Miriam E. Tucker

June 26, 2015

Washington, DC — The latest study on patent foramen ovale (PFO) closure for treatment of refractory migraine failed to achieve its primary endpoint but still leaves the door open for potential efficacy in a subset of patients, researchers say.

Final data from the Prospective, Randomized Investigation to Evaluate Incidence of Headache Reduction in Subjects With Migraine and PFO Using the AMPLATZER PFO Occluder to Medical Management (PREMIUM, St. Jude Medical), were presented here at the American Headache Society (AHS) 57th Annual Scientific Meeting.

PREMIUM is the third randomized trial to examine the question, the second to use the Amplatzer device, and also the second to use a sham procedure as a control.

In PREMIUM, implantation of the device to close the PFO along with medical management did not meet the prespecified primary endpoint of 50% reduction in migraine attack frequency in 117 patients compared to 103 who received a sham procedure and medical management.

However, some secondary endpoints did achieve significance, and 8.5% of patients — 11% in whom most migraines were accompanied by aura — had complete migraine remission.

"It is true that the PREMIUM study did not reach its primary endpoint, and this is the primary message," lead author Andrew Charles, MD, professor of neurology and director of the Headache Research and Treatment Program at the David Geffen School of Medicine at the University of California, Los Angeles, told Medscape Medical News.

"But it is intriguing in that it did reach a secondary endpoint of reduction in migraine days — a more commonly used endpoint in preventive therapy trials — and in the patients with aura the majority of their attacks clearly responded to closure of the PFO."

He added, "I guess the message from my perspective is that there are many 'switches' that can be flipped that lead to migraine in different patients, and a patent foramen ovale may be a significant 'switch' in a small but significant percentage of patients with migraine."

But some headache experts aren't buying it. Elizabeth W. Loder, MD, chief of the Division of Headache and Pain at Brigham and Women's Hospital and associate professor of neurology at Harvard Medical School, Boston, Massachusetts, told Medscape Medical News she feels the question is answered.

"I think this is fairly definitive," she said. "There is morbidity associated with closing PFOs. It's not a procedure that's free of adverse effects. So I think what this study tells us the balance of benefit to harm is not a favorable one and it should not enter practice."

Asked whether PFO closure should be studied in a subset of patients with migraine, Dr Loder replied, "I'd have to think about that, but I'm not so sure when you have other forms of treatment that are less invasive. I would be more cautious."

Controversy Continues

This topic has a long and controversial history. About a quarter of the general population has a PFO, and average PFO size has been documented to be larger in patients who have migraine with aura. Observational studies have suggested that PFO closure performed for other reasons in patients with migraine brings headache relief.

When the first of the three randomized trials, the Migraine Intervention with Starflex Technology (MIST), was presented at a cardiology conference in 2006, it appeared to have both negative and positive findings, depending on the endpoint used. Subsequent publication of the data, however, confirmed that none of the end points were positive, and numerous research improprieties were alleged.

Trial sponsor NMT Medical, which had initially planned to continue with the MIST II trial, cancelled the study and subsequently went out of business. The current Amplatzer manufacturer, St Jude Medical, closed down another trial in 2008.

And then in 2014, St. Jude's PRIMA trial — smaller than PREMIUM and MIST and not sham-controlled — also produced mixed results in patients who all had migraine with aura.

Given all that, headache expert Robert Shapiro, MD, PhD, professor in the Department of Neurological Sciences at the University of Vermont, Burlington, told Medscape Medical News that he doubts another trial will be conducted. "Obviously the primary endpoint wasn't met, but they're still hanging by a thread that maybe there may be a subgroup who may be hyper-responders. I can't imagine somebody funding another trial, but maybe somebody would."

However, Dr Shapiro noted that his view hasn't changed since he penned an editorial about this in 2006. "My opinion is that nobody should get a PFO closure device placed to treat migraine outside of a controlled clinical trial, hands down…I don't think the data are strong enough to recommend that anybody get their PFO closed outside of a clinical trial. If there are no more clinical trials, then there won't be any justification for someone to have this done."

Interventional cardiologists contacted by Medscape Medical News for comment see it differently.

"While PFO closure was not effective in the overall population, the study may have identified a subgroup of patients in whom the procedure is very effective," said James Blankenship, MD, director of the Cardiology Department and Cardiac Cath Labs at Geisinger Medical Center, Danville, Pennsylvania.

"Among my patients who have had PFO closure to prevent recurrent stroke, a few incidentally had complete relief from life-crippling migraines," he added. "For these patients, it was a life-changer. I did not give them any expectation of relief of their migraines, and when it occurred it was such a dramatic change that I cannot attribute it to the placebo effect. I think it was real."

The difficulty ahead, in his view, is to identify the patients who may benefit from PFO closure and those who will not benefit, and treat each group appropriately.

Cliff Kavinsky, MD, PhD, professor of medicine and pediatrics and director of the cardiovascular fellowship training program and the Center for Adult Structural Heart Disease at Rush University Medical Center, Chicago, Illinois, pointed out that trials of PFO closure for migraine prevention are extremely difficult to execute because of the need for rigid exclusion criteria and for effective methods of blinding to overcome the strong placebo effect.

"All of the [current randomized clinical trials] suffer to a variable extent from the above-described problems. Most students of the field feel strongly that there is a treatment effect of PFO closure on migraine symptoms, yet it is very difficult to demonstrate in a RCT [randomized controlled trial]," he said.

But Dr Loder believes that the use of a sham procedure as a control — done in both PREMIUM and MIST — addressed that issue. "That's the right way to study these interventional procedures and I think [the PREMIUM investigators] deserve a lot of credit for doing that. Too often surgical procedures enter practice without being rigorously studied and without having a credible sham procedure. That really allows you to distinguish — especially in a subjective illness like migraine — the effects of expectation and belief from the specific effects of the treatment."

Mixed Results — Again

PREMIUM, conducted in 29 US centers, enrolled patients aged 18 to 64 years who had a diagnosis of migraine with 6 to 14 attacks per month, with or without aura, and who had a PFO documented via transcranial Doppler. Participants must have not responded to trials of at least three migraine medications from two of different classes.

A total of 226 patients received femoral puncture, and those meeting all criteria were randomly assigned during the catheterization to have the device implanted or sham procedure. Both groups also received medical management. Patients and neurologists were blinded to the randomization.

Two thirds of the patients in both groups had migraine with aura, and nearly half had migraine both with and without aura.

A total of 116 device patients and 103 controls completed the 12-month follow-up. Those in the latter group were offered PFO closure at 1 year, and 87 received the device.

The primary efficacy endpoint, a 50% reduction in attacks per month based on headache diaries during months 10 to 12 after randomization, was achieved in 45 of 117 (38.5%) with the device vs 33 of 103 (32%), a difference that did not reach statistical significance (P = .32).

However, the secondary endpoint of change in mean migraine days per month was significant, with reductions of 3.4 days with the device compared with 2.0 days for the controls (P = .03).

Among patients in whom most attacks were with aura, headache days decreased significantly: 19 of 39 (49%) vs 9 of 40 in controls (22.5%) (P = .015).

In the entire study group, complete remission of migraine occurred in 10 of 117 with the device, compared with just 1 of 103 controls (8.5% vs 1.0%; P = .01). For those with a diagnosis of migraine with aura, the results were 8 of 74 (10.8%) vs 1 of 68 (1.5%) (P = .02).

The only device-related serious adverse event in all 202 who received it (including the controls after 1 year) was a single episode (0.5%) of transient atrial fibrillation. Other procedure-related adverse events included hypotension, paroxysmal ventricular tachycardia, phlebitis, vascular hematoma, vascular pain, and vasovagal response.

In closing his presentation, Dr Charles said, "Migraine with aura may have therapeutic targets that are distinct from migraine without aura" and pointed to a study presented earlier during the same conference session showing that migraine with aura may respond less well to sumatriptan.

He said that more information may be forthcoming from PREMIUM. "This is going to be a very interesting dataset to mine because it's a very long study with lots of attacks recorded. I think more may come out of this in terms of responders — could there be significant anatomic differences?"

Dr Charles currently receives honoraria/consulting fees for his role on the advisory boards of Amgen and eNeura. He received honoraria/consulting fees through December 1, 2014, for his role on the advisory board of St. Jude Medical. He also receives research grants from Takeda. Dr Shapiro receives consulting fees/honoraria from Eli Lilly and has equity/stock interests in BonTriage. Dr Loder, Dr Cavinsky, and Dr Blankenship have disclosed no relevant financial relationships.

American Headache Society (AHS) 57th Annual Scientific Meeting. Abstract LBO1. Presented June 20, 2015.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.