European Regulatory Roundup: EMA Drafts Guidelines on Controlling False Positives in Clinical Trials (6 April 2017)

Regulatory NewsRegulatory News | 06 April 2017 |  By 

Welcome to our European Regulatory Roundup, our weekly overview of the top EU regulatory news.

EMA Drafts Guidelines on Controlling Against False Positives in Clinical Trials

The European Medicines Agency (EMA) has released draft guidelines about multiplicity in clinical trials. The text deals with how to mitigate the risk of false positives arising when clinical trials look at multiple treatment groups and endpoints.

EMA’s guideline is applicable to most clinical trials. Once a study design expands beyond having two treatment groups, one predefined null hypothesis and a single primary variable — or adds an interim analysis — its sponsor needs to control for false positives. EMA cites the example of a trial that analyzes five subgroups independently. If each subgroup has a significance level of 2.5%, the likelihood of the study finding a statistically significant false positive hits 12%. As biotechs are under pressure to succeed, there is a risk they will seize on anomalies as evidence a drug works.

To ensure trial designs diminish the risk of this happening, EMA has released a 15-page guideline. The draft runs through the situations in which it is necessary to adjust for multiplicity, when it is possible to draw reliable conclusions from subgroup analyses, when success against secondary endpoints can form the basis for additional claims and how to handle composite endpoints.

Researchers well versed in designing clinical trials will be familiar with many of the agency’s recommendations. Broad messages conveyed in the guidelines include the need for sponsors to pre-specify any subgroups they plan to use to assert the safety and efficacy of a product, and the need for a study to hit its primary endpoint if success in secondary objectives is to support a regulatory claim. 

The agency covered many of these topics in earlier documents. EMA first established a position on multiplicity in clinical trials through a “points to consider” text in 2002, before going on to release a concept paper a decade later. The draft guideline released for consultation this week builds on both those texts, both by expanding on the topics they covered and incorporating advances in understanding since the first document was released 15 years ago.

New additions include a section on multiplicity issues in estimation. This section deals with what to do when the complexity of procedures makes informative clinical interpretation difficult. EMA addresses selection bias and confidence intervals in the section.

Draft Guideline

Better Multiple Myeloma Drugs Prompt EMA to Rethink Clinical Trial Endpoints

EMA is proposing to revise its guidelines on clinical trials of experimental multiple myeloma drugs in response to advances in the standard of care. With new treatments now triggering high rates of complete responses, EMA wants to incorporate minimal residual disease endpoints into its advice.

Clinical trials of multiple myeloma drugs have traditionally looked to endpoints such as the rate of complete responses and length of progression-free survival to assess efficacy. Yet, while today’s drug regimens are triggering complete responses that meet increasingly stringent criteria in more and more patients, people continue to relapse. In response, the International Myeloma Working Group (IMWG) began assessing how to determine the residual disease left behind after treatment.

The working group published its consensus criteria for minimal residual disease assessment last year, prompting EMA to revisit condition-specific guidance on the evaluation of anticancer drugs in humans. Taking its lead from the working group, EMA is exploring how the development of more sensitive assays capable of detecting minimal residual myeloma cells can facilitate the use of new endpoints that give a better picture of the depth of response triggered by a treatment.

EMA is yet to finalize its thinking on the topic. The agency views the prognostic value of achieving undetectable disease status as “undoubted.” However, EMA is unsure how differences in minimal residual disease correlate to progression-free survival. The working group grappled with this topic, too, noting, “The association between depth of response and long-term outcomes is a hotly debated topic in multiple myeloma.”

A lot rests on EMA’s ability to identify a good correlation between the extent to which a treatment eradicates residual disease and long-term outcomes. If EMA sees such a correlation, it is open to allowing the use of minimal residual disease as a primary intermediary efficacy endpoint. Sponsors using this endpoint could win early approvals before delivering progression-free survival data to confirm the effect of their drugs.

EMA is accepting comments on the concept paper until the end of June. The agency plans to release draft guidance for a six-month consultation one year after it adopts the concept paper.

Concept Paper, IMWG Paper

EMA Starts 12-Month Consultation on Comparing Drug Quality Attributes

EMA has released a draft reflection paper on comparative assessments of drug quality attributes. The document will go through an unusually long consultation before being discussed at a public workshop, an approach EMA sees as necessary given its lack of concrete positions on the topic.

In the 24-page reflection paper, EMA discusses regulatory considerations related to the statistics of comparing the quality attributes of two products. Such comparisons are particularly important to assessments of biosimilars and generics. Regulators need to know whether the quality attributes of these drugs match those of their reference products. The document also applies to companies that change their manufacturing process and need to check whether their product was affected.

Yet, while EMA knows the settings in which the document applies, it is less clear on exactly how it and the companies it regulates should handle the statistics of making the comparisons. Sections of the draft cover frequently used statistical methods, but EMA is open to using alternative approaches such as the Bayesian methodology.

Given the fluidity of EMA’s position, it has opted against setting out explicit guidance about which statistical approach is best. Instead, the agency has published a reflection paper designed to set out a framework and common language that will support discussions over the coming year. EMA wants companies to approach the Scientific Advice Working Party before the consultation closes on 31 March 2018, and then attend a planned public workshop to discuss how to proceed.  

Reflection Paper

Swissmedic Moves to Formalize, Strengthen Interactions with Medical Associations

The Swiss Agency for Therapeutic Products (Swissmedic) is working to formalize and strengthen its relationships with associations of medical professionals. Swissmedic has traditionally worked with such groups on an ad hoc basis, but is now planning periodic meetings with a set list of groups.

Officials at Swissmedic are currently drawing up a list of organizations with which they will interact going forward. Associations must meet certain criteria to get on the list. Swissmedic wants to work with associations based in Switzerland that are transparent about their sources of funding, clear about their objectives and consult with their members. 

Once Swissmedic has established the list, it plans to hold one or two meetings a year to provide updates on the progress of the collaboration and deliver other general information. In between these meetings, Swissmedic will draw on the knowledge of the associations to support its work on directives and other regulatory actions. Swissmedic hopes these interactions will ensure it includes the views of people involved in clinical practice in its positions on topics such as drug safety.

To prepare associations for the changes, Swissmedic is hosting two information meetings in the coming weeks. 

Swissmedic Statement(French)


© 2023 Regulatory Affairs Professionals Society.

Discover more of what matters to you

No taxonomy