BOSTON--The FDA faces a number of challenges regarding trial design for devices that seek to treat patients with atrial fibrillation (AF), said Randall Brockman, MD, from the division of cardiovascular devices at the FDA, during a Jan. 13 presentation at the Boston AF Symposium (BAFS).
Almost two years ago, the FDA approved the first ablation device for paroxysmal AF, which Brockman said allowed for clinical trial designs that randomize a new catheter to an approved catheter.
“There is a lot of continued interest in research for AF, and with market approval of several AF ablation devices, we have the option of adopting randomized trials with control arms where the patients wouldn’t undergo the procedure,” said Brockman, who specifically questions how to deal with device modifications, as the devices evolve over time. “At the FDA, we struggle with how much new clinical information is required when devices are modified from an older version,” he added.
For example, changes in device design or device modification could include a change in ablation tip size or the manner in which irrigation is delivered. Also, labeling changes can cause challenges, such as indications for use, instructions for use and ‘compatible’ devices, according to Brockman. Specifically, new features for AF ablation devices could include assessments of tip-tissue contact, navigation method, novel mapping systems and real-time assessments of lesion formation.
He pointed out that device-device randomized, controlled trials with an ablation control arm come with a “unique set of challenges.” One design could be superiority versus non-inferiority. And, if the trial is a non-inferiority trial, what are the appropriate non-inferiority deltas for safety and effectiveness? “These deltas have important implications for the sample size,” Brockman said.
Also, other challenges of the device-device trial are non-inferiority creep. “If you compare the newest model of a device to a previous model in a device-device trial—each with its non-inferiority profile—each device may be found non-inferior to the prior iteration, but it may not be completely representative of actual performance,” said Brockman. “In order to minimize this creep with AF ablation device, we are employing the 'minimally acceptable performance value' in our study design.” However, he cautioned the audience that the agency did not classify this measure as a performance goal.
Additionally, the need for clinical data is not always straightforward, and if it decided that clinical data are needed, how should that data best be determined: A full outcomes trial? A confirmatory trial? A surrogate outcomes trial?
“Appropriate testing and evidence needs to be available for the FDA to make a decision ... [and the agency] faces a number of challenges regarding trial designs and the amount of clinical needed,” Brockman concluded.