Industry stakeholders call for changes to ICH M14 RWD guideline

Pharmaceutical industry stakeholders from both sides of the Atlantic generally supported the International Council for Harmonization’s (ICH) M14 guidance on using real world data (RWD) for pharmacoepidemiological studies, but suggested areas for improvement, such as adding more examples of how these principles will work in practice, addressing the use of artificial intelligence and machine learning (AI/ML), and expanding its scope to include combination products.

Groups said the guideline should address some of the drawbacks of using federated data networks (FDNs) for collecting RWD in their comments to the European Medicines Agency (EMA) and the US Food and Drug Administration (FDA).

The ICH M14 guideline was issued for public consultation in May 2024, and sets out principles for using real-world data in safety assessments. The guidance sets out a multidisciplinary approach to harmonize post-approval non-interventional studies for drugs, vaccines, and other biologics. (RELATED: ICH adopts M12 guideline on drug interaction studies, releases draft M14 guidance on RWD, Regulatory Focus 28 May 2024)

Pharmaceutical industry express overall support

The guideline was generally supported by three major pharmaceutical groups in both regions: the Pharmaceutical Research and Manufacturers of America (PhRMA), the International Society for Pharmaceutical Engineering (ISPE), and the European Federation of Pharmaceutical Industries and Associations (EFPIA).

PhRMA wrote that “the guidance will help streamline the development and regulatory assessment of postmarketing pharmacoepidemiological safety studies that include Real-World Data (RWD), facilitate acceptance of study protocols and results across health authorities, and also support decision making in response to study results.”

ISPE noted in their comment to EMA that guidance “is well-written, promoting the harmonization of conducting pharmacoepidemiological studies. The high quality, clarity, and outlined recommendations reflect the hard work and dedication that went into its development. The comprehensive nature of the document, along with its clear, actionable guidance, demonstrates a deep understanding of the complexities involved. This guidance will help sponsors understand expectations for pharmacoepidemiology submissions and facilitate the assessment of the validity of such submissions by scientific reviewers.”

In their comments to the EMA, EFPIA stated that “our statisticians consider that this document, as its name suggests, is a good guide for the generalities of planning and designing pharmacoepidemiological studies using RWD.”

Practical challenges

Yet EFPIA said in its comments to EMA that more examples are needed to illustrate how the principles in the guideline can operate in practice.

The group said that “the recommendations included in this guidance are laudable but represent an ideal that is challenged at several points by the practical reality of regulatory timelines and expectations. Without concrete evaluation of the operational complexities that need to be navigated and solutions to address them, execution of at least some of the best practices outlined in this guidance will remain out of reach.”

Artificial intelligence

In their comments to FDA, the Association for Accessible Medicines (AAM) said the guidance should address the use of AI. The guidance states that because the of the “evolving nature of pharmacogenomics, artificial intelligence (AI), and other emerging technologies” relevant to RWD, these topics are out of scope.”

However, AAM argued that “it would be helpful if the final guidance would include some information and recommendations that reflect how health authorities will consider the use of these emerging technologies. We understand that this is a new and emerging area, but future studies based on RWD are likely to employ AI/ML technologies. For example, Natural Language Processing models are emerging to extract data from EHRs [electronic health records] for transformation into structured data fields. Some health authorities have begun to consider these technologies, and more detail about them would help inform deployment in studies based on RWD.”

Combination products

In their comments, PhRMA and the Biotechnology Industry Organization (BIO) suggested the scope of the guidance include combination products.

PhRMA states that “the objective of the guidance is to provide considerations for the general principles on non-interventional pharmacoepidemiological studies that utilize fit-for-purpose data for safety assessment of medicines (where medicines encompass drugs, vaccines, and other biological products); however, the draft guidance does not explicitly reference combination products.”

BIO added that “it would be helpful for the guideline to indicate whether combination products are within scope.”

Federated data networks

Both EFPIA and ISPE, in their comments to EMA, said that the guideline should present more balanced information on how sponsors can use federated data networks (FDNs) for collecting RWD.

The guidance said FDNs can “enable distributed analyses combining data or results across multiple databases” and that these networks can provide “unique advantages that can assist with addressing drug safety questions.”

EFPIA suggested that when discussing the benefits of FDNs, or other RWD sources, the guideline should also address some of the disadvantages of using these networks. Such balanced information “will enable readers to understand full context of RWD options for a given situation.”

ISPE agreed and said the guidance should address “current limitations of FDNs.” These limitations include underlying differences in original data sources and healthcare systems, operational complexity, and lack of industry access to FDA’s Sentinel system and the EMA’s and Darwin systems.

Comments to FDA; Comments to EMA

Industry stakeholders call for changes to ICH M14 RWD guideline

Related topics

Recommended content

Welcome to the new RAPS Digital Experience