A machine learning approach to predict daptomycin exposure from two concentrations based on Monte Carlo simulations

Daptomycin is a concentration-dependent lipopeptide antibiotic for which exposure/effect relationships have been shown. Machine learning (ML) algorithms, developed to predict the individual exposure to drugs, have shown very good performances in comparison to maximum a posteriori Bayesian estimation (MAP-BE). The aim of this work was to predict the area under the blood concentration curve (AUC) of daptomycin from two samples and a few covariates using XGBoost ML algorithm trained on Monte Carlo simulations. Five thousand one hundred fifty patients were simulated from two literature population pharmacokinetics models. Data from the first model were split into a training set (75%) and a testing set (25%). Four ML algorithms were built to learn AUC based on daptomycin blood concentration samples at pre-dose and 1 h post-dose. The XGBoost model (best ML algorithm) with the lowest root mean square error (RMSE) in a 10-fold cross-validation experiment was evaluated in both the test set and the simulations from the second population pharmacokinetic model (validation). The ML model based on the two concentrations, the differences between these concentrations, and five other covariates (sex, weight, daptomycin dose, creatinine clearance, and body temperature) yielded very good AUC estimation in the test (relative bias/RMSE = 0.43/7.69%) and validation sets (relative bias/RMSE = 4.61/6.63%). The XGBoost ML model developed allowed accurate estimation of daptomycin AUC using C0, C1h, and a few covariates and could be used for exposure estimation and dose adjustment. This ML approach can facilitate the conduct of future therapeutic drug monitoring (TDM) studies.

Failure to predict amikacin elimination in critically ill patients with cancer based on the estimated glomerular filtration rate: applying PBPK approach in a therapeutic drug monitoring study

PURPOSE

The aim of this work was to integrate the Therapeutic Drug Monitoring (TDM) with the model-informed precision dosing (MIPD) approach, using Physiologically-based Pharmacokinetic/Pharmacodynamic (PBPK/PD) modelling and simulation, to explore the relationship between amikacin exposure and estimated glomerular filtration rate (GFR) in critically ill patients with cancer.

METHODS

In the TDM study, samples from 51 critically-ill patients with cancer treated with amikacin were analysed. Patients were stratified according to renal function based on GFR status. A full-body PBPK model with 12 organs model was developed using Simcyp V. 21, including steady-state volume of distribution of 0.21 L/kg and renal clearance of 6.9 L/h in healthy adults. PK parameters evaluated were within the 2-fold error range.

RESULTS

During the validation step, predicted vs observed amikacin clearance values after single infusion dose in patients with normal renal function, mild and moderate renal impairment were 7.6 vs 8.1 L/h (7.5 mg/kg dose); 3.8 vs 4.5 L/h (1500 mg dose) and 2.2 vs 3.1 L/h (25 mg/kg dose), respectively. However, predicted vs observed amikacin clearance after a single dose infusion of 1400 mg in critically-ill patients with cancer were 1.46 vs 1.63 (P = 0.6406) L/h (severe), 2.83 vs 1.08 (P < 0.05) L/h (moderate), 4.23 vs 2.49 (P = 0.0625) L/h (mild) and 7.41 vs 3.36 (P < 0.05) L/h (normal renal function).

CONCLUSION

This study demonstrated that estimated GFR did not predict amikacin elimination in critically-ill patients with cancer. Further studies are necessary to find amikacin PK covariates to optimize the pharmacotherapy in this population. Therefore, TDM of amikacin is imperative in cancer patients.

Application of machine learning to predict tacrolimus exposure in liver and kidney transplant patients given the MeltDose formulation

PURPOSE

Machine Learning (ML) algorithms represent an interesting alternative to maximum a posteriori Bayesian estimators (MAP-BE) for tacrolimus AUC estimation, but it is not known if training an ML model using a lower number of full pharmacokinetic (PK) profiles (= "true" reference AUC) provides better performances than using a larger dataset of less accurate AUC estimates. The objectives of this study were: to develop and benchmark ML algorithms trained using full PK profiles to estimate MeltDose^®-tacrolimus individual AUCs using 2 or 3 blood concentrations; and to compare their performance to MAP-BE.

METHODS

Data from liver (n = 113) and kidney (n = 97) transplant recipients involved in MeltDose-tacrolimus PK studies were used for the training and evaluation of ML algorithms. "True" AUC0-24 h was calculated for each patient using the trapezoidal rule on the full PK profile. ML algorithms were trained to estimate tacrolimus true AUC using 2 or 3 blood concentrations. Performances were evaluated in 2 external sets of 16 (renal) and 48 (liver) transplant patients.

RESULTS

Best estimation performances were obtained with the MARS algorithm and the following limited sampling strategies (LSS): predose (0), 8, and 12 h post-dose (rMPE = - 1.28%, rRMSE = 7.57%), or 0 and 12 h (rMPE = - 1.9%, rRMSE = 10.06%). In the external dataset, the performances of the final ML algorithms based on two samples in kidney (rMPE = - 3.1%, rRMSE = 11.1%) or liver transplant recipients (rMPE = - 3.4%, rRMSE = 9.86%) were as good as or better than those of MAP-BEs based on three time points.

CONCLUSION

The MARS ML models developed using "true" MeltDose^®-tacrolimus AUCs yielded accurate individual estimations using only two blood concentrations.

[Therapeutic drug monitoring and pharmacokinetic models as a strategy for rational antibiotic therapy in intensive care patients]

BACKGROUND AND OBJECTIVE

Antibiotic dosing in intensive care patients is complex due to pharmacokinetic (PK) alterations. The aim of this article is to illustrate the role of therapeutic drug monitoring (TDM) and PK models to individualize antibiotic dosing.

MATERIAL AND METHODS

Guidelines and recommendations are discussed in the context of clinical practice and the prerequisites for routine TDM of different antibiotics are presented. In addition, the benefits and limitations of TDM are discussed. The advantages and disadvantages of TDM and PK models are described and the resulting future options are presented.

RESULTS

In the clinical routine, the peak or trough concentrations of antibiotics in blood are measured depending on the antibiotic class. Prerequisites for a purposeful TDM are a coordinated blood sampling and a prompt reporting of findings. As target ranges are not uniformly defined following rules, dosage adjustments are difficult. The PK models offer a valid possibility to individualize the antibiotic therapy of intensive care patients. Areas of application are the calculation of the loading dose and the combination with TDM for treatment control. For whom and how often TDM is necessary and how it can best be combined with PK models or even replace them should be investigated in the future, in addition to evaluation of the optimal target area.

CONCLUSION

The routine use of TDM for antibiotics in intensive care patients is only effective under the abovementioned conditions. By combination with PK models the treatment could be optimized in the future.