PS11: Perspectives on Using in Silico Simulation in Oncology Drug Development

Clinical trial simulation is increasingly used to enhance development efficiency and improve the probability of success. Modern oncology trials are characterized by complex designs, including evolving standards of care, delayed treatment effects, intercurrent events such as post-progression therapies, predictive subgroups, and multiple interrelated endpoints. In many such settings, closed-form analytical solutions are unavailable. Simulation provides a flexible and dynamic framework for strategic trial planning and decision-making when traditional analytic approaches are insufficient. Simulation plays a critical role in oncology drug development because the most clinically meaningful endpoint-overall survival (OS)-often takes years to mature. Consequently, early development and interim decision-making frequently rely on short-term or intermediate endpoints, such as objective response rate (ORR), duration of response (DoR), and progression-free survival (PFS), as well as biomarker-based outcomes (e.g., PSA progression in prostate cancer, ctDNA clearance, and minimal residual disease). In this context, simulation studies provide a structured framework to evaluate and project long-term OS outcomes based on earlier clinical and biomarker endpoints, which is essential for the design and success of large-scale pivotal trials. Simulation-based modeling frameworks, such as multi-state models, enable the generation of individual patient trajectories and coherent modeling of the joint distribution of multiple early endpoints and OS. These frameworks allow drug developers to project long-term outcomes from early data and, in turn, inform key design considerations, including selection of optimal testing strategies to improve power and assessment of potential OS harm at interim PFS readouts. Simulation also plays a critical role in adaptive oncology trial designs. Applications include population enrichment strategies when treatment effects vary across subgroups, interim go/no-go decisions, sample size re-estimation, dose adaptation in seamless designs, and evaluation of operating characteristics such as type I error control, probability of success, and timing of adaptation. In this session, we will bring together perspectives from industry, academia, and regulatory agencies to share practical experiences in using simulation for strategic trial planning, as well as regulatory considerations for ensuring statistical validity and interpretability of increasingly complex trial designs. Through these perspectives, we will discuss how in silico simulation can serve as a common framework to enable efficient drug development while maintaining the rigor required for regulatory decision-making.