Machine Learning as a Novel Method to Support Therapeutic Drug Management and Precision Dosing

Gelder, Teun van; Vinks, Alexander A.

doi:10.1002/cpt.2326

Cited by 10 publications

(17 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The advantage of PPK to fully account for inter-and intra-individual variability and to quantify the effect of these variations on pharmacokinetic parameters precisely compensates for the shortcomings of machine learning ( Bon et al, 2018 ; Campagne et al, 2019 ). Therefore, PPK combined with machine learning to accurately predict pharmacokinetics may be a better approach ( van Gelder and Vinks, 2021 ; Yang et al, 2022 ). Tang et al (2021) developed an individual clearance prediction model for neonatal renal clearance of drugs and successfully validated that PPK combined with machine learning can improve the prediction accuracy of drug clearance.…”

Section: Discussionmentioning

confidence: 99%

Tacrolimus pharmacokinetics in pediatric nephrotic syndrome: A combination of population pharmacokinetic modelling and machine learning approaches to improve individual prediction

et al. 2022

View full text Add to dashboard Cite

Background and Aim: Tacrolimus (TAC) is a first-line immunosuppressant for the treatment of refractory nephrotic syndrome (RNS), but the pharmacokinetics of TAC varies widely among individuals, and there is still no accurate model to predict the pharmacokinetics of TAC in RNS. Therefore, this study aimed to combine population pharmacokinetic (PPK) model and machine learning algorithms to develop a simple and accurate prediction model for TAC.Methods: 139 children with RNS from August 2013 to December 2018 were included, and blood samples of TAC trough and partial peak concentrations were collected. The blood concentration of TAC was determined by enzyme immunoassay; CYP3A5 was genotyped by polymerase chain reaction-restriction fragment length polymorphism method; MYH9, LAMB2, ACTN4 and other genotypes were determined by MALDI-TOF MS method; PPK model was established by nonlinear mixed-effects method. Based on this, six machine learning algorithms, including eXtreme Gradient Boosting (XGBoost), Random Forest (RF), Extra-Trees, Gradient Boosting Decision Tree (GBDT), Adaptive boosting (AdaBoost) and Lasso, were used to establish the machine learning model of TAC clearance.Results: A one-compartment model of first-order absorption and elimination adequately described the pharmacokinetics of TAC. Age, co-administration of Wuzhi capsules, CYP3A5 *3/*3 genotype and CTLA4 rs4553808 genotype were significantly affecting the clearance of TAC. Among the six machine learning models, the Lasso algorithm model performed the best (R2 = 0.42).Conclusion: For the first time, a clearance prediction model of TAC in pediatric patients with RNS was established using PPK combined with machine learning, by which the individual clearance of TAC can be predicted more accurately, and the initial dose of administration can be optimized to achieve the goal of individualized treatment.

show abstract

Section: Discussionmentioning

confidence: 99%

Tacrolimus pharmacokinetics in pediatric nephrotic syndrome: A combination of population pharmacokinetic modelling and machine learning approaches to improve individual prediction

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Black boxes are being applied in clinical pharmacovigilance for point-of care prediction and management of drug toxicity (e.g., via precision dosing). [25][26][27][28][29][30][31] Clinicians often require machine reasoning to somehow align with their reasoning. This requirement is understandable in any case but may have been reinforced by some highly publicized deep learning missteps.…”

Section: Explainability For Building Trustmentioning

confidence: 99%

“…Black boxes are being applied in clinical pharmacovigilance for point‐of care prediction and management of drug toxicity (e.g., via precision dosing) 25–31 . Clinicians often require machine reasoning to somehow align with their reasoning.…”

Section: Is Explainable Ai Needed In Pharmacovigilance?mentioning

confidence: 99%

Artificial intelligence in pharmacovigilance: Do we need explainability?

Hauben

2022

Pharmacoepidemiology and Drug

View full text Add to dashboard Cite

“…Considering this, MC simulation could be a most useful approach in identification of ALB concentrations significance for inter‐individual MPA pharmacokinetic variability. 25 , 26 , 27 …”

Section: Introductionmentioning

confidence: 99%

“…Considering this, MC simulation could be a most useful approach in identification of ALB concentrations significance for inter-individual MPA pharmacokinetic variability. [25][26][27] The aim was the identification of mathematical correlation and association between both, total and unbound MPA concentration in relation to ALB, BM, age and eGFR in stable kidney transplant recipients. Furthermore, investigation was conducted with the aim to clarify the role of C SAL MPA in adverse effects profile.…”

Section: Introductionmentioning

confidence: 99%

Utility of salivary mycophenolic acid concentration monitoring: Modeling and Monte Carlo validation approach

Catić‐Đorđević

Stefanović

Pavlović

et al. 2022

Pharmacology Res & Perspec

View full text Add to dashboard Cite

The results of the previous studies demonstrated an association between mycophenolic acid (MPA) exposure, serum albumin level (ALB), and adverse effects in kidney transplant patients. The aim was the identification of mathematical correlation and association between both, total and unbound MPA concentration in relation to ALB, body mass (BM), age and estimated glomerular filtration rate (eGFR) in stable kidney transplant recipients. Furthermore, investigation was conducted with the aim to clarify the role of salivary concentration (C SAL ) of MPA in adverse effect profile. In order to analyze the association between total and salivary concentration of MPA in relation to ALB, BM, age and eGFR, a least squares method for determining the correlation between these parameters was performed. In addition, derived mathematical model based on experimental data can also be performed and simulated through the Monte Carlo (MC) approach. Adverse effects were grouped according to the nature of symptoms and scored by a previously published validated system. Numerically calculated values of C SAL from the models [C SAL = f (ALB, BM, age, eGFR, C P ) = a 00 + a 10 *(ALB, BM, age, eGFR) + a 01 *C P ] were then compared with those from validation set of patients, where the best fitting model was for ALB [C SAL = 54.96–1.64*ALB +13.4*C P ]. Adverse effects estimation showed the difference in esthetic score, positively correlated with C SAL in the lower ALB group (145.41 ± 219.02 vs. 354.08 ± 262.19; with statistical significance p = .014) and almost significant for gastrointestinal score (167.69 ± 174.79 vs. 347.55 ± 320.95; p = .247). The study showed that C SAL MPA may contribute to management of adverse effects, but these findings require confirmation of clinical utility.

show abstract

Machine Learning as a Novel Method to Support Therapeutic Drug Management and Precision Dosing

Cited by 10 publications

References 24 publications

Tacrolimus pharmacokinetics in pediatric nephrotic syndrome: A combination of population pharmacokinetic modelling and machine learning approaches to improve individual prediction

Tacrolimus pharmacokinetics in pediatric nephrotic syndrome: A combination of population pharmacokinetic modelling and machine learning approaches to improve individual prediction

Artificial intelligence in pharmacovigilance: Do we need explainability?

Utility of salivary mycophenolic acid concentration monitoring: Modeling and Monte Carlo validation approach

Contact Info

Product

Resources

About