Slides
Transcript
Okay, thank you. I’m happy to be talking about this topic. I think this is perhaps the most important topic that we have in this field at the moment. And I do want to just start with a disclaimer. I would say that, you know, this is an exceedingly complicated and at times very emotional topic, and I’m going to be talking about some data from different papers and things like that in this talk. Those data represent averages, but averages are not individual people. And so, while I say some things that are in broad generalities, that doesn’t apply to every single person. Nothing that I say should really be construed to supersede decisions that individual patients make together with their physicians, because ultimately each person’s situation is different, and averages are just an aggregate representation.
So, briefly, I’m going to talk about my approach to imaging negative cases, which is:
- The brain imaging – is it really normal
- Have I done enough of a workup?
- Have I considered an alternate diagnosis that may explain symptoms?
- Should I treat anyway?
So, some other speakers have talked about this. I’m not going to dwell too long about it, but brain imaging – is it really normal? This is a case of a patient who we saw, and the brain imaging was reported as normal. If you look, there’s really not much in the way of dural enhancement, and I think if we did all the measurements in the Bern score, it would not be particularly impressive. But this patient did have a dilation of the transverse sinus, and sometimes this can be a relatively subtle sign. But sometimes this may be the only sign that there’s intracranial hypotension. So, while this patient was initially thought to have negative brain imaging, if I was looking at that, I would say, no, this is definitely positive brain imaging.
This was the patient’s CT myelogram, and you can see that there’s one of these lateral dural tears that’s partially contained. It turns out that patients with contained epidural leaks – whether they’re ventral or lateral leaks – have a much higher rate of having negative brain imaging.
This is another patient who I’m not showing you the transverse sinus on, but it was normal. There’s no brain sagging, there’s no dural enhancement. I think this case really was normal brain imaging, but this patient also had a CSF leak. So, this is the myelogram, and you can see that there’s this small dural tear laterally in the midthoracic spine.
This patient also had what I would consider to be a normal brain MRI. No real signs of brain sagging, no dural enhancement, no venous distension sign in this particular case. This patient was found to have a CSF-venous fistula.
So, sometimes there are cases where normal brain imaging is associated with abnormal findings on the myelogram. I think we have to make sure that we’re diligently looking at those. And in particular, when patients are sent to me, I want to see the first set of brain imaging that they had and review that prior to saying that the brain imaging is negative.
I think for those who don’t do this a lot, or even those who do do it a lot, the Bern score – or the score that was elaborated by a number of the people who are here today. That came out of that – that came out several years ago, in 2019. I think this can be really helpful. It can be helpful as a way of sort of systematically approaching findings. And so, particularly for those who are new to the field, I think this can be a way of helping to avoid the pitfalls of calling a study negative. And although the Bern score doesn’t provide a dichotomous assessment – positive or negative – it focuses your attention on these particular areas that are important to review.
So, the second point is: have I done enough workup? This, in particular, is a question of, you know, have we done an adequate job of doing myelography, or what we consider to be advanced myelography? And advanced myelography just means something besides the conventional myelogram – a myelogram that’s specifically targeted to look for CSF leaks.
Dr. Schievink just gave a really nice overview of this topic. Just to repeat some of the things that he talked about. They did a nice paper that came out in 2021 in Headache, and they found a 10% yield of patients who had CSF-venous fistulas but who had negative brain and spine imaging. As he mentioned, all of these patients had undergone epidural blood patch prior to their DSM. In patients who had diverticula, they found a venous fistula in 19% of those patients, and 0% without.
And the CSF-venous fistula was not always from the largest diverticulum. It was only about a third of the cases. But these were among patients who had a very high response to epidural blood patch, whether or not they did or did not ultimately have a fistula. This was a smaller study from the Mayo Clinic, and again, as Dr. Schievink mentioned, these were people with a Bern score of two or less. Much smaller number – 9%. They didn’t find any venous fistulas in that population. Again, a limitation is it didn’t address the question of, you know, how often does this happen in the negative brain MRI population. It didn’t exactly address that question. It was just in patients with low Bern scores.
Then we published this earlier. This was the work of Jacob Gibby, who’s one of the fellows at our institution. We looked at a year’s worth of data among all patients who presented – consecutive patients over the course of a year. And what we found was that among patients with positive brain imaging, we were finding a venous fistula in 73% of those. And among those with negative brain imaging, we didn’t find any definitive fistulas in that patient population. But interestingly, if you had an epidural leak seen on imaging – so if you had epidural fluid – only 78% of those patients had positive brain imaging. So, at least in this particular cohort, which is just a snapshot, all the patients who had a leak but with negative brain imaging were accounted for by epidural fluid collections.
Now, I should say that I don’t believe that 0% of patients with negative brain imaging have CSF-venous fistulas. I just showed you a case of one who did, and last week I saw another one. But I think these kinds of numbers can be helpful in counseling patients – you know, what should you expect when you’re getting a myelogram?
You know what – what is our expected diagnostic yield for that test so that you can go into that with full knowledge about what to expect. So where does that leave us? I think that the consensus would be that in strongly suspected SIH, you should have brain and spine imaging. I don’t know that we’ve reached a consensus on what type of spine imaging is sufficient for a complete workup, but certainly if you go to a CSF leak center – if you come to my center, you come to most of the other centers here – you probably will and should get advanced myelography in a lot of cases.
I think it’s really important to emphasize that negative imaging does not mean nothing is wrong. It just means that, using the technology that we have right now, we’re not able to show exactly what that pathophysiology is. And in some cases, that may be a leak that we can’t see, and in some cases, it may be something else.
More workup is not cost-free. There are medical, social, and financial risks. And we’ve heard from patients today about how difficult it can be to go through some of these things, and I think ultimately we need more science to guide.
The next thing I would say is it’s important to think about alternate diagnoses. This is a challenging case that I saw. This patient – I’m going to play this video – clear CSF-venous fistula coming off the sacrum here. This was to a venous malformation. This patient had multiple treatments, and eventually, as best as we were able to detect on imaging, had this fistula closed. The brain imaging normalized, but the symptoms persisted.
There was a very frequent syncopal component. This patient was passing out multiple times per week and feeling pre-syncopal very often. CSF pressure had gone up to 30 cm of water. And so, as best as we can tell, there wasn’t any persistent, ongoing leak. The pressure had gone up, the brain imaging had normalized. But still, this patient was symptomatic. I repeated a myelogram both under fluoro and under CT and didn’t see anything there. But because she was having all these passing out-type symptoms, I did do an active stand test on her, and her heart rate increased by 50 beats per minute, which would meet criteria for POTS. And so, that was the direction we took in terms of treating her. In my experience, patients who have really frequent syncope like that – that should be a really red flag that they may have coexistent dysautonomia.
So we haven’t done a great job of elaborating what the workup should be for patients when we’re considering alternate diagnoses. I definitely think that POTS, or postural orthostatic tachycardia syndrome, is something that we should consider. I think you can easily implement an active stand test at your institution if you’re a physician. It takes about 10 minutes – well, 20 minutes – for the patient to lie down to get a baseline and then a couple of heart rate and blood pressure measurements. And if they’re abnormal, that may be something that can be pursued, particularly among patients who have a lot of syncope or pre-syncope episodes.
We’ve seen lots of patients who have cervicogenic head pain, where they get pain when they’re upright due to axial loading. Both POTS patients and cervicogenic headache patients will often have temporary responses to blood patching. We’ve seen that a number of times, and so that can be confusing and can lead people in the wrong direction sometimes. IIH – we’ve seen lots of paradoxical cases now where people have headache that’s worse when they’re upright, and that’s not the typical pattern with IIH, but nevertheless, we have seen that in a number of cases, so it’s something to think about.
Cranio-cervical instability is something that we talk about, particularly in patients with connective tissue disorders like EDS. This is a difficult diagnosis, and I’m not going to spend too long going into it, because the treatment for this is fairly invasive – it involves fusion.
Then there’s this diagnosis of new daily persistent headache, which is a primary new daily persistent headache, and it’s really a diagnosis of exclusion.
And then, should I treat anyway? I think this is a really important question. So, if brain imaging is negative, and you haven’t been able to establish an alternate diagnosis, should I perform epidural blood patching anyway?
At least until recently, the state of research in this field – which is a super important question – has been like a desert. There just really hasn’t been that much to guide us in terms of what to do.
There was this paper from 2022 that came out of Korea, and they basically reported that in patients who had epidural blood patches, 90% to 95% of them were getting permanently cured. If this were the case, we wouldn’t be having this conference, right? Because this doesn’t match the experience, I think, in North America at least, and I think not in Europe either. There seems to be something different in Asia – maybe it’s earlier intervention or earlier treatment, or I’m not sure exactly what it is – but I don’t find this to be representative of my patient population.
But then Dr. Carroll and Dr. Callen and some others published this really important paper earlier this year, where they looked at long-term epidural patching outcomes in patients who were imaging negative. It was a really well-done study, and they had a lot of really important indicators. They found that patients received a mean of about 3.5 blood patches, and they found that there were some improvements in PROMIS physical health, mental health – although that wasn’t statistically significant – and headache impact test, both in the early and in the delayed, improving from a 67 to a 64.
Here, you can see those improvements. I think the numbers can be a little hard – I’ll show you these graphical representations. Their conclusion was that there were moderate rates of sustained clinically meaningful improvement.
These are the PROMIS measures. They’re based on what are called t-scores, which are statistical distributions. They saw small but significant changes in patients’ response to physical health after receiving blood patches, a non-significant change in mental health. And then in the headache impact test, a decrease of about three points, from 67 to 64.
Now unfortunately, on average, these patients were still in the most severe headache category after an average of 3.6 epidural blood patches. And I think whether or not you want to go through three or four blood patches – I think that question is in the eye of the beholder. That’s a question that really needs to be answered between individual physicians and their patients.
But I would say that having this information undoubtedly improves our ability to counsel patients and give them an idea of what sorts of things, on average, they can expect. I would say that wherever we are with this, it’s not far enough, and so we want to see more of this in the future.
But I think you should be optimistic because this paper represents a really strong commitment to understanding what to do when imaging is negative, and that’s something that is a high priority for our field. It was very well designed and executed. As I said, it’s the most difficult and important question in the field currently. Answers are not going to be found based on beliefs or hunches or just personal opinion, but with science like this. And I think we will see more studies like this.
So, in summary, I think epidural blood patching may be an option. I think it’s important to be clear about what to expect, what we’re going to call success, and to set those things out in advance before doing this. If it’s working, keep doing it. If it’s not working, try something different. Ultimately, I would say response to epidural blood patch is a poor diagnostic tool, at least in my opinion, but it can be helpful for some patients, and so that may be the most appropriate thing for some people.
I think it’s really important to talk about benefits and harms. This was a patient who I saw, and this patient had had seven embolizations when they were referred to me from an outside institution. The patient was not better in terms of symptoms, and so this is, I think, what we want to avoid because there are harms in over-treatment sometimes, and those include tunnel vision or failure to pursue other diagnoses. Post-procedural complications, financial implications are a real big deal, and we don’t talk about that enough. We heard from patients today about anxiety about procedures and prognosis, radiation risk, something we talked about yesterday, patient time and fatigue, and ultimately physician access – all are important.
For me, the goals in dealing with this issue are to, number one, understand the diagnostic tests that we have, and that includes both the strengths and the limitations. Understanding those strengths and limitations helps you counsel patients on what to expect.
I think it’s really important to be honest, even if that’s uncomfortable. People have been through a lot by the time they come to see you, and you owe them, as a physician, your honest opinion. But you also have to be honest with yourself about things that make you uncomfortable – situations that may be difficult for you, things that you don’t understand, and be cognizant of those things.
I think it’s really important to respect your patient’s journey because whatever is bringing them to your office usually involves a massive amount of suffering. I mean, people don’t come to see me just because they want a vacation to Durham, North Carolina, right? They’re coming because they have real big problems, and I think it’s really important to respect that. But again, you also owe them an objective assessment of what’s going on and what you think is going on with their case. I think it’s important not to blame or criticize others too harshly when we’re dealing with an environment where we really have a lot of uncertainty.
There’s no road map. There’s no textbook for this, and different people are going to approach this differently. So, I think it’s important not to be too critical. I think it’s important to set clear expectations about what’s going to happen and how we’re going to judge success, and then proceed practically. Think not just in terms of what hypothetically might be happening, but what are we actually going to do, and how are we going to take next steps to get people better. So that, in another nutshell, is my approach to negative imaging. Sorry for running over a couple minutes. Thank you very much.