University of Minnesota twin cities, Minnesota, United States
Background: Weight-based dosing is commonly used to account for the body weight effect on exposure and ensure comparable exposure across the weight ranges. However, this approach can lead to drug waste and complicate drug preparation and administration. Weight-banded dosing allows the administration of the same dose across a weight band and provides the option to use auto-injectors as an alternative dosing strategy. Bridging from weight-based dosing to weight-band dosing is feasible if the proposed weight-band dosing strategy leads to comparable exposures to the weight-based dose with clinically proven safety and efficacy. This project aims to develop an automatic R Shiny tool for pharmacokinetics (PK) comparison to improve communication and decision-making regarding weight bands and doses. Methods: This app was developed using the shiny package (v4.3.2) in R. The mrgsolve package was used to simulate various weight-based and weight-banded dosing scenarios. Data manipulation and visualization mainly used the tidyverse package suite. Results: The app requires the users to upload a population pharmacokinetics (PopPK) model file and an optional covariate dataset to define a virtual population. The app allows the users to explore different weight-banded dosing scenarios with multiple dose levels and weight cutoffs and compare the simulated exposures against the weight-based dosing as a reference. If the test population is different from the reference population, the app also provides the option to directly input summary statistics of the exposure for the reference population. To facilitate initial conceptualization of weight-band cutoffs and doses, the app features summary statistics of the weight in the virtual population and a visualization of dose by weight. Finally, the app provides comparison plots (boxplots, ribbons with median) and summary tables for exposures (e.g., Cmax, Cavg, Ctrough, AUC) for download, to demonstrate similarity in exposure between the weight-banded and weight-based scenarios. Conclusion: This user-friendly R Shiny app simplifies the simulation process and enables exploration of alternative dosing strategies and efficient communication with a broader audience. Future plans include exploring potential AI approaches to suggest tailored weight bands and initial doses.
