Hi, this is a talk on pediatric cavernous malformations and I am Ed Smith from Boston Children's Hospital. I have no disclosures to make. So in discussing cavernous malformations, I am going to be doing this as part of a team with Dr. Steinberg and Dr. Lawton. I am going to be speaking mostly about the pediatric aspects of cavernous malformations and just by way of outline, I think it is important to go over a few questions, namely what are they, why are cavernous malformations a problem, and specific to kids, what are some of the treatment issues that we face as neurosurgeons, specifically with children, what is the natural history, and if we do have kids with known cavernous malformations, how do we as surgeons select the appropriate patients for surgery, what are the indications for surgery. So those are some of the things I am going to try to cover with the discussion here today. In terms of what cavernous malformations are, basically these are slow-flow vascular malformations. You can see here a little MRI with the arrow pointing to the malformation. They have been classically described as this mulberry-like appearance. I have actually never really seen mulberries, but you can see a picture of one here and then in an intraspinal lesion, you can see how that cavernous malformation in the surgical field really looks a lot like the berry, and there are some known genetic associations with these lesions. In terms of the pathophysiology, these cerebral cavernous malformations are slow-flow, so unlike arteriovenous malformations or aneurysms, which have an arterial component with high flow, these generally tend to be lower-pressure lesions. They affect actually a very high percentage of the population, about 0.5%. They are, as I mentioned, disorganized clusters of this immature vasculature. Generally they are sort of a single layer of endothelium, and one of the big problems is they lack these tight cell-cell adhesions and smooth muscle cell layers, so essentially they are like regular capillaries that don't have the walls stick together very well. The endothelium is therefore kind of leaky. The vessels stretch and create this bubble-like appearance, and this can lead to both mass growth and then ultimately hemorrhage. In terms of the cause of these, it's a very interesting lesion. There's been some recent work involving the microbiome, where a number of studies have looked at the role of gram-negative bacteria in the gut, which release lipopalysaccharides, and it seems that this may stimulate an endothelial cell receptor, TLR4, and this interacts with this complex of proteins that regulates the normal endothelial cell adhesions, and when there are difficulties or lack of this protein, this may result with increased rates of bleeding. Genetically, the majority of cavernous malformations we see in the pediatric population are sporadic. They're solitary lesions, but about 30% are transmitted as genetic familial mutations. These are typically autosomal-dominant mutations, and most commonly, these are multiple lesions, so if you get a scan and you see many lesions throughout the brain, it is highly likely that this is an autosomal-dominant genetic lesion. There are three different mutations that are best known, CCM1, CCM2, and CCM3, and these three mutations make up about 90% of all the familial cases. In terms of the pathophysiology of these specific mutations, CCM1, which is also known as CRIT1, basically normally suppresses this Rho-A-GTPase, but when there's a mutation of this gene, this leads to an overactivation of the GTPase. This increases actin fiber activity, and there's a loss of cell-cell junction and adhesion, and this leads to bleeding from blood leaking from around the vessels. CCM2, also known as malcabin, creates this complex that regulates the same GTPase, and when CCM2 is mutated, there's similar increased activity to the CCM1 mutations, and this CCM2 also affects angiogenesis. Lastly, CCM3, PDCD10, is less well understood, but it appears to have a role in cell adhesion and also apoptosis regulation, and children with this mutation, the CCM3 mutation, tend to have the most severe phenotype. These different proteins do interact with each other and create a CCM complex, and this is involved in regulating endothelial cell adhesion. Why are these lesions a problem? Why do we care about CCMs? Well basically, they can grow. As mentioned, these bubbles basically expand, as you saw in the initial picture, and then ultimately they can leak and bleed, as you can see here in the spinal cord MRI, where the top study shows this very tiny dot in the thoracic spine, which then expanded and bled, leading to profound symptomatic problems in this young man. So if we know what these are, and if we know that they can be a problem, what do we do about them? Well, one of the difficulties with cavernous malformations, particularly in kids, is that they're very rare. There's relatively limited data on their natural history, their treatment, and their outcome, and as put together in some of the recent American Heart Association guidelines, we really need to have evidence-based data in making decisions in pretty much all that we do in neurosurgery, but especially in vascular lesions. So one of the first questions we get as surgeons, as people who intervene, is who do we treat? And so we tried to address this in a natural history study. There are other excellent studies in that literature as well, but looking specifically at children, some of the questions we ask are, how do these kids present? And in looking at our series, and again, this mirrors work that has been published by other groups around the world, many patients will present incidentally, but one of the most common presentations are with seizure, either with or without hemorrhage, and this is particularly obviously in cortical-based lesions, and another is with hemorrhage, which can present again with seizure, with headache, or with focal neurologic deficits. So in kids, really there's a very high number that present with hemorrhage, and this hemorrhage can either cause focal neurologic deficits, headache, or seizure. The other thing that people ask a lot about is, you know, if you have a cavernous malformation that presents, what is their location? Where are they? And at least in symptomatic cases in children, the vast majority are supratentorial lobar lesions. Now, this may be in part simply because the supratentorial part of the brain is the greatest volume of neurologic tissue in the human body, so just by sheer distribution of mass, this may be part of it. About 7% will be present as deep basal ganglia or thalamic lesions, and then the remainder are either cerebellar brainstem or spinal. How often do we see new cavernous malformations? Basically, the sort of development of new cavernous malformations in a patient that's previously scanned is very low, somewhere in the order of less than 1% per year per lesion. This is obviously different in the setting of patients with familial cavernous malformations, which have a much higher rate of either new or newly identified lesions, meaning that some of these lesions may have been there, but just been below the level of detection. Now one of the very big questions that at least I get commonly in my practice and certainly is an important answer is, what makes them bleed? So they'll say, hey doc, my kid has a cavernous malformation, maybe they are not an appropriate candidate for surgery, and it's being followed, and people will ask, well gee, should I limit the activity of my child? Are there things that I should or should not be doing to reduce or increase the risk of bleeding? There's a very nice study done by Dr. Froelichen at UCSF, a neurologist, looking at a population-based study, and basically it seems that a trauma probably does not significantly increase the rate of bleeding, although my practice has been to limit high-impact, high-frequency contact sports such as American football or boxing. Infection probably does not increase the rate of bleeding. Blood thinning agents and their use are unknown in their risk of causing bleeding. Probably low-dose aspirin is okay if there's a known reason to have it for other medical conditions, but in general I do advise against anticoagulation unless it's needed. Things like ibuprofen, Tylenol, which does not have a bleeding risk, I really am much more liberal. But I think that generally I have very limited types of activities that I will restrict in kids, so I let them do almost everything other than the exceptions as noted above. The other question that comes up is how often do these cavernous malformations bleed? Incidentally, the sort of incidental lesion that we typically see is a relatively low rate at about half percent per year, so in any given year the risk of bleeding is very low, but obviously in a child this is important because this risk for bleeding will be cumulative and increase over time. If the lesion has already bled, that number is substantially higher with a re-bleed rate in the first three years at about 18 percent, and then after the first three years that drops down if it hasn't bled to about 5 percent. So the point here is that hemorrhage is a really important prognostic finding for lesions that have identification in children. If you look at this, basically what I tell families is that the takeaway is that if you have a cavernous malformation, the re-bleeding rate is much higher in the first five years, and this really informs your follow-up strategy. If you look at this and you say, well, hey doc, we've had a cav mal or I have a patient with cavernous malformations, how often should I scan them? How long should I scan them? And certainly this is open to debate and question, but we've tried to use the data as outlined here in this graphic to say that the re-bleeding rate is highest in the first five years, and our practice is to get a scan once a year for five years, barring any new symptoms or problems, and then after that we graduate them to once every few years, and I think this is a practice mirrored at several other institutions around the country. So this is how I follow the patients based on the data. So the last question, if you say, okay, we have a sense of what cavernous malformations are, we have a sense of how they present, some of their genetics, what should I or should not restrict, and what is their bleeding rates? One other question is what about specific locations? You know, if we have a cavernous malformation in certain parts of the brain, which are ones that are riskier, which ones we should treat, how do they do? So one of the less common but more concerning ones are cavernous malformations of the basal ganglia in children. I know that Dr. Steinberg and Dr. Lawton are going to speak more about deep lesions in adults and surgical approaches, so I will limit my comments here essentially to those in children and what to do, but basically the take-home message here for basal ganglia lesions in kids is that if it's large, two and a half centimeters or so or more, and it's symptomatic, it's probably wise to try to surgically remove it if possible. If it's recurrently bleeding, again this is an indication for surgical resection, but if it's very small, several millimeters in size for example, if it's not bleeding and it's not symptomatic, it's usually appropriate to go ahead and leave it alone and scan it with the data as I mentioned previously. Obviously one of the other locations, sort of the most common location, are supratentorial cavernous malformations in kids. This is really the vast bulk of patients that we see in the pediatric population. The most common presentation, other than headache, would be those that present with seizures and one of the questions is, hey doc, if you take these out, what will that do for my seizure cure or my reduction in seizure? And what we found is that with surgical resection of the cavernous malformation, with or without additional resection of hemocitarin around the edge, which is important if you're in an eloquent part of the brain, essentially you have a very high rate of becoming seizure-free after surgical resection. So not only are these appropriate to remove based on symptomatology or hemorrhage or headache, but quite frankly there is an excellent chance that you will either cure or markedly reduce their seizure frequency with resection. So this is something that we highly recommend. Basically for all comers, you have in the high 90s percentage of markedly reducing or curing their seizures in the pediatric population with lobar cavernomas. So as such sort of our takeaway here to steal a phrase from Nike is that these are appropriate surgical candidates and you should just do it. So with that in mind I think that leads to the question that we're all surgeons with that idea if we have a sense of what the disease is and what surgical indications might be one of the questions that tends to get asked is okay if we know who we want to treat and we know why we want to treat them how do we treat them and how can we treat them better what are some of the advances technically we can make in the pediatric world with surgery. So with surgical advances you know one of the most important things is getting where we want to go safely. Now there will be again a much more detailed discussion with Dr. Lawton and Dr. Steinberg about how to get to difficult locations for example the brain stem. I will let them really opine and give their expertise on how to address this but suffice to say that these are surgically accessible lesions and there are ways to get to even some of the most difficult parts of the brain and one of the great advantages of cavernous malformations is they often create a very nice plane around themselves with respect to the surrounding tissue so if you can get to them they generally remove very well and this is just one of our papers from years ago showing one of the potential routes for getting these out of the brain stem. One of the other things that is particularly important in kids is being less invasive and this is sort of a surgical technique that I credit to Mike Scott and other members at our hospital about approaching deep-seated lesions for example cavernous malformations. We have really become very big advocates of using ultrasound on top of frameless stereotaxy. You can see here an example of a frameless stereotactic device, a heads-up display on the microscope showing where lesions are and for all intents and purposes these are great but why would we advocate using ultrasound or a catheter when we have these cool 3D heads-up displays and other tools? Well I think all of us as surgeons can recognize and appreciate that deep-seated brain lesions like the basal ganglia lesions with cavernous malformations can be difficult to approach. Using ultrasound as real-time imaging and to sort of coin a phrase here a little bit real-time imaging beats old time meaning that even with frameless stereotaxy there may be brain shift, this may impair accuracy and smaller kids where the volume of the brain is relatively high relative to the body the shift may be more pronounced. The other reason is that it may save time in using frameless stereotaxy. Occasionally if you use a probe you have to repeatedly determine where you are you have to put the probe in and out sometimes the imaging for the microscope can be somewhat difficult to use. Dissection particularly in deep lesions and kids can be complicated by the absence of identifiable landmarks deep to the cortisectomy so white matter tracts all look the same once you're under the microscope and while there are many different tools that can be used for this basically the idea of the ultrasound and a catheter combines intraoperative data with what we think is ease of use. How do we do this? Well basically it's pretty straightforward we mix sort of high-tech ultrasound and imaging with simple approaches. Essentially we make the small cortisectomy using ultrasound we will pass a regular ventricular catheter or probe to the edge of the lesion trying to take care not to puncture the lesion like a toothpick and an olive to promote any bleeding or problems and then it allows us direct access sort of very quickly going down along the catheter. This is just another example of how we use the ultrasound with real-time imaging. Obviously there are many ways to skin a cat here there are the various tubular retractors which can be used in a similar fashion to a simple ventricular catheter. People can use stereotaxy with a combined probe. Again we like the ultrasound because it is real-time imaging that accounts for brain shift unlike previous stereotactic imaging that is older or pre-surgery if you will and again sort of dealer's choice in terms of what sort of tools to use. The tubular retractors are very nice but the idea of the catheter is that it's available to pretty much everybody in any operating room. This is a sample case. While this is not a cavernous malformation this is an example of a deep-seated lesion in the ventricle. What we do is make a small cortisectomy at the time of surgery as you can see here then using the ultrasound you can really see where the catheter is and where the lesion is and then pass a catheter under real-time imaging. The catheter goes directly to the lesion and then we make a small cortisectomy. The point here is that while you're under the microscope you can just directly move along the catheter and get right to the lesion. This really is quite short and makes a tiny opening. In our little series of patients we use just to write this up as a surgical technique. The advantages of this approach for cavernous malformations in kids is it's real-time, it's easily applicable, it's very accurate, there are no limiting walls or barriers like you might have with the tubular dissectors and I think what's really important is that this is available pretty much at any hospital. All you really need is an ultrasound and a catheter without a lot of fancy equipment. Clearly this is not great for things like AVMs where you don't want to inadvertently puncture something and cause high-flow bleeding. The catheter can be dislodged if you're not careful and it's a straight line of attack so you can't move around corners and it's obviously applicable to lesions that's only visible with ultrasound so if you had an invisible or very tiny lesion this may not use. The takeaway really here is it's a simple tool. I mentioned it as part of this conversation with pediatric cavernous malformations because these tend to be the ones that get referred to academic centers. The simple low bar ones are easy to take out and therefore might not get referred but the deeper ones are ones that certainly we tend to see at our the academic centers. They can be customized and it's really a parsimony of surgical technique. So with these surgical advances what are other ways we can advance them and apply them to cavernous malformations in kids? We've used the 3D printing quite a bit which seems to be very helpful. Essentially we do the surgery before the surgery and this is fairly widespread at our institution. Essentially we just submit a request. We have films in the system. This really is about a 24-hour turnaround for most of our cases unless they're very complicated and it can produce 3D prints in many ways. This is an example of how this might work in an arteriovenous malformation but I think it highlights the case. Essentially you see an arteriovenous malformation here which is in the occipital lobe. The model really shows the anatomy along with the draining veins. For cavernous malformation this might help you understand where the developmental venous anomaly, a normal vein that's around it, might be. With the operative approach you really have the model to understand the size, the location, where the relevant anatomy might be. This is a view looking down. You can see the arterialized vein to the lower left and the embolized artery in the gray. It really matched very well and again for many lesions, tumors, AVMs, and in this case cavernous malformations, this can be a nice tool particularly in teaching institutions so that residents and trainees can get comfortable and allows for a very nice resection. These types of models again we've published it has a very high concordance between the model and the operative field so 98% concordance. There's less than two millimeter variation between the model and the lesion so it makes it very accurate and the advantage of this separate from the GWIS factor or from a teaching factor for those of you who have residents and fellows is that we actually showed that there's about a 12% reduction in operating room time which is a big financial advantage to the hospital and frankly better for the patient. So this is something to think about and cavernous malformations in kids where children under anesthesia and blood loss are very significant factors really can be helpful. So to wrap up here you know other than the technical advances in the operating room that are relevant to cavernous malformations, how else can we advance the care of patients with cavernous malformations? You know can we think outside the box? Can we do something different? And what I'd like to end with here are just some new ideas that are relevant. We talked about the genetics of these lesions, we talked about the indications for surgery, the bleeding rates. There are a lot of advances that hopefully will put us as surgeons out of business in the near future. One of these includes drug therapies. So obviously the rokinase inhibitors, I mentioned how the genetics of this lesion as we're increasingly becoming smarter and understanding that almost 90% of the familial cases involve some of these CCM complexes that somehow seem to involve rokinase. Getting inhibitors for this may be very helpful. Modulating inflammation, we talked a little bit about the microbiome and you know essentially really regulating in perhaps the use of antibiotics, the use of different types of microbiome simulators may affect inflammation which may affect bleeding rates. There has been studies looking at beta blockers as modulators of how these lesions may or may not bleed. Angiogenesis looking at thrombospondin and currently there's a trial again not for children but for 18 year old patients and up looking at a statin in the role of treatment of hemorrhagic cavernous malformations. And so I think there's a lot of exciting work that is ongoing currently for drug therapies which is based on the molecular understanding of these lesions as we talked about at the beginning of this conversation. This is clearly an area that is very dynamic and while there are no specific full treatments medically now and to be clear the treatments now for these lesions are surgical or observation in pretty much all cases. The hope is that drug therapies are something that are on the horizon and for patients with familial cavernous malformations is something exciting for them down the road. The other obviously interesting thing is the role of the microbiome. This is really something that in pediatric populations is certainly something I get asked about a lot. Clearly this is a field in evolution and is not without a great deal of controversy but nonetheless it is something that certainly is in the conversation and I think that going forward it will be necessary to flesh out these studies in a greater degree so that when patients ask we can have the appropriate answers about what the role is of the microbiome, how infections or antibiotics or diet may affect that and one of the other ones is the role of vitamin D and how this may be affected by absorption across the gut, dietary changes and whether that affects bleeding rates or not. So I think there's a lot of potential molecular biology and drug development that are ongoing in this field. This is particularly important in pediatrics because these are children that are going to live with these lesions for a very long time and lesions that may currently be either inappropriate for surgery or may not warrant surgery are lesions that later in life as these children grow up may transition from being surgically treatable lesions to lesions that may be treatable with medications and I think this is one of the more exciting things about pediatric cavernous malformations. So what can we say in conclusion here? I think that cavernous malformations do represent a significant health care burden for those patients that are affected and their families not just for the sporadic lesions but obviously for the familial lesions. Improving the outcomes for children will require advances from multiple disciplines innovating across traditional lines of study. It is very easy to take out a single low-bar cavernous malformation and non eloquent cortex and that alone is not that much of a challenge for most neurosurgeons but deeper seated lesions, the multiple lesions, the lesions that recurrently bleed in sensitive locations, these are very significant challenges and really push the boundaries of what we can do in neurosurgery. As such, research into these fields are very important. Stratifying risk through these population-based studies coupled with developing technologies like genetic testing to understand the underpinnings of these diseases, biomarkers that might predict which lesions are more likely to bleed or not, will really help us to clarify the best treatment plan for every individual patient and again this is particularly important in kids. Children with cavernous malformations will benefit from novel approaches to treatment including new operative techniques, some of which we outlined here, some of which Dr. Steinberg and Dr. Lawton will discuss in the companion talks here. Advances in molecular biology and I do think that cavernous malformations do represent some of the great successes in neurosurgery in really understanding the molecular underpinnings of the diseases that we treat and ultimately using this understanding in molecular biology and genetics to develop unique methods of treatment and for the surgeons unique methods of visualization and training such as 3D printing and simulation. I thank you very much for your attention and I hope this has been informative for your education here. Thank you so much.

pediatric cavernous malformations

genetics

symptoms

treatment options

surgical resection

real-time ultrasound imaging