Slides
Transcript
Thanks very much again for the invitation to be here. For this talk I’m going to go through a little bit about Chiari, a bit about syringomyelia, and then try to bring it together to see how it all fits together in SIH, and then I’ll close by just mentioning iatrogenic disorders. Some of the things we deal with are like this, right, where I’d like at the end of this talk to take you on a line like that, but I’m afraid it’s going to be like this, where it won’t be so clear.
I’m sure you all know that the so-called definition of Chiari malformation is the herniation of the cerebellar tonsils by 5 millimeters or more, and that is entirely based on this paper from 1986 studying 25 Chiari patients. I think we know that—I mean it’s a useful kind of guide—but it’s not really a definition. It works well for most of the patients that we see with Chiari malformation, and they certainly respond well to surgery, certainly in terms of resolution of headache.
But Chiari is much more complicated than that. So often it’s associated with skull-based abnormalities, platybasia, and in this case an acute clivus axis angle. So Chiari malformation but complex skull-based anatomy. For most of these patients they still actually do well with a simple decompression. Sometimes they need fusion if there’s brainstem compression, but generally speaking they do well with just decompression.
We also know that Chiari malformation is more common in patients with EDS, and of course these patients have craniocervical instability commonly and POTS, and so they get symptoms that are very similar just from that instability. So it’s a complex field and much more than that 5 mm rule. Sometimes, as in this case, the herniation is less than 5 mm but the patient is still very symptomatic, and we find assessment of those patients to be improved by doing dynamic imaging.
I hope you can see here that the cerebellum is actually pulsing with each cardiac pulsation. In this patient with very typical Chiari symptoms responded very well to posterior fossa decompression but didn’t really meet the criterion of 5 mm of herniation. When I talk about syringomyelia, I like to tell the story of the etymology, right? So this is from the Greek Metamorphoses where this is a human Syrinx is chased by the Greek god of the shepherd Pan who had the hots for her and when he tried to grab her and she wasn’t interested, she prayed to the other gods to be released from this chase. And so they changed her into a clump of reeds by the riverbank. And so Pan, when he’s trying to grab Syrinx, grabbed this clump of reeds and he was exasperated and let out a sigh which created this musical note and that is the origin of the pan pipes, his pan blowing his breath across these clumps of reeds. And Syrinx is the name for a tube, right, so the origin of the word syringe comes from Syrinx. This is my favorite Pan chasing Syrinx painting but that’s a Flemish painter. And I thought I can’t show that at a conference in the Netherlands, right? So there must be some Dutch paintings. Well, I had no trouble finding Dutch paintings of Pan chasing Syrinx. So I think it’s like in the Netherlands there’s this fascination with syringomyelia dating back to the 17th century.
But what we know is that the pathology of syringomyelia is a little complicated, right? So it can be that it’s an enlargement of the spinal cord central canal that communicates directly with the fourth ventricle, or it can be that it’s enlargement of the central canal without a communication, or it can be that the syrinx commences directly in the cord tissue outside the central canal, and that’s important in terms of the clinical features. And just to remind you, histologically the central canal is normally tiny. We don’t see it normally on an MRI, and for most, particularly I think non-surgeons who think about syringomyelia, they think of it like this, as you might see it in a postmortem specimen. But of course syringomyelia is a very different disease. It’s this very tense, dynamic, cystic enlargement inside the cord. And here this is a post-traumatic case, but opening into the syrinx and that syrinx fluid coming out under high pressure. So this is a really dynamic and high-pressure situation, not that cavitation and passive scenario that you might be led to believe.
And we’re also taught that the cape-like distribution of sensory change in syringomyelia is caused by damage to the crossing spinothalamic fibers in the center of the cord. That’s actually wrong. Right? So the truth is that the sensory disturbance is caused by rupture of the syrinx. So imagine an enlarged central canal. What happens here? It is, of course, it’s upside down, and that’s enlarged central canal, and it’s ruptured into the dorsal horn. So it’s actually disruption of the sensory fibers in the dorsal horn, not in the center of the cord. And that’s why often the sensory disturbance is actually unilateral rather than bilateral.
And so it’s not really the size of the syrinx that counts. It’s whether it’s either it’s just an expansion of the central canal, which in a sense would be analogous to hydrocephalus, where there’s not any focal neurological deficits. But if it’s gone through the central canal or it starts outside the central canal, that’s where it’s damaging the cord tissue, and that’s where you get neurological deficits. So it can be a small syrinx like this one, but very symptomatic because it’s affected the dorsal horn of the spinal cord.
And I’m going to come to all this together in a minute, but syringomyelia is often associated with some sort of spinal abnormality. And so in this case the pathological lesion is that arachnoid membrane that we can see in the dynamic imaging that’s caused that syrinx. In this case it’s arachnoiditis ossificans, that’s actually bone in the arachnoid. And in these kinds of cases the syrinx starts outside the central canal. So they can be very small but very symptomatic because it’s affecting the cord tissue, not just expanding the central canal.
They can be associated with a tumor and particularly ependymoma, for example, which really expands the central canal. So sometimes these patients are symptomatic when the syrinx ruptures through into the cord tissue, but usually they’re not symptomatic. They can be associated with tethered cord, again not usually directly symptomatic from the syrinx itself.
And sometimes we don’t really know what the cause is. For example, this one—is this a craniocervical junction abnormality, is it spinal canal stenosis, is it this membrane of dura here, is there some abnormality in the fourth ventricle? Sometimes it’s hard to tell, and sometimes we’ll never know. But in this kind of case, a syrinx in the cervical cord—very dynamic, no evidence of subarachnoid pathology, but responds well to direct shunting.
But syringomyelia is probably more certainly, in the context of what we’re talking about today, is best thought of as in association with craniocervical junction abnormality, and that’s Chiari malformation, but other abnormalities as well.
So arachnoiditis at the craniocervical junction resulting from hemorrhage, prior trauma, or prior surgery, tumors, arachnoid cysts are common at the craniocervical junction and can cause a syrinx.
So in association with Chiari malformation, it’s about half the patients depending on how you define Chiari malformation. But what I’ll draw your attention to is that for most Chiari patients, the syrinx is really quite remote from the fourth ventricle.
So what we know from pathological studies is that for most Chiari patients, it’s non-communicating. So there is not a connection from the fourth ventricle to the syrinx. Which means those original theories of Gardner and Williams of CSF coming from the fourth ventricle down the central canal into the syrinx can’t possibly be true.
Most of the patients are like this, and so some of the theories are that there’s some sort of pressure abnormality in the subarachnoid space that’s forcing CSF through the cord and expanding the syrinx back out. But that also can’t be true because if you have pressure on the outside of the cord, that can only cause the cord to flatten. It can’t cause a cyst on the inside to expand against itself. So there must be some sort of dynamic pulsatile pathophysiology going on here, and I’m not going to delve into that any further.
But we know that for Chiari patients with a syrinx, even though the syrinx might be right down here, surgery resolves the syrinx if you get a good space at the craniocervical junction. But sometimes it’s not like that. For some Chiari patients, the syrinx is actually communicating. And we can tell that because the syrinx is up here, it’s not down there, and it might be difficult to see on MR and better seen at operation through intraoperative ultrasound. This is fourth ventricle. That’s syrinx down there. And we can tell that there’s a connection between fourth ventricle and syrinx. So this is actually truly a communicating syrinx. And that happens maybe in about 5 to 10% of the Chiari-associated syrinx cases.
And usually for those cases, what we find is something that’s obstructing the CSF outflow from the fourth ventricle. In this case, it’s a membrane across the foramen of Magendie. You can tell it’s Chiari, the sclerosis of the tonsils, that’s the medulla, close medulla, now into the fourth ventricle. So we’ve opened into the fourth ventricle, and for those patients, respond. Just to clarify the pathology, it’s communicating. Often it’s small like that, sometimes they extend more inferiorly. I think what happens is the syrinx continues until it reaches a point of ependymal adhesion. As we get, as humans get older, they get multiple areas where the ependyma adheres to itself and closes the central canal. So the size or the longitudinal extent of the syrinx will be determined by where that ependymal adhesion is.
And the problem with the Chiari patients, when it’s small like this, there’s an ependymal adhesion here. They’re communicating. Those patients tend to get symptomatic syrinxes quite early because the syrinx expands. It’s small, but under high pressure, it ruptures into the surrounding cord tissue, and those patients become symptomatic. Even though the syrinx is quite small. They still respond well to surgery, right. So just with posterior fossa decompression, resecting that membrane, sometimes we have to put in a shunt from the fourth ventricle into the spinal subarachnoid space just to keep that foramen open and keep that CSF flow coming out of the fourth ventricle.
It’s not just Chiari, arachnoiditis, other things can cause syringomyelia associated with those conditions. But what’s really interesting is that for almost all of those cases, the syrinx is communicating. So rather than like in Chiari, the syrinx occurring down here, for all of the craniocervical junction abnormality cases apart from Chiari, the syrinx is like this— where it’s right up against the fourth ventricle. We can usually demonstrate communication, and they respond well to surgery with communicating the fourth ventricle into the spinal subarachnoid space.
Here’s a few more examples. Meningitis, birth trauma, hemorrhage, craniocervical junction hemorrhage from birth resulting in this dense band of arachnoid tissue obstructing CSF flow into the spinal subarachnoid space and responding well to surgery at the craniocervical junction and just maintaining that CSF flow out of the fourth ventricle into the spinal subarachnoid space, and often, as I’m showing here, putting in a ventricular catheter to keep that open, and that works very well.
Here’s a case where the patient had a posterior circulation subarachnoid hemorrhage treated surgically. Terrible posterior fossa craniocervical junction arachnoiditis, communicating syrinx, responding well to surgery.
So just to emphasize, the difference here is that there’s some sort of arachnoid pathology at the cervical junction that’s forcing CSF into the fourth ventricle and down this small channel usually, but not very far, until it expands the central canal and forms that communicating syrinx. And so perhaps those original theories of Gardner—and not Williams but the original theory of Gardner—where it’s this pulsatile flow of CSF from the fourth ventricle into the cord is the actual pathology for those cases.
So what about in SIH? So I’m getting to the point here, which is we all know that tonsil herniation is relatively common, and there’s variation in terms of how common it is in series of SIH. But it didn’t make the Bern score or the Beck score, the Bern score, the Beck score, right? It didn’t make the list, right? So although it’s common, it’s not a discriminating feature. The biggest series of syrinx in this condition that I came across was from the Mayo Clinic, where they had six patients. And so some of the things that they did was compare SIH with syrinx versus SIH with tonsil herniation but no syrinx, and compared them. Upper limb extremity weakness. The displacement through the foramen magnum was greater in the patients with syrinx. Flattening of the pons was more common in patients with syrinx. But small numbers and difficult to tease out.
And what was not clear from that paper was whether any of those symptoms were actually related to the syrinx. Had the syrinx ruptured into the cord tissue to cause any symptoms? You couldn’t tell really from that paper. The upper limb weakness I’m not sure was really attributed to the syrinx itself. So what I’ve tried to do here is, I guess this is a personal opinion mostly, but drawing upon all of that, what is the difference between Chiari malformation and tonsillar herniation, and syrinx and SIH, in terms of clinical features?
So Chiari patients don’t tend to have orthostatic headache, right? They’ve had headache for a long time, it’s not different whether they stand up. So that’s the key differentiation. But there are other clinical differences in the headache. Chiari patients have occipital headache almost entirely. They will sometimes have additional frontal headache but almost never only frontal headache. Whereas for the SIH patients it’s usually frontal or global. They can get occipital headache, and particularly those who have tonsillar herniation, they’ll have occipital headache, but it’s not a major feature.
Chiari patients have had this for a long time. Usually they’ll tell you, I thought everyone got a headache when they coughed or laughed or sneezed. I thought that was normal. I’ve had that since I was a child. They don’t come in their 30s, 40s, and 50s and say I never had a headache before and now I’ve suddenly got a headache. So it’s a really important clinical difference. The onset for Chiari patients has been long, for SIH it’s often sudden.
Cognitive effects are common in both conditions. Chiari patients get cognitive disturbance, there’s no doubt about that. It’s been well studied. There’s work trying to work out what the basis for that is, but there’s no doubt that these patients have cognitive effects. But it never gets to the point of frontal temporal dementia, all right? So if they have that, it’s not just Chiari.
Chiari patients get hearing changes. They get muffled hearing, changes in tone. That’s very common for these patients. That’s very similar with SIH. I know tremor is not common in SIH, but I’ve never seen it in Chiari, so that’s a discriminating feature. And certainly the other brain stem signs can occur in both conditions.
What about the imaging features? Well, by definition, they both have tonsillar herniation, that’s what we’re talking about. The Chiari patients don’t have other Bern signs, right? I can call them Bern signs. They don’t have flattening of the pons, reduction in the mamillopontine distance, reduction of the distance between the optic apparatus and the pituitary gland, so all pituitary enlarged venous sinuses. Chiari patients don’t have that.
However, they can have IIH signs because it often coexists. Patients with Chiari malformation will have intracranial hypertension, and patients with SIH may have had pre-existing intracranial hypertension. So that’s not a discriminating feature.
Now I can’t prove this, but I think that buckling of the medulla is a discriminating feature. This is a Chiari patient. That is the cleft between the cerebellum and the medulla. That knuckle there is actually medulla. It’s medulla that has been shaped into that by the constant pulsation from the tonsils herniation from the Chiari. And I hope that that demonstration that’s shown here on the dynamic scan.
Whereas tonsillar herniation in SIH, I’ve never seen it buckle in the medulla. I might be proved wrong, but I suspect that that’s an extremely uncommon finding in SIH. Syrinx can occur in both of these conditions. Common in Chiari malformation, but pretty uncommon in SIH. As I say, the largest series is six patients.
One of the key points, the syrinx type in Chiari malformation, the vast majority are non-communicating. In SIH, they are all communicating. At least all the ones that I’ve seen, all the ones in the literature, the one that Lalani showed this morning that was communicating.
Do they get symptoms? Chiari patients commonly not, commonly present just with a Chiari malformation, but they haven’t ruptured the syrinx into the surrounding cord. So they don’t have clinical manifestations from that. SIH, I think this is really uncommon, and I think it’s because they don’t have that same pressure of CSF going into the syrinx from the fourth ventricle to force it open and to rupture through the ependymal lining into the cord tissue and damage the cord. So I think that’s a discriminating feature.
And cerebellar slump postoperatively can occur, I think, in Chiari patients, and I’ll show an example of that, but I think it’s more common if a SIH patient with tonsillar herniation has a posterior fossa decompression—much more likely they’ll get slump.
Just an example of a patient with tremor with a syrinx. The syrinx is not the cause of the tremor. Just showing here that this is a patient with severe SIH but also with a syrinx. But it can be tricky. This is a young woman who presented with headache, and on the initial imaging I would challenge any of you to identify features of intracranial hypotension on that. She’s young, so pituitary is relatively large. T2 axial imaging showed what, that was interpreted as epidural fat. In retrospect, it’s actually CSF.
She had a posterior fossa decompression, had terrible cerebellar slump, but now has developed other signs of intracranial hypertension. And the original surgeon said, well, we need to decompress more, right? But of course when we investigated we found that she had a type 2 leak from a nerve root cyst, and what was interpreted as epidural fat was actually epidural CSF, and she responded well to surgery.
This is my last slide really, and just to admit that I’m responsible for quite a few iatrogenic intracranial hypotensions when I treat patients with intracranial hypertension. And this is a complex patient where we’ve ended up putting in a shunt in the cisterna magna to try and control her intracranial hypertension. She’s ended up getting tonsillar herniation, a syrinx, but this syrinx is caused by a membrane, an arachnoid scar membrane across the fourth ventricle. And we were able to treat that successfully by putting a shunt into the fourth ventricle.
And here I’m just showing a Seldinger technique to put the shunt in the fourth ventricle by putting a stylet down this small catheter. It was pretty difficult to get into the fourth ventricle in this case, and just a tip on how to get into the fourth ventricle when that’s a bit tricky. But that’s all we did. It’s a pretty small dural opening, just putting an extra shunt from the fourth ventricle to the spinal subarachnoid space, and that has resolved that communicating syrinx.
So in conclusion, the tonsillar descent is relatively common in SIH, but it can mimic Chiari malformation. What I didn’t mention for that case where she had imaging features typical for Chiari was that her history was typical for SIH. So who would have thought you actually got to take a history from the patient, and it tells you what the condition is, but there you go.
The clinical features are usually different. As for syringomyelia, it’s rare in SIH. I think it’s almost always, if not always, communicating. It’s almost always, if not always, actually asymptomatic, but it does improve with treatment. Just to emphasize, I think that this is a multidisciplinary team field that we work in. This is our CSF MDT, and of course we are in a beautiful city of Amsterdam, but if we ever get a chance to come to Sydney, that’s what you were saying. Thanks for your attention.