The European Medicines Agency (EMA) has announced the release of a new concept paper regarding the extrapolation of an experimental drug product's safety and efficacy data to other subgroups, explaining that a new guideline is needed to instruct industry on how to validate their conclusions.

Background

Regulators processes are an inherent balancing act between two competing forces: The need to get new therapies to patients and the need to make sure said therapies are safe and effective for those same patients. Without the latter-supported by statistically significant evidence-regulators would be unable to determine if the status quo for those patients would not in fact be better than subjecting them to a new and unproven therapy.

That leaves companies with a seemingly straightforward task that is often anything but: Gather clinical evidence to show the safety and effectiveness of a product.

The speed at which this process of gathering clinical evidence occurs is subject to a myriad of complex variables, including the severity of the disease, the intended application type, the size of the disease population, the nature of the disease itself and the regulatory requirements for testing.

And on the edges of those variables can be a complex question: Can data obtained for one purpose be extrapolated to other subpopulations?

How to Extrapolate Data?

It's a question with considerable regulatory and ethical implications. If, for example, a company is able to extrapolate that a 10mg dose of a drug given to a child will have the same safety and efficacy as if it were given to an adult, the company could not only skip the time and expense of conducting those studies, but also the ethical issues associated with testing a drug in a pediatric population (consent, safety protections, etc).

EMA explained that other subpopulations are also eligible for extrapolation, such as various ages, stages of growth and maturation, sexes, pregnancy status, co-morbidities, impaired organ functions, ethnicities, disease subtypes, medicines and the health status of patients.

What's missing, regulators said, is a cohesive framework by which industry can apply these extrapolation approaches in ways that are both reliable and scientifically valid for the purposes of supporting a drug application to EMA.

That framework will eventually take the form of a guideline, EMA said, and will "set out a structured approach to be followed for each extrapolation exercise to improve interactions with stakeholders and to standardise the decision making across EMA committees."

"Since the application of extrapolation varies by population, therapeutic area, and medicinal product, an inventory of approaches and case examples shall be collected," EMA added.

A Framework Emerges

Regulators explained that developing a framework will be no small undertaking given the number of problems associated with extrapolation. At present, EMA identified 10 "gaps in knowledge" that must first be resolved before a guideline can be developed:

• defining the impact of extrapolation in drug development and regulatory review
• considering the clinical context: how to account for feasibility and ethical restrictions for studies in specific populations
• defining and quantifying similarity of disease (progression), of PK/PD, of clinical response to treatment and safety aspects
• deciding on the quality and quantity of existing data and types of study designs to support the extrapolation concept
• weighing the strength of prior information
• integrating expert judgment in the extrapolation concept
• quantifying the uncertainty of extrapolation assumptions
• validating assumptions in the extrapolation concept
• dealing with uncertainty and risk
• analyzing and report post-authorization data to support extrapolation

An eventual framework is likely to involve at least four elements, including an "extrapolation concept" built upon and modeled off of an analysis and simulation of existing data, a plan to reduce the number of studies needed in certain populations, a way to validate the extrapolation concept using emerging data and a way to interpret the "limited data in the target population in the context" of extrapolated information.

Uncertainty Inherent

As the eventual framework will be built at a high level of understanding, EMA stipulated that it will likely rely on the use of checklists to assess the similarity of different population subtypes. The use of quantitative data will also be important, including models of disease progression, pharmacokinetics, pharmacodynamics and clinical response.

Even with a well-validated plan, however, sponsors will still be dealing with an inherent amount of uncertainty and risk. EMA said that the future framework will work to take this into account, and ask sponsors to "consolidate the reliability of conclusions based on extrapolation" to help determine whether a product is safe for use in a subpopulation without clinical testing.

EMA: Concept paper on extrapolation of efficacy and safety in medicine development