Slides
Transcript
Thank you for that kind introduction. I’m really honored to speak on a topic that we’ve done a lot of work on – and that’s fibrin glue patching. Over the next 15 minutes, we’re going to have two main learning objectives: we’re going to look at fibrin glue patching from the patient experience, and then, secondly, we’re going to move on to operator tips and tricks that we’ve learned along the way. But first, let me just share an index case to illustrate how much we, as a society and as a field, have come across in spinal CSF leak care.
This was a recent patient – presented to the ER, a 54-year-old with severe headaches, and then was sent to the ICU. You can see the bad brain sag and the subdural collections with blood. So, this was day one of the admission. On day two of the admission, we did a decubitus myelogram. We found the fistula. In the same session, we did a fibrin glue patch, and the next day, he was discharged. That just illustrates the power of how much we’ve come across the field. Patients, as you all know, have been waiting decades for something like this, and this was just a 3-day hospital course. So that really, I think, illustrates how much we’ve come as a specialty.
From a patient experience, my physician assistant, Adriana Gutierrez, and I – this is how we speak to our patients. We relay to them that fibrin glue patching is minimally invasive. We use conscious sedation for all our patients. As demonstrated in that previous case, we perform it in the same session as the myelogram. We find this convenient to patients. The fibrin glue patching takes around 15 to 20 minutes, but with the myelogram, we do it combined, it’s around 60 to 75 minutes – a one-stop shop. We relay to patients that one treatment is often needed, but more may be required, and rebound hypertension is possible. Although we don’t routinely administer Diamox, we did that initially for all patients. Now it is only on an as-needed basis. Bed rest that day. We do a one-week follow-up, and we do a brain MRI follow-up in one month.
In our discussion of the pros and cons of fibrin glue and the other modalities, we relay to patients that fibrin glue has a high success rate, is minimally invasive, we can repeat it, and there’s no permanent foreign body. Now, the cons. It may require more than one treatment. With embolization also has a high success rate, also is minimally invasive, may require more than one treatment. Additional treatments can be tough if the pathway is obstructed, and the foreign body may limit your CTM evaluation in the future.
Surgery has a very high success rate – invasive and maybe needs to be repeated if de novo fistulas develop.
So that’s the patient experience. Now, let’s move on to the operator side. Fibrin glue patching has five steps. We position the patient at prone, and we scan at the fistula level under CT. We place one to two 20-gauge needles along the course of the fistula – and I’ll talk about this more in a bit. We do a generous test dose with air. With our fibrin glue, we pull the syringe back and inject a little bit of contrast into both hubs of the glue. And then we inject two CCs right off the bat. We do a quick scan check, and then we inject one to two more CCs.
I’d like to mention that we only use the fibrin glue with the high thrombin content. I don’t think this treatment is going to be effective if you use the other type of fibrin glue with the low thrombin concept. Now, exactly where do you inject if you’re performing fibrin glue patching? Here is a CSF-venous fistula. And there are two main targets we like to patch – to target our glue. One is at the cyst-vein junction, and the second area is in this paravertebral wall. I’d say we probably do target these sites 99% of the time. So, if I were targeting this, I would maybe have one needle like in this fashion and try to get the cyst vein junction and this paravertebral component – or maybe just two needles like this, and another needle like that.
For unusual cases, such as intra-osseous CSF-venous fistulas, a direct cyst puncture may be performed. Although that’s not our first go-to for routine CSF-venous fistulas.
Now, for the operators in the room who are used to performing steroid injections for pain, they’ll know that the epidural space is a great place to inject the steroid because it provides extra benefit for the patient with pain. However, in a fistula like this, the epidural space is wasted glue. Now, not that it’s dangerous to have the glue deposited there, but you want to have the glue where the CVF is, so in this case, it would be wasted glue.
Also, just a note – if you have a routine blood patch or fibrin glue patch, the typical area we inject is in the epidural space like in this fashion. So, you can see if there was blood or glue deposited here, it would not touch the CSF-venous fistula. So, where you inject really, really makes a difference. Here’s just another illustration of this CSF-venous fistula. Target the cyst vein junction, the paravertebral wall, and/or a direct cyst puncture.
So, let’s look at a few examples of each of these different locations.
Here is a patient with a normal brain MRI. CT myelogram shows a complex fistula. I’m going to show you a little cine clip here so you can see that a little bit better. See the multiple different veins right there. We targeted the cyst vein junction. See, all this is the glue. Here’s the one-month follow-up myelogram showing resolution of the fistula, and here is how it looks on the video. You can see resolution of all those veins.
Here’s an example of the wall technique, and this is actually one of my favorite techniques to do. It was in a 50-year-old, and she had actually two definite CVFs previously to this that were diagnosed, and treated, and resolved, and now she presented with her third. The abnormal brain MRI. We performed a CT myelogram. Interestingly, we didn’t find the fistula. Two weeks later, we brought her back, and this time I did saline infusion. I don’t typically do that, but it really helped in this case. And now we see this abnormal fistula. In this case, we had two needles, and we built this paravertebral wall. And we used the wall. I think wall is effective by using the vertebral body as an anchor. The glue can smash the vein against the wall and occlude it that way. Here was the one-month follow-up MRI showing complete normalization. The patient was cured.
Lastly, the direct cyst puncture. This is in a patient also with a history of spinal CSF leaks and had a normal brain MRI and had this very bizarre CSF-venous fistula. Here is the cyst within the pedicle and these adjacent intraosseous veins. And you’ll see this basivertebral vein right here as well.
So, in this case, we directly punctured into the cyst in this fashion, and we even got some of the intraosseous veins opacified. And here is the follow-up myelogram a month later, showing resolution and resolution of all those different veins. So, a direct cyst puncture can be performed in these last resort cases.
Of course, fibrin glue patching may not work in every single patient. It didn’t in this case. Here was a patient with a fistula, had an abnormal brain MRI. We performed a fibrin glue patch. Actually, we did three of them in this case, and I was actually really happy with the coating here. I thought that was really a good technique. Though there was, on the follow-up myelogram, a faint residual. And while the brain MRI improved, there was still this dural enhancement, and the patient still had some symptoms. So, it just begs the question – what are the data on fibrin glue patching? There’s a lot of anecdotal evidence, but what are the data?
And I think this multi-institutional study from the US and the UK showed that 60% of patients had a complete clinical improvement, 35% partial, and 6% none. And I think one of the most significant variables that we found was, as mentioned, matching the glue spread to where the CVF is was one of the most specifically significant variables. And I can’t harp on that enough. You need to inject where the glue is – where the fistula is.
Now, many of you that perform fibrin glue patching may recognize that there is this “fibrin black box.” So, if you read the insert of the label, it will say, “Do not inject intravascular.” So, what about – we tried to look at our experience. Well, we’ve injected fibrin glue patching specifically for fistulas in hundreds of different attempts, and we’ve had no fibrin glue allergic reactions. That’s at least our experience. I think personally, injecting intravascular glue when the vein is small is perfectly fine and may help occlude the vein. If the vein is large, though, extrinsic compression is probably better because if the vein is large, the glue will just shoot right through the vein and won’t occlude it. So extrinsic compression may help to occlude it and also provide a better safety profile.
Here’s an example of going to be mindful of big veins. Here was a patient with SIH, and I’m going to show you the vein here – the clip. This vein is actually this big. Only a part of it is opacifying, so it was actually a pretty juicy vein. Let me show you the still images. This whole thing is the vein, even though just this part is opacified.
So, when performing the fibrin glue, I got venous return when I was in the perivertebral space, and I thought I was in the vein. So, I said, let me just inject to see what will happen, because intravascular glue was – hopefully – to include it. But when I did that, there was opacification of the pulmonary arteries, so it was a non-targeted embolization. We kept the patient overnight. She was fine. Brought her back a month later, and this time we did extrinsic compression of the vein. Got a little bit into the vein, but mostly around the cyst in the vein, and there was no PE in this case. The patient’s symptoms improved and resolved. So, be mindful of those big veins.
I’ll just share kind of one final illustration of how fibrin glue patching may be used beyond traditional CSF-venous fistulas. This is a recent case I encountered when I was volunteer faculty at UCSF, where I teach trainees, and I happened to just randomly come across this case on the list. It was a 23-year-old woman – headache, Chiari evaluation. I saw the brain MRI, saw the cerebellar tonsillar descent, and I saw this syrinx. So, I said, yeah, that could be a Chiari, but then you can see that there is pretty bad narrowing of the mamillopontine distance. So, I wondered if there was SIH present. But as many of you know, 23 years old is actually very young for SIH. We don’t see patients with SIH this young. So, I looked a little bit more into the patient’s chart and found this recent outside spine MRI, and there was this very bizarre dural ectasia. You can see we don’t – it should be like this – but you can see how outpouching it is. And on the sagittal imaging, there was this paraspinal lesion that actually I don’t think was commented on in the outside report. So, I said, “Wow, that’s really interesting. I really wonder if this has something to do with the patient’s brain MRI.” But it got more interesting.
Here was an MRI of the patient 22 years prior, when she was 1 year old, and she had this epidural enhancing lesion. At the time, this was read out as an epidural hematoma or hemangioma, and basically, the patient lived her whole life just fine until the past few months. So, I wondered if we were dealing with a vascular malformation that was fistulizing with the CSF and suggested we do a decubitus myelogram, specifically on the right side.
So, I came back to UCSF. I worked with one of my colleagues, Bill Dylan, and we did this myelogram together. You can see, again, the very bizarre dural ectasia, but you don’t see any fistulization with the malformation. We do see a phlebolith, so we know we were dealing with a venous malformation. So, what we did in this case is we just simply did a delayed scan, and now you see abnormal communication with the subarachnoid space and the venous malformation, consistent with a CSF-to-venous malformation fistula.
At this point, I was actually really happy. You know, we confirmed the diagnosis, and I was ready to call it a day, but my colleague Bill was like, “Well, do you want to treat it?” And I told him, “Well, frankly, I’ve never treated anything like this before,” but sure enough, we did. Here was fibrin glue deposited exactly at the fistula’s location, and the patient had a considerable amount of relief, for which we subsequently published. So, this was just kind of a unique example of fibrin glue patching.
So, in closing, hopefully I’ve shared with you fibrin glue patching from the patient experience, on the operator’s end, tips and tricks that we’ve learned, and some unique examples of fibrin glue patching. And with that, I thank you for your attention.