More Headaches: PREMIUM’s Publication Renews Debate Over PFO Closure for Migraine

The PREMIUM trial, which failed to show a benefit of patent foramen ovale (PFO) closure in patients with frequent migraines, is now in print. Responding to the publication this week, proponents of the procedure say they are holding out hope that future research may yet be able to identify a subset of these patients who could benefit from closure. Others, however, remain skeptical that such patients can be found and that PFO closure will ever play a meaningful role in this complex disease.

Several randomized trials have now attempted to establish that PFO closure can reduce migraines; none of them have succeeded. The problem with all of them, according to Robert Sommer, MD (NewYork-Presbyterian/Columbia University Medical Center, New York, NY), the national PI of the halted ESCAPE Migraine trial, is that while both migraines and PFOs are “super common,” not everyone with migraines has PFO and vice versa. So far, he told TCTMD, researchers have “been unable to select the patients who have the causal PFOs from the patients who have the incidental PFOs, so it’s almost impossible to have any expectation that any of these trials could have been positive.”

Nonetheless, PREMIUM’s lead author, Jonathan Tobis, MD (David Geffen School of Medicine, University of California, Los Angeles), still believes that despite the trial’s failings, the procedure’s “benefits outweigh the risks for people who have severe debilitating migraine.” These are people whose migraines disrupt their work or family life, or who in some cases become suicidal, he explained. “If you have one migraine a month, then I don't think [PFO closure] is worth it.”

Endpoints and Explanations

For the study, originally presented at TCT 2015 and published this week in the Journal of the American College of Cardiology, Tobis and colleagues screened and obtained consent from 1,653 patients with 6-14 migraine days per month who had failed at least three preventive migraine medications and had significant right-to-left shunt defined by transcranial Doppler. Ultimately, of the 230 enrolled over 7 years—a full 69% were excluded due to having too small a shunt—123 patients were randomized to PFO closure with the Amplatzer PFO occluder (Abbott) and 107 had a sham procedure. Four patients in the study arm were unable to have the procedure due to anatomical reasons, and both the patient and the neurologist were blinded as to the randomization decision for 1 year.

The trial did not meet its primary efficacy endpoint of responder rate of a 50% reduction in migraine attacks regardless of aura, with 38.5% and 32% of patients in the study and control arms, respectively, having a 50% reduction in attacks (P = 0.32). However, there was a significant decrease in the mean number of migraine days per month for patients who underwent PFO closure versus controls (4.4 vs 5.0 days; P = 0.025). There was also a difference in the numbers of device and control patients who saw an elimination of migraines at 1 year (8.5% vs 1%; P = 0.01). Notably, of the 10 study patients who reported complete remission after PFO closure, six had frequent aura and four had infrequent or no aura.

Among the 205 patients who ultimately had PFO closure—controls were given the option to undergo the procedure at the end of the study period—there were six major procedure-related events (2.9%) which were “self-limited and most represented common adverse events associated with any right heart catheterization,” according to the authors. There was one instance of device-related A-fib during the procedure in a patient who received closure.

In 2015, St. Jude Medical, the then manufacturer of the Amplatzer PFO occluder, discontinued the PREMIUM trial and its long-term follow-up as well as its plans to seek US Food and Drug Administration (FDA) approval of this device for the treatment of migraines. However, to TCTMD, Abbott said the company “remains focused on PFO closure to prevent strokes.”

The PREMIUM researchers initially wanted their primary endpoint to be related to migraine days instead of attacks, but had to change their plans based on what the FDA wanted, Tobis said. “The problem with that is there are fewer attacks than there are migraine days, and if your range is small, it's harder to prove statistical significance.”

While the secondary endpoints can only be hypothesis-generating for now, he observed, it’s important to notice how well the procedure seemed to work for migraine patients with aura, a finding commonly seen in past studies. For patients with frequent aura, complete migraine remission was achieved in 15.4% of patients in the study group compared with 2.5% of controls (P = 0.04).

Thus, Tobis said he “would support another trial that looked at that subset of patients to prove that PFO closure is effective to prevent migraines.”

Another important takeaway from the trial was the “very high” placebo effect observed—simply following patients closely and monitoring them aggressively can have an unquantifiable benefit, he said. This “is what can explain a lot of the observations that occurred during just the descriptive studies, and that's why randomized clinical trials are so important and, [for] in devices in particular, having a sham control is also critical, especially if the endpoint is a subjective endpoint like angina or pain or headaches,” Tobis added.

Would Patients Even Want This?

Others are not so sanguine as Tobis. Looking at all of the evidence to date, Nauman Tariq, MD (Johns Hopkins School of Medicine, Baltimore, MD), who runs the headache center at his institution and published a comprehensive review paper on PFOs and migraines last year, told TCTMD that PFO closure is not “a bona fide treatment for migraines.”

Citing feeble evidence of what propagates a migraine in the first place—specifically, a study on rats that observed a connection between microbubbles inserted into their carotid arteries and cortical spreading depression—he said that there are many complex mechanisms behind them that are not yet fully understood. “The whole theory behind this [study] is flawed to begin with,” Tariq added.

The whole theory behind this [study] is flawed to begin with. Nauman Tariq

In his experience, Tariq said, several of his patients have undergone PFO closure, but he has seen no improvement in their headache frequency or severity. While he’s not surprised that the trial didn’t meet its primary endpoint and said one related to migraine days instead of attacks would be “reasonable,” Tariq questioned the clinical value of the statistically significant drop in migraine days observed in PREMIUM.

An invasive procedure that is not risk-free is a tough sell for a patient having only intermittent headaches—fewer than 15 days per month, he observed. This is supported by the fact that the PREMIUM researchers had such a “hard time” enrolling patients.

“At the end of the day, you have to look at the big picture,” said Tariq. “If [PFO closure] would have taken away the migraine forever, yes sure. If the number of days are let’s say 15 days a month and if it goes down to 2 days a month, then yeah sure. One could give it some reasonable thought.” In PREMIUM, however, patients in the PFO arm had 1.4 fewer migraine days per month compared with controls. This, he continued, “is not significant enough to warrant this invasive treatment.”

Not surprisingly, Tariq said he does not endorse this procedure for his migraine patients, and the only reason he would undergo it himself would be as a last-ditch effort. If he had almost daily migraines with aura in 80-90% of them, they were not alleviated by any available intervention like drugs, Botox, and blocks, and he had “a poor quality of life and I'm missing work, then yes, I could consider that PFO closure device, only in that particular case.”

‘That’s Why You Do the Research’

On the other hand, both Tobis and Sommer said they would have the procedure if they were in a situation similar to that of the patients enrolled in PREMIUM. “A PFO closure is a ridiculously safe procedure, so if there’s a one in eight chance that we can cure your intractable migraines, then that’s worth something,” Sommer said. “The question is: could you find a) somebody to do it and b) an insurance company that's going to pay for it? Because that's the biggest stumbling block right now.”

According to Tobis, it took 2 years to simply publish the findings from this trial due to some reluctance from the neurology community. But Tobis said he’s “optimistic” that researchers will learn from this study given its “meticulous” methodology and sham control. “I believe ultimately it will be proven that there is a group of patients with migraine whose trigger is due to some substance or just low oxygen that goes from the right atrium to the left atrium through [the] PFO to trigger their migraines, and that if you close the PFO, you'll stop that trigger and reduce the migraine frequency,” he predicted.

Tobis said he is in talks with “some of the companies to sponsor another trial” and Sommer said he has been to the FDA with WL Gore to begin planning another PFO closure in migraine study. Contacted by TCTMD, WL Gore declined to confirm that such a trial was in the works saying only that “we continue to evaluate possible means for entry into these spaces recognizing that, as with PFO closure for stroke prevention, careful patient selection and trial design is critical for demonstrating the therapeutic benefit of PFO closure in these populations.”  

Removing Blinders

In an editorial accompanying the study, Brian Whisenant, MD (Intermountain Heart Institute, Salt Lake City, UT), and Mark Reisman, MD (University of Washington, Seattle), write that “PFO closure is not a cure for migraine to be applied broadly, but may be an important therapy for some. Given the tremendous unmet need of additional migraine prevention therapies, the safety of PFO closure, and ongoing observations of migraine improvement in some patients, future research must focus on removing the blinders and identifying those who may be most responsive to PFO closure.”

PFO closure is not a cure for to be applied broadly, but may be an important therapy for some. Brian Whisenant and Mark Reisman

“Migraines are so awful when you've got enough of them,” Sommer said, adding that his group has “cured” several hundred patients. “We have patients who have a completely new life—people who previously were unable to take care of their kids and they couldn't work and they couldn't go to school, and now they have a normal lifestyle,” he reported. “So, it’s very compelling.

“The key to treating patients with migraines and PFOs is developing strategies for determining which PFOs are mechanistically related to the migraines and which are incidental. I’m very optimistic,” Sommer continued. “Some of the other neurologists and cardiologists . . . are much more skeptical, but we'll see. That's why you do the research.”