On Wednesday, May 10, 2023, the Food and Drug Administration (FDA) announced the publication of a new discussion paper titled “Using Artificial Intelligence and Machine Learning in the Development of Drug and Biological Products.” The discussion paper aims to facilitate a discussion with stakeholders on the use of artificial intelligence and machine learning (AI/ML) in drug development, including the development of medical devices intended for use with drugs. The discussion paper covers three main topics: the landscape of current and potential uses of AI/ML; considerations for the use of AI/ML; and next steps and stakeholder engagement.
Here are five key takeaways:
1. FDA Acknowledges that AI/ML Applications Exist for Each Stage of Drug Development. AI/ML may have applications for use in all stages of drug development, from drug discovery to pharmaceutical manufacturing. AI/ML has been applied to real‑world data (RWD) and data from digital health technologies (DHTs) to support drug development. The first section of the discussion paper summarizes the different ways in which AI/ML can be used in drug discovery, clinical and nonclinical research, postmarketing surveillance, and advanced pharmaceutical manufacturing.
2. FDA Is Growing Its Own Experience with AI/ML for Drug Development. In recent years, the FDA has seen a growth in drug and biological product submissions that reference AI/ML. In response, FDA has taken a number of actions, including establishment of the CDER AI Steering Committee, the Innovative Science and Technology Approaches for New Drugs (ISTAND) Pilot Program, and the Model‑Informed Drug Development (MIDD) Pilot Program. For postmarket safety surveillance, the CDER Sentinel System, CBER Biologics Effectiveness and Safety (BEST) system, and CDRH National Evaluation System for health Technology (NEST) efforts, are exploring AI/ML approaches to improve existing systems.
3. FDA Understands Development of Standards and Practices for Use of AI/ML Is Critical. The U.S. government and international community have expressed an increased commitment to facilitating AI innovation and adoption. Regulators and standards organizations have developed and issued standards to facilitate the advancement of ethical AI. For example, in August 2019, the National Institute for Standards and Technology (NIST) released the “U.S. Leadership in AI: A Plan for Federal Engagement in Developing Technical Standards and Related Tools.” In addition, in October 2021, the FDA, Health Canada, and the UK’s Medicines and Healthcare products Regulatory Agency (MHRA) jointly published 10 guiding principles to inform the development of Good Machine Learning Practices (GMLP) for medical devices using AI/ML.
4. FDA Has Identified Key Questions on AI/ML in Drug Development. The FDA aims to initiate a discussion with stakeholders on the following three key areas and provides specific questions to solicit feedback.
Human-Led Governance, Accountability, and Transparency
- In what specific use cases or applications of AI/ML in drug development are there the greatest need for additional regulatory clarity?
- In your experience, what are the main barriers and facilitators of transparency with AI/ML used during the drug development process (and in what context)?
- How are pre-specification activities managed, and changes captured and monitored, to ensure the safe and effective use of AI/ML in drug development?
Quality, Reliability, and Representativeness of Data
- What additional data considerations exist for AI/ML in the drug development process?
- What are some of the key practices utilized by stakeholders to help ensure data privacy and security?
- What processes are developers using for bias identification and management?
Model Development, Performance, Monitoring, and Validation
- What practices and documentation are being used to inform and record data source selection and inclusion or exclusion criteria?
- In what context of use are stakeholders addressing explainability, and how have you balanced considerations of performance and explainability?
- What are some examples of current tools, processes, approaches, and best practices being used by stakeholders for: selecting model types and algorithms for a given use, determining when to use specific approaches for validating models and measuring performance in a given context, evaluating transparency and explainability, and increasing model transparency, etc.?
5. FDA Wants Your Feedback. The FDA is soliciting feedback on the opportunities and challenges in using AI/ML in drug and medical device development. The FDA includes a series of questions for feedback in the discussion paper, and a workshop with stakeholders is planned to provide an opportunity for further engagement. Comments must be submitted by August 9, 2023 (Docket No. FDA-2023-N-0743).