Dr. Michael Malinzak at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Mike Malinzak, assistant professor of radiology at Duke University Medical Center, presented a talk on percutaneous approaches to spinal CSF leaks 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] I’m so excited to be part of the Cedars meeting this year and really grateful to be speaking this morning on percutaneous approaches to spinal CSF leaks. I try to think about places in the context of CSF leaks, wherever I go, right? And North Carolina is very wet and flat and it kind of seems like one of those torn up dural things that just weeps all the time. I think the Hawaiian equivalent is, has gotta be the lava flows, right?
[00:00:35] I think that’s mother nature’s attempt to take what’s on the inside and patch the like, nasty stuff humans have done on the outside . All right, so we’ll go through a few topics today. We’ll touch on general considerations with blood patching. In some ways, this is kind of how to blood patch. If you’re a person that just dabbles and wants to do more, there’s some how-to components there. We’ll move on to different types of patch by various indications and leak type. We’ll touch on the controversies in patching right now. Not a ton of literature, it’s growing all the time. But in terms of really good outcomes, we’re still putting that together as a community, as we all know. Some of the factors that go into patient selection, at least sort of at our practice at Duke.
[00:01:12] And then lastly, we’ll go through a handful of the more challenging cases that, that we’ve managed primarily with patching as opposed to other interventions. So those general considerations: why do we do blood patching? Well, it’s because when we, we’re looking for a CSF leak, which is a topic near and dear, so post-dural puncture headache and spontaneous leaks.
[00:01:30] The main contraindication to blood patching is any hint of bacteremia. So any patient with a fever in the past 48 hours or pending blood cultures or a suggestion of they may be infected. It’s really not something you want to do. The concern is that the bacteria can seed this now devitalized blood clot, which is essentially a Petri dish. And you’ve taken a person that probably had a CSF leak and you’ve given them something that needs to be managed operatively, that is to say an epidural abscess. One of the last things I did before getting on the flight here was patch an interesting patient with JIA [juvenile idiopathic arthritis], Still’s disease, who developed headaches and fevers about three weeks ago, and we’ve been going in circles as to whether or not this is infection or inflammation.
[00:02:10] And when I actually met the patient, this is an inpatient, I found out that the headache phenotype changed drastically when this individual underwent about. a dozen lumbar puncture attempts in outside hospitals. So, in the end ended up treating for post-dural puncture headache, but I don’t think it was infected, right? But the fever made me take pause.
[00:02:26] There’s a question about circulating tumor cells, and do we need to worry about seeding those into the epidural space and causing epidural tumors? Reasonable data from this one paper in about 80 cancer patients specifically with hematologic malignancies, that none of those patients were found to have epidural tumor seeding.
[00:02:43] That is our most common population, right? A lot of people with leukemia, lymphoma are getting intrathecal methotrexate and other treatments. So it happens frequently. There’s less good evidence for the solid tumor population. And in that population are people who are, you know, have no effective normal circulating blood because of blast crisis. I think that’s a good time to think about just using fibrin glue alone. But that’s just more of my practice. There are some poor response indicators. We know that large neural defects tend not to hold well. That’s fairly intuitive. You can see there’s a pretty big neck here.
[00:03:15] Also chronic leaks, ones that are kind of walled off like this. I think it’s, I always tell patients it’s like putting a band aid over an ear piercing that’s healed open, right? That’s not usually going to get the job done. Scar beds are particularly difficult to patch in once the scar is established because there’s no fat plane to get your blood to flow.
[00:03:31] All these things make it hard to get a durable result. So, pre procedural considerations. This is in some ways kind of a checklist. If you’re going in to do a blood patch, you want to make sure you have a big IV that works and you want it to be new. And again, this is to mitigate that infection risk, just obsessed with infection, and you’re probably going to be okay.
[00:03:51] I prefer the prone position, but decubitus works fine depending on patient comfort. And then you want to mark your back before you prep. Once you’re prepped, we tend to draw up the blood as one of the last things we do, so it doesn’t flatten our tubes. And this arrangement here on the right side of the screen is sort of my go-to.
[00:04:06] So we have a three way stopcock hooked up to our patient’s brand new IV. We waste a bunch of blood first, we push saline in, then we suck out about 10 mLs of blood and that goes in the trash. And then we quarter turn this thing back and forth. So as not to, to contact, this tube of blood, which has a sterile exterior and can be placed on your interventional tray.
[00:04:26] A couple of options for fibrin glue and this varies by country and we’re a good international group here, but in the U. S., ARTISS and Tisseel are both Baxter fibrin glue products. I will say it’s off FDA label use to be pushing through a catheter into the epidural space. But both are used very frequently at many institutions.
[00:04:45] For a glue patch, I think it’s important to think about sedation. For a blood patch, maybe not so much. Something about glue tends to ball up and press on nerves and be uncomfortable. So we tend to do moderate sedation. Usually this is midazolam, fentanyl your sort of local house cocktail is fine.
[00:05:01] Always think about preoperatively if you’re going to have somebody that gets nausea, vomiting with sedation, because that really throws off your blood patch if they have to get sick on you. And then the thing to keep in mind is these currently available fibrin glue products have aprotinin in the mix, and repeat exposures lead to anaphylaxis. So they do have an FDA black box warning and we tend to premedicate all of our repeat fibrin glue patches with IV diphenhydramine before the procedure to again, reduce that risk.
[00:05:29] All right, now we’re on the table. What do we do? Numb up the skin, which is nice and sterile, and then pass a 22 gauge cutting tip needle. This is a good place for a cutting tip needle. Cause you’re not going through the Dura this time, right? And you place that into the epidural space, pushing a little bit of contrast. And make sure you’re in the right spot. And then proceed with the patch itself. There’s two main approaches in the spine because there’s not many holes.
[00:05:51] So you can either go from behind and the interlaminar approach, or you can go transformainal, from the side. And then depending on the leak type, you just kind of want to get as close to the leak as you can. And we’ll show some examples of that moving forward. Once that needle’s in place, you push that contrast and you do the double tap technique.
[00:06:07] So a fluoro image, as you’re injecting contrast right there, right? So this is before you’re injecting contrast. And then you know, I’m looking at this thinking, well, that’s epidural. So it gets the green light and I’m not quite sure where that is. Is that a vein? So that gets a yellow light.
[00:06:23] These guys are definitely in a vein. So that gets a red light, and that double tap, that second fluoro image will show clearance of anything that’s in the, you know, vascular compartment. So now this gets a green light, cause this is not going anywhere. So this is just some contrast that leaked out of the frame and into the fat.
[00:06:39] But now would be a good time to reposition the needle and do that again until you don’t see vascular runoff before pushing in your patch material, which is what you see on this final image. Another thing you don’t want to do with your needle is get intrathecal because for one thing, cutting tip needles through the dura I think are a bad idea based on lots of conversations from yesterday and large data but it happens, right?
[00:06:59] So here I’m going through the epidural space. And inject my first little bit of contrast and a little bit, there’s a little whiff going intrathecal there, not what I wanted. You can withdraw the needle, go in just a millimeter or two to the side, and usually things stop flowing intrathecal at that point, and you can place your patch material at that point.
[00:07:17] When we’re actually injecting patch material, this is my little sort of, you know, Johnny Cochran routine: “blood before glued, never sued.” Uh, the idea here is that intravascular spread of these glue products and aprotinin in particular leads to these severe allergic reactions. It’s speculative, but I think it stands to reason that if you put the blood in first, you compress the capillary network and the epidural venous plexus in general, and have less of a risk of intravisation of that glue. In any case, you don’t want to push the glue or blood so hard that you’re getting intravisation in general. So push it nice and slow. And if you do the blood first, at least anecdotally I get a lot less anaphylaxis that way. And there’s a little bit of research coming down the pipeline that, that sort of helps bear that out.
[00:08:02] You stop pushing material based on a couple of feedback items that you have to work with. One is the picture, if you’re doing this under CT. So this is about as comfortable as I am. So, so here we are kind of mid lower thoracic. So that’s about the right thickness for the spinal cord. You can even see a little bit of CSF space back here, but we’re getting to the point where the cord is getting a little compressed.
[00:08:23] And right about now, I’m usually asking the patient, “Hey, pressure? Pain?” You know, and the feedback coming from the patient is valuable too, and that’s why it’s nice under moderate sedation because they can tell you, “no, that hurts” or “really not bothering me at all.”
[00:08:36] The other thing is how hard it is to push. You will feel increasing sort of resistance to injection as you proceed. The discomfort usually starts kind of in the midline of the back and then proceeds distally in sort of a radicular fashion. And the more it goes down, sort of into the buttocks and the legs, the less comfortable I am with continuing to push material.
[00:08:55] Okay. Post procedure, we give all of our patients instructions. We have some talks coming up on rebound high pressure. Yeah. Feel free to grab screen captures. In general, we warn people that they’re probably going to have a backache. Most of the time, this is over the counter meds, but sometimes it requires prescription strength. Headache, so a new and different headache, particularly overnight, prop up on a few pillows, because it’s probably RIH. Usually if there’s no contraindications to acetazolamide, I will give them some acetazolamide and say, if it gets to a five out of 10, take 500 milligrams and we’ll talk in the morning.
[00:09:24] And so we have some little ticky boxes that we can check off at Duke. And then we ask everyone to do bedrest with bathroom privileges for about 24 hours. And then to limit the bending, lifting, twisting for three months. And you know, I emphasize, this is not gospel. I was in the room when we made this up. We had coffee and chicken biscuits, it, you know, didn’t descend from a mountain. So if you happen to reach for, you know, a teapot on that top shelf, don’t beat yourself up, right? Like just do your best.
[00:09:51] Okay. So what do we really worry about in terms of complications? There’s the rare stuff, the scary stuff, and this is kind of the playing with fire stuff.
[00:09:59] This is a picture from my son’s birthday party, which I realized that my child is the product of, like, my current probably overprotective parenting culture, because these boys went like Lord of the Flies, and they played with fire, and it was awesome, and you know what? The woods didn’t burn down, so, most of the time you’re going to get away with it, but every now and then you’re not. So it’s good to have your internal parents there. That’s really what I’m saying. It was my internal parents saying, don’t cause spinal cord injury. You know, you need that thing. And the ways to mess it up are intravascular injections that embolize the cord or excessive mass effect.
[00:10:33] This is a patient of mine who had a high flow leak up near the cervicothoracic junction. It wasn’t localized. And at another center you see these patches done sometime where you can—I’ve seen it done, I haven’t done it—where a catheter is placed, you know, in the lumbar spine over a wire and pushed up the epidural space, I think very much akin to the way pain pumps or or spinal cord stimulators are placed, and then blood is pushed as the catheter is withdrawn.
[00:10:59] And the patient was awake for it and says that it was the most excruciating pain he’s ever been in. And ever since the procedure he’s had paresthesias in both feet. And MRI of the spine acquired maybe 3 months later, and I don’t have a pre-procedural one, shows an area of myelomalacia here at the T6 level, and I just kind of always wonder. I mean, I sort of think those things are related, and I’m not sure if it was the vascular injury or just too much compression at that level.
[00:11:24] But I do take comfort in doing this under CT and seeing how much mass effect I’m causing along the way because that’s the sort of thing we don’t want to do, right? Of course infections, we’ve talked about that a lot, so just be sterile and don’t do this in people who probably are bacteremic.
[00:11:38] Meningitis, both infectious and just aseptic, occur sometimes when instrumenting in the system. So again just being careful. We’ve already touched on glue anaphylaxis and repeat exposures, and do your best and you can probably play with fire most of the time.
[00:11:53] The things that happen a little bit more regularly are nerve irritation. So a few times a year I’ll place a patient on gabapentin or a related drug for a little bit. If that’s not clearing it up in a week or so I’ll go on to, to a steroid dose pack and that almost always works. And then maybe like two or three times now, I’ve done a steroid injection right at the site, if none of those other things work. And that, those, that three steps has always gotten me out of trouble.
[00:12:18] This is an individual, he was about 30, had this big ventral leak. We actually temporized him for a few years with patches before, you know, he would kind of seal and re-leak before he got surgery. But this injection really irritated the T1 nerve root. He had a really good sort of radicular pattern in the forearm. I had him on gabapentin for three months and that got the job done great. This is an individual, never saw her leak. She was one of these people you’d patch and she’d be better for six to 12 months and then her symptoms would come back and we patch again. And I think those repeated patches probably set you up for this kind of risk. She got a lot of mid back pain after about the third patch. And I ended up doing a steroid injection, which resolved that problem.
[00:12:58] The thing that happens a lot is rebound intracranial hypertension. And we’ll talk about this more today in other sessions. So I won’t dwell on it. But really the treatment is very much akin to IIH. And acetazolamide is the first thing I reach for on the shelf. Probably about half of our patients get some of this, and there’s some published studies out there from people in this room, as we all know, and about one in five of our patients at Duke actually requires medication.
[00:13:21] There’s also the small minority of people that call you at 10 p.m. with a 12 out of 10 headache and they’re throwing up. Or the next day they’re starting to get dehydrated. Those are the people where it’s time to do the nuclear option and do a diagnostic and therapeutic lumbar puncture. I guess really a couple things of note: so here I’m doing a lumbar puncture, and you see the opening pressure is 24. And we all agree that’s too high. It’s not uncommon to meet a patient with SIH who has an opening pressure of 10. You patch them. And then they’re throwing up in recovery and they’re, you know, 12 out of 10 headache. And you go back and you stick a needle in and now their opening pressure’s 18, right? Which is normal, but it’s not normal for that patient. And something about that abrupt change really makes them uncomfortable. I feel again, like the analogy here is the ulnar nerve, we can all lean on it slowly. So many nerves do this, but you change something rapidly, you hit them hard, nerves hate that.
[00:14:13] So, so abrupt changes, even within the normal range, can be really symptomatic. This freaked me out the first time I saw it. When you withdraw CSF on a person that had a blood patch one, two, three days ago, it’s going to be watermelon juice color. You know, that’s normal. So there’s a little bit of heme that is getting across all of our protective membranes in these patches. And that doesn’t mean you’ve caused massive subarachnoid. That’s okay. Don’t be freaked out by that.
[00:14:39] All right. So moving into specific patches by different indications. This is probably the like generic patch. So, and I do this most commonly for people that have a positional headache, I can’t get them diagnosed with something else to explain it like POTS, and they tried everything else under the sun and they haven’t had a blood patch or they’ve only had a lumbar blood patch. At that time, we do this, and this is the, in my words, empiric—I’ve always put that out there, big and bold letters—the empiric, non-targeted thoracic epidural blood patch. And really just trying to get diffuse coating of blood all over that thoracic epidural space. Most often I achieve this with two needles, usually one around T4 ish, this one’s at T3, and one down around T11, T12, and, you know, 10 to, well, 5 to 10 milliliters of blood at both location.
[00:15:31] And we mix a little bit of ISOVUE-M 300 myelographic iodinated contrast with our blood, usually it’s about 10 milliliters of blood and one milliliter of iodinated contrast. And then you can see where it spreads, right, and I know here that I’m kind of squishing the cord a little bit at T7 and it’s getting a little tight up at T3, but nothing looks catastrophic where the spinal cord is becoming truly collapsed like a pancake. And I’m talking to the patient throughout. And if I really get nervous, I say, “Hey, move your legs.” And usually they do. I guess the other time that I’ll do this are those patients that definitely have SIH by the ICDH-3 criteria, but we know, we don’t see a leak on myelography.
[00:16:10] Again, something I did last week is, a patient with the frontotemporal dementia type features and I cannot find this individual’s leak. So I kind of just gave up and finally did a non-targeted thoracic just for now until we sort this problem out. And at the very least it can temporize in people but it, you sometimes get lucky and they just get better and presumably we’re treating a leak we haven’t seen. I’m not really sure.
[00:16:32] All right. I’d say the average volume is probably 20 milliliters, but sometimes you get up to 30 depending on how capacious the canal is and how well it spreads. So post-dural puncture headache, and there were talks on this yesterday that were quite good. So myelography is often unrevealing, but not always. And for the acute uncomplicated postdural puncture headache, you know, it’s pretty easy. You just go to the level of the bandaid, right? Somebody has made a hole there. Or in the case of my most recent patient, you know, a dozen holes and you’ve got the scabs, you’ve got the bandaids. You kind of know they were working like L2 to L4, and you get dorsal coverage at those locations.
[00:17:09] And that’s the acute, uncomplicated kind. We spoke some yesterday as a group about the complicated or the chronic postal penetrate headaches. And these individuals, by the time they get to me, have usually had blood patches at one or two other locations geographically. And so I want to offer them something new.
[00:17:25] And and in this case, we go to these circumferential epidural fibrin glue and blood patches. And this is double kind of hocus pocus, right? So, so the first like imaginary thing is that getting circumferential blood actually helps. We have no data for that, but you know, by inference it, it stands to reason that you can have a ventral dural puncture. We’ve seen them. We’ve seen them myelographically leaking. This is not my image, but you see these done where you’re using fluoroscopic guidance, you know, you’re not getting CSF return. Everything looks in line on an AP, you swing your fluoroscope laterally, and you realize the needle’s through and through the front of the sac, in this case, all the way through the disc. So by getting that ventral coverage, we’re hoping to patch any leaks that might be there.
[00:18:07] The second bit of of sort of, you know, wishful thinking is that glue works better than blood alone. We don’t have proof for that, but you know, our surgical colleagues, that’s what they put on in the OR. So it makes sense. And the way to achieve this is to try to get ventral transforaminal needle placements. And that tends to work kind of well in the mid levels. So I often get ventral transforaminal on the right, sort of, you know, between L2 and L4, and then I go down a couple levels and get ventral transforaminal on the other side and get spread in that, you know, space in front of the thecal sac.
[00:18:40] And then for the dorsal coverage, I often go to sort of the opposite ends of the lumbar spine and let it flood up and down. And you can take procedural images and the goal in the end is to achieve this bagel. And that white contrast is the blood patch and the hole in the bagel is the thecal sac.
[00:18:57] And now you know you covered everything. So, If there is a leak there that’s going to respond to patching, you know, you feel like you’ve given it your all, and obviously you follow with the patient, and sometimes this gets the job done and sometimes you have to keep thinking.
[00:19:10] Indwelling devices. These are tricky, right? So, first of all, they tend to happen on Friday at 5pm when your friend stalls and says, I just placed a catheter and the patient has a headache. Transdural catheters for baclofen pumps or pain pumps sometimes have leaks associated either with an attempted site placement or at the actual hole.
[00:19:30] Here’s one where I can see leakage at the hole. See it kind of going through? This is just like the picture the doctor showed us yesterday where I get the needle epidural next to the catheter. And when we inject contrast, we see a little bit spreading in around the catheter. That’s transdural. In this case just injected some blood. This was a child with a baclofen pump and he got better very quickly.
[00:19:52] Spinal cord stimulators. These are pushed up the dorsal epidural space, and I think you can kind of just tear up the dura on the way. So when a patient has a new, you know, positional headache or just headache after a placement of one of these, I try to get spread all along the dorsal sac wherever the operator could have possibly sort of roughed up the dura.
[00:20:13] Down here, that’s pretty easy because there’s not a lot of stuff other than the leads in the way. Up high, you get a lot of streak artifact off the actual electrodes, and and that’s a challenge, and you do your best. I always try to image search the device before doing this just to make sure there isn’t, you know, a plastic membrane between these two leads. They come in a lot of different shapes and sizes and it’s nice to know what they look like in real life as opposed to just on the CT.
[00:20:38] Surgical durotomies. You know, this is one place where we have some pretty good data from a paper at a Duke before my time that Dr. Kranz took the lead on, but if the neck of the leak is greater than five millimeters, patching tends not to work.
[00:20:50] And you know, you, you kind of want to save your patient and your surgeon and everybody else time. And and also risk. So oftentimes I’m counseling people in this case, you know, there’s a nine millimeter neck, and the way this was measured was fairly inclusively. So, you know, if it looks like it could be part of the neck, you include it in the measurement. And this is almost certainly not going to respond to a patch. You know, there’s kind of the phrase about like spitting into a hurricane. I think that’s the right idea when you’re putting a little bit of glue or blood inside this large collection: it’s just going to wash away. On the plus side, I can often guide the surgeon to the leak site because it’s often not obvious to them, particularly if you’re dealing with individuals with probably abnormal dura that are prone to leaks. You know, sometimes the leak is like on the left at L4 and the surgeon did most of their work on the right at L3. And being able to tell them that before they go into this healing sort of melange of scar and blood can really help in surgery.
[00:21:44] Okay, moving on to our classic spontaneous leak types. There are ways to treat ventral leaks using blood and fibrin glue. And really what we tend to do is try to get bilateral ventral transforaminal needle placements and spread in the epidural space, glue and blood as best we can at the leak site. Just like surgery, knowing the leak site is really key. You know, here we’ve localized to T5-6, here we’re up at T2-3 and there’s kind of a spiculated protrusion in the way. I don’t know if that works as well but we try it, particularly if it keeps the patient out of the OR. For leaks that are ruptured nerve root sleeves, as in this case, or leaks that are CSF venous fistulas, in both cases, you just want to place the needle in the foramen and try to pack pretty much as much blood and glue as you can get in there.
[00:22:28] You know, this is another example. This is one with a CVF, very obvious, you know, paraspinal venous sign, and and got complete recovery. So this does work sometimes, which launches us directly into the controversies. When do you treat percutaneously—blood patching, glue patching—versus the other options that are out there—surgery, transvenous embolization? Well, for the dural defects, you know, I think that sleeve ruptures, for reasons that aren’t entirely clear, tend to respond better to epidural patching than do the other leaks. The ventral tears respond with a reasonable enough frequency that I do perform the patching on them if they don’t appear to be chronic. And then the CSF-venous fistula, it’s just kind of all over the map. So if we just think about those alone, there’s a bunch of sort of smallish case series and then our own experience at the centers that do a lot of this. And the reported success rates, as best I can tell, because there’s variable metrics of success, range from three to 100%. And so just, you know, immediately we have no idea when you see a number like that. So I tend to quote my patients, “I think there’s about a one in three chance that you’ll get a permanent result with this.” And then the question is really kind of one of philosophy, right?
[00:23:39] Are you a single mom that needs to be better by next month or you’re going to lose your job and, you know, it’s a whole big mess and you’re healthy? Well, then let’s, you know, let’s go to surgery. Let’s do an onyx embolization. Both those are kind of in, in the, you know, 90 percent range in terms of adequate clinical improvement. [If] you’re an older person with a lot of comorbidities and you got all the time in the world and you want to stay out of the OR, well, that’s a great time for me to patch.
[00:24:02] So this is sort of my little attempt at what I advise and when, and so people that don’t definitely have SIH you know, I just start with that nontargeted thoracic patch. If they have SIH and a negative myelogram, we do a nontargeted thoracic patch. [If] they’ve got some maybe things on DSM or CTM, you know, I’ll target those and then we see how they do. Postdural puncture headache. Again, first treatment: interlaminar blood patch. When that fails, we move on to the circumferential blood and or fibrin glue. The durotomy, first we patch with blood or fibrin glue. And when that fails, surgery. And very much the same in these spontaneous leaks. In all these cases, I think the patch can be a temporizing measure. If your patient’s getting on a flight to go back to, you know, someplace where you are not, that will help tide them over before a definitive treatment.
[00:24:48] And then also, you know, things that suggest a patch might help in a ventral leak, a thin collection as opposed to the rounded edges of chronic collection. I think that tends to work better. And then also, you know, as I mentioned, patient philosophy and comorbidities really factor in. So with that in mind, let’s look at some challenging cases that we’ve seen that we managed primarily with epidural patching.
[00:25:10] So this 44 year old had very rapid cognitive decline, was admitted to our neurology service, and they were suspicious of things like CJD. I’m not a psychiatrist, but I think this was catatonia. She could only say, “where are the kids?” And her brain MRI was positive for SIH and she had this big sacral Tarlov cyst leak.
[00:25:30] You know, initially I kind of thought it was me messing up my myelogram and injecting contrast epidural, but I rolled her around enough and could see it was leaking out of the Tarlov cyst. And this was one of the more striking responses ever. And I love it. Cause a fellow walked into the room while I was doing it, and on the table, I was like, “Hey fellow, check this out. You know, you know, do you know what year it is?” And she said something silly like 1986. And as we’re patching, she turns around and goes, “You know, a moment ago, you asked the year, and I said something that was wrong, it’s 2018.” And like 45 minutes later, she was ravenously eating, didn’t remember the past four weeks of her life, and her family was all crying, and this is a time that the blood patch worked and she’s still doing great.
[00:26:14] I love this case, right? And this is one where a little bit of glue and blood fix this human’s life and she’s like a productive person and it’s wonderful. Tarlov cysts leak sometimes, but they are uncommonly the source of these giant, you know, profuse high flow leaks.
[00:26:29] Slow leaks: again, this is a nerve root sleeve leak that I think responds well to patching. This is a retired woman. She had SIH symptoms and brain MRI findings. I could never really see her leak on CTM. I believe these slow leaks sometimes we see best on MR myelography. So intrathecal gadolinium with a T1 weighted fat saturated image. There was all this staining in the epidural fat at t1. And this responded very well to a patch, and she’s still doing great. This is an 84 year old. This is one I didn’t fix. She presented with symptoms of cirrhosis, ataxia, hearing loss. We did find her leak, and she has this sort of combined hybrid thing that’s out there in the literature where there’s a little bit of an epidural leak at T2. And then that communicates. into the basivertebral vein. So it’s kind of fistualized into the venous system. There are operative choices for this, but she has severe aortic stenosis and is an anesthesia risk. So I’ve patched her a few times and I don’t think I’ve fixed her, but she’s still ticking along and we still think the risk benefit favors no surgery at present.
[00:27:29] This is a 32 year old with a very large osseus hemangioma. We found a huge CVF here at left T4 and this had been blood patched and glue patched innumerable times. And he would have these exacerbations where, you know, couldn’t stand up. So he’s an inpatient in the hospital. This was shortly before onyx embolization came to the forefront, but I passed the needle interlaminar into the intrathecal space down into the nerve root sleeve and embolized with gel foam. That gave him a really good result for about eight months. The headaches came back. We then onyx embolized him and we’re kind of in this cycle again, the headaches keep coming back. This is a little thinking outside the box, but is there percutaneous intervention, direct sticking and injecting gel foam, my neurosurgeon said wasn’t crazy, and I did it and nothing bad happened. But it was kind of a humanitarian use scenario. He was pretty sick.
[00:28:17] Post op is a great time for a patch too. This is a 41 year old unhealthy UC ulcerative colitis, chronic anticoagulation, was coughing during a PE and developed a leak, right? So he had this huge leak. He’s one of these guys that gets a little obtunded with his leaks. So he’s unwell, he’s in the ICU and he was operatively repaired, but just didn’t get better. So I passed ventral needles down—you can just sort of barely see them on these images—to this T1, T2 leak. There’s the tip of this needle. Here’s the tip of the needle on this side. Pushed about 10 milliliters combined to fibrin glue and blood. And he got better, and his brain MRI got better and he’s still doing well. He works construction, which makes me nervous, but that’s him.
[00:28:56] Lastly, this 38 year old, this individual was in a n MVC years ago, like 20 years ago, and had a cord transection, so there’s really no recognizable anatomy up here. Except there’s a lot of dural ectasia in the mid cervical levels, but he does well, you know, he’s wheelchair bound, but he’s, you know, has a good quality of life. And developed a horrible headache, and we found that he had SIH changes. CT myelogram didn’t show anything and the headache worsened. And then he got bilateral PCA infarcts as an inpatient, which I think was due to brain sagging, compressing the arteries. So now we’re moving into, you know, compassionate use again. My MRI myelogram, you know, I’ve got some question marks on some stuff that might be weeping dura. Our surgeons didn’t want to touch him.
[00:29:37] Dr. Schievink might, but at Duke, we were all pretty nervous. And so this is one where I did some transforaminal patching with blood and glue with the help of Linda Gray. And he also got better and is back to his baseline. So, all these patches can work. I think, as we come up with better other things that aren’t patches that may work better, there’s still a role for them in these patients that are too unwell, or there’s social factors that play into making it useful.
[00:30:02] So thank you for your time.