Presentation (CTI)

OBJECTIVES: This abstract introduces an integrated framework that combines the target trial emulation (TTE) framework, synthetic data (SYND) generation, and modeling & simulation (M&S) approaches, which we refer to as target trial optimization (TTO). This novel framework aims not to replicate a target trial using real-world data (RWD), but rather to optimize its design.
METHODS: We begin with TTE to identify the most relevant data source (observed data) in the SYND generation workflow, ensuring that the selected data aligns with the “baseline” trial’s estimand. Next, SYND is generated through four stages: data preprocessing, model training, data sampling, and post-processing. SYND are trained on a subset of the observed data, while a separate portion is used for validation. This process creates synthetic patient profiles that mirror the observed data. We further enhance the framework by integrating M&S techniques, allowing for the exploration of trial conditions, including population composition, treatment regimens, and design assumptions. What-if scenario analysis is used to boost the “optimal” trial parameters.
RESULTS: Applying the TTO framework using SYND and M&S enables rapid exploration of design alternatives prior to trial initiation. Starting from the “baseline” trial defined by TTE, we simulate and evaluate multiple design-analysis pairs, changing eligibility criteria, treatment strategies, follow-up, outcomes, and estimands, to define the “optimal” trial to target in development.
CONCLUSIONS: This TTE-to-TTO transition allows the refinement of key design elements based on feasibility, power, and robustness. Proactive exploration of the design space can shorten protocol development timelines. Early tuning of eligibility criteria. optimization of the study duration, outcomes and estimands can reduce screen failure rates, shorten studies and reduce required sample size, Combined, these efficiencies translate into early development savings. With fewer protocol amendments, faster first-patient-in, improved recruitment efficiency, and greater likelihood of regulatory and scientific acceptance, this framework enables more strategic and cost-effective trial design.