PREMIUM Trial: PFO Closure Primarily Benefits Migraine Patients With Aura

Percutaneous closure of patent foramen ovale (PFO) does not reduce the frequency of migraine attacks, though it can decrease the number of days per month patients are affected by migraines, researchers reported at TCT 2015. Importantly, patients whose migraines are accompanied by aura appear most responsive to the treatment. 

Jonnathan TobisFor the PREMIUM trial, Jonathan M. Tobis, MD, of the David Geffen School of Medicine at UCLA, Los Angeles, Calif., and colleagues randomly assigned 230 patients who had a PFO and medically intractable migraine with or without aura to undergo closure with the Amplatzer PFO Occluder (St. Jude Medical; n = 123) or a sham procedure (n = 107). Both groups also received medical therapy.  

The procedure was generally safe, with only one device-related serious adverse eventa transient episode of atrial fibrillation (0.5%)occurring during 1 year of follow-up.

There was no difference between the groups in the percentage of responders (primary efficacy endpoint), defined as those having at least a 50% reduction in migraine attacks per month in months 10 through 12 after randomization. However, the PFO closure group had a lower mean number of headache days per month (see Table).

Table. Outcomes in Migraine Patients


PFO Closure

(n = 123)

Sham Procedure

(n = 107)

P Value

Percentage of responders




Mean headache days/month




In patients whose migraines mostly included aura, the rate of treatment response was greater with PFO closure than with the sham procedure (49% vs. 23%; P = .015). In this subset, complete remission of migraine occurred in 10.8% of the closure group compared with 1.5% of the sham group (P = .02).

Sham Control Crucial 

“Based on all of the observational reports, we thought that more patients would respond to PFO closure,” Tobis told TCT Daily. “I was also surprised by the high placebo [response] rate, but that is why we planned this as a double blind, sham-controlled trial. It confirms that this [protocol] is important for device trials where the endpoints, such as relief of headache, are subjective.”

Not enough is understood about migraines, Tobis acknowledged, but he conjectured that the apparent benefit of PFO closure for patients who have migraine with aura may relate to a greater grade of shunt in those patients. He added that his group has unpublished data that support this hypothesis.

The study included patients who experienced 6 to 14 days of migraine per month as assessed by a headache specialist and had not responded to three preventive medications. All had a significant right-to-left cardiac shunt (Spencer grade 4-5). Randomization occurred on the catheterization table after proof of PFO was obtained.

“I believe PFO closure should be recommended for patients with aura who have frequent migraines,” Tobis concluded. However, he added, “I suspect the FDA will want another clinical trial.”


  • Tobis reports serving as a member of the PREMIUM steering committee.



We Recommend