<v Coylewright>So I'd like to introduce</v> Dr. Volpi to come up and talk about the burden of PFO associated stroke. He's gonna help us understand the problem that we have ahead of us, and then we can talk about some of the potential solutions. <v Volpi>Thank you, Dr. Coylewright.</v> Thank you, Dr. Messe. Thank you all for being here. It's really a privilege to be part of such a distinguished panel. I think when it comes to what's fun about PFO, what's fun about what we do in vascular neurology, it's diagnosis. And so much of what we do that attracts people maybe to vascular neurology, is the kind of X Games, running down to the ER, doing that kind of acute stroke intervention. But over the years, I found myself more and more drawn to the outpatient world where we see these patients, and you're kind of scratching your head about what's going on with them, and you really get to think and talk to patients, and really have the essence of shared decision making. And that's what I think I'm gonna give you a little bit of my perspective on that, and then hopefully, we'll share some more questions. So really, the first question is does stroke etiology matter? I mean, that's something that we've actually addressed in some studies not too far ago. What about our current nomenclature of cryptogenic versus ESUS? The prevalence of PFO associated stroke? Then what tools do we have? And what tools do we have for improving the diagnosis? So this is why stroke etiology matters, and why it's mattered since the late '80s, early '90s when many of these studies came out, and they showed a striking treatment effect for carotid intervention, that's on the left, showing endarterectomy versus medical management for symptomatic disease, and for anticoagulation versus, in this case, placebo for atrial fibrillation. So once we see that having certain interventions for certain diseases matter in stroke, it's clear that these indications, it matters. So when studies started to evolve after that, there was a question of whether subcategories of stroke matter. Did certain diseases like carotid stenosis versus cardioembolism behave differently if they were exposed to other interventions? And we have this residual from one of those studies, the TOAST trial, which gives us the TOAST criteria, which largely defines what we think of as the categories, the large artery atherosclerosis, cardioembolism, small vessel occlusion, stroke of other determined etiology, which would be something like a dissection or vasculitis, and then stroke of undetermined etiology. Now, you'll see that actually, the word cryptogenic isn't there, and I would say that, from a definition standpoint, cryptogenic is really a subcategory of undetermined, because you could end up at undetermined because you don't know. That would be a cryptogenic or hidden cause, but you could also end up at stroke of undetermined cause because you had competing theories. You have a patient who may have some kind of valvular disease and carotid disease, and you wonder what the actual cause was. And then within the TOAST criteria, you could have possible probable certainty, and the undetermined cause represents about 25 to 40% of the patients we discharge from the hospital. I would argue the number goes down the longer you get to know a patient, but that's usually where we end up when they complete their workup in the hospital. There have been other classification schemes, and one was the A-S-C-O, which is similar, atherosclerosis, small vessel, cardiac, or other, and then within that, there were multiple other kinds of gradations of certainty. And then the one that came on the scene most recently was this term ESUS, or embolic stroke of undetermined source. I didn't even realize this, but that actually came to us as recently as 2014 in a publication of "Lancet Neurology" looking at the new oral anticoagulants. So because the new oral anticoagulants had a safety proposition, the idea that maybe we could use these across more patients, that it would be interesting to study, and so the idea was to create this classification to study those. So this is how it started, you know, we've all seen those memes now, that cryptogenic stroke and ESUS would overlap, cardioembolic and ESUS would overlap, but they were distinct in other ways. The Venn diagram was supposed to be like this. But as you know, in clinical practice, it's not always like that. The more you see, the more paranoid you become that something's an embolic stroke, and so in some ways, cryptogenic and ESUS ate everything. And this was another publication that came out that showed, you know, you could put embolic etiologies of all different kinds and all different types that would be even, you know, no more Venn diagrams, but just complete overlap. So the idea of ESUS, the concept, the framework of an ESUS led to this question is it time to even say goodbye to the construct? And it wasn't just that this overlap issue, but also that the studies that had been designed based on this construct were not particularly useful. They showed that while safety was improved, efficacy compared to anti-platelets in the ESUS dabigatran study and the ESUS rivaroxaban study did not have a signal of benefit. So what are we left with? After all of this different diagnoses, and splitting, and lumpers, what are our level one A according to AAN guidelines for patients in whom you suspect PFO etiology? Level one A guideline must do a vessel evaluation. You really do need to do that. Almost nobody will make it out of the ER without that, and EKG, same thing. So what more can we do beyond the must dos? So PFO diagnosis is challenging. The first thought might be are we even finding the PFO? I would say you can kind of smell the PFO on the patients, you know, when you're making rounds that you think you're dealing with a PFO, but you may not find it. So when you have that, there's a concern that maybe we're underdiagnosing with TTE, but then when you do find it, sometimes it goes the other way, PFO is found, and you're not really sure if it's a pathogenic cause. So this refers to some of the data that Dr. Messe was just discussing, was that the prevalence of PFO, this is kind of a seminal article that looked at the question, turning it around. In patients under 55 who had a PFO with cryptogenic stroke versus known cause, you can see that the known cause of PFO prevalence was around 10%, and PFO was about 40% in the under 55, so significantly overrepresented with an odds ratio 4.7. And then the other point, though, is that we may have fallen into the trap over the years of calling this a stroke of the young, but really, it's a stroke of the middle aged, and if you look at the patients over 55, you still see a statistically significant overrepresentation and odds ratio that remains not as robust at 2.9, but nonetheless there for patients being overrepresented having PFOS versus those patients who had known cause. Similar meta-analysis found similar in patients over 60. What else do we know about PFO itself? Separate from the stroke literature, let's just talk for one moment about PFO anatomically, and this is going back an early autopsy series of almost 1,000 patients that found the overall incidence of 27% of patients at autopsy. Now, of course, at autopsy, you have the advantage of poking around and really seeing whether a potential space is going to open with whatever one is probing that with, and it's not quite the same, so we see less of that in our noninvasive testing. But what they did see in the autopsy series is that the incidence decreased with age. So PFOs seemed to have more prevalence in patients until about age 40, and then they start to come down after that. Most PFOs were less than 10 millimeters, and that size when present, did tend to increase with increasing age. But there are other factors that go along with aging besides simply the presence of a PFO, and one is thrombophilia. So this is again, separate from PFO, separate from stroke literature, just showing what happens with thrombophilia in aging, and venous thromboembolism, and if you were to put an inflection point on this graph, you might put it at about age 45. And I'll just ask you to kind of earmark that age, because that's an important age for PFO as well. What do we know about diseases that start to occur in life as one gets older? You're at more risk of acquired conditions that would make you higher risk of thromboembolism, especially cancer, other chronic diseases, and obesity. And then what do we know about the right heart? The right heart, and I shouldn't be saying this with cardiologists here. I don't know anything about the right heart, (participants laughing) but the reality is, from what I read, that there's a stiffening of the heart with aging and that causes right heart pressures to go up, so the degree of shunting may increase, and also this starts to occur at, you know, the mid 50s and above, so all kinds of factors that may make a person more likely to be have a stroke related to PFO with aging. So what's the general approach that I use for monitoring or bubble studies? Well, if a patient's under 60 and they have a positive TTE, so be it, you can of course use that, but if they have a negative TTE, I do recommend a bubble study to be pursued with a TEE or TCD to confirm in fact that there's no PFO, and we'll talk about that data in a minute. And then to think of, you know, atherosclerosis as a burden of risk factors over time, and if a patient lacks those risk factors, then they're unlikely, less likely to have atherosclerosis as a cause. And then the FLAIR imaging and vascular imaging gives us a clue of whether the patient actually has atherosclerosis. So in a patient who's over 60, has a low burden of atherosclerosis, normal vessels, they should have a bubble study and at least a 30 day heart monitor. Keep in mind, in patients under 60, things like hyperthyroidism, alcohol use would also make you wanna do heart monitoring, and an abnormal echo would make you do heart monitoring. So what's the keep it simple approach if you don't wanna remember all those things? Just normal heart, normal FLAIR, normal vessels, have a high suspicion for PFO at any age. I did ask you to remember the age 45, and the reason I'm coming back to that is the table one data from the REDUCE trial. I had the opportunity to be in the study. So this is some of our data, and you can see, and unfortunately, my pointer isn't showing up there, but you can see on the REDUCE, average age was 45 or the median age was 45, compared to other studies that looked at cryptogenic stroke, that specifically, the loop recorder CRYSTAL AF study here, had a median age of 61, so different populations, and also, the CRYSTAL AF study had a significant number of risk factors in their population, and relatively few risk factors in the PFO study. So considerations in choosing whether or not to do a loop recorder, for me, if a patient's over 60, it's reasonable, if they have multiple risk factors, it's reasonable, abnormal heart, it's reasonable, but not typically if they're 59 and under, and not typically if they have a normal EKG. So what tools do we have to improve diagnosis of PFO? And I'm looking at you, Mark Rubin. You've helped us with this with the bubble study. So the transthoracic echo we know is positive, and is helpful if it is positive, and same thing, TE, of course, the so-called gold standard, but it does depend on patient positioning and sedation. But this is what we can do as neurologists, and that's the transcranial doppler. So you don't have to know anything about transcranial doppler to know this is a positive study. It comes across very easy as white dots on the screen, and that's just by insinuating the middle cerebral artery and doing the bubble injection. What do we know of previous comparisons in these studies? We know that in patients who were compared who had TEE positive, all their TCDs were positive, and among patients who had TCDs negative, all the TEEs were negative, but there will be patients who have false positives, TCD positive and TEE negative. This led to the level C guideline from American Academy of Neurology saying you may do TCD as part of your workup, but this is the added diagnostic, and that's the robotic assisted transcranial doppler. And the bubble study or BUBL study was the one that my colleague, Dr. Ruben back there was the PI for and presented last year at this conference, and it used the novel technology to detect right to left shunt, and was multicenter compared to TTE. What was found, and this slide is courtesy of Dr. Rubin, is that the patients who had a positive on a standard of care TTE was 21%, and among patients who had a TCD was 63%, so three times more likely to simply find a right to left shunt on a TCD, and you may say, "Well, maybe not all of those are intervenable shunts, not all of those are shunts that you would go after." So in looking only at the patients who had intervenable shunts, the robotic assisted TCD was positive in 27%, in TTE only 10%, so keeping in mind that there's a high risk of false negatives with TTE. I think I have a few more minutes? I have one more minute, okay. So my last minute, I'll just say we may not be as clever as we think we are, and the last point I'll make is that when we look at saying is this symbolic, looking only at a DWI image, is this a small subcortical, is this atherosclerotic or embolic? Looking at these patients, both of those ended up being from known embolic sources, so it can be deceiving based only on imaging. On the other hand, you can have these distal little cortical infarcts, is that embolic? Your radiologist will say it is, but you may find, in fact, that those patients were both from known atherosclerosis and other causes, so really, it remains important for us to look beyond simply, you know, our guesswork and into this further evaluation.

PFO

stroke

diagnosis

treatment

transcranial doppler