Model Evaluation of Continuous Data Pharmacometric Models: Metrics and Graphics

Evaluating the evidence for genotype-informed Bayesian dosing of tacrolimus in children undergoing solid organ transplantation: A systematic literature review

Tacrolimus, a calcineurin inhibitor, is a highly effective immunosuppressant used in solid organ transplantation (SOT). However, it is characterized by a narrow therapeutic range and high inter-patient variability in pharmacokinetics. Standard weight-based dosing followed by empiric dose titration is suboptimal in controlling drug concentrations, increasing risk of rejection or toxicity, particularly in the initial months post transplantation. This review explores the potential of combined pre-transplant genotyping and pharmacokinetic (PK) modelling to improve tacrolimus dosing in paediatric SOT recipients. A systematic search of Medline, Embase and Cochrane databases identified studies published between March 2013 and March 2023 that investigated genotype- and PK model-informed tacrolimus dosing in children post-SOT. The Newcastle-Ottawa Scale assessed study quality. Seven studies encompassing paediatric kidney, heart, liver and lung transplants reported using genotype and model-informed dosing. A combination of clinical and genetic factors significantly impacts tacrolimus clearance and thus initial dose recommendation. Body size, transplant organ and co-medications were consistently important, while either time post-transplant or haematocrit emerged in some studies. Several models were identified, however, with limitations evident in some and with absence of evidence for their effectiveness in optimizing initial and subsequent dosing. This review highlights the development of PK models in paediatric SOT that integrate genotype and clinical covariates to personalize early tacrolimus dosing. While promising, prospective studies are needed to validate and confirm their effectiveness in improving time to therapeutic concentrations and reducing under- or overexposure. This approach has the potential to optimize tacrolimus therapy in paediatric SOT, thereby improving outcomes.

Vancomycin in Pediatric Patients with Cystic Fibrosis: Dose Optimization Using Population Pharmacokinetic Approach

BACKGROUND

An increase in Staphylococcus aureus infections has been reported in pediatric patients with cystic fibrosis (CF) over the last few years. This pathogen is commonly treated with vancomycin, an antibiotic for which therapeutic drug monitoring (TDM) is recommended. Updated guidelines were recently published regarding new targets of exposure for the TDM of vancomycin through a Bayesian approach, using population pharmacokinetic (popPK) models.

OBJECTIVES

This study aims to assess the predictive performance of vancomycin popPK models in pediatric patients with CF and to recommend optimal initial dosing regimens based on simulations.

METHODS

Patient data were collected from two centers in Canada, and a literature review was conducted to identify all published vancomycin popPK models for pediatric CF patients. External evaluation and simulations were performed according to patient and occasion of treatment.

RESULTS

A total of 53 vancomycin concentrations were collected from six pediatric CF patients. Only two popPK models of vancomycin for pediatric CF patients were identified through the literature review. The external evaluation results for both centers combined revealed a population bias of 28.1% and an imprecision of 33.7%. A re-estimation of parameters was performed to improve predictive performance. The optimal initial dosing regimen was 15 mg/kg/dose administered every 6 hours according to the per occasion remodel.

CONCLUSION

The predictive performance and identified optimal initial dosing regimens associated with the model were different depending on the data used, showing external evaluation's importance before implementing a model in clinical practice.

A small step toward precision dosing of caffeine in preterm infants: An external evaluation of published population pharmacokinetic models

BACKGROUND

Several population pharmacokinetic (PopPK) models of caffeine in preterm infants have been published, but the extrapolation of these models to facilitate model-informed precision dosing (MIPD) in clinical practice is uncertain. This study aimed to comprehensively evaluate their predictive performance using an external, independent dataset.

METHODS

Data used for external evaluation were based on an independent cohort of preterm infants. Currently available PopPK models for caffeine in preterm infants were identified and re-established. Prediction- and simulation-based diagnostics were used to assess model predictability. The influence of prior information was assessed using Bayesian forecasting.

RESULTS

120 plasma samples from 76 preterm infants were included in the evaluation dataset. Twelve PopPK models of caffeine in preterm infants were re-established based on our previously published study. Although two models showed superior predictive performance, none of the 12 PopPK models met all the clinical acceptance criteria of these external evaluation items. Besides, the external predictive performances of most models were unsatisfactory in prediction- and simulation-based diagnostics. Nevertheless, the application of Bayesian forecasting significantly improved the predictive performance, even with only one prior observation.

CONCLUSIONS

Two models that included the most covariates had the best predictive performance across all external assessments. Inclusion of different covariates, heterogeneity of preterm infant characteristics, and different study designs influenced predictive performance. Thorough evaluation is needed before these PopPK models can be implemented in clinical practice. The implementation of MIPD for caffeine in preterm infants could benefit from the combination of PopPK models and Bayesian forecasting as a helpful tool.

Model-Informed Approach to Recommend Burosumab Dosing Regimens for Pediatric and Adult Patients With the Ultrarare Disease Tumor-Induced Osteomalacia

Burosumab is approved for the treatment of hypophosphatemia in persistent tumor-induced osteomalacia. This work exemplifies a model-informed drug development approach that evaluated burosumab pharmacokinetics and pharmacokinetic/pharmacodynamics in the ultrarare tumor-induced osteomalacia population to support adult and pediatric dosing. Data from tumor-induced osteomalacia participants were combined with data from X-linked hypophosphatemia to understand pharmacokinetic and pharmacokinetic/pharmacodynamic characteristics and covariates specific to tumor-induced osteomalacia. Pharmacokinetic and pharmacokinetic/pharmacodynamic simulations were performed using final models to support dosing recommendations for adults and extrapolation to pediatric patients. Burosumab pharmacokinetics were described using a one-compartment model with first-order absorption and body weight as a significant covariate. Pharmacokinetic/pharmacodynamic relationships were described using a sigmoidal Emax model with significant covariates of baseline fibroblast growth factor 23 on baseline fasting serum phosphate and potency of burosumab response and tumor-induced osteomalacia disease state resulting in a steep slope of response; however, the covariates are not clinically meaningful. Simulations demonstrated that, in pediatric patients, starting doses of burosumab 0.3 and 0.4 mg/kg every 2 weeks at steady state would achieve normal serum phosphate levels in ≥ 30% of patients with relatively low risk of hyperphosphatemia (< 3%). In adults, burosumab 0.3 and 0.5 mg/kg every 4 weeks achieves similar percentages of responders and a relative low risk of hyperphosphatemia (< 7%). Serum phosphate titration-based burosumab dosing increased the probability of achieving normal serum phosphate levels. The models supported a model-informed drug development approach for global approvals of titration-based burosumab dosing, guided by monitoring fasting serum phosphate levels.

Population Pharmacokinetic Quantification of CYP2D6 Activity in Codeine Metabolism in Ambulatory Surgical Patients for Model-Informed Precision Dosing

BACKGROUND AND OBJECTIVE

Codeine metabolism in humans is complex due to the involvement of multiple cytochrome P450 (CYP) enzymes, and has a strong genetic underpinning, which determines the levels of relevant CYP450 enzyme expression in vivo. Polymorphic CYP2D6 metabolises codeine to morphine via O-demethylation, while a strong correlation between CYP2D6 phenotype and opioidergic adverse effects of codeine is well documented. The aim of this study was to quantify the effect of CYP2D6 genotype on the biotransformation of codeine.

METHODS

We conducted a prospective clinical trial with 1000 patients, during which ambulatory patients were administered 60 mg of codeine preoperatively and the association between CYP2D6 activity and morphine exposure across various CYP2D6 genotypes was quantified using a population pharmacokinetic model. Plasma concentration data for codeine and its primary metabolites were obtained from 997 patients and CYP2D6 genotype was screened for study subjects, and respective sums of activity scores assigned for each CYP2D6 allele were used as covariates in model development.

RESULTS

Our final model predicts the disposition of codeine and the formation of morphine, codeine-6-glucuronide and morphine-3-glucuronide adequately while accounting for variability in morphine exposure on the basis of CYP2D6 genotype. In agreement with previous results, patients with decreased function alleles (CYP2D6*10 and *41) showed varying levels of decrease in CYP2D6 activity that were inconsistent with increasing activity scores. Model simulations demonstrate that morphine concentrations in ultrarapid CYP2D6 metabolisers reach systemic concentrations that can potentially cause respiratory depression (over 9.1 ng/mL), and have 218% higher exposure (19 versus 8.7 µg · h/L, p < 0.001) to morphine than normal metabolisers. Similarly, poor and intermediate metabolisers had significantly reduced morphine exposure (1.0 and 3.7 versus 8.7 µg · h/L, p < 0.001) as compared with normal metabolisers.

CONCLUSIONS

Our final model leads the way in implementing model-informed precision dosing in codeine therapy and identifies the use of genetic testing as an integral component in the effort to implement rational pharmacotherapy with codeine.

Target-site cefiderocol pharmacokinetics in soft tissues of healthy volunteers

BACKGROUND

Cefiderocol may potentially be used to treat skin and soft tissue infections (SSTIs). However, the pharmacokinetics of cefiderocol in human soft tissues have not yet been determined. The objective of the present PK study was to investigate whether target-site concentrations of cefiderocol are sufficiently high for the treatment of SSTIs.

METHODS

In this pharmacokinetic study, a single intravenous dose of 2 g cefiderocol was administered to eight healthy male volunteers. Drug concentrations were determined in plasma, muscle and subcutis over 8 h. Free plasma concentrations were calculated using the plasma protein binding determined with ultrafiltration. Free tissue concentrations were obtained using microdialysis. Penetration ratios were calculated as AUC0-8h_free_tissue/AUC0-8h_free_plasma. A population pharmacokinetic model was developed, and the probability of target attainment (PTA) was determined using Monte Carlo simulations.

RESULTS

Cefiderocol showed good tissue penetration, with mean penetration ratios ± standard deviation of 0.99 ± 0.33 and 0.92 ± 0.30 for subcutis and muscle, respectively. Cefiderocol pharmacokinetics in plasma were best described with a two-compartment model, and tissue concentrations were described by scaling the tissue concentrations to concentrations in the peripheral compartment of the plasma model. For a thrice-daily regimen with 2 g doses intravenously infused over 3 h, PTA was ≥90% for MIC values up to 4 mg/L, both based on free plasma and soft tissue pharmacokinetics.

CONCLUSIONS

This study indicates that a dose of 2 g cefiderocol achieves concentrations in plasma considered sufficient for treating relevant bacterial species. Assuming a comparable PK/PD target for soft tissues, sufficiently high concentrations would also be achieved in soft tissues.

Population Pharmacokinetic Model of Intravenous Immunoglobulin in Patients Treated for Various Immune System Disorders

PURPOSE

Intravenous immunoglobulin (IVIG) is used to treat various immune system disorders, but the factors influencing its disposition are not well understood. This study aimed to estimate the population pharmacokinetic parameters of IVIG and to investigate the effect of genetic polymorphism of the FCGRT gene encoding the neonatal Fc receptor (FcRn) and clinical variability on the pharmacokinetic properties of IVIG in patients with immune system disorders.

METHODS

Patients were recruited from 4 hospitals in Malaysia. Clinical data were recorded, and blood samples were taken for pharmacokinetic and genetic studies. Population pharmacokinetic parameters were estimated by nonlinear mixed-effects modeling in Monolix. Age, weight, baseline immunoglobulin G concentration, ethnicity, sex, genotype, disease type, and comorbidity were investigated as potential covariates. Models were evaluated using the difference in objective function value, goodness-of-fit plots, visual predictive checks, and bootstrap analysis.

FINDINGS

A total of 292 blood samples were analyzed from 79 patients. The IVIG concentrations were best described by a 2-compartment model with linear elimination. Weight was found to be an important covariate for volume of distribution in the central compartment (Vc), volume of distribution in the peripheral compartment (Vp), and clearance in the central compartment, whereas disease type was found to be an important covariate for Vp. Goodness-of-fit plots indicated that the model fit the data adequately. Genetic polymorphism of the FCGRT gene encoding the neonatal Fc receptor did not affect the pharmacokinetic properties of IVIG.

IMPLICATIONS

This study supports the use of dosage based on weight as per current practice. The study findings highlight that Vp is significantly influenced by the type of disease being treated with IVIG. This relationship suggests that different disease types, particularly inflammatory and autoimmune conditions, may alter tissue permeability and fluid distribution due to varying degrees of inflammation. Increased inflammation can lead to enhanced permeability and retention of IVIG in peripheral tissues, reflecting higher Vp values.

Population pharmacokinetics of total and unbound valemetostat and platelet dynamics in healthy volunteers and patients with non-Hodgkin lymphoma

Valemetostat is an EZH2/1 inhibitor that has been approved in Japan for the treatment of patients with relapsed/refractory adult T-cell leukemia/lymphoma, based mainly on results from a single-arm phase II trial. It is currently under investigation worldwide for the treatment of other non-Hodgkin lymphomas (NHLs), including peripheral T-cell lymphoma, and for solid tumors. Semi-mechanistic population pharmacokinetic modeling of total and unbound valemetostat and an analysis of the platelet time course during treatment with valemetostat were conducted using data from five clinical trials (two in patients with NHL and three in healthy volunteers). Pharmacokinetic data, including 3162 total/1871 unbound valemetostat observations from 102 patients and 72 healthy volunteers, were described by a three-compartment model with sequential zero-/first-order absorption and saturable binding in the central compartment. Alpha-1-acid glycoprotein (AAG) was the most influential covariate for total valemetostat exposure, yet had little impact on unbound exposure, meaning no dose adjustment was warranted based on AAG levels. The longitudinal platelet data from 101 patients (2313 observations) were adequately described by a modified Friberg model with two proliferation compartments, which characterized unique spontaneous recovery of platelet counts without dose modifications. A model-based simulation quantitatively assessed the proposed dose-adjustment guidance in case of platelet count decreased by comparing the probability of treatment discontinuation due to platelet count decreased with or without the dose adjustment. In summary, the models described observed total and unbound valemetostat concentrations and a unique time course of platelets during treatment, which can justify the clinical dose and provide dose-adjustment guidance.

Absorption pharmacokinetics and feasibility of intranasal dexmedetomidine in patients under general anaesthesia

BACKGROUND

The use of intranasal dexmedetomidine is hampered by a limited understanding of its absorption pharmacokinetics.

METHODS

We examined the pharmacokinetics and feasibility of intranasal dexmedetomidine administered in the supine position to adult patients undergoing general anaesthesia. Twenty-eight patients between 35 and 80 years of age, ASA 1-3 and weight between 50 and 100 kg, who underwent elective unilateral total hip or knee arthroplasty under general anaesthesia were recruited. All patients received 100 μg of intranasal dexmedetomidine after anaesthesia induction. Six venous blood samples (at 0, 5, 15, 45, 60, 240 min timepoints from dexmedetomidine administration) were collected from each patient and dexmedetomidine plasma concentrations were measured. Concentration-time profiles after nasal administration were pooled with earlier data from a population analysis of intravenous dexmedetomidine (n = 202) in order to estimate absorption parameters using nonlinear mixed effects. Peak concentration (CMAX) and time (TMAX) were estimated using simulation (n = 1000) with parameter estimates and their associated variability.

RESULTS

There were 28 adult patients with a mean (SD) age of 66 (8) years and weight of 83 (10) kg. The mean weight-adjusted dose of dexmedetomidine was 1.22 (0.15) μg kg-1. CMAX 0.273 μg L-1 was achieved at 98 min after intranasal administration (TMAX). The relative bioavailability of dexmedetomidine was 80% (95% CI 75-91%). The absorption half-time (TABS = 120 min; 95% CI 90-147 min) was slower than that in previous pharmacokinetic studies on adult patients. Perioperative haemodynamics of all patients remained stable.

CONCLUSIONS

Administration of intranasal dexmedetomidine in the supine position during general anaesthesia is feasible with good bioavailability. This administration method has slower absorption when compared to awake patients in upright position, with consequent concentrations attained after TMAX for several hours.

Modelling antibody dynamics in humans after different Ad26.COV2.S vaccination schemes

AIMS

To develop a semimechanistic model that describes the kinetic profile and variability of antibody (Ab) concentrations following vaccination with Ad26.COV2.S at different doses and dosing intervals.

METHODS

Data were collected from participants randomized into 5 clinical trials receiving the Ad26.COV2.S vaccine. The model considered key elements of humoral immune response, dose proportionality and the evolutionary processes of the immune response. Interindividual variability and covariates were explored.

RESULTS

Fast and slow kinetic phases of Ab and their evolution over time were differentiated. After first and second administrations, Ab concentrations of both phases increased less than dose proportionally, indicating a saturation of B-cell production processes. Ab concentrations produced during the fast kinetic phase increased significantly after the second administration, indicating an underlying evolutive process after antigen exposures. For the slow kinetic phase, a less pronounced increase occurred after the second and third administrations but was relatively higher in subjects who had low concentrations after the first administration. Ab concentrations of the slow phase were higher in females and decreased with age. After multiple administrations, the fast phase had Ab maximum concentrations about 5 times higher than the slow phase. The limiting kinetic factors in the fast and slow phases were the elimination rates of Ab itself and Ab producing cells, respectively.

CONCLUSION

The model appears suitable to quantitatively describe the inter- and intraindividual kinetics of the immune response and the impact of covariates after multiple administrations of a vaccine.

Model-based translation of results from in vitro to in vivo experiments for afabicin activity against Staphylococcus aureus

BACKGROUND

Translation of experimental data on antibiotic activity typically relies on pharmacokinetic/pharmacodynamic (PK/PD) indices. Model-based approaches, considering the full antibiotic killing time course, could be an alternative.

OBJECTIVES

To develop a mechanism-based modelling framework to assess the in vitro and in vivo activity of the FabI inhibitor antibiotic afabicin, and explore the ability of a model built on in vitro data to predict in vivo outcome.

METHODS

A PK/PD model was built to describe bacterial counts from 162 static in vitro time-kill curves evaluating the effect of afabicin desphosphono, the active moiety of the prodrug afabicin, against 21 Staphylococcus aureus strains. Combined with a mouse PK model, outcomes of afabicin doses of 0.011-190 mg/kg q6h against nine S. aureus strains in a murine thigh infection model were predicted, and thereafter refined by estimating PD parameters.

RESULTS

A sigmoid Emax model, with EC50 scaled by the MIC described the afabicin desphosphono killing in vitro. This model predicted, without parameter re-estimation, the in vivo bacterial counts at 24 h within a ±1 log margin for most dosing groups. When parameters were allowed to be estimated, EC50 was 38%-45% lower in vivo, compared with in vitro, within the studied MIC range.

CONCLUSIONS

The developed PK/PD model described the time course of afabicin activity across experimental conditions and bacterial strains. This model showed translational capacity as parameters estimated on in vitro time-kill data could well predict the in vivo outcome for a wide variety of doses in a mouse thigh infection model.

Integrated Population Pharmacokinetics of Daprodustat in Patients with Chronic Kidney Disease with Anemia

BACKGROUND AND OBJECTIVE

Daprodustat is a first-in-class hypoxia-inducible factor prolyl hydroxylase inhibitor (HIF-PHI) approved in the USA for treatment of anemia owing to chronic kidney disease (CKD) in dialysis-dependent adults and in Japan for treatment of CKD in dialysis- and non-dialysis dependent adults. This analysis characterized the population pharmacokinetics (PopPK) of daprodustat in adults with CKD and evaluated the influence of intrinsic and extrinsic factors.

METHODS

This PopPK analysis included data from one phase 2B and four phase 3 studies comprising 707 CKD subjects dose titrated to prespecified target hemoglobin levels with daprodustat doses ranging from 1 to 24 mg once daily and 2 to 48 mg given three times a week (TIW). Model development leveraged a previous phase 1/2 PopPK model. Stepwise covariate analysis included 20 extrinsic and intrinsic factors. Model evaluation used standard goodness-of-fit and visual predictive checks.

RESULTS

Daprodustat PopPK was adequately characterized using a three-compartment distribution model with first-order elimination. The absorption phase was described using five transit compartments. Oral clearance and volume of distribution was 24.6 L/h and 26.9 L, respectively. Body weight dependence (with fixed allometric coefficients) of clearance and volume terms was a statistically significant covariate. Concomitant use of clopidogrel (moderate CYP2C8 inhibitor) decreased oral clearance, resulting in higher area under the plasma concentration-time curve (AUC) ratio of 1.59 (90% CI: 1.39-1.82), subjects' dialysis status (non-dialysis versus dialysis) had an effect on absorption, with Cmax ratio of 1.19 (90% CI: 1.09-1.30). None of the other investigated intrinsic or extrinsic covariates, including concomitant administration with phosphate binders, oral iron and acid reducing agents resulted in a significant change in daprodustat systemic exposure.

CONCLUSION

The PopPK of daprodustat in the CKD population with anemia was adequately characterized. Allometrically-scaled body weight on clearance and volume, dialysis status on absorption and clopidogrel on clearance were statistically significant covariates.

Serum bile acid change correlates with improvement in pruritus in patients with primary biliary cholangitis receiving linerixibat

BACKGROUND & AIMS

Total serum bile acid (TSBA) levels are elevated in patients with primary biliary cholangitis (PBC) and may mediate cholestatic pruritus. Linerixibat, an ileal bile acid transporter inhibitor, improved pruritus in patients with PBC. We explored the relationship between linerixibat dose, TSBA concentration, and pruritus.

METHODS

Data from Phase 1/2 trials were used to develop a population kinetic-pharmacodynamic model to characterize the linerixibat dose-TSBA relationship. Individual Bayesian parameter estimates for participants in the GLIMMER study were used to derive the area under the TSBA concentration curve over 24 h (AUC0-24). Time-matched post hoc estimates of AUC0-24 were correlated with pruritus reported on a 0-10 numerical rating scale. Baseline TSBA concentration was correlated with change from baseline (ΔBL) in monthly itch score (MIS). ΔBL in model-estimated TSBA AUC0-24 was correlated with time-matched ΔBL in weekly itch score (WIS) or MIS.

RESULTS

Linerixibat dose dependently reduced TSBA AUC0-24, reaching steady state after 5 days. Baseline TSBA levels in GLIMMER did not correlate with ΔBL in MIS. ΔBL in TSBA AUC0-24 correlated with improved WIS over 12 weeks of treatment (r = 0.52, p < 0.0001). Of participants with a ≥30% decrease in TSBA AUC0-24, 60% were pruritus responders (≥2-point improvement in WIS from baseline).

CONCLUSIONS

Linerixibat treatment leads to rapid, dose-dependent TSBA reductions. Baseline TSBA levels do not correlate with on-treatment pruritus change, suggesting they do not predict linerixibat response. Change in TSBA AUC0-24 correlates significantly with, and can be predictive of, pruritus improvement in patients with PBC.

Aging-Related CYP3A Functional Changes in Chinese Older Patients: New Findings from Model-Based Assessment of Amlodipine

Aging-related alterations in hepatic enzyme activity, particularly of the CYP3A, significantly impact drug efficacy and safety in older adults, making it essential to understand how aging affects CYP function for optimal drug therapy. The exogenous probe substrate method, a minimally invasive approach to assess liver metabolic enzyme activity in vivo, is effective in studying these changes. Amlodipine being extensively metabolized (> 90%) in the liver, primarily via cytochrome P450 enzyme CYP3A was selected as a probe to investigate and quantify the factors affecting the aging-related changes of CYP3A in the Chinese older population. Amlodipine concentration data were collected from an ongoing noninterventional clinical study conducted at Peking University Third Hospital. A physiologically-based pharmacokinetic modeling approach, grounded in population pharmacokinetic (PPK) analysis, was employed to physiologically quantify the aging-related changes in CYP3A function. A total of 132 amlodipine concentrations from 69 patients were obtained from the clinical study. PPK analysis shows that frailty phenotype but not age is a significant influence and frail patients have 37% greater plasma amlodipine exposure than nonfrail patients. This difference in CYP3A function may be attributed to a 63.2% lower CYP3A relative abundance in the frail patients, compared with that in the nonfrail patients. In the context of dose selection for older adults, focusing on frailty rather than chronological age should be recognized as a more relevant approach, because frailty might more accurately reflect the individual's biological age. Our study suggested a need to shift the research focus from chronological age to biological age.

Investigating the Influence of Covariates on Axicabtagene Ciloleucel (axi-cel) Kinetics in Patients with Non-Hodgkin's Lymphoma

BACKGROUND AND OBJECTIVE

Axicabtagene ciloleucel (axi-cel, Yescarta) is an autologous, anti-CD19, chimeric antigen receptor (CAR) T-cell therapy approved for patients with relapsed and refractory non-Hodgkin's lymphoma. Substantial inter-individual variability in cellular kinetics has been observed with CAR-T therapies and factors impacting CAR-T cellular kinetics remain poorly understood. This work reports a population cellular kinetic model of axi-cel in relapsed and patients with refractory non-Hodgkin's lymphoma and investigated the impact of covariates on early and late kinetic phases of CAR-T exposure.

METHODS

A population cellular kinetic model (NONMEM® version 7.4) for axi-cel was developed using data from 410 patients (2050 transgene observations) after a single intravenous infusion of 2 × 106 anti-CD19 CAR+ T cells/kg in patients with non-Hodgkin's lymphoma (ZUMA-1, ZUMA-5, and ZUMA-7 clinical studies). A large panel of covariates was assessed to decipher the variability of CAR-T cell kinetics including patient characteristics, product characteristics, and disease types.

RESULTS

Axi-cel cellular kinetics were well described by a piecewise model of cellular growth kinetics characterized by an exponential growth phase followed by a triphasic decline phase including a long-term persistence phase. The final cellular kinetic model retained in vitro doubling time during CAR-T cell manufacturing and total number of T cells infused as covariates impacting the duration of the growth phase, which, however, did not substantially influence maximum concentration, area under the concentration-time curve over the first 28 days, or long-term persistence. A statistically significant relationship was observed between maximum concentration and the probability to receive tocilizumab and/or corticosteroids.

CONCLUSIONS

No covariates considered in this study were found to significantly and substantially predict the exposure profile of axi-cel. Tocilizumab and steroid use were related to maximum concentration, but they were used reactively to treat toxicities that are associated with a higher maximum concentration. Further CAR-T kinetic analyses should consider additional factors to explain the observed variability in cellular kinetics or help establish a dose-exposure relationship.

CLINICAL TRIAL REGISTRATION

NCT02348216 (ZUMA-1), NCT03105336 (ZUMA-5), and NCT03391466 (ZUMA-7).

Infliximab in paediatric inflammatory bowel disease: External evaluation of population pharmacokinetic models

AIMS

Use of infliximab (IFX) has improved outcomes in children with inflammatory bowel disease (IBD). However, a proportion of patients does not respond to IFX or loses response over time. Population pharmacokinetic (PopPK) modelling is a promising approach for IFX dose optimization, but with the increasing number of PopPK models in literature, model evaluation is essential. The aims of this study are: (i) to validate the predictive performance of existing IFX PopPK models using a cohort of children with IBD; and (ii) to perform a Bayesian estimation of the most suitable model to predict the next IFX concentrations.

METHODS

PubMed was searched for IFX PopPK models in children. Selected models were rebuilt and analysed using R. Model performance was assessed through goodness-of-fit-plots, residuals against time, prediction error and prediction-corrected visual predictive checks. The validation cohort consisted of 73 children with IBD who were treated with IFX in our centre between 2017 and 2023 (340 IFX measurements).

RESULTS

We identified 9 PopPK models. Model bias for individual predicted values ranged from -9.29% to 8.01% compared to bias for population predicted values. The model by Vande Casteele et al. demonstrated superior performance (individual predicted bias 2.13, population predicted bias -6.11); upon Bayesian estimation, it predicted induction trough levels with median error of 12.95% but had a median error of -69% predicting maintenance concentrations.

CONCLUSION

The model by Vande Casteele et al. displayed superior performance in initial evaluations but had a high error in estimating next IFX levels and can only be used in practice to predict induction levels.

Visual predictive check of longitudinal models and dropout

Visual predictive checks (VPC) are commonly used to evaluate pharmacometrics models. However their performance may be hampered if patients with worse outcomes drop out earlier, as often occurs in clinical trials, especially in oncology. While methods accounting for dropouts have appeared in literature, they vary in assumptions, flexibility, and performance, and the differences between them are not widely understood. This manuscript aims to elucidate which methods can be used to handle VPC with dropout and when, along with a more informative VPC approach using confidence intervals. Additionally, we propose constructing the confidence interval based on the observed data instead of the simulated data. The theoretical framework for incorporating dropout in VPCs is developed and applied to propose two approaches: full and conditional. The full approach is implemented using a parametric time-to-event model, while the conditional approach is implemented using both parametric and Cox proportional-hazard (CPH) models. The practical performances of these approaches are illustrated with an application to the tumor growth dynamics (TGD) modeling of data from two cancer clinical trials of nivolumab and docetaxel, where patients were followed until disease progression. The dataset consisted of 3504 tumor size measurements from 855 subjects, which were described by a TGD model. The dropout of subjects was described by a Weibull or CPH model. Simulated datasets were also used to further illustrate the properties of the VPC methods. The results showed that the more familiar full approach might not provide meaningful improvement for TGD model evaluation over the naive approach of not adjusting for dropout, and could be outperformed by the conditional approach using either the Weibull model or the Cox proportional hazard model. Overall, including confidence intervals in VPC should improve interpretation, the conditional approach was shown to be more generally applicable when dropout occurs, and the nonparametric approach could provide additional robustness.

More Than a Gut Feeling─A Combination of Physiologically Driven Dissolution and Pharmacokinetic Modeling as a Tool for Understanding Human Gastric Motility

In vivo studies of formulation performance with in vitro and/or in silico simulations are often limited by significant gaps in our knowledge of the interaction between administered dosage forms and the human gastrointestinal tract. This work presents a novel approach for the investigation of gastric motility influence on dosage form performance, by combining biopredictive dissolution tests in an innovative PhysioCell apparatus with mechanistic physiology-based pharmacokinetic modeling. The methodology was based on the pharmacokinetic data from a large (n = 118) cohort of healthy volunteers who ingested a capsule containing a highly soluble and rapidly absorbed drug under fasted conditions. The developed dissolution tests included biorelevant media, varied fluid flows, and mechanical stress events of physiological timing and intensity. The dissolution results were used as inputs for pharmacokinetic modeling that led to the deduction of five patterns of gastric motility and their prevalence in the studied population. As these patterns significantly influenced the observed pharmacokinetic profiles, the proposed methodology is potentially useful to other in vitro-in vivo predictions involving immediate-release oral dosage forms.

Redefining Statin Dosage Post-Gastric Bypass: Insights from a Population Pharmacokinetics-Pharmacodynamics Link Approach

Roux-en-Y gastric bypass (RYGB) involves creating a small stomach pouch, bypassing part of the small intestine, and rerouting the digestive tract. These alterations can potentially change the drug exposure and response. Our primary aim was to assess the impact of RYGB on the pharmacokinetics of simvastatin lactone (SV) and its active metabolite, simvastatin hydroxy acid (SVA). Ultimately, we aimed to optimize dosing for this understudied population by employing a population pharmacokinetic-pharmacodynamic link approach. The study comprised patients who had undergone RYGB surgery and individuals without a previous history of RYGB. All participants received a single oral dose of simvastatin. Plasma concentration data were analyzed with a nonlinear mixed-effect modeling approach. A parent-metabolite model with first-order absorption, 2-compartments for SV and 1-compartment for SVA, linear elimination, and enterohepatic circulation best described the data. The model was linked to the turnover pharmacodynamic model to describe the SVA inhibition on LDL-cholesterol production. Our simulations indicated that following RYGB surgery, the exposure to SV and SVA decreased by 40%. Consequently, for low-intensity statin patients, we recommend increasing the dose from 10 to 20 mg in post-RYGB patients to maintain a comparable response to that of non-operated subjects. Moderate-intensity statin patients should require increasing doses to 40 or 60 mg or the addition of a non-statin medication to achieve similar therapeutic outcomes. In conclusion, individuals post-RYGB exhibit diminished exposure to SV and may benefit from increasing the dose or adjunctive therapy with non-statin drugs to attain equivalent responses and mitigate potential adverse events.

Niraparib Population Pharmacokinetics and Exposure-Response Relationships in Patients With Newly Diagnosed Advanced Ovarian Cancer

PURPOSE

Niraparib is a poly(adenosine diphosphate [ADP]-ribose) polymerase inhibitor approved for the maintenance treatment of advanced ovarian cancer (OC). Niraparib was originally approved in recurrent OC at a fixed starting dose (FSD) of 300 mg once daily (QD). This analysis characterized the population pharmacokinetics (PK) of niraparib and evaluated the relationships between exposure, efficacy, and safety to support clinical use of an individualized dosing strategy, in which the starting dose of niraparib was adjusted based on patient characteristics to improve the benefit-risk profile.

METHODS

A population PK model was developed by pooling data from four niraparib clinical trials (PN001 [n = 104], QUADRA [n = 455], NOVA [n = 403], and PRIMA [n = 480]) in patients with solid tumors, including OC. Exposure-response analyses were conducted to explore the relationships of niraparib exposure with progression-free survival (PFS) and adverse events in the PRIMA study. A multivariate logistic regression model was also developed to estimate the probability of grade ≥3 thrombocytopenia, using data from patients enrolled in PRIMA and NOVA. The impact of an individualized starting dose (ISD) regimen (200 mg QD in patients with body weight [BW] <77 kg or platelet count [PLT] <150,000/µL, or 300 mg QD in patients with BW ≥77 kg and PLT ≥150,000/µL) on systemic exposure, efficacy, and safety was assessed.

FINDINGS

Niraparib disposition was best described by a 3-compartment model with linear elimination. Key covariates included baseline creatinine clearance, BW, albumin, and age, all of which had minor effects on niraparib exposure. Comparable model-predicted exposure up to the time of disease progression/death or censoring in the 300-mg FSD and 200-/300-mg ISD groups was consistent with the lower rate of dose reduction in the ISD groups. No consistent niraparib exposure-response relationship was observed for efficacy in all PRIMA patients (first-line OC), and no statistically significant difference was seen in PFS curves for patients receiving a niraparib dose of 200 mg versus 300 mg. In the multivariate regression model, performed using combined data from PRIMA and NOVA, higher niraparib exposure (area under the concentration-time curve at steady-state [AUCss]), lower BW, and lower PLT were associated with an increased risk of grade ≥3 thrombocytopenia.

IMPLICATIONS

Population PK and exposure-response analyses support use of an ISD to improve the safety profile of niraparib, including reducing the rate of grade ≥3 thrombocytopenia, without compromising efficacy.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov, NCT01847274 (NOVA), NCT00749502 (PN001), NCT02655016 (PRIMA), NCT02354586 (QUADRA), www.

CLINICALTRIALS

gov.

Population Pharmacokinetic of the Diterpenes ent-Polyalthic Acid and Dihydro-ent-Agathic Acid from Copaifera Duckei Oil Resin in Rats

Copaifera duckei oleoresin is a plant product extensively used by the Brazilian population for multiple purposes, such as medicinal and cosmetic. Despite its ethnopharmacological relevance, there is no pharmacokinetic data on this important medicinal plant. Due to this, we determined the pharmacokinetic profile of the major nonvolatile compounds of C. duckei oleoresin. The diterpenes ent-polyalthic acid and dihydro-ent-agathic acid correspond to approximately 40% of the total oleoresin. Quantification was performed using LC-MS/MS, and the validated analytical method showed to be precise, accurate, robust, reliable, and linear between 0.57 and 114.74 µg/mL plasma and 0.09 to 18.85 µg/mL plasma, respectively, for ent-polyalthic acid and dihydro-ent-agathic acid, making it suitable for application in preclinical pharmacokinetic studies. Wistar rats received a single 200 mg/kg oral dose (gavage) of C. duckei oleoresin, and blood was collected from their caudal vein through 48 h. Population pharmacokinetics analysis of ent-polyalthic and dihydro-ent-agathic acids in rats was evaluated using nonlinear mixed-effects modeling conducted in NONMEN software. The pharmacokinetic parameters of ent-polyalthic acid were absorption constant rate = 0.47 h-1, central and peripheral apparent volume of distribution = 0.04 L and 2.48 L, respectively, apparent clearance = 0.15 L/h, and elimination half-life = 11.60 h. For dihydro-ent-agathic acid, absorption constant rate = 0.28 h-1, central and peripheral apparent volume of distribution = 0.01 L and 0.18 L, respectively, apparent clearance = 0.04 L/h, and elimination half-life = 3.49 h. The apparent clearance, central apparent volume of distribution, and peripheral apparent volume of distribution of ent-polyalthic acid were approximately 3.75, 4.00-, and 13.78-folds higher than those of dihydro-ent-agathic.

Prediction of Disease Progression and Clinical Response in Systemic Sclerosis: Experience From a Proof-of-Concept Trial

OBJECTIVE

Using the modified Rodnan skin score (mRSS) as a surrogate for disease activity, a phase 2a study in patients with systemic sclerosis (SSc) measured efficacy of the autotaxin inhibitor ziritaxestat. Mathematical modeling of mRSS was used to predict disease progression, examine candidate trial designs, and predict the probability of successfully discriminating treatment effect.

METHODS

Patients with SSc receiving 600 mg of ziritaxestat or placebo for 24 weeks were included, in addition to data up to week 52 of the open-label extension (OLE). Longitudinal mRSS data were described using a disease progression model; drug effect was a binary variable. Parameters used to predict the OLE mRSS outcome were estimated using data from the 24-week double-blind phase and validated with observed data. Three trial designs were simulated to identify which had the highest probability of detecting a treatment effect. Power to detect a treatment effect was quantified using the simulations.

RESULTS

Maximum decreases from baseline in mRSS were 50.4% (ziritaxestat) and 34.7% (placebo). Study designs based on 300 patients randomized 2:1 or 1:1 to 600 mg of ziritaxestat or placebo had similar probabilities of detecting a significant treatment effect. Power to detect a treatment effect was >80% for all simulations.

CONCLUSION

Disease progression and drug effect could be predicted beyond the range of observed data. This modeling and simulation approach may inform future trial design, including study duration, and predict the probability of success.

Effect of Interindividual Variability in Metabolic Clearance and Relative Bioavailability on Rifampicin Exposure in Tuberculosis Patients with and without HIV Co-Infection: Does Formulation Quality Matter?

The present study aims to characterise the pharmacokinetics of rifampicin (RIF) in tuberculosis (TB) patients with and without HIV co-infection, considering the formation of 25-O-desacetyl-rifampicin (desRIF). It is hypothesised that the metabolite formation, HIV co-infection and drug formulation may further explain the interindividual variation in the exposure to RIF. Pharmacokinetic, clinical, and demographic data from TB patients with (TB-HIV+ group; n = 18) or without HIV (TB-HIV- group; n = 15) who were receiving RIF as part of a four-drug fixed-dose combination (FDC) regimen (RIF, isoniazid, pyrazinamide, and ethambutol) were analysed, along with the published literature data on the relative bioavailability of different formulations. A population pharmacokinetic model, including the formation of desRIF, was developed and compared to a model based solely on the parent drug. HIV co-infection does not alter the plasma exposure to RIF and the desRIF formation does not contribute to the observed variability in the RIF disposition. The relative bioavailability and RIF plasma exposure were significantly lower than previously reported for the standard regimen with FDC tablets. Furthermore, participants weighting less than 50 kg do not reach the same RIF plasma exposure as compared to those weighting >50 kg. In conclusion, as no covariate was identified other than body weight on CL/F and Vd/F, low systemic exposure to RIF is likely to be caused by the low bioavailability of the formulation.

Understanding adefovir pharmacokinetics as a component of a transporter phenotyping cocktail

PURPOSE

Adefovir (as dipivoxil) was selected as a probe drug in a previous transporter cocktail phenotyping study to assess renal organic anion transporter 1 (OAT1), with renal clearance (CLR) as the primary parameter describing renal elimination. An approximately 20% higher systemic exposure of adefovir was observed when combined with other cocktail components (metformin, sitagliptin, pitavastatin, and digoxin) compared to sole administration. The present evaluation applied a population pharmacokinetic (popPK) modeling approach to describe adefovir pharmacokinetics as a cocktail component in more detail.

METHODS

Data from 24 healthy subjects were reanalyzed. After establishing a base model, covariate effects, including the impact of co-administered drugs, were assessed using forward inclusion then backward elimination.

RESULTS

A one-compartment model with first-order absorption (including lag time) and a combination of nonlinear renal and linear nonrenal elimination best described the data. A significantly higher apparent bioavailability (73.6% vs. 59.0%) and a lower apparent absorption rate constant (2.29 h-1 vs. 5.18 h-1) were identified in the combined period compared to the sole administration period, while no difference was seen in renal elimination. The population estimate for the Michaelis-Menten constant (Km) of the nonlinear renal elimination was 170 nmol/L, exceeding the observed range of adefovir plasma maximum concentration, while the maximum rate (Vmax) of nonlinear renal elimination was 2.40 µmol/h at the median absolute estimated glomerular filtration rate of 105 mL/min.

CONCLUSION

The popPK modeling approach indicated that the co-administration primarily affected the apparent absorption and/or prodrug conversion of adefovir dipivoxil, resulting in the minor drug-drug interaction observed for adefovir as a victim. However, renal elimination remained unaffected. The high Km value suggests that assessing renal OAT1 activity by CLR has no relevant misspecification error with the cocktail doses used.

Does Sample Size, Sampling Strategy, or Handling of Concentrations Below the Lower Limit of Quantification Matter When Externally Evaluating Population Pharmacokinetic Models?

BACKGROUND AND OBJECTIVES

Precision dosing requires selecting the appropriate population pharmacokinetic model, which can be assessed through external evaluations (EEs). The lack of understanding of how different study design factors influence EE study outcomes makes it challenging to select the most suitable model for clinical use. This study aimed to evaluate the impact of sample size, sampling strategy, and handling of concentrations below the lower limit of quantification (BLQ) on the outcomes of EE for four population pharmacokinetic models using vancomycin and tobramycin as examples.

METHODS

Three virtual patient populations undergoing vancomycin or tobramycin therapy were simulated with varying sample size and sampling scenarios. The three approaches used to handle BLQ data were to (1) discard them, (2) impute them as LLOQ/2, or (3) use a likelihood-based approach. EEs were performed with NONMEM and R.

RESULTS

Sample size did not have an important impact on the EE results for a given scenario. Increasing the number of samples per patient did not improve predictive performance for two out of the three evaluated models. Evaluating a model developed with rich sampling did not result in better performance than those developed with regular therapeutic drug monitoring. A likelihood-based method to handle BLQ samples impacted the outcomes of the EE with lower bias for predicted troughs.

CONCLUSIONS

This study suggests that a large sample size may not be necessary for an EE study, and models selected based on TDM may be more generalizable. The study highlights the need for guidelines for EE of population pharmacokinetic models for clinical use.

Population Pharmacokinetics and Absolute Oral Bioavailability of Lasalocid after Single Intravenous and Intracrop Administration in Laying Hens

Lasalocid sodium is a polyether carboxylic ionophore agent authorized by the EU for use as a coccidiostat in broilers, turkeys, and pullets up to 16 weeks of age, except for laying hens. However, laying hens are the most common nontarget species exposed to lasalocid sodium, mainly due to cross-contamination from feed mills. This exposure may result in potential drug residue deposition in eggs, which could potentially expose consumers to the drug. The breeds commonly used for commercial egg production in Poland are Isa Brown and Green-legged Partridge hens, which have been found to significantly differ in egg-laying performance. This variability may also affect the pharmacokinetics of lasalocid. Data on lasalocid plasma pharmacokinetics in laying hens are lacking. In this study, we aimed to determine typical population pharmacokinetic parameters, absolute oral bioavailability, and how breed may influence the pharmacokinetics of lasalocid. Twenty-layer hens of the two breeds were used in this study. Lasalocid was administered orally at a single dose of either 1 mg or 5 mg/kg body weight or intravenously at a dose of 1 mg/kg body weight, in a crossover design with a three-week washout period between study periods. Blood samples were collected for 72 h, and lasalocid concentrations were measured using high-performance liquid chromatography with fluorescence detection. A population pharmacokinetic analysis was conducted using nonlinear mixed effects modeling. Standard numerical and graphical criteria were used to select the best model, and a stepwise covariate modeling approach was used to determine any influencing factors. The best model was a three-compartment mammillary model with first-order absorption, transit compartments, and linear elimination. The estimated absolute oral bioavailability was low (36%). It was found that breed significantly influenced not only absorption but also the elimination of lasalocid. This study revealed that lasalocid absorption and elimination varied between the two breeds. This variability in pharmacokinetics may result in breed-related differences in drug residue accumulation in eggs, and ultimately, the risk associated with consumer exposure to drug residues may also vary.

Population pharmacokinetic analysis of sirolimus in Japanese pediatric and adult subjects receiving tablet or granule formulations

A population pharmacokinetic (PopPK) analysis was conducted using data from 215 Japanese administered oral sirolimus (tablet and granule) including healthy subjects and patients with intractable vascular anomalies and other diseases. The analysis included neonates, infants, and adults, and identified covariates that influence sirolimus pharmacokinetics (PK). The final model was used to predict sirolimus trough concentrations for various dosing regimens and covariates of interest. The results showed that sirolimus trough concentrations were predicted to increase with higher levels of hemoglobin, and that the granule formulation had a 1.23-fold higher exposure than the tablet formulation. Coadministration of CYP3A4 inducers was found to decrease trough concentrations by 54 %. The PK simulations showed that administration of the granule formulation at doses of 0.02, 0.04, 0.06, and 0.08 mg/kg/day in ages <3 months, 3 to <6 months, 6 to <12 months, and ≥1 year, respectively, resulted in >70 % target attainment within the therapeutic trough concentration range (5-15 ng/mL). In conclusion, incorporation of time-varying covariates (body weight and age) into the PopPK model appropriately predicted sirolimus concentrations in Japanese subjects from infants to adult sub-populations. This PopPK model would therefore be able to provide a reference for clinical individualization of sirolimus dosing.

Population pharmacokinetics, dosing optimization and clinical outcomes of biapenem in patients with sepsis

Introduction: Biapenem is a carbapenem antibiotic widely used in Asia, can be used for the treatment of adults and children with infections due to susceptible bacteria. Although biapenem is utilized in the treatment of a diverse range of bacterial infections, current pharmacokinetic data in the context of septic populations remain limited. Consequently, our research aims to evaluate the pharmacokinetics and efficacy of biapenem within a septic population to optimize biapenem therapy. Methods: In this study, we characterized the pharmacokinetics of biapenem in septic patients using a population pharmacokinetic (PPK) approach. The clinical PK data to develop the PPK model were obtained from 317 septic patients admitted to Nanjing Drum Tower Hospital between 2018 and 2022. All patients were randomized to the modeling and validation cohorts at a 3:1 ratio, with PPK modeling and validation performed utilizing the NONMEM software. Results: The model found to best describe the available data was a two-compartment PPK model with first-order elimination characterized by the parameters clearance (CL), central volume (V1), peripheral volume (V2), and intercompartmental clearance (Q). A covariate analysis identified that creatinine clearance (CLCR) was a significant covariate influencing biapenem CL, while blood urea nitrogen (BUN) was a significant covariate influencing biapenem Q. Accoding to the clinical outcome analyses, 70% of the time that the free antimicrobial drug concentration exceeds the MIC (fT >MIC) is associated with favourable clinical outcomes. The PPK model was then used to perform Monte Carlo simulations to evaluate the probability of attaining 70% fT >MIC. Conclusions: A final PPK model of biapenem was established for patients with sepsis. The current daily dosage regimen of 1.2 g may insufficient to achieve 70% fT >MIC in septic patients. The dosage regimen of 600 mg every 6 h appears to be the optimal choice.

Assessing potential drug-drug interactions between clofazimine and other frequently used agents to treat drug-resistant tuberculosis

Clofazimine is included in drug regimens to treat rifampicin/drug-resistant tuberculosis (DR-TB), but there is little information about its interaction with other drugs in DR-TB regimens. We evaluated the pharmacokinetic interaction between clofazimine and isoniazid, linezolid, levofloxacin, and cycloserine, dosed as terizidone. Newly diagnosed adults with DR-TB at Klerksdorp/Tshepong Hospital, South Africa, were started on the then-standard treatment with clofazimine temporarily excluded for the initial 2 weeks. Pharmacokinetic sampling was done immediately before and 3 weeks after starting clofazimine, and drug concentrations were determined using validated liquid chromatography-tandem mass spectrometry assays. The data were interpreted with population pharmacokinetics in NONMEM v7.5.1 to explore the impact of clofazimine co-administration and other relevant covariates on the pharmacokinetics of isoniazid, linezolid, levofloxacin, and cycloserine. Clofazimine, isoniazid, linezolid, levofloxacin, and cycloserine data were available for 16, 27, 21, 21, and 6 participants, respectively. The median age and weight for the full cohort were 39 years and 52 kg, respectively. Clofazimine exposures were in the expected range, and its addition to the regimen did not significantly affect the pharmacokinetics of the other drugs except levofloxacin, for which it caused a 15% reduction in clearance. A posteriori power size calculations predicted that our sample sizes had 97%, 90%, and 87% power at P < 0.05 to detect a 30% change in clearance of isoniazid, linezolid, and cycloserine, respectively. Although clofazimine increased the area under the curve of levofloxacin by 19%, this is unlikely to be of great clinical significance, and the lack of interaction with other drugs tested is reassuring.

RPEM: Randomized Monte Carlo parametric expectation maximization algorithm

Inspired from quantum Monte Carlo, by sampling discrete and continuous variables at the same time using the Metropolis-Hastings algorithm, we present a novel, fast, and accurate high performance Monte Carlo Parametric Expectation Maximization (MCPEM) algorithm. We named it Randomized Parametric Expectation Maximization (RPEM). We compared RPEM with NONMEM's Importance Sampling Method (IMP), Monolix's Stochastic Approximation Expectation Maximization (SAEM), and Certara's Quasi-Random Parametric Expectation Maximization (QRPEM) for a realistic two-compartment voriconazole model with ordinary differential equations using simulated data. We show that RPEM is as fast and as accurate as the algorithms IMP, QRPEM, and SAEM for the voriconazole model in reconstructing the population parameters, for the normal and log-normal cases.

Population Pharmacokinetic Analysis of Atomoxetine and its Metabolites in Children and Adolescents with Attention-Deficit/Hyperactivity Disorder

Atomoxetine (ATX) is a non-stimulant used to treat attention-deficit/hyperactivity disorder (ADHD) and systemic exposure is highly variable due to polymorphic cytochrome P450 2D6 (CYP2D6) activity. The objective of this study was to characterize the time course of ATX and metabolites (4-hydroxyatomoxetine (4-OH); N-desmethylatomoxetine (NDA); and 2-carboxymethylatomoxetine (2-COOH)) exposure following oral ATX dosing in children with ADHD to support individualized dosing. A nonlinear mixed-effect modeling approach was used to analyze ATX, 4-OH, and NDA plasma and urine, and 2-COOH urine profiles obtained over 24-72 hours from children with ADHD (n = 23) following a single oral ATX dose. Demographics and CYP2D6 activity score (AS) were evaluated as covariates. Simulations were performed to explore the ATX dosing in subjects with various CYP2D6 AS. A simultaneous pharmacokinetic modeling approach was used in which a model for ATX, 4-OH, and NDA in plasma and urine, and 2-COOH in urine was developed. Plasma ATX, 4-OH, and NDA were modeled using two-compartment models with first-order elimination. CYP2D6 AS was a significant determinant of ATX apparent oral clearance (CL/F), fraction metabolized to 4-OH, and systemic exposure of NDA. CL/F of ATX varied almost 7-fold across the CYP2D6 AS groups: AS 2: 20.02 L/hour; AS 1: 19.00 L/hour; AS 0.5: 7.47 L/hour; and AS 0: 3.10 L/hour. The developed model closely captures observed ATX, 4-OH, and NDA plasma and urine, and 2-COOH urine profiles. Application of the model shows the potential for AS-based dosing recommendations for improved individualized dosing.

Integrative model-based comparison of target site-specific antimicrobial effects: A case study with ceftaroline and lefamulin

OBJECTIVE

Predictions of antimicrobial effects typically rely on plasma-based pharmacokinetic-pharmacodynamic (PK-PD) targets, ignoring target-site concentrations and potential differences in tissue penetration between antibiotics. In this study, we applied PK-PD modelling to compare target site-specific effects of antibiotics by integrating clinical microdialysis data, in vitro time-kill curves, and antimicrobial susceptibility distributions. As a case study, we compared the effect of lefamulin and ceftaroline against methicillin-resistant Staphylococcus aureus (MRSA) at soft-tissue concentrations.

METHODS

A population PK model describing lefamulin concentrations in plasma, subcutaneous adipose and muscle tissue was developed. For ceftaroline, a similar previously reported PK model was adopted. In vitro time-kill experiments were performed with six MRSA isolates and a PD model was developed to describe bacterial growth and antimicrobial effects. The clinical PK and in vitro PD models were linked to compare antimicrobial effects of ceftaroline and lefamulin at the different target sites.

RESULTS

Considering minimum inhibitory concentration (MIC) distributions and standard dosages, ceftaroline showed superior anti-MRSA effects compared to lefamulin both at plasma and soft-tissue concentrations. Looking at the individual antibiotics, lefamulin effects were highest at soft-tissue concentrations, while ceftaroline effects were highest at plasma concentrations, emphasising the importance of considering target-site PK-PD in antibiotic treatment optimisation.

CONCLUSION

Given standard dosing regimens, ceftaroline appeared more effective than lefamulin against MRSA at soft-tissue concentrations. The PK-PD model-based approach applied in this study could be used to compare or explore the potential of antibiotics for specific indications or in populations with unique target-site PK.

Prediction of Individual Disease Progression Including Parameter Uncertainty in Rare Neurodegenerative Diseases: The Example of Autosomal-Recessive Spastic Ataxia Charlevoix Saguenay (ARSACS)

The aim of this study was to develop a model to predict individual subject disease trajectories including parameter uncertainty and accounting for missing data in rare neurological diseases, showcased by the ultra-rare disease Autosomal-Recessive Spastic Ataxia Charlevoix Saguenay (ARSACS). We modelled the change in SARA (Scale for Assessment and Rating of Ataxia) score versus Time Since Onset of symptoms using non-linear mixed effect models for a population of 173 patients with ARSACS included in the prospective real-world multicenter Autosomal Recessive Cerebellar Ataxia (ARCA) registry. We used the Multivariate Imputation Chained Equation (MICE) algorithm to impute missing covariates, and a covariate selection procedure with a pooled p-value to account for the multiply imputed data sets. We then investigated the impact of covariates and population parameter uncertainty on the prediction of the individual trajectories up to 5 years after their last visit. A four-parameter logistic function was selected. Men were estimated to have a 25% lower SARA score at disease onset and a moderately higher maximum SARA score, and time to progression (T50) was estimated to be 35% lower in patients with age of onset over 15 years. The population disease progression rate started slowly at 0.1 points per year peaking to a maximum of 0.8 points per year (at 36.8 years since onset of symptoms). The prediction intervals for SARA scores 5 years after the last visit were large (median 7.4 points, Q1-Q3: 6.4-8.5); their size was mostly driven by individual parameter uncertainty and individual disease progression rate at that time.

A systematic evaluation of population pharmacokinetic models for polymyxin B in patients with liver and/or kidney dysfunction

Polymyxin B (PMB) is considered a last-line treatment for multidrug-resistant (MDR) gram-negative bacterial infections. Model-informed precision dosing with population pharmacokinetics (PopPK) models could help to individualize PMB dosing regimens and improve therapy. However, the external prediction ability of the established PopPK models has not been fully elaborated. This study aimed to systemically evaluate eleven PMB PopPK models from ten published literature based on a new independent population, which was divided into four different populations, patients with liver dysfunction, kidney dysfunction, liver and kidney dysfunction, and normal liver and kidney function. The whole data set consisted of 146 patients with 391 PMB concentrations. The prediction- and simulation-based diagnostics and Bayesian forecasting were conducted to evaluate model predictability. In the overall evaluation process, none of the models exhibited satisfactory predictive ability in both prediction- and simulation-based diagnostic simultaneously. However, the evaluation of the models in the subgroup of patients with normal liver and kidney function revealed improved predictive performance compared to those with liver and/or kidney dysfunction. Bayesian forecasting demonstrated enhanced predictability with the incorporation of two to three prior observations. The external evaluation highlighted a lack of consistency between the prediction results of published models and the external validation dataset. Nonetheless, Bayesian forecasting holds promise in improving the predictive performance of the models, and feedback from therapeutic drug monitoring is crucial in optimizing individual dosing regimens.

Virtual Patient Simulation Using Copula Modeling

Virtual patient simulation is increasingly performed to support model-based optimization of clinical trial designs or individualized dosing strategies. Quantitative pharmacological models typically incorporate individual-level patient characteristics, or covariates, which enable the generation of virtual patient cohorts. The individual-level patient characteristics, or covariates, used as input for such simulations should accurately reflect the values seen in real patient populations. Current methods often make unrealistic assumptions about the correlation between patient's covariates or require direct access to actual data sets with individual-level patient data, which may often be limited by data sharing limitations. We propose and evaluate the use of copulas to address current shortcomings in simulation of patient-associated covariates for virtual patient simulations for model-based dose and trial optimization in clinical pharmacology. Copulas are multivariate distribution functions that can capture joint distributions, including the correlation, of covariate sets. We compare the performance of copulas to alternative simulation strategies, and we demonstrate their utility in several case studies. Our work demonstrates that copulas can reproduce realistic patient characteristics, both in terms of individual covariates and the dependence structure between different covariates, outperforming alternative methods, in particular when aiming to reproduce high-dimensional covariate sets. In conclusion, copulas represent a versatile and generalizable approach for virtual patient simulation which preserve relationships between covariates, and offer an open science strategy to facilitate re-use of patient data sets.

Estimating potential cardiovascular health benefits of improved population level control of LDL cholesterol through a twice-yearly siRNA-based approach: A simulation study of a health-system level intervention

BACKGROUND AND AIMS

Inclisiran, an siRNA therapy, consistently reduces low-density lipoprotein cholesterol (LDL-C) with twice-yearly dosing. Potential cardiovascular benefits of implementing inclisiran at a population level, added to statins, were evaluated through simulation.

METHODS

For each participant in the ORION-10 and ORION-11 trials comparing inclisiran with placebo, baseline 10-year cardiovascular risk was estimated using the SMART equation. The time-adjusted LDL-C difference from baseline observed 90-540 days after baseline was assumed to persist and used to estimate potential reduction in 10-year cardiovascular risk. Impact on 500,000 ORION-like individuals was simulated with Monte-Carlo.

RESULTS

Mean baseline LDL-C and predicted 10-year major vascular risk among patients randomized to inclisiran (n = 1288) versus placebo (n = 1264) were 2.66 mmol/L versus 2.60 mmol/L and 24.9% versus 24.6%, respectively. Placebo-corrected time-adjusted absolute reduction in LDL-C with inclisiran was -1.32 mmol/L (95% CI -1.37 to -1.26; p < 0.001), which predicted a 10-year cardiovascular risk of 18.1% with inclisiran versus 24.7% with placebo (absolute difference [95% CI], -6.99% [-7.33 to -6.66]; p < 0.001) NNT 15. Extrapolating to 500,000 inclisiran-treated individuals, the model predicted large population shifts towards lower quintiles of risk with fewer remaining in high-risk categories; 3350 to 471 (≥80% risk), 11,793 to 3332 (60-<80% risk), 52,142 to 22,665 (40-<60% risk), 197,752 to 141,014 (20-<40% risk), and more moving into the lowest risk category (<20%) from 234,963 to 332,518.

CONCLUSIONS

Meaningful gains in population health might be achieved over 10 years by implementing at-scale approaches capable of providing substantial and sustained reductions in LDL-C beyond those achievable with statins.

Pharmacokinetic-Pharmacodynamic Modelling in Hemophilia A: Relating Thrombin and Plasmin Generation to Factor VIII Activity After Administration of a VWF/FVIII Concentrate

BACKGROUND

Hemophilia A patients are treated with factor (F) VIII prophylactically to prevent bleeding. In general, dosage and frequency are based on pharmacokinetic measurements. Ideally, an alternative dose adjustment can be based on the hemostatic potential, measured with a thrombin generation assay (TGA), like the Nijmegen hemostasis assay.

OBJECTIVE

The objective of this study was to investigate the predicted performance of a previously developed pharmacokinetic-pharmacodynamic model for FVIII replacement therapy, relating FVIII dose and FVIII activity levels with thrombin and plasmin generation parameters.

METHODS

Pharmacokinetic and pharmacodynamic measurements were obtained from 29 severe hemophilia A patients treated with pdVWF/FVIII concentrate (Haemate P®). The predictive performance of the previously developed pharmacokinetic-pharmacodynamic model was evaluated using nonlinear mixed-effects modeling (NONMEM). When predictions of FVIII activity or TGA parameters were inadequate [median prediction error (MPE) > 20%], a new model was developed.

RESULTS

The original pharmacokinetic model underestimated clearance and was refined based on a two-compartment model. The pharmacodynamic model displays no bias in the observed normalized thrombin peak height and normalized thrombin potential (MPE of 6.83% and 7.46%). After re-estimating pharmacodynamic parameters, EC50 and Emax values were relatively comparable between the original model and this group. Prediction of normalized plasmin peak height was inaccurate (MPE 58.9%).

CONCLUSION

Our predictive performance displayed adequate thrombin pharmacodynamic predictions of the original model, but a new pharmacokinetic model was required. The pharmacodynamic model is not factor specific and applicable to multiple factor concentrates. A prospective study is needed to validate the impact of the FVIII dosing pharmacodynamic model on bleeding reduction in patients.

Population pharmacokinetic/pharmacodynamic modeling and exposure-response analysis of ciprofol in the induction and maintenance of general anesthesia in patients undergoing elective surgery: A prospective dose optimization study

AIM

This study aimed to establish a population pharmacokinetic and pharmacodynamic (PK-PD) model to explore the optimal maintenance dose and appropriate starting time of maintenance dose after induction of ciprofol and investigate the efficacy and safety of ciprofol for general anesthesia induction and maintenance in patients undergoing elective surgery.

METHOD

A total of 334 subjects with 3092 concentration measurements from nine clinical trials and 115 subjects with 5640 bispectral index (BIS) measurements from two clinical trials were used in the population PK-PD analysis. Exposure-response relationships for both efficacy endpoints (duration of anesthesia successful induction, time to recovery from anesthesia, time to respiratory recovery, and time from discontinuation to the 1st/3rd consecutive Aldrete score ≥ 9) and safety variables (hypotension, bradycardia, and injection site pain) were evaluated based on the data gathered from 115 subjects in two clinical trials.

RESULT

Ciprofol pharmacokinetics (PK) were adequately described by a three-compartment model with first-order elimination from the central compartment and redistribution from the deep and shallow peripheral compartments. An inhibitory sigmoidal Emax model best described the relationship between ciprofol effect-site concentrations and BIS measurements. Body weight, age, sex, blood sampling site, and study type (short-term infusion vs long-term infusion) were identified as statistically significant covariates on the PK of ciprofol. No covariates were found to have a significant effect on the pharmacodynamic (PD) parameters. The PK-PD simulation results showed that the optimal maintenance dose was 0.8 mg/kg/h and the appropriate time to start the maintenance dose was 4-5 mins after the induction dose of ciprofol. Within the exposure range of this study, no meaningful correlations between ciprofol exposures and efficacy or safety endpoints were observed.

CONCLUSION

A population PK-PD model was successfully developed to describe the ciprofol PK and BIS changes. Efficacy was consistent across the exposure range with a well-tolerated safety profile indicating no maintenance dose adjustment is required for patients undergoing elective surgery.

Population pharmacokinetics of methotrexate and 7-hydroxymethotrexate and delayed excretion in infants and young children with brain tumors

PURPOSE

The objectives of this study were to develop a population pharmacokinetic model of methotrexate (MTX) and its primary metabolite 7-hydroxymethotrexate (7OHMTX) in children with brain tumors, to identify the sources of pharmacokinetic variability, and to assess whether MTX and 7OHMTX systemic exposures were related to toxicity.

METHODS

Patients received 2.5 or 5 g/m2 MTX as a 24-hour infusion and serial samples were analyzed for MTX and 7OHMTX by an LC-MS/MS method. Pharmacokinetic parameters were estimated using nonlinear mixed-effects modeling. Demographics, laboratory values, and genetic polymorphisms were considered as potential covariates to explain the pharmacokinetic variability. Association between MTX and 7OHMTX systemic exposures and MTX-related toxicities were explored using random intercept logistic regression models.

RESULTS

The population pharmacokinetics of MTX and 7OHMTX were adequately characterized using two-compartment models in 142 patients (median 1.91 y; age range 0.09 to 4.94 y) in 513 courses. The MTX and 7OHMTX population clearance values were 4.6 and 3.0 l/h/m2, respectively. Baseline body surface area and estimated glomerular filtration rate were significant covariates on both MTX and 7OHMTX plasma disposition. Pharmacogenetic genotypes were associated with MTX pharmacokinetic parameters but had only modest influence. No significant association was observed between MTX or 7OHMTX exposure and MTX-related toxicity.

CONCLUSIONS

MTX and 7OHMTX plasma disposition were characterized for the first time in young children with brain tumors. No exposure-toxicity relationship was identified in this study, presumably due to aggressive clinical management which led to a low MTX-related toxicity rate.

Model-Informed Precision Dosing to Reduce Vincristine-Induced Peripheral Neuropathy in Pediatric Patients: A Pharmacokinetic and Pharmacodynamic Modeling and Simulation Analysis

BACKGROUND

Vincristine-induced peripheral neuropathy (VIPN) is a common adverse effect of vincristine, a drug often used in pediatric oncology. Previous studies demonstrated large inter- and intrapatient variability in vincristine pharmacokinetics (PK). Model-informed precision dosing (MIPD) can be applied to calculate patient exposure and individualize dosing using therapeutic drug monitoring (TDM) measurements. This study set out to investigate the PK/pharmacodynamic (PKPD) relationship of VIPN and determine the utility of MIPD to support clinical decisions regarding dose selection and individualization.

METHODS

Data from 35 pediatric patients were utilized to quantify the relationship between vincristine dose, exposure and the development of VIPN. Measurements of vincristine exposure and VIPN (Common Terminology Criteria for Adverse Events [CTCAE]) were available at baseline and for each subsequent dosing occasions (1-5). A PK and PKPD analysis was performed to assess the inter- and intraindividual variability in vincristine exposure and VIPN over time. In silico trials were performed to portray the utility of vincristine MIPD in pediatric subpopulations with a certain age, weight and cytochrome P450 (CYP) 3A5 genotype distribution.

RESULTS

A two-compartmental model with linear PK provided a good description of the vincristine exposure data. Clearance and distribution parameters were related to bodyweight through allometric scaling. A proportional odds model with Markovian elements described the incidence of Grades 0, 1 and ≥ 2 VIPN overdosing occasions. Vincristine area under the curve (AUC) was the most significant exposure metric related to the development of VIPN, where an AUC of 50 ng⋅h/mL was estimated to be related to an average VIPN probability of 40% over five dosing occasions. The incidence of Grade ≥ 2 VIPN reduced from 62.1 to 53.9% for MIPD-based dosing compared with body surface area (BSA)-based dosing in patients. Dose decreases occurred in 81.4% of patients with MIPD (vs. 86.4% for standard dosing) and dose increments were performed in 33.4% of patients (no dose increments allowed for standard dosing).

CONCLUSIONS

The PK and PKPD analysis supports the use of MIPD to guide clinical dose decisions and reduce the incidence of VIPN. The current work can be used to support decisions with respect to dose selection and dose individualization in children receiving vincristine.

Bayesian estimation in NONMEM

Bayesian estimation is a powerful but underutilized tool for answering drug development questions. In this tutorial, the principles of Bayesian model development, assessment, and prior selection will be outlined. An example pharmacokinetic (PK) model will be used to demonstrate the implementation of Bayesian modeling using the nonlinear mixed-effects modeling software NONMEM.

A tutorial for model-based evaluation and translation of cardiovascular safety in preclinical trials

Assessment of drug-induced effects on the cardiovascular (CV) system remains a critical component of the drug discovery process enabling refinement of the therapeutic index. Predicting potential drug-related unintended CV effects in the preclinical stage is necessary for first-in-human dose selection and preclusion of adverse CV effects in the clinical stage. According to the current guidelines for small molecules, nonclinical CV safety assessment conducted via telemetry analyses should be included in the safety pharmacology core battery studies. However, the manual for quantitative evaluation of the CV safety signals in animals is available only for electrocardiogram parameters (i.e., QT interval assessment), not for hemodynamic parameters (i.e., heart rate, blood pressure, etc.). Various model-based approaches, including empirical pharmacokinetic-toxicodynamic analyses and systems pharmacology modeling could be used in the framework of telemetry data evaluation. In this tutorial, we provide a comprehensive workflow for the analysis of nonclinical CV safety on hemodynamic parameters with a sequential approach, highlight the challenges associated with the data, and propose respective solutions, complemented with a reproducible example. The work is aimed at helping researchers conduct model-based analyses of the CV safety in animals with subsequent translation of the effect to humans seamlessly and efficiently.

Preliminary pharmacokinetics and patient experience of jet-injected dexmedetomidine in healthy adults

Jet injection is a drug delivery system without a needle. A compressed liquid drug formulation pierces the skin, depositing the drug into the subcutaneous or intramuscular tissues. We investigated the pharmacokinetics and patient experience of dexmedetomidine administered using jet injection in six healthy adult study participants. This needleless jet injection device was used to administer dexmedetomidine 0.5 μg/kg to the subcutaneous tissues overlying the deltoid muscle. Serum concentrations of dexmedetomidine were assayed at approximately 5 minutes, 15 minutes, 30 minutes, 1 hour and 4 hours after administration. Pharmacokinetic interrogation of concentration time profiles estimated an absorption half time for jet-injected dexmedetomidine of 21 minutes (coefficient of variation 69.4%) with a relative bioavailability assumed unity. In our samples the measured median peak (range) concentration was 0.164 μg/l (0.011-0.325 μg/l), observed in the sample taken at a median (range) of 13.5 minutes (11-30 minutes). The Richmond agitation sedation scale was used to assess the sedative effect, and scored 0 (alert and calm) or -1 (drowsy) in all participants. Five of the six participants stated they would prefer jet injection to needle injection in the future and one had no preference. The findings suggest that the use of a larger dose (>2 μg/kg) would be required to achieve the clinically relevant target concentration of 1 μg/l necessary to achieve deeper sedation (Richmond agitation sedation scale ≤3).

Nonlinear multilevel joint model for individual lesion kinetics and survival to characterize intra-individual heterogeneity in patients with advanced cancer

In advanced cancer patients, tumor burden is calculated using the sum of the longest diameters (SLD) of the target lesions, a measure that lumps all lesions together and ignores intra-patient heterogeneity. Here, we used a rich dataset of 342 metastatic bladder cancer patients treated with a novel immunotherapy agent to develop a Bayesian multilevel joint model that can quantify heterogeneity in lesion dynamics and measure their impact on survival. Using a nonlinear model of tumor growth inhibition, we estimated that dynamics differed greatly among lesions, and inter-lesion variability accounted for 21% and 28% of the total variance in tumor shrinkage and treatment effect duration, respectively. Next, we investigated the impact of individual lesion dynamics on survival. Lesions located in the liver and in the bladder had twice as much impact on the instantaneous risk of death compared to those located in the lung or the lymph nodes. Finally, we evaluated the utility of individual lesion follow-up for dynamic predictions. Consistent with results at the population level, the individual lesion model outperformed a model relying only on SLD, especially at early landmark times and in patients with liver or bladder target lesions. Our results show that an individual lesion model can characterize the heterogeneity in tumor dynamics and its impact on survival in advanced cancer patients.

Rifampicin reduces plasma concentration of linezolid in patients with infective endocarditis

BACKGROUND

Linezolid in combination with rifampicin has been used in treatment of infective endocarditis especially for patients infected with staphylococci.

OBJECTIVES

Because rifampicin has been reported to reduce the plasma concentration of linezolid, the present study aimed to characterize the population pharmacokinetics of linezolid for the purpose of quantifying an effect of rifampicin cotreatment. In addition, the possibility of compensation by dosage adjustments was evaluated.

PATIENTS AND METHODS

Pharmacokinetic measurements were performed in 62 patients treated with linezolid for left-sided infective endocarditis in the Partial Oral Endocarditis Treatment (POET) trial. Fifteen patients were cotreated with rifampicin. A total of 437 linezolid plasma concentrations were obtained. The pharmacokinetic data were adequately described by a one-compartment model with first-order absorption and first-order elimination.

RESULTS

We demonstrated a substantial increase of linezolid clearance by 150% (95% CI: 78%-251%), when combined with rifampicin. The final model was evaluated by goodness-of-fit plots showing an acceptable fit, and a visual predictive check validated the model. Model-based dosing simulations showed that rifampicin cotreatment decreased the PTA of linezolid from 94.3% to 34.9% and from 52.7% to 3.5% for MICs of 2 mg/L and 4 mg/L, respectively.

CONCLUSIONS

A substantial interaction between linezolid and rifampicin was detected in patients with infective endocarditis, and the interaction was stronger than previously reported. Model-based simulations showed that increasing the linezolid dose might compensate without increasing the risk of adverse effects to the same degree.

Multivariate Exact Discrepancy: A New Tool for PK/PD Model Evaluation

BACKGROUND

Pharmacokinetic models are evaluated using three types of metrics: those based on estimating the typical pharmacokinetic parameters, those based on predicting individual pharmacokinetic parameters and those that compare data and model distributions. In the third groups of metrics, the best-known methods are Visual Predictive Check (VPC) and Normalised Prediction Distribution Error (NPDE). Despite their usefulness, these methods have some limitations, especially for the analysis of dependent concentrations, i.e., evaluated in the same patient.

OBJECTIVE

In this work, we propose an evaluation method that accounts for the dependency between concentrations.

METHODS

Thanks to the study of the distribution of simulated vectors of concentrations, the method provides one probability per individual that its observations (i.e., concentrations) come from the studied model. The higher the probability, the better the model fits the individual. By examining the distribution of these probabilities for a set of individuals, we can evaluate the model as a whole.

RESULTS

We demonstrate the effectiveness of our method through two examples. Our approach successfully detects misspecification in the structural model and identifies outlier kinetics in a set of kinetics.

CONCLUSION

We propose a straightforward method for evaluating models during their development and selecting a model to perform therapeutic drug monitoring. Based on our preliminary results, the method is very promising but needs to be validated on a larger scale.

Topotecan clearance based on a single sample and a population pharmacokinetic model: Application to a pediatric high-risk neuroblastoma clinical trial

BACKGROUND

Topotecan, an antitumor drug with systemic exposure (SE)-dependent activity against many pediatric tumors has wide interpatient pharmacokinetic variability, making it challenging to attain the desired topotecan SE. The study objectives were to update our topotecan population pharmacokinetic model, to evaluate the feasibility of determining individual topotecan clearance using a single blood sample, and to apply this approach to topotecan data from a neuroblastoma trial to explore exposure-response relationships.

PROCEDURE

Our previous population pharmacokinetic and covariate model was updated using data from 13 clinical pediatric studies. A simulation-based Bayesian analysis was performed to determine if a single blood sample could be sufficient to estimate individual topotecan clearance. Following the Bayesian approach, single pharmacokinetic samples collected from a Children's Oncology Group Phase III clinical trial (ANBL0532; NCT0056767) were analyzed to estimate individual topotecan SE. Associations between topotecan SE and toxicity or early response were then evaluated.

RESULTS

The updated population model included the impact of patient body surface area (BSA), age, and renal function on topotecan clearance. The Bayesian analysis with the updated model and single plasma samples showed that individual topotecan clearance values were estimated with good precision (mean absolute prediction error ≤16.2%) and low bias (mean prediction error ≤7.2%). Using the same approach, topotecan SE was derived in patients from ANBL0532. The exposure-response analysis showed an increased early response after concomitant cyclophosphamide and topotecan up to a topotecan SE of 45 h ng/mL.

CONCLUSIONS

A simple single-sample approach during topotecan therapy could guide dosing for patients, resulting in more patients reaching target attainment.

Bridging the Worlds of Pharmacometrics and Machine Learning

Precision medicine requires individualized modeling of disease and drug dynamics, with machine learning-based computational techniques gaining increasing popularity. The complexity of either field, however, makes current pharmacological problems opaque to machine learning practitioners, and state-of-the-art machine learning methods inaccessible to pharmacometricians. To help bridge the two worlds, we provide an introduction to current problems and techniques in pharmacometrics that ranges from pharmacokinetic and pharmacodynamic modeling to pharmacometric simulations, model-informed precision dosing, and systems pharmacology, and review some of the machine learning approaches to address them. We hope this would facilitate collaboration between experts, with complementary strengths of principled pharmacometric modeling and flexibility of machine learning leading to synergistic effects in pharmacological applications.

Population Pharmacodynamic Modelling of the CD19+ Suppression Effects of Rituximab in Paediatric Patients with Neurological and Autoimmune Diseases

BACKGROUND

Limited pharmacotherapy and the failure of conventional treatments in complex pathologies in children lead to increased off-label use of rituximab. We aimed to characterize the time course of CD19+ B lymphocytes (CD19+) under treatment with intravenous rituximab in children with neurologic and autoimmune diseases and to evaluate the impact of covariates (i.e., demographics, diagnosis and substitution between innovator and biosimilar product) on rituximab pharmacodynamics and disease activity.

METHODS

Pre- and post-drug infusion CD19+ in peripheral blood were prospectively registered. A population pharmacodynamic model describing the time course of CD19+ was developed with NONMEM v7.4. Simulations of three different rituximab regimens were performed to assess the impact on CD19+. Logistic regression analysis was performed to identify predictors of clinical response recorded through disease activity scores.

RESULTS

281 measurements of CD19+ lymphocyte counts obtained from 63 children with neurologic (n = 36) and autoimmune (n = 27) diseases were available. The time course of CD19+ was described with a turn-over model in which the balance between synthesis and degradation rates is disrupted by rituximab, increasing the latter process. The model predicts half-lives (percent coefficient of variation, CV(%)) of rituximab and CD19+ of 11.6 days (17%) and 173.3 days (22%), respectively. No statistically significant effect was found between any of the studied covariates and model parameters (p > 0.05). Simulations of different regimens showed no clinically significant differences in terms of CD19+ repopulation times. A trend towards a lack of clinical response was observed in patients with lower CD19+ repopulation times and higher areas under the CD19+ versus time curve.

CONCLUSIONS

Rituximab pharmacodynamics was described in a real-world setting in children suffering from autoimmune and neurologic diseases. Diagnosis, substitution between innovator rituximab and its biosimilars or type of regimen did not affect rituximab-induced depletion of CD19+ nor the clinical response in this cohort of patients. According to this study, rituximab frequency and dosage may be chosen based on clinical convenience or safety reasons without affecting CD19+ repopulation times. Further studies in larger populations are required to confirm these results.

Mechanistic modeling projections of antibody persistence after homologous booster regimens of COVID-19 vaccine Ad26.COV2.S in humans

Mechanistic model-based simulations can be deployed to project the persistence of humoral immune response following vaccination. We used this approach to project the antibody persistence through 24 months from the data pooled across five clinical trials in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2)-seronegative participants following vaccination with Ad26.COV2.S (5 × 1010 viral particles), given either as a single-dose or a homologous booster regimen at an interval of 2, 3, or 6 months. Antibody persistence was quantified as the percentage of participants with detectable anti-spike binding and wild-type virus neutralizing antibodies. The projected overall 24-month persistence after single-dose Ad26.COV2.S was 70.5% for binding antibodies and 55.2% for neutralizing antibodies, and increased after any homologous booster regimen to greater than or equal to 89.9% for binding and greater than or equal to 80.0% for neutralizing antibodies. The estimated model parameters quantifying the rates of antibody production attributed to short-lived and long-lived plasma cells decreased with increasing age, whereas the rate of antibody production mediated by long-lived plasma cells was higher in women relative to men. Accordingly, a more pronounced waning of antibody responses was predicted in men aged greater than or equal to 60 years and was markedly attenuated following any homologous boosting regimen. The findings suggest that homologous boosting might be a viable strategy for maintaining protective effects of Ad26.COV2.S for up to 24 months following prime vaccination. The estimation of mechanistic modeling parameters identified the long-lived plasma cell pathway as a key contributor mediating antibody persistence following single-dose and homologous booster vaccination with Ad26.COV2.S in different subgroups of recipients stratified by age and sex.