So, the next topic is AVMs, and when I was a resident watching Dr. Spetzler operate on AVMs, I found myself a lot of times just wondering what was going on and how he was thinking about these cases and what the plan was, and it was difficult for me to really figure out AVM surgery. So, it's much more difficult to understand than, say, aneurysm surgery, where you're working from a particular point on the circle of Willis out to your lesion. You're gaining proximal control, distal control, dissecting the neck, and applying the clip. This is a far more difficult task to deal with this kind of a problem here and figure out how to take yourself through the steps, and so I spent a lot of time thinking about this and trying to develop concepts and a system for doing AVM surgery, and that's based on this experience. I'm up to about 670 or so, so it's a lot of AVMs, and I think a lot of these things really didn't start coming to me until I committed myself to breaking it down, and so I just wanted to share some of those thoughts with you. So, first of all, it's like a war. The mentality of an aneurysm case is more like a ballet where you're in there and you're doing these choreographed steps, and it's very clean, it's very elegant. This is different because you're up against a very fearsome enemy. You have to view yourself more as a soldier in a war, and really the keys are learning the battlefield, which means defining your anatomy very carefully, developing very strategic battle plans, and knowing the enemy. Learning some ways of distinguishing this vast array of AVM pathology and kind of boiling it down into types, what I call types and subtypes. So I've broken these down, in my mind, into seven AVM types, and it's all regional based on the lobar anatomy or ventricular deep brain stem or cerebellar. So first off, if you just think about these seven different regions of the brain, that's the place to begin. And then the next thing to think about is the subtype, and the subtype really is just about what surface does the AVM live on. If you think about AVMs, they are based on a surface. Just think about the convexity AVM. It's got a broad surface, and then it funnels down as a cone down to the ventricle. So if you think about every AVM in that same way, there's going to be a surface that it's based on, and it's usually conical, and it's based on that surface. So frontal lobe has different surfaces, and all the different regions, the seven regions I talked to you about, have surfaces, and these are what they look like. The frontal AVM has this lateral surface. It has a medial surface. It has a basal surface, and it has a sylvian surface. I've also included what I call a paramedian AVM, because a lot of these AVMs come at the intersection between two of those planes, and particularly for the convexity at the lateral and medial surfaces. So there's this paramedian. So there are five different frontal AVMs. If you look at the temporal AVMs, this is just a different view showing that lateral medial. These are cross-sectional views showing the same thing. The temporal has the same concept. The temporal lobe has a lateral surface, a basal surface, a medial surface, and a sylvian surface, and those define the four different types of temporal AVMs you're going to see. And here it is in cross-section. Parietal occipital, I put those together because if you think about the surgery for that, they're very similar. The steps that you take in dealing with the parietal AVM are very much like what you deal with for an occipital, so there's really no need to break them up artificially. There's really more of a front and a back to these AVMs. And these are the different surfaces, once again, with that paramedian one thrown in there at the confluence of those two surfaces. Going deeper, the ventricular AVMs break down into the colossal ones, those actually in the different parts of the ventricular system, be it the body, the atrium, or the temporal horn. Now, as we go to the deep AVMs, they can be in either the basal ganglia or thalamus. And looking at it a little differently in cross-section, you see the basal ganglia and thalamic AVMs, but there's also these insular AVMs and these pure sylvian AVMs. So these are these deep regions, and you get these six different variations there. This is another view looking from the top down, and you can see how you work your way from sylvian to insular, and continuing more inward, you go basal ganglia to basal ganglia to thalamus and thalamus here. Top of the thalamus here, medial wall of thalamus, and here we've got the caudate head and here we've got, sorry, this is putamen, this is caudate head. So those are the different variations for the deeps. And the brainstem, there's six ones that I've seen. I've actually now seen seven, so I might have to modify this, but these are broken down by midbrain pons and medulla, and again, surfaces either anterior, lateral, or posterior. And this is just another view showing the different perspectives. Cerebellum, same idea. There are four cerebellar surfaces. There's the suboccipital, tentorial, patrosal, actually three surfaces. Those are the surfaces. And then the vermis and the tonsil in the midline that form the other subtypes. And these are just different views of the same. So that's the enemy, okay? If you think about all the different AVMs that you see, and you think about what type is it, what subtype is it, you're going to be able to start putting these into categories or boxes that make sense to you. And that way, when you are dealing with a different AVM, you don't have to rethink it every time. You can just say, look, this is a particular type and subtype. I'm going to be able to figure this thing out. Now the next thing is to, we're surgeons, we do steps, we do techniques that are broken down in steps. We break down AVM resection into something that's methodical, that we can recognize every step of the way. And so these are the eight steps for AVM resection, and I'll take you through each one. Now exposure is all about this box. You want to think about the AVM as a box with six sides. And you've got to get an exposure that's going to bring as many of those six sides into your surgical corridor as possible. Now for a convexity AVM, it's very easy. You just have to make a wide craniotomy that is right over the lesion. But it gets more challenging for different lesions, for example, on the medial frontal lobe or on these deep, deep areas. But the other nice thing about the box is that you can take the anatomy and you can sort of impose the anatomy onto this construct conceptually and define for every type of AVM the particular anatomy. And so we'll see how that's useful in just a minute. But there are a couple other concepts with this exposure idea. If you can bring that box into a, what I call a parallel approach, where you're facing everything parallel or direct on, it's going to facilitate the resection because you can see the superficial side. You can get around the four sides. And the only thing that will be difficult is that deep side on the deep surface here. It's very different from the perpendicular approach where you're going more tangentially. And I've selected this sub-temporal or this basal temporal AVM as an example. If you're going sub-temporally and you're trying to resect the AVM at an off angle, it's going to mean viewing one way with your approach, but then turning another way to get around the sides of the cube. So that's what I call a perpendicular approach, and it's a far more difficult resection. But the advantage is it avoids having to go through brain that would otherwise be eloquent or be in your way. So this becomes something that we're trying to achieve, but it makes it difficult for us. The other idea with the box is that sometimes you have free sides. So for example, this is a frontal convexity AVM. This requires no dissection. That's one side of your cube that's right there. As soon as you peel the dura back, it's already free. Whereas if you look on the flip side, the basal ganglia AVM, there are no free sides. You've got to go through brain, no matter how you come at this, to get to a border of the AVM. And these are other examples. This is two free sides. This is three free sides. But these free sides, the more that you have, the less dissection you do to get it exposed. Now step two is about subarachnoid dissection. That means after you've done the crani and opened the dura, this is the work that it takes to really get to that surface or that side of the AVM. There are a whole host of different approaches that you can use. It depends on where you're trying to get to, but these are the different approaches that are available. And the other key thing about subarachnoid dissection is that what you're doing really is you're taking all this angiographic anatomy and you're figuring it all out. You're applying that and imposing that onto your surgical field. And that helps you define these margins here between normal pia and AVM. It helps to develop these planes around the AVM. It may be splitting fissures like this one in the sylvian fissure. Or you may be looking for this one little arterialized vein here that's going to lead you to the AVM, which otherwise can't be seen on the cortical surface. Now third is the draining vein, and that's probably the most important thing conceptually. You've got to have a vein that you've identified as your outflow that you're going to preserve throughout the resection because that's what's going to keep you safe. If you compromise one of your only outflow vessels, then this thing will rupture, obviously. So that's the critical concept. But the other thing is that these provide a compass. Sometimes all you see is that cortical vein and you don't see the nidus, so you've got to follow that arterialized vein down to even get to the AVM. So it provides you with guidance. It's an odometer. When that thing's red, you've got a long way to go. When that thing starts to turn blue, you're almost to the end of your journey. So it's a guide for you determining how far along the process you are. The key is recognition. You've got to be able to look at a big red vessel and distinguish artery from vein because it's so tempting to see a big red vessel and think, ah, I'm going to just take that and this thing is going to all of a sudden turn blue and I'm going to have a nice easy time of it, when in fact it could be one of these veins and all of a sudden your AVM becomes tense and explodes. So you've got to recognize very carefully by looking at the wall, seeing if there's any of those medial elements, those muscular elements, looking for very subtle variations in the color. You can often see turbulence through the wall of the vessel, which would be an indication that you're on the venous side. These are the things that under the microscope, under high magnification, you can start to visualize. Another concept is this idea of primary veins and secondary veins. If you've got this AVM here that drains both back to LaBay and forward to the superficial sylvian veins, you don't have to preserve both. You can take one or the other, but you've got to designate which of those veins is the primary. If this is the primary, then you can take this, but you've got to be really guarded about this one. And typically what you want to think about is which way is your dissection moving? Is it moving from front to back? Are we going to roll this AVM this way as we go underneath, or are we going to roll it this way to get underneath? And that has to guide that decision. Now feeding arteries, there are different variations. There are your classic terminal feeding artery that comes in and ends in the AVM here. That's the beautiful vessel that you want to take early because that will be an easy victory. But there are others that are much more challenging. This is a transit artery. It starts here, it feeds the AVM, and then it continues on. If you don't protect this distal outflow end of that artery, you'll end up with an infarct. So you've got to recognize that transit vessel and treat it differently than you would this terminal vessel. Now on the other end, you've got these perforators. You've got white matter perforators, and you've got ventricular perforators. And they're very small, but they're very thin-walled, and they're very difficult to coagulate. So these, I think, are some of the more difficult ones to deal with and can really ruin your resection if you don't manage those carefully. The bystander just goes right on by. It has no contribution, and you have to recognize it because if you view this one and think it's a terminal vessel, you could end up with an infarct. So think about those five or so different types of feeding arteries, and here's an example of how that's relevant. The transit vessel, you want to dissect from distal to proximal. If you're worried about preserving the distal end, you've got to find that first and work your way this way. If you do the other thing and start dissecting this way, you might think that one of those ends is going distally when, in fact, it might be the other one. So the answer is to really follow it from distal to proximal, and you'll end up preserving the end that you need. So step number five is peel the section. All the stuff before is just about sort of figuring it out, kind of surveying the anatomy and getting things ready for the battle. Here's where we actually engage, and it's easiest to start in the pia. This is where these intersections between peel feeders and the nidal margin are coagulated and interrupted. There are usually clear points where you can get these early victories. Those are easy. That's all on that superficial side of the AVM. It gets harder on the deeper sides, the four sides of the box here. This is where we actually have to get into the parenchyma a little bit and start working around. This is where you really have to gauge this dissection distance. So what I mean by that is if you drift too far into parenchyma, it's safe. It keeps you away from the AVM and the bleeding that might come from getting into the AVM, but you're starting to get into eloquence, and so you have to balance that with wanting to hug this very tightly, which preserves your brain tissue but can get you into the nidus. So there's a constant evaluation of that dissection distance which you have to think about your eloquence as well. It's also relevant for compact and diffuse AVMs. The diffuse AVMs, it's very easy, sorry, very easy with compact to judge that distance. For the diffuse, it's very difficult because you've got this zone here where it's very hard to see where AVM starts and where brain ends, and here it's about defining that distance, which is a challenge. This is to just make the point about eloquence. You've got to know your eloquent areas, and you've got to respect those. There's some variability depending upon the individual anatomy. Sometimes functional MRI will show that anatomy has been shifted. Sometimes you have to just go with the traditional anatomic eloquence, but it's important to know where you are and to respect that. Now the seventh step is about ependymal dissection, and I think of it as the dark side of the moon because if you've come around all the five sides of our cube and you're now on that sixth side, it's dark. You can't see exactly what you need to see, particularly with a large AVM that's obstructing your view, and so this last side of the box, which is always the scoop in these drawings is tough. You've got to mobilize the AVM to one side or the other. This is where you're going to encounter all those deep white matter perforators that are tough to cauterize, and this is where you can sometimes have an AVM get away from you. So how do you deal with that? Well, these are some tricks. You want to line your cavity with Telfa so that you protect the parenchyma. You want to use more retraction for this last step than you would otherwise because that's going to help you see. You want to retract the AVM. Better to move the AVM than move the tissue, so be aggressive with that. And finally, use these clips. When you can't cauterize one of those feeders, those microclips are real lifesavers. Now this illustration is important because if you misjudge the point where you come across the tip, you can end up with a choroidal feeding artery, an appendable vein, and a small AVM remnant that will show up on your angiogram, and you'll be forced to go to your patient the next day and say, look, we've got to go back to the OR because we left something. It's very humbling. It's very embarrassing. If you make that judgment correctly and you go all the way around, usually down to the ventricle, you won't miss that piece. So it's very important not to jump the gun. Finally, the eighth step, resecting this thing. That's when things turn blue. You can pick up the AVM and roll it out, but you want to make sure that you haven't done that too soon. And if you have any redness in the vein, you want to be careful. It's usually that you've missed something. Hemostasis, couple comments about this. You want to make sure that you maintain absolute hemostasis all the way through. When you have bleeders, it's all about just suction and cautery. When that fails you, then go to those microclips that I showed you. Never pack things off. You never leave a bleeder. You always want to stay on top of it because if you pack it, what happens is you just don't see the bleeding. It bleeds back into the parenchyma, and you'll end up with a big clot that you don't know about. Once in a while, things just get out of control, and there's no other way than to just get that thing out as quickly as possible. That's what I call the command over section. If things are getting away from you, sometimes you just have to bite the bullet. Those are the steps. There are eight steps to think about. Each one of the AVMs that you take out, you'll be going through those different steps so you can see where you are in your mind's eye before you go on to the next. This is where we put it all together, and these are the battle plans. If you think about your type or your enemy, and then you think about your different steps, and you fuse those concepts together, what you end up with is the ability to build battle plans for all the different subtypes. This is just a chart that shows you the frequency of some of these different subtypes. The frontal ones are some of the more common, so I just wanted to take you through these. Here's an example of the lateral frontal one. This is probably the most common AVM out there. You'll recognize it here. Here's the frontal pole. The temporal lobe is here just to orient you. Here's our cube. As we build that construct, what we can do is we can impose the anatomy onto the cube and know that these feeders are going to be coming up from below, from the sylveon fissure, your prefrontal, your precentral, and your central arteries. You're going to find those all along this edge of the cube. The veins are going to be draining superiorly to the superior sagittal sinus. There'll be very little on the deep side to worry about. It's really about dissecting this plane here, preserving these veins here, and when you transition to parenchymal dissection, being aware of these deep feeders that are going to be coming up from the lenticular strides from below, sometimes the recurrent artery or cubin from below, but dealing with those on the deep side. Here's a case example. This is a lateral frontal AVM shown here. This is what it looks like. Again, you have none of those arrows. You have none of those numbers in real life, but if you think about these concepts, you can take this daunting view here and convert that to a series of steps that take you to the feeders, preserve these veins. You see your odometer here, nice and blue, and you can get that AVM out very easily. Here's a medial frontal AVM. Here we are. We've just moved a little bit more medially, not very far, but look how everything changes with this AVM. Here, we've got the head turned laterally, like I showed you for the bypass. The midline's horizontal. We're going to go down the inner hemispheric fissure because these AVMs come to that medial surface. Our anatomy here is callosomarginal and pericallosal, all fed from below on the cube, so we have to be aware of that side as our vital side that feeds the AVM. The drainage is upwards to the sagittal sinus, these medial frontal veins, and as we do our dissection, it's all in the inner hemispheric fissure. Gravity is retracting. We get a perpendicular view. This is not one of those parallel views, so we're seeing the surface, but we're kind of angled in a way that takes us off the sides of the cube. We have to draw this AVM up into the fissure to get this out, and that's what's demonstrated here. Here's our surgical trajectory, but we're dissecting in this way to get into the parenchyma, so you have to draw it out a bit into the fissure to get those deep sides out. Here's our case example. Here's that medial AVM. It's right above the cingulate gyrus here. Here are the ACA feeders coming from below, and here's the view interoperatively. The gravity retraction opens up the fissure. You can see the margins of the AVM. The feeders come along this plane here. Here's one example of that feeder. After we've come across that, you can see the blueness in the veins, and it simplifies the resection. Here's a paramedian frontal. This is the combination of the two. Once again, different subtype, different strategy. Instead of midline horizontal like I showed in the last one, now we've got the nose up because what we want is both the interhemispheric fissure side and the lateral side, and with table rotation, we can move things to achieve that, but you have to get the patient positioning just right. Here's our cube, so you can see how the anatomy lays out. You've got lateral feeders. You've got medial feeders from the ACA, and the drainage is all the way medial to the sagittal sinus. Here's how that looks strategically. Here's the deep dissection. Here's going across the ependymal side, and you can see some connection down to the ventricle. Here's an example of this one. This is, once again, a large AVM. We've got the nose up, so here's the phalx. Here's the frontal pole. Here are the feeders coming from the ACA side. Nice access to that down the interhemispheric fissure, but we also have access laterally along this margin here because of the way we've positioned the table. The basal frontal one is shown here. This is one where you want to get low on the inferior surface of the frontal lobe, so an orbital zygomatic is helpful. Here's the cube showing the anatomy laid out, mainly feeders from MCA, but also feeders from ACA, both from below and from the interhemispheric fissure coming from the side, which we see here. There are three different types of feeders that come to this one, and here are different steps from one side to the other. Here's a case example. Nice view underneath the frontal lobe. Here's our dissection plane, olfactory tract, and you can see for yourself how that lays out. Sylveon frontal. These are really nice AVMs. These are in the sylveon fissure on the frontal side fed by MCA vessels that go along its lateral edge inferiorly. It requires a sylveon fissure split getting into the frontal lobe, parenchymally on that side, and here's a nice example. Here you can see it's on the frontal side of the sylveon fissure fed by these branches that come up and out of the sylveon fissure, so as we split the fissure, we see these feeders coming in here, and this one had been previously radiated, so there's that nice gliotic margin, and there's a nice one. I'm going to show you this one just because there's some video. It's a ventricular body AVM, and so like the gravity retraction cases I've shown you, the head is turned to allow gravity to open things up. You can see it's mainly ACA fed, deep venous drainage down to the internal cerebral veins. This is the interhemispheric fissure. This is the phalx up here. This is the contralateral frontal lobe, but if we separate the two, we can start to see the collosal marginals here. Here are the pericollosals. Here's the corpus callosum down here. You can see that bright white of the corpus callosum. As we move forward, we can see the pericollosal vessels here and here, and the feeders are essentially coming off from the genu all the way back. They come off the genu, come off the rostrum, and they go right into the ventricular spaces. Here we're going through the corpus callosum, and this will get us down to the ventricle. Here's clot in the ventricle. As we get that clot evacuated, we can start to see the AVM anatomy, which is primarily in corpus callosum and septum. Now we're in the ventricular system. There's the septum here. There's the view in the ventricle. Now you start to see AVM here. This is the pericollosals up above, sending feeders downward. This is the view into the ventricle. Just by following those feeders along, we can close them down. Let me speed this up a little bit. Working our way into the ventricle there, you can see skeletonization of the pericollosal vessels. You see those feeders as they drop down and into the corpus callosum. This is working our way across the front end of the nidus. The trick here is the back end of the nidus, because at the back end of the nidus, what do we have to worry about if we're in the septum? We have to worry about the fornix. Here's the fornix right here, coming around. We have to stay off of the fornix, because fornix is very intolerant of any kind of manipulation, so this is being rolled away from the fornix. Foramen of Monroe is back here, back in this space here. You can start to see now the vein is turning blue there. We've gone across most of our feeders at this point, and the vein is going back towards the foramen of Monroe and to the vellum interposinum, roof of the third ventricle. These are some of the last feeders here. Now, this thing is pretty much done. There's the venus tail. Things are nice and blue. And there it comes out. So you can see how that AVM occupied the corpus callosum and septum. There's the view, complete resection, and she ended up doing just fine. I'm going to skip through this in the interest of time. I think the course is going to give you a copy of my book that has all this stuff laid out for you. So you can read this in as much detail as you want. Let me just go to the end here. So these are my concluding thoughts for AVMs. Each AVM is not unique. So this is a way that I used to think about AVMs. I used to think of them all as being different. But I think for you guys learning your techniques, gaining your confidence, what I think would be helpful is to think of them as not unique, as being all categorizable into these types and subtypes. So that when you come across one, if you work at trying to identify one, you can go to the book and review that and think about the anatomy, more importantly, to give yourself a sense of your enemy. It'll help you put together the eight steps for the resection and defining that battle plan. So that when you go to the OR, you've got this all clear in your mind what you're up against and how you should deal with this. I can't emphasize this point enough, study the angiographic, radiographic anatomy, because really what makes us good in the OR is our understanding of the anatomy, both the pathology and the normal anatomy, and removing all those surprises. If you have spent your time going over that stuff in detail, there should be no surprises in the OR. And it's really a matter of recognition rather than interpretation. So you're really just kind of going in there. And it's like the quarterbacks that review game tape before the game. They know what their opponent is going to do. And so there are no surprises. And they can read the situation quickly. And I think if you view these cases in the same perspective, then you'll be well prepared. And it's all about the angiography. We've got to be very careful about our outcomes now with Aruba. I think we've been challenged in much the same way that ISAT challenged us to improve our results with aneurysms. So we really have to be careful about our outcomes and make sure that we get excellent results. And finally, be courageous. These are not easy cases. And I think it's very easy for us to say no to AVM surgery, because it can be painful. It's not just these technical steps that I've outlined for you. You are literally one mistake away from disaster. And I think these are some of the most painful cases for me personally, bar none. So you really have to have courage, I think, to say yes to a tough AVM and to take this to the OR. You have to be willing to handle those complications. And remember this. It's not just about what we're capable of doing, but what we choose to do. And I think we really have to choose to take these cases on. I think we've got to be careful on how we select our patients. But I really think we need to stay in the game. I know that people have put pressure on me not to be as aggressive with elective, unruptured AVMs because of ARUBA. And I tell them, look, I'm completely comfortable in doing these. I know I can get good results. I think this is the right choice for the patient. And it's going to be harder for you guys. It's easy for me to say that with 650 of these under my belt. It's hard for you to say that with 10 or whatever your number is. But I think you have to believe in where we've come from, what we've developed, and just be courageous and take these on. So that's it. Thank you.

Dr. Spetzler

arteriovenous malformations

AVM surgery

brain location

surgical strategies

anatomy understanding

AVM resection