EMA addresses use of registry-based studies

Regulatory NewsRegulatory News | 24 September 2020 |  By 

The European Medicines Agency (EMA) has issued a new guideline on registry-based studies, giving detailed advice for use of patient registries in both the pre- and post-authorization phases for medicinal products. 
After committee approval and a consultation period with the EU Regulatory Network, the guideline has been released for public consultation, with comments due before the end of 2020.
The guideline’s focus is on the use of patient registries for studies by marketing authorisation applicants and holders (MAAs/MAHs). The guideline’s scope includes disease and condition registries, such as those for patients with a particular disease or disease characteristic, a condition such as pregnancy, or molecular or genomic features.
Registries are particularly important, notes the guideline, for patients with rare diseases who may receive advanced therapy medicinal products (ATMPs) such as gene therapy.
In the post-authorization phase, patient registry-based studies can be useful for both safety and efficacy studies. Registry-based post-authorization studies can help in such situations as assessing effectiveness of adapted drug dosing schemes. More broadly, registry-based studies can aid in understanding of the “effectiveness and safety of medicinal products in a broader clinical disease-related context and a more heterogenous patient population,” according to the guidance.
In the case of ATMPs, registry-based post-authorization studies can be particularly important since approval is often based on small patient numbers. These studies “are frequently and preferentially performed on the basis of existing disease registries," says EMA, noting that post-authorization activity planning should begin in the pre-authorization phase for ATMPs.
Considerations in planning a registry-based study include identifying the registry and setting up a collaboration agreement with the operators of the registry; additionally, sponsors should consider data extraction and data quality control issues at the outset.
In designing a registry-based study, considerations should include how relevant the registry is to the specific questions being asked, as well as the quality and completeness of data contained in the registry. The guidance recommends consultation with national competent authorities before beginning a registry-based study.
For pre-authorization studies, observational evidence from patient registries can be helpful, says EMA, since the registries can supply information about population characteristics and can help identify valid surrogate endpoints. 
"Studies based on patient registries may also contextualize the results of uncontrolled trials,” as well as providing comparator groups when randomized controlled trials (RCTs) are infeasible or unethical. EMA’s scientific advice should be sought when “deviations from a traditional RCT design are considered.” 
The EMA PRIority MEdicines (PRIME) procedure can be used during the pre-authorization phase, if its use is applicable. “It is the responsibility of the MAA/MAH to include in the discussion the holders of the registry(-ies) intended to be used,” says the agency.
The guideline outlines components of a feasibility analysis that MAAs/MAHs should consider performing before completing the study design. Among the components of the feasibility analysis are an analysis of the availability, quality and completeness of the data elements needed for the study; how missing data will be handled; how adverse events will be addressed; specifics of how the registry characteristics will impact patient recruitment; potential confounders; data privacy issues; and other considerations.
The study protocol for a registry-based study should follow regulatory requirements including guidelines E6, E8 and E9 from the International Council on Harmonization. The method of data collection should be specified in the study protocol, and be clear about whether use will be primary or secondary.
The study population selection should ensure all individual centers can enroll the population of interest, and data for non-enrolling or withdrawing patients should still consent to providing “a small set of baseline data” to allow analysis for selection bias and generalizability. 
Considerations for data collection including planning for such entails as sensitivity analyses as early as possible, and for using risk-based methodologies for data quality management. In particular, the study plan should anticipate how to address missing or incomplete data within the registry. 
Data analysis should ensure the estimands of interest are aligned with adequate estimation and testing methods, with sensitivity analyses planned to test robustness.
The guideline also gives guidance in how to address the possibility for non-randomized treatment allocation, including how to measure incidence of outcomes of interest, and addressing the fact that “characteristics of patient groups given different treatments are likely to differ.” Also, the potential for immortal time bias and other time-related biases should be accounted for: for example, consideration should be given to enrolling patient with only incident, and not prevalent, treatment. 
Data reporting plans should include registering results and the final study report with applicable national and EU authorities, and legal and regulatory requirements in the pertinent jurisdictions should be met. The guideline includes a table outlining the applicable legal and regulatory requirements.
The guideline also includes an annex detailing general considerations on constructing and using patient registries. 


© 2022 Regulatory Affairs Professionals Society.

Discover more of what matters to you

No taxonomy