Slides
Transcript
Insights into Lateral Leaks — Dr. Ajay Madhavan
Moderator: Right, welcome back everyone. It is my distinct pleasure to introduce our next two speakers who will be discussing new research related to lateral dural tears, a type of spontaneous spinal CSF leak. We will hear first from Dr. Ajay Madhavan, neuroradiologist from Mayo Clinic Rochester. We will then hear from Dr. Wouter Schievink, professor of neurosurgery and director of the microvascular neurosurgery program at Cedars-Sinai Hospital. My name is Kyle Spawn, and I will be the patient moderator for this session. I’m a current volunteer on the Board of Directors for the Spinal CSF Leak Foundation and a patient with a repaired iatrogenic spinal CSF leak found above the shoulder of a lumbar nerve root. While I am not a spontaneous lateral tear patient, these talks will be of particular interest to me, and I know they will be of interest to you. With that, it is my distinct pleasure to welcome Dr. Madhavan to the screen.
Dr. Madhavan: All right. Thanks so much, Kyle. I’m just gonna go ahead and share my screen here. Like Kyle said, I’m Ajay Madhavan. I’m a professor of radiology, and I have a lot of interest in spinal CSF leaks, and I’m going to be talking about new insights into lateral leaks.
I don’t have any disclosures. I do want to thank Dr. Callen for inviting me here to speak, the Spinal CSF Leak Foundation, and all the patient speakers who are here today. It’s really an honor to be able to talk to you all.
So I’ll start just by revisiting the different types of spinal CSF leaks. I think this is really important. If you read about spinal CSF leaks, there are a lot of classification schemes out there for how people view these different types of leaks. It doesn’t really matter which one you use, but I do think it’s important that we all have sort of a common understanding of what the different pathologies at play are, because I really think I’ve come to think of them as different diseases that all lead to the same common pathway of spinal fluid loss.
The first or kind of classic type of spinal CSF leak is a ventral dural tear. So that’s a spontaneous hole in the front of the dura that usually happens because of a little bone spur that comes from the vertebral column, and it just pokes a hole in the dura. If you do a Spine MRI in these patients, you’ll almost always see this big amount of extradural CSF. You see spinal fluid outside of the space where it’s supposed to be, and that can manifest either as a collection, so many people call that a SLEC, or it can be more just diffused free-floating fluid, and it kind of depends on whether it’s more of an acute or a chronic case. These patients usually have pretty normal looking dura, at least on imaging. They don’t necessarily have a lot of meningeal diverticula, and that’s in contradistinction to patients with CSF-venous fistulas that Dr. Callen just talked about. These patients usually have many meningeal diverticula. We don’t always know what causes these fistulas, but there’s some very good papers out there, one from Dr. Mamlouk that talks about how many of these patients have an elevated BMI, and they may have initially had intracranial hypertension before they developed a leak, and another one from Dr. Callen and his group that shows that some of these fistulas tend to occur at the same level as prominent spinal degenerative changes.
And then this is another great paper from Dr. Mamlouk as well that discussed a very different type of spontaneous leak caused by vascular malformations. So it turns out if you get vascular malformations near the spinal canal, they can sort of elicit a spinal CSF leak, and this was one example from this paper of a leak that was going into a lymphatic malformation. These two images really nicely show contrast leaking out, and this is not into a vein. This is into the thoracic duct, which is sort of the main lymphatic drainage pathway. So there are these more rare and unusual types of leaks too.
And then there’s lateral dural tears, which is what I’m going to be talking about. These are probably less common than ventral dural tears and CSF-venous fistulas. They’re more common than leaks into vascular malformations. They tend to happen in patients of a younger age, but they can happen in anyone, and they’re a little bit more common in women compared to men.
This was a nice table from one of Dr. Schievink’s recent papers showing some of the demographic features in these patients. I won’t go over this in all that in detail, but you can see that they happen in younger patients in particular.
And then the other really important thing about these leaks that I really want to highlight is that lateral dural tears are usually associated with this interesting finding of an arachnoid outpouching or herniation. This is actually the initial description of that phenomenon from many years ago. Dr. Beck and his group published this paper, and this is a really beautiful illustration from them, and I like this because it shows exactly what’s happening.
In these patients you typically get a lateral dural tear somewhere near the nerve root, and then you get this arachnoid, which is the layer just underneath the dura, which sort of herniates out of the dura, and it looks just like a diverticulum. But what’s interesting is that the spinal fluid is not leaking from this diverticulum. It’s almost always leaking from the margin of the dural tear. So it’s a little bit more medial, and this wasn’t really fully recognized until pretty recently. And so it’s not like a ruptured diverticulum. It’s really a leak from the lateral dura, and this is more of a secondary manifestation.
This is an example of what this sort of arachnoid herniation might have looked like on a conventional myelogram, something that we would have done many years ago. So this is a patient with a lateral leak. They have this big epidural CSF collection, so you can see contrast outside in the epidural space. But the finding that’s a little bit more subtle is that there’s also this arachnoid diverticulum here. It’s a little bit hard to see, a little bit easier with the arrows, but that’s a secondary finding that can go unrecognized.
This is what that looks like at surgery. This is from Dr. Schievink. And so this is a patient with a lateral dural tear. This is where the nerve root is right here, and then just under the nerve root you see this big arachnoid herniation. But the key thing is you really have to retract that herniation. You pull it off to the side, and then you’re able to visualize the dural defect.
Now this is another example of the same thing from Dr. Beck. And so in this case, this is the common dural tube here. This is the nerve root coming out, and then this white structure here is the arachnoid diverticulum. So in this case the lateral leak is probably a little bit more along the nerve root. So remember that there’s dura along the nerve root sleeve too, and so the dural tear can occur a little bit more laterally as well.
And then the other interesting thing is Dr. Schievink published this paper recently, and this is a really nice illustration describing the different locations that lateral leaks can occur in. The classic location has always been that it happens under the nerve root. So this is the nerve root here, and then the dural tear is right under the nerve root, and you can see the arachnoid herniation. But it can also happen right above the nerve root sleeve, so it can happen at what’s called the shoulder location, and it can even happen a little bit higher too at the level of the pedicle, which is just one of the bones that connect the vertebrae to the posterior elements. And you can see from Dr. Schievink’s table here the axillary location right under the nerve root sleeve is the most common.
These are just some examples of that on digital subtraction myelography. So this is an example of the most classical type where the leak is happening at the level of the axilla. This is the nerve root sleeve right here, and you can see it’s an epidural contrast leak that’s coming out right under the nerve root sleeve, and over time it accumulates in the epidural space.
This is the shoulder variant. In this case, here’s the nerve root sleeve right here, and the leak is right above that. It’s at the shoulder location. These are three examples that Dr. Schievink gave me. This is the rarest location for lateral leaks, but they can also happen at the level of the pedicle. In this case, this is one of the nerve root sleeves. This is another nerve root sleeve, and the leak is right halfway between those, which is where the pedicle is. You can see they all look the same. These are two other examples of similar types of pedicle leaks.
That’s the introduction, but going from there I just want to show some cases, because I think how I have come to understand this is by doing myelograms on patients and looking at the images, and I think that gives you a good appreciation for the anatomy.
This would be a very classic example of a lateral dural tear. This was a patient I had about a year ago. This is a cone beam CT. That’s something we usually do right after a digital subtraction myelogram, and it gives us more high-resolution cross-sectional images. You’re looking at a coronal view, so you’re looking at the patient from the front. What you can see here is that this is intact dura here. This is intact dura here. And there’s a dural tear right here. It’s this little hole, and this big membrane here is the arachnoid herniation that’s protruding out of the dural defect. Right above that you can see the nerve root coming down and exiting the thecal sac right here. This is the nerve root sleeve. And then you can see that there’s a lot of contrast that has exited out into the epidural space by this point. That is a typical look for what a lateral dural tear would look like.
But not all cases are quite that straightforward. This was a more challenging case I had recently. This is a young woman with some very subtle epidural fluid on her spine MRI. This is a sagittal T2-weighted image. The spinal fluid is bright, and you can see the arrows pointing to very small amounts of epidural CSF there in the dorsal epidural space. A lot of times we like to do high-resolution T2-weighted images, and this is a coronal image. You can appreciate the epidural CSF a little bit better there. But I’ll tell you, even with these images, it’s not entirely obvious that there’s something abnormal there. So these can be hard to detect.
In this patient, we opted to do a digital subtraction myelogram. This is a right decubitus digital subtraction myelogram. You don’t have to do a DSM, you can certainly do a CTM. I think both DSM and CTM are good techniques to find these.
And so as I scroll through here, you can see the first thing is that there’s this diverticulum that’s filling up with contrast. As I continue to go, eventually you start to see the subtle accumulation of epidural contrast that’s leaking out both above and below that diverticulum. That’s a typical look here. The unsubtracted images look the same. It’s a little bit harder to see without subtraction, but that would be a standard look for a lateral dural tear on a DSM.
In this case, we did a cone beam CT to get a little better characterization as well, and I think that shows the anatomy well. So these are the spinal nerves in the thecal sac. And as I scroll down here, you can see this is normal intact dura here. Coming down, you can see the two nerves coming down as well. And this is right where the nerves exit the thecal sac. Here you can see intact dura here, intact dura here, and then right next to the nerve root sleeve there’s this very subtle little dural hole here. So this is the lateral dural tear.
And then you might ask, well, where’s the arachnoid herniation? As I continue to go down here, it’s right here. A lot of times on CT, these arachnoid herniations will fill less with contrast than all the other structures. That’s how you know that’s what the arachnoid herniation is. You can see it a little bit better on this coronal reconstruction. This is again looking at the patient from the front. This thing right here is the arachnoid herniation. The point is they can be a little bit hard to see. This is that same patient, just another view of that arachnoid herniation.
The teaching point of that case is that these arachnoid herniations can be very subtle on imaging. I think probably as I’ve seen more and more of these, they’re likely underrecognized. I don’t know that they happen in every patient with a lateral dural tear, but it may be more than we realize.
Here’s another challenging case. This was another patient with a lateral dural tear. You can see on their spine MRI, this was done without fat suppression, which, if you ever get a spine MRI, I’d encourage your provider to use fat suppression if they have it available. It just improves the sensitivity of finding epidural CSF collections. On this MRI, I couldn’t really appreciate any obvious epidural CSF. There might be some here. It’s a little bit hard to tell, but what I’ll draw your attention to is at this level, at T11-T12, there’s this unusual-looking diverticulum here. And the thing that’s unusual about it is it sort of has this broad base with the dura. That’s not how most diverticula look. Normally they’re a little bit more rounded.
On this patient’s CT myelogram, she also had a regular conventional CT myelogram done on the outside. And you don’t really see anything too obviously abnormal there. There’s no big epidural collection or anything like that. You can again see that unusual-looking diverticulum at T11-T12, which sort of has this broad base with the dura. And so anytime I see something that looks a little bit abnormal at one specific level, you wonder if that’s where the leak is going to end up being.
So we did a decubitus digital subtraction myelogram on this patient. This is T11-T12 here at that level where the unusual diverticulum was. And so initially you don’t see anything. On this first DSM run, I imaged this patient for over a minute, and I used a good amount of contrast, and I didn’t see anything obviously abnormal. I didn’t see any leak into the epidural space. Again, you just see this diverticulum. You maybe wonder if there’s some hazy stuff right here.
But then I repeated another imaging acquisition. We injected more contrast and continued imaging at that same level. And what I’ll show you here at that same T11-T12 level now, you can pretty quickly see that there’s contrast leaking into the epidural space. You can see it both above and below the nerve root sleeve. And then over time it accumulates more. And this is just the spot image about 30 seconds after that. And now you can see that there’s a lot more epidural contrast out there. But this is a very clear lateral leak. When you do a cone beam CT, you can kind of get a better sense of what’s going on. That’s why I like getting these with DSM. So again, you see that there’s a lateral dural tear here at that T11–T12 level. But now you get a better sense that there’s this more unusual structure here that’s not really showing with contrast, so you wonder what that is.
If you look at the coronal image, you get a better sense of things. This is the unusual structure. This is like the normal nerve root sleeve. And in this case, this ends up being the arachnoid diverticulum. The dural tear is probably somewhere right around here. I show this case to highlight the different anatomy, right? This is the normal nerve root sleeve. This is a normal nerve root sleeve diverticulum. You can see this patient has additional diverticula at the levels above and below that. But this is a clearly distinct thing. This is the arachnoid herniation. It’s different, and it looks different from a normal nerve root sleeve diverticulum. This is what it looks like on a sagittal image.
The teaching points from that case are many things. One, these arachnoid herniations can be pretty big. You can see in that last case it was fairly large. And I would speculate that when they get that big, they can probably cause these lateral leaks to slow down a little bit. When I first learned about dural tears, they were explained to me as always being fast leaks, meaning that some people have used those two terms synonymously, but that’s not always the case. Sometimes dural tears can be very slow, and I think this is one explanation for why lateral dural leaks in particular can be a little bit slower.
I’ve learned that when I do myelography on these patients, sometimes I have to slow down a little bit and get more delayed imaging and perhaps use a little bit more contrast as well. The other thing is that in these patients it’s possible that epidural spinal fluid collections might be absent on MRI or very subtle, and that again might be because that arachnoid herniation is essentially blocking spinal fluid from constantly leaking out into the epidural space. And then the last thing is that broad-based diverticulum I showed you. I think that’s an important feature that can sometimes help us predict that there’s a lateral leak there.
This was a really nice paper from Dr. Lützen in Germany and some of his colleagues, and they looked at the timing of how long it took for these lateral leaks to show up on DSM, and it was pretty variable. But one interesting finding was that some of the slower leaks tended to have larger arachnoid herniations. So I think there may be something potentially real there and helpful in the future.
Here’s another case of a lateral leak. This one’s a little bit different. In this case, you can see contrast coming up, and you pretty much immediately see contrast leaking into the epidural space. This is a much faster leak. But the interesting finding in this patient is that there’s also this unusual structure here that’s filling up with contrast. And if you do any kind of spinal work or spinal imaging, you might recognize that that is what the facet joint looks like. So the facet joints are these joints that connect the vertebrae.
This is what that patient’s cone beam looked like. And what I’ll point out here: again, you can see the lateral leak. You can see the arachnoid herniation there. Notice that the arachnoid herniation is kind of remodeling the bone. That’s probably been there for a while.
And then as I continue to scroll down here, you can even see some contrast going into the facet joint. That corresponds to that finding on the DSM. Sometimes when these lateral leaks have been there for a while, they can erode into the facet joint and cause a direct leak into the facet joint. This is the first time I’ve ever seen that, but since this case I’ve seen it a few times, so it’s possible I was just not noticing it.
This is another case from Dr. Lützen showing the same thing. This is the facet joint right here. This patient had a lateral leak, and you can really beautifully see the contrast going into the facet joint on the cone beam CT that he did afterwards. And here’s that arachnoid herniation that’s remodeling the bone.
So why could this be clinically useful? This is one of those patient’s spine MRIs, and if you look back at their spine MRI, you can actually see fluid in the facet joint. People can get fluid in the facet joint for all kinds of reasons. Sometimes it’s just from arthritis, but I think if you see a localized effusion of fluid into the facet joint, that might help us predict where a spinal fluid leak is going to be. I think that’s something I’m interested to look at in the future.
So the teaching point there is that these lateral leaks can cause arachnoid herniations which then go on to erode into the facet joints. And you might ask why that is clinically useful? I think it probably indicates a chronic leak. It might help us detect leaks better on MRI. And the last thing here too is that I think that could facilitate blood patching for these patients.
This was a paper where we wrote up some of these patients. My excellent colleague Michelle Kodet and several of my collaborators from other institutions helped with this. In one of those patients we actually tried to inject contrast directly into the facet joint. You can see here this is the normal facet joint. This is how it should normally fill, but there’s no contrast that’s supposed to go beyond this margin. It’s supposed to stop there. But in this patient who had a lateral leak that was eroding into the facet joint, the facet joint actually communicated with the epidural space. That was helpful for my purposes in this case because I injected fibrin and blood directly into the joint, and that gave me an easy way to access the epidural space that was very safe and helped this patient.
Okay, so just a couple more cases to go here. This is another patient with a lateral leak. This is one of the earliest patients I had. And you can see she has this dorsal epidural fluid collection. On the axial images, you can probably appreciate already that there’s going to be an arachnoid herniation here.
So we did a decubitus myelogram looking for a lateral leak, and she did indeed have a lateral leak at that level. But the other finding you can see here is that there’s a CSF-venous fistula which seems to be coming directly from the arachnoid herniation. That’s a little bit unusual. This patient had two different types of leaks that were both visible on the same myelogram.
This is another patient from Dr. Lützen who had a very similar finding. This patient had a diffuse epidural CSF collection on their MRI. On a digital subtraction myelogram, you can see they had a very prominent lateral leak. This is an arachnoid herniation right here. But the other finding that was observed here is on a delayed CT myelogram. There was also a CSF-venous fistula that seemed to be coming from that arachnoid herniation.
This is another interesting finding that we’ve seen a few times. Sometimes with these lateral leaks you get an associated CSF-venous fistula, and that’s the teaching point from those cases. And that’s very interesting, and it might affect how we treat this in the future.
I do think in most of these patients the primary problem is the lateral dural tear, but it’s certainly possible that the CSF-venous fistula might exacerbate things, and it’s something that I continue to look for.
Okay. And then the last thing, I won’t go into this too much because Wouter Schievink is probably going to touch on it quite a bit, but treatment options for lateral dural tears – how do we treat these? The easiest way is a non-targeted blood patch, which might work early on. As with most CSF leaks, it won’t work if the leak has progressed too far along the line, and then you probably need a more targeted approach, and that might include surgery. It might include a targeted blood patch or fibrin injection.
I think Dr. Callen and many of the other folks on this paper, Dr. Carlton Jones and many others that I mentioned, have done some really good work in looking into how we treat these lateral leaks. It turns out that having arachnoid herniations that come out of the dural defect probably makes it a little bit harder to treat these with fibrin and blood. One thing they observed in this study, where they looked at a lot of patients who were treated with blood patching, was that when you could see a big arachnoid pouch or herniation on imaging, it made it such that the leak was less amenable or less likely to resolve with targeted blood patching or fibrin injection.
To summarize, these lateral dural tears can occur in multiple locations. I think we’ve learned a lot more about what they look like on imaging in recent years. They can happen at the axilla, but also at the shoulder of the nerve root sleeve, and they can even happen at the level of a pedicle, as I’ve shown. Many times they’re associated with these herniated arachnoid diverticula or outpouchings, which is a very important feature of these leaks. And I think importantly, MRI does not always show an epidural fluid collection. I showed a couple cases like that, and I think that that’s something that could cause lateral leaks to be underrecognized.
These are not always fast leaks. I showed the one case where the leak was a little bit slower, and I’ve seen that a number of times, and so I think it’s very important if you’re getting a myelogram that early and delayed imaging be done. And then there’s some interesting features that accompany these lateral dural tears, including leaks into the facet joint and CSF-venous fistulas. I think that’s something we need to learn more about, see how they affect the symptomatology in these patients.
I will stop there. Thanks so much for your attention, and if you have any questions, I’m happy to answer them.
Moderator: Thank you very much, Dr. Madhavan, for your interesting talk. We will now turn to Dr. Wouter Schievink for his talk from the surgical side of treating lateral dural tears.
Insights into Lateral Leaks — Dr. Wouter Schievink
So the real title of my talk is Is That a Diverticulum? New Insights in Lateral Leaks Part 1: Pathophysiology and Anatomy From a Neurosurgical Perspective.
So that’s a picture of a digital subtraction myelogram, which is really kind of the workhorse of neuroradiology here at Cedars-Sinai, and this is kind of what I thought was the clinical scenario that we are dealt with on a really frequent basis. So on the left side you see an MRI scan of the brain, subdural thickening, hygromas, or the middle MRI looks pretty normal except the optic nerve sheath diameter is a little diminished, or the scenario on the right side where a really good doctor like Ian Carroll will call you and say, you know what, I have a patient that has normal imaging but a really good story for a leak. That’s the clinical scenario. So either abnormal brain imaging or a really good history and maybe physical for a CSF leak.
And a lot of those patients undergo many, many different radiographic imaging problems. And why do I say problems? We’ll talk about it a little bit later. Digital myelograms, digital myelograms with anesthesia, dynamic CT myelograms, cone beam CT myelograms, or photon counting CT myelograms. And let’s say all of those tests are normal, right? So what are you going to do?
So then, number one, you say, well, maybe it’s not a leak, of course, but there are not that many doctors who can help the patient or can help you. So maybe the leak is not in the spine. Less than one out of a thousand people who have this have a leak at the level of the skull base, so you can look into that.
Maybe it’s pooling of CSF, right? So particularly patients with Marfan syndrome, Loeys-Dietz syndrome, neurofibromatosis. Maybe it’s not that they have a leak, they just have these very baggy collections of CSF possible. And then there are the really unusual cases. Is it azygos vein stenosis, IVC stenosis? Is there a history of a craniectomy and an inadequate cranioplasty? And you can look into that. Or maybe is there what we, and with we I mean me and my colleagues in neuroradiology, call small innocuous-looking lateral dural CSF collections. That’s a bit of a mouthful. So what did Dr. Callen call it? Is that a diverticulum? So that’s what I’d like to talk about. So is that a diverticulum?
And this is what that looks like on a dynamic CT myelogram on the left, not from our institution, somewhere else, or what we do a lot of, a digital subtraction myelogram.
And I’ll be talking mostly about two or three different papers. One of them already has been published, that’s this one, talking about lateral leaks. Dr. Ajay has talked about it already. Most of the time lateral leaks are at the level of the axilla. Axilla is right below where the nerve root comes out of the spinal cord, or it can be a little bit above where the nerve root comes out of the spinal cord at the level of the shoulder of the spinal nerve root, or what we call the pedicular type of lateral leak, which is where it’s not related to a nerve root. Pretty uncommon but not that rare.
This is an example of these pedicular leaks. In our experience a lot of these pedicular leaks are really, really tiny types of CSF collections lateral side of the dural sac. When you do an MRI, MRI myelogram, or you do a post-myelogram CT scan, you don’t really see them. On regular imaging these are almost invisible. And what do these look like under the microscope? Why are they invisible? Why is it not like a big extradural CSF collection? It’s because they’re tiny tears on the lateral side of the dural sac, and often times the arachnoid, which is incredibly variable.
When you do a lot of these surgeries, in some patients the arachnoid, which is the layer underneath the dura. Most people, the dura, dura mater, tough mother, your dura mater matters. It’s almost impenetrable, but the arachnoid can be anywhere from a very weak type of spidery, arachnoid, spidery-type membrane to more or less a very well-formed membrane, and it’s difficult for CSF to penetrate that membrane. So that’s what that looks like in this particular patient.
Sometimes that membrane is almost non-existent. This is somebody who has a lateral leak at the level of the pedicle, as you can see there on the picture on the left. It starts off as just a tiny collection, but then it leaks out, and it becomes a lateral, typical extradural collection. Some people call that a SLEC.
We have been operating on some of those patients for the last five, six years, and we wrote that up. It’ll be published in a few months. We looked at 31 patients who have one or sometimes more of these tiny little CSF collections lateral to the dural sac, and we found that the vast majority of those, even though you don’t see the vein, they actually represent CSF-venous fistulas. They’re just not very exuberant. They’re kind of shy. It’s difficult to find them. But some of those patients actually have a lateral tear, and it’s really difficult to differentiate those.
These are eight patients. Two of them have a lateral tear. Six of them have a CSF-venous fistula, and I have asked our neuroradiologist people I’ve worked with for 20 or 25 years to look at those. I’ve also asked other neuroradiologists to look at them, see if they can actually find which one is a lateral tear, which one is a fistula, and it’s essentially impossible for them to figure that out.
These are six of those eight happen to be CSF-venous fistulas, and these are some of these pictures that you can see, what we find in the operating room. Right? What we find in the OR, the veins are very small. They’re generally less than 1 millimeter. They all arise from a tiny little diverticulum.
This is one I operated on a couple of days ago. The arrow on the left, you see there’s a tiny little diverticulum just caudal. 95% of those are caudal, rarely cranial to the nerve root sleeve, but almost always caudal, one or two millimeters. There’s just a little bit of contrast just lateral to the dural sac.
In this particular patient, it happened to be a normal MRI, just no CSF around the optic nerves. And then on the right, you see what it looks like under the microscope. This is a little bit of better projection. You see there’s the diverticulum, and just below the diverticulum you actually see the vein. Right?
But these are incredibly high-resolution images. Right? What I want you to realize is that the resolution on a DSM is less than 0.1 millimeters. And I thought this was a really good illustration about how what we find in the operating room, where we work with microscopes with 20 or 30 times magnification, compared to what we see on a digital subtraction myelogram. Of course, you also can see it on a really good lateral dynamic CT myelogram.
But let’s talk about the ones where we actually see a lateral leak. So the one on the top is a patient who we had followed for a few years, very extensive extradural CSF collection. Then over the years it became smaller, smaller, smaller, ends up with a tiny little focus of contrast lateral to the dural sac. That’s what that looks like. Or the one at the bottom, two figures where you just see maybe 1 or 1.5 millimeters of extradural contrast. The patient has been told for many years all their scans are normal, but then actually in the operating room you see there’s a tiny little tear on the side of the dural sac, and they’re fairly safe, very safe and easy to fix. Just like with ventral leaks, right, you can have a very tiny tear, you can have more of an extensive tear. It’s the same with these lateral leaks.
This is somewhat a good comparison. On the left side, in the very tiny picture you see there with the arrow, there’s a tiny little area of contrast outside of the dural sac. But on the right side, you see there’s an enormous amount of contrast lateral to the dural sac. That happens to be a patient, he’s in his 40s. When he was a little kid, his mom told me, “Oh, he used to have these terrible headaches. He would lie down. He would play puzzles.” This is before iPads. No more leak, I mean no more symptoms of a leak. And then, as time went on, years, decades, he developed weakness in his leg.
You can see on the MRI a pretty big extradural collection. Digital myelogram shows a very, very, very tiny area where there is the leak, and then in the operating room we find very unusual, never seen that before. There’s a tiny little hole right where the radicular artery, most likely the artery of Adamkiewicz, at the left L1 level penetrates the dura from extradural to intradural. It’s very unusual because the artery of Adamkiewicz, of course, is supposed to be all intradural. And that’s what it looks like after we put a little clip on the hole to repair that.
I’ve been looking into some patients who are diagnosed as a congenital arachnoid cyst, but I think they actually are lateral dural tears, especially when their moms or dads tell me, “Oh, you know, when they were young, they used to have these really terrible headaches.” They would say, “Oh, when I’m up and about, my head is on fire, and when I lie down, the headache would dissipate.”
This is another patient who also actually presented to one of my colleagues with lower extremity, bowel, and bladder symptoms from spinal cord deflection. There’s an extradural collection, and on this digital subtraction myelogram not only does it tell you exactly where the communication is between normal CSF space and extradural CSF space, but also in this particular one you also see that in addition to that there’s a little CSF-venous fistula.
What about surgery for lateral tears? You know, Dr. Beck presented these data, 100% success rate. I’m not nearly that good. I think they’re a little bit less successful as far as surgery is concerned, probably somewhere between 90 and 95 percent success rate. But that’s still quite a bit better than the reported rate of one out of three, or what we usually counsel patients with as a 50/50 percent of curing their lateral leak.
And then finally, let me just share a little bit about the research that Cassie Parks, who’s an MD PhD student at Hopkins, has done. This is really fascinating research on lateral leaks that we’ve done for the last seven or eight years, and this is hopefully going to be published soon.
We looked at 42 patients that we have treated here at Cedars. About one out of four of those 42 patients turned out to have a mutation in the Fibrillin-2 gene. Very fascinating. These are three-dimensional models on where of this gene product the mutations are.
Fibrillin-2 gene mutations can cause this really devastating congenital disorder called contractural congenital arachnodactyly. They do not have this. Right. So when these people, usually young people, usually young women in their 20s or 30s, have these lateral leaks, they don’t have any other problems. There’s no other manifestation of their connective tissue disorder. They just have an inborn fragility of the lateral part of their dura, and it’s because the mutation in their gene is in specific areas of the gene itself. And when Cassie Parks and other people at Hopkins developed an infusion model in rats, these are knockout rats — I’m sorry, mice, knockout mice — that lack the fibrillin-2 gene, she found that unlike Marfan knockout mice, where the capacity of the dural membrane is really excessive so they don’t generally actually develop leaks, they just keep on growing their dural sacs, the ones with the fibrillin-2 gene, they actually rupture at a much earlier time period than those who have normal dural membranes.
And then, of course, we always have to think about the aftermath. We see so many people who we think, or who we know, have a leak. They go through not one, two, five, ten invasive myelograms. They’re not any better. We don’t find a leak. Maybe it’s because we have caused the leak. Like in this patient who has undergone all of these treatments, and then finally we find that they’re still having symptoms because of the leak that we have caused with our procedures.
All right. Thank you very much. At the end of our talks, we like to put down our email address so you can call us. But I would like to really thank Dr. Callen for putting this conference together. I’m really, really grateful to him and to the Spinal CSF Leak Foundation.