FDA Neurological Devices Panel Addresses Clinical Trial Designs for Endovascular Aneurysm Devices

April 21, 2015—The US Food and Drug Administration issued a summary of the Neurological Devices Panel of the Medical Devices Advisory Committee meeting that discussed the current knowledge regarding the conduct of clinical studies and evaluation of clinical study data for endovascular devices for the treatment of aneurysms, including flow diverter technology. The FDA convened this committee on April 17 to seek expert opinion on scientific and clinical considerations relating to study design and existing clinical studies for endovascular devices.

Presenting for the FDA, Jeffery Toy, PhD; Samuel Raben, PhD; Mohamad Bydon, MD; and Laura Thompson, PhD, provided current knowledge on the scientific, clinical, and regulatory information on aneurysm treatment devices and considerations relating to their study design and existing clinical studies.

During the open public hearing period, presentations were given by representatives from the American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Section on Cerebrovascular Surgery, Covidien/Medtronic plc, and the Society of NeuroInterventional Surgery.

The deliberation of the panel discussion included imaging modalities, different aneurysm subgroups (eg, characteristics, location, size, morphology, etc), clinical trial designs, safety and effectiveness endpoints, and durability of treatment. 

The following statements are the panel’s consensus recommendations in response to eight questions that the meeting addressed:

First, when designing clinical trials of aneurysm treatment, the following subgroups are important to consider: anatomic location, morphology (wide neck vs nonwide neck, saccular vs fusiform vs dissecting), size, perforator vessel density, parent vessel characteristics, and patient characteristics.

Second, from a safety perspective, the anterior circulation aneurysms are dissimilar to posterior circulation aneurysms, and aneurysm sizes may be grouped as either small and medium or large and giant. Additionally, aneurysms located in perforator-rich regions of the vasculature are different than those located in perforator-poor locations, and patient characteristics play a further role in distinguishing aneurysms.

Third, in measuring effectiveness by occlusion rates, aneurysm sizes may also be grouped as either small and medium or large and giant. Additionally, wide-neck aneurysms should not be separate from nonwide-neck aneurysms; saccular aneurysms will have a different occlusion rate from fusiform/dissecting aneurysms; and parent characteristics could lead to dissimilarity in occlusion rates between treated aneurysms.

Fourth, performance goal trial designs should be well justified and should be reserved for situations where randomized control studies and/or cohort studies (comparator can be current or historical) are not feasible. If this trial design is warranted, the performance goals should be updated as new information becomes available.

There are concerns with performance goal trial designs, including those for effectiveness for patients treated in the intracranial anterior circulation independent of other aneurysm characteristics, a single performance goal for effectiveness for patients treated in the intracranial posterior circulation independent of other aneurysm characteristics, and a single performance goal for effectiveness that pertains to the complete neurovasculature independent of other aneurysm characteristics.

Fifth, randomized controlled trials should always be considered. However, when randomized control trials or cohort studies are not feasible, a performance goal study may also be considered. When designing a study, the following should be considered: well-defined questions, target populations, focused studies upon specific aneurysm types, and multiple treatment arms.

Sixth, the Raymond Scale and Target Aneurysm Retreatment Rate are considered appropriate as effectiveness endpoints; the panel was less in favor of the Meyers, Kamran, and Szikora Scales as effectiveness endpoints.

Seventh, in general, several of the safety and effectives endpoints identified by the FDA were relevant when designing a clinical trial. The secondary effectiveness endpoints identified by the FDA are retreatment rate, recanalization, change in modified Rankin Scale, change in Raymond Scale, < 50% stenosis, aneurysm occlusions of 90% or 95%, and improvement in symptoms. The secondary safety endpoints identified by the FDA are death, neurological death, stroke, neurological deficit, transient ischemic attack, aneurysm rupture, neuropsychological effects, complications of cerebral angiography, complications of antiplatelet therapy, and silent brain imaging changes.

Additionally, the following factors are to be considered during a clinical trial evaluating endoluminal aneurysm treatment safety and effectiveness: functional outcome score (modified Rankin Scale or Barthel Index), residual filling of aneurysms, vessel occlusion, periprocedural complications, and other safety-related events as appropriate.

The panel agreed that different safety endpoints have different severities but did not provide direct feedback on a composite endpoint for a performance goal-based study.

Finally, the panel was asked to address the length of premarket follow-up that is sufficient to capture the major adverse events and demonstrate that the majority of the aneurysm healing has reached a steady state, given that the endoluminal occlusion of aneurysm by flow diverters is delayed (which could lead to delayed aneurysm rupture). The panel recommended that 1-year premarket follow-up is sufficient to capture major adverse events. The panel advised that 5 years postmarket may be necessary to capture other adverse events during follow-up and that primary care physicians may also be following patients for their lifetime.