Dr. Wouter Schievink at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Wouter Schievink, Professor of Neurosurgery at Cedars-Sinai in Los Angeles, CA, presented this talk on “Types of Spontaneous Leaks: Anatomy and Pathogenesis” at the 2023 Cedars-Sinai Intracranial Hypotension Conference on July 8, 2023. The conference was hosted by Cedars-Sinai with generous support from the Spinal CSF Leak Foundation in Kohala Coast, Hawaii.
Slides from the talk
View a PDF of Dr. Schievink’s slides here.
Transcript
Dr. Wouter Schievink on “Types of Spontaneous Leaks: Anatomy and Pathogenesis”:
[00:00:09] What do we really mean by a spontaneous CSF leak at the level of the spine, or does it always have to be at the spine? So the reason that we still call it a spontaneous leak, even though we know that at least one out of three patients will tell us there’s some kind of more or less trivial traumatic precipitating event, is that we really want to differentiate it from a leak that’s caused by let’s say spine surgery or an epidural or after a dural puncture. So that’s why we still call it spontaneous.
[00:00:53] But what exactly is spontaneous? So this is a doctor I saw a few years ago, she lives in Los Angeles. She’s originally from the Ukraine, and in the 70s, so about 40 years before we saw her, she tells me she had a spinal cord tumor removed.
[00:01:11] And then one day to the next, 40 years later, she started having this terrible orthostatic headache. It took her about a year to seek medical attention. She’s a doctor, right? We can all relate to that. And we did some imaging, and it showed, as you can see here, it shows that she has a pseudomeningocele, right? So there was a leak at the time of surgery that was closed either with sutures or whatever they had at the time, or maybe it just closed by itself. But then outside of this pseudomeningocele or bleb coming out of the dural sac, there’s a leak, right?
[00:01:47] You can see that on the MRI. You can see that on the CT myelogram. This digital subtraction myelogram really just shows the pseudomeningocele. And then at surgery, this is what the pseudomeningocele looked like. And there’s a tiny little pinhole that we fixed just with a single suture. So do you call that spontaneous or do you call that traumatic?
[00:02:09] So what does ICHD 3 say? So they say that you cannot diagnose a patient with SIH, with a spontaneous leak, if they have had any procedure or trauma known to be able to cause CSF leakage within the prior month. So they said, if it’s more than a month, then you can call it spontaneous. But I think, you know, most of us would not call that spontaneous.
[00:02:34] So really after that, that patient, we became interested in, you know, how often do we see that? So Dr. Tay, Dr. Angelique Tay, who should be here in the audience, she’s one of our chief residents, started looking at this. So and this will be presented in September at a neurosurgery meeting.
[00:02:50] So we looked at almost 250 patients, new patients—we had never seen them before, so there was less bias—over an 18-month period. And about one out of 30 of those patients were sent to us for a traumatic CSF leak, but it actually turned out to be a spontaneous leak. They all happened to be women.
[00:03:10] They were in their, you know, between the third and sixth decade of life. There, you can see what type of procedures they had done. And I’ll just show you two examples. You know, obviously this is just eight patients, but this was a young lady who has this really bad spondylolisthesis where L5, lumbar five, slips on sacral one.
[00:03:32] She had a lot of pain and she underwent a facet block. Now, usually that doesn’t really help this kind of pain, but that’s what she underwent. And she told me this was done in a surgery center at 3:30 in the afternoon. And literally within three hours, she had a terrible orthostatic headache.
[00:03:49] So she was sent to me by the scoliosis spine pediatric orthopedist at Cedars because her brain MRI had shown some meningeal thickening and enhancement, mostly frontal, and an MRI scan showed extradural CSF, basically from the sacrum up to the cervical spine. And she already had had a single epidural blood patch that didn’t really make her any better.
[00:04:13] So we did some more imaging and then it actually turned out that she had a spontaneous leak at the thoracic four level, a lateral leak, right, those lateral leaks typically occur in younger people. We suspect maybe they have some connective tissue disorder. So she eventually ended up needing surgery to fix that leak.
[00:04:36] And then this is a different patient who we saw I think a year ago, 57-year-old lady who had orthostatic headaches after undergoing lumbar spinal decompression and fusion. And she said it took a few weeks and then she realized she had orthostatic headaches. They had done some blood patches and actually she felt quite a bit better, but only transiently.
[00:04:59] Then she underwent a digital subtraction myelogram somewhere else in the country. And they thought this showed a leak, but we really thought that was an iatrogenic leak. And then the CT at the site of the surgery showed that there’s a bleb. Right. It’s like a little pseudo meningocele, you know, that’s something that that we see not uncommonly in people who undergo spine surgery.
[00:05:23] So I wasn’t really convinced that that’s where her leak was, number one because you know that CT after the myelogram doesn’t really show a leak. And more importantly is that her brain MRI showed quite a bit of brain sagging. And you usually don’t see that with real, you know, post surgical or post, you know, let’s say, iatrogenic leaks.
[00:05:43] So we did a digital subtraction myelogram, and it turns out that she had a mid thoracic CSF venous fistula and her brain MRI, just half a day after surgery already looked a lot better. So I think, you know, obviously you don’t need to image everybody with a post-dural puncture headache, but you know, if they are refractory to treatment, then maybe it’s a good idea to at least think about imaging the remainder of the spine.
[00:06:09] So that’s kind of spontaneous versus traumatic. And then of course, you know, is it the leak in the spine or is it a CSF, rhinorrhea type of, and I can see Dr. Hepworth getting excited here. So, this is, we all know what that looks like. This is a CSF rhinorrhea. Does that ever cause SIH?
[00:06:31] So I’ve never really seen a well documented case, where you see brain sagging, meningeal enhancement, caused by CSF rhinorrhea or any other kind of skull based CSF leak. So like maybe 10 years ago, dr. Maya and I, and some of our collaborators, we published this article where we said, you know, there’s really no association between skull-based CSF leak and SIH.
[00:06:55] And of course we were wrong because then we saw a young patient a few years ago. He was 13, had acquired Chiari, he was diagnosed with that. That was really brain sagging. And then, as you can see on these beautiful digital subtraction myelograms that Dr. Maya performed, is that he was leaking from the posterior fossa through the middle ear out into this collection in his neck.
[00:07:23] So, uh, it’s not only in the spine, but this literally is like, you know, less than one out of a thousand patients who we have seen with SIH. And then, of course, you know, a year after we saw that patient, we saw another 13 year old boy who had also a skull based CSF leak, but this was traumatic one at the level of the lower clivus.
[00:07:46] So let’s go to, what are the different types of leaks that we see in the spine, right? That’s, that’s really where, like I said, 99. 9 percent of, of our patients live. There are three main types. So type one is any type of dural tear. So almost always this is associated with a really extensive extradural CSF collection.
[00:08:09] Some people call it SLEC or something. We usually don’t call it that, but it’s extradural CSF. Or it could be type two. So those are people with either meningeal diverticula or dural ectasia. And then type three is this CSF venous fistula. So type 1A, these are dural tears in front of the spinal cord.
[00:08:32] So we call them ventral leaks. And they’re almost always associated with a little piece of bone. Here, this is an intraoperative ultrasound. This is the spinal cord. And then here you see through the dura, there’s this little spike of bone that we already knew was going to be there because this is what the post digital myelogram CT showed. So this is the spinal cord. This is the vertebral body, and then here’s this little tooth that sticking through the dura that has caused the tear and this is what it looks like at surgery looking from the back. This is the spinal cord up there. This is the dura with little dural vessels.
[00:09:11] And then here’s this sharp piece of bone. So, you know, we’ve known about this since last century, but at least for myself, I thought that was something that was pretty rare up until like 15 years ago, where it really became apparent to all of us that this is actually really common. Probably over 99 percent of people with ventral leaks have some sort of bony abnormality. So I think it’s important to classify these leaks, because even these leaks in front of the spinal cord that are associated with a bone spur—and it can be pretty big bone spur—if these people come in through your ER, you can cure them just with doing regular blood patches and then they seal around this piece of bone, but you have to do it early, right?
[00:09:54] So I think that’s really important. If it’s been there for a couple of months, then I think, you know, if, if the interventional neuroradiologist is comfortable with doing ventral patching, then I think that’s a really good idea. But at least in, in our institution, if people have had it for more than three or six months, we just take them directly to surgery.
[00:10:13] So the other type of dural tear is what we call the type 1B tear, or a lateral leak. Sometimes it’s posterior, but usually it’s a lateral leak and you can see some examples there. So it can have the appearance of a cyst. And I used to call this leaking cysts, but the primary pathology is really the dural tear that you can see here.
[00:10:36] This is the common thecal sac. This is a suction device. Obviously this is high magnification. And then you see, this is the tear. And then this is just arachnoid billowing out through the tear. And usually it’s not leaking through the arachnoid, but it’s leaking from the sides of this arachnoid
[00:10:56] that’s billowing out. So even if people have had this type of leak for, let’s say, a decade, I think there’s still a role for non surgical treatment. I think if it’s been there for a long time, you know, your body has formed connective tissue membranes around it. So I think just doing a regular blood patch probably won’t work on it unless you get to it early on.
[00:11:16] But then you can do either, you know, a direct puncture at the level of the tear. And you can inject blood or you can inject some type of, uh, fibrin sealant. And we approach that extra durably, and you can either suture it or the way that I usually fix it is with one or two or three titanium clips. And these lateral tears are always associated with the nerve root sleeve, right? So where the nerve root comes out, the common thecal sac, and it’s usually inferior to it. And rarely it’s more at the shoulder of where that nerve root comes out. And then the cysts, uh, type what we call 2A, those are simple cysts that you can see here a couple of examples.
[00:12:00] This is the MR myelo, post-myelo CT. This is what it looks like at surgery. And, you know, we still pay a lot of attention to it, of course, but Dr Kranz quite a few years ago now already showed us that you know this also is really common in what we all refer to as the general population. So about two out of five people in this audience would have one or more of these simple cysts.
[00:12:28] I think that is really rare for these simple cysts to rupture. And then if they do, oftentimes it’s not just confined to the spinal canal, uh, but actually ruptures more out into the plural space. So this was a lady, I think she has Marfan syndrome, uh, and she was diagnosed with hydrothorax that was beta two transferrin positive about 12 years ago, but she didn’t really have that many headaches. But then her headache started and her brain MRI became grossly abnormal. And then dural ectasia, right? These are very difficult to treat type of leaks. You know, sometimes they look more like a Tarlov cyst. And Dr. Schrot and Dr. Murphy will talk about that, uh, tomorrow about the treatment.
[00:13:13] But what I would just like to say is that, you know, although that the surgery for simple cysts, that’s, that’s really pretty simple. But surgery for these, I know like dural ectasia, complex cysts can be, you know, very complex. And the, the main complication is that you cause a leak, right? And then you end up with like a really large pseudomeningocele.
[00:13:37] That’s, you know, that’s and also you can see how beautiful, you know, spinal fluid looks, it’s my favorite bodily fluid.
[00:13:51] But that’s, you know, so try not to operate on those, unless you really need to. And then, you know, the type three leaks are the CSF venous fistula, you know, it was first discovered about 10 years ago, totally fortuitously by the way. It’s not that anybody was looking for it. And as all of you know, the main difference of the fistula compared to the other ones is that there’s no spinal fluid outside of the dura that you can detect because it just gets sucked out by this vein.
[00:14:21] Now how common are those fistulas? We have a little bit of information about that. So if you have people with an abnormal brain MRI, we can detect that, at least 75 percent of the time, if it’s people with a normal brain MRI, but they have orthostatic headaches and they’re sent by a headache specialist, we find it about 10 percent of the time.
[00:14:41] So that’s not very often. And then, as I mentioned earlier today, if you have this sagging brain syndrome with FTD, we only are able to discover it about 40 percent of the time. They’re also different types. I apologize for those drawings. I made those. Um, so you can have, uh, the most common type is really just where there’s the fistula is very distal along the nerve root sleeve.
[00:15:08] Sometimes it’s very proximal so that’s easy to fix with surgery. And sometimes it actually arises from the common thecal sac. So, and then this VM plus or VM one if you want to do a sub-typing of these fistulas, that relates to whether or not they’re associated with a vascular malformation, and a lot of people have described that already.
[00:15:32] So this is a young girl who was sent to us a few years ago, was born with this very extensive lymphatic malformation with orthostatic headaches, brain sagging. This is what her MRI looked like. This is the sacrum here, this pelvis. It’s a very extensive vascular anomaly. This is a digital subfection myelogram Dr.. Maya did. And you see here, this is the CSF venous fistula that arises from a little diverticulum into this vascular malformation. I was not able to fix that with surgery. As a matter of fact, I had to abort the surgery. Luckily Dr. Maya was able to fix it with direct punctures translaminar into this vascular malformation along where these fistulas connections were and injected it with onyx.
[00:16:22] So that was great. And then this is somebody we saw recently, uh, who’d had really bad orthostatic headaches for a long time. He has some other type of widespread vascular anomalies really from, you know, tip to toe, you can see here in his tongue and the bottom of his feet. Uh, this is what his MRI shows. And then this is what is digital subtraction myelogram showed. Here you can start seeing this little fistulous connection. And then there it becomes clear that this is, you know, pretty complex CSF venous fistula. And I think while we were at our meeting in Naples, uh, Dr. Maya fixed that with endovascular treatment.
[00:17:02] This is a lady with a fistula that had been there literally for 40 years. As you can see, she’s had multiple treatments, laminectomies, because some really good neurovascular surgeons thought she had a angiographically occult arteriovenous malformation of the cervical spine because of these giant dilated veins, right?
[00:17:25] So that’s intradural, extramedullary dilated veins, but that’s just from the leak, right? So there’s not enough CSF. Uh, how does the body react to that? Just like meningeal enhancement, they can engorge their venous system. This was the vascular malformation. This is a digital myelogram. We went straight to surgery because they already had tried to treat it with embolization.
[00:17:49] And they had to have a few direct glue injections, and then, you know, after fixing this fistula, which was, you know, super easy to do, this AVM that had been bothering her, apparently she’d had four surgeries for that, just, you know, went away 24 hours after surgery. So what we have learned, at least we think we have learned, is that for these fistulas, blood patches don’t really work that well.
[00:18:16] So unless we get to them really early on, which is unusual, right, because you have to do all this sophisticated imaging, we just go directly to more invasive testing trying to figure out where the leak is. And then there really are three good ways of treating it. You can inject glue, you can embolize it with onyx, or you can do surgery.
[00:18:38] And it’s really up to the patient, right, because all these treatments have advantages and disadvantages. And one of the main things that I like about the endovascular treatment of it, not only, you know, it’s less invasive. Appears to be. But it also has really attracted a whole new group of physicians to the field of CSF leaks, because endovascular radiologist neurosurgeons now become interested in it. And of course, you know, they do try to attract patients so they sometimes, you know, say things that that are a little bit against surgery, like you have to do a laminectomy and a facetectomy, ligate the nerve root— that’s really not the case, you don’t need to do that invasive of a surgery.
[00:19:24] Or they might say things that, you know, that fibrin glue injections are only effective in a tiny minority. That’s not true either. This is from Dr. Mamlouk’s paper; uh, Dr. Carlton Jones has had the same experience. So that’s not really true. So it’s not that, you know, at Cedars we’re like, you know, surgery versus embolization. It’s not like that. Or at Duke, it might be, you know, like, like glue versus endovascular embolization. Uh, that that’s really not the way it should be, right? So especially here in America, that’s not the way it should be. Uh, and it really has to be an individualized type of treatment.
[00:20:04] Thank you.