Dr. Geoffrey Parker at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Geoffrey Parker, of Macquarie University Hospital in Australia, presented this talk on “SIH Cranial Imaging: Qualitative and Quantitative” 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. Parker’s slides here.
Transcript
Dr. Geoffrey Parker on “SIH Cranial Imaging: Qualitative and Quantitative”:
[00:00:09] Thank you. Thank you very much. And thanks Wouter and Marcus for the invitation. So my job today is to talk about cranial imaging in SIH. So we start off with the definition of intracranial hypotension, and this is the International Headache Society, and that the definition is that, you have to have a pressure less than six and those clinical manifestations we’ve been talking about this morning.
[00:00:48] So the problem is that this definition was written in order to be very specific, and there are a great many patients who either haven’t read this definition or don’t fit this definition. So that’s a real problem. How do you find a patient? So Wouter’s group in 2011 came up with an alternative definition and they said that they hoped that these criteria would be, could make a basis for change when the criteria were revised, and their criteria are presence of orthostatic headache, which we’ve heard this morning, even that is not completely specific.
[00:01:29] Presence of at least one of the following being low opening pressure, less than six, sustained improvement after epidural blood patching, demonstration of an active CSF leak, and most importantly, and what we’re here for this morning, is cranial magnetic resonance imaging changes of intracranial hypotension like brain sagging or pachymenogeneal enhancement.
[00:01:54] And you also needed to have no recent history of dural puncture and not attributable to another disorder. So that was written in 2011. Uh, and the criteria were revised in 2018, and they’re still much the same, so we still, we still have this problem. So the cranial imaging features include subdural collections, acquired tonsillar ectopia, dural venous sinus distension, and this is on CT.
[00:02:27] Uh, and the so called layer cake skull, which is diffuse thickening of the inner table of the skull, and one reference is that this occurs in up to 14 percent of patients with very long-standing SIH. Um, so that’s CT. And on MR, the most common quality you’re finding is pachymeningeal thickening and enhancement, followed by dural venous engorgement, tonsillar herniation, and subdural collections. However, as we’ve heard this morning, these features are not always present, which is why quantitative features are helpful in making an accurate diagnosis. So these are qualitative and quantitative features. We’ll follow along with now, uh, here’s the, this is pachymeningeal thickening on the right hand side.
[00:03:20] You can see an obvious case. And it’s, it’s quite interesting to see how, how much clearer this is seen when you use T1 weighted images, T1 gadolinium enhanced images with fat suppression, it’s beautifully seen. It’s also very interesting, I think, that the pachymeningeal thickening is thicker anteriorly than posteriorly, which is presumably a gravitational effect. Sometimes on the left hand side, you can see, a normal patient with just a whisker of meningeal enhancement, and that can be over interpreted as dural enhancement. And also people do tend to get confused sometimes with vessels adjacent to the pachymeninges, which they say are pachymeningeal enhancement, but it has to be, it has to be thick. It has to be fairly smooth. And that’s a good example on the, on the right.
[00:04:15] So that’s pachymeningeal thickening. What about dural enhancement? Well, here’s a case where there is on the left hand side, uh, quite marked distention of the straight sinus, vein of Galen straight sinus, and a cross sectional image through the sagittal sinus shows that the sinus walls have a convex outer margin.
[00:04:37] So what we’re looking at here is really just a simple barometer. The sagittal sinus here is, if you like, sucked open, and that tells you that the patient has intracranial hypotension and compare that to a patient with BIH, where the sagittal sinus has concave margins because the sagittal sinus is compressed by high intracranial pressure.
[00:05:05] Tonsillar herniation, another qualitative finding. This, I think this is extraordinarily interesting because what seems to be happening is that the brainstem becomes concertinaed. So the midbrain becomes concertinaed into the pons with this very strange vector of movement.
[00:05:25] So the, the midbrain comes down and it ends up going posterior to the pons. The pons gets pushed forward and compressed against the clivus, and the pontomesencephalic angle gets compressed and the pituitary enlarges. So that, that’s all, it’s this strange rotation, strange rotational vector, and as well as that, the tonsils descend.
[00:05:51] So here’s a list of the qualitative signs, a fuller list, pachymeningeal enhancement, pituitary enlargement, subdural thickening, diffuse cerebral edema, brain sagging, acquired ectopia, drooping of the splenium, and as well as what we heard this morning about a decreased size of the superior ophthalmic veins, you can also get reduced amount of fluid in the optic nerve sheath.
[00:06:21] So what about quantitative signs? Well, these are more measurable and therefore potentially more useful for statistical studies. If the pontomammillary distance should be more than 5.5, if it’s less, that, that’s a good, that’s a sign. The pontomesencephalic angle less than 50 degrees, the interpeduncular angle less than 40 degrees.
[00:06:48] The third ventricle always has a normal upslope and, uh, that’s, if that is less than 20 degrees, that is, that is abnormal. The supracellar cistern should be greater than or equal to 4 millimetres and the prepontine cistern should be greater than or equal to 5 millimetres in normal patients. And we’ve heard about the SEEPS mnemonic from this morning.
[00:07:16] Now, so why does the brain stem sag? Well, I’ve told you about that strange curved vector. The cerebellar tonsils seem to disappear through the foramen magnum. A bit like things going down the plug hole in the bath. And this is an example of a reservoir, the Ladybower Reservoir in the UK.
[00:07:41] And because of the town just down below, they couldn’t have a spillway in this reservoir. So they built this very strange funnel like arrangement, which is a bit like the boundary zone of a black hole. You know, if you, if you get too close to it, you’re utterly finished, unless you can travel faster than the speed of light.
[00:08:01] And that is, that seems to be why the tonsils and the brainstem perhaps get sucked down through the foramen magnum. Here’s an example of a patient where we, where we found a large CSF-venous fistula. Jeff, Jeffrey Brennan in the audience. operated on this patient and the fistula was closed and so that’s her MR brain at the time of diagnosis.
[00:08:33] You can see the brainstem descent, the tonsillar descent, um, the, that concertina arrangement. And six weeks after the fistula was closed, that’s what she looks like on the right hand side, which is just, I think, utterly amazing.
[00:08:52] The other quantitative measures are mamillopontine distance, so that’s supposed to be more than six in the normal patient on the left at 7.7, and the height of the suprasellar cistern is six millimeters in the abnormal patient. On the right hand side, the mamillopontine distance is down to four, and the suprasellar cistern is, is uh, three millimeters. Pontomesencephalic angle, nor normally should be more than 50 in the patient.
[00:09:23] In the normal patient on the left, it’s 52, and in the patient on the right, uh, it’s 37 degrees because, because of that how that vector is starting to rotate the structures of the brainstem. The interpeduncular angle I think is, is very useful in a normal patient. There’s a normal patient on the right hand side, 55 degrees in a patient with SIH.
[00:09:52] The angle is reduced to 29 degrees in another patient, even worse that the angle is, is less than 10. So what is happening is that the clam shell of the cerebral peduncles is closing as the brainstem descends. And those two convex surfaces come together. Compare that to a patient with progressive supranuclear palsy where there is midbrain atrophy, and you can see the interpeduncular angle in the PSP patient is 90 degrees and there’s the typical convexity of the outer margin of the midbrain. So, okay, we’ve given you, we’ve given you qualitative, we’ve given you quantitative, but even those measurements aren’t enough. And there’s been this effort that Tom Dobrocky from Switzerland has come up with to formalize assessment with the Bern SIH score.
[00:10:48] And that was, that came out in 2019. And what they did was they looked at 152 patients. And they evaluated nine quantitative and seven qualitative signs, and they came up with a predictive diagnostics score, which was statistically validated with some prospective patients who were also suspected of having SIH.
[00:11:14] And the final scoring system they came up with was, there were major criteria, two points. You’ve got two points for pachymeningeal enhancement, two points for venous sinus engorgement. Uh, two points for effacement of the suprasellar cistern to less than four, and you’ve got one point for subdural fluid collection.
[00:11:35] One point for pre-pontine cistern less than five and one point for mamillopontine distance of 6.5 or less, and so they said, well, if you if you get two points or less, you’ve got a low probability of a spinal CSF leak. If you get three to four, there’s an intermediate probability, and more than five, five or more, beg your pardon, you’ve got a high probability.
[00:12:01] And they showed that this was true in their validation. This is just a couple of their images showing those compressive features that we’ve mentioned. These are their statistics. So on the left hand side is their derivation cohort, and you can see that patients on the far left, patients with SIH who have a positive leak, the vast majority have a score greater than equal to five, but also, but look, there were four patients in that group who had a score of less than or equal to two.
[00:12:35] In the control group, the vast majority of patients had a score less than or equal to two and a few, only a few had higher scores. And in patients with orthostatic headache, who were leak negative, most of them had a score of less than two and in the validation cohort that showed much the same thing.
[00:12:56] So, I think it’s a very, very valuable initiative. I think that further refinements or changes in the threshold or weightings might be able to increase the discriminatory ability of this score. For example, we’ve heard, um, this morning, Dr. Kranz published that the pachymeningeal enhancement decreases over time.
[00:13:22] So I just wonder why that is given two points in that validation scheme, because when that disappears, you’re, you know, you’re starting to lose the chance to get, to get over five. And I think what needs to happen is, is that this score needs to be validated in a large multicenter patient cohort.
[00:13:42] So that’s all I’ve got and I’ll leave it at that. Thank you.