PS06: Collaboration Between Health Tech and Biopharma to Advance Patient-Centric Medical Product Development with Digital Endpoints

The number of digital endpoints collected in registered interventional trials to assess efficacy and safety has greatly increased over the last five years from 64 endpoints collected by 12 sponsors, to 505 endpoints collected by 80 sponsors, according to the crowdsourced library by the Digital Medicine Society. The expanded adoption of digital health technology (DHT)-derived endpoints is driven by technological advancements in wearable sensors and new AI/ML techniques to process raw sensor data. It is also the result of the growing emphasis on patient-focused development of medical products by regulatory guidelines resulting from FDA's commitments to PDUFA VII. However, despite the growing adoption of digital endpoints, the majority of them are still used as exploratory endpoints in clinical trials and there is an evidentiary gap that limits digital endpoints from becoming primary/secondary endpoints. Closing this gap requires close collaboration between health tech and biopharma, and early engagement with regulatory agencies to determine an appropriate evidence generation and evaluation plan. The proposed evidentiary checklist to close this gap includes: device verification, fit-for-purpose justification, analytical validation in the specific clinical population, reliability characterization, and clinical validation consisting of sensitivity in capturing longitudinal changes, sensitivity in detecting interventional effects, and establishing anchor-based minimum clinical important difference (MCID). This session aims to engage key stakeholders in the field of digital endpoints from industry and FDA to share their experiences and insights through case studies. The session consists of three presentations. The first speaker from the FDA will share regulatory perspectives and recent guidance to ensure the proper development, validation, and application of DHT-derived endpoints. The second speaker from health tech will share first-hand experience deriving digital measures from raw sensor data and clinically validating them as fit-for-purpose endpoints for clinical trial deployment. They will also share a recent case study in quantifying Parkinson's disease progression where digital measures outperform gold-standard clinician rating in terms of clinical effect size and test-retest reliability. The third speaker from biopharma will discuss the unmet needs of measuring disease modification with conventional endpoints in neurological disorders, and the efforts to build novel measures that can enable efficient drug development through pre-competitive collaborations. They will also share perspective on utilizing digital measures for internal decision making purposes. This session is particularly relevant to this year's workshop theme of "Future in statistics: collaboration and innovation in the AI/ML Era" by sharing successful case studies of tech-pharma collaboration in leveraging AI/ML innovation to advance patient-centric medical product development, and subsequently, encouraging wider and proper adoption of technological innovation to facilitate therapeutic discovery.