(s), under exclusive licence to Springer Nature Switzerland AG.

Références

Résumé des caractéristiques du produit - DAPTOMYCINE ACCORD 350 mg, poudre pour solution injectable/pour perfusion - Base de données publique des médicaments [Internet]. [cited 2023 Jan 24]. Available from: https://base-donnees-publique.medicaments.gouv.fr/affichageDoc.php?specid=69082149&typedoc=R
Benvenuto M, Benziger DP, Yankelev S, Vigliani G. Pharmacokinetics and tolerability of daptomycin at doses up to 12 milligrams per kilogram of body weight once daily in healthy volunteers. Antimicrob Agents Chemother. 2006;50:3245–9.
doi: 10.1128/AAC.00247-06 pubmed: 17005801 pmcid: 1610083
Jones TW, Jun AH, Michal JL, Olney WJ. High-dose daptomycin and clinical applications. Ann Pharmacother. 2021;55:1363–78.
doi: 10.1177/1060028021991943 pubmed: 33535792 pmcid: 8573721
Safdar N, Andes D, Craig WA. In vivo pharmacodynamic activity of daptomycin. Antimicrob Agents Chemother. 2004;48:63–8.
doi: 10.1128/AAC.48.1.63-68.2004 pubmed: 14693519 pmcid: 310158
Falcone M, Russo A, Cassetta MI, Lappa A, Tritapepe L, d’Ettorre G, et al. Variability of pharmacokinetic parameters in patients receiving different dosages of daptomycin: is therapeutic drug monitoring necessary? J Infect Chemother. 2013;19:732–9.
doi: 10.1007/s10156-013-0559-z pubmed: 23361566
Bhavnani SM, Rubino CM, Ambrose PG, Drusano GL. Daptomycin exposure and the probability of elevations in the creatine phosphokinase level: data from a randomized trial of patients with bacteremia and endocarditis. Clin Infect Dis Off Publ Infect Dis Soc Am. 2010;50:1568–74.
doi: 10.1086/652767
Dvorchik BH, Damphousse D. The pharmacokinetics of daptomycin in moderately obese, morbidly obese, and matched nonobese subjects. J Clin Pharmacol. 2005;45:48–56.
doi: 10.1177/0091270004269562 pubmed: 15601805
Pai MP, Norenberg JP, Anderson T, Goade DW, Rodvold KA, Telepak RA, et al. Influence of morbid obesity on the single-dose pharmacokinetics of daptomycin. Antimicrob Agents Chemother. 2007;51:2741–7.
doi: 10.1128/AAC.00059-07 pubmed: 17548489 pmcid: 1932544
Chagnac A, Weinstein T, Korzets A, Ramadan E, Hirsch J, Gafter U. Glomerular hemodynamics in severe obesity. Am J Physiol-Ren Physiol. 2000;278:F817–22.
doi: 10.1152/ajprenal.2000.278.5.F817
Butterfield-Cowper JM, Lodise TP, Pai MP. A fixed versus weight-based dosing strategy of daptomycin may improve safety in obese adults. Pharmacotherapy. 2018;38:981–5.
doi: 10.1002/phar.2157 pubmed: 29906315
Dvorchik B, Arbeit RD, Chung J, Liu S, Knebel W, Kastrissios H. Population pharmacokinetics of daptomycin. Antimicrob Agents Chemother. 2004;48:2799–807.
doi: 10.1128/AAC.48.8.2799-2807.2004 pubmed: 15273084 pmcid: 478529
Garreau R, Bricca R, Gagnieu M-C, Roux S, Conrad A, Bourguignon L, et al. Population pharmacokinetics of daptomycin in patients with bone and joint infection: minimal effect of rifampicin co-administration and confirmation of a sex difference. J Antimicrob Chemother. 2021;76:1250–7.
doi: 10.1093/jac/dkab006 pubmed: 33550409
Woillard J-B, Labriffe M, Prémaud A, Marquet P. Estimation of drug exposure by machine learning based on simulations from published pharmacokinetic models: the example of tacrolimus. Pharmacol Res. 2021;167: 105578.
doi: 10.1016/j.phrs.2021.105578 pubmed: 33775863
Ponthier L, Ensuque P, Destere A, Marquet P, Labriffe M, Jacqz-Aigrain E, et al. Optimization of vancomycin initial dose in term and preterm neonates by machine learning. Pharm Res. 2022;39:2497–506.
doi: 10.1007/s11095-022-03351-6 pubmed: 35918452
Elmokadem A, Riggs MM, Baron KT. Quantitative systems pharmacology and physiologically-based pharmacokinetic modeling with mrgsolve: a hands-on tutorial. CPT Pharmacomet Syst Pharmacol. 2019;8:883–93.
doi: 10.1002/psp4.12467
Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, et al. Welcome to the tidyverse. J Open Source Softw. 2019;4:1686.
doi: 10.21105/joss.01686
Kuhn M, Wickham H. Tidymodels: a collection of packages for modeling and machine learning using tidyverse principles. [Internet]. 2020. Available from: https://www.tidymodels.org
Tuloup V, Millet A, Taricco A, Parant F, Ferry T, Goutelle S. Evaluation of limited sampling strategies for bayesian estimation of daptomycin area under the concentration-time curve: a short communication. Ther Drug Monit. 2023;45:562–5.
doi: 10.1097/FTD.0000000000001070 pubmed: 36728573
Chaves RL, Chakraborty A, Benziger D, Tannenbaum S. Clinical and pharmacokinetic considerations for the use of daptomycin in patients with Staphylococcus aureus bacteraemia and severe renal impairment. J Antimicrob Chemother. 2014;69:200–10.
doi: 10.1093/jac/dkt342 pubmed: 24030545
Labriffe M, Woillard J, Debord J, Marquet P. Machine learning algorithms to estimate everolimus exposure trained on simulated and patient pharmacokinetic profiles. CPT Pharmacomet Syst Pharmacol. 2022;11:1018–28.
doi: 10.1002/psp4.12810

Auteurs

Florence Rivals (F)

Service de Pharmacologie, Toxicologie et Pharmacovigilance, CHU Limoges, France.

Sylvain Goutelle (S)

Service de Pharmacie, Hospices Civils de Lyon, Groupement Hospitalier Nord, Lyon, France.
UMR CNRS 5558, Laboratoire de Biométrie et Biologie Évolutive, Univ Lyon, Université Claude Bernard Lyon 1, Villeurbanne, France.
Faculté de Médecine et de Pharmacie de Lyon, Univ Lyon, Université Claude Bernard Lyon 1, Lyon, France.

Cyrielle Codde (C)

Service de Maladies Infectieuses et Tropicales, CHU Limoges, Limoges, France.
Inserm, Univ. Limoges, CHU Limoges, Pharmacology & Transplantation, U 12482 rue du Pr Descottes, 87000, Limoges, France.

Romain Garreau (R)

Service de Pharmacie, Hospices Civils de Lyon, Groupement Hospitalier Nord, Lyon, France.
UMR CNRS 5558, Laboratoire de Biométrie et Biologie Évolutive, Univ Lyon, Université Claude Bernard Lyon 1, Villeurbanne, France.
Faculté de Médecine et de Pharmacie de Lyon, Univ Lyon, Université Claude Bernard Lyon 1, Lyon, France.

Laure Ponthier (L)

Inserm, Univ. Limoges, CHU Limoges, Pharmacology & Transplantation, U 12482 rue du Pr Descottes, 87000, Limoges, France.

Pierre Marquet (P)

Service de Pharmacologie, Toxicologie et Pharmacovigilance, CHU Limoges, France.
Inserm, Univ. Limoges, CHU Limoges, Pharmacology & Transplantation, U 12482 rue du Pr Descottes, 87000, Limoges, France.

Tristan Ferry (T)

Faculté de Médecine et de Pharmacie de Lyon, Univ Lyon, Université Claude Bernard Lyon 1, Lyon, France.
Service des Maladies Infectieuses et Tropicales, Centre de Référence pour la prise en charge des Infections Ostéo-Articulaires complexes (CRIOAc Lyon), Hospices Civils de Lyon, Groupement Hospitalier Nord, Hôpital de la Croix-Rousse, Lyon, France.
CIRI-Centre International de Recherche en Infectiologie, Inserm, U1111, Université́ Claude Bernard Lyon 1, CNRS, UMR5308, Ecole Normale Supérieure de Lyon, Univ Lyon, 69007, Lyon, France.

Marc Labriffe (M)

Service de Pharmacologie, Toxicologie et Pharmacovigilance, CHU Limoges, France.
Inserm, Univ. Limoges, CHU Limoges, Pharmacology & Transplantation, U 12482 rue du Pr Descottes, 87000, Limoges, France.

Alexandre Destere (A)

Service de Pharmacologie et Pharmacovigilance, CHU, Nice, France.

Jean-Baptiste Woillard (JB)

Service de Pharmacologie, Toxicologie et Pharmacovigilance, CHU Limoges, France. jean-baptiste.woillard@unilim.fr.
Inserm, Univ. Limoges, CHU Limoges, Pharmacology & Transplantation, U 12482 rue du Pr Descottes, 87000, Limoges, France. jean-baptiste.woillard@unilim.fr.

Classifications MeSH