Prior information for population pharmacokinetic and pharmacokinetic/pharmacodynamic analysis: overview and guidance with a focus on the NONMEM PRIOR subroutine

Tutorial on Conditional Simulations With a Tumor Size-Overall Survival Model to Support Oncology Drug Development

The gold standard for regulatory approval in oncology is overall survival (OS). Because OS data are initially limited, early drug development decisions are often based on early efficacy endpoints, such as objective response rate and progression-free survival. Tumor size (TS)-OS models provide a framework to support decision-making on potential late-stage success based on early readouts, through leveraging TS data with limited follow-up and treatment-agnostic TS-OS link functions, to predict longer-term OS. Conditional simulations (also known as Bayesian forecasting) with TS-OS models can be used to simulate long-term OS outcomes for an ongoing study, conditional on the available TS and OS data at interim data cuts of the same study. This tutorial provides a comprehensive overview of the steps involved in using such conditional simulations to support better informed drug development decisions in oncology. The tutorial covers the selection of the TS-OS framework model; applying the TS-OS model to the interim data; performing conditional simulations; generating relevant output; as well as correct interpretation and communication of the output for decision making.

Palbociclib exposure in relation to efficacy and toxicity in patients with advanced breast cancer

BACKGROUND

Data on exposure-response or exposure-toxicity relationships of cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) are limited and inconclusive. We aimed to investigate whether there is an association between palbociclib exposure and progression-free survival (PFS), adverse events (AEs) and dose reductions.

MATERIALS AND METHODS

Data were retrieved from the prospective, multicentre SONIA trial in which patients with advanced estrogen receptor-positive, human epidermal growth factor receptor 2-negative breast cancer were randomised to receive CDK4/6i treatment in first versus second line. Blood for pharmacokinetics (PK) was taken at day 15 of cycles 1 and 2 during CDK4/6i treatment. Individual trough concentrations and plasma area under the curves of palbociclib were constructed using a population PK model. Associations with palbociclib exposure were tested using Cox regression for PFS and chi-square tests for AEs or dose reductions.

RESULTS

PK data were available for 344 patients. No association between palbociclib exposure and PFS was found. Although patients with higher palbociclib exposure had more dose reductions during their entire CDK4/6i treatment course, this was not reflected by a higher incidence of grade 3-4 AEs in the first 3 months.

CONCLUSION

The absence of an association between palbociclib exposure and PFS and the presence of the association between palbociclib exposure and dose reductions suggest that dose reductions may safely be carried out in case of palbociclib-related toxicity.

Implementing a Bayesian approach using Stan with Torsten: Population pharmacokinetics analysis of somatrogon

Fully Bayesian approaches are not commonly implemented for population pharmacokinetic (PK) modeling. In this paper, we evaluate the use of Stan with R and Torsten for population PK modeling of somatrogon, a recombinant long-acting growth hormone approved for the treatment of growth hormone deficiency. As a software for Bayesian inference, Stan provides an easy way to conduct MCMC sampling for a wide range of models with efficient sampling algorithms, and there are several diagnostic tools to evaluate the MCMC convergence and other potential issues. Three different sets of priors were evaluated for estimation and prediction: a weakly informative uniform set, a moderately informative set, and a very informative set of priors. All three prior sets showed good performance and all chains mixed well. There were some minor differences in the final parameter posterior distributions while implementing different prior sets, but the posterior predictions covered the observations nicely, not only for the individuals included in posterior sampling but also for new individuals. The impact of a centered versus non-centered parameterization were evaluated, with the non-centered approach improving the estimation time, but it was still computationally intensive. Computational resources had the biggest impact on sampling time. Stan took approximately 2.5 h total for the MCMC sampling on a high-performance computing platform (6 cores) and may be reduced further with additional computational resources. The model and comparisons presented show that with adequate computational resources, the Bayesian approaches using Stan and Torsten are useful for population PK analysis, especially for the analysis of special populations, small sample datasets, and when complex model structures are needed.

Bayesian Population Pharmacokinetic Modeling of Ondansetron for Neonatal Opioid Withdrawal Syndrome

Ondansetron is an anti-emetic 5-HT3 receptor antagonist being investigated for treating neonatal opioid withdrawal syndrome (NOWS). Sparse PK data were analyzed from a multicenter, double-blind clinical trial with 98 mother/neonate dyads. Pregnant women with opioid use disorder were randomized to receive either placebo or ondansetron 8 mg intravenously within 4 h of delivery. Neonates born to mothers who were randomized to ondansetron received 0.07 mg/kg orally once every 24 h for up to five doses. Using current PK data, model parameters from a two-compartmental structural model from the literature (i.e., a priori model) were updated with the Metropolis-Hastings Markov-chain Monte Carlo estimation algorithm in NONMEM. The updated Bayesian model indicated a slower absorption rate (KA) but no differences in model parameters (CL, V, V2, Q) after including body weight and postmenstrual age. Sensitivity analyses on CL prior revealed statistical improvement favoring larger body weights, but not changes in postmenstrual age. However, further model development using larger body weights did not illustrate superior performance through visual inspection of diagnostic plots. Overall, a cumulative AUC of at least 1000 ng*h/mL appears to be the threshold for reductions in symptom severity. Exposure-response analyses suggest the total number of doses to be the primary driver for efficacy with respect to AUC, which reasonably aligns with the literature. Overall, it is suggested that at least three doses of the current oral ondansetron regimen are required to reduce symptom severity in neonates.

Population pharmacokinetics and dosing simulations of meropenem in septic critically ill patients with complicated intra-abdominal infection or pneumonia

OBJECTIVES

Meropenem pharmacokinetics (PK) may be altered in septic critically ill patients with complicated intra-abdominal infections (cIAI) and pneumonia. We aimed to evaluate the covariates affecting meropenem PK and the performance of different dosing regimens to optimize the PK/pharmacodynamic target.

METHODS

Population PK analysis was performed using non-linear mixed-effects modeling. The final model was validated and used to simulate meropenem exposure to assess the probability of attaining the 100 %ƒT>MIC target.

RESULTS

Forty-six and 14 patients were respectively enrolled for PK analysis and external validation. A one-compartment linear model adequately described the data of 226 concentrations. The typical clearance (CL) and volume of distribution (Vd) were 9.69 L/h and 27.4 L, respectively. Septic shock from cIAI (cIASS) and actual body weight were significant covariates for meropenem Vd in addition to the influential covariates of creatinine clearance (CLCR-CG) and augmented renal clearance for CL. External validation showed the robustness and accuracy of this model. Simulation results proposed continuous infusion (CI) dosing regimens of meropenem against pathogens with MICs ≥ 2 mg/L in patients with cIASS and CLCR-CG ≥ 60 mL/min.

CONCLUSIONS

For the patients with cIASS and CLCR-CG ≥ 60 mL/min, CI meropenem is proposed for treatment of less sensitive pathogens with MICs ≥ 2 mg/L.

Within-chain parallelization-Giving Stan Jet Fuel for population modeling in pharmacometrics

Stan is a powerful probabilistic programming language designed mainly for Bayesian data analysis. Torsten is a collection of Stan functions that handles the events (e.g., dosing events) and solves the ODE systems that are frequently present in pharmacometric models. To perform a Bayesian data analysis, most models in pharmacometrics require Markov Chain Monte Carlo (MCMC) methods to sample from the posterior distribution. However, MCMC is computationally expensive and can be time-consuming, enough so that people will often forgo Bayesian methods for a more traditional approach. This paper shows how to speed up the sampling process in Stan by within-chain parallelization through both multi-threading using Stan's reduce_sum() function and multi-processing using Torsten's group ODE solver. Both methods show substantial reductions in the time necessary to sufficiently sample from the posterior distribution compared with a basic approach with no within-chain parallelization.

Vedolizumab Clearance as a Surrogate Marker for Remission in Inflammatory Bowel Disease Patients: Insights from Real-World Pharmacokinetics

Background/Objectives: Vedolizumab (VDZ) is approved in the treatment of patients with moderate to severe ulcerative colitis (UC) or Crohn's disease (CD). VDZ exhibits considerable variability in its pharmacokinetic (PK) profile, and its exposure-response relationship is not yet fully understood. The aim was to investigate the variability in VDZ trough levels and PK parameters, to assess the relationship between VDZ PK and biochemical response, as well as clinical and endoscopic outcomes. Methods: We included 61 UC and 45 CD patients. Patients' data and trough VDZ concentrations were retrospectively obtained. Population PK analysis was performed using non-linear mixed-effects modelling with NONMEM (version 7.5). Graphs and statistical analyses were performed using R (version 4.1.3). Results: In total, 116 trough VDZ concentrations from 106 patients were described by a two-compartment model. For a typical patient, clearance (CL) was estimated at 0.159 L/day, while in patients previously treated with anti-TNFα agents, VDZ CL increased by 26.4% on average. In univariate binary logistic regression, VDZ trough concentration was not statistically significant predictor of remission, whereas CL was. Moreover, combined CL and faecal calprotectin (FCP) were a statistically significant predictors of remission. The hazard ratio (HR) for CL above 0.1886 L/day was 0.35 (p = 0.05) and for FCP below 250 µg/g was 2.66 (p = 0.02) in a time-to-event analysis. Conclusions: Our population PK model incorporates the effect of prior anti-TNFα agents on CL, suggesting its association with more severe forms of IBD. VDZ CL emerged as a more robust and clinically relevant predictor of remission in IBD patients than trough concentration.

Population PK modeling of certolizumab pegol in pregnant women with chronic inflammatory diseases

Certolizumab pegol (CZP; CIMZIA™) is the only Fc-free tumor necrosis factor inhibitor with data from a clinical study demonstrating no to minimal placental transfer. The pharmacokinetics (PK) of certolizumab pegol during pregnancy and postpartum in women with chronic inflammatory diseases were assessed using a population PK model based on data from the CHERISH study (NCT04163016), a longitudinal, prospective, open-label PK phase IB study. Model development was performed in NONMEM using a frequentist prior approach, with prior information based on a population PK model for certolizumab pegol in non-pregnant adult patients (NCT04740814). A one-compartment model with first-order absorption (Ka = 0.236 1/day) and linear elimination (CL/F = 0.416 L/day) from the central compartment (V/F = 7.86 L) best described certolizumab pegol PK in the CHERISH study. The structural model parameters were estimated with good precision (RSE < 25%). Baseline BW was included as a covariate on CL/F and V/F. Pregnancy trimester and time-varying log-transformed anti-drug antibody (ADA) titer were identified as the only significant covariates for CL/F with a comparable influence on CL/F. Individuals with higher ADA titer (75th percentile) during pregnancy exhibited CL/F up to 1.43-fold higher relative to individuals postpartum that showed median levels of ADA titer. However, the confidence interval for the combined effect of pregnancy stage and ADA titer effects on CL/F overlapped with the CL/F range of the typical individual postpartum. In addition, simulations showed a large overlap in certolizumab pegol concentrations between pregnant and non-pregnant adults. The findings of this population PK analysis support the maintenance of established certolizumab pegol dosing regimens throughout pregnancy.

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.

The Influence of Subclinical Active Inflammation on IFX Pharmacokinetic Modeling and Disease Progression Assessment: Findings from a Prospective Real-World Study in Inflammatory Bowel Disease Patients

BACKGROUND AND AIMS

Effective management of inflammatory bowel disease (IBD) relies on a comprehensive understanding of infliximab (IFX) pharmacokinetics (PK). This study's primary goal was to develop a robust PK model, identifying key covariates influencing IFX clearance (CL), while concurrently evaluating the risk of disease progression during the maintenance phase of IBD treatment.

METHODS

The multicenter, prospective, real-world DIRECT study was conducted in several care centers, which included 369 IBD patients in the maintenance phase of IFX therapy. A two-compartment population PK model was used to determine IFX CL and covariates. Logistic and Cox regressions were applied to elucidate the associations between disease progression and covariates embedded in the PK model.

RESULTS

The PK model included the contributions of weight, albumin, antidrug antibody (ADA), and fecal calprotectin (FC). On average, higher ADA, FC concentration and weight, and lower albumin concentration resulted in higher IFX CL. In the multivariate regression analyses, FC levels influenced the odds of disease progression in the majority of its definitions, when adjusted for several confounding factors. Additionally, alongside FC, both IFX and CL demonstrated a significant impact on the temporal aspect of disease progression.

CONCLUSION

In this 2-year real-world study, readily available clinical covariates, notably FC, significantly impacted IFX availability in IBD patients. We demonstrated that subclinical active inflammation, as mirrored by FC or CRP, substantially influenced IFX clearance. Importantly, FC emerged as a pivotal determinant, not only of IFX pharmacokinetics but also of disease progression. These findings underscore the need to integrate FC into forthcoming IFX pharmacokinetic models, amplifying its clinical significance.

Isavuconazole Pharmacokinetics in Critically Ill Patients: Relationship with Clinical Effectiveness and Patient Safety

Isavuconazole is used to treat fungal infections. This study aims to describe isavuconazole pharmacokinetics in critically ill patients and evaluate their relationship with clinical efficacy and patient safety. We conducted a prospective, observational study in patients treated with intravenous isavuconazole. Samples were collected at predose (Cmin), 1 h (Cmax) and 12 h (C50) after the last dose. The plasma concentration was determined by high-performance liquid chromatography. The relationship between plasma concentration and clinical and microbiological outcomes and safety was evaluated. The influence of covariates (age, sex, weight, SAPS3, creatinine, liver enzymes and extracorporeal devices: continuous renal replacement therapy (CRRT) and extracorporeal membrane oxygenation (ECMO)) was analysed. Population pharmacokinetic modelling was performed using NONMEN®. A total of 71 isavuconazole samples from 24 patients were analysed. The mean Cmin was 1.76 (1.02) mg/L; 87.5% reached the optimal therapeutic target and 12.5% were below 1 mg/L. Population pharmacokinetics were best described by a one-compartment model with first-order elimination. No factor had a significant impact on the plasma concentration or pharmacokinetic parameters. Thus, isavuconazole could be safely used in a critically ill population, even in those treated with CRRT and ECMO, from a pharmacokinetic standpoint. Therefore, routine therapeutic drug monitoring may not be strictly necessary in daily clinical practice.

Semi-physiological Pharmacokinetic Model of Clozapine and Norclozapine in Healthy, Non-smoking Volunteers: The Impact of Race and Genetics

BACKGROUND AND OBJECTIVES: Clozapine is the medication of choice for treatment-resistant schizophrenia. However, it has a complex metabolism and unexplained interindividual variability. The current work aims to develop a pharmacokinetic model of clozapine and norclozapine in non-smokers and assess the impact of demographic and genetic predictors.

METHODS

Healthy volunteers were recruited in a population pharmacokinetic study. Blood samples were collected at 30 min and 1, 2, 3, 5 and 8 h following a single flat dose of clozapine (12.5 mg). The clozapine and norclozapine concentrations were measured via high-performance liquid chromatography-ultraviolet method. A semi-physiological pharmacokinetic model of clozapine and norclozapine was developed using nonlinear mixed-effects modeling. Clinical and genetic predictors were evaluated, including CYP1A2 (rs762551) and ABCB1 (rs2032582), using restriction fragment length polymorphism.

RESULTS

A total of 270 samples were collected from 33 participants. The data were best described using a two-compartment model for clozapine and a two-compartment model for norclozapine with first-order absorption and elimination and pre-systemic metabolism. The estimated (relative standard error) clearance of clozapine and norclozapine were 27 L h-1 (31.5 %) and 19.6 L h-1 (30%), respectively. Clozapine clearance was lower in sub-Saharan Africans (n = 4) and higher in Caucasians (n = 9) than Asians (n = 20). Participants with CYP1A2 (rs762551) (n = 18) and ABCB1 (rs2032582) (n = 12) mutant alleles had lower clozapine clearance in the univariate analysis.

CONCLUSIONS

This is the first study to develop a semi-physiological pharmacokinetic model of clozapine and norclozapine accounting for the pre-systemic metabolism. Asians required lower doses of clozapine as compared with Caucasians, while clozapine pharmacokinetics in sub-Saharan Africans should be further investigated in larger trials.

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.

Practical Applications of a Nausea and Vomiting Model in the Clinical Development of Additional Doses of Dulaglutide

Dulaglutide 3.0 and 4.5 mg weekly doses were approved for additional glycemic control in adult patients with type 2 diabetes inadequately controlled with metformin and 0.75 or 1.5 mg weekly doses of dulaglutide. Effects such as nausea and vomiting are commonly reported with dulaglutide and other glucagon-like peptide-1 receptor agonist therapies. Based on a pharmacokinetic/pharmacodynamic model-informed approach, a stepwise dose-escalation scheme with 4-week intervals between dose increments was suggested to mitigate gastrointestinal events for dulaglutide. These gastrointestinal events are dose dependent and attenuate over time with repeated dosing. A Markov chain Monte Carlo pharmacokinetic/pharmacodynamic joint model was developed using AWARD-11 data (N = 1842) to optimize dulaglutide dose escalation to 3.0 and 4.5 mg to mitigate gastrointestinal events. Model simulations evaluated probabilities of nausea and vomiting events for various dosing scenarios in patients needing higher doses for additional glycemic control. The model indicated that patients may dose escalate from 1.5 to 3.0 mg, then 4.5 mg weekly after at least 4 weeks on each dose. No clinically meaningful differences in nausea or vomiting events were expected when patients escalated to 3.0 or 4.5 mg following initiation at 0.75 or 1.5 mg dulaglutide. Based on the findings of this model, a minimum 4-week duration at each dose before escalation was appropriate to reduce gastrointestinal events of dulaglutide, consistent with observed gastrointestinal events data from the AWARD-11 study and supporting the currently recommended dose-escalation regimen of dulaglutide doses of 3.0 and 4.5 mg for additional glycemic control.

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.

Evolution of Ritlecitinib Population Pharmacokinetic Models During Clinical Drug Development

BACKGROUND

Ritlecitinib is an oral Janus kinase 3/tyrosine kinase expressed in hepatocellular carcinoma family inhibitor undergoing parallel clinical development for alopecia areata, vitiligo, ulcerative colitis, Crohn's disease, and rheumatoid arthritis.

OBJECTIVE

As studies read out simultaneously, strategic planning of population pharmacokinetic model development and evaluation is required to ensure timely decisions.

METHODS

Data from healthy participants and patients from 12 clinical trials between December 2014 and July 2021 were included: seven phase I studies in healthy participants and organ impairment, five phase II/III studies in patients with rheumatoid arthritis, ulcerative colitis, alopecia areata, and vitiligo. Population pharmacokinetic models consisted of stepwise procedures to accommodate data availability and the model's application to answering clinical development questions. At each iteration of the model update, parameters of the next model were re-estimated by leveraging previous information and new data.

RESULTS

Three model development lifecycle iterations of the ritlecitinib population pharmacokinetic model were conducted to support alopecia areata, vitiligo, and ulcerative colitis study readouts. Initial structural modeling based on healthy participant data (and some rheumatoid arthritis and alopecia areata data) in iteration 1 provided a platform for comprehensive covariate testing during iteration 2, and model evaluation and implementation of the frequentist prior approach in iteration 3. The final model was a two-compartment model with first-order absorption and direct-response non-stationary clearance and bioavailability driven by concentrations in the peripheral compartment.

CONCLUSIONS

The present approach demonstrated the evolution of three population pharmacokinetic models with accumulating data, addressed clinical drug development questions related to systemic exposures of ritlecitinib, and informed the approved product label.

CLINICAL TRIAL REGISTRATION

NCT02309827, NCT02684760, NCT02958865, NCT02969044, NCT03232905, NCT03732807, NCT04016077, NCT03715829, NCT04037865, NCT04004663, NCT04634565, NCT02974868.

Population pharmacokinetics of ivermectin after mass drug administration in lymphatic filariasis endemic communities of Tanzania

Ivermectin (IVM) is a drug of choice used with albendazole for mass drug administration (MDA) to halt transmission of lymphatic filariasis. We investigated IVM pharmacokinetic (PK) variability for its dose optimization during MDA. PK samples were collected at 0, 2, 4, and 6 h from individuals weighing greater than 15 kg (n = 468) receiving IVM (3-, 6-, 9-, or 12 mg) and ALB (400 mg) during an MDA campaign in Tanzania. Individual characteristics, including demographics, laboratory/clinical parameters, and pharmacogenetic variations were assessed. IVM plasma concentrations were quantified by liquid-chromatography tandem mass spectrometry and analyzed using population-(PopPK) modeling. A two-compartment model with transit absorption kinetics, and allometrically scaled oral clearance (CL/F) and central volume (Vc /F) was adapted. Fitting of the model to the data identified 48% higher bioavailability for the 3 mg dose compared to higher doses and identified a subpopulation with 97% higher mean transit time (MTT). The final estimates for CL/F, Vc /F, intercompartment clearance, peripheral volume, MTT, and absorption rate constant for a 70 kg person (on dose other than 3 mg) were 7.7 L/h, 147 L, 20.4 L/h, 207 L, 1.5 h, and 0.71/h, respectively. Monte-Carlo simulations indicated that weight-based dosing provides comparable exposure across weight bands, but height-based dosing with capping IVM dose at 12 mg for individuals with height greater than 160 cm underdoses those weighing greater than 70 kg. Variability in IVM PKs is partly explained by body weight and dose. The established PopPK model can be used for IVM dose optimization. Height-based pole dosing results in varying IVM exposure in different weight bands, hence using weighing scales for IVM dosing during MDA is recommended.

Impact of single-nucleotide variants and nutritional status on population pharmacokinetics of Doxorubicin, and its effect on cardiotoxicity in children with leukemia

PURPOSE

 Doxorubicin is an important antineoplastic agent with wide interindividual variability in response to treatment and in its cardiotoxic effects. To determine the effect of genotypic status of three single-nucleotide variants in ABCC1, NCF4, and CBR3 genes and nutritional status assessed by body mass index, on the population pharmacokinetics of Doxorubicin and its cardiotoxic effects in pediatric patients with leukemia.

PATIENTS AND METHODS

Seventy pediatric patients treated with Doxorubicin were studied, in which 189 biological samples were obtained to determine Doxorubicin concentrations (1 to 3 samples per patient) at different times, for 20 h.

RESULTS

Low body mass index and age ≤ 7 years were associated with decreased clearance of Doxorubicin, and female gender was associated with increased clearance of Doxorubicin. Low BMI and low height were associated with a decrease and increase, respectively, in the intercompartmental clearance (Q) of Doxorubicin. TT homozygosity of the single-nucleotide variant rs3743527 of the ABCC1 gene was associated with an increase in clearance and decreased area under the curve, AA homozygosity of the single-nucleotide variant rs1883112 of the NCF4 gene was associated with a decrease in the volume of distribution in the peripheral compartment (V2), and GG homozygosity of CBR3 rs1056892 with increasing area under the curve.

CONCLUSION

Some covariates studied are directly related to the increase or decrease of the pharmacokinetic parameters of Doxorubicin. Decreased clearance, V2, and increased area under the curve were associated with systolic dysfunction, and decreased Q and V2 were associated with diastolic dysfunction. These results may contribute to the effective and safe use of Doxorubicin in pediatric patients with leukemia.

Predicting cytokine kinetics during sepsis; a modelling framework from a porcine sepsis model with live Escherichia coli

BACKGROUND

Describing the kinetics of cytokines involved as biomarkers of sepsis progression could help to optimise interventions in septic patients. This work aimed to quantitively characterise the cytokine kinetics upon exposure to live E. coli by developing an in silico model, and to explore predicted cytokine kinetics at different bacterial exposure scenarios.

METHODS

Data from published in vivo studies using a porcine sepsis model were analysed. A model describing the time courses of bacterial dynamics, endotoxin (ETX) release, and the kinetics of TNF and IL-6 was developed. The model structure was extended from a published model that quantifies the ETX-cytokines relationship. An external model evaluation was conducted by applying the model to literature data. Model simulations were performed to explore the sensitivity of the host response towards differences in the input rate of bacteria, while keeping the total bacterial burden constant.

RESULTS

The analysis included 645 observations from 30 animals. The blood bacterial count was well described by a one-compartment model with linear elimination. A scaling factor was estimated to quantify the ETX release by bacteria. The model successfully described the profiles of TNF, and IL-6 without a need to modify the ETX-cytokines model structure. The kinetics of TNF, and IL-6 in the external datasets were well predicted. According to the simulations, the ETX tolerance development results in that low initial input rates of bacteria trigger the lowest cytokine release.

CONCLUSION

The model quantitively described and predicted the cytokine kinetics triggered by E. coli exposure. The host response was found to be sensitive to the bacterial exposure rate given the same total bacterial burden.

External validation of population pharmacokinetic models of gentamicin in paediatric population from preterm newborns to adolescents

The aim of this study was to externally validate the predictive performance of published population pharmacokinetic models of gentamicin in all paediatric age groups, from preterm newborns to adolescents. We first selected published population pharmacokinetic models of gentamicin developed in the paediatric population with a wide age range. The parameters of the literature models were then re-estimated using the PRIOR subroutine in NONMEM^®. The predictive ability of the literature and the tweaked models was evaluated. Retrospectively collected data from a routine clinical practice (512 concentrations from 308 patients) were used for validation. The models with covariates characterising developmental changes in clearance and volume of distribution had better predictive performance, which improved further after re-estimation. The tweaked model by Wang 2019 performed best, with suitable accuracy and precision across the complete paediatric population. For patients treated in the intensive care unit, a lower proportion of patients would be expected to reach the target trough concentration at standard dosing. The selected model could be used for model-informed precision dosing in clinical settings where the entire paediatric population is treated. However, for use in clinical practice, the next step should include additional analysis of the impact of intensive care treatment on gentamicin pharmacokinetics, followed by prospective validation.

Clinical Pharmacokinetics of Radiopharmaceuticals from SPECT/CT Image Acquisition by Contouring in Patients with Gastroenteropancreatic Neuroendocrine Tumors: Lu-177 DOTATATE (Lutathera®) Case

BACKGROUND AND OBJECTIVE

Lu-177 DOTATATE (Lutathera^®) is a radiolabeled analog of somatostatin administered intravenously in patients with somatostatin receptor-positive gastroenteropancreatic neuroendocrine tumors. Biodistribution of Lu-177 DOTATATE in tumor and healthy tissues can be monitored by serial post-injection scintigraphy imaging. Patient exposure to the drug is variable with the recommended fixed dosage, and hence there is a variable response to treatment. The aim of this work was to study the pharmacokinetics of Lu-177 DOTATATE by a population modeling approach, based on single-photon emission computed tomography (SPECT)/computed tomography (CT) images used as surrogate of plasma concentrations to study the interindividual variability and finally optimize an individual dosage.

METHODS

From a retrospective study, SPECT/CT images were acquired at 4 h, 24 h, 72 h, and 192 h postadministration. From these images, volumic activities were calculated in blood and bone marrow. An individual non-compartmental pharmacokinetic analysis was performed, and the mean pharmacokinetic parameters of each tissue were compared together and with reference data. Blood volumic activities were then used to perform a population pharmacokinetic analysis (NONMEM).

RESULTS

The pharmacokinetic parameters (non-compartmental analysis) obtained from blood (clearance [CL] = 2.65 L/h, volume of distribution at steady state [Vss] = 309 L, elimination half-life [t_1/2] = 86.3 h) and bone marrow (CL =1.68 L/h, Vss = 233 L, t_1/2 = 98.8 h) were statistically different from each other and from reference values (CL = 4.50 L/h, Vss = 460 L, t_1/2 = 71.0 h) published in the literature. SPECT/CT blood images were used as a surrogate of plasma concentrations to develop a population pharmacokinetic model. Weight was identified as covariate on volume of the central compartment, reducing the interindividual variability of all population pharmacokinetic parameters.

CONCLUSION

This study is a proof of concept that obtaining pharmacokinetic parameters with image-based blood concentration is possible. Obtaining observed concentrations from SPECT/CT images, without the need for blood sampling, is a real advantage for the patient and the drug monitoring. Pharmacokinetic modeling could be combined with a deep learning model for automatic contouring and allow precise patient-specific dose adjustment in a non-invasive manner.

External evaluation and systematic review of population pharmacokinetic models for high-dose methotrexate in cancer patients

Several population pharmacokinetic (PPK) models have been established to optimize the therapeutic regimen and reduce the toxicity of high-dose methotrexate (HDMTX) in patients with cancer. However, their predictive performance when extrapolated to different clinical centers was unknown. In this study, we aimed to externally evaluate the predictive ability of HDMTX PPK models and determine the potential influencing factors. We searched the literature and determined the predictive performance of the selected models using methotrexate concentrations in 721 samples from 60 patients in the First Affiliated Hospital of the Navy Medical University. Prediction-based diagnostics and simulation-based normalized prediction distribution errors (NPDE) were used to evaluate the predictive performance of the models. The influence of prior information was also assessed using Bayesian forecasting, and the potential factors affecting model predictability were investigated. Thirty models extracted from published PPK studies were assessed. Prediction-based diagnostics showed that the number of compartments potentially influenced model transferability, and simulation-based NPDE indicated model misspecification. Bayesian forecasting significantly improved the predictive performance of the models. Various factors, including bioassays, covariates, and population diagnosis, influence model extrapolation. The published models were unsatisfactory for all prediction-based diagnostics, except for the 24 h methotrexate concentration monitoring and simulation-based diagnostics, making them inappropriate for direct extrapolation. Moreover, Bayesian forecasting combined therapeutic drug monitoring could improve the predictive performance of the models.

Population Pharmacokinetic-Pharmacodynamic Modeling and Probability of Target Attainment Analysis of Rocuronium and Sugammadex in Children Undergoing Surgery

BACKGROUND AND OBJECTIVES

Probability of target attainment (PTA) curves are commonly used to support dose recommendations of antibiotics for different patient groups. In this study we propose PTA analysis to optimize sugammadex dosing in children.

METHODS

This study involved data from an observational cohort study of 30 American Society of Anesthesiologists (ASA) Physical Status I and II children undergoing surgery requiring muscle relaxation. All patients received 0.6 mg/kg rocuronium, with sugammadex administered at the end of surgery in three different doses (0.5, 1.0, and 2.0 mg/kg) to reverse the neuromuscular blockade.

RESULTS

The data were analyzed using a population Bayesian-based approach. The developed model was used to simulate pharmacokinetic-pharmacodynamic profiles for different patient groups and dosing regimens before the PTA analysis was performed to translate these simulations into a clinically useful measure. The target was defined as neuromuscular blockade reversal measured by Train-of-Four (TOF ratio > 90%) at 1.5, 3, and 5 min post sugammadex dose. The sugammadex doses leading to 90% PTA were determined for different patients' body weights, rocuronium doses, and time gaps between rocuronium and sugammadex administration assuming the model, priors, and gathered data. For comparison, PTA curves for a range of clinical scenarios are provided to illustrate the usefulness of PTA analysis in selecting the appropriate dose for a given patient.

CONCLUSIONS

The proposed PTA analysis is useful to support the sugammadex dose selection in different clinical scenarios.

TRIAL REGISTRATION

The study was registered by ClinicalTrials.gov under number NCT04851574 on 21 April 2021.

Optimization of trial duration to predict long-term HbA1c change with therapy: A pharmacometrics simulation-based evaluation

Glycated hemoglobin (HbA1c) is the main biomarker of diabetes drug development. However, because of its delayed turnover, trial duration is rarely shorter than 12 weeks, and being able to predict long-term HbA1c with precision using data from shorter studies would be beneficial. The feasibility of reducing study duration was therefore investigated in this study, assuming a model-based analysis. The aim was to investigate the predictive performance of 24- and 52-week extrapolations using data from up to 4, 6, 8 or 12 weeks, with six previously published pharmacometric models of HbA1c. Predictive performance was assessed through simulation-based dose-response predictions and model averaging (MA) with two hypothetical drugs. Results were consistent across the methods of assessment, with MA supporting the results derived from the model-based framework. The models using mean plasma glucose (MPG) or nonlinear fasting plasma glucose (FPG) effect, driving the HbA1c formation, showed good predictive performance despite a reduced study duration. The models, using the linear effect of FPG to drive the HbA1c formation, were sensitive to the limited amount of data in the shorter studies. The MA with bootstrap demonstrated strongly that a 4-week study duration is insufficient for precise predictions of all models. Our findings suggest that if data are analyzed with a pharmacometric model with MPG or FPG with a nonlinear effect to drive HbA1c formation, a study duration of 8 weeks is sufficient with maintained accuracy and precision of dose-response predictions.

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

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

Improved Confidence in a Confirmatory Stage by Application of Item-Based Pharmacometrics Model: Illustration with a Phase III Active Comparator-Controlled Trial in COPD Patients

PURPOSE

The current study aimed to illustrate how a non-linear mixed effect (NLME) model-based analysis may improve confidence in a Phase III trial through more precise estimates of the drug effect.

METHODS

The FULFIL clinical trial was a Phase III study that compared 24 weeks of once daily inhaled triple therapy with twice daily inhaled dual therapy in patients with chronic obstructive pulmonary disease (COPD). Patient reported outcome data, obtained by using The Evaluating Respiratory Symptoms in COPD (E-RS:COPD) questionnaire, from the FULFIL study were analyzed using an NLME item-based response theory model (IRT). The change from baseline (CFB) in E-RS:COPD total score over 4-week intervals for each treatment arm was obtained using the IRT and compared with published results obtained with a mixed model repeated measures (MMRM) analysis.

RESULTS

The IRT included a graded response model characterizing item parameters and a Weibull function combined with an offset function to describe the COPD symptoms-time course in patients receiving either triple therapy (n = 907) or dual therapy (n = 894). The IRT improved precision of the estimated drug effect compared to MMRM, resulting in a sample size of at least 3.64 times larger for the MMRM analysis to achieve the IRT precision in the CFB estimate.

CONCLUSION

This study shows the advantage of IRT over MMRM with a direct comparison of the same primary endpoint for the two analyses using the same observed clinical trial data, resulting in an increased confidence in Phase III.

Model-Predicted Impact of ECG Monitoring Strategies During Bedaquiline Treatment

Background

The M2 metabolite of bedaquiline causes QT-interval prolongation, making electrocardiogram (ECG) monitoring of patients receiving bedaquiline for drug-resistant tuberculosis necessary. The objective of this study was to determine the relationship between M2 exposure and Fridericia-corrected QT (QTcF)-interval prolongation and to explore suitable ECG monitoring strategies for 6-month bedaquiline treatment.

Methods

Data from the PROBeX study, a prospective observational cohort study, were used to characterize the relationship between M2 exposure and QTcF. Established nonlinear mixed-effects models were fitted to pharmacokinetic and ECG data. In a virtual patient population, QTcF values were simulated for scenarios with and without concomitant clofazimine. ECG monitoring strategies to identify patients who need to interrupt treatment (QTcF > 500 ms) were explored.

Results

One hundred seventy patients were included, providing 1131 bedaquiline/M2 plasma concentrations and 1702 QTcF measurements; 2.1% of virtual patients receiving concomitant clofazimine had QTcF > 500 ms at any point during treatment (0.7% without concomitant clofazimine). With monthly monitoring, almost all patients with QTcF > 500 ms were identified by week 12; after week 12, patients were predominantly falsely identified as QTcF > 500 ms due to stochastic measurement error. Following a strategy with monitoring before treatment and at weeks 2, 4, 8, and 12 in simulations with concomitant clofazimine, 93.8% of all patients who should interrupt treatment were identified, and 26.4% of all interruptions were unnecessary (92.1% and 32.2%, respectively, without concomitant clofazimine).

Conclusions

Our simulations enable an informed decision for a suitable ECG monitoring strategy by weighing the risk of missing patients with QTcF > 500 ms and that of interrupting bedaquiline treatment unnecessarily. We propose ECG monitoring before treatment and at weeks 2, 4, 8, and 12 after starting bedaquiline treatment.

Avelumab Dose Selection for Clinical Studies in Pediatric Patients with Solid Tumors

BACKGROUND AND OBJECTIVE

A phase I/II trial evaluated the safety, antitumor activity, and pharmacokinetics of avelumab (anti-PD-L1 antibody) in pediatric patients with refractory/relapsed solid tumors (NCT03451825). This study aimed to inform avelumab dose selection in pediatric populations using population pharmacokinetic modeling and simulations.

METHODS

Patients aged < 18 years with refractory/relapsed solid tumors enrolled in phase I received avelumab 10 or 20 mg/kg intravenously every 2 weeks. A pediatric population pharmacokinetic model was developed via the frequentist prior approach.

RESULTS

Pharmacokinetic parameters from 21 patients who received avelumab 10 mg/kg (n = 6) or 20 mg/kg (n = 15) were analyzed. Patients had a wide range of weights and ages (medians, 37.3 kg and 12 years). Exposures with 10-mg/kg dosing were lower vs adult dosing, particularly in patients weighing < 40 kg, whereas 20-mg/kg dosing achieved or exceeded adult exposures, irrespective of body weight. A two-compartment linear model with time-varying clearance using body weight as a covariate, with the frequentist prior approach, best described pediatric data. In this model, optimal overlap in exposure with adult data was achieved with 800 mg every 2 weeks for patients aged ≥ 12 years and weighing ≥ 40 kg, and 15 mg/kg every 2 weeks for patients aged < 12 years or weighing < 40 kg.

CONCLUSIONS

Based on exposure matching, the recommended doses for further avelumab studies, including combination studies, are 15 mg/kg every 2 weeks for pediatric patients aged < 12 years or weighing < 40 kg and the adult flat dose of 800 mg every 2 weeks for pediatric patients aged ≥ 12 years and weighing ≥ 40 kg.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov NCT03451825.

Pharmacokinetic equations versus Bayesian guided vancomycin monitoring: Pharmacokinetic model and model-informed precision dosing trial simulations

The recently released revised vancomycin consensus guideline endorsed area under the concentration-time curve (AUC) guided monitoring. Means to AUC-guided monitoring include pharmacokinetic (PK) equations and Bayesian software programs, with the latter approach being preferable. We aimed to evaluate the predictive performance of these two methods when monitoring using troughs or peaks and troughs at varying single or mixed dosing intervals (DIs), and evaluate the significance of satisfying underlying assumptions of steady-state and model transferability. Methods included developing a vancomycin population PK model and conducting model-informed precision dosing clinical trial simulations. A one-compartment PK model with linear elimination, exponential between-subject variability, and mixed (additive and proportional) residual error model resulted in the best model fit. Conducted simulations demonstrated that Bayesian-guided AUC can, potentially, outperform that of equation-based AUC predictions depending on the quality of model diagnostics and met assumptions. Ideally, Bayesian-guided AUC predictive performance using a trough from the first DI was equivalent to that of PK equations using two measurements (peak and trough) from the fifth DI. Model transferability diagnostics can guide the selection of Bayesian priors but are not strong indicators of predictive performance. Mixed versus single fourth and/or fifth DI sampling seems indifferent. This study illustrated cases associated with the most reliable AUC predictions and showed that only proper Bayesian-guided monitoring is always faster and more reliable than equations-guided monitoring in pre-steady-state DIs in the absence of a loading dose. This supports rapid Bayesian monitoring using data as sparse and early as a trough at the first DI.

Novel strategy to personalise use of ibuprofen for closure of patent ductus arteriosus in preterm neonates

OBJECTIVE

Exploration of a novel therapeutic drug monitoring (TDM) strategy to personalise use of ibuprofen for closure of patent ductus arteriosus (PDA) in preterm neonates.

DESIGN

Prospective, single-centre, open-label, pharmacokinetics study in preterm neonates.

SETTING

Neonatal intensive care unit at McMaster Children's Hospital.

PATIENTS

Neonates with a gestational age ≤28⁺⁶ weeks treated with oral ibuprofen for closure of a PDA.

METHODS

Population pharmacokinetic parameters, concentration-time profiles and exposure metrics were obtained using pharmacometric modelling and simulation.

MAIN OUTCOME MEASURE

Association between ibuprofen plasma concentrations measured at various sampling time points on the first day of treatment and attainment of the target exposure over the first 3 days of treatment (AUC_0-72h >900 mg·hour/L).

RESULTS

Twenty-three preterm neonates (median birth weight 780 g and gestational age 25.9 weeks) were included, yielding 155 plasma ibuprofen plasma samples. Starting from 8 hours' postdose on the first day, a strong correlation between ibuprofen concentrations and AUC_0-72h was observed. At 8 hours after the first dose, an ibuprofen concentration >20.5 mg/L was associated with a 90% probability of reaching the target exposure.

CONCLUSION

We designed a novel and practical TDM strategy and have shown that the chance of reaching the target exposure (AUC_0-72h >900 mg·hour/L) can be predicted with a single sample collection on the first day of treatment. This newly acquired knowledge can be leveraged to personalise ibuprofen dosing regimens and improve the efficacy of ibuprofen use for pharmacological closure of a PDA.

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

AIMS

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

METHODS

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

RESULTS

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

CONCLUSIONS

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

Model-Based Exposure-Response Assessment for Spectinamide 1810 in a Mouse Model of Tuberculosis

Despite decades of research, tuberculosis remains a leading cause of death from a single infectious agent. Spectinamides are a promising novel class of antituberculosis agents, and the lead spectinamide 1810 has demonstrated excellent efficacy, safety, and drug-like properties in numerous in vitro and in vivo assessments in mouse models of tuberculosis. In the current dose ranging and dose fractionation study, we used 29 different combinations of dose level and dosing frequency to characterize the exposure-response relationship for spectinamide 1810 in a mouse model of Mycobacterium tuberculosis infection and in healthy animals. The obtained data on 1810 plasma concentrations and counts of CFU in lungs were analyzed using a population pharmacokinetic/pharmacodynamic (PK/PD) approach as well as classical anti-infective PK/PD indices. The analysis results indicate that there was no difference in the PK of 1810 in infected compared to healthy, uninfected animals. The PK/PD index analysis showed that bacterial killing of 1810 in mice was best predicted by the ratio of maximum free drug concentration to MIC (fC_max/MIC) and the ratio of the area under the free concentration-time curve to the MIC (fAUC/MIC) rather than the cumulative percentage of time that the free drug concentration is above the MIC (f%T_MIC). A novel PK/PD model with consideration of postantibiotic effect could adequately describe the exposure-response relationship for 1810 and supports the notion that the in vitro observed postantibiotic effect of this spectinamide also translates to the in vivo situation in mice. The obtained results and pharmacometric model for the exposure-response relationship of 1810 provide a rational basis for dose selection in future efficacy studies of this compound against M. tuberculosis.

Upadacitinib pharmacokinetics and exposure-response analyses of efficacy and safety in psoriatic arthritis patients - Analyses of phase III clinical trials

Upadacitinib is an oral Janus kinase inhibitor approved for the treatment of rheumatoid arthritis (RA) and recently approved by the European Medicines Agency for the treatment of psoriatic arthritis (PsA). The efficacy and safety profile of upadacitinib in PsA have been established in the SELECT‐PsA program in two global phase III studies, which evaluated upadacitinib 15 and 30 mg q.d. The analyses described here characterized upadacitinib pharmacokinetics and exposure‐response relationships for efficacy and safety endpoints using data from the SELECT‐PsA studies. Upadacitinib pharmacokinetics in patients with PsA were characterized through a Bayesian population analysis approach and were comparable to pharmacokinetics in patients with RA. Exposure‐response relationships for key efficacy and safety endpoints were characterized using data from 1916 patients with PsA. The percentage of patients achieving efficacy endpoints at week 12 (American College of Rheumatology [ACR]50 and ACR70), 16 and 24 (sIGA0/1) increased with increasing upadacitinib average plasma concentration over a dosing interval, whereas no clear exposure‐response trend was observed for ACR20 at week 12 or ACR20/50/70 at week 24 within the range of plasma exposures evaluated in the phase III PsA studies. No clear trends for exposure‐response relationships were identified for experiencing pneumonia, herpes zoster infection, hemoglobin less than 8 g/dl, lymphopenia (grade ≥ 3), or neutropenia (grade ≥ 3) after 24 weeks of treatment. Shallow relationships with plasma exposures were observed for serious infections and hemoglobin decrease greater than 2 g/dl from baseline at week 24. Based on exposure‐response analyses, the upadacitinib 15 mg q.d. regimen is predicted to achieve robust efficacy in patients with PsA and to be associated with limited incidences of reductions in hemoglobin or occurrence of serious infections.

Population Pharmacokinetics and Pharmacodynamics of Meropenem in Critically Ill Patients: How to Achieve Best Dosage Regimen According to the Clinical Situation

BACKGROUND AND OBJECTIVES

Meropenem is frequently used for the treatment of severe bacterial infections in critically ill patients. Because critically ill patients are more prone to pharmacokinetic variability than other patients, ensuring an effective blood concentration can be complex. Therefore, describing this variability to ensure a proper use of this antibiotic drug limits the rise and dissemination of antimicrobial resistance, and helps preserve the current antibiotic arsenal. The aims of this study were to describe the pharmacokinetics of meropenem in critically ill patients, to identify and quantify the patients' characteristics responsible for the observed pharmacokinetic variability, and to perform different dosing simulations in order to determine optimal individually adapted dosing regimens.

METHODS

A total of 58 patients hospitalized in the medical intensive care unit and receiving meropenem were enrolled, including 26 patients with renal replacement therapy. A population pharmacokinetic model was developed (using NONMEM software) and Monte Carlo simulations were performed with different dosing scenarios (bolus-like, extended, and continuous infusion) exploring the impact of clinical categories of residual diuresis (anuria, oliguria, and preserved diuresis) on the probability of target attainment (MIC: 1-45 mg/L).

RESULTS

The population pharmacokinetic model included five covariates with a significant impact on clearance: glomerular filtration rate, dialysis (continuous and semi-continuous), renal function status, and volume of residual diuresis. The clearance for a typical patient in our population is 4.20 L/h and volume of distribution approximately 44 L. Performed dosing regimen simulations suggested that, for equivalent doses, the continuous infusion mode (with loading dose) allowed the obtaining of the pharmacokinetic/pharmacodynamic target for a larger number of patients (100% for MIC ≤ 20 mg/L). Nevertheless, for the treatment of susceptible bacteria (MIC ≤ 2 mg/L), differences in the probability of target attainment between bolus-like, extended, and continuous infusions were negligible.

CONCLUSIONS

Identified covariates in the model are easily accessible information in patient health records. The model highlighted the importance of considering the patient's overall condition (renal function and dialysis) and the pathogen's characteristics (MIC target) during the establishment of a patient's dosing regimen.

Can We Predict Individual Concentrations of Tacrolimus After Liver Transplantation? Application and Tweaking of a Published Population Pharmacokinetic Model in Clinical Practice

BACKGROUND

Various population pharmacokinetic models have been developed to describe the pharmacokinetics of tacrolimus in adult liver transplantation. However, their extrapolated predictive performance remains unclear in clinical practice. The purpose of this study was to predict concentrations using a selected literature model and to improve these predictions by tweaking the model with a subset of the target population.

METHODS

A literature review was conducted to select an adequate population pharmacokinetic model (L). Pharmacokinetic data from therapeutic drug monitoring of tacrolimus in liver-transplanted adults were retrospectively collected. A subset of these data (70%) was exploited to tweak the L-model using the $PRIOR subroutine of the NONMEM software, with 2 strategies to weight the prior information: full informative (F) and optimized (O). An external evaluation was performed on the remaining data; bias and imprecision were evaluated for predictions a priori and Bayesian forecasting.

RESULTS

Seventy-nine patients (851 concentrations) were enrolled in the study. The predictive performance of L-model was insufficient for a priori predictions, whereas it was acceptable with Bayesian forecasting, from the third prediction (ie, with ≥2 previously observed concentrations), corresponding to 1 week after transplantation. Overall, the tweaked models showed a better predictive ability than the L-model. The bias of a priori predictions was -41% with the literature model versus -28.5% and -8.73% with tweaked F and O models, respectively. The imprecision was 45.4% with the literature model versus 38.0% and 39.2% with tweaked F and O models, respectively. For Bayesian predictions, whatever the forecasting state, the tweaked models tend to obtain better results.

CONCLUSIONS

A pharmacokinetic model can be used, and to improve the predictive performance, tweaking the literature model with the $PRIOR approach allows to obtain better predictions.

The Role of Mathematical Models in Immuno-Oncology: Challenges and Future Perspectives

Immuno-oncology (IO) focuses on the ability of the immune system to detect and eliminate cancer cells. Since the approval of the first immune checkpoint inhibitor, immunotherapies have become a major player in oncology treatment and, in 2021, represented the highest number of approved drugs in the field. In spite of this, there is still a fraction of patients that do not respond to these therapies and develop resistance mechanisms. In this sense, mathematical models offer an opportunity to identify predictive biomarkers, optimal dosing schedules and rational combinations to maximize clinical response. This work aims to outline the main therapeutic targets in IO and to provide a description of the different mathematical approaches (top-down, middle-out, and bottom-up) integrating the cancer immunity cycle with immunotherapeutic agents in clinical scenarios. Among the different strategies, middle-out models, which combine both theoretical and evidence-based description of tumor growth and immunological cell-type dynamics, represent an optimal framework to evaluate new IO strategies.

Improved Decision-Making Confidence Using Item-Based Pharmacometric Model: Illustration with a Phase II Placebo-Controlled Trial

This study aimed to illustrate how a new methodology to assess clinical trial outcome measures using a longitudinal item response theory–based model (IRM) could serve as an alternative to mixed model repeated measures (MMRM). Data from the EXACT (Exacerbation of chronic pulmonary disease tool) which is used to capture frequency, severity, and duration of exacerbations in COPD were analyzed using an IRM. The IRM included a graded response model characterizing item parameters and functions describing symptom-time course. Total scores were simulated (month 12) using uncertainty in parameter estimates. The 50th (2.5th, 97.5th) percentiles of the resulting simulated differences in average total score (drug minus placebo) represented the estimated drug effect (95%CI), which was compared with published MMRM results. Furthermore, differences in sample size, sensitivity, specificity, and type I and II errors between approaches were explored. Patients received either oral danirixin 75 mg twice daily (n = 45) or placebo (n = 48) on top of standard of care over 52 weeks. A step function best described the COPD symptoms-time course in both trial arms. The IRM improved precision of the estimated drug effect compared to MMRM, resulting in a sample size of 2.5 times larger for the MMRM analysis to achieve the IRM precision. The IRM showed a higher probability of a positive predictive value (34%) than MMRM (22%). An item model–based analysis data gave more precise estimates of drug effect than MMRM analysis for the same endpoint in this one case study.

Model-Informed Therapeutic Drug Monitoring of Meropenem in Critically Ill Patients: Improvement of the Predictive Ability of Literature Models with the PRIOR Approach

BACKGROUND AND OBJECTIVE

To improve the predictive ability of literature models for model-informed therapeutic drug monitoring (TDM) of meropenem in intensive care units, we propose to tweak the literature models with the "prior approach" using a subset of the data. This study compares the predictive ability of both literature and tweaked models on TDM concentrations of meropenem in critically ill patients.

METHODS

Blood samples were collected from patients of an intensive care unit treated with intravenous meropenem. Data were split six times into an "estimation" and a "prediction" datasets. Population pharmacokinetic (popPK) models of meropenem were selected from literature. These models were run on the "estimation" dataset with the $PRIOR subroutine in NONMEM to obtain tweaked models. The literature and tweaked models were used a priori (with covariate only) and with Bayesian fitting to predict each individual concentration from the previous concentration(s). Their respective predictive abilities were compared using median relative prediction error (MDPE%) and median absolute relative prediction error (MDAPE%).

RESULTS

The total dataset was composed of 115 concentrations from 58 patients. For each of the six splits, the "estimation" and the "prediction" datasets were respectively composed of 44 and 14 patients or 45 and 13 patients. Six popPK models were selected in the literature. MDPE% and MDAPE% were globally lower for the tweaked than for the literature models, especially for a priori predictions.

CONCLUSION

The "prior approach" could be a valuable tool to improve the predictive ability of literature models, especially for a priori predictions, which are important to optimize dosing in emergency situations.

Population-based meta-analysis of chloroquine: informing chloroquine pharmacokinetics in COVID-19 patients

AIMS

Chloroquine (CQ) has been repurposed to treat coronavirus disease 2019 (COVID-19). Understanding the pharmacokinetics (PK) in COVID-19 patients is essential to study its exposure-efficacy/safety relationship and provide a basis for a possible dosing regimen optimization.

SUBJECT AND METHODS

In this study, we used a population-based meta-analysis approach to develop a population PK model to characterize the CQ PK in COVID-19 patients. An open-label, single-center study (ethical review approval number: PJ-NBEY-KY-2020-063-01) was conducted to assess the safety, efficacy, and pharmacokinetics of CQ in patients with COVID-19. The sparse PK data from 50 COVID-19 patients, receiving 500 mg CQ phosphate twice daily for 7 days, were combined with additional CQ PK data from 18 publications.

RESULTS

A two-compartment model with first-order oral absorption and first-order elimination and an absorption lag best described the data. Absorption rate (ka) was estimated to be 0.559 h-1, and a lag time of absorption (ALAG) was estimated to be 0.149 h. Apparent clearance (CL/F) and apparent central volume of distribution (V2/F) was 33.3 l/h and 3630 l. Apparent distribution clearance (Q/F) and volume of distribution of peripheral compartment (Q3/F) were 58.7 l/h and 5120 l. The simulated CQ concentration under five dosing regimens of CQ phosphate were within the safety margin (400 ng/ml).

CONCLUSION

Model-based simulation using PK parameters from the COVID-19 patients shows that the concentrations under the currently recommended dosing regimen are below the safety margin for side-effects, which suggests that these dosing regimens are generally safe. The derived population PK model should allow for the assessment of pharmacokinetics-pharmacodynamics (PK-PD) relationships for CQ when given alone or in combination with other agents to treat COVID-19.