Transcript
Question: Thank you for these wonderful talks. I had two questions regarding the radiological follow-up after treatment. So the first one, as Dr. Carlton Jones said, if there’s no spinal epidural fluid left radiologically then we are happy, and we also know that this has been touched upon, that some patients remain symptomatic after treatment of the leak. I’ve sometimes had difficulty convincing my patients that the leak is closed because they’re still symptomatic, and I usually do this by repeating the spinal MRI, and if the fluid is gone that’s for me the reason, the evidence the leak has gone. But do you think that a normalized spinal MRI in SLEC-positive patients is sufficient to prove the leak is sealed, or do you repeat myelograms in these persistently symptomatic patients?
Answer:
Dr. Carlton Jones: That is a great question. I think we do have a proportion of patients who are still symptomatic even after surgical repair, who don’t have fluid radiologically, and we will do myelograms for these patients and we don’t seal it. I think that has been the experience of some other centers, that sometimes patients can get better initially and then they become symptomatic again. But I think as a radiologist the best tool for excluding a leak is myelography, and so sometimes we will do that.
Dr. Schievink: We get an MRI on every patient within 48 hours before they’re discharged from the hospital, at least after surgery, right? For any type of leak that’s caused by a dural tear, right? So ventral, lateral, dorsal. Of course, not everybody wants to do that within two days of surgery, right? Their back hurts. So then we get it when they get home. I think MRI is really, really good for looking at extradural fluid. If we’re still convinced that there’s a chance that they’re still leaking, then we will repeat a digital subtraction myelogram. But I can’t really remember anybody who had a leak after the MRI was normal. Can you? For CSF-venous fistulas, of course it’s a little bit of a different scenario, because to make certain that that’s gone we have to do another invasive procedure, because we don’t have non-invasive imaging looking for CSF-venous fistulas. So if we feel that there might still be a fistula, then we would repeat the invasive myelogram. I know that Dr. Mamlouk, who’s here in the audience, he routinely does invasive myelography to make sure that the fistulas that he treats with fibrin glue injections are completely gone. Isn’t that right, Dr. Mamlouk?
Dr. Mamlouk: We used to start off with post-myelogram after patching, but then now we typically use MR brain just as a surrogate. But if there is a question, like as you mentioned, a high concern for recurrence, then we’ll do a follow-up myelogram.
Question: Thank you. My other question was about the reverse scenario. In some patients I’ve seen that they had very good clinical results with patching, but then their spinal MRI after three months, after a year, still has this large epidural fluid collection and it does not resolve. I am convinced that these patients still have a leak, but some of them really are symptom free. And because of these worrying long-term consequences, the superficial siderosis, I doubt. Should we still refer those asymptomatic or oligosymptomatic patients for surgery to definitively close the leak and to avoid these long-term complications? And if not, how do you follow them up? Do you do yearly MRIs, myelograms, stuff like that?
Answer:
Dr. Schievink: Yeah, we published a couple of articles about the long-term risks of developing one or more of those complications, particularly in patients with ventral leaks. And the vast majority of those had no symptoms anymore from their leak, or just very minor symptoms. And then after 15 years, right? So if you have a ventral leak, after 15 years there’s about a 50% of having developed superficial siderosis. The risk of spinal cord herniation is about 10%, and the risk of this bibrachial amyotrophy is about 10%. As Dr. Brouwer showed this morning in their series, I think you said like 10–20%. All right, you’re sitting there. 10-20% of patients either just had a history of a headache or never really had a headache that they remember.
Dr. Brouwer: That’s correct, yeah.
Dr. Schievink: So if it’s somebody who is fairly young, let’s say less than 80, and they have a persistent leak, then I do recommend surgery for that, just because we know what the risk is of developing these complications. We know that the risk of surgery is relatively low. If for some reason they prefer not to have surgery, then we get yearly MRIs to look for superficial siderosis. We more or less carefully explain to them what symptoms to look for. And I think your idea of doing maybe yearly audiograms is a really good idea as well.
Dr. Carlton Jones: Have you had any cases of lateral type leaks with longer-term complications such as siderosis? Because obviously in our paper we didn’t see that. So I do agree that the most important follow-up in these dural tear type leaks of any cause is the spine to see if there’s still fluid there. But I’m interested to hear Wouter’s experience.
Dr. Schievink Yeah, we haven’t seen anybody with a lateral leak develop superficial siderosis. And we think that might be because when you have a ventral leak, right, there’s an epidural plexus there, that’s a really robust epidural plexus. So that tends to bleed more, I think, we have really no scientific proof for that. There’s some type of calcification that might injure the vessels there that prevents that from clotting. We now have seen a few dorsal leak patients with superficial siderosis, and we also can get superficial siderosis from CSF-venous fistula. So that means that it’s not just one direction, it’s not just from CSF space to vascular space, but you also can get flow from vascular to intradural space.
Question: We see patients after neuraxial procedures who have PDPH, and we typically—I’m an anesthesiologist—we typically treat them ultimately with blood patches, sometimes a second blood patch. Fortunately, most of them get better, but not all of them. When should we be referring them to you?
Answer:
Dr. Beck: So I have to take this difficult question. We don’t know. And these patients that are reluctant to the third or second or fourth and do a blood patch usually end up sometime in our service in a specialized center. We do extensive imaging, and really basically we do not find anything that we know how to handle. There’s this dinosaur sign. We don’t know the sensitivity and specificity of it. We find—rarely you find—findings that you can have, and there’s one exception. The needle punched through both dural layers also to the ventral side of the dura. In these cases, you almost ever see a ventral SLEC, a small tiny ventral SLEC. Then we are happy because surgery is easy, straightforward, we close it, the patient is fine. In a non-shot-through ventral leak, post-dural puncture patient, it’s difficult. And then we have the group of patients that end up with surgery, and surgery is not dangerous, it’s easy, it’s extra-durally, and we almost always find something. But it’s difficult—sticky membranes, a lot of vessels. We are happy we can remove, but then still patients only improve in 60% maybe 70%. So, difficult cases. We are trying to write this up, to work this up, to build registries. I don’t have a definitive answer, but only shot-through ventral leaks—this is straightforward, close it in patients.
Question: So going back to the previous discussion on Dr. Carlton Jones, you mentioned that there is increasing experience and evidence of sort of serial CSF-venous fistulas, that you patch one, you find another, which to me raises the question of whether after the initial procedure there’s some role for pressure testing. Should we be looking for a way to identify other weak spots? Is there any experience with that? What’s your opinion of that?
Answer:
Dr. Carlton Jones: I would say that we have found more of these. I’d say probably my myelos are a lot better than they were several years ago. So maybe that’s why I’m seeing so many of them. Speaking to a lot of other people at other centers, I think that is also the experience. And it does raise the question what is physiological for that patient. When we have these multiple CSF-venous fistula patients, we don’t treat all of the fistulas. We treat what we think is the biggest. Whether that’s right or not, we don’t know. But that’s the question that I raised in the end, which is, if these patients are going to keep recurring, are we going to do five, six, seven embolizations or bone removal? No matter, even if it’s a really tiny hole, it’s still bone removal. And maybe we need to tread gently with some of these patients. This is not all, this is just some patients.
Dr. Schievink: You make that sound like, “Oh my god, there’s some bone removal.” I think endovascular embolization, right, when you see those pictures afterwards, like the whole venous—well not the whole venous system—but so much of the venous system is gone, right? It’s full of this Onyx. There’s no flow going through it. And I think, Dr. Maya, I think that’s the reason why there’s probably a higher frequency of rebound, because you just got rid of part of your venous system that absorbs normal spinal fluid. But I recommend patients who have had a recurrence somewhere else along the spine. I tell them either they can take lifelong Diamox, acetazolamide. Most people don’t like that because that has a lot of side effects. So they can take a homeopathic dose. Especially here in Europe people are very much into homeopathic medicine, even in veterinary medicine actually. So they can take a homeopathic dose, or they can drink some dandelion tea. Dandelion tea is a really mild diuretic, and that’s much better tolerated than Diamox. We’ve been looking at if there’s a relationship between compression of the internal jugular veins or any kind of venous stenosis, that those patients might be at higher risk of a recurrent leak. But the follow-up of that has just been a year, maybe 18 months. So maybe in the 2035 leak session we’ll have some results of that.
Question: Hello. After a blood patch, some patients will improve immediately and some patients do not improve their symptoms immediately. They tend to wait days or weeks before improvement or do not improve at all. In your experience, do you have a delay to consider after a blood patch or any therapeutic procedure until considering they failed the treatment? In our experience sometimes we wait a month until considering they fail, but maybe it’s too long. What is your experience about considering the failure of the treatment?
Answer:
Dr. Urbach: In my office they should improve immediately. If they do not improve immediately, something is wrong.
Unknown Speaker: Even some people… Immediately, you mean one day, two days?
Dr. Urbach: One day, next day, yeah.
Dr. Mamlouk: At least in our experience it’s usually not immediate, and it can be up to about a couple of weeks or so, particularly for the spontaneous leaks. I think if someone recently had a dural puncture and then you do a patch, I think those patients can respond the next day. But at least in our experience it’s usually not the next day.
Dr. Schievink: I think it’s just a different mechanism, right? When you place that blood patch, you basically introduce a hematoma in the spinal canal and the pressure will go up. So I think a lot of patients have an immediate effect, like what Dr. Urbach talked about. And I think there also can be a bit of a delayed effect, like maybe some inflammation around the dural sac, some scar formation, that the CSF might still be leaking but just doesn’t escape into the soft tissues anymore. I think it can be kind of bimodal. I tell patients if you’re not any better after two weeks, at least you don’t have to take precautions anymore after your blood patch.
Question: Thanks all the speakers for their excellent talks. I’m originally from South Korea and I have one comment and one question. My comment is on the application of the ICHD-3 as mentioned by Dr. Brouwer. Actually, the ICHD has no effort, because the purpose of the ICHD is classifying headache into secondary and primary headaches. We apply this to patients with headache, and if the patients have a headache and some kind of disorder, we classify if this headache is attributed to this disorder. That is the purpose and structure of the ICHD. So, it contains a headache in the first line, but it’s not for diagnosing SIH itself. So, please don’t be confused by it. And my question is to Dr. Carlton Jones. Do you recommend the fibrin sealant patch as a kind of first line to patients with ventral dural defect over the blood, or do you think it’s a kind of second line after blood patch?
Answer:
Dr. Carlton Jones: I don’t do any blood patching for fistulas. I don’t think that they respond. I don’t do any non-targeted blood patching really at all when we know what we’re going to find, really. I think non-targeted blood patching has a role in certain patients like very sick unwell patients, brain sag dementia patients where we can’t find the leak, and some post-dural puncture patients. For fistulas, we know that we’re going to find a fistula most of the time, so we have already had this discussion with the patient as to what the available treatments are. Then when we find the fistula we can offer this the same day in not all but a majority of patients. Some patients say no, I don’t want to do that, I want to go straight for a surgical repair or a transembolization, and that’s fine. We work with the patient for whatever their wishes are, but most patients at least in the UK want to have fibrin glue injection.
Unknown Speaker: Can I just step into this question of blood versus fibrin? Isn’t it just the compression of the vein that causes a leak to close? So could blood be an option when you are scared of using fibrin in this location?
Question: Hello, thank you for the excellent talk. I have a question regarding CSF-venous fistula as well. We heard about fibrin glue, CT-guided, and endovascular embolization. My question is if you plan to do the fibrin glue and you hit the vein, what do you do? I had one patient we used glue [inaudible] in the end because we didn’t want to inject the glue into the venous system. But is it feasible in your opinion?
Answer:
Dr. Carlton Jones: I showed examples there where we have had intravascular spread, and most of the time the fibrin will clog in that vein. Sometimes we have seen it, like I showed those examples of tiny asymptomatic silent PEs. We see that with embolization as well. We obviously monitor the patient if that happens. But that’s why I said we also don’t just give 4 mL of fibrin in the foramen, we do it as a staged. We give a little, we make sure we do test ep air and epidurogram contrast, and we move the needle appropriately depending on what we’re seeing our spread being. We haven’t had any adverse effects from this.
Question: This goes to Professor Schievink—just a technical question for your endoscopic technique. Did I get it right that it’s a strictly extradural approach, or you go transdurally as well?
Answer:
Dr. Schievink: No, it’s strictly, strictly extradural. But because it’s a 30-degree angled scope, you can actually look underneath the dura, underneath the dural sac, and that’s why you can visualize the dural tear from the outside rather than the inside.
Unknown Speaker: And do you try to squeeze a piece of TachoSil inside from intradurally, or you just patch it from extradurally?
Dr. Schievink: We try to squeeze it through the tear. But again, we’ve done it less than 10 times I think. But tomorrow at the very end of the day, there’ll be a talk about that from the experience from South Korea where they’ve treated quite a few patients that way.
Question: Yes, I have a question to Dr. Maya. You opened a new field with azygos vein stenosis, and you said it’s really important to image the venous system, the azygos vein system, using iron oxide particles. Unfortunately, they are not allowed in Europe, so it makes a difference. But if you do an angiography of the azygos vein system, we sometimes have the impression there could be also valves, venous valves, vasospasm. What would you say? If you have the venograms and the angiograms, are they compatible, or is it if you inject via the catheter you feel the spasm that’s different from the venogram, from MRI perspective?
Answer:
Dr. Maya: We found that the MRV is correlating very well with the angiogram. The question though is, do we see sometimes azygos veins which are looking narrow but in reality, you do the angiogram, there is no narrowing? That can happen, and that has happened not just with azygos but also in the internal jugular vein and SVC. And we tried pressure measurements, which may or may not be helpful — not very helpful in reality. So there’s no convincing answer. There is good correlation, but there are some false positives yes.
Dr. Urbach: Who uses MR Venograms in Europe? It’s not allowed. I think it could be an off-label use. Is there anyone in the room who uses MR venograms with iron oxide particles?
Unknown Speaker: And without iron oxide particles?
Dr. Urbach: Obviously not. But in the US many use it. To the US community: who uses it in the USA? Only, only Cedars.
Dr. Maya: That’s correct. I think other centers I know they use CTV to a certain extent, and it’s quite good. We chose MRV because number one, it’s less radiation. Number two, we thought it would be a very good depiction of it, better than CTV, and it doesn’t rely on the iodinated contrast. So we have not had head-to-head comparison to compare these two, but CTV is an option. Thank you.
Question: So we have the persisting observation that the presentation of brain sag dementia or bibrachial amyotrophy is far more common in men than women. So other than the obvious point demonstrated by our panel here, that women are constitutionally smarter and stronger than men, does anyone have any thoughts about anatomy underlying this phenomenon?
Answer:
Dr. Schievink: Yeah, I mean I don’t know either, but it is kind of an interesting observation because some of the other long-term complications, like what Dr. Brouwer mentioned with the bibrachial amyotrophy, that’s also much more common in men than it is in women. While if you take the entire population of patients with spontaneous leaks, women are clearly overrepresented. That’s one of the 25,000 questions that we have about CSF leaks that remains unanswered.
Unknown Speaker: Actually, I think we should not only consider differences in anatomy, but differences in physiology between men and women. They might be more responsible for the differences we find.
Question: But actually I had a different question about the venograms. You do these in lying down position. You do these with ventilation—positive pressure ventilation. Are the venograms you produce not affected by these overextended venous systems because of the position and of the positive pressure ventilation?
Answer:
Dr. Maya: For the MRVs?
Unknown Speaker: It’s easier to embolize.
Dr. Maya: The patients are awake, they’re breathing on their own, they are lying down. There is no upright MRI. But for the venograms, they’re under anesthesia. That’s correct.
Unknown Speaker. It makes the embolization easier of course, but maybe it visualizes if you don’t embolize you see maybe more than there is.
Dr. Urbach: But it makes a difference. They do it in general anesthesia, so we get higher range of CSF-venous fistula. We do it with pre-pressurization, get the same or this higher range with that increasing the intrathoracic pressure.
Dr. Beck: and during surgery it’s the other way around. When the patient is under positive pressure ventilation and lying on his belly, you almost don’t find these lesions. It’s a lesion that is so clearly depicted it acts like a vessel. You see how the contrast is oozing out like in a vessel. In the CT myelogram, patient is awake. Patient under positive pressure airway ventilation on the belly, you don’t. I had to push the contrast into the vein because everything—I think it’s vice versa. You find it, it’s more difficult to find it in positive pressure ventilation, and even you do it with a forced inspiration in CT. In CT it’s the opposite I think.
Question: I’m a patient. I have kind of a two-part question. One is a few of the talks sort of highlighted the differences between the terminology used between institutions. So the type 1, 2, and 3 slightly different across institutions. Is anyone working towards standardization of data and classification systems internationally? And do you think there might be potential for clinical phenotyping of cases based on leak type and location, that eventually we might understand purely from the clinical presentation whether the patient is likely to have a specific type of leak or a specific location of leak, and that that would streamline their diagnostic process so that they use fewer resources of the healthcare system, for example. Sorry, it’s an open-ended question to all of you.
Answer:
Dr. Schievink: I think Dr. Beck, who gave an excellent presentation about different types of leak, will answer your question.
Dr. Beck: Thank you, Wouter. Of course, we are trying to work to harmonize this and as Wouter has shown you, probably we need to take into account also the anatomical view. And what I really did like is Horst’s depiction of the tube tire, that we need to discriminate whether it’s an external force or it’s a weakness of the tissue. Sorry for not being faster. Of course, we try to get better. But I think we need neurologists to answer this question finally.
Dr. Brouwer: So we’ll be happy to work on this in the coming years.
Dr. Matharu: No, I think what Brouwer said is right, that we do need to define this better. And even though different people are calling it different things, I think at the end of the day the bottom line is a lot of us know what we are talking about. But there is no doubt there’s a need for harmonization. I think the next question speaks to is the presentation of the different types of leaks going to be different, and can we actually start predicting who is going to be in which category.
The challenge over there is going to be for somebody who actually works in AI. A lot of, in fact, this is a sidekick for what I do. A lot of my work is in fact right now in AI. And we have, for example, phenotyped 2,000 patients with migraine, for example, and the kind of thing we’re trying to do is can we, from phenotypes, actually start picking out what a person is going to actually have as a final diagnosis?
And the difficulty, and I think this is also being graded as been mentioned by ICHD criteria, which is really based on headaches only. And we know that, in fact, the study that I was presenting was that only 75% of the patients actually presenting with orthostatic headaches, despite what the literature says, and that’s just a bias in literature. What we have is very diverse manifestations, and that is the point I was making in my presentation earlier on. And the question that we are really going to need to answer in the long term is going to be, is there going to be a particular phenotype that is going to go with a particular radiological finding?
And the problem is going to be that there are going to be huge overlaps no matter what. And with those overlaps, our confidence intervals of saying it is more likely this and that are going to be very, very small. So I don’t think that is going to really give us the answer in the long term, but it is something that is going to really need to be done for us to get to that point.
I think the more important issue is that where we need to do the education is with our neurology colleagues, who often just go to the conclusion that if you do not have an orthostatic headache, you do not have a CSF leak. And that is what we need them to move away from and start appreciating—this is a very heterogeneous presentation. There’s also the point that was being made earlier on that not having an orthostatic headache does not mean that you do not have a leak, yeah? That there’s a disjoint between the radiology and the clinical presentation, and that just needs to be appreciated better.
And just coming back to a point that was being made earlier on, the reason—just to coming back to your question in terms of the reason there’s a difference between male and female—I think is actually physiology. I give an example: men and women or boys and girls before menses kicks in women have the same rates of having headaches, primary headaches. So that’s migraine, because the genetic load is exactly the same. Hormones kick in, and from the time that the women are menstruating to the time that they stop, women outnumber men in terms of presentation of headache by a factor of 3 to 1. And yet the genetic load is the same. Once menopause kicks in, the rates for women drop down again.
So what has happened? The biological, the load, that genetic load you had is exactly the same in migraine. But what was driving it and what drives that extra manifestation of headache is really the hormonal and draining that is going on. So there’s an excellent study that has come out here from the east where they’ve actually looked at SIH presentations in men versus women. And what comes through is regardless of which type of a leak you have, men have fewer symptoms, present later, and because they present later very often have poorer outcomes because they’re just not being picked up early enough. And they’re much more likely to, for example, present with a subdural hematoma. That’s because I think the smoother, the easier ones are actually being missed, and those authors have made the point that you need to be a bit more vigilant in men than women. But again, I think it speaks to that it is probably the physiology, and the physiology that’s entraining the pain system is being by the hormonal side of things. Those hormonal side of things are making women more likely to present with the headache side of things than men do. I think that’s what’s underlying the difference between men and women. It’s the underlying pain physiology that is actually driving that difference in presentation. No, no, I agree. I agree. I agree. It doesn’t explain everything. It doesn’t explain everything, but I think it explains part of the things.