At a Glance
- To manage the potential data security and information accuracy risks associated with large language models, the European Medicines Agency’s guiding principles emphasize that medical staff using these tools should apply critical thinking and human cross-checking of outputs, while also encouraging regulatory agencies to make efforts to provide training to medical staff and monitor potential risks of these models.
- A recently updated FDA guidance addressing “Control of Nitrosamine Impurities in Human Drugs” gives FDA’s current views on how drug product manufacturers as well as supply-chain entities should seek to eliminate or acceptably mitigate nitrosamine risks in drug products. New information includes descriptions of nitrosamine drug substance-related impurities, potential root causes, and strategies to prevent or reduce the presence of these compounds.
- The Biden administration’s new federal research strategy identifies four strategic goals that will drive future federal research and development efforts on PFAS.
- The FDA’s Center for Drug Evaluation and Research (CDER) has established a Quantitative Medicine Center of Excellence, which will function across CDER as a point of coordination for various quantitative medicine-related efforts progressing through the agency. FDA appears to be responding to the rapid innovations and applications of quantitative medicine methods, tools and approaches in submissions as well as to help improve and streamline drug development.
With the rapidly evolving regulatory landscape, staying on top of the latest updates is crucial for companies in biopharma and medical devices. In this briefing, we highlight recent regulatory and legislative developments that should be top-of-mind for companies in these sectors.
Regulatory and Legislative Developments
Guidance on Use of Large Language Models (LLMs) in Medical Practice
The European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) recently published guiding principles and recommendations for the use of LLMs in medical practice across the European Medicines Regulatory Network (EMRN). LLMs, a type of generative artificial intelligence (AI) model, may have a variety of uses in medical practice, including automating data mining processes, documenting general regulatory or administrative information, as virtual assistants, or other everyday administrative tasks.
To manage the potential data security and information accuracy risks associated with LLMs, the EMA’s guiding principles emphasize that medical staff using these tools should apply critical thinking and human cross-checking of outputs, while also encouraging regulatory agencies to make efforts to provide training to medical staff and monitor potential risks of these models. These principles are part of the broader deliverables of the EMA’s multiannual AI work plan through 2028 and support requirements under the EU AI Act.
FDA Guidance on Control of Nitrosamine Impurities in Human Drugs
Although nitrosamine levels in drug products have been a focus of regulation for many years, the U.S. Food and Drug Administration (FDA) and other regulatory agencies have increased their scrutiny over the last eight years when unacceptable levels were found in medicines, including some for high blood pressure, heartburn and diabetes. The FDA regulates nitrosamine levels due to their potential carcinogenicity, and has set revised acceptable human intake limits for various types of nitrosamines.
A recently updated FDA guidance addressing “Control of Nitrosamine Impurities in Human Drugs" gives FDA’s current views on how drug product manufacturers as well as supply-chain entities should seek to eliminate or acceptably mitigate nitrosamine risks in drug products. Importantly the guidance describes considerations for evaluating and mitigating the risk of nitrosamine formation in the synthesis of active pharmaceutical ingredients and excipients, and in the formulation and manufacture of the final drug product. The guidance also includes considerations for applying confirmatory testing based on risk assessment outcomes, and application of current acceptable intake limits for nitrosamines. New information includes descriptions of nitrosamine drug substance-related impurities (NDSRIs), potential root causes, and strategies to prevent or reduce the presence of these compounds.
A related page describes acceptable intake limits for a number of specific NDSRIs; non-API (active pharmaceutical ingredient) related nitrosamine impurities; considerations for communications with FDA; and the possibility of interim limits for nitrosamines found in currently marketed drugs, to reduce the chances of drug shortages.
The FDA Center for Drug Evaluation has a nitrosamine information hub with links to guidance, patient information, frequently asked questions and updates.
New Federal Research Strategy on PFAS
On September 3, 2024, the Biden administration released a new federal research strategy on per- and polyfluoroalkyl substances (PFAS). The strategy is another action taken by the Biden administration on PFAS that should be considered along with the multiple regulatory and federal procurement actions the administration has taken throughout the past four years. The new federal research strategy identifies four strategic goals that will drive future federal research and development (R&D) efforts on PFAS:
- Providing scientific data to increase the understanding of PFAS exposures to inform federal decisions
- Effectively communicating federal work and results by engaging the public and Tribal, state and local agencies
- Identifying research and technologies to address PFAS contamination and mitigating adverse impacts on communities
- Generating information that helps federal procurement decisions by federal agencies, manufacturers and consumers regarding products that contain PFAS
CDER Quantitative Medicine Center of Excellence
The FDA Center for Drug Evaluation and Research (CDER) has established a Quantitative Medicine Center of Excellence, which will function across CDER as a point of coordination for various quantitative medicine-related efforts progressing through the agency, including agency research and application of aspects of quantitative medicine in drug product reviews.
FDA defines quantitative medicine as the “application of quantitative methods and computational techniques to better understand diseases and treat patients,” and can include a large number of applications across the drug development process, including artificial intelligence / machine learning approaches to model and better understand the effect of quality control parameters and pharmacokinetic modeling to support bioequivalence determinations.
FDA appears to be responding to the rapid innovations and applications of quantitative medicine methods, tools and approaches in submissions as well as to help improve and streamline drug development.
