Chair: Kevin Gan, PhD (ViiV Healthcare)
Abstract: This session will explore how Bayesian model-based approaches are transforming the way we understand drug exposure and optimize dosing strategies across therapeutic areas. We’ll discuss how integrating prior knowledge with emerging short-term data enables early prediction of long-term exposure, supports strategic decision-making, and addresses key challenges in trial design and regulatory planning.
Speaker: Sarah Ji, PhD (Glaxo Smith Kline | ViiV Healthcare)
Title: Bayesian Estimation of Time to Steady State Plasma Trough Concentrations in HIV
Abstract: Estimation of time to steady-state drug concentrations in plasma is typically performed retrospectively, after all study data have been collected. We identified a Bayesian PK modeling approach in the literature that enables prospective estimation of the time to steady-state drug concentrations for study designs where the same maintenance dose is administered at constant dosing intervals. Building on the Bayesian framework from the referenced paper, our team explored various adjustments to this model to better suit study designs for long-acting injectables for HIV prevention (PrEP).
Speaker: Hongtao Zhang, PhD (Merck)
Title: Prior Effective Sample Size When Borrowing on the Treatment Effect Scale
Abstract: With the robust uptick in the applications of Bayesian external data borrowing, eliciting a prior distribution with the proper amount of information becomes increasingly critical. The prior effective sample size (ESS) is an intuitive and efficient measure for this purpose. The majority of ESS definitions have been proposed in the context of borrowing control information. Meanwhile, Bayesian borrowing is frequently conducted on the treatment effect scale to extrapolate evidence in pediatric or global trials. While many Bayesian models can be naturally extended to leveraging external information on the treatment effect scale, very little attention has been directed to computing the prior ESS in this setting. In this research, we bridge this methodological gap by extending the popular expected local information ratio (ELIR) ESS definition. We lay out the general framework, and derive the ESS for various types of endpoints and treatment effect measures. The desirable predictive consistency property of ELIR ESS is examined and found to only be preserved for the difference between two normal endpoints.
Speaker: Hongmei Zhang, PhD (University of Memphis)
Title: Clustering with Respect to Undirected Networks
Abstract: It has been suggested that genes work in concert and such concerted work can be described using networks, directed or undirected, to reflect the connection and strength of connection among genes or epigenetic sites such as Cytosine-phosphate-Guanine (CpG) sites. Evidence has shown that networks among genetic or epigenetic factors can be different for different populations. Cluster analyses can help identify potential underlying joint activities among nodes (e.g., genes) unique to a population and findings will benefit prediction of disease risk at a much early stage. We propose a clustering approach driven by network structures and strength of network edges to cluster observations, denoted as Multi-center Graph Clustering (McGC). In the present work, we focus on undirected networks to learn joint activities among genes and, to ease the incorporation of prior knowledge into the clustering process, the approach is developed under the Bayesian framework. Simulations are applied to assess the feasibility of McGC. We apply McGC to DNA methylation data collected in a birth cohort established on the Isel of Wight cohort in the United Kingdom to identify clusters of subjects based on joint activities of DNA methylation loci and examine the association of the clusters with allergic health conditions.
