A prediction method for the individual serum concentration and therapeutic effect for optimizing adalimumab therapy in inflammatory bowel disease.

OBJECTIVES: Adalimumab (ADM) therapy is effective for inflammatory bowel disease (IBD), but a significant number of IBD patients lose response to ADM. Thus, it is crucial to devise methods to enhance ADM's effectiveness. This study introduces a strategy to predict individual serum concentrations and therapeutic effects to optimize ADM therapy for IBD during the induction phase. METHODS: We predicted the individual serum concentration and therapeutic effect of ADM during the induction phase based on pharmacokinetic and pharmacodynamic (PK/PD) parameters calculated using the empirical Bayesian method. We then examined whether the predicted therapeutic effect, defined as clinical remission or treatment failure, matched the observed effect. RESULTS: Data were obtained from 11 IBD patients. The therapeutic effect during maintenance therapy was successfully predicted at 40 of 47 time points. Moreover, the predicted effects at each patient's final time point matched the observed effects in 9 of the 11 patients. CONCLUSION: This is the inaugural report predicting the individual serum concentration and therapeutic effect of ADM using the Bayesian method and PK/PD modelling during the induction phase. This strategy may aid in optimizing ADM therapy for IBD.

Mixed effect estimation in deep compartment models: Variational methods outperform first-order approximations.

This work focusses on extending the deep compartment model (DCM) framework to the estimation of mixed-effects. By introducing random effects, model predictions can be personalized based on drug measurements, enabling the testing of different treatment schedules on an individual basis. The performance of classical first-order (FO and FOCE) and machine learning based variational inference (VI) algorithms were compared in a simulation study. In VI, posterior distributions of the random variables are approximated using variational distributions whose parameters can be directly optimized. We found that variational approximations estimated using the path derivative gradient estimator version of VI were highly accurate. Models fit on the simulated data set using the FO and VI objective functions gave similar results, with accurate predictions of both the population parameters and covariate effects. Contrastingly, models fit using FOCE depicted erratic behaviour during optimization, and resulting parameter estimates were inaccurate. Finally, we compared the performance of the methods on two real-world data sets of haemophilia A patients who received standard half-life factor VIII concentrates during prophylactic and perioperative settings. Again, models fit using FO and VI depicted similar results, although some models fit using FO presented divergent results. Again, models fit using FOCE were unstable. In conclusion, we show that mixed-effects estimation using the DCM is feasible. VI performs conditional estimation, which might lead to more accurate results in more complex models compared to the FO method.

Population pharmacokinetics of TLD-1, a novel liposomal doxorubicin, in a phase I trial.

STUDY OBJECTIVES: TLD-1 is a novel pegylated liposomal doxorubicin (PLD) formulation aiming to optimise the PLD efficacy-toxicity ratio. We aimed to characterise TLD-1's population pharmacokinetics using non-compartmental analysis and nonlinear mixed-effects modelling. METHODS: The PK of TLD-1 was analysed by performing a non-compartmental analysis of longitudinal doxorubicin plasma concentration measurements obtained from a clinical trial in 30 patients with advanced solid tumours across a 4.5-fold dose range. Furthermore, a joint parent-metabolite PK model of doxorubicinentrapped, doxorubicinfree, and metabolite doxorubicinol was developed. Interindividual and interoccasion variability around the typical PK parameters and potential covariates to explain parts of this variability were explored. RESULTS: Medians  ± standard deviations of dose-normalised doxorubicinentrapped+free Cmax and AUC0-∞ were 0.342 ± 0.134 mg/L and 40.1 ± 18.9 mg·h/L, respectively. The median half-life (95 h) was 23.5 h longer than the half-life of currently marketed PLD. The novel joint parent-metabolite model comprised a one-compartment model with linear release (doxorubicinentrapped), a two-compartment model with linear elimination (doxorubicinfree), and a one-compartment model with linear elimination for doxorubicinol. Body surface area on the volumes of distribution for free doxorubicin was the only significant covariate. CONCLUSION: The population PK of TLD-1, including its release and main metabolite, were successfully characterised using non-compartmental and compartmental analyses. Based on its long half-life, TLD-1 presents a promising candidate for further clinical development. The PK characteristics form the basis to investigate TLD-1 exposure-response (i.e., clinical efficacy) and exposure-toxicity relationships in the future. Once such relationships have been established, the developed population PK model can be further used in model-informed precision dosing strategies. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov-NCT03387917-January 2, 2018.

Model-Based Interspecies Scaling for Predicting Human Pharmacokinetics of CB 4332, a Complement Factor I Protein.

Interspecies scaling of the pharmacokinetics (PK) of CB 4332, a 150 kDa recombinant complement factor I protein, was performed using traditional and model-based approaches to inform first-in-human dose selection. Plasma concentration versus time data from four preclinical PK studies of single intravenous and subcutaneous (SC) CB 4332 dosing in mice, rats and nonhuman primates (NHPs) were modeled simultaneously using naive pooling including allometric scaling. The human-equivalent dose was calculated using the preclinical no observed adverse effect level (NOAEL) as part of the dose-by-factor approach. Pharmacokinetic modeling of CB 4332 revealed species-specific differences in the elimination, which was accounted for by including an additional rat-specific clearance. Signs of anti-drug antibodies (ADA) formation in all rats and some NHPs were observed. Consequently, an additional ADA-induced clearance parameter was estimated including the time of onset. The traditional dose-by-factor approach calculated a maximum recommended starting SC dose of 0.9 mg/kg once weekly, which was predicted it to result in a trough steady-state concentration lower than the determined efficacy target range for CB 4332 in humans. Model simulations predicted the efficacy target range to be reached using 5 mg/kg once weekly SC dosing.

Determining tissue distribution of the oral antileishmanial agent miltefosine: a physiologically-based pharmacokinetic modeling approach.

Miltefosine (MTS) is the only approved oral drug for treating leishmaniasis caused by intracellular Leishmania parasites that localize in macrophages of the liver, spleen, skin, bone marrow, and lymph nodes. MTS is extensively distributed in tissues and has prolonged elimination half-lives due to its high plasma protein binding, slow metabolic clearance, and minimal urinary excretion. Thus, understanding and predicting the tissue distribution of MTS help assess therapeutic and toxicologic outcomes of MTS, especially in special populations, e.g., pediatrics. In this study, a whole-body physiologically-based pharmacokinetic (PBPK) model of MTS was built on mice and extrapolated to rats and humans. MTS plasma and tissue concentration data obtained by intravenous and oral administration to mice were fitted simultaneously to estimate model parameters. The resulting high tissue-to-plasma partition coefficient values corroborate extensive distribution in all major organs except the bone marrow. Sensitivity analysis suggests that plasma exposure is most susceptible to changes in fraction unbound in plasma. The murine oral-PBPK model was further validated by assessing overlay of simulations with plasma and tissue profiles obtained from an independent study. Subsequently, the murine PBPK model was extrapolated to rats and humans based on species-specific physiological and drug-related parameters, as well as allometrically scaled parameters. Fold errors for pharmacokinetic parameters were within acceptable range in both extrapolated models, except for a slight underprediction in the human plasma exposure. These animal and human PBPK models are expected to provide reliable estimates of MTS tissue distribution and assist dose regimen optimization in special populations.

Synergizing pharmacometrics and pharmacovigilance for medication error management: the case of secukinumab.

PURPOSE: To demonstrate the effective integration of pharmacometrics and pharmacovigilance in managing medication errors, highlighted by a case involving secukinumab in a patient with hidradenitis suppurativa. METHODS: We present the case of a 41-year-old male with progressive hidradenitis suppurativa, unresponsive to multiple antibiotic regimens and infliximab treatment. Due to a medication error, the patient received 300 mg of secukinumab daily for 4 days instead of weekly, totaling 1200 mg. The regional pharmacovigilance center assessed potential toxicity, and a pharmacometric analysis using a population pharmacokinetic model was performed to inform dosing adjustments. RESULTS: Clinical data indicated that the received doses were within a non-toxic range. No adverse effects were observed. Pharmacometric simulations revealed a risk of underexposure due to the dosing error. Based on these simulations, it was recommended to restart monthly secukinumab injections on day 35 after the initial dose. Measured plasma concentrations before re-administration confirmed the model's accuracy. CONCLUSION: This case highlights the crucial collaboration between clinical services, pharmacovigilance, and pharmacometrics in managing medication errors. Such interdisciplinary efforts ensure therapeutic efficacy and patient safety by maintaining appropriate drug exposure levels.

Simulating Healthy Participant Pharmacokinetics for Renal and Hepatic Impairment Studies: Retrospective Assessment of the Approach.

The recruitment of a parallel, healthy participants (HPs) arm in renal and hepatic impairment (RI and HI) studies is a common strategy to assess differences in pharmacokinetics. Limitations in this approach include the underpowered estimate of exposure differences and the use of the drug in a population for which there is no benefit. Recently, a method was published by Purohit et. al. (2023) that leveraged prior population pharmacokinetic (PopPK) modeling-based simulation to infer the distribution of exposure ratios between the RI/HI arms and HPs. The approach was successful, but it was a single example with a robust model having several iterations of development and fitting to extensive HP data. To test in more studies and models at different stages of development, our catalogue of RI/HI studies was searched, and those with suitable properties and from programs with available models were analyzed with the simulation approach. There were 9 studies included in the analysis. Most studies were associated with models that would have been available at the time (ATT) of the study, and all had a current, final model. For 3 studies, the HP PK was not predicted well by the ATT (2) or final (1) models. In comparison to conventional analysis of variance (ANOVA), the simulation approach provided similar point estimates and confidence intervals of exposure ratios. This PopPK based approach can be considered as a method of choice in situations where the simulation of HP data would not be an extrapolation, and when no other complicating factors are present.

Predicting overall survival from tumor dynamics metrics using parametric statistical and machine learning models: application to patients with RET-altered solid tumors.

In oncology drug development, tumor dynamics modeling is widely applied to predict patients' overall survival (OS) via parametric models. However, the current modeling paradigm, which assumes a disease-specific link between tumor dynamics and survival, has its limitations. This is particularly evident in drug development scenarios where the clinical trial under consideration contains patients with tumor types for which there is little to no prior institutional data. In this work, we propose the use of a pan-indication solid tumor machine learning (ML) approach whereby all three tumor metrics (tumor shrinkage rate, tumor regrowth rate and time to tumor growth) are simultaneously used to predict patients' OS in a tumor type independent manner. We demonstrate the utility of this approach in a clinical trial of cancer patients treated with the tyrosine kinase inhibitor, pralsetinib. We compared the parametric and ML models and the results showed that the proposed ML approach is able to adequately predict patient OS across RET-altered solid tumors, including non-small cell lung cancer, medullary thyroid cancer as well as other solid tumors. While the findings of this study are promising, further research is needed for evaluating the generalizability of the ML model to other solid tumor types.

Application of machine learning techniques in population pharmacokinetics/pharmacodynamics modeling.

Population pharmacokinetics/pharmacodynamics (pop-PK/PD) consolidates pharmacokinetic and pharmacodynamic data from many subjects to understand inter- and intra-individual variability due to patient backgrounds, including disease state and genetics. The typical workflow in pop-PK/PD analysis involves the determination of the structure model, selection of the error model, analysis based on the base model, covariate modeling, and validation of the final model. Machine learning is gaining considerable attention in the medical and various fields because, in contrast to traditional modeling, which often assumes linear or predefined relationships, machine learning modeling learns directly from data and accommodates complex patterns. Machine learning has demonstrated excellent capabilities for prescreening covariates and developing predictive models. This review introduces various applications of machine learning techniques in pop-PK/PD research.

ChatGPT and Gemini large language models for pharmacometrics with NONMEM: comment.

This is a correspondence on "Evaluation of ChatGPT and Gemini large language models for pharmacometrics with NONMEM". Additional concern on using ChatGPT and Gemini is provided.

Guiding Dyslipidemia Treatment: A Population Pharmacokinetic-Pharmacodynamic Framework for Obicetrapib.

Obicetrapib is a selective inhibitor of cholesteryl ester transfer protein that is currently in phase 3 of development for the treatment of dyslipidemia as adjunct therapy. The purpose of this study was to comprehensively characterize the pharmacokinetic (PK) and pharmacodynamic (PD) disposition of obicetrapib. Data from 7 clinical trials conducted in healthy adults and those with varying degrees of dyslipidemia were included for model development. The structural model that best described obicetrapib PK was a 3-compartment model with 4-compartment transit absorption and first-order elimination. Body weight was the only covariate found to significantly explain observed variability and was therefore included using allometric scaling on all disposition parameters. For a typical patient weighing 75 kg, the estimated apparent total body clearance and apparent volume of distribution of the central compartment was 0.81 L/h and 36.1 L, respectively. The final PK model parameters were estimated with good precision and were ultimately leveraged to sequentially inform 2 turnover models that describe obicetrapib's effect on low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C) concentrations. The maximum stimulatory effect of obicetrapib on LDL-C loss was estimated to be 1.046, while the maximum inhibitory effect of obicetrapib on HDL-C loss was 0.691. This corresponds to a predicted typical maximum percent change from baseline LDL-C and HDL-C of 51.1% and 224%, respectively. The final sequential model described obicetrapib PKPD well and was ultimately able to both demonstrate evidence of internal consistency and support decision-making throughout the development lifecycle.

Machine-Learning Assisted Screening of Correlated Covariates: Application to Clinical Data of Desipramine.

Stepwise covariate modeling (SCM) has a high computational burden and can select the wrong covariates. Machine learning (ML) has been proposed as a screening tool to improve the efficiency of covariate selection, but little is known about how to apply ML on actual clinical data. First, we simulated datasets based on clinical data to compare the performance of various ML and traditional pharmacometrics (PMX) techniques with and without accounting for highly-correlated covariates. This simulation step identified the ML algorithm and the number of top covariates to select when using the actual clinical data. A previously developed desipramine population-pharmacokinetic model was used to simulate virtual subjects. Fifteen covariates were considered with four having an effect included. Based on the F1 score (an accuracy measure), ridge regression was the most accurate ML technique on 200 simulated datasets (F1 score = 0.475 ± 0.231), a performance which almost doubled when highly-correlated covariates were accounted for (F1 score = 0.860 ± 0.158). These performances were better than forwards selection with SCM (F1 score = 0.251 ± 0.274 and 0.499 ± 0.381 without/with correlations respectively). In terms of computational cost, ridge regression (0.42 ± 0.07 seconds/simulated dataset, 1 thread) was ~20,000 times faster than SCM (2.30 ± 2.29 hours, 15 threads). On the clinical dataset, prescreening with the selected ML algorithm reduced SCM runtime by 42.86% (from 1.75 to 1.00 days) and produced the same final model as SCM only. In conclusion, we have demonstrated that accounting for highly-correlated covariates improves ML prescreening accuracy. The choice of ML method and the proportion of important covariates (unknown a priori) can be guided by simulations.

PmWebSpec: An Application to Create and Manage CDISC-Compliant Pharmacometric Analysis Dataset Specifications.

A well-documented pharmacometric (PMx) analysis dataset specification ensures consistency in derivations of the variables, naming conventions, traceability to the source data, and reproducibility of the analysis dataset. Lack of standards in creating the dataset specification can lead to poor quality analysis datasets, negatively impacting the quality of the PMx analysis. Standardization of the dataset specification within an individual organization helps address some of these inconsistencies. The recent introduction of the Clinical Data Interchange Standards Consortium (CDISC) Analysis Data Model (ADaM) Population Pharmacokinetic (popPK) Implementation Guide (IG) further promotes industry-wide standards by providing guidelines for the basic data structure of popPK analysis datasets. However, manual implementation of the standards can be labor intensive and error-prone. Hence, there is still a need to automate the implementation of these standards. In this paper, we present PmWebSpec, an easily deployable web-based application to facilitate the creation and management of CDISC-compliant PMx analysis dataset specifications. We describe the application of this tool through examples and highlight its key features including pre-populated dataset specifications, built-in checks to enforce standards, and generation of an electronic Common Technical Document (eCTD)-compliant data definition file. The application increases efficiency, quality and semi-automates PMx analysis dataset, and specification creation and has been well accepted by pharmacometricians and programmers internally. The success of this application suggests its potential for broader usage across the PMx community.

Evaluation of ChatGPT and Gemini large language models for pharmacometrics with NONMEM.

To assess ChatGPT 4.0 (ChatGPT) and Gemini Ultra 1.0 (Gemini) large language models on NONMEM coding tasks relevant to pharmacometrics and clinical pharmacology. ChatGPT and Gemini were assessed on tasks mimicking real-world applications of NONMEM. The tasks ranged from providing a curriculum for learning NONMEM, an overview of NONMEM code structure to generating code. Prompts in lay language to elicit NONMEM code for a linear pharmacokinetic (PK) model with oral administration and a more complex model with two parallel first-order absorption mechanisms were investigated. Reproducibility and the impact of "temperature" hyperparameter settings were assessed. The code was reviewed by two NONMEM experts. ChatGPT and Gemini provided NONMEM curriculum structures combining foundational knowledge with advanced concepts (e.g., covariate modeling and Bayesian approaches) and practical skills including NONMEM code structure and syntax. ChatGPT provided an informative summary of the NONMEM control stream structure and outlined the key NONMEM Translator (NM-TRAN) records needed. ChatGPT and Gemini were able to generate code blocks for the NONMEM control stream from the lay language prompts for the two coding tasks. The control streams contained focal structural and syntax errors that required revision before they could be executed without errors and warnings. The code output from ChatGPT and Gemini was not reproducible, and varying the temperature hyperparameter did not reduce the errors and omissions substantively. Large language models may be useful in pharmacometrics for efficiently generating an initial coding template for modeling projects. However, the output can contain errors and omissions that require correction.

Pharmacokinetic Modeling of Bepotastine for Determination of Optimal Dosage Regimen in Pediatric Patients with Allergic Rhinitis or Urticaria.

Bepotastine, a second-generation antihistamine for allergic rhinitis and urticaria, is widely used in all age groups but lacks appropriate dosing guidelines for pediatric patients, leading to off-label prescriptions. We conducted this study to propose an optimal dosing regimen for pediatric patients based on population pharmacokinetic (popPK) and physiologically based pharmacokinetic (PBPK) models using data from two previous trials. A popPK model was built using NONMEM software. A one-compartment model with first-order absorption and absorption lag time described our data well, with body weight incorporated as the only covariate. A PBPK model was developed using PK-Sim software version 10, and the model well predicted the drug concentrations obtained from pediatric patients. Furthermore, the final PBPK model showed good concordance with the known properties of bepotastine. Appropriate pediatric doses for different weight and age groups were proposed based on the simulations. Discrepancies in recommended doses from the two models were likely due to the incorporation of age-dependent physiological factors in the PBPK model. In conclusion, our study is the first to suggest an optimal oral dosing regimen of bepotastine in pediatric patients using both approaches. This is expected to foster safer and more productive use of the drug.

Risk Factors for COVID-19 and Respiratory Tract Infections during the Coronavirus Pandemic.

(1) Background: Some individuals are more susceptible to developing respiratory tract infections (RTIs) or coronavirus disease (COVID-19) than others. The aim of this work was to identify risk factors for symptomatic RTIs including COVID-19 and symptomatic COVID-19 during the coronavirus pandemic by using infection incidence, participant baseline, and regional COVID-19 burden data. (2) Methods: Data from a prospective study of 1000 frontline healthcare workers randomized to Bacillus Calmette-Guérin vaccination or placebo, and followed for one year, was analyzed. Parametric time-to-event analysis was performed to identify the risk factors associated with (a) non-specific symptomatic respiratory tract infections including COVID-19 (RTIs+COVID-19) and (b) symptomatic RTIs confirmed as COVID-19 using a polymerase chain reaction or antigen test (COVID-19). (3) Results: Job description of doctor or nurse (median hazard ratio [HR] 1.541 and 95% confidence interval [CI] 1.299-1.822), the reported COVID-19 burden (median HR 1.361 and 95% CI 1.260-1.469 for 1.4 COVID-19 cases per 10,000 capita), or a BMI > 30 kg/m2 (median HR 1.238 and 95% CI 1.132-1.336 for BMI of 35.4 kg/m2) increased the probability of RTIs+COVID-19, while positive SARS-CoV-2 serology at enrollment (median HR 0.583 and 95% CI 0.449-0.764) had the opposite effect. The reported COVID-19 burden (median HR 2.372 and 95% CI 2.116-2.662 for 1.4 COVID-19 cases per 10,000 capita) and a job description of doctor or nurse (median HR 1.679 and 95% CI 1.253-2.256) increased the probability of developing COVID-19, while smoking (median HR 0.428 and 95% CI 0.284-0.648) and positive SARS-CoV-2 serology at enrollment (median HR 0.076 and 95% CI 0.026-0.212) decreased it. (4) Conclusions: Nurses and doctors with obesity had the highest probability of developing RTIs including COVID-19. Non-smoking nurses and doctors had the highest probability of developing COVID-19 specifically. The reported COVID-19 burden increased the event probability, while positive SARS-CoV-2 IgG serology at enrollment decreased the probability of RTIs including COVID-19, and COVID-19 specifically.

Pharmacokinetics-Based Pediatric Dose Evaluation and Optimization Using Saliva - A Case Study.

Understanding pharmacokinetics (PK) in children is a prerequisite to determine optimal pediatric dosing. As plasma sampling in children is challenging, alternative PK sampling strategies are needed. In this case study we evaluated the suitability of saliva as alternative PK matrix to simplify studies in infants, investigating metamizole, an analgesic used off-label in infants. Six plasma and 6 saliva PK sample collections were scheduled after a single intravenous dose of 10 mg/kg metamizole. Plasma/saliva pharmacometric (PMX) modeling of the active metabolites 4-methylaminoantipyrine (4-MAA) and 4-aminoantipyrine (4-AA) was performed. Various reduced plasma sampling scenarios were evaluated by PMX simulations. Saliva and plasma samples from 25 children were included (age range, 5-70 months; weight range, 8.7-24.8 kg). Distribution of metamizole metabolites between plasma and saliva was without delay. Estimated mean (individual range) saliva/plasma fractions of 4-MAA and 4-AA were 0.32 (0.05-0.57) and 0.57 (0.25-0.70), respectively. Residual variability of 4-MAA (4-AA) in saliva was 47% (28%) versus 17% (11%) in plasma. A simplified sampling scenario with up to 6 saliva samples combined with 1 plasma sample was associated with similar PK parameter estimates as the full plasma sampling scenario. This case study with metamizole shows increased PK variability in saliva compared to plasma, compromising its suitability as single matrix for PK studies in infants. Nonetheless, rich saliva sampling can reduce the number of plasma samples required for PK characterization, thereby facilitating the conduct of PK studies to optimize dosing in pediatric patients.

Predictive performance of population pharmacokinetic models of imatinib in chronic myeloid leukemia patients.

BACKGROUND AND AIM: Chronic myeloid leukemia is a myeloproliferative neoplasm associated with the specific chromosomal translocation known as the Philadelphia chromosome. Imatinib is a potent BCR-ABL tyrosine kinase inhibitor, which is approved as the first line therapy for CML patients. There are various population pharmacokinetic studies available in the literature for this population. However, their use in other populations outside of their cohort for the model development has not been evaluated. This study was aimed to perform the predictive performance of the published population pharmacokinetic models for imatinib in CML population and propose a dosing nomogram. METHODS: A systematic review was conducted through PubMed, and WoS databases to identify PopPK models. Clinical data collected in adult CML patients treated with imatinib was used for evaluation of these models. Various prediction-based metrics were used for assessing the bias and precision of PopPK models using individual predictions. RESULTS: Eight imatinib PopPK model were selected for evaluating the model performance. A total of 145 plasma imatinib samples were collected from 43 adult patients diagnosed with CML and treated with imatinib. The PopPK model reported by Menon et al. had better performance than all other PopPK models. CONCLUSION: Menon et al. model was able to predict well for our clinical data where it had the relative mean prediction error percentage ≤ 20%, relative median absolute prediction error ≤ 30% and relative root mean square error close to zero. Based on this final model, we proposed a dosing nomogram for various weight groups, which could potentially help to maintain the trough concentrations in the therapeutic range.