Application of optimal design methodologies in clinical pharmacology experiments

In silico modeling and simulation of organ-on-a-chip systems to support data analysis and a priori experimental design

Organ-on-a-chip (OoC) systems are a promising new class of in vitro devices that can combine various tissues, cultured in different compartments, linked by media flow. The properties of these novel in vitro systems linked to increased physiological relevance of culture conditions may lead to more in vivo-relevant cell phenotypes, enabling better in vitro pharmacology and toxicology assessment. Improved cell activities combined with longer lasting cultures offer opportunities to improve the characterization of absorption, distribution, metabolism, and excretion (ADME) processes, potentially leading to more accurate prediction of human pharmacokinetics (PKs). The inclusion of barrier tissue elements and metabolically competent tissue types results in complex concentration-time profiles (in vitro PK) for test drugs and their metabolites that require appropriate mathematical modeling of in vitro data for parameter estimation. In particular, modeling is critical to estimate in vitro ADME parameters when multiple different tissues are combined in a single device. Therefore, sophisticated in silico data analysis and a priori experimental design are highly recommended for OoC experiments in a manner not needed with standard ADME screening. The design of the experiment should be optimized based on an investigation of the structural characteristics of the in vitro system, the ADME features of the test compound and any available knowledge of cell phenotypes. This tutorial aims to provide such a modeling framework to inform experimental design and refine parameter estimation in a Gut-Liver OoC (the most studied multi-organ systems to predict the oral drug PKs) to improve translatability of data generated in such complex cellular systems.

Streamlining tablet lubrication design via model-based design of experiments

In oral solid dosage production through direct compression powder lubrication must be carefully selected to facilitate the manufacturing of tablets without degrading product manufacturability and quality (e.g. dissolution). To do so, several semi-empirical models relating compression performance to process operating conditions have been developed. Among them, we consider an extension of the Kushner and Moore model (Kushner and Moore, 2010, International Journal Pharmaceutics, 399:19) that is useful for the purpose, but requires an extensive experimental campaign for parameters identification. This implies the preparation and compression of multiple powder blends, each one with a different lubrication extent. In turn, this translates into a considerable consumption of Active Pharmaceutical Ingredient (API), and into time-consuming experiments. We tackled this issue by proposing a novel model-based design of experiments (MBDoE) approach, which minimizes the number of optimal blends for model calibration, while obtaining statistically sound parameters estimates and model predictions. Both sequential and parallel MBDoE configurations were compared. Experimental results involving two placebo blends with different lubrication sensitivity showed that this methodology is able to reduce the experimental effort by 60-70% with respect to the standard industrial practice independently of the formulation considered and configuration (i.e. parallel vs. sequential) adopted.

Population Pharmacokinetics of Teicoplanin in Preterm and Term Neonates: Is It Time for a New Dosing Regimen?

Our objective was to develop a population pharmacokinetic (PK) model in order to evaluate the currently recommended dosing regimen in term and preterm neonates. By using an optimal design approach, a prospective PK study was designed and implemented in 60 neonates with postmenstrual ages (PMA) of 26 to 43 weeks. A loading dose of 16 mg/kg was administered at day 1, followed by a maintenance dose of 8 mg/kg daily. Plasma concentrations were quantified by high-pressure liquid chromatography-mass spectrometry. Population PK (popPK) analysis was performed using NONMEM software. Monte-Carlo (MC) simulations were performed to evaluate currently recommended dosing based on a pharmacodynamic index of area under the concentration-time curve (AUC)/MIC ratio of ≥400. A two-compartment model with linear elimination best described the data by the following equations: clearance (CL) = 0.0227 × (weight [wt]/1,765)^0.75 × (estimated creatinine clearance [eCRCL]/22)^0.672, central compartment volume of distribution (V1) = 0.283 (wt/1,765), intercompartmental clearance (Q) = 0.151 (wt/1,765)^0.75, and peripheral compartment volume (V2) = 0.541 (wt/1,765). The interindividual variability estimates for CL, V1, and V2 were 36.5%, 45.7%, and 51.4%, respectively. Current weight (wt) and estimated creatinine clearance (eCRCL) significantly explained the observed variability. MC simulation demonstrated that, with the current dosing regimen, an AUC/MIC ratio of ≥400 was reached by only 68.5% of neonates with wt of <1 kg when the MIC was equal to 1 mg/kg, versus 82.2%, 89.7%, and 92.7% of neonates with wt of 1 to <2, 2 to <3, or ≥3 kg, respectively. Augmentation of a maintenance dose up to 10 or 11 mg/kg for preterm neonates with wt of 1 to <2 or <1 kg, respectively, increases the probability of reaching the therapeutic target; the recommended doses seem to be adequate for neonates with wt of ≥2 kg. Teicoplanin PK are variable in neonates, with wt and eCRCL having the most significant impact. Neonates with wt of <2 kg need higher doses, especially for Staphylococcus spp. with an MIC value of ≥1 mg/liter.

GAPPS (Grading and Assessment of Pharmacokinetic-Pharmacodynamic Studies) a critical appraisal system for antimicrobial PKPD studies - development and application in pediatric antibiotic studies

Introduction: There are limited data on optimal dosing of antibiotics in different age groups for neonates and children. Clinicians usually consult pediatric formularies or online databases for dose selection, but these have variable recommendations, are usually based on expert opinion and are not graded based on the existing pharmacokinetic-pharmacodynamic (PKPD) studies. We describe here a potential new tool that could be used to grade the strength of evidence emanating from PKPD studies.Areas covered: A scoring system was developed (GAPPS tool) to quantify the strength of each PK assessment and rate the studies quality in already published articles. GAPPS was evaluated by applying it to pediatric PKPD studies of antibiotics from the 2019 Essential Medicines List for children (EML_C), identified through a search of PubMed.Expert opinion: Evidence for most antibiotic dose selection decisions was generally weak, coming from individual PK studies and lacked PKPD modeling and simulations. However, the quality of evidence appears to have improved over the last two decades.Incorporating a formal grading system, such as GAPPS, into formulary development will provide a transparent tool to support decision-making in clinical practice and guideline development, and guide PKPD authors on study designs most likely to influence guidelines.

Generating Robust and Informative Nonclinical In Vitro and In Vivo Bacterial Infection Model Efficacy Data To Support Translation to Humans

In June 2017, the National Institute of Allergy and Infectious Diseases, part of the National Institutes of Health, organized a workshop entitled “Pharmacokinetics-Pharmacodynamics (PK/PD) for Development of Therapeutics against Bacterial Pathogens.” The aims were to discuss details of various PK/PD models and identify sound practices for deriving and utilizing PK/PD relationships to design optimal dosage regimens for patients. Workshop participants encompassed individuals from academia, industry, and government, including the United States Food and Drug Administration.

In June 2017, the National Institute of Allergy and Infectious Diseases, part of the National Institutes of Health, organized a workshop entitled “Pharmacokinetics-Pharmacodynamics (PK/PD) for Development of Therapeutics against Bacterial Pathogens.” The aims were to discuss details of various PK/PD models and identify sound practices for deriving and utilizing PK/PD relationships to design optimal dosage regimens for patients. Workshop participants encompassed individuals from academia, industry, and government, including the United States Food and Drug Administration. This and the accompanying review on clinical PK/PD summarize the workshop discussions and recommendations. Nonclinical PK/PD models play a critical role in designing human dosage regimens and are essential tools for drug development. These include in vitro and in vivo efficacy models that provide valuable and complementary information for dose selection and translation from the laboratory to human. It is crucial that studies be designed, conducted, and interpreted appropriately. For antibacterial PK/PD, extensive published data and expertise are available. These have been leveraged to develop recommendations, identify common pitfalls, and describe the applications, strengths, and limitations of various nonclinical infection models and translational approaches. Despite these robust tools and published guidance, characterizing nonclinical PK/PD relationships may not be straightforward, especially for a new drug or new class. Antimicrobial PK/PD is an evolving discipline that needs to adapt to future research and development needs. Open communication between academia, pharmaceutical industry, government, and regulatory bodies is essential to share perspectives and collectively solve future challenges.

Modelling performance and skeletal muscle adaptations with exponential growth functions during resistance training

System theory is classically applied to describe and to predict the effects of training load on performance. The classic models are structured by impulse-type transfer functions, nevertheless, most biological adaptations display exponential growth kinetics. The aim of this study was to propose an extension of the model structure taking into account the exponential nature of skeletal muscle adaptations by using a genetic algorithm. Thus, the conventional impulse-type model was applied in 15 resistance trained rodents and compared with exponential growth-type models. Even if we obtained a significant correlation between actual and modelled performances for all the models, our data indicated that an exponential model is associated with more suitable parameters values, especially the time constants that correspond to the positive response to training. Moreover, positive adaptations predicted with an exponential component showed a strong correlation with the main structural adaptations examined in skeletal muscles, i.e. hypertrophy (R² = 0.87, 0.96 and 0.99, for type 1, 2A and 2X cross-sectional area fibers, respectively) and changes in fiber-type composition (R² = 0.81 and 0.79, for type 1 and 2A fibers, respectively). Thus, an exponential model succeeds to describe both performance variations with relevant time constants and physiological adaptations that take place during resistance training.

Useful pharmacodynamic endpoints in children: selection, measurement, and next steps

Pharmacodynamic (PD) endpoints are essential for establishing the benefit-to-risk ratio for therapeutic interventions in children and neonates. This article discusses the selection of an appropriate measure of response, the PD endpoint, which is a critical methodological step in designing pediatric efficacy and safety studies. We provide an overview of existing guidance on the choice of PD endpoints in pediatric clinical research. We identified several considerations relevant to the selection and measurement of PD endpoints in pediatric clinical trials, including the use of biomarkers, modeling, compliance, scoring systems, and validated measurement tools. To be useful, PD endpoints in children need to be clinically relevant, responsive to both treatment and/or disease progression, reproducible, and reliable. In most pediatric disease areas, this requires significant validation efforts. We propose a minimal set of criteria for useful PD endpoint selection and measurement. We conclude that, given the current heterogeneity of pediatric PD endpoint definitions and measurements, both across and within defined disease areas, there is an acute need for internationally agreed, validated, and condition-specific pediatric PD endpoints that consider the needs of all stakeholders, including healthcare providers, policy makers, patients, and families.

Choosing an optimal input for an intravenous glucose tolerance test to aid parameter identification

OBJECTIVE

The minimal model is used to estimate insulin sensitivity in patients with diabetes, following an intravenous glucose tolerance test (IVGTT). Issues have been reported regarding parameter estimation, including correlation between insulin sensitivity and action parameters. The objective was to reduce these issues, by modifying the input of glucose in the test.

METHODS

Data were available for 24 volunteers following an IVGTT and glucose clamp test. Correlation between parameters was explored using likelihood heatmaps. An integrated glucose-insulin model was used to simulate glucose and insulin concentrations following new glucose inputs. The improved input for the test was selected by finding the minimum inverse of the determinant of the Fisher information matrix.

KEY FINDINGS

When the minimal model was fitted to the IVGTT data, there was clear correlation between the insulin parameters. With the glucose clamp, all parameters were correlated and badly estimated. The modified input, a bolus dose followed by constant infusion, resulted in improvement in parameter estimation and reduction in parameter correlation.

CONCLUSIONS

It is possible to reduce the issues with parameter estimation in the minimal model by modifying the glucose input, leading to a simplified test deign and a reduction in the total amount of glucose infused.

Study design and population pharmacokinetic analysis of a phase II dose-ranging study of interleukin-1 receptor antagonist

Interleukin-1 receptor antagonist, a naturally-occurring antagonist to the pro-inflammatory cytokine Interleukin-1, is already in clinical use. In experimental models of stroke, Interleukin-1 receptor antagonist in cerebrospinal fluid has been associated with cerebral neuroprotection and in a phase I clinical trial in patients with subarachnoid haemorrhage it crosses the blood-cerebrospinal fluid barrier. The aims of the current work were to design a dose-ranging clinical study in patients and to analyse the plasma and cerebrospinal fluid data obtained using a population pharmacokinetic modelling approach. The study was designed using prior information: a published population pharmacokinetic model and associated parameter estimates. Simulations were carried out to identify combinations of intravenous bolus and 4 h infusion doses that could achieve a concentration of 100 ng/ml in cerebrospinal fluid within approximately 30 min. The most informative time points for plasma and cerebrospinal fluid were obtained prospectively; optimisation identified five sampling time points that were included in the 15 time points in the present study design. All plasma and cerebrospinal fluid concentration data from previous and current studies were combined for updated analysis. The result of the simulations showed that a dosage regimen of 500 mg intravenous bolus and 10 mg/kg/h could achieve the target concentration, however four other regimens that represent a stepwise increase in maximum concentration were also selected. Analysis of the updated data showed improvement in parameter accuracy and predictive performance of the model; the percentage relative standard errors for fixed and random-effects parameters were <15 and 35% respectively. A dose-ranging study was successfully designed using modelling and simulation.

Dependence of treatment planning accuracy in peptide receptor radionuclide therapy on the sampling schedule

Background

Peptide receptor radionuclide therapy (PRRT) plays an important role in the treatment of neuroendocrine tumors (NET). Pre-therapeutic dosimetry using the area under the measured time-activity curve (AUC) is important. The sampling schedule for this dosimetry determines the accuracy and reliability of the obtained AUC.

The aim of this study was to investigate the effect of reduced number of measurement points (i.e., gamma camera image acquisition or serum measurements) on treatment planning accuracy in PRRT using ¹¹¹In-labeled-diethylenetriaminopentaacetic acid-octreotide (DTPAOC; Octreoscan™).

Methods

Pre-therapeutic biokinetic data of 15 NET patients were investigated using a recently developed physiologically based pharmacokinetic (PBPK) model. Two parameter sets were determined (standard or iterative approach) and used for calculation of time-integrated activity coefficients (TIACs) for the tumor, kidneys, liver, spleen, serum, and whole body. TIACs obtained using the full data sets were used as reference. To evaluate the effect of sampling on individual treatment planning, reduced sampling schedules were generated omitting either 1, 2, 3, or 4 organ and serum measurements or all serum measurements for each patient. Relative deviations (RDs) between these and reference TIACs were calculated and used as criterion for treatment planning accuracy. An RD < 0.1 was considered acceptable.

Results

When omitting serum measurements, TIAC accuracy remained acceptable (RD < 0.1) for the standard approach. The kidney TIACs could be estimated for both approaches with acceptable RDs using two time points (t = 4 h; 2 d); tumor RDs were <0.3. The iterative approach reduced the range of RD, but did not further reduce the number of needed measurement points (i.e., to achieve an RD <0.1). For both approaches RDs for liver, spleen and whole body were larger than 0.1. However, in the clinical setting these RDs are less relevant as liver and spleen are not organs at risk due to the low absorbed doses.

Conclusions

When using a priori information of a PBPK model structure combined with Bayesian information about PBPK model parameter distribution, the administered activity could be determined with acceptable accuracy using only two time points (4 h, 2 d) and thus allow a considerable reduction of needed data for individual dosimetry.

Model-based clinical drug development in the past, present and future: a commentary

Clinical drug development remains a mostly empirical, costly enterprise, in which decision-making is often based on qualitative assessment of risk, without properly leveraging all the relevant data collected throughout the development programme. Model-based drug development (MBDD) has been proposed by regulatory agencies, academia and pharmaceutical companies as a paradigm to modernize drug research through the quantification of risk and combination of information from different sources across time. We present here a historical account of the use of MBDD in clinical drug development, the current challenges and further opportunities for its application in the pharmaceutical industry.

Optimal design in pediatric pharmacokinetic and pharmacodynamic clinical studies

It is not trivial to conduct clinical trials with pediatric participants. Ethical, logistical, and financial considerations add to the complexity of pediatric studies. Optimal design theory allows investigators the opportunity to apply mathematical optimization algorithms to define how to structure their data collection to answer focused research questions. These techniques can be used to determine an optimal sample size, optimal sample times, and the number of samples required for pharmacokinetic and pharmacodynamic studies. The aim of this review is to demonstrate how to determine optimal sample size, optimal sample times, and the number of samples required from each patient by presenting specific examples using optimal design tools. Additionally, this review aims to discuss the relative usefulness of sparse vs rich data. This review is intended to educate the clinician, as well as the basic research scientist, whom plan on conducting a pharmacokinetic/pharmacodynamic clinical trial in pediatric patients.

Modeling the responses to resistance training in an animal experiment study

The aim of the present study was to test whether systems models of training effects on performance in athletes can be used to explore the responses to resistance training in rats. 11 Wistar Han rats (277 ± 15 g) underwent 4 weeks of resistance training consisting in climbing a ladder with progressive loads. Training amount and performance were computed from total work and mean power during each training session. Three systems models relating performance to cumulated training bouts have been tested: (i) with a single component for adaptation to training, (ii) with two components to distinguish the adaptation and fatigue produced by exercise bouts, and (iii) with an additional component to account for training-related changes in exercise-induced fatigue. Model parameters were fitted using a mixed-effects modeling approach. The model with two components was found to be the most suitable to analyze the training responses (R(2) = 0.53; P < 0.001). In conclusion, the accuracy in quantifying training loads and performance in a rodent experiment makes it possible to model the responses to resistance training. This modeling in rodents could be used in future studies in combination with biological tools for enhancing our understanding of the adaptive processes that occur during physical training.

Using Animal Instincts to Design Efficient Biomedical Studies via Particle Swarm Optimization

Particle swarm optimization (PSO) is an increasingly popular metaheuristic algorithm for solving complex optimization problems. Its popularity is due to its repeated successes in finding an optimum or a near optimal solution for problems in many applied disciplines. The algorithm makes no assumption of the function to be optimized and for biomedical experiments like those presented here, PSO typically finds the optimal solutions in a few seconds of CPU time on a garden-variety laptop. We apply PSO to find various types of optimal designs for several problems in the biological sciences and compare PSO performance relative to the differential evolution algorithm, another popular metaheuristic algorithm in the engineering literature.

A mixed-effects model of the dynamic response of muscle gene transcript expression to endurance exercise

Altered expression of a broad range of gene transcripts after exercise reflects the specific adjustment of skeletal muscle makeup to endurance training. Towards a quantitative understanding of this molecular regulation, we aimed to build a mixed-effects model of the dynamics of co-related transcript responses to exercise. It was built on the assumption that transcript levels after exercise varied because of changes in the balance between transcript synthesis and degradation. It was applied to microarray data of 231 gene transcripts in vastus lateralis muscle of six subjects 1, 8 and 24 h after endurance exercise and 6-week training on a stationary bicycle. Cluster analysis was used to select groups of transcripts having highest co-correlation of their expression (r > 0.70): Group 1 comprised 45 transcripts including factors defining the oxidative and contractile phenotype and Group 2 included 39 transcripts mainly defined by factors found at the cell periphery and the extracellular space. Data from six subjects were pooled to filter experimental noise. The model fitted satisfactorily the responses of Group 1 (r (2) = 0.62 before and 0.85 after training, P < 0.001) and Group 2 (r (2) = 0.75 and 0.79, P < 0.001). Predicted variation in transcription rate induced by exercise yielded a difference in amplitude and time-to-peak response of gene transcripts between the two groups before training and with training in Group 2. The findings illustrate that a mixed-effects model of transcript responses to exercise is suitable to explore the regulation of muscle plasticity by training at the transcriptional level and indicate critical experiments needed to consolidate model parameters empirically.

Optimal design in population kinetic experiments by set-valued methods

We propose a new method for optimal experimental design of population pharmacometric experiments based on global search methods using interval analysis; all variables and parameters are represented as intervals rather than real numbers. The evaluation of a specific design is based on multiple simulations and parameter estimations. The method requires no prior point estimates for the parameters, since the parameters can incorporate any level of uncertainty. In this respect, it is similar to robust optimal design. Representing sampling times and covariates like doses by intervals gives a direct way of optimizing with rigorous sampling and dose intervals that can be useful in clinical practice. Furthermore, the method works on underdetermined problems for which traditional methods typically fail.