Dr. Michael Malinzak at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Mike Malinzak, assistant professor of radiology at Duke University Medical Center, presented this talk on diagnostic imaging in idiopathic intracranial hypertension at the 2023 Cedars-Sinai Intracranial Hypotension Conference on July 9, 2023. The conference was hosted by Cedars-Sinai with generous support from the Spinal CSF Leak Foundation in Kohala Coast, Hawaii.
Transcript
[00:00:12] Hello, everyone. So, diagnostic imaging in IIH, idiopathic intracranial hypertension. Um, that’s a big sella. Look at that. Okay, so we’ll go through a few things. The role of imaging in IIH diagnosis, which is a little bit limited in some ways.
[00:00:32] The imaging features we’re looking for in IIH, imaging pitfalls, and then complications seen on imaging. And I’m sure you all gathered that during the break we’ve really switched gears. Now we’re talking about high pressure, right? There are a variety of diagnostic schema for calling IIH, and there’s an entire section of like famous headache neurologists down there.
[00:00:51] You just yell if I say something dumb, especially you, Dr. Friedman, you know. But as I understand the Dandy criteria, the oldest ones. And they seem very inclusive, right? It basically says like any signs or symptoms of elevated intracranial pressure and an opening pressure greater than 25 centimeters of CSF or water, but they’re pretty much the same as far as gravity is concerned, right?
[00:01:13] And then the ICDH3 criteria. Which I find interesting in, in that neuroimaging can be like supportive in the way it’s written, but it’s not at all part of the actual diagnostic criteria, which are papilledema, pulsatile tinnitus, either one of those, plus an open pressure of 25, if you really boil it down.
[00:01:30] Of course, you have to read through the, all of the ICDH3, like legal code, you know, I feel like a lawyer, always it’s like, “notwithstanding criteria seven and eight and not better explained by—” there’s a lot of like triple negatives. But as you sort of really squeeze it down, that’s what you ring out. And then the AAN criteria which requires opening pressure of greater than 25 centimeters of water and kindly includes children. And it recognizes that people under 18 often have higher opening pressures.
[00:01:55] So 28 and above is where it becomes abnormal. This really breaks it down into the people with papilledema, and those who have elevated pressures and other problems, but not papilledema. Sixth nerve palsy is, pretty much all these can be treated as sort of a papilledema equivalent, I feel, and that it’s like definitely wrong, I can see it’s wrong with you, it’s a sign, not a symptom. And then this actually makes allocations for imaging features, so three of the following would rise to the diagnostic threshold.
[00:02:23] One thing we get along about and we get to high five about is is some of the negative stuff, right? All of the criteria are going to say that your brain MRIs and your MRVs can’t have other causes of elevated intracranial pressure on them. Right? And in many ways, imaging is there to exclude other causes, right? That’s why we’re doing this. We’re making sure it’s not tumor cerebri, not pseudo tumor, but you actually had a brain tumor, or that you have a thrombosed vein. We also want to see no localizing symptoms outside of the sixth nerve palsy.
[00:02:56] We want to see normal CSF labs. The things that cause this that aren’t truly idiopathic because we have a reason: you know, prior trauma, prior CNS infection, anything that raises venous pressure. So, the thrombosed dural venous sinuses, any kind of fistula in the neuraxis, right? Heart failure, which, you know, it’s everywhere. Metabolic causes, so, renal failure liver failure, hypercarbia. Hormonal disruptions. One of the more fun things for me was I got sent a person, internationally, I believe, for a CT myelogram and I found a huge parathyroid adenoma just like on the side. And it’s supposed to be hypoparathyroidism that leads to IIH, but this individual had, you know, headaches and we ultimately think it’s related to hyperparathyroidism.
[00:03:37] So, lots of things can mess with this. Of course, retinoids and tetracyclines are the classic meds. I’ve seen Deb speak and list a big, bigger list than that, I believe. And certain genetic causes, classically Down syndrome and Turner syndrome are elevated ICP disorders. So what do we see on imaging, these things that are supportive but not in themselves diagnostic?
[00:04:02] These are all changes related to the CSF volume and pressure, pretty much squishing other things, right. And in a lot of ways it’s the opposite of SIH, not entirely, but in a lot of ways, right. So there’s a bunch of things that happen to the optic nerve, a nice illustration of several of those things from one of our group papers. There’s a lot of changes in the veins. There can be tonsillar descent, which is confusing. We’ll talk about that plenty. And then enlargement of the CSF spaces and remodeling of the bone, particularly at the skull base. Right? I think this is one of these things that’s just simpler when you just put it in a box. You know, we don’t see a lot of enlargement of, of CSF spaces up here, and IIH is all kind of down low, which I think helps if you’re just getting, you know, your head around the imaging.
[00:04:43] Real quick though, you know, this is a big list. Here is a little bit of a summary that looks at pooled specificities of these findings. This is a group out of Emory from 2015, a really nice review paper on this. I’m not sure if I totally believe these numbers and because they’re pooled specificities, the way these things were, you know, included or not included varies among the different papers, but, you know, to my mind, I like stoplights, the green and yellow are pretty suggestive, and pretty much anything that happens to the optic nerve is pretty suggestive. When we see that transverse sinus narrowing, that’s about 90 percent specific. Cerebellar tonsillar descent is listed at about a 90 percent specificity.
[00:05:21] I don’t think I believe that, because there’s two other diseases that cause it, and it’s very easy to confuse them, but that’s out there. And then lastly, empty sella, expanded sella. You know, a little bit less specific. I’ve seen certainly much lower numbers in a variety of papers. And the skull base CSF space enlargement I would think is pretty telling, if you’re calling it right, you know, but this is sort of a rough idea.
[00:05:42] Alright, so let’s go through these findings individually and get familiar with what they look like if you don’t look at a ton of MRIs. So the optic nerve head. Normal tends to be in this pretty kind of, you know, Hawaiian ocean blue and and the pumpkin orange is, it means bad, right? And abnormal.
[00:05:58] So the optic nerve head should be kind of flat in the back of the globe. And it shouldn’t enhance like a funny little, you know, lump on the back of the surface of the globe here, as we see here. So when we see that it’s pretty darn specific. That to me is a papilledema equivalent almost.
[00:06:15] And it’s just not very common. So it’s a low specificity. In that same vein and of the same degree of specificity, on the 95 percent ish level, is protrusion of the optic nerve head without enhancement, either because he didn’t give contrast or just because it didn’t enhance. Again, some normal up here. It’s often helpful to look on the sagittal image. For whatever reason you sometimes see these findings much better on the sagittal than the axial and occasionally vice versa, but you know, I try to look at both. Every MRI that’s done for headache, I’m looking for this, right? Just because maybe I can throw that out there. You know, if I’ve got a great headache neurologist, they’re usually all over it, but we read for everybody everywhere. You know, sometimes it’s helpful to weigh in.
[00:06:53] So here’s that optic nerve head protrusion on a sagittal view. And here it is on axial. It can be really subtle, right? Even just a little bit of flattening of the globe with a bunch of other findings kind of pushes me to suggest it. And so I suggest that this is like the mild version, the moderate version, and the severe version.
[00:07:09] And that’s what we’re looking for, that sort of protrusion in the back. The nerve sheath can also distend and it can also become tortuous. The tortuosity is mostly described on a sagittal image, but you can see it on an axial too. So, you know, you want a relatively straight nerve without a ton of CSF around it, maybe a little bit’s okay, but this is getting to be a bit much and it’s getting to be a little too squiggly.
[00:07:31] Right? So that’s what we’re looking for. And it is pretty subjective. Yes, measurements have been published. No, I don’t use them in a routine practice. And the main fooler is, in many things in radiology like this, you’re making a ratio, right? Like it, it looks like this sheath is too big for this nerve, but what if the nerve is too small for the sheath? Having looked at enough MRIs at some point, you realize, nah, that’s about the right thickness of an optic nerve, right?
[00:07:56] Cause there’s some heritable forms of optic atrophy, there’s septo optic dysplasia that gives you small nerves, prior bouts of optic neuritis, all these things will cause your nerve to shrink and your sheath to stay the same and might mix you up. And we’ll look at a case of that later. All right.
[00:08:10] Venous findings. So this transverse sinus narrowing we’re really talking about the lateral aspects of the transverse sinuses way out far to the side. So on this coronal CTV image, you just don’t see the transverse sinuses. But as you scroll back and forth, you come to appreciate that there’s not thrombus in there. They kind of smoothly taper down to slit-like things. And that’s the stenosis we’re looking for. There are grading schema and it’s a little bit a little bit much for me. So I go with the 50 percent rule. So we know that if you look at people with IIH, 95% percent of them will have 50 percent narrowing of the transverse sinuses in that lateral most aspect compared to 5 percent of normal controls.
[00:08:50] So there’s some help in suggesting this is the problem, whether these stenoses are fixed stenotic things, or whether they are dynamic, or whether it’s just kind of a hypoplastic portion of vein with compensatory large vein on the other side. There’s a fair amount of subjectivity and you know, I think talking amongst ourselves is usually the right way to go there.
[00:09:11] Again, mild, I wouldn’t even say this rises to 50%, but there’s a little bit of narrowing there. Here we are moderate. It’s getting pretty thin. And then severe. I kind of lose the vein in that segment. And you can look at this with CTV or MRV, or honestly contrast enhanced MR is shown to be almost as good if you’re getting, you know, high resolution imaging, which often we are.
[00:09:32] So this is the idea of that dynamic compression that I find pretty compelling. So whether the stenosis was fixed or dynamic in the first place, once it’s present, we have backup of blood upstream, and the rest of our neural venous sinuses over the top of our head. And that decreases that pressure gradient between the CSF, which is higher pressure, and the venous compartment, which is lower pressure.
[00:09:55] So you stop absorbing CSF—well, you slow down how much you’re absorbing CSF. Consequently, your CSF pressure goes up, and there’s a thought that that further compresses the transverse sigmoid junction, and you’re stuck in a bad feedback loop. Right. And then if there’s an underlying stenosis, you know, probably something that predilects for this. That’s how I understand it. So, other venous findings, these are not specific, but they’re out there, and I find they’re kind of true and I like them in my age population. Occipital emissary veins are kind of unusual, right? This funny little vein punching across here.
[00:10:30] An occipital sinus, which we all have in, in early like fetal life. And many of us lose, we get older, these sorts of veins that, that either recanalize or hypertrophy to get blood around those transverse sinuses. The more you see of them in an individual, the more I think like, maybe we’re dealing with IIH. And this is a person trying to reroute their flow around the traffic jam happening at the transverse sinuses.
[00:10:53] Cerebellar tonsillar descent. Well, this is hard, right? I mean, we just spent an entire conference talking about SIH and how the tonsils get low in low pressure. What are they doing low in high pressure? Yeah, thanks universe. So, it happens, and I don’t really get it. And it also happens in Chiari 1. One important thing is to notice that in SIH, you know, brain sagging is really a midbrain phenomenon. We’re looking for that crumpling in the midbrain, the Bern score, all the other metrics we talked about.
[00:11:21] And to me tonsillar descent is just kind of the afterthought among all the other findings. And when I meet a new headache patient and they have isolated tonsillar descent, you know, I’m not thinking SIH, primarily. It could be, but it’s just one thing, right? IIH, we just want to see that. That’s it. And this is one of my patients. She’s in her mid thirties. Five years ago, she got a brain MRI for something kind of unrelated and her tonsils look normal. And then when I met her, you know, she had these headaches that sound like IIH, pituitary gland has flattened, and now her tonsils are about 10 millimeters below the foramen magnum. And so, turns out this is this is what she had. We can also, you know, lean heavily on our clinical history in these settings.
[00:12:03] All right, moving on to the skull base, CSF space enlargement, partially empty sella. This is also a subjective thing. There are grading systems. I find them a little bit complex. I look for a pituitary that’s about halfway squished down, right? That’s what’s out there in the literature. In reality, you need to gauge for age. 19 year olds should have big pituitary. So this is too small.
[00:12:21] 42 year olds should have small pituitaries, but this is too small. The sella itself will enlarge in IIH patients. It’s about one millimeter per decade. Here’s an individual I saw 10 years apart on imaging, and it measures bigger and it looks bigger, and it seems to be real. Keep in mind the bony changes stay forever. So you might no longer have IIH, but you still got the bony changes, right? Other things that expand: Meckel’s cave, which is that area surrounding the fifth nerve. It really usually just looks like this little circle when it starts to take on this bi-lobed appearance on coronal, that’s pretty suggestive that we’re dealing, you know, with a high pressure scenario.
[00:12:54] As it gets really big, it erodes bone, it often digs down to the clivus. This individual actually entrapped their fifth nerve in it and had a new trigeminal nerve neuralgia associated, which is confusing from a headache phenotype standpoint. Arachnoid pits, very similar, a little bit more lateral in the stenoid, and they protrude down here. And this one has a little knuckle of brain in it.
[00:13:13] But they just kind of enlarge over time. And this is another skull based phenomenon. Meningoceles, very much the same thing, but, you know, into into a defined space on the other side, we can rupture through the cribiform plate into the nasal cavity, we can rupture into the middle ear and these are, you know, meningeal contained diverticula at this point. And they sometimes go out the foramen ovale, which is kind of wild, right along the V3 branch.
[00:13:37] So pitfalls, what can fool us? I mentioned optic neuritis, in particular papillitis, which is just that isolated autoimmune, inflammatory, post viral, you name it, inflammation of the optic nerve head. You know, that looks a lot like that optic nerve head swelling and enhancement, but in this case, you know, it’s T2 bright, and it’s focally enhancing, and it’s asymmetric, and the history fits.
[00:13:58] Similarly, I always want to be looking out for real tumors, and less the big brain tumors and more things in the dura. This individual had this funny thing back here along the transverse sinus and the optic nerve had swelling and she turned out to have metastatic sarcoma, which was news. And our NF2 population and people with meningiomas in general tended to, you know, collapse their sinus or occlude or thrombose their sinuses due to these meningiomas producing a secondary high pressure state.
[00:14:24] All right. Lastly, we’ll touch on a few complications. This is rare. It’s out there in case reports. This was a patient of mine who I met as an inpatient. He was 47. He lapsed on his meds and everything got a whole lot worse, and his vision was awful and his opening pressure was in the 50s. He has the optic nerve head swelling, and the transverse sinuses are definitely squished, and they’re not just squished, they’re thrombosed. He’s also a very hypercoagulable person with prior DVT, and I think by squishing those lateral transverse sinuses enough, he just clotted.
[00:14:53] Encephaloceles, so those, we mentioned the meningoceles. Once the brain herniates into them, you’ve got bigger problems, because that brain can kind of scar down, become gliotic, and become a seizure focus. I think all of us in neuroradiology have seen this where it’s like the fifth seizure study done on a patient. And you’re like, well, the good news is we have the positive seizures. The bad news is we missed it four times because we’re not always looking for brain herniating into the skull base, but that absolutely can be a pattern. Sometimes these things leak CSF as the dura kind of degrades.
[00:15:23] So here’s a knuckle of the the middle, you know, of the temporal lobe through the middle cranial fossa through an arachnoid pit into the sphenoid air cell. And once we give sternographic contrast, we can see that this fluid in the nasal cavity has iodine in it now. So this is a skull based CSF leak at this point. And then lastly, there’s this horrible overlap with spinal CSF leaks. We’re very clear that elevated intracranial pressure causes skull base leaks.
[00:15:48] It’s very well documented. The spinal CSF leaks, less so, but we kind of all think it happens, right? And here’s a time that I saw it in person. It’s a 14 year old, known IIH, moved across state boundaries, lost her headache neurologist, ran out of acetazolamide and her headache phenotype changed overnight one night from kind of a high pressure, constant thing to very positional.
[00:16:08] Her last documented opening pressure, about six months before, it was 37. We did a brain MRI. I don’t see signs of SIH, but her opening pressure when I checked it was now, I believe 6. I don’t see that written there, but that’s my recollection. And Oh, there it is. So I did a myelogram. And she had a ruptured nerve root sleeve here at L2 on the left, a little bit easier to see as you scroll back and forth.
[00:16:29] But here it is leaking out at L1. I glue patched this and she reverted back to her chronic headache phenotype and is still being cared for for IIH. So when you’re a high pressure state, I feel like you’re a water balloon that’s a little bit at risk of bursting. And we need to be cognizant of that too.
[00:16:44] That’s the rundown. Thank you so much.