Complete PFO Closure May Be Key for Migraine Reduction

In patients with migraine who undergo transcatheter closure of patent foramen ovale (PFO), the presence of aura and of residual right-to-left shunt appear to influence the amount of migraine benefits, new data show.

After PFO closure, migraine symptoms were completely abolished in nearly half of patients (48%), with this outcome being much more likely in patients with aura (OR 4.34; 95% CI 1.53-12.30), according to researchers led by Eyal Ben-Assa, MD (Tel Aviv Sourasky Medical Center, Israel, and Massachusetts General Hospital, Boston).

Moreover, migraine burden was cut by more than half in 87.3% of patients after PFO closure, with an even greater likelihood of seeing that level of improvement when there was no residual interatrial shunt (OR 4.61; 95% CI 1.32-16.10), Ben-Assa et al report in a study published online ahead of the February 10, 2020, issue of JACC: Cardiovascular Interventions.

The findings, from a retrospective, single-center study, need to be considered in the context of the migraine-PFO literature—observational studies have supported an improvement in migraine after PFO closure and three randomized trials failed to meet their primary endpoints but provided evidence suggestive of a benefit of closure.

Ben-Assa told TCTMD that none of the randomized trials took the presence of residual shunt after PFO closure into account. It could be, he explained, that “effective closure” that allows for the presence of mild shunt after the procedure may be good enough to prevent stroke but not migraine. Proposed—but not proven—mechanisms to explain the link between PFO and migraine are that the PFO allows microemboli or vasoactive substances like serotonin to travel to the brain and cause migraine symptoms. Ensuring complete closure of the PFO would eliminate those processes.

“Our data strengthens those pathophysiological questions, because those microemboli and vasoactive factors can percolate through very small shunts also,” Ben-Assa said. “That is why we think that it is good to have effective closure for stroke but if you want to improve migraine you need to strive for complete closure.”

Complex Connection

Ben-Assa described the association between PFO and migraine as complex. Migraine occurs in about 10% of the general population and is roughly two to three times more common in patients with PFO. In the clinic, patients with PFO and migraine often ask what will happen with the migraine symptoms after PFO closure, and there isn’t a definitive answer to the question, he added.

Although the randomized trials failed to show a significant benefit on their primary endpoints, there were positive signs. “We’re looking at this complex connection between [PFO and migraine] and saying that there might be more than meets the eye, and you cannot say no it doesn’t work according to three studies that do show benefit in certain populations,” Ben-Assa said.

He and his colleagues tried to tease some of that nuance out in the current study. Of 474 patients who underwent transcatheter PFO closure for cryptogenic stroke/TIA (91%) or hypoxia (9%) at Massachusetts General Hospital, 110 patients had migraine. Most of the migraineurs (mean age 43 years; 67% women) had aura (77%).

We should strive for complete closure and if someone has aura we can tell them that the literature suggests that they will have higher likelihood to improve. Eyal Ben-Assa

During a median follow-up of 3.2 years, migraine symptoms—including frequency, duration, and burden—improved both in patients with and without aura. Migraine burden—the number of days with headache per month multiplied by the average duration of each episode—was reduced by more than 50% in nearly nine out of every 10 patients, with no difference based on the presence of aura. Patients with aura were more likely, however, to have symptoms completely abolished after PFO closure (55% vs 24%; P = 0.006). Only four patients (3.6%) had worsening migraine symptoms.

At 6 months after closure, residual right-to-left shunt on transthoracic echocardiography was present in 26% of patients, with no difference based on the presence of aura. Patients without residual shunt were more likely to have an improvement in migraine burden.

“Our study adds another layer of information and a new perspective on the relationship of PFO and migraine,” Ben-Assa et al write in their paper.

More Trials Needed

Commenting for TCTMD, Gretchen Tietjen, MD (University of Toledo, OH), said the study has all of the limitations that go along with a retrospective, single-center analysis but called it “a real contribution to our knowledge base that I think just really gives more ammunition for continuing to do research in patients with PFO and particularly in patients who have migraine and migraine with aura.”

She said she was not very surprised to see that patients with aura were more likely to have a complete elimination of migraine symptoms after closure because she has adhered to the belief that a substantial proportion of migraine with aura is triggered by microemboli or vasoactive substances traveling to the brain.

More surprising, Tietjen said, was the large reduction in migraine burden in patients without aura and the effect of residual shunt. Patients with mild shunt still had a reduction in migraine symptoms, but not as much as when it was absent. “What that suggests to me is that it may not be just microemboli, which has always been my own bias, inducing an aura,” she noted. “It points to the presence of other vasoactive substances or hypoxia in the shunted blood playing a role.”

The study investigators called for future trials of PFO closure in migraineurs that incorporate these findings and aim to identify the specific subsets of patients who will benefit.

“Investigators may consider including not only the refractory migraineurs populations, consider less stringent primary endpoints, and consider [following] patients for longer periods,” they write. “Regarding efficacy of device closure, we recommend evaluating residual right-to-left shunt at 6 to 12 months postclosure and considering including residual shunt analysis in the outcome metrics, as well as evaluating the role of ‘complete’ versus ‘effective’ closure.”

What that suggests to me is that it may not be just microemboli . . . inducing an aura. It points to the presence of other vasoactive substances or hypoxia in the shunted blood playing a role. Gretchen Tietjen

In an accompanying editorial, Stéphane Noble, MD (University Hospital of Geneva, Switzerland), highlights the importance of careful patient selection for future studies.

“There are patients with PFO without migraine, and there are migraineurs without PFO. There are still many questions that should be addressed in the migraine population in collaboration with neurologists. The abolishment of migraine symptoms seems repeatedly more frequent in patients with aura as shown again in the report by Ben-Assa et al,” he says.

“Further studies in PFO closure could consider the presence of aura or P2Y12 inhibitor responsiveness among migraineurs as selection criteria,” he continues. “Finally, the safety and efficacy of the procedure with high complete closure rates are essential in order to treat this population; therefore, device selection, as well as procedural technique, is important. Transesophageal echocardiography or at least intracardiac echography should be used to guide the procedure.”

Ben-Assa indicated, however, that the current study could have an impact right now.

“It strengthened at least my clinical perspective that there is a specific population that will benefit from PFO closure in terms of their migraines,” he said. “I think that we should strive for complete closure and if someone has aura we can tell them that the literature suggests that they will have higher likelihood to improve.”