FDA Releases Draft Guidance on Adaptive Clinical Trial Design

The U.S. Food and Drug Administration (FDA) recently released a draft guidance on adaptive design for medical device studies.

“An adaptive design for a medical device clinical study is defined as a clinical trial design that allows for prospectively planned modifications based on accumulating study data without undermining the trial’s integrity and validity. Adaptive designs, when properly implemented, can reduce resource requirements and/or increase the chance of study success,” according to the FDA.

This latest document from the agency discusses advantages and limitations, special considerations for design and even outlines what should be included in a submission.

To provide some clarity about the guidance document and adaptive trials, in general, MPO spoke with Vicki Anastasi, vice president global head of medical device and diagnostics for Icon plc, a global provider of outsourced development services to the pharmaceutical, biotechnology and medical device industries.

MPO: How was FDA handling clinical data gathered via adaptive designs prior to this guidance? Did the agency have a good grasp of the practice?

Anastasi: Adaptive designs are not new to the FDA. The CDRH (the FDA’s Center for Devices and Radiological Health) has been evaluating adaptive designs and submissions utilizing adaptive design since the early 2000s. From 2007 to 2012 the CDRH received 120 submissions that utilized adaptive designs.

This draft guidance is really intended to drive home the point that the FDA encourages adaptive designs, when appropriate, for medical device studies. Device manufacturers should take note if they have not already done so.

MPO: Why do you think adaptive trials are underused?

Anastasi: Some within the industry believe that adaptive designs increase the complexity of the trial and the burden placed on sponsors. Though adaptive designs require upfront planning and simulation, special infrastructure, and subject matter experts, I would argue that they do not actually make a trial more complex. In fact, these designs drive manufacturers to collect clinical data to support regulatory submissions and reimbursement in the most efficient and ethical way possible.

For those that realize the advantages of using an adaptive design, the biggest barrier is infrastructure. Many know that adaptive designs involve interim analysis and modifications to the study. But what is actually required for proper design and execution of adaptive trials remains largely unknown outside of the statistics community and a handful of CROs (contract research organizations) and manufacturers.

Importantly, one cannot simply borrow protocols from adaptive designs utilized in the pharmaceutical industry. Medical device studies that utilize adaptive designs will need experienced staff able to simulate, design, and execute these studies with medical device-specific SOPs (standard operating procedures). Operationally speaking, adaptive designs require validated, regulatory-compliant technological infrastructure that has data firewalls and other failsafes in place to reduce operational bias and maintain the validity and integrity of the trial. It must also capture and clean data in real-time to make it “interim analysis ready.”

MPO: How far will a guidance such as this, once finalized, go toward increasing the use of adaptive trial designs? 

Anastasi: The FDA now has a resource that demonstrates its proposed position on adaptive design. Even though the FDA has been open in its encouragement of adaptive design, when appropriate, the draft guidance is a more formal statement of this viewpoint. As a result, I expect to see a steady increase in the use of adaptive designs in medical device.

However, the ball is now in manufacturers’ courts to educate themselves, hire statisticians and experts in adaptive design, and take the time to determine whether an adaptive design is appropriate for trials in their portfolios. To achieve maximum benefit, all of this effort needs to take place as early as possible in the design stage. The FDA strongly recommends that device manufacturers schedule a pre-submission meeting to discuss the adaptive design prior to initiation of the study.

Adaptive design improves decision-making and increases the quality of clinical evidence collected to support regulatory, reimbursement, and market access. Along with risk based monitoring and a robust assessment of health economic drivers, adaptive designs should be part of every manufacturer’s strategy to stay competitive.

MPO: How sufficient is the current draft as-is? What are the areas you would comment on or change?

Anastasi: This is a well-written draft guidance that came to be as a result of a collaborative effort with industry to find ways to efficiently collect regulatory-acceptable clinical data. The discussion of when and when not to choose and adaptive design, types, challenges, and regulatory considerations make this document an excellent tool to level the playing field, so to speak, by giving manufacturers clarity on the FDA’s expectations.

MPO: What types of studies are adaptive trial designs best suited for? When are they not a good fit?

Anastasi: Adaptive designs work best on studies whose timelines can accommodate the pauses for interim analysis. In device, this will typically be orthopedic/spine and cardiovascular studies, as well studies evaluating IVDs (in-vitro diagnostic devices). Because of the time requirements, quick enrolling studies will not be a good fit.

Regardless of indication, adaptive design offers strong protection for the bottom line and for patients. Though there are many types of adaptive designs, group sequential designs are the most well known because they can yield early stops for futility or success. These enable manufacturers reallocate resources to more promising devices or submit for regulatory approval earlier than expected, respectively. Other applicable adaptive designs allow manufacturers to adjust sample size to save a trial or reduce the incidence of overpowered studies, as well compress trial duration by combining pilot and pivotal studies.