Interactive Pharmacometric Applications Using R and the Shiny Package

gPKPDviz: A flexible R shiny tool for pharmacokinetic/pharmacodynamic simulations using mrgsolve

GPKPDviz is a Shiny application (app) dedicated to real-time simulation, visualization, and assessment of the pharmacokinetic/pharmacodynamic (PK/PD) models. Within the app, gPKPDviz is capable of generating virtual populations and complex dosing and sampling scenarios, which, together with the streamlined workflow, is designed to efficiently assess the impact of covariates and dosing regimens on PK/PD end points. The actual population data from clinical trials can be loaded into the app for simulation if desired. The app-generated dosing regimens include single or multiple dosing, and more complex regimens, such as loading doses or intermittent dosing. When necessary, the dosing regimens can be defined externally and loaded to the app for simulation. Using mrgsolve as the simulation engine, gPKPDviz is typically used for population simulation, however, with a slight modification of the mrgsolve model, gPKPDviz is capable of performing individual simulations with individual post hoc parameters, individual dosing logs, and individual sampling timepoints through an external dataset. A built-in text editor has a debugging feature for the mrgsolve model, providing the same error messages as model compilation in R. GPKPDviz has had stringent validation by comparing simulation results between the app and using mrgsolve in R. GPKPDviz is a member of the suite of Modeling and Simulation Shiny apps developed at Genentech to facilitate the typical modeling work in Clinical Pharmacology. For broader access to the Pharmacometric community, gPKPDviz has been published as an open-source application in GitHub under the terms of GNU General Public License.

CITIES: Clinical trials with intercurrent events simulator

Although clinical trials are often designed with randomization and well-controlled protocols, complications will inevitably arise in the presence of intercurrent events (ICEs) such as treatment discontinuation. These can lead to missing outcome data and possibly confounding causal inference when the missingness is a function of a latent stratification of patients defined by intermediate outcomes. The pharmaceutical industry has been focused on developing new methods that can yield pertinent causal inferences in trials with ICEs. However, it is difficult to compare the properties of different methods developed in this endeavor as real-life clinical trial data cannot be easily shared to provide benchmark data sets. Furthermore, different methods consider distinct assumptions for the underlying data-generating mechanisms, and simulation studies often are customized to specific situations or methods. We develop a novel, general simulation model and corresponding Shiny application in R for clinical trials with ICEs, aptly named the Clinical Trials with Intercurrent Events Simulator (CITIES). It is formulated under the Rubin Causal Model where the considered treatment effects account for ICEs in clinical trials with repeated measures. CITIES facilitates the effective generation of data that resemble real-life clinical trials with respect to their reported summary statistics, without requiring the use of the original trial data. We illustrate the utility of CITIES via two case studies involving real-life clinical trials that demonstrate how CITIES provides a comprehensive tool for practitioners in the pharmaceutical industry to compare methods for the analysis of clinical trials with ICEs on identical, benchmark settings that resemble real-life trials.

Role of Artificial Intelligence in Drug Discovery and Target Identification in Cancer

Drug discovery and development (DDD) is a highly complex process that necessitates precise monitoring and extensive data analysis at each stage. Furthermore, the DDD process is both timeconsuming and costly. To tackle these concerns, artificial intelligence (AI) technology can be used, which facilitates rapid and precise analysis of extensive datasets within a limited timeframe. The pathophysiology of cancer disease is complicated and requires extensive research for novel drug discovery and development. The first stage in the process of drug discovery and development involves identifying targets. Cell structure and molecular functioning are complex due to the vast number of molecules that function constantly, performing various roles. Furthermore, scientists are continually discovering novel cellular mechanisms and molecules, expanding the range of potential targets. Accurately identifying the correct target is a crucial step in the preparation of a treatment strategy. Various forms of AI, such as machine learning, neural-based learning, deep learning, and network-based learning, are currently being utilised in applications, online services, and databases. These technologies facilitate the identification and validation of targets, ultimately contributing to the success of projects. This review focuses on the different types and subcategories of AI databases utilised in the field of drug discovery and target identification for cancer.

Improving the usability of open health service delivery simulation models using Python and web apps

One aim of Open Science is to increase the accessibility of research. Within health services research that uses discrete-event simulation, Free and Open Source Software (FOSS), such as Python, offers a way for research teams to share their models with other researchers and NHS decision makers. Although the code for healthcare discrete-event simulation models can be shared alongside publications, it may require specialist skills to use and run. This is a disincentive to researchers adopting Free and Open Source Software and open science practices. Building on work from other health data science disciplines, we propose that web apps offer a user-friendly interface for healthcare models that increase the accessibility of research to the NHS, and researchers from other disciplines. We focus on models coded in Python deployed as streamlit web apps. To increase uptake of these methods, we provide an approach to structuring discrete-event simulation model code in Python so that models are web app ready. The method is general across discrete-event simulation Python packages, and we include code for both simpy and ciw implementations of a simple urgent care call centre model. We then provide a step-by-step tutorial for linking the model to a streamlit web app interface, to enable other health data science researchers to reproduce and implement our method.

Improving pharmacometrics analysis efficiency using DataCheQC: An interactive, Shiny-based app for quality control of pharmacometrics datasets

DataCheQC is an interactive application based on the R Shiny framework developed for the purposes of performing quality control (QC) checks on pharmacometric datasets, and thereby supporting the implementation of model-informed drug development. Features include visual inspection of variables and data entries for errors and/or anomalies, and ensuring structural integrity through comparison with a dataset specification file. The app, which requires no programming knowledge to operate, allows the user to collect all findings into a summary report downloadable directly from the app itself. The source code for the app is freely available on GitHub under an open-source license (https://github.com/DotanOr/DataCheQC) and can also be accessed online (https://dotanor.shinyapps.io/DataCheQC/).

Pharmacokinetic Pharmacodynamic Modelling Contributions to Improve Paediatric Anaesthesia Practice

The use of pharmacokinetic-pharmacodynamic models has improved anaesthesia practice in children through a better understanding of dose-concentration-response relationships, developmental pharmacokinetic changes, quantification of drug interactions and insights into how covariates (e.g., age, size, organ dysfunction, pharmacogenomics) impact drug prescription. Simulation using information from these models has enabled the prediction and learning of beneficial and adverse effects and decision-making around clinical scenarios. Covariate information, including the use of allometric size scaling, age and consideration of fat mass, has reduced population parameter variability. The target concentration approach has rationalised dose calculation. Paediatric pharmacokinetic-pharmacodynamic insights have led to better drug delivery systems for total intravenous anaesthesia and an expectation about drug offset when delivery is stopped. Understanding concentration-dependent adverse effects have tempered dose regimens. Quantification of drug interactions has improved the understanding of the effects of drug combinations. Repurposed drugs (e.g., antiviral drugs used for COVID-19) within the community can have important effects on drugs used in paediatric anaesthesia, and the use of simulation educates about these drug vagaries.

Greenhouse gas dynamics in an urbanized river system: influence of water quality and land use

Rivers act as a natural source of greenhouse gases (GHGs). However, anthropogenic activities can largely alter the chemical composition and microbial communities of rivers, consequently affecting their GHG production. To investigate these impacts, we assessed the accumulation of CO₂, CH₄, and N₂O in an urban river system (Cuenca, Ecuador). High variation of dissolved GHG concentrations was found among river tributaries that mainly depended on water quality and land use. By using Prati and Oregon water quality indices, we observed a clear pattern between water quality and the dissolved GHG concentration: the more polluted the sites were, the higher were their dissolved GHG concentrations. When river water quality deteriorated from acceptable to very heavily polluted, the mean value of pCO₂ and dissolved CH₄ increased by up to ten times while N₂O concentrations boosted by 15 times. Furthermore, surrounding land-use types, i.e., urban, roads, and agriculture, could considerably affect the GHG production in the rivers. Particularly, the average pCO₂ and dissolved N₂O of the sites close to urban areas were almost four times higher than those of the natural sites while this ratio was 25 times in case of CH₄, reflecting the finding that urban areas had the worst water quality with almost 70% of their sites being polluted while this proportion of nature areas was only 12.5%. Lastly, we identified dissolved oxygen, ammonium, and flow characteristics as the main important factors to the GHG production by applying statistical analysis and random forests. These results highlighted the impacts of land-use types on the production of GHGs in rivers contaminated by sewage discharges and surface runoff.

Design of decision support system incorporating data mining algorithms for strengthening maternal and child health systems: Inclusion of systems-thinking approach

Reduction of maternal and infant mortality rates has been recognisedas one of the important goals of this century. Both coverageimprovement and inequity reduction have been set up asmillennium targets. Despite the availability of effective interventions,maternal and child healthcare conditions are not improvingin developing countries because of inefficiently functioninghealth systems. Knowledge generation about behaviors ofhealth system building blocks on the implementation of severalhealthcare interventions will help policymakers to design situation-specific and strategic interventions. A decision supportsystem has been devised incorporating data mining algorithmswhich would help to understand the condition of maternal andchild healthcare indicators; educational, socio, and economicsituations; healthcare status; and healthcare service blocksand their relationships with each other. In this paper, the designof the DSS has been discussed elaborately. To enhance a system-wide understanding of the healthcare system, all healthcare-related factors have been incorporated into this system.Three knowledge generation modules have been prepared byutilizing different visualization and data mining algorithms.

A Machine Learning Model to Predict the Triple Negative Breast Cancer Immune Subtype

Background

Immune checkpoint blockade (ICB) has been approved for the treatment of triple-negative breast cancer (TNBC), since it significantly improved the progression-free survival (PFS). However, only about 10% of TNBC patients could achieve the complete response (CR) to ICB because of the low response rate and potential adverse reactions to ICB.

Methods

Open datasets from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) were downloaded to perform an unsupervised clustering analysis to identify the immune subtype according to the expression profiles. The prognosis, enriched pathways, and the ICB indicators were compared between immune subtypes. Afterward, samples from the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) dataset were used to validate the correlation of immune subtype with prognosis. Data from patients who received ICB were selected to validate the correlation of the immune subtype with ICB response. Machine learning models were used to build a visual web server to predict the immune subtype of TNBC patients requiring ICB.

Results

A total of eight open datasets including 931 TNBC samples were used for the unsupervised clustering. Two novel immune subtypes (referred to as S1 and S2) were identified among TNBC patients. Compared with S2, S1 was associated with higher immune scores, higher levels of immune cells, and a better prognosis for immunotherapy. In the validation dataset, subtype 1 samples had a better prognosis than sub type 2 samples, no matter in overall survival (OS) (p = 0.00036) or relapse-free survival (RFS) (p = 0.0022). Bioinformatics analysis identified 11 hub genes (LCK, IL2RG, CD3G, STAT1, CD247, IL2RB, CD3D, IRF1, OAS2, IRF4, and IFNG) related to the immune subtype. A robust machine learning model based on random forest algorithm was established by 11 hub genes, and it performed reasonably well with area Under the Curve of the receiver operating characteristic (AUC) values = 0.76. An open and free web server based on the random forest model, named as triple-negative breast cancer immune subtype (TNBCIS), was developed and is available from https://immunotypes.shinyapps.io/TNBCIS/.

Conclusion

TNBC open datasets allowed us to stratify samples into distinct immunotherapy response subgroups according to gene expression profiles. Based on two novel subtypes, candidates for ICB with a higher response rate and better prognosis could be selected by using the free visual online web server that we designed.

A Near Real-Time Hydrological Information System for the Upper Danube Basin

The multi-national catchment of the Upper Danube covers an area of more than 100,000 km2 and is of great ecological and economic value. Its hydrological states (e.g., runoff conditions, snow cover states or groundwater levels) affect fresh-water supply, agriculture, hydropower, transport and many other sectors. The timely knowledge of the current status is therefore of importance to decision makers from administration or practice but also the interested public. Therefore, a web-based, near real-time hydrological information system was conceptualized and developed for the Upper Danube upstream of Vienna (Upper Danube HIS), utilizing ERA5 reanalysis data (ERA5) and hydrological simulations provided by the semi-distributed hydrological model COSERO. The ERA5 reanalysis data led to comparatively high simulation performance for a total of 65 subbasins with a median NSE and KGE of 0.69 and 0.81 in the parameter calibration and 0.63 and 0.75 in the validation period. The Upper Danube HIS was implemented within the R programming environment as a web application based on the Shiny framework. This enables an intuitive, interactive access to the system. It offers various capabilities for a hydrometeorological analysis of the 65 subbasins of the Upper Danube basin, inter alia, a method for the identification of hydrometeorological droughts. This proof of concept and system underlines how valuable information can be obtained from freely accessible data and by the means of open source software and is made available to the hydrological community, water managers and the public.

Easy and reliable maximum a posteriori Bayesian estimation of pharmacokinetic parameters with the open-source R package mapbayr

Pharmacokinetic (PK) parameter estimation is a critical and complex step in the model‐informed precision dosing (MIPD) approach. The mapbayr package was developed to perform maximum a posteriori Bayesian estimation (MAP‐BE) in R from any population PK model coded in mrgsolve. The performances of mapbayr were assessed using two approaches. First, “test” models with different features were coded, for example, first‐order and zero‐order absorption, lag time, time‐varying covariates, Michaelis–Menten elimination, combined and exponential residual error, parent drug and metabolite, and small or large inter‐individual variability (IIV). A total of 4000 PK profiles (combining single/multiple dosing and rich/sparse sampling) were simulated from each test model, and MAP‐BE of parameters was performed in both mapbayr and NONMEM. Second, a similar procedure was conducted with seven “real” previously published models to compare mapbayr and NONMEM on a PK outcome used in MIPD. For the test models, 98% of mapbayr estimations were identical to those given by NONMEM. Some discordances could be observed when dose‐related parameters were estimated or when models with large IIV were used. The exploration of objective function values suggested that mapbayr might outdo NONMEM in specific cases. For the real models, a concordance close to 100% on PK outcomes was observed. The mapbayr package provides a reliable solution to perform MAP‐BE of PK parameters in R. It also includes functions dedicated to data formatting and reporting and enables the creation of standalone Shiny web applications dedicated to MIPD, whatever the model or the clinical protocol and without additional software other than R.

Development and Validation of a LASSO Prediction Model for Better Identification of Ischemic Stroke: A Case-Control Study in China

Background

Timely diagnosis of ischemic stroke (IS) in the acute phase is extremely vital to achieve proper treatment and good prognosis. In this study, we developed a novel prediction model based on the easily obtained information at initial inspection to assist in the early identification of IS.

Methods

A total of 627 patients with IS and other intracranial hemorrhagic diseases from March 2017 to June 2018 were retrospectively enrolled in the derivation cohort. Based on their demographic information and initial laboratory examination results, the prediction model was constructed. The least absolute shrinkage and selection operator algorithm was used to select the important variables to form a laboratory panel. Combined with the demographic variables, multivariate logistic regression was performed for modeling, and the model was encapsulated within a visual and operable smartphone application. The performance of the model was evaluated on an independent validation cohort, formed by 304 prospectively enrolled patients from June 2018 to May 2019, by means of the area under the curve (AUC) and calibration.

Results

The prediction model showed good discrimination (AUC = 0.916, cut-off = 0.577), calibration, and clinical availability. The performance was reconfirmed in the more complex emergency department. It was encapsulated as the Stroke Diagnosis Aid app for smartphones. The user can obtain the identification result by entering the values of the variables in the graphical user interface of the application.

Conclusion

The prediction model based on laboratory and demographic variables could serve as a favorable supplementary tool to facilitate complex, time-critical acute stroke identification.

Amikacin initial dosage in patients with hypoalbuminaemia: an interactive tool based on a population pharmacokinetic approach

OBJECTIVES

To characterize amikacin population pharmacokinetics in patients with hypoalbuminaemia and to develop a model-based interactive application for amikacin initial dosage.

METHODS

A population pharmacokinetic model was developed using a non-linear mixed-effects modelling approach (NONMEM) with amikacin concentration-time data collected from clinical practice (75% hypoalbuminaemic patients). Goodness-of-fit plots, minimum objective function value, prediction-corrected visual predictive check, bootstrapping, precision and bias of parameter estimates were used for model evaluation. An interactive model-based simulation tool was developed in R (Shiny and R Markdown). Cmax/MIC ratio, time above MIC and AUC/MIC were used for optimizing amikacin initial dose recommendation. Probabilities of reaching targets were calculated for the dosage proposed.

RESULTS

A one-compartment model with first-order linear elimination best described the 873 amikacin plasma concentrations available from 294 subjects (model development and external validation groups). Estimated amikacin population pharmacokinetic parameters were CL (L/h) = 0.525 + 4.78 × (CKD-EPI/98) × (0.77 × vancomycin) and V (L) = 26.3 × (albumin/2.9)-0.51 × [1 + 0.006 × (weight - 70)], where CKD-EPI is calculated with the Chronic Kidney Disease Epidemiology Collaboration equation. AMKdose is a useful interactive model-based application for a priori optimization of amikacin dosage, using individual patient and microbiological information together with predefined pharmacokinetic/pharmacodynamic (PKPD) targets.

CONCLUSIONS

Serum albumin, total bodyweight, estimated glomerular filtration rate (using the CKD-EPI equation) and co-medication with vancomycin showed a significant impact on amikacin pharmacokinetics. A powerful interactive initial dose-finding tool has been developed and is freely available online. AMKdose could be useful for guiding initial amikacin dose selection before any individual pharmacokinetic information is available.

Spatial and temporal variations of greenhouse gas emissions from a waste stabilization pond: Effects of sludge distribution and accumulation

Due to regular influx of organic matter and nutrients, waste stabilization ponds (WSPs) can release considerable quantities of greenhouse gases (GHGs). To investigate the spatiotemporal variations of GHG emissions from WSPs with a focus on the effects of sludge accumulation and distribution, we conducted a bathymetry survey and two sampling campaigns in Ucubamba WSP (Cuenca, Ecuador). The results indicated that spatial variation of GHG emissions was strongly dependent on sludge distribution. Thick sludge layers in aerated ponds and facultative ponds caused substantial CO₂ and CH₄ emissions which accounted for 21.3% and 78.7% of the total emissions from the plant. Conversely, the prevalence of anoxic conditions stimulated the N₂O consumption via complete denitrification leading to a net uptake from the atmosphere, i.e. up to 1.4±0.2 mg-N m^-2 d^-1. Double emission rates of CO₂ were found in the facultative and maturation ponds during the day compared to night-time emissions, indicating the important role of algal respiration, while no diel variation of the CH₄ and N₂O emissions was found. Despite the uptake of N₂O, the total GHG emissions of the WSP was higher than constructed wetlands and conventional centralized wastewater treatment facilities. Hence, it is recommended that sludge management with proper desludging regulation should be included as an important mitigation measure to reduce the carbon footprint of pond treatment facilities.

Model-Optimized Fluconazole Dose Selection for Critically Ill Patients Improves Early Pharmacodynamic Target Attainment without the Need for Therapeutic Drug Monitoring

Fluconazole has been associated with higher mortality compared with the echinocandins in patients treated for invasive candida infections. Underexposure from current fluconazole dosing regimens may contribute to these worse outcomes, so alternative dosing strategies require study. The objective of this study was to evaluate fluconazole drug exposure in critically ill patients comparing a novel model-optimized dose selection method with established approaches over a standard 14-day (336-h) treatment course. Target attainment was evaluated in a representative population of 1,000 critically ill adult patients for (i) guideline dosing (800-mg loading and 400-mg maintenance dosing adjusted to renal function), (ii) guideline dosing followed by therapeutic drug monitoring (TDM)-guided dose adjustment, and (iii) model-optimized dose selection based on patient factors (without TDM). Assuming a MIC of 2 mg/liter, free fluconazole 24-h area under the curve (fAUC24) targets of ≥200 mg · h/liter and <800 mg · h/liter were used for assessment of target attainment. Guideline dosing resulted in underexposure in 21% of patients at 48 h and in 23% of patients at 336 h. The TDM-guided strategy did not influence 0- to 48-h target attainment due to inherent procedural delays but resulted in 37% of patients being underexposed at 336 h. Model-optimized dosing resulted in ≥98% of patients meeting efficacy targets throughout the treatment course, while resulting in less overexposure compared with guideline dosing (7% versus 14%) at 336 h. Model-optimized dose selection enables fluconazole dose individualization in critical illness from the outset of therapy and should enable reevaluation of the comparative effectiveness of this drug in patients with severe fungal infections.

Evaluation of Current Amikacin Dosing Recommendations and Development of an Interactive Nomogram: The Role of Albumin

This study aimed to evaluate the potential efficacy and safety of the amikacin dosage proposed by the main guidelines and to develop an interactive nomogram, especially focused on the potential impact of albumin on initial dosage recommendation. The probability of target attainment (PTA) for each of the different dosing recommendations was calculated through stochastic simulations based on pharmacokinetic/pharmacodynamic (PKPD) criteria. Large efficacy and safety differences were observed for the evaluated amikacin dosing guidelines together with a significant impact of albumin concentrations on efficacy and safety. For all recommended dosages evaluated, efficacy and safety criteria of amikacin dosage proposed were not achieved simultaneously in most of the clinical scenarios evaluated. Furthermore, a significant impact of albumin was identified: The higher is the albumin, (i) the higher will be the PTA for maximum concentration/minimum inhibitory concentration (Cmax/MIC), (ii) the lower will be the PTA for the time period with drug concentration exceeding MIC (T_>MIC) and (iii) the lower will be the PTA for toxicity (minimum concentration). Thus, accounting for albumin effect might be of interest for future amikacin dosing guidelines updates. In addition, AMKnom, an amikacin nomogram builder based on PKPD criteria, has been developed and is freely available to help evaluating dosing recommendations.

STRait Razor Online: An enhanced user interface to facilitate interpretation of MPS data

Since 2013, STRait Razor has enabled analysis of massively parallel sequencing (MPS) data from various marker systems such as short tandem repeats, single nucleotide polymorphisms, insertion/deletions, and mitochondrial DNA. In this paper, STRait Razor Online (SRO), available at https://www.unthsc.edu/straitrazor, is introduced as an interactive, Shiny-based user interface for primary analysis of MPS data and secondary analysis of STRait Razor haplotype pileups. This software can be accessed from any common browser via desktop, tablet, or smartphone device. SRO is available also as a standalone application and open-source R script available at https://github.com/ExpectationsManaged/STRaitRazorOnline. The local application is capable of batch processing of both fastq files and primary analysis output. Processed batches generate individual report folders and summary reports at the locus- and haplotype-level in a matter of minutes. For example, the processing of data from ∼700 samples generated with the ForenSeq Signature Preparation Kit from allsequences.txt to a final table can be performed in ∼40 min whereas the Excel-based workbooks can take 35-60 h to compile a subset of the tables generated by SRO. To facilitate analysis of single-source, reference samples, a preliminary triaging system was implemented that calls potential alleles and flags loci suspected of severe heterozygote imbalance. When compared to published, manually curated data sets, 98.72 % of software-assigned allele calls without manual interpretation were consistent with curated data sets, 0.99 % loci were presented to the user for interpretation due to heterozygote imbalance, and the remaining 0.29 % of loci were inconsistent due to the analytical thresholds used across the studies.

BCO App: tools for generating BioCompute Objects from next-generation sequencing workflows and computations

The BioCompute Object (BCO) standard is an IEEE standard (IEEE 2791-2020) designed to facilitate the communication of next-generation sequencing data analysis with applications across academia, government agencies, and industry. For example, the Food and Drug Administration (FDA) supports the standard for regulatory submissions and includes the standard in their Data Standards Catalog for the submission of HTS data. We created the BCO App to facilitate BCO generation in a range of computational environments and, in part, to participate in the Advanced Track of the precisionFDA BioCompute Object App-a-thon. The application facilitates the generation of BCOs from both workflow metadata provided as plaintext and from workflow contents written in the Common Workflow Language. The application can also access and ingest task execution results from the Cancer Genomics Cloud (CGC), an NCI funded computational platform. Creating a BCO from a CGC task significantly reduces the time required to generate a BCO on the CGC by auto-populating workflow information fields from CGC workflow and task execution results. The BCO App supports exporting BCOs as JSON or PDF files and publishing BCOs to both the CGC platform and to GitHub repositories.

An interactive application for malaria elimination transmission and costing in the Asia-Pacific

Leaders in the Asia-Pacific have endorsed an ambitious target to eliminate malaria in the region by 2030. The emergence and spread of artemisinin drug resistance in the Greater Mekong Subregion makes elimination urgent and strategic for the global goal of malaria eradication. Mathematical modelling is a useful tool for assessing and comparing different elimination strategies and scenarios to inform policymakers. Mathematical models are especially relevant in this context because of the wide heterogeneity of regional, country and local settings, which means that different strategies are needed to eliminate malaria. However, models and their predictions can be seen as highly technical, limiting their use for decision making. Simplified applications of models are needed to allow policy makers to benefit from these valuable tools. This paper describes a method for communicating complex model results with a user-friendly and intuitive framework. Using open-source technologies, we designed and developed an interactive application to disseminate the modelling results for malaria elimination. The design was iteratively improved while the application was being piloted and extensively tested by a diverse range of researchers and decision makers. This application allows several target audiences to explore, navigate and visualise complex datasets and models generated in the context of malaria elimination. It allows widespread access, use of and interpretation of models, generated at great effort and expense as well as enabling them to remain relevant for a longer period of time. It has long been acknowledged that scientific results need to be repackaged for larger audiences. We demonstrate that modellers can include applications as part of the dissemination strategy of their findings. We highlight that there is a need for additional research in order to provide guidelines and direction for designing and developing effective applications for disseminating models.

Statistical Methods for Adjusting Estimates of Treatment Effectiveness for Patient Nonadherence in the Context of Time-to-Event Outcomes and Health Technology Assessment: A Systematic Review of Methodological Papers

Introduction. Medication nonadherence can have a significant negative impact on treatment effectiveness. Standard intention-to-treat analyses conducted alongside clinical trials do not make adjustments for nonadherence. Several methods have been developed that attempt to estimate what treatment effectiveness would have been in the absence of nonadherence. However, health technology assessment (HTA) needs to consider effectiveness under real-world conditions, where nonadherence levels typically differ from those observed in trials. With this analytical requirement in mind, we conducted a review to identify methods for adjusting estimates of treatment effectiveness in the presence of patient nonadherence to assess their suitability for use in HTA. Methods. A "Comprehensive Pearl Growing" technique, with citation searching and reference checking, was applied across 7 electronic databases to identify methodological papers for adjusting time-to-event outcomes for nonadherence using individual patient data. A narrative synthesis of identified methods was conducted. Methods were assessed in terms of their ability to reestimate effectiveness based on alternative, suboptimal adherence levels. Results. Twenty relevant methodological papers covering 12 methods and 8 extensions to those methods were identified. Methods are broadly classified into 4 groups: 1) simple methods, 2) principal stratification methods, 3) generalized methods (g-methods), and 4) pharmacometrics-based methods using pharmacokinetics and pharmacodynamics (PKPD) analysis. Each method makes specific assumptions and has associated limitations. Five of the 12 methods are capable of adjusting for real-world nonadherence, with only g-methods and PKPD considered appropriate for HTA. Conclusion. A range of statistical methods is available for adjusting estimates of treatment effectiveness for nonadherence, but most are not suitable for use in HTA. G-methods and PKPD appear to be more appropriate to estimate effectiveness in the presence of real-world adherence.

Dynamic predictive model for postoperative nausea and vomiting for intravenous fentanyl patient-controlled analgesia

Postoperative nausea and vomiting is the most common side-effect of opioid-based intravenous patient-controlled analgesia. Apfel's simplified risk score is popular but it has some limitations. We developed and validated a dynamic predictive model for nausea or vomiting up to 48 postoperative hours, available as an online web application. Fentanyl was used by 22,144 adult patients for analgesia after non-cardiac surgery under general anaesthesia: we randomly divided them into development (80%) and validation (20%) cohorts, repeated 100 times. We used linear discriminant analysis to select variables for multivariate logistic regression. The incidences of postoperative nausea or vomiting were: 0-48 h, 5691/22,144 (26%); 0-6 h, 2749/22,144 (12%); 6-12 h, 2687/22,144 (12%); 12-18 h, 2624/22,144 (12%); 18-24 h, 1884/22,144 (9%); and 24-48 h, 1082/22,144 (5%). The median (95%CI) area under the receiver operating characteristic curve was 0.72 (0.71-0.73) up to 48 postoperative hours compared with 0.65 (0.64-0.66) for the Apfel model, p < 0.001. The equivalent areas for 0-6 h, 6-12 h, 12-18 h, 18-24 h and 24-48 h were: 0.70 (0.69-0.72); 0.71 (0.69-0.73); 0.69 (0.68-0.71); 0.70 (0.67-0.72); and 0.69 (0.66-0.71), respectively. Our web application allows clinicians to calculate incidences of nausea and vomiting in patients receiving intravenous fentanyl for patient-controlled analgesia.

Converting from Transdermal to Buccal Formulations of Buprenorphine: A Pharmacokinetic Meta-Model Simulation in Healthy Volunteers

Objective

To develop a model to predict buprenorphine plasma concentrations during transition from transdermal to buccal administration.

Design

Population pharmacokinetic model-based meta-analysis of published data.

Methods

A model-based meta-analysis of available buprenorphine pharmacokinetic data in healthy adults, extracted as aggregate (mean) data from published literature, was performed to explore potential conversion from transdermal to buccal buprenorphine. The time course of mean buprenorphine plasma concentrations following application of transdermal patch or buccal film was digitized from available literature, and a meta-model was developed using specific pharmacokinetic parameters (e.g., absorption rate, apparent clearance, and volumes of distribution) derived from analysis of pharmacokinetic data for intravenously, transdermally, and buccally administered buprenorphine.

Results

Data from six studies were included in this analysis. The final transdermal absorption model employed a zero-order input rate that was scaled to reflect a nominal patch delivery rate and time after patch application (with decline in rate over time). The transdermal absorption rate constant became zero following patch removal. Buccal absorption was a first-order process with a time lag and bioavailability term. Simulations of conversion from transdermal 20 mcg/h and 10 mcg/h to buccal administration suggest that transition can be made rapidly (beginning 12 hours after patch removal) using the recommended buccal formulation titration increments (75-150 mcg) and schedule (every four days) described in the product labeling.

Conclusions

Computer modeling and simulations using a meta-model built from data extracted from publications suggest that rapid and straightforward conversion from transdermal to buccal buprenorphine is feasible.

Optimizing Vancomycin Dosing in Chronic Kidney Disease by Deriving and Implementing a Web-Based Tool Using a Population Pharmacokinetics Analysis

Background: Chronic kidney disease (CKD) patients requiring intravenous vancomycin bear considerable risks of adverse outcomes both from the infection and vancomycin therapy itself, necessitating especially precise dosing to avoid sub- and supratherapeutic vancomycin exposure.

Methods: In this retrospective study, we performed a population pharmacokinetic analysis to construct a vancomycin dose prediction model for CKD patients who do not require renal replacement therapy. The model was externally validated on an independent cohort of patients to assess its prediction accuracy. The pharmacokinetic parameter estimates and the equations were productized into a Web application (VancApp) subsequently implemented in routine care. The association between VancApp-based dosing and time-to-target concentration attainment, 30-day mortality, and nephrotoxicity were assessed postimplementation.

Results: The model constructed from an initial cohort (n = 80) revealed a population clearance and volume of distribution of 1.30 L/h and 1.23 L/kg, respectively. External model validation (n = 112) demonstrated a mean absolute prediction error of 1.25 mg/L. Following 4 months of clinical implementation of VancApp as an optional alternative to usual care [VancApp (n = 22) vs. usual care (n = 21)], patients who had received VancApp-based dosing took a shorter time to reach target concentrations (median: 66 vs. 102 h, p = 0.187) and had fewer 30-day mortalities (14% vs. 24%, p = 0.457) compared to usual care. While statistical significance was not achieved, the clinical significance of these findings appear promising.

Conclusion: Clinical implementation of a population pharmacokinetic model for vancomycin in CKD can potentially improve dosing precision in CKD and could serve as a practical means to improve vital clinical outcomes.

Integration of Food Animal Residue Avoidance Databank (FARAD) empirical methods for drug withdrawal interval determination with a mechanistic population-based interactive physiologically based pharmacokinetic (iPBPK) modeling platform: example for flunixin meglumine administration

Violative chemical residues in animal-derived food products affect food safety globally and have impact on the trade of international agricultural products. The Food Animal Residue Avoidance Databank program has been developing scientific tools to provide appropriate withdrawal interval (WDI) estimations after extralabel drug use in food animals for the past three decades. One of the tools is physiologically based pharmacokinetic (PBPK) modeling, which is a mechanistic-based approach that can be used to predict tissue residues and WDIs. However, PBPK models are complicated and difficult to use by non-modelers. Therefore, a user-friendly PBPK modeling framework is needed to move this field forward. Flunixin was one of the top five violative drug residues identified in the United States from 2010 to 2016. The objective of this study was to establish a web-based user-friendly framework for the development of new PBPK models for drugs administered to food animals. Specifically, a new PBPK model for both cattle and swine after administration of flunixin meglumine was developed. Population analysis using Monte Carlo simulations was incorporated into the model to predict WDIs following extralabel administration of flunixin meglumine. The population PBPK model was converted to a web-based interactive PBPK (iPBPK) framework to facilitate its application. This iPBPK framework serves as a proof-of-concept for further improvements in the future and it can be applied to develop new models for other drugs in other food animal species, thereby facilitating the application of PBPK modeling in WDI estimation and food safety assessment.

An interactive application for malaria elimination transmission and costing in the Asia-Pacific

Leaders in the Asia-Pacific have endorsed an ambitious target to eliminate malaria in the region by 2030. The emergence and spread of artemisinin drug resistance in the Greater Mekong Subregion makes elimination urgent and strategic for the global goal of malaria eradication. Mathematical modelling is a useful tool for assessing and comparing different elimination strategies and scenarios to inform policymakers. Mathematical models are especially relevant in this context because of the wide heterogeneity of regional, country and local settings, which means that different strategies are needed to eliminate malaria. However, models and their predictions can be seen as highly technical, limiting their use for decision making. Simplified applications of models are needed to allow policy makers to benefit from these valuable tools. This paper describes a method for communicating complex model results with a user-friendly and intuitive framework. Using open-source technologies, we designed and developed an interactive application to disseminate the modelling results for malaria elimination. The design was iteratively improved while the application was being piloted and extensively tested by a diverse range of researchers and decision makers. This application allows several target audiences to explore, navigate and visualise complex datasets and models generated in the context of malaria elimination. It allows widespread access, use of and interpretation of models, generated at great effort and expense as well as enabling them to remain relevant for a longer period of time. It has long been acknowledged that scientific results need to be repackaged for larger audiences. We demonstrate that modellers can include applications as part of the dissemination strategy of their findings. We highlight that there is a need for additional research in order to provide guidelines and direction for designing and developing effective applications for disseminating models.

A Survey of Software Tool Utilization and Capabilities for Quantitative Systems Pharmacology: What We Have and What We Need

Quantitative systems pharmacology (QSP) is a rapidly emerging discipline with application across a spectrum of challenges facing the pharmaceutical industry, including mechanistically informed prioritization of target pathways and combinations in discovery, target population, and dose expansion decisions early in clinical development, and analyses for regulatory authorities late in clinical development. QSP's development has influences from physiologic modeling, systems biology, physiologically‐based pharmacokinetic modeling, and pharmacometrics. Given a varied scientific heritage, a variety of tools to accomplish the demands of model development, application, and model‐based analysis of available data have been developed. We report the outcome from a community survey and resulting analysis of how modelers view the impact and growth of QSP, how they utilize existing tools, and capabilities they need improved to further accelerate their impact on drug development. These results serve as a benchmark and roadmap for advancements to the QSP tool set.

Toward Better Understanding of Insulin Therapy by Translation of a PK-PD Model to Visualize Insulin and Glucose Action Profiles

Insulin replacement therapy is a fundamental treatment for glycemic control for managing diabetes. The engineering of insulin analogues has focused on providing formulations with action profiles that mimic as closely as possible the pattern of physiological insulin secretion that normally occurs in healthy individuals without diabetes. Hence, it may be helpful to practitioners to visualize insulin concentration profiles and associated glucose action profiles. Expanding on a previous analysis that established a pharmacokinetic (PK) model to describe typical profiles of insulin concentration over time following subcutaneous administration of various insulin formulations, the goal of the current analysis was to link the PK model to an integrated glucose‐insulin (IGI) systems pharmacology model. After the pharmacokinetic‐pharmacodynamic (PK‐PD) model was qualified by comparing model predictions with clinical observations, it was used to project insulin (PK) and glucose (PD) profiles of common insulin regimens and dosing scenarios. The application of the PK‐PD model to clinical scenarios was further explored by incorporating the impact of several hypothetical factors together, such as changing the timing or frequency of administration in a multiple‐dosing regimen over the course of a day, administration of more than 1 insulin formulation, or insulin dosing adjusted for carbohydrates in meals. Visualizations of insulin and glucose profiles for commonly prescribed regimens could be rapidly generated by implementing the linked subcutaneous insulin PK‐IGI model using the R statistical program (version 3.4.4) and a contemporary web‐based interface, which could enhance clinical education on glycemic control with insulin therapy.

gPKPDSim: a SimBiology®-based GUI application for PKPD modeling in drug development

Modeling and simulation (M&S) is increasingly used in drug development to characterize pharmacokinetic-pharmacodynamic (PKPD) relationships and support various efforts such as target feasibility assessment, molecule selection, human PK projection, and preclinical and clinical dose and schedule determination. While model development typically require mathematical modeling expertise, model exploration and simulations could in many cases be performed by scientists in various disciplines to support the design, analysis and interpretation of experimental studies. To this end, we have developed a versatile graphical user interface (GUI) application to enable easy use of any model constructed in SimBiology^® to execute various common PKPD analyses. The MATLAB^®-based GUI application, called gPKPDSim, has a single screen interface and provides functionalities including simulation, data fitting (parameter estimation), population simulation (exploring the impact of parameter variability on the outputs of interest), and non-compartmental PK analysis. Further, gPKPDSim is a user-friendly tool with capabilities including interactive visualization, exporting of results and generation of presentation-ready figures. gPKPDSim was designed primarily for use in preclinical and translational drug development, although broader applications exist. gPKPDSim is a MATLAB^®-based open-source application and is publicly available to download from MATLAB^® Central™. We illustrate the use and features of gPKPDSim using multiple PKPD models to demonstrate the wide applications of this tool in pharmaceutical sciences. Overall, gPKPDSim provides an integrated, multi-purpose user-friendly GUI application to enable efficient use of PKPD models by scientists from various disciplines, regardless of their modeling expertise.

Pharmacometric models simulation using NONMEM, Berkeley Madonna and R

In this tutorial, we introduce a differential equation simulation model for use in pharmacometrics involving NONMEM, Berkeley Madonna, and R. We report components of the simulation code and similarities/differences between software, rather than how to use each software. Depending on the purpose of the simulation, an appropriate tool can be selected for effective communication.

Modeling Pharmacokinetic Profiles of Insulin Regimens to Enhance Understanding of Subcutaneous Insulin Regimens

Insulin pharmacokinetics following subcutaneous administration were modeled, simulated, and displayed through an interactive and user‐friendly interface to illustrate the time course of administered insulins frequently prescribed, providing a simple tool for clinicians through a straightforward visualization of insulin regimens. Pharmacokinetic data of insulin formulations with different onset and duration of action from several clinical studies, including insulin glargine, regular insulin, neutral protamine Hagedorn (NPH), insulin lispro, and premixed preparations of NPH with regular insulin (Mix 70/30), and insulin lispro protamine suspension with insulin lispro (Mix 50/50, Mix 75/25), were used to develop a predictive population pharmacokinetic model of insulins with consideration of factors such as insulin formulation, weight‐based dosing, body‐weight effect on volume of distribution, and administration time relative to meals, on the insulin time‐action profile. The model‐predicted insulin profile of each insulin was validated and confirmed to be comparable to observed data via an external validation method. Model‐based simulations of clinically relevant insulin‐dosing scenarios to cater to specific initial patient and prescribing conditions were then implemented with differential equations using the R statistical program (version 3.2.2). The R package Shiny was subsequently applied to build a web browser interface to execute and visualize the model simulation outputs. The application of insulin pharmacokinetic modeling enabled informative visualization of insulin time‐action profiles and provided an efficient and intuitive educational tool to quickly convey and interactively explore many insulin time‐action profiles to ease the understanding of insulin formulations in clinical practice.

Population Pharmacokinetic Model of Doxycycline Plasma Concentrations Using Pooled Study Data

The literature presently lacks a population pharmacokinetic analysis of doxycycline. This study aimed to develop a population pharmacokinetic model of doxycycline plasma concentrations that could be used to assess the power of bioequivalence between Doryx delayed-release tablets and Doryx MPC. Doxycycline pharmacokinetic data were available from eight phase 1 clinical trials following single/multiple doses of conventional-release doxycycline capsules, Doryx delayed-release tablets, and Doryx MPC under fed and fasted conditions. A population pharmacokinetic model was developed in a stepwise manner using NONMEM, version 7.3. The final covariate model was developed according to a forward inclusion (P < 0.01) and then backward deletion (P < 0.001) procedure. The final model was a two-compartment model with two-transit absorption compartments. Structural covariates in the base model included formulation effects on relative bioavailability (F), absorption lag (ALAG), and the transit absorption rate (KTR) under the fed status. An absorption delay (lag) for the fed status (FTLAG2 = 0.203 h) was also included in the model as a structural covariate. The fed status was observed to decrease F by 10.5%, and the effect of female sex was a 14.4% increase in clearance. The manuscript presents the first population pharmacokinetic model of doxycycline plasma concentrations following oral doxycycline administration. The model was used to assess the power of bioequivalence between Doryx delayed-release tablets and Doryx MPC, and it could potentially be used to critically examine and optimize doxycycline dose regimens.

Pharmacokinetic models of morphine and its metabolites in neonates:: Systematic comparisons of models from the literature, and development of a new meta-model

Morphine is commonly used for pain management in preterm neonates. The aims of this study were to compare published models of neonatal pharmacokinetics of morphine and its metabolites with a new dataset, and to combine the characteristics of the best predictive models to design a meta-model for morphine and its metabolites in preterm neonates. Moreover, the concentration-analgesia relationship for morphine in this clinical setting was also investigated. A population of 30 preterm neonates (gestational age: 23-32weeks) received a loading dose of morphine (50-100μg/kg), followed by a continuous infusion (5-10μg/kg/h) until analgesia was no longer required. Pain was assessed using the Premature Infant Pain Profile. Five published population models were compared using numerical and graphical tests of goodness-of-fit and predictive performance. Population modelling was conducted using NONMEM® and the $PRIOR subroutine to describe the time-course of plasma concentrations of morphine, morphine-3-glucuronide, and morphine-6-glucuronide, and the concentration-analgesia relationship for morphine. No published model adequately described morphine concentrations in this new dataset. Previously published population pharmacokinetic models of morphine, morphine-3-glucuronide, and morphine-6-glucuronide were combined into a meta-model. The meta-model provided an adequate description of the time-course of morphine and the concentrations of its metabolites in preterm neonates. Allometric weight scaling was applied to all clearance and volume terms. Maturation of morphine clearance was described as a function of postmenstrual age, while maturation of metabolite elimination was described as a function of postnatal age. A clear relationship between morphine concentrations and pain score was not established.

A Tutorial on RxODE: Simulating Differential Equation Pharmacometric Models in R

This tutorial presents the application of an R package, RxODE, that facilitates quick, efficient simulations of ordinary differential equation models completely within R. Its application is illustrated through simulation of design decision effects on an adaptive dosing regimen. The package provides an efficient, versatile way to specify dosing scenarios and to perform simulation with variability with minimal custom coding. Models can be directly translated to Rshiny applications to facilitate interactive, real‐time evaluation/iteration on simulation scenarios.

MatVPC: A User-Friendly MATLAB-Based Tool for the Simulation and Evaluation of Systems Pharmacology Models

Quantitative systems pharmacology (QSP) models are progressively entering the arena of contemporary pharmacology. The efficient implementation and evaluation of complex QSP models necessitates the development of flexible computational tools that are built into QSP mainstream software. To this end, we present MatVPC, a versatile MATLAB-based tool that accommodates QSP models of any complexity level. MatVPC executes Monte Carlo simulations as well as automatic construction of visual predictive checks (VPCs) and quantified VPCs (QVPCs).

Population pharmacokinetic modeling of itraconazole and hydroxyitraconazole for oral SUBA-itraconazole and sporanox capsule formulations in healthy subjects in fed and fasted states

Itraconazole is an orally active antifungal agent that has complex and highly variable absorption kinetics that is highly affected by food. This study aimed to develop a population pharmacokinetic model for itraconazole and the active metabolite hydroxyitraconazole, in particular, quantifying the effects of food and formulation on oral absorption. Plasma pharmacokinetic data were collected from seven phase I crossover trials comparing the SUBA-itraconazole and Sporanox formulations of itraconazole. First, a model of single-dose itraconazole data was developed, which was then extended to the multidose data. Covariate effects on itraconazole were then examined before extending the model to describe hydroxyitraconazole. The final itraconazole model was a 2-compartment model with oral absorption described by 4-transit compartments. Multidose kinetics was described by total effective daily dose- and time-dependent changes in clearance and bioavailability. Hydroxyitraconazole was best described by a 1-compartment model with mixed first-order and Michaelis-Menten elimination for the single-dose data and a time-dependent clearance for the multidose data. The relative bioavailability of SUBA-itraconazole compared to that of Sporanox was 173% and was 21% less variable between subjects. Food resulted in a 27% reduction in bioavailability and 58% reduction in the transit absorption rate constant compared to that with the fasted state, irrespective of the formulation. This analysis presents the most extensive population pharmacokinetic model of itraconazole and hydroxyitraconazole in the literature performed in healthy subjects. The presented model can be used for simulating food effects on itraconazole exposure and for performing prestudy power analysis and sample size estimation, which are important aspects of clinical trial design of bioequivalence studies.