Fundamentals of population pharmacokinetic modelling: validation methods

The Role of Pharmacometrics in Advancing the Therapies for Autoimmune Diseases

Autoimmune diseases (AIDs) are a group of disorders in which the immune system attacks the body's own tissues, leading to chronic inflammation and organ damage. These diseases are difficult to treat due to variability in drug PK among individuals, patient responses to treatment, and the side effects of long-term immunosuppressive therapies. In recent years, pharmacometrics has emerged as a critical tool in drug discovery and development (DDD) and precision medicine. The aim of this review is to explore the diverse roles that pharmacometrics has played in addressing the challenges associated with DDD and personalized therapies in the treatment of AIDs. Methods: This review synthesizes research from the past two decades on pharmacometric methodologies, including Physiologically Based Pharmacokinetic (PBPK) modeling, Pharmacokinetic/Pharmacodynamic (PK/PD) modeling, disease progression (DisP) modeling, population modeling, model-based meta-analysis (MBMA), and Quantitative Systems Pharmacology (QSP). The incorporation of artificial intelligence (AI) and machine learning (ML) into pharmacometrics is also discussed. Results: Pharmacometrics has demonstrated significant potential in optimizing dosing regimens, improving drug safety, and predicting patient-specific responses in AIDs. PBPK and PK/PD models have been instrumental in personalizing treatments, while DisP and QSP models provide insights into disease evolution and pathophysiological mechanisms in AIDs. AI/ML implementation has further enhanced the precision of these models. Conclusions: Pharmacometrics plays a crucial role in bridging pre-clinical findings and clinical applications, driving more personalized and effective treatments for AIDs. Its integration into DDD and translational science, in combination with AI and ML algorithms, holds promise for advancing therapeutic strategies and improving autoimmune patients' outcomes.

External evaluation of neonatal vancomycin population pharmacokinetic models: Moving from first-order equations to Bayesian-guided therapeutic monitoring

INTRODUCTION

Guidelines for vancomycin therapeutic monitoring recommend using a Bayesian approach with a population pharmacokinetic model to estimate the 24 h area under the concentration-time curve over first-order equations. Thus, we performed an external evaluation of population pharmacokinetic models of vancomycin in neonates and compared Bayesian results with those observed in clinical practice via pharmacokinetic equations to improve therapeutic monitoring by proposing optimized initial dosing nomograms and assessing the feasibility of reduced blood sampling strategies using the most predictive models.

METHODS

Models were identified from the literature and evaluated via an external neonatal population. A priori predictive performance was first assessed by prediction-based diagnostics, then by simulation-based diagnostics and a posteriori analyses only if deemed satisfactory; model-informed vancomycin exposure was also compared with reference first-order pharmacokinetic equations. The best-performing models were ultimately subjected to Monte Carlo simulations to develop new initial dosing nomograms offering the highest probability of achieving therapeutic target.

RESULTS

A total of 28 population pharmacokinetic models were evaluated in the external dataset, which includes 72 neonates and 380 vancomycin concentrations. Eleven models had an adequate predictive performance with bias ≤ ± 15% and imprecision ≤ 30%, while the Bayesian approach yielded over 75% agreement with reference exposure values in most cases. Nonetheless, Capparelli et al. and Mehrotra et al. models performed the best overall, showing the lowest imprecisions of 16.8% and 16.9%, respectively; both models recommended higher dosage regimens than the theoretical nomogram currently applied to favor therapeutic target attainment.

DISCUSSION

We externally evaluated numerous neonatal population pharmacokinetic models of vancomycin and used the most predictive ones to advocate new initial dosing nomograms. Clinical implementation of the Bayesian approach could reduce the time needed to reach therapeutic target and limit the number of blood samples in newborns compared with traditional pharmacokinetic equations.

Recommended approaches for integration of population pharmacokinetic modelling with precision dosing in clinical practice

Current methods of dose determination have contributed to suboptimal and inequitable health outcomes in underrepresented patient populations. The persistent demand to individualise patient treatment, alongside increasing technological feasibility, is leading to a growing adoption of model-informed precision dosing (MIPD) at point of care. Population pharmacokinetic (popPK) modelling is a technique that supports treatment personalisation by characterising drug exposure in diverse patient groups. This publication addresses this important shift in clinical approach, by collating and summarising recommendations from literature. It seeks to provide standardised guidelines on best practices for the development of popPK models and their use in MIPD software tools, ensuring the safeguarding and optimisation of patient outcomes. Moreover, it consolidates guidance from key regulatory and advisory bodies on MIPD software deployment, as well as technical requirements for electronic health record integration. It also considers the future application and clinical impact of machine learning algorithms in popPK and MIPD. Ultimately, this publication aims to facilitate the incorporation of high-quality precision-dosing solutions into standard clinical workflows, thereby enhancing the effectiveness of individualised dose selection at point of care.

Dose Individualisation of Antimicrobials from a Pharmacometric Standpoint: The Current Landscape

Successful antimicrobial therapy depends on achieving optimal drug concentrations within individual patients. Inter-patient variability in pharmacokinetics (PK) and differences in pathogen susceptibility (reflected in the minimum inhibitory concentration, [MIC]) necessitate personalised approaches. Dose individualisation strategies aim to address this challenge, improving treatment outcomes and minimising the risk of toxicity and antimicrobial resistance. Therapeutic drug monitoring (TDM), with the application of population pharmacokinetic (popPK) models, enables model-informed precision dosing (MIPD). PopPK models mathematically describe drug behaviour across populations and can be combined with patient-specific TDM data to optimise dosing regimens. The integration of machine learning (ML) techniques promises to further enhance dose individualisation by identifying complex patterns within extensive datasets. Implementing these approaches involves challenges, including rigorous model selection and validation to ensure suitability for target populations. Understanding the relationship between drug exposure and clinical outcomes is crucial, as is striking a balance between model complexity and clinical usability. Additionally, regulatory compliance, outcome measurement, and practical considerations for software implementation will be addressed. Emerging technologies, such as real-time biosensors, hold the potential for revolutionising TDM by enabling continuous monitoring, immediate and frequent dose adjustments, and near patient testing. The ongoing integration of TDM, advanced modelling techniques, and ML within the evolving digital health care landscape offers a potential for enhancing antimicrobial therapy. Careful attention to model development, validation, and ethical considerations of the applied techniques is paramount for successfully optimising antimicrobial treatment for the individual patient.

Rutin and Physalis peruviana Extract: Population Pharmacokinetics in New Zealand Rabbits

Background/Objectives: An extract of calyces from Physalis peruviana with hypoglycemic activity is being considered as a potential herbal medicine. Preclinical pharmacokinetics (PK) studies of the extract in rats, focusing on plasma concentrations of its main compound, rutin, and its metabolites, revealed PK interactions in the extract matrix that improved the absorption of rutin metabolites compared to the pure compound, among other PK effects. This research aimed to study the PK of rutin alone and in the extract and assess potential PK interactions in the extract matrix on the flavonoid and its metabolites in rabbits, a nonrodent species; Methods: Animals received pure rutin or extract orally and intravenously. The PK analysis used noncompartmental and population pharmacokinetics (popPK) methods, and simple allometry was applied to predict human PK parameters; Results: The rutin concentration-time profile fit a two-compartment model with first-order elimination, while its metabolites fit a double first-order absorption model. The extract matrix led to increased absorption, distribution, and elimination of rutin as well as increased bioavailability of its metabolites in rabbits; Conclusions: The popPK model defined the equations for PK parameters describing these findings, and the increased volume of distribution and clearance of rutin was maintained in human predictions. These results will support the development of a new herbal medicine.

Toward Personalized Salbutamol Therapy: Validating Virtual Patient-Derived Population Pharmacokinetic Model with Real-World Data

Interindividual variability, influenced by patient-specific factors including age, weight, gender, race, and genetics, among others, contributes to variations in therapeutic response. Population pharmacokinetic (popPK) modeling is an essential tool for pinpointing measurable factors affecting dose-concentration relationships and tailoring dosage regimens to individual patients. Herein, we developed a popPK model for salbutamol, a short-acting β2-agonist (SABA) used in asthma treatment, to identify key patient characteristics that influence treatment response. To do so, synthetic data from physiologically-based pharmacokinetic (PBPK) models was employed, followed by an external validation using real patient data derived from an equivalent study. Thirty-two virtual patients were included in this study. A two-compartment model, with first-order absorption (no delay), and linear elimination best fitted our data, according to diagnostic plots and selection criteria. External validation demonstrated a strong agreement between individual predicted and observed values. The incorporation of covariates into the basic structural model identified a significant impact of age on clearance (Cl) and intercompartmental clearance (Q); gender on Cl and the constant rate of absorption (ka); race on Cl; and weight on Cl in the volume of distribution of the peripheral compartment (V2). This study addresses critical challenges in popPK modeling, particularly data scarcity, incompleteness, and homogeneity, in traditional clinical trials, by leveraging synthetic data from PBPK modeling. Significant associations between individual characteristics and salbutamol's PK parameters, here uncovered, highlight the importance of personalized therapeutic regimens for optimal treatment outcomes.

Optimization of Vancomycin Initial Dosing Regimen in Neonates Using an Externally Evaluated Population Pharmacokinetic Model

BACKGROUND

Vancomycin therapeutic monitoring guidelines were revised in March 2020, and a population pharmacokinetics-guided Bayesian approach to estimate the 24-hour area under the concentration-time curve to the minimum inhibitory concentration ratio has since been recommended instead of trough concentrations. To comply with these latest guidelines, we evaluated published population pharmacokinetic models of vancomycin using an external dataset of neonatal patients and selected the most predictive model to develop a new initial dosing regimen.

METHODS

The models were identified from the literature and tested using a retrospective dataset of Canadian neonates. Their predictive performance was assessed using prediction- and simulation-based diagnostics. Monte Carlo simulations were performed to develop the initial dosing regimen with the highest probability of therapeutic target attainment.

RESULTS

A total of 144 vancomycin concentrations were derived from 63 neonates in the external population. Five of the 28 models retained for evaluation were found predictive with a bias of 15% and an imprecision of 30%. Overall, the Grimsley and Thomson model performed best, with a bias of -0.8% and an imprecision of 20.9%; therefore, it was applied in the simulations. A novel initial dosing regimen of 15 mg/kg, followed by 11 mg/kg every 8 hours should favor therapeutic target attainment.

CONCLUSIONS

A predictive population pharmacokinetic model of vancomycin was identified after an external evaluation and used to recommend a novel initial dosing regimen. The implementation of these model-based tools may guide physicians in selecting the most appropriate initial vancomycin dose, leading to improved clinical outcomes.

Therapeutic drug monitoring in inflammatory bowel diseases. Position statement of the Spanish Working Group on Crohn's Disease and Ulcerative Colitis

The treatment of inflammatory bowel disease has undergone a significant transformation following the introduction of biologic drugs. Thanks to these drugs, treatment goals have evolved from clinical response and remission to more ambitious objectives, such as endoscopic or radiologic remission. However, even though biologics are highly effective, a significant percentage of patients will not achieve an initial response or may lose it over time. We know that there is a direct relationship between the trough concentrations of the biologic and its therapeutic efficacy, with more demanding therapeutic goals requiring higher drug levels, and inadequate exposure being common. Therapeutic drug monitoring of biologic medications, along with pharmacokinetic models, provides us with the possibility of offering a personalized approach to treatment for patients with IBD. Over the past few years, relevant information has accumulated regarding its utility during or after induction, as well as in the maintenance of biologic treatment, in reactive or proactive strategies, and prior to withdrawal or treatment de-escalation. The aim of this document is to establish recommendations regarding the utility of therapeutic drug monitoring of biologics in patients with inflammatory bowel disease, in different clinical practice scenarios, and to identify areas where its utility is evident, promising, or controversial.

Quantifying the Effect of Methotrexate on Adalimumab Response in Psoriasis by Pharmacokinetic-Pharmacodynamic Modeling

Previously, we showed that the combination of methotrexate and adalimumab treatment leads to less antidrug antibody development. In this study, we quantify the pharmacokinetics/pharmacodynamics (PK/PD) of adalimumab and evaluate the influence of methotrexate cotreatment. A population PK-PD model was developed using prospective data from 59 patients with psoriasis (baseline PASI = 12.6) receiving adalimumab over 49 weeks. Typical PK and PD parameters and their corresponding interpatient variability were estimated. We performed a covariate analysis to assess whether interpatient variability could be explained by addition of methotrexate and other covariates. In total, 330 PASIs, 252 adalimumab serum concentrations, and 247 antidrug antibody titers were available. Presence of antidrug antibodies (adalimumab group = 46.7%, adalimumab + methotrexate group = 38.7%; P = .031) was correlated with increased adalimumab apparent clearance (P < .001). In the PD model, the use of concomitant methotrexate was borderline to significantly correlated with a decreased half-maximal inhibitory concentration (adalimumab concentration for which clinical response score is reduced by half; P < .10). On the basis of our PK-PD model, concomitant use of methotrexate indirectly increases adalimumab concentration, partially through less antidrug antibodies formation, which may result in better efficacy.

Advancing Precision Medicine: A Review of Innovative In Silico Approaches for Drug Development, Clinical Pharmacology and Personalized Healthcare

The landscape of medical treatments is undergoing a transformative shift. Precision medicine has ushered in a revolutionary era in healthcare by individualizing diagnostics and treatments according to each patient's uniquely evolving health status. This groundbreaking method of tailoring disease prevention and treatment considers individual variations in genes, environments, and lifestyles. The goal of precision medicine is to target the "five rights": the right patient, the right drug, the right time, the right dose, and the right route. In this pursuit, in silico techniques have emerged as an anchor, driving precision medicine forward and making this a realistic and promising avenue for personalized therapies. With the advancements in high-throughput DNA sequencing technologies, genomic data, including genetic variants and their interactions with each other and the environment, can be incorporated into clinical decision-making. Pharmacometrics, gathering pharmacokinetic (PK) and pharmacodynamic (PD) data, and mathematical models further contribute to drug optimization, drug behavior prediction, and drug-drug interaction identification. Digital health, wearables, and computational tools offer continuous monitoring and real-time data collection, enabling treatment adjustments. Furthermore, the incorporation of extensive datasets in computational tools, such as electronic health records (EHRs) and omics data, is also another pathway to acquire meaningful information in this field. Although they are fairly new, machine learning (ML) algorithms and artificial intelligence (AI) techniques are also resources researchers use to analyze big data and develop predictive models. This review explores the interplay of these multiple in silico approaches in advancing precision medicine and fostering individual healthcare. Despite intrinsic challenges, such as ethical considerations, data protection, and the need for more comprehensive research, this marks a new era of patient-centered healthcare. Innovative in silico techniques hold the potential to reshape the future of medicine for generations to come.

Comparison Pharmacokinetic Dosing Tools in Hemophilia A Children

Prophylaxis is the gold standard for the management of hemophilia A patients. It has been shown that prophylaxis regulated with pharmacokinetic (PK) data reduces frequency of bleeding and cost of treatment. To determine the best prophylaxis regimen, PK dosing tools using the Bayesian method have been developed. We aimed to compare two PK dosing tools. Blood samples were drawn before, 4, 24, and 48 h after FVIII infusions from patients with severe hemophilia A and inhibitor negative. FVIII levels were measured by PTT-based one-stage assay method. PK parameters obtained using WAPPS and myPKFiT, which are web-accessible PK dosing tools using Bayesian algorithm, and daily prophylaxis dose estimated by the programs were compared. Forty-two hemophilia A patients [median age 13 years (IQR 8.9-16.4)] included in the study. There was no difference between the daily dose of FVIII given for prophylaxis and the dose recommended by the myPKFiT for the 1% trough level; whereas, a significant difference was found with the WAPPS. The half-lives of FVIII did not differ between the two dosing tools; however, significant differences were found in the estimated dose, clearances, and times to 1% trough level. There was no significant difference between PK data of patients who received Advate® and those who received non-Advate® factor concentrates. Choice of PK dosing tool can affect recommended FVIII dose. However, target trough levels should be individualized according to bleeding phenotype and daily activity of patient.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12288-023-01671-0.

Evidence supporting the use of therapeutic drug monitoring of ganciclovir in transplantation

PURPOSE OF REVIEW

This review describes current knowledge of ganciclovir (GCV) and valganciclovir (ValGCV) pharmacokinetic/pharmacodynamic characteristics, highlighting the likely contribution from host genetic factors to interpatient variability. The evidence and challenges surrounding optimization of drug dosing through therapeutic drug monitoring (TDM) are examined, with recommendations made.

RECENT FINDINGS

Pharmacokinetic studies of current dosing guidelines have shown high interindividual and intraindividual variability of GCV concentrations. This is sometimes associated with a slow decline in cytomegalovirus (CMV) viral load in some transplant recipients. A high incidence of GCV-associated myelosuppression has limited the use of this drug in the transplant setting. Patient groups identified to benefit from GCV TDM include pediatric patients, cystic fibrosis with lung transplantation, obese with kidney transplantation, and patients with fluctuating renal function or on hemodialysis. The emergence of refractory resistant CMV, particularly in immune compromised patients, highlights the importance of appropriate dosing of these antivirals. Host genetic factors need to be considered where recently, two host genes were shown to account for interpatient variation during ganciclovir therapy. Therapeutic Drug Monitoring has been shown to improve target antiviral-level attainment. The use of TDM may guide concentration-based dose adjustment, potentially improving virological and clinical outcomes. However, evidence supporting the use of TDM in clinical practice remains limited and further study is needed in the transplant cohort.

SUMMARY

Further studies examining novel biomarkers are needed to guide target concentrations in prophylaxis and treatment. The use of TDM in transplant recipients is likely to improve the clinical efficacy of current antivirals and optimize outcomes in transplant recipients.

Tutorial on model selection and validation of model input into precision dosing software for model-informed precision dosing

There has been rising interest in using model-informed precision dosing to provide personalized medicine to patients at the bedside. This methodology utilizes population pharmacokinetic models, measured drug concentrations from individual patients, pharmacodynamic biomarkers, and Bayesian estimation to estimate pharmacokinetic parameters and predict concentration-time profiles in individual patients. Using these individualized parameter estimates and simulated drug exposure, dosing recommendations can be generated to maximize target attainment to improve beneficial effect and minimize toxicity. However, the accuracy of the output from this evaluation is highly dependent on the population pharmacokinetic model selected. This tutorial provides a comprehensive approach to evaluating, selecting, and validating a model for input and implementation into a model-informed precision dosing program. A step-by-step outline to validate successful implementation into a precision dosing tool is described using the clinical software platforms Edsim++ and MwPharm++ as examples.

Real world population pharmacokinetic study in children and young adults with inflammatory bowel disease discovers novel blood and stool microbial predictors of vedolizumab clearance

BACKGROUND

Vedolizumab for inflammatory bowel disease (IBD) is often intensified based on distinct pharmacokinetics in children. Prior adult-specific population pharmacokinetic models have identified limited covariates of drug clearance.

AIMS

To establish a population pharmacokinetic model for children and young adults to identify novel covariates of drug clearance to better account for paediatric-specific inter-patient variability in vedolizumab pharmacokinetics; a key secondary exploratory aim was to identify microbial signatures of pharmacokinetic outcomes in a subset of patients.

METHODS

The study included data from 463 observed vedolizumab concentrations (59 peaks and 404 troughs) from 74 patients with IBD (52 with Crohn's disease and 22 with ulcerative colitis or unclassified IBD, median age 16 years). Pharmacokinetic analysis was conducted with non-linear mixed effects modelling. For the evaluation of the exposure-response relationship, clinical outcomes were evaluated by trough levels, clearance and vedolizumab exposure. Whole-genome metagenomic sequencing was conducted at baseline and week 2.

RESULTS

A two-compartment population pharmacokinetic model was identified with a clear correlation between CL and weight, erythrocyte sedimentation rate, and hypoalbuminemia. Trough concentrations before infusion 3 (37 μg/ml) and before infusion 4 (20 μg/ml) best predicted steroid-free clinical remission at infusion 4. Using faecal metagenomics, we identified an early (baseline and week 2) abundance of butyrate-producing species and pathways that were associated with an infusion 4 trough concentration >20 μg/ml.

CONCLUSIONS

This novel paediatric vedolizumab pharmacokinetic model could inform precision dosing. While additional studies are needed, an abundance of faecal butyrate producers is associated with early response to vedolizumab, suggesting that microbial analysis may be beneficial to biological selection.

Beta-Lactam Probability of Target Attainment Success: Cefepime as a Case Study

Introduction: Probability of target attainment (PTA) analysis using Monte Carlo simulations has become a mainstay of dose optimization. We highlight the technical and clinical factors that may affect PTA for beta-lactams. Methods: We performed a mini review in adults to explore factors relating to cefepime PTA success and how researchers incorporate PTA into dosing decisions. In addition, we investigated, via simulations with a population pharmacokinetic (PK) model, factors that may affect cefepime PTA success. Results: The mini review included 14 articles. PTA results were generally consistent, given the differences in patient populations. However, dosing recommendations were more varied and appeared to depend on the definition of pharmacodynamic (PD) target, definition of PTA success and specific clinical considerations. Only 3 of 14 articles performed formal toxicological analysis. Simulations demonstrated that the largest determinants of cefepime PTA were the choice of PD target, continuous vs. intermittent infusion and creatinine clearance. Assumptions for protein binding, steady state vs. first dose, and simulating different sampling schemes may impact PTA success under certain conditions. The choice of one or two compartments had a minimal effect on PTA. Conclusions: PTA results may be similar with different assumptions and techniques. However, dose recommendation may differ significantly based on the selection of PD target, definition of PTA success and considerations specific to a patient population. Demographics and the PK parameters used to simulate time-concentration profiles should be derived from patient data applicable to the purpose of the PTA. There should be strong clinical rationale for dose selection. When possible, safety and toxicity should be considered in addition to PTA success.

Population Pharmacokinetic/Pharmacodynamic Modelling of Daptomycin for Schedule Optimization in Patients with Renal Impairment

The aims of this study are (i) to develop a population pharmacokinetic/pharmacodynamic model of daptomycin in patients with normal and impaired renal function, and (ii) to establish the optimal dose recommendation of daptomycin in clinical practice. Several structural PK models including linear and non-linear binding kinetics were evaluated. Monte Carlo simulations were conducted with a fixed combination of creatinine clearance (30-90 mL/min/1.73 m2) and body weight (50-100 kg). The final dataset included 46 patients and 157 daptomycin observations. A two-compartment model with first-order peripheral distribution and elimination kinetics assuming non-linear protein-binding kinetics was selected. The bactericidal effect for Gram+ strains with MIC ≤ 0.5 mg/L could be achieved with 5-12 mg/kg daily daptomycin based on body weight and renal function. The administration of 10-17 mg/kg q48 h daptomycin allows to achieve bactericidal effect for Gram+ strains with MIC ≤ 1 mg/L. Four PK samples were selected as the optimal sampling strategy for an accurate AUC estimation. A quantitative framework has served to characterize the non-linear binding kinetics of daptomycin in patients with normal and impaired renal function. The impact of different dosing regimens on the efficacy and safety outcomes of daptomycin treatment based on the unbound exposure of daptomycin and individual patient characteristics has been evaluated.

Machine learning advances the integration of covariates in population pharmacokinetic models: Valproic acid as an example

Background and Aim: Many studies associated with the combination of machine learning (ML) and pharmacometrics have appeared in recent years. ML can be used as an initial step for fast screening of covariates in population pharmacokinetic (popPK) models. The present study aimed to integrate covariates derived from different popPK models using ML.

Methods: Two published popPK models of valproic acid (VPA) in Chinese epileptic patients were used, where the population parameters were influenced by some covariates. Based on the covariates and a one-compartment model that describes the pharmacokinetics of VPA, a dataset was constructed using Monte Carlo simulation, to develop an XGBoost model to estimate the steady-state concentrations (Css) of VPA. We utilized SHapley Additive exPlanation (SHAP) values to interpret the prediction model, and calculated estimates of VPA exposure in four assumed scenarios involving different combinations of CYP2C19 genotypes and co-administered antiepileptic drugs. To develop an easy-to-use model in the clinic, we built a simplified model by using CYP2C19 genotypes and some noninvasive clinical parameters, and omitting several features that were infrequently measured or whose clinically available values were inaccurate, and verified it on our independent external dataset.

Results: After data preprocessing, the finally generated combined dataset was divided into a derivation cohort and a validation cohort (8:2). The XGBoost model was developed in the derivation cohort and yielded excellent performance in the validation cohort with a mean absolute error of 2.4 mg/L, root-mean-squared error of 3.3 mg/L, mean relative error of 0%, and percentages within ±20% of actual values of 98.85%. The SHAP analysis revealed that daily dose, time, CYP2C19*2 and/or *3 variants, albumin, body weight, single dose, and CYP2C19*1*1 genotype were the top seven confounding factors influencing the Css of VPA. Under the simulated dosage regimen of 500 mg/bid, the VPA exposure in patients who had CYP2C19*2 and/or *3 variants and no carbamazepine, phenytoin, or phenobarbital treatment, was approximately 1.74-fold compared to those with CYP2C19*1/*1 genotype and co-administered carbamazepine + phenytoin + phenobarbital. The feasibility of the simplified model was fully illustrated by its performance in our external dataset.

Conclusion: This study highlighted the bridging role of ML in big data and pharmacometrics, by integrating covariates derived from different popPK models.

Development and evaluation of uncertainty quantifying machine learning models to predict piperacillin plasma concentrations in critically ill patients

BACKGROUND

Beta-lactam antimicrobial concentrations are frequently suboptimal in critically ill patients. Population pharmacokinetic (PopPK) modeling is the golden standard to predict drug concentrations. However, currently available PopPK models often lack predictive accuracy, making them less suited to guide dosing regimen adaptations. Furthermore, many currently developed models for clinical applications often lack uncertainty quantification. We, therefore, aimed to develop machine learning (ML) models for the prediction of piperacillin plasma concentrations while also providing uncertainty quantification with the aim of clinical practice.

METHODS

Blood samples for piperacillin analysis were prospectively collected from critically ill patients receiving continuous infusion of piperacillin/tazobactam. Interpretable ML models for the prediction of piperacillin concentrations were designed using CatBoost and Gaussian processes. Distribution-based Uncertainty Quantification was added to the CatBoost model using a proposed Quantile Ensemble method, useable for any model optimizing a quantile function. These models are subsequently evaluated using the distribution coverage error, a proposed interpretable uncertainty quantification calibration metric. Development and internal evaluation of the ML models were performed on the Ghent University Hospital database (752 piperacillin concentrations from 282 patients). Ensuing, ML models were compared with a published PopPK model on a database from the University Medical Centre of Groningen where a different dosing regimen is used (46 piperacillin concentrations from 15 patients.).

RESULTS

The best performing model was the Catboost model with an RMSE and [Formula: see text] of 31.94-0.64 and 33.53-0.60 for internal evaluation with and without previous concentration. Furthermore, the results prove the added value of the proposed Quantile Ensemble model in providing clinically useful individualized uncertainty predictions and show the limits of homoscedastic methods like Gaussian Processes in clinical applications.

CONCLUSIONS

Our results show that ML models can consistently estimate piperacillin concentrations with acceptable and high predictive accuracy when identical dosing regimens as in the training data are used while providing highly relevant uncertainty predictions. However, generalization capabilities to other dosing schemes are limited. Notwithstanding, incorporating ML models in therapeutic drug monitoring programs seems definitely promising and the current work provides a basis for validating the model in clinical practice.

Influence of N-acetyltransferase 2 (NAT2) genotype/single nucleotide polymorphisms on clearance of isoniazid in tuberculosis patients: a systematic review of population pharmacokinetic models

Purpose

Significant pharmacokinetic variabilities have been reported for isoniazid across various populations. We aimed to summarize population pharmacokinetic studies of isoniazid in tuberculosis (TB) patients with a specific focus on the influence of N-acetyltransferase 2 (NAT2) genotype/single-nucleotide polymorphism (SNP) on clearance of isoniazid.

Methods

A systematic search was conducted in PubMed and Embase for articles published in the English language from inception till February 2022 to identify population pharmacokinetic (PopPK) studies of isoniazid. Studies were included if patient population had TB and received isoniazid therapy, non-linear mixed effects modelling, and parametric approach was used for building isoniazid PopPK model and NAT2 genotype/SNP was tested as a covariate for model development.

Results

A total of 12 articles were identified from PubMed, Embase, and hand searching of articles. Isoniazid disposition was described using a two-compartment model with first-order absorption and linear elimination in most of the studies. Significant covariates influencing the pharmacokinetics of isoniazid were NAT2 genotype, body weight, lean body weight, body mass index, fat-free mass, efavirenz, formulation, CD4 cell count, and gender. Majority of studies conducted in adult TB population have reported a twofold or threefold increase in isoniazid clearance for NAT2 rapid acetylators compared to slow acetylators.

Conclusion

The variability in disposition of isoniazid can be majorly attributed to NAT2 genotype. This results in a trimodal clearance pattern with a multi-fold increase in clearance of NAT2 rapid acetylators compared to slow acetylators. Further studies exploring the generalizability/adaptability of developed PopPK models in different clinical settings are required.

A Review of Population Pharmacokinetic Analyses of Linezolid

In recent years, many studies on population pharmacokinetics of linezolid have been conducted. This comprehensive review aimed to summarize population pharmacokinetic models of linezolid, by focusing on dosage optimization to maximize the probability of attaining a certain pharmacokinetic-pharmacodynamic parameter in special populations. We searched the PubMed and EMBASE databases for population pharmacokinetic analyses of linezolid using a parametric non-linear mixed-effect approach, including both observational and prospective trials. Of the 32 studies, 26 were performed in adults, four in children, and one in both adults and children. High between-subject variability was determined in the majority of the models, which was in line with the variability of linezolid concentrations previously detected in observational studies. Some studies found that patients with renal impairment, hepatic failure, advanced age, or low body weight had higher exposure and adverse reactions rates. In contrast, lower concentrations and therapeutic failure were associated with obese patients, young patients, and patients who had undergone renal replacement techniques. In critically ill patients, the inter-individual and intra-individual variability was even greater, suggesting that this population is at an even higher risk of underexposure and overexposure. Therapeutic drug monitoring may be warranted in a large proportion of patients given that the Monte Carlo simulations demonstrated that the one-size-fits-all labeled dosing of 600 mg every 12 h could lead to toxicity or therapeutic failure for high values of the minimum inhibitory concentration of the target pathogen. Further research on covariates, including renal function, hepatic function, and drug–drug interactions related to P-glycoprotein could help to explain variability and improve linezolid dosing regimens.

Recent lessons learned from population pharmacokinetic studies of mycophenolic acid: physiological, genomic, and drug interactions leading to the prediction of drug effects

INTRODUCTION

Mycophenolic acid (MPA) is a widely used immunosuppressant in transplantation and autoimmune disease. Highly variable pharmacokinetics have been observed with MPA, but the exact mechanisms remain largely unknown.

AREAS COVERED

The current review provided a critical, comprehensive update of recently published population pharmacokinetic/dynamic models of MPA (n=16 papers identified from PubMed and Embase, inclusive from January 2017 to August 2021), with specific emphases on the intrinsic and extrinsic factors influencing the pharmacology of MPA. The significance of the identified covariates, potential mechanisms, and comparisons to historical literature have been provided.

EXPERT OPINION

While select covariates affecting the population pharmacokinetics of MPA are consistently observed and mechanistically supported, some variables have not been regularly reported and/or lacked mechanistic explanation. Very few pharmacodynamic models were available, pointing to the need to extrapolate pharmacokinetic findings. Ideal models of MPA should consist of: i) utilizing optimal sampling points to allow the characterizations of absorption, re-absorption, and elimination phases; ii) characterizing unbound/total MPA, MPA metabolites, plasma/urinary concentrations, and genetic polymorphisms to facilitate mechanistic interpretations; and iii) incorporating actual outcomes and pharmacodynamic data to establish clinical relevance. We anticipate the field will continue to expand in the next 5 to 10 years.

Validation of population pharmacokinetic models: a comparison of internal and external validation approaches for hydrochlorothiazide

1. Model evaluation is an important issue in population analyses. Our aim was to perform and illustrate metrics and techniques for internal and external evaluation with an application to population pharmacokinetics of hydrochlorothiazide (HCTZ).2. A nonlinear mixed effects model was used to study the pharmacokinetics of HCTZ. In addition, different types of internal assessment tools and external metrics were used for model evaluation. External evaluation was performed using an alternative dataset that included data from an independent group of subjects. For comparison, a previously published population pharmacokinetic model for HCTZ was applied to the same data.3. A two-compartment model with first-order oral absorption using a constant time delay between administration and absorption and first-order elimination best described HCTZ pharmacokinetics. Age had a statistically significant effect on HCTZ clearance. The final model performed adequately in the internal and external assessment tests. The final model showed better predictive performance than the other previously published HCTZ model.4. Finally, a robust population pharmacokinetic model for HCTZ in adults was constructed and validated internally and externally. Incorporating analytical assessment of nonlinear pharmacokinetics into the modelling may be a promising approach to improve the predictive power of the model.

The PHARMACOM-EPI Framework for Integrating Pharmacometric Modelling Into Pharmacoepidemiological Research Using Real-World Data: Application to Assess Death Associated With Valproate

In pharmacoepidemiology, it is usually expected that the observed association should be directly or indirectly related to the pharmacological effects of the drug/s under investigation. Pharmacological effects are, in turn, strongly connected to the pharmacokinetic and pharmacodynamic properties of a drug, which can be characterized and investigated using pharmacometric models. Recently, the use of pharmacometrics has been proposed to provide pharmacological substantiation of pharmacoepidemiological findings derived from real-world data. However, validated frameworks suggesting how to combine these two disciplines for the aforementioned purpose are missing. Therefore, we propose PHARMACOM-EPI, a framework that provides a structured approach on how to identify, characterize, and apply pharmacometric models with practical details on how to choose software, format dataset, handle missing covariates/dosing data, how to perform the external evaluation of pharmacometric models in real-world data, and how to provide pharmacological substantiation of pharmacoepidemiological findings. PHARMACOM-EPI was tested in a proof-of-concept study to pharmacologically substantiate death associated with valproate use in the Danish population aged ≥ 65 years. Pharmacological substantiation of death during a follow-up period of 1 year showed that in all individuals who died (n = 169) individual predictions were within the subtherapeutic range compared with 52.8% of those who did not die (n = 1,084). Of individuals who died, 66.3% (n = 112) had a cause of death possibly related to valproate and 33.7% (n = 57) with well-defined cause of death unlikely related to valproate. This proof-of-concept study showed that PHARMACOM-EPI was able to provide pharmacological substantiation for death associated with valproate use in the study population.

Assessing the predictive performance of population pharmacokinetic models for intravenous polymyxin B in critically ill patients

Polymyxin B (PMB) has reemerged as a last‐line therapy for infections caused by multidrug‐resistant gram‐negative pathogens, but dosing is challenging because of its narrow therapeutic window and pharmacokinetic (PK) variability. Population PK (POPPK) models based on suitably powered clinical studies with appropriate sampling strategies that take variability into consideration can inform PMB dosing to maximize efficacy and minimize toxicity and resistance. Here we reviewed published PMB POPPK models and evaluated them using an external validation data set (EVD) of patients who are critically ill and enrolled in an ongoing clinical study to assess their utility. Seven published POPPK models were employed using the reported model equations, parameter values, covariate relationships, interpatient variability, parameter covariance, and unexplained residual variability in NONMEM (Version 7.4.3). The predictive ability of the models was assessed using prediction‐based and simulation‐based diagnostics. Patient characteristics and treatment information were comparable across studies and with the EVD (n = 40), but the sampling strategy was a main source of PK variability across studies. All models visually and statistically underpredicted EVD plasma concentrations, but the two‐compartment models more accurately described the external data set. As current POPPK models were inadequately predictive of the EVD, creation of a new POPPK model based on an appropriately powered clinical study with an informed PK sampling strategy would be expected to improve characterization of PMB PK and identify covariates to explain interpatient variability. Such a model would support model‐informed precision dosing frameworks, which are urgently needed to improve PMB treatment efficacy, limit resistance, and reduce toxicity in patients who are critically ill.

Novel models for the prediction of drug-gene interactions

INTRODUCTION

Adverse drug reactions (ADRs) are among the leading causes of death, and frequently associated with drug-gene interactions (DGIs). In addition to pharmacogenomic programs for implementation of genetic preemptive testing into clinical practice, mathematical modeling can help to understand, quantify and predict the effects of DGIs in vivo. Moreover, modeling can contribute to optimize prospective clinical drug trial activities and to reduce DGI-related ADRs.

AREAS COVERED

Approaches and challenges of mechanistical DGI implementation and model parameterization are discussed for population pharmacokinetic and physiologically based pharmacokinetic models. The broad spectrum of published DGI models and their applications is presented, focusing on the investigation of DGI effects on pharmacology and model-based dose adaptations.

EXPERT OPINION

Mathematical modeling provides an opportunity to investigate complex DGI scenarios and can facilitate the development process of safe and efficient personalized dosing regimens. However, reliable DGI model input data from in vivo and in vitro measurements are crucial. For this, collaboration among pharmacometricians, laboratory scientists and clinicians is important to provide homogeneous datasets and unambiguous model parameters. For a broad adaptation of validated DGI models in clinical practice, interdisciplinary cooperation should be promoted and qualification toolchains must be established.

External evaluation of population pharmacokinetic models of imatinib in adults diagnosed with chronic myeloid leukaemia

AIMS

Imatinib is considered the standard first-line treatment in newly diagnosed patients with chronic-phase myeloid leukaemia (CML). Several imatinib population pharmacokinetic (popPK) models have been developed. However, their predictive performance has not been well established when extrapolated to different populations. Therefore, this study aimed to perform an external evaluation of available imatinib popPK models developed mainly in adult patients, and to evaluate the improvement in individual model-based predictions through Bayesian forecasting computed by each model at different treatment occasions.

METHODS

A literature review was conducted through PubMed and Scopus to identify popPK models. Therapeutic drug monitoring data collected in adult CML patients treated with imatinib was used for external evaluation, including prediction- and simulated-based diagnostics together with Bayesian forecasting analysis.

RESULTS

Fourteen imatinib popPK studies were included for model-performance evaluation. A total of 99 imatinib samples were collected from 48 adult CML patients undergoing imatinib treatment with a minimum of one plasma concentration measured at steady-state between January 2016 and December 2020. The model proposed by Petain et al showed the best performance concerning prediction-based diagnostics in the studied population. Bayesian forecasting demonstrated a significant improvement in predictive performance at the second visit. Inter-occasion variability contributed to reducing bias and improving individual model-based predictions.

CONCLUSIONS

Imatinib popPK studies developed in Caucasian subjects including α1-acid glycoprotein showed the best model performance in terms of overall bias and precision. Moreover, two imatinib samples from different visits appear sufficient to reach an adequate model-based individual prediction performance trough Bayesian forecasting.

Population pharmacokinetic modeling of factor concentrates in hemophilia: an overview and evaluation of best practice

The accuracy of pharmacokinetic (PK)-guided dosing depends on the clinical and laboratory data used to construct a population PK model, as well as the patient’s individual PK profile. This review provides a detailed overview of data used for published population PK models for factor VIII (FVIII) and factor IX (FIX) concentrates, to support physicians in their choices of which model best suits each patient. Furthermore, to enhance detailed data collection and documentation, we do suggestions for best practice. A literature search was performed; publications describing prophylactic population PK models for FVIII and FIX concentrates based on original patient data and constructed using nonlinear mixed-effect modeling were included. The following data were collected: detailed demographics, type of product, assessed and included covariates, laboratory specifications, and validation of models. Included models were scored according to our recommendations for best practice, specifically scoring the quality of data documentation as reported. Respectively, 20 models for FVIII and 7 for FIX concentrates were retrieved. Although most models (22/27) included pediatric patients, only 4 reported detailed demographics. The wide range of body weights suggested that overweight and obese adults were represented. Twenty-six models reported the assay applied to measure factor levels, whereas only 16 models named reagents used. Eight models were internally validated using a data subset. This overview presents detailed information on clinical and laboratory data used for published population PK models. We provide recommendations on data collection and documentation to increase the reliability of PK-guided prophylactic dosing of factor concentrates in hemophilia A and B.

Creation of an inventory of quality markers used to evaluate pharmacokinetic literature: A systematic review

WHAT IS KNOWN AND OBJECTIVE

Robust critical appraisal tools for clinical pharmacokinetic studies are limited. Before development of such a tool is possible, quality markers (items deemed important for credibility of study results) must be identified. We aim to create an inventory of quality markers intended for the appraisal of clinical pharmacokinetic studies and to categorize identified markers into associated domains of study quality.

METHODS

Medline via ProQuest central (1946-Sep 2020, EMBASE (1974-Sep 2020), Cochrane database of systematic reviews, Google and Google Scholar were searched using the following search categories: pharmacokinetics, reporting guidelines and quality markers. Reference lists of the identified articles were searched manually. Any article (review, study or guideline) reporting quality markers related to the appraisal of pharmacokinetic literature was eligible for inclusion. Articles were further screened and limited to those reported in English on human subjects only. Cell-based and animal-based pharmacokinetic studies were excluded. Extracted data from included articles included identified or perceived markers of quality and baseline article data. Identified quality markers were then categorized according to manuscript reporting domains (abstract, introduction/background, methodology, results, discussion and conclusion).

RESULTS AND DISCUSSION

Of 789 studies identified, 17 articles were included for extraction of quality markers. A total of 35 quality markers were identified across eight categories. The most frequently reported quality markers were related to method (13/35) and result sections (6/35). Quality markers encompassed all aspects of study design and reporting and were both similar and different to established reporting checklists for clinical pharmacokinetic studies.

WHAT IS NEW AND CONCLUSION

The inventory of quality markers is now suitable to undergo further testing for inclusion in a tool designed for the appraisal of clinical pharmacokinetic studies.

Population Pharmacokinetics and Dosing Simulation of Vancomycin Administered by Continuous Injection in Critically Ill Patient

Background: Vancomycin is widely used for empirical antimicrobial therapy in critically ill patients with sepsis. Continuous infusion (CI) may provide more stable exposure than intermittent infusion, but optimal dosing remains challenging. The aims of this study were to perform population pharmacokinetic (PK) analysis of vancomycin administered by CI in intensive care unit (ICU) patients to identify optimal dosages. Methods: Patients who received vancomycin by CI with at least one measured concentration in our center over 16 months were included, including those under continuous renal replacement therapy (CRRT). Population PK was conducted and external validation of the final model was performed in a dataset from another center. Simulations were conducted with the final model to identify the optimal loading and maintenance doses for various stages of estimated creatinine clearance (CRCL) and in patients on CRRT. Target exposure was defined as daily AUC of 400-600 mg·h/L on the second day of therapy (AUC24-48 h). Results: A two-compartment model best described the data. Central volume of distribution was allometrically scaled to ideal body weight (IBW), whereas vancomycin clearance was influenced by CRRT and CRCL. Simulations performed with the final model suggested a loading dose of 27.5 mg/kg of IBW. The maintenance dose ranged from 17.5 to 30 mg/kg of IBW, depending on renal function. Overall, simulation showed that 55.8% (95% CI; 47-64%) of patients would achieve the target AUC with suggested dosages. Discussion: A PK model has been validated for vancomycin administered by CI in ICU patients, including patients under CRRT. Our model-informed precision dosing approach may help for early optimization of vancomycin exposure in such patients.

Pharmacokinetics of perioperative FVIII in adult patients with haemophilia A: An external validation and development of an alternative population pharmacokinetic model

INTRODUCTION

Haemophilia A patients require perioperative clotting factor replacement to limit excessive bleeding. Weight-based dosing of Factor VIII (FVIII) does not account for inter-individual pharmacokinetic (PK) variability, and may lead to suboptimal FVIII exposure.

AIM

To perform an external validation of a previously developed population PK (popPK) model of perioperative FVIII in haemophilia A patients.

METHODS

A retrospective chart review identified perioperative haemophilia A patients at the University of North Carolina (UNC) between April 2014 and November 2019. Patient data was used to externally validate a previously published popPK model proposed by Hazendonk. Based on these validation results, a modified popPK model was developed to characterize FVIII PK in our patients. Dosing simulations were performed using this model to compare FVIII target attainment between intermittent bolus (IB) and continuous infusion (CI) administration methods.

RESULTS

A total of 521 FVIII concentrations, drawn from 34 patients, were analysed. Validation analyses revealed that the Hazendonk model did not fully capture FVIII PK in the UNC cohort. Therefore, a modified one-compartment model, with weight and age as covariates on clearance (CL), was developed. Dosing simulations revealed that CI resulted in improved target attainment by 16%, with reduced overall FVIII usage by 58 IU/kg, compared to IB.

CONCLUSION

External validation revealed a previously published popPK model of FVIII did not adequately characterize UNC patients, likely due to differences in patient populations. Future prospective studies are needed to evaluate our model prior to implementation into clinical practice.

To Estimate or to Forecast? Lessons From a Comparative Analysis of Four Bayesian Fitting Methods Based on Nonparametric Models

ABSTRACT

Using pharmacokinetic (PK) models and Bayesian methods in dosing software facilitates the analysis of individual PK data and precision dosing. Several Bayesian methods are available for computing Bayesian posterior distributions using nonparametric population models. The objective of this study was to compare the performance of the maximum a posteriori (MAP) model, multiple model (MM), interacting MM (IMM), and novel hybrid MM(HMM) in estimating past concentrations and predicting future concentrations during therapy. Amikacin and vancomycin PK data were analyzed in older hospitalized patients using 2 strategies. First, the entire data set of each patient was fitted using each of the 4 methods implemented in BestDose software. Then, the 4 methods were used in each therapeutic drug monitoring occasion to estimate the past concentrations available at this time and to predict the subsequent concentrations to be observed on the next occasion. The bias and precision of the model predictions were compared among the methods. A total of 406 amikacin concentrations from 96 patients and 718 vancomycin concentrations from 133 patients were available for analysis. Overall, significant differences were observed in the predictive performance of the 4 Bayesian methods. The IMM method showed the best fit to past concentration data of amikacin and vancomycin, whereas the MM method was the least precise. However, MM best predicted the future concentrations of amikacin. The MAP and HMM methods showed a similar predictive performance and seemed to be more appropriate for the prediction of future vancomycin concentrations than the other models were. The richness of the prior distribution may explain the discrepancies between the results of the 2 drugs. Although further research with other drugs and models is necessary to confirm our findings, these results challenge the widely accepted assumption in PK modeling that a better data fit indicates better forecasting of future observations.

Identification of optimal dosing schedules of dacomitinib and osimertinib for a phase I/II trial in advanced EGFR-mutant non-small cell lung cancer

Despite the clinical success of the third-generation EGFR inhibitor osimertinib as a first-line treatment of EGFR-mutant non-small cell lung cancer (NSCLC), resistance arises due to the acquisition of EGFR second-site mutations and other mechanisms, which necessitates alternative therapies. Dacomitinib, a pan-HER inhibitor, is approved for first-line treatment and results in different acquired EGFR mutations than osimertinib that mediate on-target resistance. A combination of osimertinib and dacomitinib could therefore induce more durable responses by preventing the emergence of resistance. Here we present an integrated computational modeling and experimental approach to identify an optimal dosing schedule for osimertinib and dacomitinib combination therapy. We developed a predictive model that encompasses tumor heterogeneity and inter-subject pharmacokinetic variability to predict tumor evolution under different dosing schedules, parameterized using in vitro dose-response data. This model was validated using cell line data and used to identify an optimal combination dosing schedule. Our schedule was subsequently confirmed tolerable in an ongoing dose-escalation phase I clinical trial (NCT03810807), with some dose modifications, demonstrating that our rational modeling approach can be used to identify appropriate dosing for combination therapy in the clinical setting.

The Role of PK/PD Analysis in the Development and Evaluation of Antimicrobials

Pharmacokinetic/pharmacodynamic (PK/PD) analysis has proved to be very useful to establish rational dosage regimens of antimicrobial agents in human and veterinary medicine. Actually, PK/PD studies are included in the European Medicines Agency (EMA) guidelines for the evaluation of medicinal products. The PK/PD approach implies the use of in vitro, ex vivo, and in vivo models, as well as mathematical models to describe the relationship between the kinetics and the dynamic to determine the optimal dosing regimens of antimicrobials, but also to establish susceptibility breakpoints, and prevention of resistance. The final goal is to optimize therapy in order to maximize efficacy and minimize side effects and emergence of resistance. In this review, we revise the PK/PD principles and the models to investigate the relationship between the PK and the PD of antibiotics. Additionally, we highlight the outstanding role of the PK/PD analysis at different levels, from the development and evaluation of new antibiotics to the optimization of the dosage regimens of currently available drugs, both for human and animal use.

Fixed Dose of Cemiplimab in Patients with Advanced Malignancies Based on Population Pharmacokinetic Analysis

Introduction

This study outlined cemiplimab intravenous (IV) dosing strategy to move from body weight (BW)-based 3 mg/kg every-2-week (Q2W) dosing in first-in-human study (study 1423; NCT02383212) to fixed 350 mg every-3-week (Q3W) dosing, utilizing population pharmacokinetics (PopPK) modeling and simulations, and supported by a limited dataset from a phase 2 study (study 1540; NCT02760498).

Methods

Cemiplimab concentration data from a total of 505 patients were pooled from study 1423 in advanced malignancies and study 1540 in advanced cutaneous squamous cell carcinoma (CSCC). All patients received weight-based cemiplimab dose (1, 3, 10 mg/kg Q2W or 3 mg/kg Q3W) except 4% who received 200 mg Q2W. A linear two-compartment PopPK model incorporating covariates that improved goodness-of-fit statistics was developed to compare cemiplimab exposure at 350 mg Q3W versus 3 mg/kg Q2W. Upon availability, observed cemiplimab concentration at 350 mg Q3W in study 1540 was then compared with the simulated values.

Results

Post hoc estimates of cemiplimab exposure and variability (505 patients; weight range 30.9–156 kg; median 76.1 kg) at steady state were found to be similar at 350 mg Q3W and 3 mg/kg Q2W. Effect of BW on cemiplimab exposure was described by exposure versus BW plots and at extreme BW. Overlay of individual observed cemiplimab concentrations in 51 patients with metastatic CSCC on simulated concentration–time profiles in 2000 patients at 350 mg Q3W confirmed cemiplimab exposure similarity and demonstrated the robustness of dose optimization based on PopPK modeling and simulations.

Conclusions

Cemiplimab 350 mg Q3W is being further investigated in multiple indications.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-021-01638-5.

Validation of a perioperative population factor VIII pharmacokinetic model with a large cohort of pediatric hemophilia a patients

AIMS

Population pharmacokinetic (PK) models are increasingly applied to perform individualized dosing of factor VIII (FVIII) concentrates in haemophilia A patients. To guarantee accurate performance of a population PK model in dose individualization, validation studies are of importance. However, external validation of population PK models requires independent data sets and is, therefore, seldomly performed. Therefore, this study aimed to validate a previously published population PK model for FVIII concentrates administrated perioperatively.

METHODS

A previously published population PK model for FVIII concentrate during surgery was validated using independent data from 87 children with severe haemophilia A with a median (range) age of 2.6 years (0.03-15.2) and body weight of 14 kg (4-57). First, the predictive performance of the previous model was evaluated with MAP Bayesian analysis using NONMEM v7.4. Subsequently, the model parameters were (re)estimated using a combined dataset consisting of the previous modelling data and the data available for the external validation.

RESULTS

The previous model underpredicted the measured FVIII levels with a median of 0.17 IU mL^-1 . Combining the new, independent and original data, a dataset comprising 206 patients with a mean age of 7.8 years (0.03-77.6) and body weight of 30 kg (4-111) was obtained. Population PK modelling provided estimates for CL, V1, V2, and Q: 171 mL h^-1  68 kg^-1 , 2930 mL 68 kg^-1 , 1810 mL 68 kg^-1 , and 172 mL h^-1  68 kg^-1 , respectively. This model adequately described all collected FVIII levels, with a slight median overprediction of 0.02 IU mL^-1 .

CONCLUSIONS

This study emphasizes the importance of external validation of population PK models using real-life data.

Evaluation of the MeroRisk Calculator, A User-Friendly Tool to Predict the Risk of Meropenem Target Non-Attainment in Critically Ill Patients

BACKGROUND

The MeroRisk-calculator, an easy-to-use tool to determine the risk of meropenem target non-attainment after standard dosing (1000 mg; q8h), uses a patient's creatinine clearance and the minimum inhibitory concentration (MIC) of the pathogen. In clinical practice, however, the MIC is rarely available. The objectives were to evaluate the MeroRisk-calculator and to extend risk assessment by including general pathogen sensitivity data.

METHODS

Using a clinical routine dataset (155 patients, 891 samples), a direct data-based evaluation was not feasible. Thus, in step 1, the performance of a pharmacokinetic model was determined for predicting the measured concentrations. In step 2, the PK model was used for a model-based evaluation of the MeroRisk-calculator: risk of target non-attainment was calculated using the PK model and agreement with the MeroRisk-calculator was determined by a visual and statistical (Lin's concordance correlation coefficient (CCC)) analysis for MIC values 0.125-16 mg/L. The MeroRisk-calculator was extended to include risk assessment based on EUCAST-MIC distributions and cumulative-fraction-of-response analysis.

RESULTS

Step 1 showed a negligible bias of the PK model to underpredict concentrations (-0.84 mg/L). Step 2 revealed a high level of agreement between risk of target non-attainment predictions for creatinine clearances >50 mL/min (CCC = 0.990), but considerable deviations for patients <50 mL/min. For 27% of EUCAST-listed pathogens the median cumulative-fraction-of-response for the observed patients receiving standard dosing was < 90%.

CONCLUSIONS

The MeroRisk-calculator was successfully evaluated: For patients with maintained renal function it allows a reliable and user-friendly risk assessment. The integration of pathogen-based risk assessment substantially increases the applicability of the tool.

Machine learning in pharmacometrics: Opportunities and challenges

The explosive growth in medical devices, imaging and diagnostics, computing, and communication and information technologies in drug development and healthcare has created an ever-expanding data landscape that the pharmacometrics (PMX) research community must now traverse. The tools of machine learning (ML) have emerged as a powerful computational approach in other data-rich disciplines but its effective utilization in the pharmaceutical sciences and PMX modelling is in its infancy. ML-based methods can complement PMX modelling by enabling the information in diverse sources of big data, e.g. population-based public databases and disease-specific clinical registries, to be harnessed because they are capable of efficiently identifying salient variables associated with outcomes and delineating their interdependencies. ML algorithms are computationally efficient, have strong predictive capabilities and can enable learning in the big data setting. ML algorithms can be viewed as providing a computational bridge from big data to complement PMX modelling. This review provides an overview of the strengths and weaknesses of ML approaches vis-à-vis population methods, assesses current research into ML applications in the pharmaceutical sciences and provides perspective for potential opportunities and strategies for the successful integration and utilization of ML in PMX.

Predictive performance of parent-metabolite population pharmacokinetic models of (S)-ketamine in healthy volunteers

Purpose

The recent repurposing of ketamine as treatment for pain and depression has increased the need for accurate population pharmacokinetic (PK) models to inform the design of new clinical trials. Therefore, the objectives of this study were to externally validate available PK models on (S)-(nor)ketamine concentrations with in-house data and to improve the best performing model when necessary.

Methods

Based on predefined criteria, five models were selected from literature. Data of two previously performed clinical trials on (S)-ketamine administration in healthy volunteers were available for validation. The predictive performances of the selected models were compared through visual predictive checks (VPCs) and calculation of the (root) mean (square) prediction errors (ME and RMSE). The available data was used to adapt the best performing model through alterations to the model structure and re-estimation of inter-individual variability (IIV).

Results

The model developed by Fanta et al. (Eur J Clin Pharmacol 71:441–447, 2015) performed best at predicting the (S)-ketamine concentration over time, but failed to capture the (S)-norketamine C_max correctly. Other models with similar population demographics and study designs had estimated relatively small distribution volumes of (S)-ketamine and thus overpredicted concentrations after start of infusion, most likely due to the influence of circulatory dynamics and sampling methodology. Model predictions were improved through a reduction in complexity of the (S)-(nor)ketamine model and re-estimation of IIV.

Conclusion

The modified model resulted in accurate predictions of both (S)-ketamine and (S)-norketamine and thereby provides a solid foundation for future simulation studies of (S)-(nor)ketamine PK in healthy volunteers after (S)-ketamine infusion.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00228-021-03104-1.

Population pharmacokinetic modeling and clinical application of vancomycin in Chinese patients hospitalized in intensive care units

Management of vancomycin administration for intensive care units (ICU) patients remains a challenge. The aim of this study was to describe a population pharmacokinetic model of vancomycin for optimizing the dose regimen for ICU patients. We prospectively enrolled 466 vancomycin-treated patients hospitalized in the ICU, collected trough or approach peak blood samples of vancomycin and recorded corresponding clinical information from July 2015 to December 2017 at Tai Zhou Hospital of Zhejiang Province. The pharmacokinetics of vancomycin was analyzed by nonlinear mixed effects modeling with Kinetica software. Internal and external validation was evaluated by the maximum likelihood method. Then, the individual dosing regimens of the 92 patients hospitalized in the ICU whose steady state trough concentrations exceeded the target range (10–20 μg/ml) were adjusted by the Bayes feedback method. The final population pharmacokinetic model show that clearance rate (CL) of vancomycin will be raised under the conditions of dopamine combined treatment, severe burn status (Burn-S) and increased total body weight (TBW), but reduced under the conditions of increased serum creatinine (Cr) and continuous renal replacement therapy status; Meanwhile, the apparent distribution volume (V) of vancomycin will be enhanced under the terms of increased TBW, however decreased under the terms of increased age and Cr. The population pharmacokinetic parameters (CL and V) according to the final model were 3.16 (95%CI 2.83, 3.40) L/h and 60.71 (95%CI 53.15, 67.46). The mean absolute prediction error for external validation by the final model was 12.61% (95CI 8.77%, 16.45%). Finally, the prediction accuracy of 90.21% of the patients’ detected trough concentrations that were distributed in the target range of 10–20 μg/ml after dosing adjustment was found to be adequate. There is significant heterogeneity in the CL and V of vancomycin in ICU patients. The constructed model is sufficiently precise for the Bayesian dose prediction of vancomycin concentrations for the population of ICU Chinese patients.

Real-World Infliximab Pharmacokinetic Study Informs an Electronic Health Record-Embedded Dashboard to Guide Precision Dosing in Children with Crohn's Disease

Standard-of-care infliximab dosing regimens were developed prior to the routine use of therapeutic drug monitoring and identification of target concentrations. Not surprisingly, subtherapeutic infliximab concentrations in pediatric Crohn's disease (CD) are common. The primary aim was to conduct a real-world pharmacokinetic (PK) evaluation to discover blood biomarkers of rapid clearance, identify exposure targets, and a secondary aim to translate PK modeling to the clinic. In a multicenter observational study, 671 peak and trough infliximab concentrations from 78 patients with CD were analyzed with a drug-tolerant assay (Esoterix; LabCorp, Calabasas, CA). Individual area under the curve (AUC) estimates were generated as a measure of drug exposure over time. Population PK modeling (nonlinear mixed-effect modeling) identified serum albumin, antibody to infliximab, erythrocyte sedimentation rate (ESR), and neutrophil CD64 as biomarkers for drug clearance. Week 14 and week 52 biochemical remitters (fecal calprotectin < 250 µg/g) had higher infliximab exposure (AUC) throughout induction. The optimal infliximab AUC target during induction for week 14 biochemical remission was 79,348 µg*h/mL (area under the receiver operating characteristic curve (AUROC) 0.77, [0.63-0.90], 85.7% sensitive, and 64.3% specific) with those exceeding the AUC target more likely to achieve a surgery-free week 52 biochemical remission (OR 4.3, [1.2-14.6]). Pretreatment predictors for subtherapeutic week 14 AUC included neutrophil CD64 > 6 (OR 4.5, [1.4-17.8]), ESR > 30 mm/h (OR 3.8, [1.4-11]), age < 10 years old (OR 4.2, [1.2-20]), and weight < 30 kg (OR 6.6, [2.1-25]). We created a decision-support PK dashboard with an iterative process and embedded the modeling program within the electronic health record. Model-informed precision dosing guided by real-world PKs is now available at the bedside in real-time.

Concept and utility of population pharmacokinetic and pharmacokinetic/pharmacodynamic models in drug development and clinical practice

Due to frequent clinical trial failures and consequently fewer new drug approvals, the need for improvement in drug development has, to a certain extent, been met using model-based drug development. Pharmacometrics is a part of pharmacology that quantifies drug behaviour, treatment response and disease progression based on different models (pharmacokinetic - PK, pharmacodynamic - PD, PK/PD models, etc.) and simulations. Regulatory bodies (European Medicines Agency, Food and Drug Administration) encourage the use of modelling and simulations to facilitate decision-making throughout all drug development phases. Moreover, the identification of factors that contribute to variability provides a basis for dose individualisation in routine clinical practice. This review summarises current knowledge regarding the application of pharmacometrics in drug development and clinical practice with emphasis on the population modelling approach.

Dosing Recommendations for Lamotrigine in Children: Evaluation Based on Previous and New Population Pharmacokinetic Models

Lamotrigine is a broad-spectrum antiepileptic drug with high interindividual variability in serum concentrations in children. The aims of this study were to evaluate the predictive performance of pediatric population pharmacokinetic (PPK) models published on lamotrigine, to build a new model with our monitoring data and to evaluate the current recommended doses. A validation cohort included patients treated with lamotrigine who had a serum level assayed during therapeutic drug monitoring (TDM). PPK models published in the literature were first applied to the validation cohort. We assessed their predictive performance using mean prediction errors, root mean squared errors, and visual predictive checks. A new model was then built using the data. Dose simulations were performed to evaluate the doses recommended. We included 270 lamotrigine concentrations ranging from 0.5 to 17.9 mg/L from 175 patients. The median (range) age and weight were 11.8 years (0.8-18 years) and 32.7 kg (8-110 kg). We tested 6 PPK models; most had acceptable bias and precision but underestimated the variability of the cohort. We built a 1-compartment model with first-order absorption and elimination, allometric scaling, and effects of inhibitor and inducer comedications. In our cohort, 22.6% of trough concentrations were below 2.5 mg/L. In conclusion, we proposed a PPK model that can be used for TDM of lamotrigine in children. In our population, a high percentage of children had low trough concentrations of lamotrigine. As the intervals of recommended doses are large, we suggest aiming at the higher range of doses to reach the target concentration.

Comparisons of Four Protein-Binding Models Characterizing the Pharmacokinetics of Unbound Phenytoin in Adult Patients Using Non-Linear Mixed-Effects Modeling

BACKGROUND AND OBJECTIVE

Phenytoin is extensively protein bound with a narrow therapeutic range. The unbound phenytoin is pharmacologically active, but total concentrations are routinely measured in clinical practice. The relationship between free and total phenytoin has been described by various binding models with inconsistent findings. Systematic comparison of these binding models in a single experimental setting is warranted to determine the optimal binding behaviors.

METHODS

Non-linear mixed-effects modeling was conducted on retrospectively collected data (n = 37 adults receiving oral or intravenous phenytoin) using a stochastic approximation expectation-maximization algorithm in MonolixSuite-2019R2. The optimal base structural model was initially developed and utilized to compare four binding models: Winter-Tozer, linear binding, non-linear single-binding site, and non-linear multiple-binding site. Each binding model was subjected to error and covariate modeling. The final model was evaluated using relative standard errors (RSEs), goodness-of-fit plots, visual predictive check, and bootstrapping.

RESULTS

A one-compartment, first-order absorption, Michaelis-Menten elimination, and linear protein-binding model best described the population pharmacokinetics of free phenytoin at typical clinical concentrations. The non-linear single-binding-site model also adequately described phenytoin binding but generated larger RSEs. The non-linear multiple-binding-site model performed the worst, with no identified covariates. The optimal linear binding model suggested a relatively high binding capacity using a single albumin site. Covariate modeling indicated a positive relationship between albumin concentration and the binding proportionality constant.

CONCLUSIONS

The linear binding model best described the population pharmacokinetics of unbound phenytoin in adult subjects and may be used to improve the prediction of free phenytoin concentrations.

Genetic polymorphisms of ABCB1 (P-glycoprotein) as a covariate influencing daptomycin pharmacokinetics: a population analysis in patients with bone and joint infection

BACKGROUND

Daptomycin has been recognized as a therapeutic option for the treatment of bone and joint infection (BJI). Gene polymorphism of ABCB1, the gene encoding P-glycoprotein (P-gp), may influence daptomycin pharmacokinetics (PK).

OBJECTIVES

We aimed to examine population PK of daptomycin and its determinants, including genetic factors, in patients with BJI.

PATIENTS AND METHODS

We analysed data from patients who received daptomycin for BJI between 2012 and 2016 in our regional reference centre and who had measured daptomycin concentrations and P-gp genotyping. A population approach was used to analyse PK data. In covariate analysis, we examined the influence of three single nucleotide variations (SNVs) of ABCB1 (3435C > T, 2677G > T/A and 1236C > T) and that of the corresponding haplotype on daptomycin PK parameters. Simulations performed with the final model examined the influence of covariates on the probability to achieve pharmacodynamic (PD) targets.

RESULTS

Data from 81 patients were analysed. Daptomycin body CL (CLDAP) correlated with CLCR and was 23% greater in males than in females. Daptomycin central V (V1) was allometrically scaled to body weight and was 25% lower in patients with homozygous CGC ABCB1 haplotype than in patients with any other genotype. Simulations performed with the model showed that sex and P-gp haplotype may influence the PTA for high MIC values and that a dosage of 10 mg/kg/24 h would optimize efficacy.

CONCLUSIONS

Daptomycin dosages higher than currently recommended should be evaluated in patients with BJI. Gender and P-gp gene polymorphism should be further examined as determinants of dosage requirements.

Pharmacokinetics and population pharmacokinetics in pediatric oncology

Pharmacokinetic research has become increasingly important in pediatric oncology as it can have direct clinical implications and is a crucial component in individualized medicine. Population pharmacokinetics has become a popular method especially in children, due to the potential for sparse sampling, flexible sampling times, computing of heterogeneous data, and identification of variability sources. However, population pharmacokinetic reports can be complex and difficult to interpret. The aim of this article is to provide a basic explanation of population pharmacokinetics, using clinical examples from the field of pediatric oncology, to facilitate the translation of pharmacokinetic research into the daily clinic.

Population Pharmacokinetics of Clozapine: A Systematic Review

BACKGROUND AND OBJECTIVE

Clozapine is a second-generation antipsychotic drug that is considered the most effective treatment for refractory schizophrenia. Several clozapine population pharmacokinetic models have been introduced in the last decades. Thus, a systematic review was performed (i) to compare published pharmacokinetics models and (ii) to summarize and explore identified covariates influencing the clozapine pharmacokinetics models.

METHODS

A search of publications for population pharmacokinetic analyses of clozapine either in healthy volunteers or patients from inception to April 2019 was conducted in PubMed and SCOPUS databases. Reviews, methodology articles, in vitro and animal studies, and noncompartmental analysis were excluded.

RESULTS

Twelve studies were included in this review. Clozapine pharmacokinetics was described as one-compartment with first-order absorption and elimination in most of the studies. Significant interindividual variations of clozapine pharmacokinetic parameters were found in most of the included studies. Age, sex, smoking status, and cytochrome P450 1A2 were found to be the most common identified covariates affecting these parameters. External validation was only performed in one study to determine the predictive performance of the models.

CONCLUSIONS

Large pharmacokinetic variability remains despite the inclusion of several covariates. This can be improved by including other potential factors such as genetic polymorphisms, metabolic factors, and significant drug-drug interactions in a well-designed population pharmacokinetic model in the future, taking into account the incorporation of larger sample size and more stringent sampling strategy. External validation should also be performed to the previously published models to compare their predictive performances.

Population Pharmacokinetic Analysis of Immediate-Release Oral Tacrolimus Co-administered with Mycophenolate Mofetil in Corticosteroid-Free Adult Kidney Transplant Recipients

BACKGROUND AND OBJECTIVE

Tacrolimus is the mainstay calcineurin inhibitor frequently administered with mycophenolic acid with or without corticosteroids to prevent graft rejection in adult kidney transplant recipients. The primary objective of this study was to develop and evaluate a population pharmacokinetic model characterizing immediate-release oral tacrolimus co-administered with mycophenolate mofetil (a pro-drug of mycophenolic acid) in adult kidney transplant recipients on corticosteroid-free regimens. The secondary objective was to investigate the effects of clinical covariates on the pharmacokinetics of tacrolimus, emphasizing the interacting effects of mycophenolic acid.

METHODS

Population modeling and evaluation were conducted with Monolix (Suite-2018R1) using the stochastic approximation expectation-maximization algorithm in 49 adult subjects (a total of 320 tacrolimus whole-blood concentrations). Effects of clinical variables on tacrolimus pharmacokinetics were determined by population covariate modeling, regression modeling, and categorical analyses.

RESULTS

A two-compartment, first-order absorption with a lag-time, linear elimination, and constant error model best represented the population pharmacokinetics of tacrolimus. The apparent clearance value for tacrolimus was 17.9 l/h (6.95% relative standard error) in our model, which is lower compared with similar subjects on corticosteroid-based therapy. The glomerular filtration rate had significant effects on the apparent clearance and central compartment volume of distribution. Conversely, mycophenolic acid did not affect the apparent clearance of tacrolimus.

CONCLUSION

We have developed and internally evaluated a novel population pharmacokinetic model for tacrolimus co-administered with mycophenolate mofetil in corticosteroid-free adult kidney transplant patients. These findings are clinically important and provide further reasons for conducting therapeutic drug monitoring in this specific population.

Population Pharmacokinetics of Mycophenolic Acid: An Update

The most recent comprehensive reviews on the population pharmacokinetics of mycophenolic acid (MPA) were published in 2014. Since then, several population pharmacokinetic studies on MPA have been published. The majority of literature is still focused on the kidney transplant population, although studies have also been conducted in liver and lung transplantation, autoimmune diseases, and hematopoietic stem cell transplant. While the majority of the model building is still based on parametric non-linear mixed-effects modeling, recent studies suggest the suitability of other methodologies. Additionally, instead of just focusing on pharmacokinetic modeling, a trend toward describing the relationships between pharmacokinetic and pharmacodynamic parameters is observed. Given the importance of enterohepatic recirculation (EHR) in the pharmacokinetics of MPA, more authors have attempted to characterize this process in their models. Overall, the recent models have become more sophisticated and incorporate EHR, pharmacodynamic relationships, and metabolites while maintaining many of the population values and covariates identified previously. However, the number of MPA population pharmacokinetic models describing the enteric-coated formulation of MPA (EC-MPA) is still limited. Given the increasing use of EC-MPA, more studies are needed to fill this literature gap. In addition, few studies are yet available characterizing free MPA concentration or MPA metabolites. Given the extensive protein binding, low to intermediate extraction, and intrinsic clearance characteristics of MPA in humans, including these variables would improve the population structural models.