Seattle, WA - Adding both more clarity and more confusion to a field already peppered with conflicting data, a new study has found that people who have migraine with aura are more likely to experience migraine relief following patent foramen ovale (PFO) closure than people who get migraines without aura [1]. But, more puzzlingly, migraine relief can occur even if residual shunting occurs after the PFO is closed. The counterintuitive findings come from a retrospective analysis of 77 patients treated at the Swedish Medical Center, in Seattle, WA, between April 2001 and June 2005.

"This retrospective study included a group of patients with ischemic stroke, who also had a history of active migraine at the time of PFO closure," Dr Jill Jesurum (Swedish Medical Center, Seattle, WA) told heartwire. "In this sample of patients, we found that those who had migrainous aura had an odds ratio of 3.5 that they would more likely experience migraine relief following PFO closure than those who did not have aura."

Jesurum notes that while the MIST trial—the only completed randomized study to date addressing this question—enrolled only patients with migraine with aura, other trials still ongoing, ESCAPE and PREMIER, are enrolling migraineurs with and without aura. "The jury is still out, basically, as to the migraine profile most likely to benefit from PFO closure and whether PFO closure is efficacious in reducing migraine frequency," Jesurum said.

But the more perplexing finding in Jesurum et al's analysis was that full closure of the PFO was not a prerequisite for migraine reduction. "That was very surprising for us as well, and I think that has not been previously reported, at least in long-term follow-up," she said. "A possible explanation is that migraineurs have increased cortical hypersensitivity, and a large right-to-left shunt permits a greater volume of vasoactive chemicals and microaggregates, such as activated platelets, to cross the PFO and trigger migraine. It is speculation that a significant reduction of shunt and transmission of these chemicals and microaggregates may reduce the threshold of cortical hypersensitivity and therefore reduce migraine frequency."

The notion that migraine might be linked to PFO stems from observations that people who had undergone PFO closure for decompression illness or secondary prevention of cryptogenic stroke reported fewer or no migraines postprocedure. Investigators have speculated that some migraines might be caused by microemboli or chemicals normally filtered out in the lungs, traveling via the PFO to the brain. The disappointing MIST trial results indicated that PFO closure had no significant impact on migraine cure or on reducing migraine frequency despite a low rate of residual shunting postprocedure. As reported by heartwire, however, one of the investigators for the MIST study has disputed the residual shunt data and is currently in the middle of a legal imbroglio with the study sponsor, NMT Medical.

In Jesurum et al's study, investigators reviewed late follow-up data, including transcranial Doppler (TCD) imaging results six and 12 months after PFO closure and migraine questionnaires from 77 patients who had received PFO-closure devices and who reported migraines at baseline. In all, 71% of patients had migraine with aura, while the remainder reported experiencing migraine without aura at the start of the study.

Of the 67 patients in whom both PFO follow-up and migraine information was available, 54% of patients reported complete migraine relief and an additional 25% experienced "substantial" relief (more than a 50% reduction in migraines per month). Overall, more than one-third had residual right-to-left shunts, defined as 30 embolic tracks on TCD at a median of 366 days postprocedure. Strikingly, there were no significant differences in migraine relief between people who experienced no residual shunting vs some residual shunting.

Where differences in migraine relief were seen, however, was according to type of migraine at baseline: 86% of subjects with migraine with aura at baseline experienced migraine relief at follow-up, compared with only 59% of subjects reporting migraine without aura. Of note, residual shunt rates were similar between the two groups. Overall, say the authors, "Migraineurs with aura were 4.6 times more likely than migraineurs without aura to experience relief after PFO closure."

To heartwire, Jesurum emphasized that the analysis was underpowered, retrospective, and not designed with migraine reduction as a primary end point. "This needs to be repeated prospectively and with an adequate sample size before any definitive conclusions can be made," she said.

She also pointed out that TCD measures embolic tracks in the cerebral vasculature from contrast that's been in injected into an antecubital vein, but it is not specific to contrast passing through the PFO. "Most of the time, a positive TCD is indicative of a PFO, but this is not absolute," she explained. "A positive TCD can also be due to a pulmonary arteriovenous malformation. So there can be other sources of shunt, although the most probable source is a PFO. That is another limitation of this study. We assumed the residual shunt was related to incomplete PFO closure, but that was not confirmed in all cases."

As such, even a patient with successful PFO closure could continue to have right-to-left shunting via another source. Indeed, in other research that Jesurum and colleagues are in the process of preparing for publication, they have found that almost one-third of patients referred to the cath lab for PFOs also have a secondary source of right-to-left shunt.

Jesurum, for the time being, says she can't even speculate on whether PFO closure will ever prove to be a solution for migraine sufferers. "I definitely do not believe that all migraine or even all migraine with aura can be attributed to PFO. We do know from epidemiology studies that migraineurs who have aura have a higher prevalence of large PFOs than the general population. But I don't believe we can say that this is the be-all and end-all treatment for all migraineurs. There may be a subpopulation of migraineurs for whom there is some potential causal mechanism between PFO and migraine and who may benefit from PFO closure, but the mechanism has yet to be determined, as does the efficacy of PFO closure."

Commenting on the study for heartwire, Dr Jonathan Tobis (University of California, Los Angeles [UCLA]) acknowledged that the theory of a "threshold" effect based on shunt size is a plausible hypothesis to pursue. "If a little bit of chemical goes through, you don't get a migraine, but if a lot goes through, you do get a migraine: I think that's a reasonable theory," he said.

But emphasizing a point also made by Jesurum, Tobis highlighted the fact that the randomized, controlled trials of PFO closure for migraine are not necessarily enrolling the same kinds of patients as the ones studied in Jesurum et al's study or, for that matter, most of the studies that first led to the migraine-PFO hypothesis evolving in the first place.

"The population that seems to respond to PFO closure is those people with episodic migraine, who came in with cryptogenic stroke or some neurologic symptoms, their PFOs were closed, and their migraines got better," Tobis explained. "The problem is the FDA won't let us treat these kinds of patients with a permanent device." Instead, the FDA is asking for companies to study people who are severely disabled by migraine, meaning six or more days per month, Tobis said.

"Because of the risk/benefit ratio associated with permanent implants, they want us to study only the most disabled patients, which is reasonable. The problem is we don't know if that will actually work or not, particularly with the results of MIST weighing on everybody's minds."

He notes that while the PREMIER and ESCAPE trials are still ongoing, the study sponsors have gone back to the FDA to explain why they are having difficulty recruiting patients and to ask whether adjustments can be made to the eligibility criteria. "So hopefully things will be a little bit liberalized, and they will be able to increase enrollment," Tobis said.