Interventionalist vs Neurologist: RESPECT for Both Sides of the PFO Closure Debate

Extended findings from the RESPECT trial suggesting some longer term benefit to closing patent foramen ovale (PFO) are breathing new life into a controversy that has dogged the procedure for more than a decade.

On the one hand, interventional cardiologists participating in press conference at TCT 2015 unanimously said they would have a PFO closed if they suffered a stroke of unknown origin. Others, however, remain skeptical, saying that the late-term RESPECT data will not be enough to persuade the FDA to take a second look.

Interventionalist vs Neurologist: RESPECTing Both Sides of the PFO Closure Debate

In the original RESPECT results, the rate of the primary endpoint—recurrent nonfatal ischemic stroke, fatal ischemic stroke, or early death after randomization in the intention-to-treat population of 980 patients—was similar among patients diagnosed with cryptogenic stroke who were randomized to PFO closure with the Amplatzer device (St. Jude Medical; n = 499) compared with guideline-directed medical therapy (n = 481) over 270 days.

In the new data presented at TCT 2015, however, investigators zeroed in specifically on the outcome of cryptogenic stroke, showing a 54% relative risk reduction in its recurrence for patients treated with closure compared with medical therapy at mean follow-up periods of 5.5 and 4.9 years, respectively.

Disparate Reactions

Steven R. Messé, MD, a neurologist at the Hospital of the University of Pennsylvania (Philadelphia, PA), said he is unable to let go of the fact that RESPECT did not show a statistically significant difference for the primary endpoint, either in the trial’s original intention-to-treat analysis or in the extended follow-up.

While the original published data are suggestive of a benefit—a significant difference was seen among the prespecified per protocol and as-treated patients—the new data “doesn’t move me one way or the other,” he told TCTMD. The Kaplan-Meier curves “seem to come together completely by 5 or 6 years,” Messé said, “and if you look even at the cryptogenic stroke cohort, [the curves don’t] seem to be continuing to widen, which is what we’d expect if we’d really stopped the reason why they had their stroke in the first place, which was this paradoxical embolism.”

But RESPECT principal investigator John D. Carroll, MD, of the University of Colorado, Denver (Aurora, CO), insists that understanding of the link between PFO and cryptogenic stroke continued to evolve over the long course of the clinical trials. As such, it is only in hindsight that researchers can truly know why the original primary endpoint “needed to be refined,” he told TCTMD. “The good news is that with long-term follow up you get a lot of data, and the consequence is that people age and that people start having strokes from the normal mechanisms.”

Of note, the longer term data from RESPECT suggested that subjects under age 60 showed an even greater benefit of PFO closure.

However, Messé described himself as being “not that impressed with the age of the population” at follow-up. Specifically, he said, only 20% of the RESPECT study population is older than 60, “and for a stroke cohort, that’s remarkably young.”

Additionally, Messé pointed out that patients who received the Amplatzer device had 17 pulmonary embolism events at follow-up compared with 3 in the medical management arm. Because of this, he said, concluding that closure is clearly beneficial is “troubling.”

While most experts agreed that the pulmonary embolism findings warrant a closer look, the overwhelming response from physicians at TCT was that the negative findings from RESPECT deserve a rethink.

Simply dismissing RESPECT as a “negative trial,” is akin to “throwing the baby out with the bathwater,” Robert J. Sommer, MD, of Columbia University Medical Center (New York, NY), told TCTMD. “We’re never going to be able to do another trial like this.”

Lessons From RESPECT

According to Sommer and others, the biggest lesson from RESPECT and other randomized PFO closure trials relates to their problems with enrollment and retention—the same things that will make future trials impossible to perform.

“The problem has been, throughout all of these PFO trials, that there was off-label availability of the closure [device],” said Sommer. “If the hospital up the block is willing to do the procedure off-label, nobody wanted to be randomized. Plus, with the cardiologists’ insistence on getting these things closed and the neurologists’ insistence that they shouldn’t be closed, the [referring] doctors weren’t that anxious to randomize their patients either.”

Calling the length of time RESPECT took to enroll “unfortunate,” Messé said the “apparent biologic plausibility of PFO closure and perhaps, to some extent, economic incentive led to rampant off-label PFO closure in the past which slowed enrollment. Stroke is a very heterogeneous disease and many biologically plausible interventions have failed before.”

Despite this “paradox,” RESPECT “is the only evidence that we’re going to get,” Jonathan M. Tobis, MD, of the David Geffen School of Medicine at UCLA, (Los Angeles, CA), told TCTMD.

In fact, 2 other trials are still struggling to enroll their quota of patients. REDUCE is testing 2 different septal occluders (Gore) for PFO closure compared against medical therapy, but is “also underpowered in terms of the number of cases,” Tobis said. “It makes it extremely difficult, and I think that’s got to be taken into account. I can’t imagine that any other company is going to put the money into perform [another] trial like this.”

There is also the French CLOSE study, which has now finished enrolling patients.

“It would be nice to see some additional confirmatory evidence just as proof-of-concept that closing the PFO reduces stroke risk and is safe,” Messé said. But the notion that PFO plays a causal role in stroke has not been definitively established. “Having a stroke as a young person and finding a PFO and no other cause of that stroke puts you at incredibly low risk for recurrence. So I don’t think there’s any mad rush to go off and close lots of PFOs. I think that we need to prove that it really works and determine who is most likely to benefit.”

By contrast, Carroll asserted that the new data strongly support moving forward with PFO closure in highly selected patients, in whom all other potential sources of stroke have been conclusively ruled out. Past challenges with RESPECT “should not impair us from now bringing this treatment with now supportive data to formal approval and making it available for these young patients who are looking for scientifically valid, safe therapies to reduce the risk of having another stroke,” he said.

Enough for the FDA?

In the mid-2000s, the FDA withdrew Humanitarian Device Exemption marketing approval for the 2 PFO closure devices already on the market, essentially shutting the door on off-label use. Whether the agency ultimately decides to consider the RESPECT follow-up data and rethink its position remains to be seen.

Tobis pointed out that if PFO closure for cryptogenic stroke prevention is ultimately reconsidered, the field would change. Insurance companies would start reimbursing for the procedure and patients would no longer have to go through the arduous process of appeals and approvals, he said.

On the other hand, if the FDA denies the approval, “I see no future for [the procedure],” Sommer said. “At that point, PFO closure as we know it is over. Even patients who have high risks are not going to be able to get the therapy, because nobody is going to pay for it.”

In Sommer’s opinion, the data as they stand are “incredibly compelling.” Any reconsideration of PFO closure approval would be “political, … not medical,” he said, adding that while neurologists have “been a big stumbling block in the past,” the RESPECT data should “convince many of them that this is the right thing to do for a specific group of patients.”

A post-approval registry that would “continue to collect data just like they do with any other PMA device” could assess additional risks and long-term complications, he said.

“There is a market out there for closure, obviously,” Messé conceded, but he called for more confirmatory data on the best possible candidates and insisted a new study is not out of the question. Now that off-label closure “is happening much less,” an appropriately-powered study “would get done in a reasonable time frame,” he stressed.

Carroll rebutted that “in this circumstance, I think we have a reasonable amount of data…. We as clinicians always have to make decisions on an individualized basis and this provides the overall guidance.”

Asked to comment on how much the new insights from RESPECT could potentially influence regulators, Harlan M. Krumholz, MD, SM, of Yale-New Haven Hospital (New Haven, CT), called the extended analysis “interesting, but the level of evidence is far short of a positive clinical trial. The results raise future hypotheses to test, but are certainly not definitive evidence of benefit and leave substantial uncertainty about the efficacy of the procedure.”

Several regulatory bodies outside the United States have allowed the use of PFO closure devices when specific conditions are met. During a TCT 2015 press conference, Thierry Lefèvre, MD, of Hopital Privé Jacques Cartier (Massy, France), noted that, in France, patients found to have a PFO are required to have not 1 but 2 strokes of unknown origin before a device can be considered.

That kind of approach does not sit well with many physicians. According to Tobis, if he had a PFO and had already experienced a single event, he “would not want to sit around waiting for Damocles’ sword to fall and have a second stroke. I think that requirement is chilling.”

Sommer agreed. “We’re not talking about a disease that’s got a 5-year life course,” he commented. “We’re talking about 20 year-olds who have strokes who are going to have these PFOs for the rest of their lives.”

Messé, for his part, seemed leery of making decisions based on the data available to date and resisted the idea that young stroke patients are being put at risk while this therapy is unavailable. There is no solid proof that patients already treated with this device are any better off, he emphasized. “Perhaps we will look back and say that many of those were inappropriately done.”

Carroll J. RESPECT extended follow-up results. Presented at: Transcatheter Cardiovascular Therapeutics; October 15, 2015; San Francisco, CA.

Related Stories:


  • RESPECT was funded by St. Jude Medical.
  • Carroll reports receiving consultant fees from St. Jude Medical.
  • Messé reports serving as a local PI for REDUCE, a subinvestigator for CLOSURE I, and on the PFO guideline development subcommittee of the American Academy of Neurology.
  • Sommer reports holding stock options in Coherex and serving as a PI on several PFO closure trials.
  • Tobis reports serving on the steering committee for the PREMIUM trial and as a local investigator for RESPECT.
  • Krumholz reports no relevant conflicts of interest.