Dr. Peter Kranz at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Peter Kranz, Associate Professor, Department of Radiology, Duke University Medical Center in Durham, NC, presented this talk titled “Rebound Intracranial Hypertension: Diagnosis, Imaging, and Natural History” 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] Thank you very much. I’m going to be talking about rebound intracranial hypertension, which is actually a really important topic that we probably don’t talk about enough.
[00:00:22] I’m going to be specifically talking about the diagnosis and the natural history and then the next two speakers will be talking about the ins and the outs and the what have you’s of treatment. And so my talk is going to be specifically confined to diagnosis and natural history.
[00:00:41] And so we’re going to start with a case example. And so this is a patient who I saw a number of years ago, and you can see that the top image here is a brain MRI and below is her brain MRI from three years prior. And you can see there’s brain sagging here, so this is a clear example of spontaneous intracranial hypotension.
[00:01:04] This was her myelogram and you can see she has a lateral dural tear with the leak into the epidural space. And at that time, her initial CSF opening pressure was seven centimeters of water. And so, I patched her at the spot where the leak occurred. And then she did okay.
[00:01:23] She went home, um, but overnight she developed some new symptoms and this included a non positional headache. It was worse when she was lying in bed. She was so uncomfortable lying flat that she had to actually sit up in bed, which was, you know, went against our sort of typical instructions.
[00:01:43] And she also developed some pretty significant nausea with a couple episodes of vomiting and some subjective complaints of blurred vision. I talked to her the next morning. She was doing pretty poorly and she came into the emergency room. And then we did a lumbar puncture because she was having 10 out of 10 headache.
[00:02:02] And at that time 24 hours later, her opening pressure was 28 centimeters of water. She was admitted to the hospital after I drained her down to 9 centimeters of water for observation. She did okay for the next couple hours, and again, that following evening, her headaches came back. You know, the nausea, vomiting, blurred vision.
[00:02:23] And we repeated the lumbar puncture again the next day. Again, the CSF opening pressure was 29 centimeters of water, drained her down to seven. Again, she had a little bit of a rocky night but maybe not quite as bad during this period of time. We had put her on escalating doses of acetazolamide, and eventually after about four days or so, she went home.
[00:02:47] But, and this is a particularly impressive case of rebound intracranial hypertension. But the reason I include it is because I think this really encapsulates some of the major clinical features of it. And it also points out, we don’t do CSF pressure measurements in many of the patients now who have rebound high pressure.
[00:03:07] We make that diagnosis clinically. But you can see there are clear alterations in CSF pressure that occur in some of these patients, and sometimes they’re frankly above normal CSF pressures and sometimes they’re relative increases. So, it was, like with a lot of things in this field, Bahram Mokri had originally described some of these cases. This was a paper from 2002 where he described, I believe, four cases of people who had elevated CSF pressure after treatment of CSF leaks. And there were a couple of intermittent articles in the literature. This was a short note, also in 2002, where they talk about a patient who got an epidural blood patch, and they say that the opening pressure that they measured after the epidural blood patch was 390 millimeters of water, which is quite elevated, and they attributed this to inadvertent injection of blood into the subarachnoid space, which they didn’t directly observe, but they thought that was the most likely putative hypothesis.
[00:04:15] And then here’s another paper from 2006 describing a single case report of a patient with rebound high pressure. So some of these cases have, you know, were sort of scattered around but hard to, you know, sort of glean take home points from them. So, back in 2014 we collected a couple of these cases and described them in this AJNR article, and basically this rebound intracranial hypertension is characterized clinically by a new headache type that develops after blood patching. And notably there’s a change in the phenotype. So oftentimes the most reproducible thing is that it goes from being an orthostatic headache, sort of a leak type headache, to a non orthostatic or a reverse orthostatic headache, which is worse when laying down. When you measure CSF pressure, which we don’t always do, there’s typically an increase compared with baseline, and sometimes that’s greater than 20 centimeters of water and sometimes it’s not.
[00:05:19] So a single CSF pressure measurement without any context of where what came before is not particularly helpful in either making or excluding this diagnosis. Most commonly, we see this occur within the first 12 to 24 hours after an epidural blood patch, and far and away the most common presentation is the night of the procedure, people will develop new symptoms, oftentimes, you know, one in the morning, two in the morning, after they’ve been laying down for a couple hours.
[00:05:52] So from this paper we had nine patients, and I just want to point out a couple of, I think, relevant factors, because these are all patients in whom, you know, we were just getting sort of attuned to the diagnosis, and so we did lumbar punctures on all of these patients.
[00:06:08] And what you can see is that all of them had an elevation in their CSF pressure compared to the baseline. And in, in these cases, all of them were elevated above 20 centimeters. Some of them only to 22, and then there was one patient who was greater than 55 centimeters of water. So, but relatively speaking, all of these patients had increases in their CSF pressure. Some of them we studied very early, you can see, as quickly as two hours afterwards. And some of them were substantially longer periods of time but mostly within the first the first 48 hours or so.
[00:06:44] Most of these patients also noticed a change in their headache pattern. And SIH headaches are most commonly posterior or occipital in nature, although by no means is that always true. Whereas the headaches associated with rebound high pressure, there was most commonly a shift, often to either holocephalic or frontal. A lot of patients describe pain around their orbits, or they’ll say it feels like it’s deep behind their eyes. It’s a very common complaint. At least one patient did not feel a change in their headache location.
[00:07:19] And so these are the schematic representations of where people usually feel either SIH type headaches on the left or rebound high pressure headaches on the right. Again, these are stereotypes. They’re not going to be true in every situation. What you can see in these columns is that nausea, vomiting is very common among these patients, as is subjective complaints of blurred vision. This is also relatively agnostic to how much volume you use. So this is not purely driven by a volume phenomenon.
[00:07:52] Some of these patients had volumes as low as five milliliters for blood patches and some of them had much higher volumes. And most patients in this particular series were on acetazolamide for several weeks although some patients were on them for substantially longer. So, one of the things we want to know is not just, you know, what happened in these nine patients, right?
[00:08:15] We want to know how often this occurs. And how often should we expect that patients who undergo an epidural blood patch will have symptoms like this? So, this is some unpublished internal data from Duke. And this is all consecutive patients under a certain period of time who underwent epidural blood patch.
[00:08:33] And they all got followed up with a structured note that occurred on the post procedure day. Number one, and we graded their headache severity as either mild, zero to four; moderate, five to eight; or severe, a nine or 10. And what you can see is that in patients with SIH, about a little over 50 percent of patients had some level of rebound intracranial hypertension. But it was only about 20 or 25 percent of patients that had moderate or greater headache. So, a lot of patients would have mild, very transient symptoms that would resolve by the next by the next morning and really require no treatment other than head elevation and maybe some Tylenol.
[00:09:19] And then we also looked at patients without a verifiable CSF leak. For lack of a better term, I call these CDHO, which is Chronic Daily Headache with Orthostatic Features. We’re not really sure what their ultimate diagnosis is, but we know that we haven’t been able to confirm a CSF leak, either by brain imaging or spine imaging. And the proportions were about the same: in 84 patients who we followed up who had an empiric epidural blood patch, they developed rebound intracranial hypertension in about the same proportions and with about the same number of people having moderate or greater severity.
[00:09:56] And we also looked at our patients who had post dural puncture headache and who had epidural blood patches for that. Now these are a smaller N, so this is only an N of 10, in this particular case, but you know, still about 50 percent of these patients had moderate or severe rebound intracranial hypertension that required treatment with either acetazolamide or lumbar puncture.
[00:10:20] And so, you know, what’s sort of the take home message from this data. What I tell folks is that the incidence of rebound intracranial hypertension is about 50 to 60%. And that most people who get epidural blood patches are either going to have no rebound or mild rebound. Self-limited that we don’t need to treat. About 20 percent are going to require some sort of intervention, and that’s usually starting off with acetazolamide or methazolamide, which you’ll hear more about from our next speakers. And that this really occurs whenever epidural blood patch is performed, regardless of whether or not we see a leak on imaging. So it’s probably to some extent, some non-specific effects of compressing the epidural space with blood.
[00:11:11] Now, in this particular data set, we didn’t really— because all of all these assessments occurred day one, we didn’t really assess how often the onset was after day one. And we didn’t really assess how often or how long the symptoms last. But Dr. Schievink and his group published this article, and they looked at all cases of SIH, either surgical or percutaneous treatment. In this paper, they defined rebound as a reverse orthostatic headache that was different from the original headache, which makes sense. They also included that it had to resolve with acetazolamide. So all of these patients had to meet the threshold for acetazolamide, which is important because the mild cases that, you know, had resolved by the next morning, they wouldn’t have necessarily included those patients in their statistics in this paper. So the numbers are seemingly a little different, but in the end, they turn out to be pretty close to the same, and not better accounted for by an another cause.
[00:12:08] And they looked at the effect of venous stenosis. In this case they found some differences. That’s not really what I want to focus on. For this talk, I want to look at the clinical characteristics of rebound. So they also found that the headache phenotype changed in most patients. They went from occipital or sub occipital to frontal. Twelve patients. Generalized headaches to frontal in five patients. suboccipital to generalized in three patients, and then a couple of other sort of smaller variations.
[00:12:40] But you can see, again, most of the patients either had generalized or suboccipital headaches in the beginning, and most of them noticed a change in headache phenotype. And then, what do they find about the onset? Well, 74 percent of those patients in their study were within the first 72 hours. 19 patients were three to seven days afterwards and 6 percent were between a week and a month. Or an overall incidence of 27%.
[00:13:14] Again, those would have been the patients who we classified as moderate or severe, which again was about, you know, 20-25 percent in our series so pretty similar numbers, I think. They looked at papilledema. This is often a question: do you have to have papilledema to get this diagnosis? Only two of 25 patients in their group who had fundoscopic exam had papilledema. So the minority of patients.
[00:13:39] And then what about resolution? Well, most resolved within six weeks. A couple lasted beyond six weeks, up to three months. And then one patient got a shunt and one patient was on still on acetazolamide after 15 months follow up. So, you know, I think in terms of the incidents, I think this is common. I think we can reassure patients that most of the time, when this occurs, it’s going to be mild and relatively self-limited, but there are some patients who have more moderate or severe symptoms. And we have to be prepared to treat that.
[00:14:17] And we also have to recognize that it’s not rare. It occurs with relative frequency. And so when it occurs, it tends to occur early in the treatment. So you should really think about sending your patients home with some sort of treatment plan, if they’re treated as an outpatient, for what to do if this occurs, because there’s a reasonable chance that it might.
[00:14:40] So what are the questions that I ask? I usually just start out by asking, you know, how’s your head feeling today? This is after they’ve had their procedure, just sort of an open-ended question. If they say they have headache, I ask them, does this feel like what you had before the patch or is it different, right? Because we know that rebound intracranial hypertension often has a change in character. Is it better or worse in any particular position. And I try not to ask too many leading questions but, you know, if it’s worse when laying down or really has lost the positional component, I think about that as sort of a change in the character because rebound is worse with recumbency or is non-positional. Where do you feel it in your head, to get a sense of, you know, where it’s located. Very often, but not always, rebound is worse in the frontal or periorbital locations. Any changes in the vision? And any nausea, because those are two common complaints as well.
[00:15:40] When people present with rebound intracranial hypertension, I think it’s good to at least consider two other potentially serious complications. And those include venous sinus thrombosis and subdural hematomas. Now these complications typically occur as a consequence of undertreated CSF leaks. Both of these are associated with persistent low pressure, but both of them can cause focal neurologic deficits and subsequent high pressure.
[00:16:11] So if I know I’ve really done a really good blood patch or something like that, I know that these are less likely to be the causative explanation the next day. But certainly, you know, I would want you, if you’re not totally familiar with rebound high pressure, to think about questions like, does this patient have new focal neurologic deficit? Is there something about the headaches that are really severe?
[00:16:32] If the headaches are really severe, you may want to consider doing some neuroimaging, just to make sure that they haven’t had one of these complications. I don’t think that routine neuroimaging is needed for the vast majority of patients with rebound intracranial hypertension. I think those patients can just be treated on the basis of their symptoms alone.
[00:16:51] So in summary, I think I would just say, regarding the natural history and the diagnosis, that rebound is common, but usually mild. About 20 percent of patients will need to be treated. So you need to be prepared to treat those patients and you need to counsel patients to be on the lookout for those symptoms so they’re not surprised by them and they know what to expect.
[00:17:12] Look for a change in the headache phenotype, which occurs in the majority of patient. The onset and peak typically happens very early, so you want to have a treatment available at discharge. And it can happen after any type of epidural blood patch regardless of diagnosis. So it’s not particularly helpful from a diagnostic standpoint.
[00:17:31] And you should not assume that if you’re doing an epidural blood patch for post dural puncture headache or something else like that, that you will not have to deal with this problem because you will see it. So, thank you very much for your attention.