S5:Ep2 – The Relationship Between Patent Foramen Ovale and Migraine



Thank you to our 2023 education sponsor, Lundbeck!


Voice-over: Welcome to Spotlight on Migraine hosted by the Association of Migraine Disorders. Join us for fresh perspectives by medical experts and advocates as we explore the spectrum of migraine and dig deeper into this complex disease. How are patent foramen ovale and migraine connected? Studies show that adults with this cardiovascular condition are more likely to have migraine. Stay tuned as Dr. Alexander Postalian dives into the latest research exploring how PFO closure procedures may or may not benefit a person living with both conditions.

Dr. Alexander Postalian: Hello, my name is Alexander Postalian. I’m an interventional cardiologist at Texas Heart Institute here, in Houston, Texas. First of all, I want to thank the organizers for the invitation. As always, it is an honor and a privilege.

And I’m going to talk a little bit about patent foramen ovale, or PFO, and migraine headaches, which is a controversial topic. Where do we stand today? First, some disclosures. And I have to say, I have a hammer, from the saying, if you have a hammer everything looks like a nail. I am an interventionalist. Procedures are a big part of what I do. I get paid to close PFOs. But having said that, I am also an enthusiast of doing the right thing for patients every single time. So I’m going to say, I’m a reasonable hammer, and I’m going to try to provide a nonbiased view on the topic.

And to prove that, I’m going to start by saying, don’t do it. These are the conclusions. Don’t close PFOs in patients with migraines, yet. Maybe never but it seems like things are changing a little bit, and these are exciting times. So let’s go through it.

A PFO is a communication between the left and the right side of the heart at the atrial level, which is something that should not be present in adults. But it is necessary for fetal oxygenation. However, this communication does not close in about 25% of people once they become adults. And we’ve seen this time and time again in autopsy studies and echocardiographic studies.

And this concern was a communication between blood, essentially, crossing from the right side to the left, which should normally not happen. And a clue that this might be an issue is that patients that have a stroke for no identifiable cause – meaning they have no atrial fibrillation, no carotid stenosis and nothing specific that may cause a stroke – about half of those patients end up having a PFO, which is more than 25%. So that’s a clue that some things may be brewing in that channel that may not be good.

So when do we close these, officially, or in current practice? The number one thing is in patients that have stroke of undetermined cause, as I just discussed. And the patients that benefit the most are the ones that have recurrent strokes. If there’s enough blood crossing from one side to the other, enough to cause deoxygenation and sometimes this is positional, then we can close it.

We close other types of shunts like ASDs, which are atrial septal defects, which are just holes, as opposed to a PFO, which is kind of like a valved communication, and we close those more often for deoxygenation and other reasons, and right-heart strain, and things of this sort.

It’s important to note, though, that closing a PFO, this day and age, is a relatively straightforward and safe procedure, which was not the case in the beginning of the experience.

These are the guidelines for stroke management and TIA in 2021. And they state that it is now considered reasonable to percutaneously close a patent foramen ovale in patients with certain criteria and have a stroke of undetermined cause. So this is what we do.

And how do we do it? Just real brief, we stick the femoral vein with ultrasound guidance. We insert a 9 French sheath, which is about a 3 mm diameter sheath, which is where through which we deliver the device. We cross from the right side to the left atrium using a wire. And we plug the device under echocardiographic guidance, which might be intravascular echo or transesophageal echo.

It’s relatively quick and straightforward. This is just a representation from the company. We cross with a catheter. And then once we have a catheter on the left side, we deploy the device, which once you take it out of the catheter, it takes its shape. And that was the left side. And this is the right side. The device does this on its own. Then we release it. And that’s it.

So over time, this device gets covered by endothelial cells, so it’s not recognized as a foreign body, and it prevents thrombi from crossing from the right side to the left side. That’s the idea.

Now onto migraine headaches. And you know what migraine headaches are, you’re experts. But do you, though, do you really? Because this thing is kind of complicated. We’ve gone through many theories in the past, vascular. I mean, right now, we know it’s a neuronal phenomenon that starts first, and then results in maybe vascular changes that explain the headache. But it’s complex. And when things are this complex, it usually means that we don’t fully, fully understand it. Especially when we use so many drugs with so many different mechanisms to treat this.

It’s a very common disease. It can sometimes be disabling. And we know that vascular changes at least play some role in the development or sustaining of a migraine headache. And if we get into the aura discussion of migraine, then perfusion seems to play an even more significant role.

This is an old study from the 90s that shows that right before you get the headache, but during the aura phenomenon, you get hypoperfusion in the occipital lobe, and you can see it on the right, the black areas. And the occipital area is the one that processes visual stimuli. So it makes sense to have visual aura there.

And then, during the headache, you get hyperperfusion, and then you return to normal cerebral blood flow. So there’s some role of blood flow and vascular changes played here.

PF1 migraines have been associated for quite some time, but this is one of the studies that sort of started to show a more serious link. This was in 2008. And it showed that if you have a PFO, the likelihood of you having aura, migraine with aura, was about 3 times higher than if you didn’t. And if you have migraine, the likelihood of you having a PFO was about 2.5 times greater than if you didn’t have the migraine. So, quite a significant link there.

Then we looked at association of PFO and migraine in patients that had a cryptogenic stroke. Some interesting things can be derived from this study, which was published in 2018. The prevalence of PFO, in patients that have cryptogenic stroke, was about 50%. Which is what I described earlier, more than 25%.

Now if you had a cryptogenic stroke and a migraine, then 79% of those patients had a PFO. And if you had a cryptogenic stroke, migraine, and aura, then 93% of those patients had a PFO. So, it’s a pretty strong link. I mean, causalities are hard to establish, but it’s a strong association.

Now enter white matter lesions. These are lesions that we can see in multiple different scenarios, including in patients with migraine. These are more common. And about five times greater than in controls. The hypothesis is, are these white matter lesions maybe ministrokes that could be triggering migraine attacks. It’s hard to say.

So why does it clearly make sense that PFOs are linked to migraine headaches? Well, there’s a somewhat strong statistical association. It is possible that small emboli, crossing from the right side to the left side, can trigger a neuronal event that results in a migraine headache, and white matter lesions could be proof of this.

Maybe substances that are usually metabolized by the lungs but can now skip the lungs, because they go from the right atrium to the left atrium, without going through the lungs, such as serotonin, may get to the brain, and cause a migraine headache. And if you take rats and you inject microemboli into their carotids directly, it can trigger cortical-spreading depression which is thought to be the initiating factor of migraine. So, definitely yes, but maybe not so much, not so quick.

There are studies that have shown that there is no statistical association, even though most have shown there is some association. The white matter lesions can be anywhere in the brain. And not necessarily in areas that correspond with aura. For example, aura can be visual, but you can have white matter lesions in the frontal lobe. So, that makes no pathophysiologic sense.

Some of these theories have been relatively disproven, especially the ones that involve substances crossing from one side to the other. And the overall quality of the evidence is low. It’s mostly observational studies.

I think it’s fair to say that there seems to be an association, but what happens if we fix the PFO. This is the final proof, in theory, it should be, so enter the clinical trials. Number one was the MIST trial. This one was done in 2008 and it published in Circulation and done in Europe. It used the STARFlex device, which is no longer in use, versus a sham procedure. So a very high-quality control group. And this what they found.

They had patients that had migraine with aura that failed prophylactic therapy. The primary endpoint was important to mention because they looked at no headache at three to six months. That’s a pretty, pretty hefty goal. No headache, that’s hard to achieve with any therapy. They had many secondary endpoints. About 75 patients in each group, and they found no difference in the primary endpoint or most secondary endpoints.

But they did show a small reduction in total migraine days. Now, this is Dr. Dowson, and he’s a neurologist and it looks like for him it’s a no, about the association. However, there are some issues with the trial, as with any other study, the primary endpoint was difficult to achieve. Follow-up was short. They had few patients. And this device is no longer available. There were two studies, this one with migraines and another one with cryptogenic stroke that showed no efficacy, high complications, so the company went bankrupt. This is not the device we use today.

The second trial, the PRIMA trial. This study was published in the European Heart Journal, done in Europe and Canada, published in 2016. This one compared the Amplatzer PFO Occluder, which is one of the more common devices used today, versus medical therapy. Not sham. So the control group is not as high quality.

It included patients that had migraine and migraine with aura that failed prophylactic therapy. This had a more reasonable primary endpoint of reduction in migraine days at 9 to 12 months with many secondary endpoints. About 100 patients.

And they found an overall reduction in migraine parameters in both groups, the intervention and the medical therapy group, but more so in the PFO closure group. Although it didn’t reach statistical significance, as you see on the P wave here on the right.

But, if you look at patients that had a migraine with aura, there was a statistically significant difference. And of note, in 10% of the patients that had PFO closure, there was complete cessation of headache, and no patient in the medical treatment group had complete cessation of headache. Ten percent is not zero, so, it’s also an important finding.

This trial also had issues. It stopped early because of low, slow enrollment. A few patients, and it is difficult to enroll for trials like these. Not sham controlled. And the device failed to eliminate shunting in about 12% of patients. But honestly, in clinical practice, sometimes we see some shunting remaining from left to right, right to left, but it can decrease massively after closure, versus before. 22% of patients were lost to follow-up.

So, this is Dr. Meier, one of the investigators and it looks like for him, it’s a maybe. And he’s an interventional cardiologist.

Number three, the PREMIUM trial. This study was published in 2017 in JACC. It was done in the US. And also compared the Amplatzer PFO occluder versus medical treatment, the same as the previous one. It looks like I already said it’s a maybe but anyway.

It had patients that had migraine with and without aura. Primary endpoint was assessing responder rates, which was 50% reduction in migraine attacks. Multiple secondary endpoints. Had more patients, about 230. There was no difference in the primary endpoint, but there was a significant reduction in migraine days. And the rates of reduction was higher if the patients had aura.

On the right is Dr. Jonathan Tobis, also, it’s a maybe for him. And he’s also an interventional cardiologist.

Now these are the three bigger trials that we have. But we have other data, and some are more recent. There’s a pooled analysis, published in 2021, that analyzed the results of PRIMA and PREMIUM which are the ones that used the Amplatzer PFO occluder versus medical therapy and excluded the STARFlex device.

And it found, well, I did an intention-to-treat analysis, so basically, it takes all of the patients from both trials and reanalyzes the data together. And they found a significant reduction in migraine attacks, and higher responder rates, although that didn’t reach statistical significance. Complete cessation of headache was more common in patients that had PFO closure versus those that did not. Procedural complications were less than 1%, significant procedural complications.

The issue with a pooled analysis is that it’s a pooled analysis, it is not a new trial. There’s bias, so there could be spin. When you read the study you can somewhat tell that the authors had sort of maybe an idea of what they should find. And that reflects in the text, in the conclusions, and maybe perhaps how the data were analyzed. So, that’s important to keep in mind.

The two studies had similar, but not the same patient populations. So, it looks like for Dr. Mojadidi, well he did this as a fellow, so I said, we just want to finish, and Dr. Tobis was also involved in this study. No, actually, I kid, he’s a very prolific author and investigator. It’s also a maybe yes for these two.

There was a meta-analysis and I’m going to go quickly through this one, because it’s not the best data, published in 2022. About 1200 patients were analyzed from observational studies, cohort, case series and what not, and they found that PFO closure could reduce the frequency of headaches. And is an efficient treatment for migraine attacks with aura, which seems a little bit like overreaching, but strong conclusion.

This is the most recent study, significant at least, published on the matter and this is a little bit more basic science trying to find what may be the reason these two are associated. Published in JACC, basic science, done in Italy. And what they assume is that migraine with aura…the patients that have migraine with aura have more platelet activation, more inflammation, and this perhaps results in more pain.

So they had 78 patients that were taking aspirin that had migraine headaches with aura, that had PFO closure as well. And they compared them with 12 healthy subjects, that were also on aspirin to adjust for the confounding issue that what affect might aspirin have by itself.

The official indication for closure in these patients were lesions on brain MRI that were assumed to be strokes. It could have been white matter lesions for other cause, but this is real practice. They checked the blood of the patients the day before the procedure and six months after, and they found that there was complete resolution of migraine headaches with aura in about 70% of the patients, and significant improvement in the rest. Essentially they looked at many parameters of migraine improvement, all of these are subjective, but they found statistically significant improvements in all of the areas that they looked at.

And the assumption, if you look here at the bottom, is that platelet activation might be addressed by aspirin, but for some reason PFO closure also improves the prothrombotic milieu and oxidative stress in those patients. They did a lot of analyses, and they tried to show this.

The assumption is that for some reason these phenomenon might be closed by PFO closure. And the data are compelling, but it’s one piece of the puzzle. Let’s put it that way.

What issues does the study have? A small number of patients. Obviously, this is not a real trial to establish efficacy of the procedure. It’s a proof-of-concept study. And, of course, if there’s no real control group, no real good control group, it’s hard to assess for placebo effect.

One thing that they hypothesize is that maybe the P2Y12 receptor, if you were to block it, it would replicate some of the effects seen by PFO closure. And we use P2Y12 blockers very frequently in cardiology. Clopidogrel, prasugrel, ticagrelor for patients that have stents, so maybe that could be used as something to look at, which we haven’t done yet.

This is Dr. Trabattoni, in Italy. It seems like she is pretty convinced when you go through the paper, and she’s a cardiologist.

Now, something important to keep in mind is that all of these trials and studies have been conducted over the last few years, a couple of decades. But now things don’t look the same as they did before. We have many different therapies to treat migraine headaches, from traditional therapy and new agents, like calcitonin gene-related peptide antagonists, to aerobic exercise, acupuncture, etc., etc., all of which have shown some improvement in migraine headaches.

So, it’s a little bit different environment now than it was in the past. And this PFO with migraine headaches, it’s actually something, that is, when you get into that field, you can tell it’s a quite passionate discussion sometimes.

So, what about the court of public opinion? Facebook. I go through these groups every now and then to understand the patient experience and to better communicate with patients, so there are PFO and migraine groups on Facebook that you can go through, and you can read things like this.

So, on the left side, somebody had a PFO closure about six months and now the patient is saying that they don’t have to take anything more than a baby aspirin a day, since getting the Amplatzer Occluder, which by the way, we have three devices, I’ll talk about a little bit later. Haven’t had any migraines.

On the right side, again, it’s my first PFO closure, definitely happy I did it, I haven’t had another stroke and my shortness of breath and migraines have subsided. So, it’s reassuring.

But, you can also see things like since I had my PFO closure I have had over 67 auras without migraine. Then on the right side, I had my PFO closure done in 2015, and I have since had an increase in every possible symptom and more. Migraines, severe chronic fatigue, etc., etc.

And then even in the same thread, somebody asks, have you had a PFO closure for migraines and what happened. I had mine closed, haven’t had any migraines. No migraines since. And then on the bottom, closure did not help my migraines at all. On the right side, migraines are gone.

So, again, this is not helpful. This is level Z evidence. Some patients feel strongly about this. Some doctors feel strongly about this. We have to navigate this with care.

So, it doesn’t seem like the relationship between PFO and migraines is a straight line. If anything, it is a very convoluted communication and where we have to tease out why these two are related and what we can do about it, and how can we do it best.

So what gives. How can we conclude? Don’t close PFOs to treat migraine headaches, yet. The more recent data is actually encouraging, that there might be a link, but it’s a lower quality data than the three clinical trials that I showed.

There is a possible role for a select group of patients. If somebody has frequent and disabling migraine with aura that responds to no medical treatment. And they have a PFO that has significant shunting, I think, it is reasonable to explore this idea. And if you have white matter lesions in the brain that could potentially be embolic events, that patient may benefit from PFO closure. But again, this cannot be recommended in general.

And there’s another trial that is being done that is the RELIEF trial using the Gore device. We have, in general, there are three devices, the Amplatzer PFO very commonly used, the Gore device and then the NobleStitch, which leaves less material in the heart. The one that has the most data is the Amplatzer PFO Occluder. But, anyway, this trial is coming. Not anytime soon. It’s still a few years away. But it hopefully will help answer the question.

Thank you very much for your attention. This is my name and my email. Please feel free to share questions, comments, or concerns, and I’d be more than happy to respond. See you next time, maybe.

Voice-over: Thank you for tuning into Spotlight on Migraine. For more information on migraine disease, please visit MigraineDisorders.org.


*The contents of this podcast are intended for general informational purposes only and do not constitute professional medical advice, diagnosis, or treatment. Always seek the advice of a physician or other qualified health provider with any questions you may have regarding a medical condition. The speaker does not recommend or endorse any specific course of treatment, products, procedures, opinions, or other information that may be mentioned. Reliance on any information provided by this content is solely at your own risk.