Ph.D Student University of Florida - Center for Pharmacometrics & Systems Pharmacology Orlando, United States
Disclosure(s):
Anthony Gebhart: No financial relationships to disclose
Objectives: In the realm of drug development, understanding the relationship between drug exposure and response is paramount for optimizing therapeutic outcomes and minimizing adverse effects[1,2]. We aimed to create a user-friendly Shiny application tailored for exposure-response analysis. By providing researchers, regardless of their programming proficiency, with an accessible tool, the goal is to expedite decision-making in early-phase drug development, accelerating the review of data from clinical trials, facilitating the identification of optimal dosing regimens and enhancing the understanding of drug efficacy and safety profiles.
Methods: An automated exploration app in R-Shiny was developed to standardize and assist the exploratory exposure-response analysis. The versatile functionalities provided by bslib, RStudio, and RMarkdown to develop dynamic and interactive dashboards were utilized. Taking advantage of the integrated features of these tools, engaging and customizable dashboard interfaces can be created. bslib enriches Shiny apps with enhanced appearances and responsiveness by integrating Bootstrap themes. Coupled with Shiny's reactive capabilities, it allows to create visually appealing and reactive interfaces that can respond dynamically to changes in data or user input in real-time. We have employed the adaptable and robust ggplot2 package to design visually captivating and informative plots. By leveraging its extensive range of functionalities, we have crafted dynamic visualizations that effectively communicate essential insights and trends present within the dataset. This comprehensive approach enables us to create engaging graphics that enhance the understanding of key findings, providing a valuable tool for data analysis and interpretation.
Results: The R-Shiny app “ERpmx” incorporates diverse exposure-response models, including Emax, logistic, and Kaplan-Meier, enabling researchers to assess dose-response relationships across therapeutic contexts comprehensively. Key to its functionality is an intuitive interface allowing effortless visualization of exposure metrics distributions across different doses. Interactive features enable users to display simultaneously multiple endpoints and multiple metrics like AUC or Cmax, providing nuanced insights into drug efficacy and toxicity profiles. Additionally, the application facilitates the projection of exposure metrics distributions onto the model fit, offering a deeper understanding of the relationship between exposure and response. It also facilitates data stratification by time or covariates, enhancing treatment tailoring for specific patient subpopulations, and supports partitioning data into quantile groups for refined analysis. With the capability to load multiple datasets, it enhances its utility in comparative analyses across experimental conditions and patient cohorts, thereby reinforcing the evidence base for drug development decisions.
Citations: [1] L.A. Cox Jr. What is an exposure-response curve. Global Epidemiology 6 (2023). [2] R.V. Overgaard et al. Establishing Good Practices for Exposure-Response Analysis of Clinical Endpoints in Drug development. CPT Pharmacometrics Syst. Pharmacol. 4 (2015).
