Transcript
Perfect, thanks a lot for the introduction, and thanks a lot for pronouncing my name correctly – it’s a tough job, I know. So, congratulations to the faculty around Andrew for setting up this meeting. It’s fantastic, and this mode is really something we can maybe have more of in the future. So, thanks again for this.
So, I always show this picture of the Swiss Alps to impress the audience, but I don’t think if this really works in Denver, where we have some Rocky Mountains around the corner, but let me say I tried.
So, I have nothing to disclose regarding this talk.
And so, the next step now – we found the fistula as Peter showed us, or the colleagues from yesterday showed as well. So, I want to present you the technique on how we would perform transvenous embolization of the CSF-venous fistula and how we want to get there and what kind of materials you use to really close that leak.
The next step will be – that’s just my opinion a little bit sure – you will hear the bias – what it takes to perform transvenous embolization. And the next step, as my title is called, is why or why not to embolize the CSF-venous fistula embolization. So, when I talk about transvenous embolization, I always refer to this first publication of the first cases from the colleagues from Mayo and from Toronto, and there are a ton of variations for transvenous embolization. I will just show some of them.
First case to present the technique. It’s a classic story: 66-year-old woman with sudden onset of orthostatic headache. They performed the MRI of the brain, and you see classic findings. And this is something we always discuss about findings of SIH and brain imaging. So, it can be hard when the signs are just subtle, but this is something nobody should really overlook, and that’s what we advocate for – really, really look through the scan that you don’t have a negative case or report negatively, but it’s really, really striking. Spinal MRI is just no epidural fluid, so we think about CSF-venous fistula. And then what we perform is lateral decubitus myelography. And then we keep the patient free breathing so they can have all the inspiration cycle or expiration respiration cycle. Then you see this – and you really see this small vein where the contrast media is flushing away. And when you have CT, you see the same as well – that you have these small veins, paraspinal veins, filling with contrast media. So, this is a clear CSF-venous fistula.
What we do now is that you know where you have to go – now you just have to get there. And what we do is a trans-femoral axis through the vein, and then we just advance our catheters and our wires upwards through the right atrium of the heart. Then we go to the azygos vein. The good thing is, we know where the azygos vein is because we always perform myelography in these patients and then can just look at where the origin is. Then, you go in there with your wire. This is pretty straightforward.
Afterward, we mostly do a venogram just to see how the anatomy is. And then you have to look where your nerve root or your paraspinal vein is. You want to get into it, so you can count upwards from 12 to 8 and think, “Okay, somewhere there it must be.” Then, you look for where the vein is, do another run, well there it is. You see then already all the veins inside the neuroforamen, you see the epidural plexus. This is just a lateral view, and this is your target right – that’s where you want to go to.
Afterwards, you advance your microcatheter or any materials you have that you’re familiar with working with in doing the embolization. Another run, so you’re perfectly in the right position to do your embolization because some of those veins are the venous part of the CSF-venous fistula, and this is what you want to occlude. Then afterwards, you take whatever embolic agent you have and are allowed to use in your country – because that’s an issue, especially in the United States, as I’ve heard. In this case, we use this embolic agent, then just inject and try to fill all the parts of the CSF-venous fistula to really occlude the venous draining part of it.
We mostly perform afterwards fat pattern imaging just to see where the embolic agent really is, and then you can really compare it to your myelography. Here, you see, “Okay, this is where all the veins are opacifying with contrast media through the fistula.” When we compare it to where the embolic agent is, you see, “Okay, we had the embolic agent inside the vessels we tried to occlude.” So, this is something we always aim for.
We bring all patients back for clinical visits and for MRI four weeks later. Then you can see, okay, this is prior to embolization. This is afterwards. And you see, okay, the MRI brain normalized, and the symptoms dissolves. So, this is the, let’s say, only imaging-proven sign we have that maybe includes the fistula itself. That’s a problem for CSF-venous fistula, because we cannot just prove that the epidural fluid is gone. We just have to rely on clinical findings and on imaging signs from brain MRI, if there’s some that we can really relate to.
There are some variations. Just one of them is called the “pressure cooker technique.” That could take – you take a catheter with a small balloon on the tip, and then you just have more forward pressure when injecting your embolic agent, just to really be able to penetrate the small vessels in the neuroforamen, in the epidural plexus, where the fistula is. This looks like this – so this is a balloon that’s inflated, and then you see really how the embolic agent is going around the nerve because you really want to penetrate deep in the anterior and posterior part around the nerve root to really get all the vessels responsible for draining the fistula itself.
Another case where we have a 57-year-old male, and we perform lateral decubitus myelography again. And what you can see – this really big nerve root cyst. You really can see nicely how the contrast media is just running out of the cyst. But when you look closely, you see, like, a level below that, there’s no fistula. So, this is a case where we have two CSF-venous fistulas in one patient at the first myelography.
Now, the question is always – like, how to treat those? And it’s pretty easy. You just do the same, the same access, and you just treat both of them. And the combination of how you do it quite depends on what your technique is because due to the big cyst, we see that there’s not a really nice paraspinal vein at the upper level of both CSF-venous fistula. So, what you can do – and this is something you can easily do in most of the cases – is you just enter the lower level, and then you just pass through the epidural plexus at the level above, do your embolization there. And then afterwards, you just pull your catheter back. And then, through the same access, you just inject on the other level where the other fistula is, your embolic agent again. And then you have treated both levels with the same catheter in the same session. And it takes roughly 10 to 15 minutes extra per level you have to treat.
This is how it looks, and this is something we really want to achieve. You have just circular embolic agent around the nerve root, and this is something where we consider it should close the venous part of the fistula.
This is just one example of how to access the fistula. There are tons of variations. It’s a very, very nice picture from the colleagues from Mayo. It’s really the standard way through the azygos vein is mostly considered the standard way you want to go to for, let’s say, 80% of the fistulas, but here, there are tons of variations. If they are upper cervical spine or lower, you can pass through the epidural plexus, so there’s always a way to go to the fistula itself – a very, very nice paper.
So, next point – what does it take to perform a transvenous embolization? This is a little bit – because there’s no data or guideline for this – it’s a little bit my opinion on what I think should be necessary. And what you need, sure, you need a neuro-interventionalist with a proper angio suite that you can perform neurovascular stuff. You need a neuro-interventionalist that is dedicated to the disease because, sure, it’s not an aneurysm, it’s not stroke, and it’s not an AVM. Some neuro-interventionalists I know would consider SIH being boring – because it’s not just this fancy brain intervention stuff.
And this neuro-interventionalist should be part of a team involved in diagnostic workup and discussing cases, not just like mercenary-style interventionalist to perform the procedures itself. What’s really helpful is you need to know how to handle the necessary materials for transvenous embolization. But luckily, they are identical to what we use for aneurysms and AVM. So, this is quite helpful when you do these kinds of treatments, you have the techniques already at your hand from the other ones.
And the next thing you really need to know is you have to have some anatomical understanding of how to get to the epidural fistula and to the epidural plexus, and how you safely can perform venous intervention around the spinal cord. That’s really, really important.
So, next part now. Like I said, the title of my talk – why or why not to embolize. Because there are no comparative studies, it’s a little bit biased from my side, sure, but the other colleagues will have the chance afterwards to talk about percutaneous patching and surgery.
So, when you have a CSF-venous fistula, and this is sometimes made for patients and even referring physicians, it’s like you have a CSF-venous fistula diagnosis made, localization is there, and then you shove everything, all the information you have, in a black box, and suddenly this comes out. It’s always a question – yeah, why should we do embolization, surgery, or patching? And that’s a good question – a lot of patients ask myself, “Why should I do this, and why shouldn’t I do the other stuff?” And I think there are some reasons why we consider advising patients on what would be a good idea.
And I think one point depends on what you have at your hand, right? Do you have, like, a spinal cord surgeon that is capable of treating a CSF-venous fistula or a dural tear? Do you have a neurointerventionalist that knows how to treat this? And another factor – what is coming out of an interdisciplinary board, for example. It’s not my opinion that I say we need to do embolization. We discuss any case or every case with all the colleagues, and then mostly we give one advice to the patient on what we consider would be the best option for you.
Another thing that’s really important is patient preference. So, a lot of patients come to the hospital and see you, and then they have a clear preference for what they prefer, and what not.
But everything we talk about and every therapy we offer the patients should be focused around the available data we have, right? And this is just – at the moment, we have the available data for these three kinds of treatments, and I think we should consider informing the patient around what we know, right? And I will not cover the other ones. I will just cover my part.
Now, straight to the question: why should we embolize? And, like I said, it’s my opinion a little bit – and excuse me for that. I think it’s a very, very fast and quick procedure. This is just one case we had in October. From the groin puncture to the femoral vein until we just injected the embolic agent and finished the intervention, it was roughly a little bit more than 20 minutes. We can do this in conscious sedation, so there’s no need for general anesthesia, but a lot of patients prefer general anesthesia. That’s why we discuss it with the patient themselves, but you can do it in conscious sedation.
Another thing is, we do this quite regularly, it’s an outpatient treatment. It’s just purely transvenous. After the treatment, mostly the patient has some sore area where you inject your embolic agent, but mostly that’s it. And then we send them home in the evening. So this is a really, really big advantage of this procedure, and that’s why I think it’s a good reason to embolize.
Another thing is, it’s safe and efficient. So, we have some literature, especially from the colleagues from Mayo, that it improves imaging findings. So, when you have, like, a positive finding from a brain MRI, afterwards the signs mostly normalize, and the brain – the headache – improves as well. So, we have some data about that. The complication rate is quite low. We see quite regularly patients having, like, back pain – a local back pain – and sometimes, like, some kind of radiculopathy. but mostly it dissolves after four weeks. And then rebound hypertension is nothing exclusive to transvenous embolization. I think we see it in any treatment we do.
And then the Onyx emboli in the lung is not clinically relevant. It doesn’t look as nice when you do, like, a CT from the thorax, but this is something we see and consider a complication, but mostly without any risk. There’s new data available for more patients, but the results are practically the same.
Next reason why to embolize: because when you have a patient with multiple fistulas, it’s just easy. It just adds 10 to 15 minutes of intervention time. You just take another route to another level and then do embolization as well. But, sure, you have to know where the fistula is. You cannot just say, “Okay, we will prevent recurrence,” and just do, like, five, six, seven injections just to prevent recurrence. That doesn’t make any sense. But when you know that there are more fistulas, you can treat them in the same session easily. And when there’s an eloquent nerve root, mostly it’s not a problem treating them as well transvenously. Then you just do the same technique, and it’s an eloquent nerve root, but you don’t really care about that.
Another thing – and this is a little bit counterintuitive as a reason maybe for performing transvenous embolization, because this is a patient where we did transvenous embolization, and we found that there is a recurrence at the same level. The patient had recurrence and symptoms. We saw, okay, there’s another fistula, and there’s a nice paper – one from Dr. Schievink – that proved that when you perform transvenous embolization and you have a recurrence, you still can perform surgery. So, I could argue, okay, then we should always start with transvenous embolization because we still can perform surgery. This is like a little bit provocative, but that could be one argument of why we should perform transvenous embolization first.
So, the other way around – why we should maybe not perform embolization – and like I said, it’s my opinion. So, when you have a recurrent fistula at the same level we treated already, like the case we had before, sometimes it gets a little bit annoying. Because if the patient we did myelography, clear fistula, went in for transvenous embolization, but we had a hard time getting our embolic agent in the posterior part of the neural foramen to really occlude all fistula-bearing veins. And we just injected a lot of embolic agent, went into the epidural plexus to the contralateral side, and then we said we’d stop and see if it works.
The patient came back with a recurrent fistula. You see it in myelography. There is still a vein that’s opacifying with contrast agent, and it’s exactly where we thought – it’s in the posterior part of the neural foramen. So, we could really not penetrate it. We went in again, said, okay, now we go trans through the epidural plexus and try again to embolize it. It still didn’t work. We still saw a fistula afterward, and this is something where we consider – maybe in a recurrent fistula at the same level, we should go straight for surgery.
Another thing – no proven data, but this is something I sometimes see – is when you have an extensive fistula, it’s sometimes hard getting the embolic agent really to the upper level part of the venous plexus, because it’s just really big and you see already like remodeling of the bone. So maybe this is not a good case for transvenous embolization. This is what I say – other centers would say differently – but that could be one reason.
Last couple of slides. So, some say, okay, when you inject like Onyx – like crazy, we will never have a chance afterward to find the fistula anymore because we have a ton of artifacts. But the good thing is there are new agents are arriving where we know we can use different stuff or materials. It’s not allowed at the moment, but it will become available next year in Switzerland as well – this Obtura stuff. And the good thing about this – we know from neurovascular interventions – just a dense part of the embolic agent just dissolves, that will give free vision again afterwards to the area wherever they embolized. This will be perfect for transvenous embolization of CSF-venous fistulas as well.
So, last point is, when you have problems with access, maybe you should consider changing your approach or change it to surgery as well. So, when you have, like, thrombosis of the femoral veins or variants, maybe you could consider going for surgery or finding a different way to your fistula itself.
Another pretty obvious part is, so when all that I told you about what is necessary to perform transvenous embolization – when everything to those – is, like, you have to answer with no, maybe that’s not a good idea of performing transvenous embolization.
So, take-home points. I haven’t talked about it, but diagnostic workup is really essential, right? You just go in when you really, really know where it is. Otherwise, it doesn’t make any sense. There are tons of techniques and routes to go to the femoral veins – you just have to pick one you’re familiar with and the techniques and materials you have in your department itself.
It’s really helpful when you have the alternatives of surgery and percutaneous alternatives in your hospital because then you are flexible in choosing what’s best for your patient. And you need an experienced interventionalist. Otherwise, you have a very, very low cure rate, and your technique is being discredited. That’s a problem in itself.
If it fails and the patient still has symptoms. Don’t stop – look again, do another myelogram, and maybe try to do another therapy again, or send to surgery. And keep recurrence in mind with CSF-venous fistula.
And now, final slides. Thanks to all my colleagues, especially from Bern and Freiburg, like Jurgen and Tomas, and all the friends I’ve met throughout the years working in this field. Thanks for the journey – it is a pleasure still. And if anybody has questions, just send me an email or text. Thank you.