In 2023, the U.S. Food and Drug Administration (FDA) established the Digital Health Advisory Committee (DHAC) to consider topics related to the use of digital tools in drug development. On November 20 and 21st, 2024 the committee convened for the first time to discuss total product lifecycle (TPLC) considerations for generative artificial intelligence (GenAI) enabled medical devices. The establishment of the DHAC and this meeting come in response to growing interest and use of these technologies in healthcare and signal a recognition of the unique regulatory challenges presented by GenAI- and AI-enabled devices.
Source: U.S. FDA. Executive Summary for the Digital Health Advisory Committee Meeting: Total Product Lifecycle Considerations for Generative AI-Enabled Devices. November 20-21, 2024.
The two-day meeting highlighted industry, regulator, patient, and academic perspectives on the risks and complexities of GenAI-enabled devices that challenge current approaches to evidence generation and regulatory evaluation and monitoring. Presenters and panelists considered the unique evidentiary needs and methods for premarket performance evaluation and validation, novel risks associated with the use of GenAI-enabled devices in healthcare and approaches for mitigating such risks, and metrics and strategies for postmarket performance monitoring and re-evaluating GenAI-enabled devices. The common thread throughout the meeting was that the current TPLC approach to regulating medical devices may need to be adapted to account for the unique risks and evidentiary needs of GenAI-enabled devices.
Friends work to date in this area aims to inform the use of AI-enabled tools and data sets for assessing performance of AI-enabled devices used in oncology. In 2023, we convened a multi-stakeholder working group to evaluate existing regulatory frameworks and developed proposals to support robust development of AI-enabled computational pathology platforms used in oncology drug development. The proposals outlined by the working group include a risk-based approach to assessing evidentiary needs for tool validation. Friends continues to lead collaborative projects aimed at supporting the establishment of common reference data sets and exploring their use in evaluating the performance of tools, thereby supporting future platform development and cross-validation.
The DHAC meeting provided an insightful discussion around evidentiary and regulatory needs of GenAI-enabled, and AI-enabled, devices and Friends looks forward to future collaborative opportunities to ensure the safe and effective use of these transformative technologies.
