Model-informed Precision Dosing Research Articles

Challenges and Possible Solutions to Direct-Acting Oral Anticoagulants (DOACs) Dosing in Patients with Extreme Bodyweight and Renal Impairment

This article aims to highlight the dosing issues of direct oral anticoagulants (DOACs) in patients with renal impairment and/or obesity in an attempt to develop solutions employing advanced data-driven techniques. DOACs have become widely accepted by clinicians worldwide because of their superior clinical profiles, more predictable pharmacokinetics, and hence more convenient dosing relative to other anticoagulants. However, the optimal dosing of DOACs in extreme bodyweight patients and patients with renal impairment is difficult to achieve using the conventional dosing approach. The standard dosing approach (fixed-dose) is based on limited data from clinical studies. The existing formulae (models) for determining the appropriate doses for these patient groups leads to suboptimal dosing. This problem of mis-dosing is worsened by the lack of standardized laboratory parameters for monitoring the exposure to DOACs in renal failure and extreme bodyweight patients. Model-informed precision dosing (MIPD) encompasses a range of techniques like machine learning and pharmacometrics modelling, which could uncover key variables and relationships as well as shed more light on the pharmacokinetics and pharmacodynamics of DOACs in patients with extreme bodyweight or renal impairment. Ultimately, this individualized approach-if implemented in clinical practice-could optimise dosing for the DOACs for better safety and efficacy.

Therapeutic drug monitoring for cytotoxic anticancer drugs: Principles and evidence-based practices.

Cytotoxic drugs are highly efficacious and also have low therapeutic index. A great degree of caution needs to be exercised in their usage. To optimize the efficacy these drugs need to be given at maximum tolerated dose which leads to significant amount of toxicity to the patient. The fine balance between efficacy and safety is the key to the success of cytotoxic chemotherapeutics. However, it is possibly more rewarding to obtain that balance for this class drugs as the frequency of drug related toxicities are higher compared to the other therapeutic class and are potentially life threatening and may cause prolonged morbidity. Significant efforts have been invested in last three to four decades in therapeutic drug monitoring (TDM) research to understand the relationship between the drug concentration and the response achieved for therapeutic efficacy as well as drug toxicity for cytotoxic drugs. TDM evolved over this period and the evidence gathered favored its routine use for certain drugs. Since, TDM is an expensive endeavor both from economic and logistic point of view, to justify its use it is necessary to demonstrate that the implementation leads to perceivable improvement in the patient outcomes. It is indeed challenging to prove the utility of TDM in randomized controlled trials and at times may be nearly impossible to generate such data in view of the obvious findings and concern of compromising patient safety. Therefore, good quality data from well-designed observational study do add immense value to the scientific knowledge base, when they are examined in totality, despite the heterogeneity amongst them. This article compiles the summary of the evidence and the best practices for TDM for the three cytotoxic drug, busulfan, 5-FU and methotrexate. Traditional use of TDM or drug concentration data for dose modification has been witnessing a sea change and model informed precision dosing is the future of cytotoxic drug therapeutic management.

β-lactam precision dosing in critically ill children: Current state and knowledge gaps.

There has been emerging interest in implementing therapeutic drug monitoring and model-informed precision dosing of β-lactam antibiotics in critically ill patients, including children. Despite a position paper endorsed by multiple international societies that support these efforts in critically ill adults, implementation of β-lactam precision dosing has not been widely adopted. In this review, we highlight what is known about β-lactam antibiotic pharmacokinetics and pharmacodynamics in critically ill children. We also define the knowledge gaps that present barriers to acceptance and implementation of precision dosing of β-lactam antibiotics in critically ill children: a lack of consensus on which subpopulations would benefit most from precision dosing and the uncertainty of how precision dosing changes outcomes. We conclude with opportunities for further research to close these knowledge gaps.

Delineating gene-environment effects using virtual twins of patients treated with clozapine.

Studies that focus on individual covariates, while ignoring their interactions, may not be adequate for model-informed precision dosing (MIPD) in any given patient. Genetic variations that influence protein synthesis should be studied in conjunction with environmental covariates, such as cigarette smoking. The aim of this study was to build virtual twins (VTs) of real patients receiving clozapine with interacting covariates related to genetics and environment and to delineate the impact of interacting covariates on predicted clozapine plasma concentrations. Clozapine-treated patients with schizophrenia (N=42) with observed clozapine plasma concentrations, demographic, environmental, and genotype data were used to construct VTs in Simcyp. The effect of increased covariate virtualization was assessed by performing simulations under three conditions: "low" (demographic), "medium" (demographic and environmental interaction), and "high" (demographic and environmental/genotype interaction) covariate virtualization. Increasing covariate virtualization with interaction improved the coefficient of variation (R2 ) from 0.07 in the low model to 0.391 and 0.368 in the medium and high models, respectively. Whereas R2 was similar between the medium and high models, the high covariate virtualization model had improved accuracy, with systematic bias of predicted clozapine plasma concentration improving from -138.48 ng/ml to -74.65 ng/ml. A high level of covariate virtualization (demographic, environmental, and genotype) may be required for MIPD using VTs.

Evaluation of Neonatal and Paediatric Vancomycin Pharmacokinetic Models and the Impact of Maturation and Serum Creatinine Covariates in a Large Multicentre Data Set.

Infants and neonates present a clinical challenge for dosing drugs with high interindividual variability due to these patients' rapid growth and the interplay between maturation and organ function. Model-informed precision dosing (MIPD), which can account for interindividual variability via patient characteristics and Bayesian forecasting, promises to improve individualized dosing strategies in this complex population. Here, we assess the predictive performance of published population pharmacokinetic models describing vancomycin in neonates and infants, and analyze the robustness of these models in the face of clinical uncertainty surrounding covariate values. The predictive precision and bias of nine pharmacokinetic models were compared in a large multi-site data set (N = 2061 patients, 5794 drug levels, 28 institutions) of patients aged 0-365 days. The robustness of model predictions to errors in serum creatinine measurements and gestational age was assessed by using recorded values or by replacing covariate values with 0.3, 0.5 or 0.8 mg/dL or with 40 weeks, respectively. Of the nine models, two models (Dao and Jacqz-Aigrain) resulted in predicted concentrations within 2.5 mg/L or 15% of the measured values for at least 60% of population predictions. Within individual models, predictive performance often 2 differed in neonates (0-4 weeks) versus older infants (15-52 weeks). For preterm neonates, imputing gestational age as 40 weeks reduced the accuracy of model predictions. Measured values of serum creatinine improved model predictions compared to using imputed values even in neonates ≤1 week of age. Several available pharmacokinetic models are suitable for MIPD in infants and neonates. Availability and accuracy of model covariates for patients will be important for guiding dose decision-making.

Developmental pharmacokinetics of indomethacin in preterm neonates: Severely decreased drug clearance in the first week of life.

Indomethacin is used commonly in preterm neonates for the prevention of intracranial hemorrhage and closure of an abnormally open cardiac vessel. Due to biomedical advances, the infants who receive this drug in the neonatal intensive care unit setting have become younger, smaller, and less mature (more preterm) at the time of treatment. To develop a pharmacokinetics (PK) model to aid future dosing, we designed a prospective cohort study to characterize indomethacin PK in a dynamically changing patient population. A population PK base model was created using NONMEM, and a covariate model was developed in a primary development cohort and subsequently was tested for accuracy in a validation cohort. Postnatal age was a significant covariate for hepatic clearance (CLH ) and renal clearance (CLR ). The typical value of the total clearance (CL, the sum of CLR and CLH ) was 3.09 ml/h and expressed as CL/WTmedian =3.96 ml/h/kg, where WTmedian is the median body weight. The intersubject variability of CLR and CLH were 61% and 207%, respectively. The typical value of the volume of distribution Vp =366 ml (Vp /WTmedian =470 ml/kg), and its intersubject variability was 38.8%. Half-life was 82.1h. Compared with more mature and older preterm populations studied previously, indomethacin CL is considerably lower in this contemporary population. Model-informed precision dosing incorporating important covariates other than weight alone offers an opportunity to individualize dosing in a susceptible patient undergoing rapid change.

A Model-Informed Precision Busulfan Dosing in Pediatric Patients Undergoing Hematopoietic Cell Transplantation- Real-World Data

Targeted i.v. busulfan dosing in pediatric non-malignant and malignant disorders undergoing HCT has been challenging due to the unpredictable PK of busulfan in young children. In 59 children with b-thalassemia undergoing HCT, we have previously reported that Bu TDM is feasible, and targeted IV Bu resulted in better EFS than untargeted oral bu in these patients (Balasubramanian et al., TCT 2019). However, the incidence of graft rejection was still high, probably due to age-related clearance of busulfan in very young children (Gaziev J et al. Blood. 2010). Precision dosing models result in less inter-patient variability in systemic exposure and better treatment outcomes (Shukla P et al. Front Pharmacol. 2020). We carried out a pilot study of precision busulfan dosing in pediatric patients in collaboration with InsightRX(California, USA). We aimed to evaluate the performance of the InsightRX precision dosing model for busulfan in young patients undergoing HCT and compare with the incidence of mixed chimerism on days 30, 60, and 90 post HCT [> 5% recipient cells], the time to neutrophil and platelet engraftment and the incidence of regimen related toxicities. Thirteen patients were enrolled in the study from June 2021 to May 2022, and the demographics are listed in Table 1. Initial Bu dosing was based on EMA guidelines (Body weight-based dosing, Palmer J et al. BBMT. 2016). Busulfan levels were monitored after doses 1, 5, 9, and 13 using a validated LC-MS/MS method (Desire et al., 2013). The values were entered in the InsightRX TDM algorithm, and a cumulative AUC of 80-90mg.hr/L was targeted. The mean Bu dose per EMA guideline and the actual dose received after precision dosing was compared. Based on InsightRX prediction, ten patients received higher cumulative Bu dosing, and 3 received lower cumulative Bu dosing compared to fixed weight-based dosing (Fig 1A). All patients had lower Bu AUC (<4.6 mg*hr/L) after the 1st dose, while at the end of the 13th dose, most patients achieved therapeutic Bu AUC (Median: 4.8 (3.6-5.8 mg*hr/L, Fig 1B). The interpatient variability in systemic exposure was only 12% with precision dosing, compared to 33% with the trapezoidal rule-based method (Balasubramanian et al., TCT 2019). Irrespective of the busulfan formulation, GST genotype, and underlying disease, the total cumulative target AUC was achieved in all the patients; two patients rejected the graft (one rejected within day+90 and the other rejected late after day+90, both had unrelated donor source and lowered cumulative Bu AUC compared to the other patients), and one developed Sinusoidal obstruction syndrome (despite lower cumulative Bu AUC, therefore cannot be attributed to Bu exposure). Targeting higher cumulative Bu exposure using a model-informed precision dosing software resulted in better HCT outcomes. Our study suggests the feasibility of improved precision of dosing of busulfan in young children undergoing HCT resulting in improved outcomes. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Model-informed precision dosing of beta-lactam antibiotics and ciprofloxacin in critically ill patients: a multicentre randomised clinical trial

PurposeIndividualising drug dosing using model-informed precision dosing (MIPD) of beta-lactam antibiotics and ciprofloxacin has been proposed as an alternative to standard dosing to optimise antibiotic efficacy in critically ill patients. However, randomised clinical trials (RCT) on clinical outcomes have been lacking.MethodsThis multicentre RCT, including patients admitted to the intensive care unit (ICU) who were treated with antibiotics, was conducted in eight hospitals in the Netherlands. Patients were randomised to MIPD with dose and interval adjustments based on monitoring serum drug levels (therapeutic drug monitoring) combined with pharmacometric modelling of beta-lactam antibiotics and ciprofloxacin. The primary outcome was ICU length of stay (LOS). Secondary outcomes were ICU mortality, hospital mortality, 28-day mortality, 6-month mortality, delta sequential organ failure assessment (SOFA) score, adverse events and target attainment.ResultsIn total, 388 (MIPD n = 189; standard dosing n = 199) patients were analysed (median age 64 [IQR 55–71]). We found no significant differences in ICU LOS between MIPD compared to standard dosing (10 MIPD vs 8 standard dosing; IRR = 1.16; 95% CI 0.96–1.41; p = 0.13). There was no significant difference in target attainment before intervention at day 1 (T1) (55.6% MIPD vs 60.9% standard dosing; p = 0.24) or at day 3 (T3) (59.5% vs 60.4%; p = 0.84). There were no significant differences in other secondary outcomes.ConclusionsWe could not show a beneficial effect of MIPD of beta-lactam antibiotics and ciprofloxacin on ICU LOS in critically ill patients. Our data highlight the need to identify other approaches to dose optimisation.

S824 Cost-Effectiveness of Precision-Guided Dosing in Adult Crohn’s Disease Patients Initiating Infliximab Maintenance Therapy

Introduction: Crohn’s disease (CD) patients that lose response to biologics experience reduced quality of life (QoL) and costly hospitalizations. Precision-guided dosing provides clinicians with a comprehensive pharmacokinetic (PK) profile that allows for the next biologic dose to be personalized. We analyzed the cost-effectiveness of infliximab (IFX) Precision-guided dosing relative to two IFX dose intensification strategies (DIS). Methods: We developed a hybrid (Markov and decision tree) model of CD patients who had a clinical response to IFX induction and entered IFX maintenance in “remission” or “mild symptoms” health states. The analysis took a US payer perspective, a time horizon of 2 years in the base case, and a cycle length of 4 weeks. There were 3 comparators for IFX dosing: Precision-guided dosing, dose intensification based on symptoms, inflammatory markers, and trough IFX concentration (DIS1), and IFX dose intensification based on symptoms alone (DIS2). Patients that failed IFX initiated ustekinumab (UST), followed by vedolizumab, and conventional therapy. Transition probabilities for IFX were estimated from real-world clinical PK data and interventional clinical trial (PMID: 34978325; 29317275) patient-level data. All other transition probabilities were derived from published randomized clinical trials and cost-effectiveness analyses. Utility values were sourced from previous health technology assessments. Direct costs included biologic acquisition and infusion, surgeries and procedures, conventional therapy, and lab testing. The primary outcomes were total discounted costs, total quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs). The robustness of results was assessed via one-way sensitivity and scenario analyses. Results: Total costs and QALYs at 2 years are presented in Table. The ICERs of Precision-guided dosing relative to DIS1 and DIS2 were 171,810 and 127,990, respectively. One-way sensitivity analyses (Figure) demonstrated that the cost-effectiveness of Precision-guided dosing was most sensitive to the time between IFX doses. Precision-guided dosing had the lowest proportion of patients requiring a new biologic through 2 years (0.5% vs 2.4% and 27.7% for DIS1 and DIS2, respectively). Conclusion: Precision-guided dosing provides substantial clinical and QoL benefits for adult CD patients by maintaining clinical remission and avoiding IFX failure; it is cost-effective relative to other DISs at a WTP of $175,000/QALY.Figure 1.: Precision-guided Dosing ICER Relative to DIS2 † Varied by ±10% * Varied by ±5% CD: Crohn's disease, DIS: dose intensification strategy, ICER: incremental cost-effectiveness ratio, IFX: infliximab, M-S: moderate-severe, PMPM: per member per month, Rem: remission Table 1. - Base Case Results (2-year Horizon Discounted) DIS Total QALYs Total Costs ICER Relative to Precision-guided Dosing ICER Relative to DIS1 ICER Relative to DIS2 Incremental NMB vs DIS1 (WTP $150,000/QALY) Incremental NMB vs DIS1 (WTP $50,000/QALY) Incremental NMB vs DIS2 (WTP $150,000/QALY) Incremental NMB vs DIS2 (WTP $50,000/QALY) Precision-guided Dosing* 1.572 $50,753 - 171,810 127,990 -$765 -$4,274 $1,816 -$6,436 DIS1† 1.537 $44,725 171,810 - 95,581 - - $2,581 -$2,162 DIS2 ‡ 1.489 $40,191 127,990 95,581 - -$2,581 $2,162 - - DIS: dose intensification strategy, ICER: incremental cost-effectiveness ratio, NMB: net monetary benefit, QALY: quality-adjusted life year, WTP: willingness to pay.*Model informed precision dosing with homogenous mobility shift assay (Prometheus Laboratories).†Dose intensification based on a combination of symptoms, inflammatory markers and proactive therapeutic drug monitoring. Corresponds to cohorts 1 and 2 of the TAILORIX clinical trial.‡Dose intensification reactive on symptoms only. Corresponds to cohort 3 of the TAILORIX clinical trial.

Implementation and Cross-Validation of a Pharmacokinetic Model for Precision Dosing of Busulfan in Hematopoietic Stem Cell Transplanted Children.

Busulfan, a drug used in conditioning prior to hematopoietic stem cell transplantation (HSCT) in children, has a narrow therapeutic margin. The model-informed precision dosing (MIPD) of busulfan is desirable, but there is a lack of validated tools. The objective of this study was to implement and cross-validate a population pharmacokinetic (PK) model in the Tucuxi software for busulfan MIPD in HSCT children. A search of the literature was performed to identify candidate population PK models. The goodness of fit of three selected models was assessed in a dataset of 178 children by computing the mean error (ME) and root-mean-squared error of prediction (RMSE). The best model was implemented in Tucuxi. The individual predicted concentrations, the area under the concentration-time curve (AUC), and dosage requirements were compared between the Tucuxi model and a reference model available in the BestDose software in a subset of 61 children. The model from Paci et al. best fitted the data in the full dataset. In a subset of 61 patients, the predictive performance of Tucuxi and BestDose models was comparable with ME values of 6.4% and −2.5% and RMSE values of 11.4% and 13.6%, respectively. The agreement between the estimated AUC and the predicted dose was good, with 6.6% and 4.9% of the values being out of the 95% limits of agreement, respectively. To conclude, a PK model for busulfan MIPD was cross-validated and is now available in the Tucuxi software.

Dosing for Personalized Prophylaxis in Hemophilia A Highly Varies on the Underlying Population Pharmacokinetic Models.

Model-informed personalized prophylaxis with factor VIII (FVIII) replacement therapy aimed at higher trough levels is becoming indispensable for patients with severe hemophilia A. This study aimed to identify the most suitable population pharmacokinetic (PK) models for personalized prophylaxis using various FVIII products and 2 clinical assays and to implement the most suitable one in open-access software. Twelve published population PK models were systematically compared to predict the time above target (TaT) for a reference dosing occasion. External validation was performed using a 5-point PK data from 39 adult patients with hemophilia A with FVIII measured by chromogenic substrate (CSA) and 1-stage assays (OSAs) using NONMEM under 3 different conditions: a priori (with all FVIII samples blinded), a posteriori (with 1 trough sample), and general model fit (with all FVIII samples including the reference dosing occasion provided). On average, the baseline covariate models overpredicted TaT (a priori; bias -3.8 hours to 49.6 hours). When additionally including 1 previous trough FVIII sample before the reference dosing occasion (a posteriori), only 50% of the models improved in bias (-1.0 hours to 36.5 hours) and imprecision (22.4 hours and 60.7 hours). Using all the time points (general model fit), the models accurately predicted (individual TaT less than ±12 hours compared with the reference) 62%-90% and 33%-74% of the patients using CSA and OSA data, respectively. Across all scenarios, predictions using CSA data were more accurate than those using the OSA data. One model performed best across the population (bias: -3.8 hours a priori, -1.0 hours a posteriori , and 0.6 hours general model fit ) and acceptably predicted 44% (a priori) to 90% ( general model fit ) of the patients. To allow the community-based evaluation of patient-individual FVIII dosing, this model was implemented in the open-access model-informed precision dosing software "TDMx."

Evaluating and Improving Neonatal Gentamicin Pharmacokinetic Models Using Aggregated Routine Clinical Care Data.

Model-informed precision dosing (MIPD) can aid dose decision-making for drugs such as gentamicin that have high inter-individual variability, a narrow therapeutic window, and a high risk of exposure-related adverse events. However, MIPD in neonates is challenging due to their dynamic development and maturation and by the need to minimize blood sampling due to low blood volume. Here, we investigate the ability of six published neonatal gentamicin population pharmacokinetic models to predict gentamicin concentrations in routine therapeutic drug monitoring from nine sites in the United State (n = 475 patients). We find that four out of six models predicted with acceptable levels of error and bias for clinical use. These models included known important covariates for gentamicin PK, showed little bias in prediction residuals over covariate ranges, and were developed on patient populations with similar covariate distributions as the one assessed here. These four models were refit using the published parameters as informative Bayesian priors or without priors in a continuous learning process. We find that refit models generally reduce error and bias on a held-out validation data set, but that informative prior use is not uniformly advantageous. Our work informs clinicians implementing MIPD of gentamicin in neonates, as well as pharmacometricians developing or improving PK models for use in MIPD.

Physiologically-based pharmacokinetic modelling and dosing evaluation of gentamicin in neonates using PhysPK.

Each year, infections caused around the 25% of neonatal deaths. Early empirical treatments help to reduce this mortality, although optimized dosing regimens are still lacking. The aims were to develop and validate a gentamicin physiologically-based pharmacokinetic (PBPK) model and then potentially explore dosing regimens in neonates using pharmacokinetic and pharmacodynamic criteria. The PBPK model developed consisted of 2 flow-limited tissues: kidney and other tissues. It has been implemented on a new tool called PhysPK, which allows structure reusability and evolution as predictive engine in Model-Informed Precision Dosing (MIPD). Retrospective pharmacokinetic information based on serum levels data from 47 neonates with gestational age between 32 and 39 weeks and younger than one-week postnatal age were used for model validation. The minimal PBPK model developed adequately described the gentamicin serum concentration-time profile with an average fold error nearly 1. Extended interval gentamicin dosing regimens (6 mg/kg q36h and 6 mg/kg q48h for term and preterm neonates, respectively) showed efficacy higher than 99% with toxicity lower than 10% through Monte Carlo simulation evaluations. The gentamicin minimal PBPK model developed in PhysPK from literature information, and validated in preterm and term neonates, presents adequate predictive performance and could be useful for MIPD strategies in neonates.

Pharmacogenomics in psychiatry - the challenge of cytochromeP450 enzyme phenoconversion and solutions to assist precision dosing.

Pharmacogenomic (PGx) testing of cytochrome P450 (CYP) enzymes may improve the efficacy and/or safety of some medications. This isfacilitated by increased availability and affordability of genotyping, the development of clinical practice PGx guidelines and regulatory support. However, the common occurrence of CYP phenoconversion, a mismatch between genotype-predicted CYP phenotype and the actual CYP phenotype, currently limits the application of PGx testing for precision dosing in psychiatry. This review proposes a stepwise approach to assist precision dosing in psychiatry via the introduction of PGx stewardship programs and innovative PGx education strategies. A future perspective on delivering precision dosing for psychiatrists is discussed thatinvolves innovative clinical decision support systems powered by model-informed precision dosing.

An interpretable stacking ensemble learning framework based on multi-dimensional data for real-time prediction of drug concentration: The example of olanzapine.

Background and Aim: Therapeutic drug monitoring (TDM) has evolved over the years as an important tool for personalized medicine. Nevertheless, some limitations are associated with traditional TDM. Emerging data-driven model forecasting [e.g., through machine learning (ML)-based approaches] has been used for individualized therapy. This study proposes an interpretable stacking-based ML framework to predict concentrations in real time after olanzapine (OLZ) treatment. Methods: The TDM-OLZ dataset, consisting of 2,142 OLZ measurements and 472 features, was formed by collecting electronic health records during the TDM of 927 patients who had received OLZ treatment. We compared the performance of ML algorithms by using 10-fold cross-validation and the mean absolute error (MAE). The optimal subset of features was analyzed by a random forest-based sequential forward feature selection method in the context of the top five heterogeneous regressors as base models to develop a stacked ensemble regressor, which was then optimized via the grid search method. Its predictions were explained by using local interpretable model-agnostic explanations (LIME) and partial dependence plots (PDPs). Results: A state-of-the-art stacking ensemble learning framework that integrates optimized extra trees, XGBoost, random forest, bagging, and gradient-boosting regressors was developed for nine selected features [i.e., daily dose (OLZ), gender_male, age, valproic acid_yes, ALT, K, BW, MONO#, and time of blood sampling after first administration]. It outperformed other base regressors that were considered, with an MAE of 0.064, R-square value of 0.5355, mean squared error of 0.0089, mean relative error of 13%, and ideal rate (the percentages of predicted TDM within ± 30% of actual TDM) of 63.40%. Predictions at the individual level were illustrated by LIME plots, whereas the global interpretation of associations between features and outcomes was illustrated by PDPs. Conclusion: This study highlights the feasibility of the real-time estimation of drug concentrations by using stacking-based ML strategies without losing interpretability, thus facilitating model-informed precision dosing.

Systematic Evaluation of Voriconazole Pharmacokinetic Models without Pharmacogenetic Information for Bayesian Forecasting in Critically Ill Patients.

Voriconazole (VRC) is used as first line antifungal agent against invasive aspergillosis. Model-based approaches might optimize VRC therapy. This study aimed to investigate the predictive performance of pharmacokinetic models of VRC without pharmacogenetic information for their suitability for model-informed precision dosing. Seven PopPK models were selected from a systematic literature review. A total of 66 measured VRC plasma concentrations from 33 critically ill patients was employed for analysis. The second measurement per patient was used to calculate relative Bias (rBias), mean error (ME), relative root mean squared error (rRMSE) and mean absolute error (MAE) (i) only based on patient characteristics and dosing history (a priori) and (ii) integrating the first measured concentration to predict the second concentration (Bayesian forecasting). The a priori rBias/ME and rRMSE/MAE varied substantially between the models, ranging from −15.4 to 124.6%/−0.70 to 8.01 mg/L and from 89.3 to 139.1%/1.45 to 8.11 mg/L, respectively. The integration of the first TDM sample improved the predictive performance of all models, with the model by Chen (85.0%) showing the best predictive performance (rRMSE: 85.0%; rBias: 4.0%). Our study revealed a certain degree of imprecision for all investigated models, so their sole use is not recommendable. Models with a higher performance would be necessary for clinical use.

Population pharmacokinetic modeling of caffeine in preterm infants with apnea of prematurity: New findings from concomitant erythromycin and AHR genetic polymorphisms

Current standard-dose caffeine therapy results in significant intersubject variability. The aims of this study were to develop and evaluate population pharmacokinetic (PPK) models of caffeine in preterm infants through comprehensive screening of covariates and then to propose model-informed precision dosing of caffeine for this population. A total of 129 caffeine concentrations from 96 premature neonates were incorporated into this study. Comprehensive medical record and genotype data of these neonates were collected for analysis. PPK modeling was performed by a nonlinear mixed effects modeling program (NONMEM). Final models based on the current weight (CW) or body surface area (BSA) were evaluated via multiple graphic and statistical methods. The model-informed dosing regimen was performed through Monte Carlo simulations. In addition to CW or BSA, postnatal age, coadministration with erythromycin (ERY), and aryl hydrocarbon receptor coding gene (AHR) variant (rs2158041) were incorporated into the final PPK models. Multiple evaluation results showed satisfactory prediction performance and stability of the CW- and BSA-based models. Monte Carlo simulations demonstrated that trough concentrations of caffeine in preterm infants would be affected by concomitant ERY therapy and rs2158041 under varying dose regimens. For the first time, ERY and rs2158041 were found to be associated with the clearance of caffeine in premature infants. Similar predictive performance and stability were obtained for both CW- and BSA-based PPK models. These findings provide novel insights into caffeine precision therapy for preterm infants.

Implementation of Model-Based Dose Adjustment of Tobramycin in Adult Patients with Cystic Fibrosis.

Therapeutic drug monitoring (TDM) of tobramycin is widely performed in patients with cystic fibrosis (CF), but little is known about the value of model-informed precision dosing (MIPD) in this setting. We aim at reporting our experience with tobramycin MIPD in adult patients with CF. We analyzed data from adult patients with CF who received IV tobramycin and had model-guided TDM during the first year of implementation of MIPD. The predictive performance of a pharmacokinetic (PK) model was assessed. Observed maximal (Cmax) and minimal (Cmin) concentrations after initial dosing were compared with target values. We compared the initial doses and adjusted doses after model-based TDM, as well as renal function at the beginning and end of therapy. A total of 78 tobramycin courses were administered in 61 patients. After initial dosing set by physicians (mean, 9.2 ± 1.4 mg/kg), 68.8% of patients did not achieve the target Cmax ≥ 30 mg/L. The PK model fit the data very well, with a median absolute percentage error of 4.9%. MIPD was associated with a significant increase in tobramycin doses (p < 0.001) without significant change in renal function. Model-based dose suggestions were wellaccepted by the physicians and the expected target attainment for Cmax was 83%. To conclude, the implementation of MIPD was effective in changing prescribing practice and was not associated with nephrotoxic events in adult patients with CF.

Daptomycin Physiology-Based Pharmacokinetic Modeling to Predict Drug Exposure and Pharmacodynamics in Skin and Bone Tissues.

Daptomycin has been recommended in the treatment of bone and joint infection. Previous work showed that the approved dosage of daptomycin may be insufficient to achieve optimal exposure in patients with bone and joint infection. However, those studies assumed that bone exposure was similar to steady-state daptomycin-free plasma concentrations. We sought to establish a physiologically based pharmacokinetic (PBPK) model of daptomycin to describe the dynamics of daptomycin disposition in bone and skin tissue. A PBPK model of daptomycin was built using PK-Sim®. Daptomycin concentrations in plasma and bone were obtained from three previously published studies. Physicochemical drug characteristics, mass balance, anthropometrics, and experimental data were used to build and refine the PBPK model. Internal validation of the PBPK model was performed using the usual diagnostic plots. The final PBPK model was then used to run simulations with doses of 6, 8, 10, and 12 mg/kg/24 h. Pharmacokinetic profiles were simulated in 1000 subjects and the probabilities of target attainment for the area under the concentration-time curve over the bacterial minimum inhibitory concentration were computed in blood, skin, and bone compartments. The final model showed a good fit of all datasets with an absolute average fold error between 0.5 and 2 for all pharmacokinetic quantities in blood, skin and bone tissues. Results of dosing simulations showed that doses ≥10 mg/kg should be used in the case of bacteremia caused by Staphylococcus aureus with a minimum inhibitory concentration >0.5 mg/L or Enterococcus faecalis with a minimum inhibitory concentration >1 mg/L, while doses ≥12 mg/kg should be used in the case of bone and joint infection or complicated skin infection. When considering a lower minimum inhibitory concentration, doses of 6-8 mg/kg would likely achieve a sufficient success rate. However, in the case of infections caused by E. faecalis with a minimum inhibitory concentration >2 mg/L, a higher dosage and combination therapy would be necessary to maximize efficacy. We developed the first daptomycin PBPK/pharmacodynamic model for bone and joint infection, which confirmed that a higher daptomycin dosage is needed to optimize exposure in bone tissue. However, such higher dosages raise safety concerns. In this setting, therapeutic drug monitoring and model-informed precision dosing appear necessary to ensure the right exposure on an individual basis.

Daptomycin Population Pharmacokinetics in Patients Affected by Severe Gram-Positive Infections: An Update.

Daptomycin pharmacokinetics may not depend on renal function only and it significantly differs between healthy volunteers and severely ill patients. Herein, we propose a population pharmacokinetics model based on 424 plasma daptomycin concentrations collected from 156 patients affected by severe Gram-positive infections during a routine therapeutic drug monitoring protocol. Model building and validation were performed using NONMEM 7.2 (ICON plc), Xpose4 and Perl-speaks-to-NONMEM. The final pop-PK model was a one-compartment first-order elimination model, with a 2.7% IIV for drug clearance (Cl), influence of creatinine clearance on drug clearance and of sex on distribution volume. After model validation, we simulated 10,000 patients with the Monte-Carlo method to predict the efficacy and tolerability of different daptomycin daily dosages. For the most common 6 mg/kg daily dose, the simulated probability of overcoming the toxic minimum concentration (24.3 mg/L) was 14.8% and the efficacy (expressed as a cumulative fraction of response) against methicillin-resistant S. aureus, S. pneumoniae and E. faecium was 95.77%, 99.99% and 68%, respectively. According to the model-informed precision dosing paradigm, pharmacokinetic models such as ours could help clinicians to perform patient-tailored antimicrobial dosing and maximize the odds of therapy success without neglecting toxicity risks.