Hello, everyone. I'd like to welcome you to yet another episode of the Front Row series, the new educational series of the AANS NeuroU, which is the brand new educational website of the AANS. Today's webinar will be focusing on cerebrovascular. For those who are not familiar with this series, again, this is a series that features renowned experts from around the world discussing several themes in neurosurgery, including skull-based cerebrovascular tumor radiosurgery, as well as many more. Participants can have access both to the live events as well as the archive events with CME. And for the first time, participants can submit cases for expert discussion. Today's experts, again, need no introduction. We have Dr. Greg Ziffel. Dr. Ziffel is not only one of the most renowned cerebrovascular neurosurgeons around the world, but he also heads our educational online committee. He's actually responsible for a lot of the work that's being done, including this series. Again, Dr. Ziffel doesn't need any introduction. He is the Ralph Dacey Distinguished Professor of Neurosurgery and the Chairman at Wash U in St. Louis. And today he's going to be discussing the evidence behind the treatment of duraluveficulitis. On the other hand, we have Dr. Bradley Gross. Dr. Bradley Gross is a rising star in cerebrovascular neurosurgery, specifically in the field of endovascular neurosurgery. He's been a friend for a long time now, and he is really the go-to person at our department for anything complex related to endovascular. And Brad always has this gift of deconstructing seemingly very complicated situations and disease into something very simple. So I asked him today to, again, make duraluveficulitis, which are a very complicated subject for most normal neurosurgeons, and give us his two cents about how he goes around to make things simple. Dr. Ziffel is still stuck in the OR today, so we're actually going to start with Dr. Gross, and hopefully Dr. Ziffel will be joining us as we go along. So Brad, please take it away. Thank you, George. I appreciate the kind introduction. Can you see my screen now? Not yet. Okay, let me share it right over there. How's that? Here we go. Yes. Okay, great. Well, thanks again, George, for the introduction. I appreciate the opportunity to talk about a personal passion of mine and interest that is duralutrebinas fischl, as I think every case you learn something from, and I will try to distill it as simply as possible to make things as clear as I can, but again, every case is a learning opportunity for us. My relevant conflicts of interest and disclosures are here. I am a consultant for Medtronic that makes Onyx, which is frequently used by many practitioners to treat these and consult for my prevention as well. My talk will be divided into five key pieces. First, try to make a very clear definition of duralutrebinas fischl, then explore their natural history, describe their anatomy, their diagnosis, methods of diagnosis, and then treatment. So simply, what is a dural AVF, a duralutrebinas fischl? So a fischl is an abnormal connection. An arteriovenous fischl is an abnormal connection between an artery and a vein. And you can see that I've lineated that here in this angiogram on the right, the supraselective angiogram. A dural AVF obviously occurs within the dura, within the lining of the brain and spinal cord. So you have this abnormal connection that's happening in the dura. Thus, it is unique to the neuraxis because it is in the dura. It is a dural AVF, and it is this abnormal connection that defines the pathology in that white lining of the brain. So why are duralutrebinas fischl a problem? Well, simply, if you have this abnormal connection between an artery and a vein in the lining of your brain, your sensate innervated lining of your brain, you can have flow-related symptoms such as pulsatile tinnitus. These are particularly common with a fairly common variant of these, the transverse sigmoid fischl. But in addition, the great concern occurs when there is impairment of the normal venous drainage. And if that impacts the brain, that can result in hemorrhage, hemorrhagic stroke, or it can also mimic ischemic stroke. So the brain can dysfunction as a result of an inability to drain properly. It's a venous phenomenon akin to a venous sinus thrombosis. If this occurs in the spinal cord, if you have impairment of normal venous drainage in the spinal cord, you have dysfunction of the spinal cord as a result of edema and impaired outflow, which can result in paralysis. And if this impacts the eye or the outflow of the eye, it can result in chemosis and or proptosis, diplopia, or frank vision loss. And so the emphasis with dura or trevenous fischl is venous. It is the impairment of venous drainage that is really the crux of the issue for when these become particularly dangerous or concerning. So this is a diagram from an old Netter atlas that basically lineates obviously the dura, the lining of the brain. This is not a pathologic picture, and it's illustrating the sinus, which is where dura or trevenous fischl can potentially exclusively drain where they stay in the dura and they drain into the sinus. Or alternatively, the fistula can pressurize a cortical vein, which is lineated here with the red arrow. And that is, again, what can lead to these concerning venous symptoms. It's this dichotomy that really is the key defining feature of fistulas, whether they stay in the dura, whether there's a shunt in the dura and it stays in the dura, i.e. drains into a sinus, or instead, whether it decides to drain into a vein that traverses the subneural space and then pressurizes the brain, and that's the concerning type. And that leads me into discussing their natural history, obviously, as I emphasize the importance of venous drainage, which is really what the natural history is predicated upon. There are other features that we'll also discuss. The importance of this venous drainage, this dichotomy that I lineate, was first described in this paper in 1972. A very familiar author here, Al Roten, when he was younger. This is a paper from the Mayo Clinic that clearly lineates this dichotomy of fistulas that simply drain into a sinus, those that stay in the dura, those that are, quote, benign, versus those that cause cortical venous reflux. You can see that here in this bottom right picture here. This was a very seminal paper lineating this important dichotomy. And really, though, in modern times, the natural history discussion of Professor René Gendron is really what we still use today. Unfortunately, we do not credit him with that, as these classification schemes are credited to other people, but it's important to emphasize that Professor René Gendron was a real visionary, luminary in the field of endovascular. He was a French interventionalist who really is the father of our understanding, a lot of meningeal anatomy, dural AV fistulas, spinal anatomy, spinal fistulas, and embolization. And in this classic book that I recommend anybody read for phenomenal anatomic insights and just phenomenal information about fistulas and other pathology of the ECA, he clearly lineates four types of dural AVS. And this is really a classification scheme that we use today, the Xinjiang classification scheme. So type 1 fistulas are those that drain only into a venous sinus. Like I said, those are the ones that stay in the neuro. They're thought to be, quote, benign or not harboring really a risk of hemorrhage. Type 2 fistulas are those that drain into a sinus, but then cause cortical venous reflux. So if you can see, this is the original figure from the paper. What is lineated as A is type 1. What is lineated as B is type 2, where there's drainage into a venous sinus, but there's reflux into normal cortical veins. And this poses a risk of venous hypertension and consequent ischemic-like symptoms or frank hemorrhage in the brain. Type 3 fistulas are those that drain directly into cortical veins. That's D over here. And type 4 are those that drain directly into cortical veins and the vein itself harbors a nectasia. This has actually been proven to be an independent risk factor for hemorrhage. And so this distinction between type 3 and type 4, and really this pattern of classification is really a very robust one even today. So in 1995, the general neurosurgery, Dr. Bill Shukar was a senior author on this paper that we often cite for cranial fistulas, but it's actually a very common misquote or misclassification in neurosurgery. Because if you actually read the paper, and I've just exerted parts of the paper, first of all, the purpose of the paper, a proposed classification for spinal and cranial dural ABS. This was supposed to be a paper to link Xin Zhang's classification scheme of cranial fistulas to spinal fistulas, although it never really caught on because it didn't really make sense for spinal fistulas. There was a lot of emphasis on the epidural fistula, but it doesn't really carry on. And so what they did was they called type 1 fistulas, Xin Zhang type 1 fistula in the brain was the same as an epidural spinal ABS without intradural drainage. So the cortical reflux is the same as that intradural drainage of spinal fistulas. And so they drew that analogy in as much as these epidural fistulas in the spine do not really pose a risk of venous hypertension in the spinal cord. And you can see here the type 2 cranial fistula of Xin Zhang was drawn analogy to epidural spinal ABS that have intradural drainage. That's a known, it's actually the more common variant for epidural ABS. And then the type 3 Xin Zhang cranial fistulas were cited to be like a typical spinal dural ABS where there's just direct drainage into quote, it's not cortical vein, but an intradural vein of the spinal cord that results in venous hypertension. But again, this scheme for spinal ABS didn't catch on, but most of the neurosurgical community now refers to the Xin Zhang type 1, 2, and 3 as Borden 1, 2, and 3. And the disadvantage of that is you lose the type 4 notation for that with venous ectasia, that's actually relevant. So we still really use the original Xin Zhang scheme, whether or not we refer to it as his scheme or Borden or what have you. Now his pupil, you can see on the original textbook, his junior author was Jean-Jacques Merlin, who was another phenomenal interventionalist, who then updated Xin Zhang's classification scheme with Christophe Cognard, who was the first author on this paper. And often everyone knows about the Cognard classification, which was again an expansion of Xin Zhang's classification where they stratified the type 2 fistulas into type 2A, which is that that has retrograde flow in the sinus, but no cortical venous reflux. This can be thought of as a potentially intermediary fistula or one that can cause potentially cerebral venous hypertension as a result of high venous pressures and pseudotumor-like symptomatology, but not Frank Hemorrhage. And then the type 2B is the more typical as we think of as type 2. Type 2A plus B is one with retrograde flow and cortical venous reflux. These are the more common pair. Type 3 is direct cortical venous drainage, like Xin Zhang said, as is type 4. And type 5 was a unique classification for those with spinal perimedullary venous drainage, which is a useful notation. So again, this is another modern classification scheme that is useful that's a dovetail of Xin Zhang's classification. So about 10 years ago, when I was a resident with Dr. Hu in Boston, we decided to look at a case series of 80 fistulas and then amalgamate that data with about 320 fistulas from the literature for a total of 400 to try to lineate demographic and frank natural history data for these fistulas. To this point, there was a lot of work done on the natural history of derartuvian fistulas from the University of Toronto. There was a paper out of Wash U and a couple of papers out of Europe that all independently showed the importance of particularly the venous drainage and also presentation. And this was basically an amalgamation of that with our own data. So what we found was that type 1 fistulas, you can see here no significant sex predilection, often occurred in the transverse sigmoid sinus about half the time and otherwise were the carotid indirect cavernous fistulas and they had no risk of hemorrhage if they maintain their type 1 status. Type 2 fistulas affected slightly older patients without a significant sex predilection were still commonly in the transverse sigmoid sinus, but they harbored a 6% annual hemorrhage risk altogether, all taken together. The type 3 fistulas, slight male sex predilection, and that's somewhat akin to spinal derartuvian fistulas. You can see their distribution more so in the tentorium pitreous ridge or obviously in the intracranial fossa notorious location, and they had a 10% annual hemorrhage risk. Then the type 4 fistula, that with venous ectasia, you can see there's a substantial increase in hemorrhage risk to 21%. This is per year. And that was certainly an interesting finding that lineated not only importance of venous drainage, but of venous ectasia. There was a paper before this, besides Jin Zhang's recognition of this importance, there's a paper from Bolters et al. Neurosurgery that recognized this important feature as well. And we saw this in our data, put that together. When we combined all type 2 to 4, those that have cortical venous reflux together, you can see here that the mean age and the sex predilection. And when we looked at the modality of presentation, so if the patient presented, now asymptomatic is a bit of a misnomer. It simply means they didn't have stroke-like symptoms or venous hypertension symptoms. So they could have had ocular symptoms or tinnitus or no symptoms. So those fistulas altogether had a 2% annual risk of bleeding. Now, if the patient had a prior non-hemorrhagic neurologic deficit, so focal neurologic signs due to venous hypertension, the risk of hemorrhage was 10%. And if the patient had a prior hemorrhage, that annual risk increased to 46%, which is a substantially more malignant natural history. And so this is a fairly relevant finding that encourages the expedient treatment of ruptured dural archaevenous fistulas. Not quite, but somewhat akin to even a ruptured aneurysm. We then subsequently did a multicenter study. Now, one of the limitations of this prior natural history study, when we amalgamated data from literature, is it incorporated some data of partially treated fistulas. About half of the natural history studies had partially treated fistulas. So here, this was a pure multicenter series of untreated fistulas that we published in the Journal of Neurosurgery, where we also lineated annual rates in gray. These are non-hemorrhagic neurologic deficit rates. So event rates that aren't hemorrhagic and then hemorrhagic rates. You can see for type 1, that rate is 0%. For type 2, it's about 3% to 4%. Type 3, maybe 4% to 6%. Type 4, the hemorrhage rate increased up to 9%. And then when we did that designation that we did in that other paper, where we looked at asymptomatic, albeit these are not truly asymptomatic, you can have tinnitus or ocular symptoms, the annual rate of non-hemorrhagic neurologic deficit was 2.3%. The annual risk of hemorrhage was 2.9%. Interestingly, when we compared patients that presented with non-hemorrhagic neurologic deficit, their recurrent rate of non-hemorrhagic neurologic deficit was 22%, although their hemorrhage rate was only 3%. So non-hemorrhagic neurologic deficit, so if you have a focal neurological symptom, a hemiparesis, a seizure, perhaps, you are more likely to have recurrent non-hemorrhagic neurologic deficit events even more so than hemorrhage, whereas patients who presented with hemorrhage were simply more likely to simply re-hemorrhage at a rate upwards of 46% per year, which is certainly a high rate. I do think that this is probably potentially a slight overestimate due to the short follow-up. So these natural history studies are done pseudo-prospectively, where you follow the patient from diagnosis until treatment, and that can be fairly short, a matter of weeks to months, and in general, in cerebrovascular, the natural history is always the worst early on. So if you do 100 patient years following 100 patients for one year, that's different than following one patient for 100 years, even though we think of it as the same thing. In general, we think that the natural history often gets better with time, and so this may be a bit of an overestimate just when we talk about natural history studies in general. So I hope that that lineates some of the important features of the natural history of derivative venous fistulas. Again, to summarize, the most important feature is whether the fistula has cortical venous drainage, drainage into a cortical vein. That's type 2, type 3, type 4. A second important feature is whether there's venous ectasia, and I should clarify that is greater than 5 millimeter dilation of the vein. That's an independent risk factor for hemorrhage as well. A third and important risk factor for hemorrhage is the modality of presentation. So if a patient presents with a non-hemorrhagic neurologic deficit or hemorrhage, they are more likely to have that in the future at a substantially increased rate. A gluttonous and somewhat enjoyable personal pleasure of mine is to review the anatomy of these. I think the anatomy of derivative venous fistulas is rather fascinating and interesting, and I think it's helpful to understand what are the potential arterial conduits into these so that when you're evaluating them, you understand their complete anatomy and particularly after you think you've effectively treated them, you know what potential alternative arterial conduits can recanalize them and be a source of recurrence, which is relevant particularly for endovascular treatment. So when it comes to the anatomy of derivative venous fistulas, location, location, location, there are certain common locations that fistulas occur, and these locations have, number one, they have certain tendencies with their venous drainage, and number two, the location dictates what meningeal arteries can typically feed the fistula. So one location is in the front along the anterior fossa or sometimes called ethmoidal fistulas, periviform plate fistulas. Because of really the absence of a frank sinus in this location, these almost ubiquitously have cortical venous drainage, and so they often present with hemorrhage. They're often very silent and otherwise are asymptomatic unless they present with hemorrhage, not even non-hemorrhagic neurologic deficit, really hemorrhage, which is interesting. We'll talk about their arterial anatomy. Indirect carotid cavernous fistulas, which are very different from direct fistulas that are due to aneurysms or traumatic avulsions, these are a dural AVF of the cavernous sinus is a shunt within the leaflets of the cavernous sinus that's completely distinct and has nothing to do with a direct fistula. That's a dural AVF of the cavernous sinus. They most commonly present with frank ocular symptoms due to reflux or pressurization of the superior ophthalmic vein, sometimes the interior ophthalmic vein as well. About a third of them will harbor cortical venous drainage, and I have had a couple that have presented with hemorrhage. But again, typically their presentation is related to their ocular venous hypertension. Now fistulas that occur further back, as we work our way back from the anterior fossa to the cavernous sinus, now back to the tentorium or the petrous ridge or the deep venous system. So these fistulas are notoriously those that harbor cortical venous drainage, and therefore they have a malignant natural history largely comprised of presentation with non-hemorrhagic neurologic deficit or hemorrhage. Now the most common type of fistula, those that occur in the transverse and or sigmoid sinus, these tend to present with potentially more benign symptoms like tinnitus, but can harbor cortical venous drainage about half of the time, and therefore they can also present with hemorrhage or non-hemorrhagic neurologic deficits. I do think I have a couple interesting examples of that. A unique and multi-named fistula location, I'm using the term marginal sinus, that's the venous system around the foramen magnum. These are also referred to, these are totally the same as those that are referred to as hypoglossal canal or condylar fistulas. These are again those around the foramen magnum. They're distinct from those in the tentorium or those of the cavernous sinus, although they can mimic cavernous fistulas if they reflux into the inferior petrosal sinus. So these are also generally a slightly more benign variant. They can present with tinnitus or again ocular symptoms and mimic indirect carotid cavernous fistulas, about a quarter of them have cortical venous drainage. Then another fairly common type of fistula, one that occurs on the suprasaggial sinus, sometimes referred to as convexity fistulas. These actually tend to drain directly into cortical veins, even though they're associated with that sinus system, unlike transverse sigmoid fistulas, and so they can have a more malignant natural history. So in the bottom pane here, you can see there's an ethmoidal fistula at the very far left of the screen here. So these are a little ethmoidal branch of the ophthalmic artery supplying this fistula. You can see, interestingly, as it can, it can hook into an olfactory vein and have deep venous drainage, and so this is certainly a malignant fistula. This is a fairly typical appearing indirect carotid cavernous or dural fistula of the cavernous sinus with fairly profound venous drainage into exclusively the superior ophthalmic vein. So this would be not one that would be denoted as one with cortical venous drainage, but still can present with fairly debilitating symptomatology. And here's a fairly typical appearing transverse sigmoid fistula. There's no cortical venous drainage, so this would be a type 1 fistula, perhaps presenting in a patient as either asymptomatic or presenting with pulsatile tinnitus. So the relevant arterial anatomy, as we said, for these anterior fossa dural AVFs are really ethmoidal branches of the ophthalmic. They can rarely have supply as well from branches of the internal maxillary artery and middle meningeal artery, but they're mainly ethmoidal branches of the ophthalmic. You can see in these rodent pictures here, you can see the ophthalmic artery and small ethmoidal branches piercing the cribriform plate, and these are the relevant branches, the most relevant branches that supply these fistulas. Often as a result of this, your endovascular options, if you don't have a transvenous option, can be limited and have a risk of potentially causing blindness, so these are probably still most commonly treated through microsurgery, but certainly do have endovascular options when well selected. So that's one key piece of the ICA. When we think of dural AVFs, there's really not a lot of key pieces of the internal carotid artery. There's really just the ophthalmic ethmoidal branches and cavernous branches that I'll highlight. So everyone knows about the meningeal hypothesial trunk that gives rise to the tentorial trunk. Mainly we think of the medial tentorial artery, the artery of Bernasconi and Casanare, that is often involved in a variety of fistulas along the tentorium as well as transverse sigmoid fistulas. Another fairly preponderant artery that's involved, particularly in fistulas that can occur along the clivus, like marginal sinus fistulas, is the dorsal meningeal artery. This travels with the sixth nerve in Dorello's canal. The inferior hypothesial artery is obviously not implicated, it's a branch to the hypothesis. And then the second relevant branch for these dural AVFs is the infralateral trunk that supplies the dura of the cavernous sinus. It can also supply, in this case, this was a fistula I treated that was in the tentorium with a fairly enlarged infralateral trunk. This is an alternative vessel that's very important to recognize because if you're using liquid embolizate, it's not a vessel you want to be refluxing into. And so it's very important to recognize the distinction between the MHT, the ILT, and their supply to these dural AVFs, particularly when you're treating cavernous sinus fistulas. This is another interesting illustration of the recurrent meningeal artery, where the recurrent meningeal artery is supplying the middle meningeal artery to supply a convexity dural AVF. That's a fairly rare variant, but it's important to recognize that not all patients have a middle meningeal artery emanating from the internal maxillary artery. Their middle meningeal artery may in fact be off of their ophthalmic artery, and this is a case where this fistula is exclusively supplied by this branch of the ophthalmic artery. And again, this is the summary of really the relevant dural internal carotid artery branches for dural AVFs, the cavernous branch of the MHT and the ILT, and the ophthalmic branches as well. In terms of the most important anatomy of the ECA for dural AVFs, there's really only three key pieces. You can really distill this fairly simply into relevant branches of the internal maxillary artery. We'll talk about the middle meningeal artery and other relevant branches, the occipital artery, and then very importantly, the ascending pharyngeal artery, particularly the neuromeningeal trunk. All right, so the middle meningeal artery is becoming a fairly popular endovascular conduit in the era of these MMA embolizations that are getting more and more preponderant. So just to very importantly review the typical anatomy of the middle meningeal, because all of these pieces can be very much involved in dural trivenous fistulas, right? So you have your middle meningeal artery. This would be a lateral view. This is just a sketch that has a classic 90-degree turn right above the foramen spinosum. This is a fairly constant vascular finding that facilitates identification of the MMA for some of our younger audience. Very often, very early off of the middle meningeal artery coursing posteriorly just above foramen spinosum is the petrosal branch. This is a no-go branch. This is a branch you should not ever get liquid embolizate into, as you will cause seventh nerve palsies if you do that. Petrosal branches are often supplied to petrosal dural ADFs, and so if you're embolizing a petrosal dural ADF and there's petrosal branch supply, you have to be very, very careful to ensure your embolizate does not reflux into a petrosal branch, as you will unexpectedly cause a seventh nerve palsy. Now another important branch to be aware of is the sphenoidal branch early on. Often, the sphenoidal branch only supplies the lacrimal gland and does not actually go on to supply the ophthalmic artery, but this can be an anastomotic vessel to the ophthalmic artery, so you have to be very careful about embolizing that, but otherwise, the remainder of the branches are excellent endovascular conduits, frontal parietal branches are often very safe. If the only time you get in trouble with the petroschmosal branch is an endovascular conduit, and this is involved in a plethora of dural ADFs, is if there's no obvious petrosal branch and you see tiny little branches going down to the petrosal territory, but these MMA branches are implicated in almost any fissile you can think of the brain, what be it in the convexity, the tentorium, even as I said, these ethmoidal fissiles can often steal some sphenoidal branches, so these are some important MMA branches just to review. This is just an interesting variant of a patient who actually did not have a central retinal artery, so you can see here, this is the middle meningeal artery, classic 90 degree anteriorly oriented turn, fairly typical appearing petroschmosal branch with parietal and frontal branches, but here the sphenoidal branch is in fact supplying the coroidal blush, this patient actually is supplying her orbit through a meningo-ophthalmic branch of the sphenoidal, and so this is a very important finding, you do not want to do anything to this middle meningeal artery endovascularly whatsoever. But it's important to recognize when we talk about carotid cavernous fissiles, two other important branches of the internal maxillary artery that are often, if not more involved in the middle meningeal artery in carotid cavernous fissiles, and that's the accessory meningeal artery and the artery foramen rotunda, and knowing this scenario is important not only for dural ABS, but for JNAs and anything you're embolizing in this region. So the accessory meningeal artery can emanate directly from the internal maxillary artery immediately adjacent to the middle meningeal artery, or it will come off the middle meningeal artery while it's still extracranial. The accessory meningeal artery travels through foramen ovale, and then supplies the fifth nerve and can supply the cavernous sinus, so it can often be involved in these indirect carotid cavernous fistulas. The artery of foramen rotundum has a slightly more distal takeoff, so it travels through foramen rotundum. So these are your two foraminal arteries, your artery of foramen rotundum is your foramen rotundum artery, your accessory meningeal artery is your foramen ovale artery, but both of these arteries basically can anastomose at the ILT, which is why they're involved in carotid cavernous fistulas, and why, in general, embolizing through them, you have to be very careful and understand these very important anastomoses that are fairly common. So here's a fairly typical carotid indirect or dural carotid cavernous fistula. So you can see here, so this is the middle meningeal artery with its classic 90-degree turn, there's actually a nice petrosal branch here, but much bigger than that, here's AMA accessory meningeal artery is supplying this carotid cavernous fistula, along with this artery of foramen rotundum. Here's just a zoomed in view, here's, you could see here, this, the M is for middle meningeal artery, and here are the arrows on the accessory meningeal artery, that's the main conduit to this carotid cavernous fistula, in fact, sometimes it can get so big that you can actually go through a transarterial and get into the cavernous sinus to serendipitously treat these fistulas. And here, this is a different carotid cavernous fistula, you could see the A for accessory meningeal artery, so here's middle meningeal, which is smaller, here's the accessory meningeal artery going into the fistula, along with supply from the artery of foramen rotundum. So again, this anatomy and these anastomosis with the ILT are very important bullets to understand when you're deconstructing, particularly carotid cavernous fistulas or other pathology in this region. The occipital artery is another very common conduit to a plethora of durotribute fistulas, in particular these transverse sigmoid fistulas. It used to be a fairly vexing conduit where you would think that you can get really good penetration and cure the fistula through the occipital artery, but the transosseous branches would always get you and not get very good penetration. That's changing now in the era of balloon microcatheters, which I'll talk about. And then finally, the ascending pharyngeal artery, which is involved in a variety of durotribute fistulas, particularly those around the marginal sinus. Again, these marginal sinus fistulas are referred to in the literature as hypoglossal canal, condylar, they're all synonyms. And so the ascending pharyngeal artery bifurcates into a pharyngeal and neuromeningeal trunk. The pharyngeal trunk supplies the pharynx, although the superior pharyngeal artery is a source of anastomosis to the carotid, and particularly in patients who have ischemic issues. But really of relevance in dural AVFs is the neuromeningeal trunk that nourishes the lower cranial nerves, 9 through 11, through the jugular and hypoglossal division, but also has a clival branch that can supply carotid cavernous fistulas, and additionally can often be implicated in posterior phosphorylation, or those in the tentorium, as the neuromeningeal trunk can continue on as a posterior meningeal artery. So here's a carotid cavernous fistula supplied by, you can see here, so here's the classic turn of the neuromeningeal trunk. Here's the clival branch going up into the carotid cavernous fistula right here. So that's a carotid cavernous fistula supplied by the clival branch of the ascending pharyngeal artery. These are some marginal sinus fistulas. Again, they have a proclivity to be benign, so they don't tend to have cortical venous drainage. Sometimes they drain into the IPS, though, and can cause ocular symptoms. You can see here, in fact, the nuclear supply to these marginal sinus, condylar, hypoglossal fistulas is often the neuromeningeal trunk of the ascending pharyngeal artery. So it's important to understand that anatomy. And here's basically a summary slide of this anatomy that we've sort of whirled in through over the past 10 minutes that lineates the relevant parent artery and the relevant meningeal branch and the typical fistula locations that these particular meningeal arteries can supply. This is exerted from our textbook. I didn't really talk about some of the interesting posterior circulation branches, but often people talk about the artery of Davidoff and Schechter, which is a PCA neural branch that can often be involved in tentorial fistulas. So you can see here, this is a vertebral injection. This is a fairly classic artery of Davidoff and Schechter that's going into the tentorium to supply this tentorial fistula. There's a sibling artery called the artery of Wollschleger and Wollschleger. Wollschleger and Wollschleger refers to the SCA neural branch that is often misnamed the artery of Davidoff and Schechter. Wollschleger and Wollschleger were actually the people that also named the artery of Davidoff and Schechter, who were their mentors actually, but they were the ones that sort of did the microanatomic dissections to lineate that SCA neural branch. But we'll leave it at, this is just a useful summary slide of the relevant meningeal arterial anatomy to dural artery venous fistulas and the relevant locations. So let's move on to relevant, the diagnosis and relevant imaging features of dural AVS. I'll use this illustrative case to reinforce that. So this was a patient that I encountered in residency who was billed as having a prior and now recurrent right temporal stroke with hemorrhagic conversion. The patient had a history of atrial fibrillation, was on Coumadin. So common things being common, that's what they thought this was. Again with this recurrent event, the patient presented with new confusion and hemiparesis and was admitted for a new stroke. This is the MRI. This is an enhanced MRI. Some people thought this represented a glioblastoma. But this has the fairly typical appearance of a dural artery venous fistula with cortical venous reflux. These enhancing serpentine veins here. Now this is not entirely specific for dural AVFs. This is exactly what you can see in venous sinus thrombosis, which is very similar to dural AVFs because venous sinus thrombosis cause venous hypertension, just like dural AVFs do. And so it's a very analogous process, but these corkscrew enhancing veins are fairly concerning for dural AVFs. That's what you look for an MRI. The first thing I go to on an MRI, if I'm looking for a dural AVF is the T1 post to look for these little corkscrew veins. And so this patient subsequently had an angiogram that demonstrated this type 3 dural AVF that was embolized successfully transarterially. So on a CTA angiogram, in contrast to an arterial venous malformation, where you have often enlarged peel arteries or a frank nidal area, instead on a CTA, all you have are these wandering veins. So here in the CTA, you can see here, this is a dural AVF with cortical reflux. And all you see on the CTA are the refluxing veins. You often don't see the arteries unless they're enormous transosseous arteries, but you see these veins and there's no nidus, there's no enlarged peel arteries. And that's how you can distinguish on CTA AVMs from dural AVFs in a simplistic fashion. So here's a type 4 dural AVF. You can see here a convexity type 4 dural AVF, and all that you're seeing here are the refluxing cortical veins. You're not seeing arteries, and that's the telltale sign of these on CTA. And again, here's the original MRI show. This was another patient that I treated that has all these, like, wispy enhancing veins that you see in dural AVFs with cortical venous reflux. And here's just a summary slide. So in terms of clinical features and neuroimaging findings in patients with specific types of dural AVFs. So with the cavernous dural AVFs, patients can present with chemosis, proptosis, or ophthalmoparesis. Clinically, on neuroimaging, you will often have a dilated SOV. So that's lineated down here. You can see there's a fairly large SOV with the green arrow, and or preponderant vessels in the region of cavernous sinus. Now, I'll tell you, patients that have trauma or patients, depending on the phasing of the CTA, CTV, you can often psych yourself out into thinking a lot of people have indirect carotid cavernous fischlias. One sinus lights up more than the other. So the CTV findings can be a little bit sensitive, but not specific. But nevertheless, these are the findings that you could see for cavernous fischlias. For those along the suprasaggial sinus or convexity, so again, these have a high proclivity to really just have cortical drainage. So they only present with focal deficits or non-hemorrhagic neurologic deficits or frank hemorrhage. In this case, you'll see corkscrew veins at the convexity, somewhat analogous to that picture I showed earlier where there were those temporal veins. In MRI, patients can have unexplained flare signal at the convexity due to venous hypertension. Transverse sigmoid fischlias often present with pulsatile tinnitus, or alternatively, they can present with malignant symptomatology due to venous hypertension, focal deficits, or non-hemorrhagic neurologic deficits. On these fischlias, you'll see preponderant vessels or enlarged occipital branches in the occipital or suboccipital region. On a non-contrast head CT, you'll see these large transosseous channels for the occipital branches, and you'll see these sometimes in the temporal lobe, these corkscrew veins. Marginal sinus fischlias can present with pulsatile tinnitus. Sometimes they can have cavernous symptoms if there's reflux in the inferior petrosal sinus or if they do have cortical venous drainage, they will present potentially with cerebellar hemorrhage. A little piece of advice for these, sometimes you can miss these if you do an external injection and the ascending pharyngeal comes out approximately, so oftentimes it's good to do a common injection to ensure you're capturing the ascending pharyngeal or select the ascending pharyngeal to ensure you're evaluating it if there's a high suspicion for these. We already talked about anterior fossa and tentorial petrosal fischlias, so we'll conclude putting this together with treatment. George, is this supposed to be a half hour or an hour long, just so I don't go... It's supposed to be half an hour, but we really don't have like a... No problem, we'll be done in 30 minutes. So in terms of endovascular treatment, so endovascular is a means to both diagnose and treat the lesion at the same time. This is probably how most sterile AVFs are treated in most institutions, I'd say about 80% of the time. It's important to do a full six-vessel angiogram, cover all the vessels, and do not forget the ascending pharyngeal. This was a case of a patient who presented with a posterior fossa hemorrhage. You can see the MRI has these corkscrew veins, and we had a suspicion that they had actually a marginal sinus high-risk fistula. And very subtly, you could see here, supplied by the ascending pharyngeal, we embolized it transarterially for treatment. And it's important to emphasize that endovascularly, the introduction of onyx into the treatment armamentarium is extremely important. The ability to inject a liquid embolizate over a prolonged period in contrast to glue really revolutionized the way that we can treat fistulas endovascularly, right? So when I was in Phoenix, my mentors did a very high-volume dural AVFs, you know, between them, UCSF, and places that do a lot of these. There's a lot to learn from. And when we compare fistulas that were embolized prior to the introduction of onyx to after the introduction of onyx, you can see that the obliteration rate went up from 60 to 76%. The ability to treat these things from a single arterial pedicle went up substantially, need for subsequent surgery went down. It really, onyx was a big, a big advance. And so here's just a typical transarterial embolization where we're going through the middle meningeal artery, casting the vein with onyx, you must get onyx into the vein, you must fill the vein thoroughly with onyx. A very important finding is you should see the onyx not only hit the vein, but reflux into other suppliers to the fistula. So that, and that's why it's important to know what those suppliers are. If you're not refluxing into other arteries that supply the fistula, it's going to come back. So it's not only casting the vein, but it's refluxing into other arterial supply to the fistula. If you're going to try to treat these transarterially and successfully embolize them for a long-term treatment result. Here's a case of a ruptured trans or sigmoid fistula. That's really primarily supplied by these tiny little occipital branches. We used a modified pressure cooker technique transarterially to treat this. So the problem is these occipital branches before the introduction of balloon microcatheters really were not, did not allow for adequate penetration of the vein. And even in this case, we have all these tiny little side branches. If I use a balloon microcatheter, I'm just going to glue the heck out of the scalp. So what we did was we put a bunch of coils into the main exiting scalp branch to basically pressure cook the fistula. So we inflated a balloon, injected Onyx, and you can see the Onyx penetrated the vein through this modified pressure cooker technique using coils that when we talk about pressure cooker technique endovascularly that often refers to gluing in the catheter with glue and then using Onyx. This is an alternative approach. You can see there was a long-term cure. Transvenous embolization I think is really the way to go if you're treating carotid cavernous fistulas. I think trans-arterially treating them is really dangerous in terms of potential and that reflux into ILT and into the carotid artery. And so these are fairly safely done through either the IPS, sometimes through the SOV. And again, coils, sometimes with a little Onyx on top of it transvenously, people have different levels of interest doing that, is really an effective way to treat these. This was, as I alluded to, some of these carotid cavernous fistulas could present with hemorrhage. So this was a carotid cavernous fistula that drained into the deep middle cerebral vein that presented with this hemorrhage. We transvenously coiled it, fairly technically straightforward. Here's another one. This patient actually presented with subarachnoid hemorrhage with this type four fistula supplied. You can see that the substantial supply here by this enlarged artery frame and rotundum, this was transvenously coiled and things went well. But again, endovascular is not a panacea. Not everything is amenable to endovascular therapy, but I do think treating it, trying to do things endovascularly first, about 80% of the time works. That accepts anterior fossa fistulas that I still think are often best treated surgically. Surgical interruption here is really a very effective treatment, especially for type three and four fistulas, where you cut the vein and often try to cut the dura with it. It's incredibly facile, much more simple than AVMs. This is just a pooled analysis we did of surgery for ethmoidal fistulas, very high obliteration rate, very low complication rates. This is a case that Dr. Spetzler did through an eyebrow approach that did very well. Finally, radiosurgery is another final alternative in the treatment of fistulas that can be fairly effective, particularly in fistulas without cortical venous drainage. Everyone has their meta-analysis of this. We did one about 10 years ago where you could see the obliteration rate for radiosurgery in general is somewhat comparable to AVMs, but I think you need to select the right ones for radiosurgery. We certainly have a nice series of these out of our institution here in Pittsburgh that have done fairly well. So this is a wonderful alternative and well-selected fistulas. The most important factors that influence SRS results is, first of all, location. So carotid cavernous fistulas in patients with tolerable intraocular pressures are really good candidates for this therapy. It's a fairly focal area that you can target. Fistulas that do not have cortical venous drainage, which makes sense clinically as well, tended to do better. And like anything for radiosurgery, smaller ones tended to do better as well. We had a couple of papers that looked at SRS for dural AVS without cortical reflux, and those that had cortical reflux but did not present with malignant symptoms, again, this is Dr. Lunsford and his partners' phenomenal experience with these, and this is an important alternative for dural AVS. So when we think about management of dural AVS, really embolization is often first line in the hands of whether interventional neuroradiologists, neurosurgeons, radiologists, neurologists. They're really the most experienced and see these the most. However, if a patient has a high-risk fistula and you're worried about eloquent branches, atrosal branches, sphenoidal branches that can be eloquent with the eye, or there's really not a good conduit to the fistula, surgery is an excellent option. I've had a lot of success with it, Dr. Zinonos, Dr. Gardner, my partners have certainly disconnected a lot of these very successfully with good long-term results. And then stereotactic radiosurgery is another excellent option, particularly if there's no cortical venous drainage, the patient's really not a great surgical candidate. My partner here, Dr. Lunsford, has an expansive experience with this, and as the angiographer, I see the success from it. So here's our final slide. So in summary, so durotrevenous fistulas are abnormal connections in the dura between arteries and veins. They're unique to the neuroraxis. They occur in the dura. And essentially, they can stay in the dura, so drain into a vein in the dura, and that's type one. That's a benign fistula that can present with tinnitus or ocular symptoms, but that's it. Or alternatively, if they have cortical venous drainage, if the drain exits the dura and pressurizes the peel space, okay, you can have impairment of venous outflow, and that's where you can have hemorrhage or consequent symptoms. We talked about how their natural history is predicated primarily on their venous drainage, but in addition, on their modality of presentation and the presence of venous hectasia and the importance of the Xinjiang classification to alineate that. That's really the classification we still use today, although we refer to it as Borden or Kocknerd. Their anatomy and diagnosis are reviewed in detail. And then the treatment. So endovascular is often first line, if it's feasible and safe. Otherwise, surgical disconnection, and particularly in high-grade fistulas, can be very safe and effective and fairly fast, particularly compared to AVMs. And then finally, radiosurgery is an excellent alternative for low-grade and particularly carotid cavernous fistulas. So I thank you for your attention, and we can move on. Brad, such a fantastic overview, going from the very basics to, you know, many, many pearls. A plug-in for Dr. Gross's book. You've seen some of the many nice tables. I encourage everyone to take a look at it. Everyone is interested in this disease. Please, for the live participants, if you have any questions, just put them in the chat and we'll get to them as soon as we're done with the presentations. We're very lucky that Dr. Zipfel made it out of the OR, done saving lives. Again, Dr. Zipfel needs no introduction. He's one of the most renowned cerebrovascular neurosurgeons in the world, but also the fearless leader of the online educational committee. And again, he is responsible for most of the content that is being produced, including this series. It's been enjoyed. Dr. Zipfel, again, is gonna speak to us today about the data behind the treatment of the dural AVFs. Dr. Zipfel, please take it away. Thank you, Georges. I appreciate it. I'm sorry I'm a little late. And Bradley, I caught a fair amount of what you had to say. So a very nice talk. And I enjoy being on the same panel with you with all your work that you've done in fistulas. So it's great to be here. So I'll just share my screen, hopefully. Does that look like full presenter mode? Does that look okay? Perfect. It looks perfect. So I'm gonna focus, Bradley gave a great overview and a lot of details about management and some treatment recommendations. I'm just gonna concentrate solely on a couple of papers, one that's published and one that will be published soon coming from a consortium that we started that Bradley's a part of and others from Condor. It's called Condor. So this international consortium is called the International Consortium for Dural-Ave fistula Outcomes Research. We started several years ago. It took a while to kind of get everything in terms of the patient entered and all the annotation and the analysis. And finally this year, the papers are coming out and there's been a couple that have been published. And then there's a series of five that are gonna be published in the journal of surgery. They've been accepted for about six months, but I guess the JNS is behind with their publication. This is the group. And this includes Wash U as the central site and many others as you see lifted there, most from North America, but we do have a few sites from around the world. And most of these individuals and centers had already published in the area of Dural-Ave fistulas and they were invited to be part of this group and excited to present some data from it. This is just a paper that will be published here soon in JNS that describes our cohort. It's now 1,077 patients from 16 centers. You see the years that we found these patients and had data entered. It's all retrospective. We do have interest in making this repository a prospective moving forward, something that we're working on. All these fistulas were andrographically confirmed and then there's about 153 clinical and radiographic variables that were entered for every single patient. I'm sure Bradley went through some of this. I probably don't need to harp on this too much, but I will highlight the importance of corticopoietin drainage. This is an anterior fossa flora fistula fed by an enlarged ophthalmic artery and then the anterior and posterior ethmoids are going to break into a cortical vein that ultimately winds up in the superior sagittal sinus. But in this particular vein is dilated. So it's a cortical venous drainage and it's also venous ectasia, which I'll bring up. These are very important in terms of natural history and understanding when you should treat. I'm sure we went through this, the Borden-Schuchart type one, which does not have cortical venous drainage, the type two that has cortical venous drainage and sinus drainage and the type three that go directly into a vein without involving the sinus. The type twos and threes are important in terms of a worse natural history. And I'll talk about that. Another concept just to get started that's important is this idea of motive presentation. So there's now several papers from our group, from Groves when Bradley, when he was at Brigham with Rose Du and also another nice paper from a couple of nice papers from Europe that show that motive presentation matters in terms of predicting natural history for dural aviary fistulas. So what does that mean? Well, there are symptoms that are purely referable to the high, the increased pressure and pulsatile flow into the sinus itself. And those symptoms are tinnitus for the most common transverse sigmoid. You can have tinnitus with other dural aviary fistulas as well. And if the fistula involves the orbital, the cavernous sinus, then you can have orbital phenomena. But these really are due to increased sinus drainage, not due to the cortical venous hypertension that results from the cortical venous drainage. Then there are symptoms referable to the cortical venous hypertension because of the cortical venous drainage. And these are the more aggressive symptoms that includes cerebral hemorrhage. It can also involve a group of symptoms that are often referred to as non-hemorrhagic neurological deficits, what we call seizures, focal cortical deficits, dementia, symptoms of increased intracranial pressure. All those things are due to either focal or global cortical venous hypertension, but they're a harbinger of a poor acting dural aviary fistula and we'll talk about that. And then finally, increasingly, there are fistulas that present without symptoms at all. The MRI is done or MRA or CTA is done for other reasons and these are identified. I don't know if Bradley went through this, but I'll just really quickly summarize that the bottom line is if there's no cortical venous drainage, a lot of studies coming out of Toronto here, Dr. Gross's work and others shows that these are really very benign lesions that rarely cause neurological events, never cause a significant mortality or mortality. And even the up conversion from a type one to a more aggressive fistula, a type two or three, that's even pretty uncommon. And last I looked, all the reports of an up conversion, meaning a type one fistula changing to a type two or three where they had cortical venous drainage, all of those, at least those in the literature have occurred with a change in symptoms. Either the patient develops new symptoms or a marked change in their symptoms, or they'll have a spontaneous loss of their symptoms like tinnitus. And when the angiograms are obtained, then this up conversion occurs. So the natural history of type one fistula is very benign and that needs to be taken into account in terms of treatment recommendations. The type twos and threes, and I'm not gonna parse into it, there is some data that suggests that there may be a difference between type two versus type three in terms of the natural history, but generally they're grouped together in terms of the natural history studies. And what you'll see here is that those that present with aggressive symptoms, hemorrhage, seizure, focal deficits, things of that nature, their annual event rate, neurological morbidity rate is high and the mortality per year is high. So these are bad actors that need to be treated and I think need to be treated pretty aggressively and early. But those that present with benign symptoms, they only have tinnitus or they have no symptoms at all. The morbidity is significant, but not near what an aggressive duraleve fistula presentation is like. And the mortality is zero. So I think you can tailor your treatment recommendations based on this type of natural history data. And the other thing is increasingly venous ectasia, which I mentioned before, has also been associated with increased or worsening natural history as it relates to hemorrhagic or non-hemorrhagic neurological deficits. And so we need to use these findings in consultation with the duraleve fistula, the angiographic type in terms of stratifying risk and making treatment recommendations. Here are the two papers I just want to focus in on that I think are challenging a little bit of our approach, or at least this one will challenge some of our approach to treatment. The next one will kind of affirm and confirm an approach that a lot of us have been taking with those fistulas. So there's two papers I want to talk about. This one's been published. This is the work of Jason Sheehan out of Virginia as part of the Condor Group. Dr. Chen was the first author on this. And what we're looking at here is we're looking at low-grade fistulas, so those that do not have cortical venous drainage, and looking at outcomes. And we took our retrospective 1,000 patients. We figured out which ones had low-grade fistulas. And then we just looked at those that were observed as a treatment approach versus those that were treated with endovascular treatment, surgery, or radiosurgery, or a combination. And we looked at the outcomes and see if there's any difference in a variety of ways. And so the number of patients we're talking about here is over 300. About two-thirds of them were treated, and about a third of them were observed. All were type 1 fistulas, and this analysis did include CC fistulas, but I'll make a special comment about CC fistulas at the end. The primary outcome measure that was utilized was modified Rankin scale at last follow-up, but there are a number of important secondary outcome measures, including did the symptoms improve? Again, these patients, if they're gonna get treated because they have such a benign natural history, they're getting treated because of the symptoms that are intolerable or are real bothersome to the patients. That's the main reason that type 1 fistulas are being treated. And so looking at how the symptoms do with observation versus treatment is important. We also looked at mortality and angiographic cure. We looked at the unmatched cohorts, but also, I'm sorry, looked at unmatched and propensity, yeah, we looked at the unmatched cohort, and then we did propensity-matched cohorts to try to get really apple-to-apple comparisons. And then we did logistic regression analyses to determine if outcome was changed by treatment. And here are the findings. This is the unmatched cohort initially. And so what you'll see is that of all the primary endpoint, modified Rankin scale, in all the secondary outcomes, including symptom improvement, there was no difference between the observed arm and the treated arm. The only thing that was different is the rate of angiographic cure, which was significantly higher in those that got treated versus those that did not. But otherwise, there was no difference. If you look at the propensity-matched cohort, we're really trying to make sure we're comparing apples-to-apples, same binding, no real difference in the primary endpoint, no real difference in the secondary endpoints, including whether the symptoms improved, there was no difference in that. The angiographic cure rate was higher, but the other outcome measures, there was no difference. By subgroup analysis, if you just only looked at those patients that presented with symptoms or getting treated because of the symptoms that are intolerable, again, no difference in anything other than the angiographic cure. And in that group, the procedural complication rate was 19%, but a lot of those things are temporary or technical and things that really don't impact the patient. But there was a 2% permanent neurological morbidity rate. So that's not, I mean, that's not common, but that's real and it's permanent. And that type of neurological morbidity needs to be taken into account for what is a very, known to be a very benign duralated fistula, right? Our group, or at least, you know, I think a lot of our group, you know, in terms of the conclusions of this is that treatment of duralated fistulas is not associated with increased functional disability compared to conservative management. Although the angiographic obliteration rates were higher in the treatment group, the rate of symptomatic improvement were no different between the two. And there was significant morbidity, relatively low, but there was a permanent morbidity rate of, you know, two to 3%. And this really challenges the idea that we should be treating low grade duralated fistulas with a lot of frequency. Again, two thirds of our fistulas were treated, but this analysis suggests that maybe we should be more cautious and have a higher bar for treatment than what we, you know, based on this kind of analysis. That challenges, I think, a little bit the conventional wisdom. Now, there are caveats to that. This is retrospective. You know, these are, this is not a, you know, no way were patients, you know, picked and then randomized to different treatment arms. You know, so there are some, you know, significant caveats to this, but this does challenge, I think, our general approach to these to some degree. Let me move on. This is also from Dr. Sheehan from the Virginia group. This is not published yet. It's waiting for Jane to have to publish all these papers. I'm sure it's coming. And this time we're going to look at a different group. These are the high grade fistulas, but they're unruptured. These are the ones that have not caused hemorrhage. They have not caused subarachnoid hemorrhage. They're presenting in a more benign fashion and looking to see, again, a similar type of analysis. How does intervention hold up versus conservative management? So this group is over 400 patients, and you can see there is a very strong bias of the participants in our group to treat these type of fistula. And that's based on the natural history that I showed you with even when they present in a benign unruptured fashion, there's still a one and a half percent neurologic event rate per year. So that obviously weighed in on our treatment recommendations, but there were about 30 patients that were observed. You can see the treatment approach was primarily embolization with a few with surgery and radiosurgery. We did the same thing where there was a, the primary outcome was modified Rankin scale. Secondary outcome included symptom improvement and also good functional outcome. Now, these patients, again, are unruptured. They have not had a significant neurologic event yet. And so seeing how they do modified Rankin scale and functional, but also seeing if any symptoms improve is important. Then we did the unmatched and propensity matched cohorts, and we did logistic regression analyses to determine a potential impact of treatment versus observation. And what you see here, different than the first paper, in that there are a number of endpoints that were better with treatment. And that included angiographic cure, which was significantly higher in those that were treated versus those that were not. Their mortality was lower in the embolization group. And again, embolization was by far the most common treatment arm. So embolization lowered the mortality of unruptured high-grade fistula. The hemorrhage rate were statistically lower for embolization versus observation. And there was a trend that surgery and radiosurgery also reduced hemorrhage rates. Again, the numbers in these two groups are lower, and we lost power, and that may have accounted for the lack of statistical significance, but certainly the trend was there. Interestingly, non-hemorrhagic neurological deficits, there was no difference between those. But the mortality and the hemorrhage rates were definitely better in the treated group, and the cure rates were obviously higher. So the conclusion here, I think, is that embolization and surgery for unruptured fistula affords a greater likelihood that you will obliterate or cure angiographically the fistula. Most importantly, is embolization reduced the hemorrhage and death rate compared to conservative management? And surgery and radiosurgery tended to improve risk of hemorrhage. And overall, we thought that these conclusions from this large study really supports a conventional wisdom that these fistulas should be treated, often with embolization, and there is a functional benefit for doing so. So with those two data points, let me just review my treatment recommendations. Bradley already, I think, in some way, gave his, and I'm sure these will match up pretty well, but let me just review it for the group. Again, type 1 fistulas really should be held out for only patients who really have intractable symptoms that are life-altering. And I think, in my mind, since the data has come out, in my own practice here at Wash U, our bar for treatment of these has gone up because there is a real permanent more neurologic morbidity rate with these, although it's low, but we aren't, it's not clear to us how well we are doing with symptom resolution or not. Again, like I mentioned previously, if you are following these patients and there's a change up or down with their symptoms, it really does mandate bathroom imaging to figure out if the type 1 fistula has changed to a type 2 or 3 fistula with corcal venous drainage. I didn't focus on this. We actually have not used CONDOR in a meaningful way to look at CC fistulas, so I don't want to make any change in conventional wisdom about how to treat these. I think there is a lot of data that suggests that these, although the symptoms are, quote, benign, they're not hemorrhages, they're not focal cortical deficits, they can be very bothersome to the patient, painful, loss of function of the eye, and I think those type of symptoms do benefit from intervention. So I think if we do an analysis of CC fistulas and the CONDOR group specifically, I suspect that we will see some data that would support that conventional wisdom. The type 2, 3 fistulas. Now, these, I think, get parsed in my mind in terms of how to treat these and how quickly. If you have a type 2 or 3 fistula with corcal venous drainage and the patient presents with seizure, focal deficits, hemorrhage, things of that nature, that's an aggressive symptomatology. The data is overwhelming that they are going to do bad and they're going to have recurrent events if not treated. They should be treated almost always, and I think they should be treated early, and to me, that means you don't see someone in clinic, see them with either get an angiogram or once the angiogram's obtained and then you treat and then you schedule them from surgery or into vascular treatment in two or three or four weeks. In my mind, once the fistula is identified as a type 2 or 3 and the patient has aggressive symptoms, they should be treated really within 24, 48, 72 hours of presentation to avoid recurrent events while you're waiting. Now, if you have a type 2, 3 fistula that has venous acasia like this example I showed before, similarly, I think the data suggests that these are bad actors. They need to be treated almost in all cases and they probably should also be treated early rather than in a very elective fashion. This is the group that I think you really need to focus in on, in my opinion. These are the patients that have type 2 or 3 fistulas. They're not presenting with aggressive symptoms. They're only presenting with tinnitus or no symptoms at all, and they don't have venous acasia. That's the group where I think there's a lot of patients but not all should be treated, but they can be treated electively. You have time, you can have discussions, you can make sure they understand what the risks and benefits are and work around your schedule and their schedule. And I do think there are a select group of patients who probably shouldn't be treated at all. I think if you have a 75-year-old who has a one to 1.5% event rate, there's a case to be made in that patient population for no treatment. There's also a case to be made for radiosurgery, I think, in some of these patients, especially if the endovascular or surgical treatment approach you have is gonna be particularly complicated because of the angiographic characteristics of that particular fistula. So here, I think elective treatment in a lot of patients, but not all, and I think a more selective approach to patients who are particularly old or medically infirm or that have complex fistula architecture that would increase the risk of your procedure. And with that, I'll just thanks Georgios and Bradley for the invitation and look forward to a further conversation on this webinar. So thank you very much. Dr. Zipfel, thank you so, so much for this fantastic overview. I like the educational vibe of your presentation where you just really crystallize things at the end. I think that's exactly what everyone is looking for. We're gonna hold any questions for now and then go on to go over some cases and see what our experts think. I have a couple of cases to review. Can you see my screen? Yes. Okay. All right, so a couple of cases that are not following exactly the straight path and we'll see what you think. Dr. Gross knows everything about me. So it's a little bit hard to hide anything from him. He knows everything. So he knows these cases, but we'll see what Dr. Zipfel thinks and we'll ask Dr. Gross to elaborate a little bit as well. So this case is a case that was a little bit, I don't wanna say confusing, but not straightforward. This is a 34 year old man, young man that presented with a headache and left side hemiparesis that was found to have this small hemorrhage in the posterior limb of the internal capsule. He did have a past narrow history of a pulmonary AVM as a child. And at that point they had done extensive genetic workup, the workup that was available again 30 years ago or so. and at the time they didn't find anything. The MRI again replicated that he had this small bleed in that area. But there's no evidence of a cavernous malformation in the classical corn appearance or like anything or any indirect signs such as a draining vein, et cetera. So he went on to have an angiogram. And it was clear on the angiogram on the external runs there was actually a type III fistula that was fed by some of the sphenoidal branches of the MMA and drained into the sylvia veins. You see the fistula's point here. This is the lateral again. This is ipsilar to the lesion. Again, right-sided, both right-sided. And in addition to that, on the internal carotid, it's very, very subtle. And these are the things that always Dr. Gross picks up on and I don't immediately pick up on. But there's a very early, very slight but somewhat early drainage of the bay as well. So again, with these findings, we have this type III fistula. We have this ipsilateral kind of punctate bleed, a vague history. There's no really other findings that is grossly an exam that suggests any syndrome. Dr. Zipfel, what are your thoughts? And please let me know if you want me to go back and go over something. The lateral one, that one again. So there's the fistula's point here. Yeah. Well, am I on? Yeah, you can hear me, right? Yeah, yeah. So I mean, it's ipsilateral to the hemorrhage. The hemorrhage seems a little remote compared to where the fistula itself is. And you don't see any T2 or flare changes. I mean, you see like a rim of it right around the hemorrhage, but you don't see a more elaborate T2 or flare change in the hemisphere around that that would kind of clearly indicate corticolvenous hypertension. But so it's not a great match, but I would assume that somehow the fistula and the corticolvenous hypertension that there led to that. 34-year-old, no other reason to have a hemorrhage. So I would consider it a hemorrhagic, symptomatic, aggressive presentation for a fistula, even though it doesn't completely match as well as sometimes you'd like. And then you look into treatment options. I certainly would want to see what my interventional colleagues would think. We typically do endovascular first for these fistula. And if that can be done, that's great. We would do it early, within 24 or so hours of presentation. And then only if there was a problem treating it endovascularly, we would consider surgery. If surgery was considered, I usually would do a CTA to try to identify that dilated arterialized vein so I really know where the target is for the disconnection. I mainly do a disconnection. If there's dura in the region that looked arterialized or there was engorgement, I will coagulate it. But I personally don't, I'm not super aggressive about doing more than disconnection because there's good data, I think, out of University of Toronto where disconnection is very effective at treating any recurrent events associated with a type 2 or 3 fistula. So those would be my initial thoughts. And let's see what our interventional folks would say. Right. I mean, the main question here, which it was a little bit puzzling, was whether the two were associated. It's hard to ignore, but it is a type 2 fistula. It's ipsilateral. And again, even with the fistula alone, without the bleed based on your data, we were leaning towards treatment. Brad, why don't you tell us a little bit what you were thinking about this case. I mean, I agree with everything Dr. Zipfel said. I think it's, I think just this is one, so we have the, we're fortunate to be a part of Dr. Zipfel's Chondra Consortium, which is really a unique and one time opportunity for fistulas. And actually, it's interesting, in the database, when I put this into our database, it's very thoughtful because it'll actually say hemorrhagic presentation and then possible and probable. Is it really possible? There's actually a, and I think this is one I put possible because shucks, I mean, the drainage is into this inferior temporal vein. And so it's possible, right? But I think the conclusion is still, I mean, let's pretend it was totally not related. You're still going to treat this type 3 fistula in a young patient. So I put this as a possible, but shucks, it's just so hard. And I think that thalamic hemorrhage, so this patient, you know, we did a completion angiogram on him. He has a small AVM or capillary, you know, these HHT patients, they have these capillary malformations too that are sort of half AVM, half capillary phalangitis. So it's possible that thalamic thing's a little cav mal, it's going to pop out. But I don't know. So I certainly agree, you know, like Dr. Zipfel says, you know, you see if you could take an endovascular swing at it, I would certainly treat it. And, you know, we talked about, I think the subtlety here, you know, the shadow catcher on the endovascular side to some of these cases is you need to look at the normal venous anatomy. And so in this case, that temporal vein, so you can't treat this transarterial. So to just briefly talk about the treatment of this, those little sphenoidal branches, you're not going to be able to penetrate transarterially. You'd have to pressure cook it, take the MMA distally, and that would potentially have vision issues and cause blindness. So the way to treat this endovascular would be transvenously. But the problem with transvenous embolization here is that that whole temporal vein that drains the fistula is also draining the normal temporal lobe. And, you know, this is right on the sphenoid ridge. As a fellow neurosurgeon, I can't think of a simpler, you know, George gets the terion in about 15 minutes. So he very successfully disconnected it. And I think it was probably the right thing to do as opposed to try to flirt with a venous infarct. So that's what ended up happening here. Yeah, so we discussed it with Brad. And for a number of reasons, we decided to do it like a mini teriornal because it's a small, short video, but because exactly he was young and it wasn't clear whether the origin or I did try to separate the temporal dura and try and resect it. So we did a little bit of peeling. Here are the two. When we did the, one of the downsides and perhaps a mistake is when we separate the temporal dura. When we did this upsurgery ICG, it was hard to see exactly the fistula's point. But there are only two draining veins going anteriorly in the temporal dura. And we disconnected it. And then we went on to, again, peel some of the, I'm not going to belabor this, but peel some of that temporal dura and essentially resect the large patch, going all the way just lateral to the uptake and down to the pitrochromatic fold all the way down to E3, all that dura that's medial over the sphenoid and the cavernous. There were this corkscrew veins here. We did another ICG. They did not seem to be arterialized or really feeling afterwards. And he did fine, but we did an angiogram right after. That showed, indeed, that the fistula's point was gone from the external run. But he still had this very subtle but early venous drainage in la baie. So again, we're discussing what the cause for that is and likely maybe a small micro AVM or something along those lines. But it's still a little bit of a mystery. We referred him to genetics, but there was no clear outcome from that as of yet. Any other thoughts, Dr. Zipfel, why that early drainage? Obviously, we tried to do a story of a disconnection as well as resection of the dura there. Is there anything else that you would do? No, I would have just done the disconnection, rather than just, like I mentioned before. I would definitely do a delayed angio. I don't know how long this goes, just to see if anything shows up in a delayed MRI, to see if a cav mal pops up more clearly later. Those are the follow-up things I would do. The HHT thing, we have a really big HHT clinic here. I see most of those patients as they get screened, because we screen them here. And I don't remember that many fistulas. There's the AVMs and telangiectasias, but I don't remember a lot of fistulas, actually. So I actually don't know literature on this. But my gut is that HHT, I get to make some sense that there may be fistulas, brain fistulas, but I just don't remember that many. They're almost always AVMs, usually smaller size. Some are these kind of these, you can't even see them on axial imaging. But when you do the angiogram, then they show up, and the telangiectasias. But genetic testing makes sense to see. So great. One other quick case, and I know everyone had a long day. And then we'll wrap it up. But there's an older lady that presented with a subarachnoid hemorrhage. Mostly, the hemorrhage was down in the cervical medullary junction. And then on the CTA, it did seem that she had some corkscrew vessels right around the cervical medullary junction again. So she went on to have an angiogram, of course. We had a little bit of a discussion on how to call this. But at this point, what are you thinking, Dr. Zipfel? Yeah, clearly, hemorrhage related to a fistula and need to be treated. These are fun surgically to do. And usually, I would do a far lateral just to make sure you have all the exposure you need. The far laterals I do, I think that they never get destabilized. No, I mean, I think it's almost like just drilling a sphenoid wing. And from my perspective, we're not going to push it. So I don't think there's any. I do it very frequently for lesions in this area. And I don't think it adds virtually any risk at all. And I'm not worried about destabilization. So if I operated on it, it would be through a far lateral to that side. And again, looking for the arterialized vein coming out of the dura and just dividing it and repeating an angio intraoperatively to make sure you have it. And that's what I would do. Probably do some lower cranial nerve monitoring. But I would want to see if interventional could treat this. So I'll just mention that a lot of these can be treated interventionally, has been my experience. I'll just mention two areas where I think surgery often is primarily used. And that's the anterior fossil floor, dural avial fistulas, a lot of centers will only, surgery is the first choice for those because they're fed by the anterior and posterior ethmoids coming out the ophthalmic. So you're going to have to cannulate and embolize through the ophthalmic. There are a few people, few centers that have published on interventional treatment of anterior fossil floor. Ricardo Hannel has a nice series and others. But I think most centers and certainly ours, we treat those almost exclusively with surgery. And then, and the other one is the tentorial region dural avial fistulas. Those lesions, often it's difficult to treat interventionally because the circuitous route of the dural pedia is coming off the cavernous ICA. And so those are ones that we, I don't mind if they think they have a chance in trying, but more often, quite often, surgery is required for those. But other than those two locations, like Dr. Groth mentioned earlier, a lot of interventional approaches can be used for a lot of these fistulas. Brad, what specific concerns about endoscopic treatment of, I guess, this specific one? I think one of the concerns we have is potentially a mixed. So if you have mixed, so if you have a dural fistula that has mixed peel and dural supply, the endovascular dictum is that you can get in trouble just taking the vein. So the way, you can't treat this transarterially. So the way you treat this endovascularly is you would go transvenously through the straight sinus. I measured the length of the vein. We may have run out of length. You'd have to do a direct jugular puncture, which is the end of the world. But the concern potentially was some peel supply as well. If you don't take, so some of the people in Europe that have done like four or 500 fistula immunizations will tell you that if there's peel, it's really not peel. It's transdural peel supply where the peel arteries traverse the subdural space into the fistula. Sometimes if you take the vein, they can hemorrhage after, kind of like a mixed AVM. So that was one of my concerns with this. And also actually getting, again, there's no transarterial option, it was transvenous. So I wasn't, I think our discussion was, look, I could take a swing at this transvenously and we could see, but I didn't have a warm and fuzzy feeling because of possible peel supply. And so I knew you'd do a very good job with this and you did, so take it away. And you know, if this is a pure fistula, these usually go very well surgically as we know as surgeons. Yeah, as Dr. Zippo said, this is not really a really challenging case to do once you're there with the far lateral and the dissection. There was actually, we weren't worried in the lower cranial nerve. Sometimes it's helpful when there's thick blood. And there were a few official points around the C1 nerve root, just dividing the dentate. But putting clips all first, moderating neurophys, and again, there were, it seemed like there were more than one official points there, but one by one, putting the clips on, just ensuring with neurophys that there were no changes to make sure that we weren't actually disconnecting something with direct supply to the cord. There's still one more, and total three official points that were disidentified there. Dr. Gross performed the angio-radial approach, perhaps the most difficult part of the procedure doing this here. Early days of radial here. And actually, yeah, she did quite well. The one here follow up obliteration of the fistula. Just a couple of questions, and then we'll wrap it for the day. Dr. Zipfel, you've been at the cutting edge of producing data for multiple centers for this. Again, as you said, many of them are retrospective. What do you think is the role for randomized trial or do you think there's a subpopulation that's worth actually performing a randomized trial for dual AVS in their treatment? Yeah, I think there's a couple of populations. I certainly think the low-grade fistulas with symptoms, at least our data doesn't clearly support treatment at the level that I think we are treating. And so I think that's something that could be examined. And then I think the other one is the low-grade, I mean, the high-grade dual AV fistulas without vena tectasia and without aggressive presentation, I think there could be a case made for randomizing that patient population. So those are the two areas that I think are, I mean, I don't personally think it's controversial and I wouldn't advocate a randomized trial for ruptured high-grade dual AV fistulas into, I mean, you might compare treatment approaches, but you're not gonna observe versus treat in that realm. And I also, although the vena tectasia data I don't think is as strong as the motor presentation data, I think there is quite a bit of data that would suggest that there are higher risk. And I personally wouldn't focus my efforts on that particular type of fistula. I would focus it on the benign presentation without vena tectasia. That's the area that I think you could. Excellent. Brad, as we go forward, hopefully less and less, we will need to cut people open to do this. What do you think are the advances down in the endorhazical world? You briefly touched upon a couple of things that may allow us to treat this more and more endovascular, even more complex lesions or distal ones, sort of just push the envelope endovascularly. Right. I think one of the things that is, so there've been several iterative advances when we think about endovascular treatment. The first was onyx. I think the second big advance is the introduction of balloon microcatheters like SEPTR, Eclipse for embolization, where you can really pressure cook these things and use the occipital artery much more effectively than we ever could before, particularly for those. I mean, no one likes a type two transverse. I mean, those are not ones I would ever send, I would embolize those till it hurts before I send one of those to the OR because those are very treacherous. You look at some of the old data from Dr. Suntz when he would operate on those things. I mean, so that's been a big advance. I think definitely some of the limitations are still in those ethmoidal fistulas and really the moral is getting there transvenously, which you often have to do jugular punctures with. I still think ethmoidal fistulas and tentorial ones, just like Dr. Zipfel said, are really still just anatomically very vexing. I can't get my brain around it unless you can get there transvenously. The transerterial conduits are not only too serpentine as Dr. Zipfel says, but they also have a risk of refluxing into petrosal branches and doing a lot of damage trying to hammer some of those deep tentorial ones, not just failing, but hurting the patient. So I think that is still an endovascular frontier. One plug that really, I think, I'm just a small part of Dr. Zipfel's consortium, but it really, when you look at the literature on dural AVFs, back in the 70s, you had Gin John's, big innovation, and then into the 90s where you started having some big data come out of Toronto and then Dr. Zipfel's paper about the importance of symptomatic presentation. But I really think that this condor opportunity is really brilliant. I mean, these are very rare. These are very, very rare. And Dr. Zipfel's plan to get us all, everyone together, literally a large number of centers to get some real data on this. I mean, I would really look forward to the epoch, the literature epoch in the next few years from what comes out of this condor consortium, because it's going to be very hard to replicate for this very rare, very, very rare pathology. I think we're going to learn a lot. We may realize some things we thought were right were wrong and vice versa. So I really, you know, my hat's off to Dr. Zipfel for doing this and it's going to be very exciting. Fantastic. Well, I want to thank everyone again for their precious time. I certainly learned a lot every time I hear you speak. And I know everyone will as well when this is on the website and people can access it again and again, just to absorb everything. I also want to thank Shannon for her precious work in getting us going and making this happen. And I wish everyone a good night and see you again soon. Thank you, everybody. Thanks. Great job, Georgios. Thank you.

dural arteriovenous fistulas

DAVFs

neuroimaging findings

treatment options

venous drainage

endovascular treatment

embolization

Onyx

transvenous embolization

surgical intervention

high-grade DAVFs

stereotactic radiosurgery

individualized treatment plans