Model-Informed Precision Dosing: Background, Requirements, Validation, Implementation, and Forward Trajectory of Individualizing Drug Therapy.

Model-informed precision dosing (MIPD) has become synonymous with modern approaches for individualizing drug therapy, in which the characteristics of each patient are considered as opposed to applying a one-size-fits-all alternative. This review provides a brief account of the current knowledge, practices, and opinions on MIPD while defining an achievable vision for MIPD in clinical care based on available evidence. We begin with a historical perspective on variability in dose requirements and then discuss technical aspects of MIPD, including the need for clinical decision support tools, practical validation, and implementation of MIPD in health care. We also discuss novel ways to characterize patient variability beyond the common perceptions of genetic control. Finally, we address current debates on MIPD from the perspectives of the new drug development, health economics, and drug regulations.

Prospective validation of a model-informed precision dosing tool for vancomycin treatment in neonates.

We recruited 48 neonates (50 vancomycin treatment episodes) in a prospective study to validate a model-informed precision dosing (MIPD) software. The initial vancomycin dose was based on a population pharmacokinetic model and adjusted every 36-48 h. Compared with a historical control group of 53 neonates (65 episodes), the achievement of a target trough concentration of 10-15 mg/L improved from 37% in the study to 62% in the MIPD group (P = 0.01), with no difference in side effects.

Serum ganciclovir drug exposure in children receiving standard ganciclovir dosing.

Intravenous ganciclovir (GCV) is used for the treatment of cytomegalovirus (CMV) infection in immunocompromised children. Although the therapeutic target for treatment is unclear, studies have shown a serum area under the concentration-time curve (AUC24h) ≥40 mg/L·h correlates with effective CMV prevention. This study aimed to externally validate existing GCV population pharmacokinetic (PopPK) models and develop a model if needed and evaluate the serum AUC24h achieved with standard GCV dosing and propose an optimized dosing strategy for immunocompromised children. Ganciclovir drug monitoring data from two pediatric hospitals were retrospectively collected, and published pediatric PopPK models were externally validated. The population AUC24h with standard GCV dosing (5 mg/kg twice daily) was calculated, and an optimized dosing strategy was determined using Monte Carlo simulations to achieve an AUC24h between 40 and 100 mg/L·h. Overall, 161 samples from 23 children with a median (range) age of 9.0 years (0.4-17.0) and weight of 28.2 kg (5.6-73.3) were analyzed. Transferability of published pediatric PopPK models was limited. Thus, a one-compartment model with first-order absorption and elimination with weight and serum creatinine as covariates was developed. The median (5th-95th percentiles) steady state AUC24h with standard dosing was 38.3 mg/L·h (24.8-329.2) with 13 children having an AUC24h <40 mg/L·h, particularly those aged <4 years (8/13). An optimized simulated GCV dosing regimen, ranging from 2 to 13 mg/kg twice daily for children with normal renal function, achieved 61%-78% probability of target attainment. Standard GCV dosing likely results in inadequate drug exposure in more than half of the children, particularly those aged <4 years. An optimized dosing regimen has been proposed for clinical validation.

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.

Kinetic-pharmacodynamic model of warfarin for prothrombin time-international normalized ratio in Japanese patients.

AIMS: Genotype-guided dosing algorithms can explain about half of the interindividual variability in prothrombin time-international normalized ratio (PT-INR) under warfarin treatment. This study aimed to refine a published kinetic-pharmacodynamic model and guide warfarin dosage for an optimal PT-INR based on renal function. METHODS: Using a retrospective cohort of adult patients (>20 years) who were administered warfarin and underwent PT-INR measurements, we refined the kinetic-pharmacodynamic model with age and the genotypes of cytochrome P450 2C9 and vitamin K epoxide reductase complex subunit 1 using the PRIOR subroutine in the nonlinear-mixed-effect modelling programme. We searched the significant covariates for parameters, such as the dose rate for 50% inhibition of coagulation (EDR50 ), using a stepwise forward and backward method. Monte Carlo simulation determined a required daily dose of warfarin with a target range of PT-INR (2.0-3.0 or 1.6-2.6) based on the significant covariates. RESULTS: A total of 350 patients with 2762 PT-INR measurements were enrolled (estimated glomerular filtration rate [eGFR]: 47.5 [range: 2.6-199.0] mL/min/1.73 m2 ). The final kinetic-pharmacodynamic model showed that the EDR50 changed power functionally with body surface area, serum albumin level and eGFR. Monte Carlo simulation revealed that a lower daily dose of warfarin was required to attain the target PT-INR range as eGFR decreased. CONCLUSIONS: Model-informed precision dosing of warfarin is a valuable approach for estimating its dosage in patients with renal impairment.

Eculizumab dose tapering should take into account the nonlinearity of its pharmacokinetics.

AIMS: Eculizumab is a monoclonal antibody targeting complement protein C5 used in renal diseases. As recommended dosing regimen leads to unnecessarily high concentrations in some patients, tailored dosing therapeutic drug monitoring was proposed to reduce treatment cost. The objectives of the present work were (i) to investigate the target-mediated elimination of eculizumab and (ii) whether a pharmacokinetic model integrating a nonlinear elimination allows a better prediction of eculizumab concentrations than a linear model. METHODS: We analysed 377 eculizumab serum concentrations from 44 patients treated for atypical haemolytic uraemic syndrome and C3 glomerulopathy with a population pharmacokinetic approach. Critical concentrations (below which a non-log-linear decline of concentration over time is evidenced) were computed to estimate the relevance of the target-mediated elimination. Simulations of dosing regimens were then performed to predict probabilities of target attainment (i.e. trough >100 mg/L). RESULTS: Pharmacokinetics of eculizumab was nonlinear and followed a mixture of first-order (CL = 1.318 mL/day/kg) and Michaelis-Menten elimination (Vmax = 26.07 mg/day, Km = 24.06 mg/L). Volume of distribution (72.39 mL/kg) and clearance were weight-dependent. Critical concentrations (Vmax/CL) ranged from 144.7 to 759.7 mg/L and were inversely related to body weight (P = .013). Nonlinearity was thus noticeable at therapeutic concentrations. Simulations predicted that 1200 mg of eculizumab every 21 days would allow 85% and 76% of patients to maintain a therapeutic exposure, for 50 or 90 kg body weight, respectively. CONCLUSIONS: Our study investigates the nonlinear elimination of eculizumab and discusses the importance of accounting for eculizumab target-mediated elimination in therapeutic drug monitoring.

The potential for therapeutic drug monitoring of belatacept and other biologicals in solid organ transplantation.

In solid organ transplantation (SOT), biologicals such as recombinant therapeutic proteins, monoclonal antibodies, fusion proteins and conjugates are increasingly used for immunosuppression, desensitization, ABO (blood group) incompatibility, antibody-mediated rejections and atypical haemolytic uremic syndrome. In this paper, we review the medical evidence available for biologicals used in SOT and the potential for improvement by the application of therapeutic drug monitoring (TDM) and model-informed precision dosing. Biologicals are used for off-label indications within the field of SOT, building on the experience from their use on labelled indications. Dosing is currently mostly standard, and experience vs. effect and toxicity is limited. Pharmacokinetic characteristics of these large, partly also immunogenic molecules differ from those of traditional small molecules. Individualization by concentration measurements and modelling has mostly been proof-of-concept or feasibility studies that lack the power to provide evidence for improvement in clinical outcome. For some drugs such as alemtuzumab, eculizumab, rituximab, tocilizumab and belatacept, studies have demonstrated significant interindividual variability in pharmacokinetics. Variability in absorption from subcutaneous administration may increase interindividual variability. There is also an economic aspect of appropriate dosing that needs to be pursued. Available assays and models to refine interpretation are in place, but trials of adequate size to document the usefulness of TDM and MIPD are scarce. Collaboration within the TDM community seems mandatory to establish studies of sufficient strength to provide evidence for the use of biologicals that are currently used off-label in SOT and furthermore to identify the settings where TDM may be beneficial.

A simulation study on model-informed precision dosing of amikacin for achieving target area under the concentration-time curve.

AIMS: Amikacin requires therapeutic drug monitoring for optimum efficacy; however, the optimal model-informed precision dosing strategy for the area under the concentration-time curve (AUC) of amikacin is uncertain. This simulation study aimed to determine the efficient blood sampling points using the Bayesian forecasting approach for early achievement of the target AUC range for amikacin in critically ill patients. METHODS: We generated a virtual population of 3000 individuals using 2 validated population pharmacokinetic models identified using a systematic literature search. AUC for each blood sampling point was evaluated using the probability of achieving a ratio of estimated/reference AUC at steady state in the 0.8-1.2 range. RESULTS: On day 1, the 1-point samplings for population pharmacokinetic models showed a priori probabilities of 26.3 and 45.6%, which increased to 47.3 and 94.4% at 23 and 15 h, respectively. Using 2-point sampling at the peak (3 and 4 h) and trough (24 h) on day 1, these probabilities further increased to 72.3 and 99.5%, respectively. These probabilities were comparable on days 2 and 3, regardless of 3 and 6 sampling points or estimated glomerular filtration rate. These results indicated the higher predictive accuracy of 2-point sampling than 1-point sampling on day 1 for amikacin AUC estimation. Moreover, 2-point sampling was a more reasonable approach than rich sampling. CONCLUSIONS: This study contributes to the development of an efficient model-informed precision dosing strategy for early targeting of amikacin AUC in critically ill patients.

Immunosuppressant adherence in adult outpatient hematopoietic cell transplant recipients.

INTRODUCTION: Medication nonadherence continues to be challenging for allogeneic hematopoietic cell transplant (HCT) recipients. The risk and severity of chronic graft-versus-host disease (GVHD) are associated with low immunosuppressant concentrations (which can be improved with model-informed precision dosing (MIPD)) and with immunosuppressant nonadherence (which can be improved with acceptable interventions). METHODS: With the goals of improving adherence and achieving therapeutic concentrations of immunosuppressants to eliminate GVHD, we characterized the feasibility of using the Medication Event Monitoring (MEMS®) Cap in adult HCT recipients. RESULTS: Of the 27 participants offered the MEMS® Cap at the time of hospital discharge, 7 (25.9%) used it, which is below our a priori threshold of 70%. These data suggest the MEMS® Cap is not feasible for HCT recipients. The MEMS® Cap data were available for a median of 35 days per participant per medication (range: 7-109 days). The average daily adherence per participant ranged from 0 to 100%; four participants had an average daily adherence of over 80%. CONCLUSIONS: MIPD may be supported by MEMS® technology to provide the precise time of immunosuppressant self-administration. The MEMS® Cap was used by only a small percentage (25.9%) of HCT recipients in this pilot study. In accordance with larger studies using less accurate tools to evaluate adherence, immunosuppressant adherence varied from 0% to 100%. Future studies should establish the feasibility and clinical benefit of combining MIPD with newer technology, specifically the MEMS® Button, which can inform the oncology pharmacist of the time of immunosuppressant self-administration.

Maximum a posteriori Bayesian methods out-perform non-compartmental analysis for busulfan precision dosing.

Dose personalization improves patient outcomes for many drugs with a narrow therapeutic index and high inter-individuality variability, including busulfan. Non-compartmental analysis (NCA) and model-based methods like maximum a posteriori Bayesian (MAP) approaches are two methods routinely used for dose optimization. These approaches vary in how they estimate patient-specific pharmacokinetic parameters to inform a dose and the impact of these differences is not well-understood. Using busulfan as an example application and area under the concentration-time curve (AUC) as a target exposure metric, these estimation methods were compared using retrospective patient data (N = 246) and simulated precision dosing treatment courses. NCA was performed with or without peak extension, and MAP Bayesian estimation was performed using either the one-compartment Shukla model or the two-compartment McCune model. All methods showed good agreement on real-world data (correlation coefficients of 0.945-0.998) as assessed by Bland-Altman plots, although agreement between NCA and MAP methods was higher during the first dosing interval (0.982-0.994) compared to subsequent dosing intervals (0.918-0.938). In dose adjustment simulations, both NCA and MAP estimated high target attainment (> 98%) although true simulated target attainment was lower for NCA (63-66%) versus MAP (91-93%). The largest differences in AUC estimation were due to different assumptions for the shape of the concentration curve during the infusion phase, followed by how the methods considered time-dependent clearance and concentration-time points collected in earlier intervals. In conclusion, although AUC estimates between the two methods showed good correlation, in a simulated study, MAP lead to higher target attainment. When changing from one method to another, or changing infusion duration and other factors, optimum estimated exposure targets may require adjusting to maintain a consistent exposure.

Subgroup identification-based model selection to improve the predictive performance of individualized dosing.

Currently, model-informed precision dosing uses one population pharmacokinetic model that best fits the target population. We aimed to develop a subgroup identification-based model selection approach to improve the predictive performance of individualized dosing, using vancomycin in neonates/infants as a test case. Data from neonates/infants with at least one vancomycin concentration was randomly divided into training and test dataset. Population predictions from published vancomycin population pharmacokinetic models were calculated. The single best-performing model based on various performance metrics, including median absolute percentage error (APE) and percentage of predictions within 20% (P20) or 60% (P60) of measurement, were determined. Clustering based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm was used to group neonates/infants according to their best-performing model. Subsequently, classification trees to predict the best-performing model using clinical and demographic characteristics were developed. A total of 208 vancomycin treatment episodes in training and 88 in test dataset was included. Of 30 identified models from the literature, the single best-performing model for training dataset had P20 26.2-42.6% in test dataset. The best-performing clustering approach based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm had P20 44.1-45.5% in test dataset, whereas P60 was comparable. Our proof-of-concept study shows that the prediction of the best-performing model for each patient according to the proposed model selection approaches has the potential to improve the predictive performance of model-informed precision dosing compared with the single best-performing model approach.

Model-Informed Precision Dosing of Vancomycin in Adult Patients Undergoing Hemodialysis.

Model-informed precision dosing (MIPD) maximizes the probability of successful dosing in patients undergoing hemodialysis. In these patients, area under the concentration-time curve (AUC)-guided dosing is recommended for vancomycin. However, this model is yet to be developed. The purpose of this study was to address this issue. The overall mass transfer-area coefficient (KoA) was used for the estimation of vancomycin hemodialysis clearance. A population pharmacokinetic (popPK) model was developed, resulting in a fixed-effect parameter for nonhemodialysis clearance of 0.316 liters/h. This popPK model was externally evaluated, with a resulting mean absolute error of 13.4% and mean prediction error of -0.17%. KoA-predicted hemodialysis clearance was prospectively evaluated for vancomycin (n = 10) and meropenem (n = 10), with a correlation equation being obtained (slope of 1.099, intercept of 1.642; r = 0.927, P < 0.001). An experimental evaluation using an in vitro hemodialysis circuit validated the developed model of KoA-predicted hemodialysis clearance using vancomycin, meropenem, vitamin B6, and inulin in 12 hemodialysis settings. This popPK model indicated a maximum a priori dosing for vancomycin-a loading dose of 30 mg/kg, which achieves the target AUC for 24 h after first dose with a probability of 93.0%, ensured by a predialysis concentration of >15 µg/mL. Maintenance doses of 12 mg/kg after every hemodialysis session could achieve the required exposure, with a probability of 80.6%. In conclusion, this study demonstrated that KoA-predicted hemodialysis clearance may lead to an upgrade from conventional dosing to MIPD for vancomycin in patients undergoing hemodialysis.

Systematic Evaluation of Pharmacokinetic Models for Model-Informed Precision Dosing of Meropenem in Critically Ill Patients Undergoing Continuous Renal Replacement Therapy.

The altered pharmacokinetics of renally cleared drugs such as meropenem in critically ill patients receiving continuous renal replacement therapy (CRRT) might impact target attainment. Model-informed precision dosing (MIPD) is applied to individualize meropenem dosing. However, most population pharmacokinetic (PopPK) models developed to date have not yet been evaluated for MIPD. Eight PopPK models based on adult CRRT patients were identified in a systematic literature research and encoded in NONMEM 7.4. A data set of 73 CRRT patients from two different study centers was used to evaluate the predictive performance of the models using simulation and prediction-based diagnostics for i) a priori dosing based on patient characteristics only and ii) Bayesian dosing by including the first measured trough concentration. Median prediction error (MPE) for accuracy within |20%| (95% confidence intervals including zero) and median absolute prediction error (MAPE) for precision ≤ 30% were considered clinically acceptable. For a priori dosing, most models (n = 5) showed accuracy and precision MPE within |20%| and MAPE <35%. The integration of the first measured meropenem concentration improved the predictive performance of all models (median MAPE decreased from 35.4 to 25.0%; median MPE decreased from 21.8 to 4.6%). The best predictive performance for intermittent infusion was observed for the O'Jeanson model, including residual diuresis as covariate (a priori and Bayesian dosing MPE within |2%|, MAPE <30%). Our study revealed the O'Jeanson model as the best-predicting model for intermittent infusion. However, most of the selected PopPK models are suitable for MIPD in CRRT patients when one therapeutic drug monitoring sample is available.

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.

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.

Model-informed precision dosing of antimicrobial drugs in pediatrics: experiences from a pilot scale program.

Antibiotics are among the most utilized drugs in pediatrics. Nonetheless, there is a lack in pharmacokinetics information for this population, and dosing criteria may vary between healthcare centers. Physiological variability associated with maturation in pediatrics makes it challenging to reach a consensus on adequate dosing, which is further accentuated in more vulnerable groups, such as critically ill or oncology patients. Model-informed precision dosing is a useful practice that allows dose optimization and attainment of antibiotic-specific pharmacokinetic/pharmacodynamic targets. The aim of this study was to evaluate the needs of model-informed precision dosing of antibiotics in a pediatrics unit, at a pilot scale. Pediatric patients under antibiotic treatment were monitored with either a pharmacokinetic/pharmacodynamic optimized sampling scheme or through opportunistic sampling. Clindamycin, fluconazole, linezolid, meropenem, metronidazole, piperacillin, and vancomycin plasma concentrations were quantified through a liquid chromatography coupled to mass spectrometry method. Pharmacokinetic parameters were estimated using a Bayesian approach to verify pharmacokinetic/pharmacodynamic target attainment. A total of 23 pediatric patients aged 2 to 16 years were included, and 43 dosing regimens were evaluated; 27 (63%) of them required adjustments as follows: 14 patients were underdosed, 4 were overdosed, and 9 patients needed infusion rate adjustments. Infusion rate adjustments were mostly recommended for piperacillin and meropenem; daily doses were augmented for vancomycin and metronidazole, meanwhile linezolid was adjusted for under- and overdosing. Clindamycin and fluconazole regimens were not adjusted at all.  Conclusion: Results showcase a lack of antibiotic pharmacokinetic/pharmacodynamic target attainment (particularly for linezolid, vancomycin, meropenem, and piperacillin), and the need for model-informed precision dosing in pediatrics. This study provides pharmacokinetic evidence which can further improve antibiotic dosing practices. What is Known: • Model-informed precision dosing is performed in pediatrics to optimize the treatment of antimicrobial drugs such as vancomycin and aminoglycosides, while its usefulness is debated for other groups (beta-lactams, macrolides, etc.). What is New: • Vulnerable pediatric subpopulations, such as critically ill or oncology patients, can benefit the most from model-informed precision dosing of antibiotics. • Model-informed precision dosing of linezolid, meropenem, piperacillin, and vancomycin is particularly useful in pediatrics, and further research may improve dosing practices altogether.

Comparison of the predictive accuracy of the physiologically based pharmacokinetic (PBPK) model and population pharmacokinetic (PPK) model of vancomycin in Japanese patients with MRSA infection.

INTRODUCTION: The latest therapeutic drug monitoring guidelines for vancomycin (VCM) recommend that area under the concentration-time curve is estimated based on model-informed precision dosing and used to evaluate efficacy and safety. Therefore, we predicted VCM concentrations in individual methicillin-resistant Staphylococcus aureus-infected patients using existing a physiologically based pharmacokinetic (PBPK) model and 1- and 2-compartment population pharmacokinetic (PPK) models and confirmed and verified the accuracy of the PBPK model in estimating VCM concentrations with the PPK model. METHODS: The subjects of the study are 20 patients, and the predicted concentrations were evaluated by comparing the observed and predicted trough and peak values of VCM concentrations for individual patients. RESULTS: The results showed good correlation between the observed and predicted trough and peak concentrations of VCM was observed generally in the PBPK model, R2 values of 0.72, 0.62, and 0.40 with trough values of 0.49, 0.40, and 0.34 with peak values for PBPK model, 1-compartment, and 2-compartment model, respectively. CONCLUSIONS: Although the performance of the PBPK model is not as predictive as the PPK model, generally similar predictive trends were obtained, suggesting that it may be a valuable tool for rapid and accurate prediction of AUC for VCM.

Implementation of a Vancomycin Dose-Optimization Protocol in Neonates: Impact on Vancomycin Exposure, Biological Parameters, and Clinical Outcomes.

Vancomycin dosing used in neonates results frequently in insufficient concentrations. A vancomycin dose-optimization protocol consisting of an individualization of loading and maintenance doses (administered during continuous infusion) through a previously validated pharmacokinetic model was implemented in our center. This monocenter retrospective study aimed to compare vancomycin average concentration (Cavg) in the therapeutic range (15 to 25 mg/L) and biological and clinical parameters before and after implementation of this protocol. A total of 60 and 59 courses of vancomycin treatment in 45 and 49 patients were analyzed in groups before and after implementation, respectively. Initial vancomycin Cavg were more frequently in the therapeutic range in the group after implementation (74.6% versus 28.3%, P < 0.001), with 1.6-fold higher Cavg (20.3 [17.0-22.2] mg/L versus 12.9 [11.3-17.0] mg/L, P < 0.001). Considering all Cavg during longitudinal therapeutic drug monitoring (TDM), the frequency of therapeutic Cavg was higher in the group after implementation (74.8% [n = 103] versus 31% [n = 116], P < 0.001). The dose optimization protocol was also associated with a reduced time to obtain a negative blood culture (P < 0.001) and fewer antibiotic switches (P = 0.025), without increasing the frequency of nephrotoxicity. Clinical outcomes also appeared to be improved, with less periventricular leukomalacia (P = 0.021), trended toward less respiratory instability (P = 0.15) and a shorter duration of vasoactive drug use (P = 0.18) for neonates receiving personalized doses of vancomycin. This personalized vancomycin dose protocol improves vancomycin exposure in neonates, with good safety, and suggests an improvement in biological and clinical outcomes.

A Machine Learning Approach to Predict Interdose Vancomycin Exposure.

INTRODUCTION: Estimation of vancomycin area under the curve (AUC) is challenging in the case of discontinuous administration. Machine learning approaches are increasingly used and can be an alternative to population pharmacokinetic (POPPK) approaches for AUC estimation. The objectives were to train XGBoost algorithms based on simulations performed in a previous POPPK study to predict vancomycin AUC from early concentrations and a few features (i.e. patient information) and to evaluate them in a real-life external dataset in comparison to POPPK. PATIENTS AND METHODS: Six thousand simulations performed from 6 different POPPK models were split into training and test sets. XGBoost algorithms were trained to predict trapezoidal rule AUC a priori or based on 2, 4 or 6 samples and were evaluated by resampling in the training set and validated in the test set. Finally, the 2-sample algorithm was externally evaluated on 28 real patients and compared to a state-of-the-art POPPK model-based averaging approach. RESULTS: The trained algorithms showed excellent performances in the test set with relative mean prediction error (MPE)/ imprecision (RMSE) of the reference AUC = 3.3/18.9, 2.8/17.4, 1.3/13.7% for the 2, 4 and 6 samples algorithms respectively. Validation in real patient showed flexibility in sampling time post-treatment initiation and excellent performances MPE/RMSE<1.5/12% for the 2 samples algorithm in comparison to different POPPK approaches. CONCLUSIONS: The Xgboost algorithm trained from simulation and evaluated in real patients allow accurate and precise prediction of vancomycin AUC. It can be used in combination with POPPK models to increase the confidence in AUC estimation.

Population pharmacokinetic and pharmacodynamic analysis of rivaroxaban in Chinese patients with non-valvular atrial fibrillation.

Rivaroxaban, a direct factor Xa inhibitor, is widely used for stroke prevention in patients with non-valvular atrial fibrillation (NVAF). The aim of this study was to conduct a population pharmacokinetic-pharmacodynamic (PK-PD) analysis of rivaroxaban in Chinese patients with NVAF to assess ethnic differences and provide model-based precision dosing. A total of 256 rivaroxaban plasma concentrations and 244 prothrombin time (PT) measurements were obtained from 195 Chinese NVAF patients from a prospective clinical trial. The population PK-PD model was developed using nonlinear mixed effects modeling (NONMEM) software. The PK of rivaroxaban was adequately described using a one-compartment model with first-order adsorption and elimination. Estimated glomerular filtration rate (eGFR) was identified as a major covariate for apparent clearance. No single nucleotide polymorphism was identified as a significant covariate. PT exhibited a linear relationship with rivaroxaban concentration. Total bilirubin (TBIL) and eGFR were identified as significant covariates for baseline PT. According to the Monte Carlo simulation, 15 mg for Chinese patients with eGFR ≥50 mL/min and normal liver function yielded an exposure comparable to 20 mg for Caucasian patients. Patients with moderately impaired renal function may require a lower dose of rivaroxaban to avoid overexposure. Moreover, there was an approximate 26% increase in PT levels in patients with TBIL of 34 µmol/L and eGFR of 30 mL/min, which could increase the risk of major bleeding. The established population PK-PD model could inform individualized dosing for Chinese NVAF patients who are administered rivaroxaban.