Quantitative Benefit-Risk Assessment in Rare Disease

Traditionally a benefit-risk assessment of a novel therapy involves the separate assessment of each critical benefit and risk endpoint on a population-level, followed by the combination of these marginal assessments into a value judgement.
However, several important challenges arise by separately assessing each benefit and risk.
First, the association between the benefits and risk is ignored, preventing crucial assessments, such as: Are patients benefiting from the therapy more or less likely to experience risks?
Second, the benefit and risk may not always be assessed in the same population, which raises the question to which population the benefit-risk assessment is applicable. Next, clinical events may be double-counted, potentially over-or underestimating the true benefit-risk balance. Finally, the process of preference elicitation—determining how different benefits and risks should be weighted relative to each other—is complex and can introduce subjectivity into the analysis.
The generalized pairwise comparisons (GPC) method form a very flexible class of non-parametric techniques for the analysis of prioritized endpoints, which includes benefit-risk assessments. Individual-level benefit-risk assessment within prioritized GPC overcomes the limitations of traditional population level assessments. It involves ranking outcomes, which is a more straightforward process than weighting, and it effectively addresses issues related to double-counting and the association between outcomes. Moreover, it allows for sensitivity analyses with varying priorities and thresholds, providing a more comprehensive view of the benefit-risk balance, and facilitating patient-centric decision-making.
Theoretical developments point to the good small sample properties of GPC, making it a useful tool in rare diseases. Additionally, patient-centric endpoints, such as patient-reported outcomes can be incorporated in a prioritized GPC endpoint, enhancing decision-making and aligning drug development and administration with patient needs and priorities.
In this presentation, we illustrate the potential of prioritized GPC as a valuable evaluation tool for the benefit-risk assessment of novel therapies, especially in the context of rare diseases. Through re-analyses of clinical trial data, including the integration of patient-centric endpoints such as PROs, we demonstrate how GPC can support more informed and patient-centered decision-making in rare diseases. By leveraging GPC's ability to handle multivariate outcomes at the individual level, researchers and clinicians can gain a clearer, more accurate understanding of a therapy's potential impact, ultimately advancing the pursuit of precision medicine and improving outcomes for patients with rare and complex conditions
 

Assessing the benefit-risk profile is essential throughout the entire lifecycle of a medical product to ensure that the benefits outweigh the risks (FDA. 2023. Benefit-risk assessment (BRA) for new drug and biological products: Guidance for industry: Final guidance). Conducting rigorous, evidence-based BRA can be challenging, as it involves making these assessments transparent and entirely measurable during development, approval, and post-marketing stages. These difficulties are more pronounced for rare diseases due to limited availability of clinical data during their development phases. Furthermore, the lack of control data adds complexity to BRA. This presentation addresses how to achieve a transparent and fully measurable benefit-risk balance in such cases using an innovative Bayesian approach. The approach utilizes both stakeholder preference elicitation and clinical trial results and accounts for uncertainties in those measures. A discussion on an R-shiny application for implementing this methodology will also be included.