Posted 02 March 2022 | By Mary Ellen Schneider
Pharmaceutical manufacturers and patient advocates want to see more details from the US Food and Drug Administration as it finalizes guidance on the use real-world data (RWD) from registries, as well as alignment with three other RWD guidance documents.
The comments were in response to the FDA’s draft guidance “Real-World Data: Assessing Registries to Support Regulatory Decision-Making for Drug and Biological Products,” which was issued in November 2021 (RELATED: FDA offers draft guidance for registries as RWD, Regulatory Focus 29 November 2021). The comment period closed on 28 February 2022.
Harmonizing RWD guidance
The draft guidance spells out considerations for sponsors when proposing to design a registry or use an existing registry to support regulatory decisions on drugs and biologics, but does not address study design or type of statistical analysis that should be employed when using registry data. Several commenters said the registry guidance should be better aligned with other recent draft guidance documents on RWD and real-world evidence (RWE) to support regulatory decision making in drug and biological products, data standards for drugs and biological product submissions contain RWD, and RWD in assessing electronic health records (EHRs) and medical claims to support regulatory decision making.
The Pharmaceutical Research and Manufacturers of America (PhRMA) commented on the confusion created by recommendations in the draft guidance about linking registries to external data sources are different from what is outlined in the guidance on EHRs.
Officials at Takeda also noted the importance of increasing harmonization among the various guidance documents on RWD that the FDA has issued recently. “FDA’s recent draft guidance ‘Real-World Data: Assessing Electronic Health Records and Medical Claims Data to Support Regulatory Decision-Making for Drug and Biological Products’ includes specific recommendations on the validation of outcomes that are pertinent (and not discussed) in this guidance. Additionally, many of the considerations in this, and the other RWD guidances, are applicable to both registries and other data sources,” Takeda commented.
The Biotechnology Innovation Organization (BIO) suggested that FDA include more considerations for hypothesis testing and validation in the guidance on registries, since these topics are discussed in detail in the guidance on data from EHRs and claims. The agency could also offer an explanation about why some considerations may be more important for one RWE data source than another. “Topics that are likely relevant across RWD sources such as claims, EHRs and registries (e.g., validation, linkage, hypothesis testing, provenance, etc) are not treated equally across the guidance documents, which may be a source of confusion for sponsors. Alignment across guidance documents (to the extent possible) would be welcomed by sponsors,” BIO commented.
Data control, patient-level data
Regeneron Pharmaceuticals urged FDA to acknowledge that data collection, curation, management, and storage associated with registries may be considered proprietary and outside the control of the drug/biologic sponsor. The guidance should provide recommendations of best practices when information on data accrual, curation, and transformation is not available, the company commented. Regeneron also suggested that FDA encourage data owners to share their master file with the FDA and receive the agency’s approval for use of their data in certain applications. This approach would likely “reduce development delays,” Regeneron commented.
Takeda echoed the concern that the guidance implies that sponsors control the data in registries. The company recommended that the agency consider advocating for registries that allow for multi-stakeholder participation and data sharing by directly engaging with registry owners or including them in the audience for the guidance. “This would both be more efficient and would avoid creating a fractured registry landscape where individual sponsors communicate different interpretations of this draft guidance to different registry owners,” Takeda commented.
Takeda also pointed out that patient-level data sharing is not a common practice with existing registries for many reasons, including privacy concerns. The company encouraged FDA to exercise flexibility in its recommendation that sponsors provide patient-level data to support a product’s effectiveness or safety. Meanwhile, Amgen called on the FDA to convene a public workshop to solicit views on how to share patient-level data with sponsors and the agency.
Data heterogeneity, global registries
The RWE Alliance, a coalition of RWD and analytics organizations, encouraged FDA to offer recommendations for mitigating the limitations of data heterogeneity, such as through data curation, ontological harmonization, and subgroup analyses. The group also recommended clarifying that there are additional considerations related to data heterogeneity that come up when using multiple registries in a single study, including pooling data or reconciling registry-specific results.
The RWE Alliance urged FDA to include information on how data heterogeneity offers potential advantages, not just limitations. “Heterogeneity in population characteristics can provide additional insight into diverse and underrepresented groups, which in turn can help advance the important public health objective of addressing disparities in healthcare access, treatment, and outcomes,” the group commented.
The RWE Alliance also suggested additions to the considerations surrounding relevancy of registry data. Assessment of relevance for a particular research question should include not just inclusion/exclusion criteria but other factors that impact patient participation, such as disease severity or location of care. Other factors could include the extent and pattern of patient loss to follow up, which can impact the potential for missing outcome information. The group also urged FDA to add biomarkers to the data to include in a registry. “[Biomarkers] can be accurate, valid, reproducible, and objective indicators of a patient’s medical state. Moreover, the inclusion of biomarker data in registries is one factor distinguishing registries from other data sources that do not include biomarkers, such as medical claims,” the RWE Alliance commented.
The LUNGevity Foundation, a lung cancer patient advocacy organization, asked for additional details on whether a registry is fit for use. Specifically, the group is seeking more clarity on the data elements required to support different uses of registry data, such as external control arms in clinical trials. The foundation also recommend that FDA provide specific examples across disease types of how stakeholders can design registries for specific purposes. They are seeking information on necessary data elements, data sources, and data linkage plans of registries that could be used to support an externally-controlled interventional trial.
Some commenters called on FDA to clarify that it will allow the use of registries from outside the US. The National Hemophilia Foundation (NHF) said global registries, such as the World Federation of Hemophilia Gene Therapy Registry, are critical tools for achieving sufficient patient enrollment and statistical power in the case of ultra-rare conditions like hemophilia.