Slides
Transcript
So usually people think that surgery is kind of the last option, and that’s certainly not always true. As Dr. Maya has shown you, we work together a lot, and we’ve been doing it for a long, long time. I like to put neurosurgery on top of the program, unlike Dr. Maya, and in this slide everybody’s name is actually spelled correctly. I’m super grateful to all the neuroradiologists here and in general, because you can’t really do surgery without having a great neuroradiology team. It’s just really not possible.
I’ve been doing surgery for these spontaneous leaks since last century when I was a resident. As you can see here, this was the first, I don’t know, about 10 patients who had surgery when I was a resident. Basically the surgery just consisted of packing of the epidural space. Right? I think 7 out of 10 patients were just treated with a surgical blood patch. In the short term the results were pretty good. It was a really, really good epidural patch, but in the long term the results certainly were not that good.
In our group we like to use this digital subtraction myelography a lot. Dr. Maya will talk about that tomorrow because he thinks that’s vastly superior compared to dynamic CT myelography that a lot of other neuroradiologists use, but I don’t discriminate like he does. I’d be super happy to operate on anybody where we know exactly where the leak is and what type of leak this is, and this happens to be a group of patients who had their imaging done by Dr. Carlton Jones.
Before we can really talk about surgery, we have to talk about the different types of leaks. Dr. Beck already talked about that quite a bit, but I think it’s kind of important to simplify it, right? Because we call it type 1, 2, 3, 4. And then what we called 4, Dr. Farb in Toronto did not call that 4. Then Dr. Lützen and Dr. Beck came up with this CSF lymphatic fistula that we had never seen before. So I think just to have a good concept of what the different types of leaks are, I want to divide it just a little bit different. Right, so you either can have a tear of that sac, that sac that contains the spinal fluid, or at least is supposed to contain the spinal fluid, and that can be in front, it can be lateral, or it can be posterior. One of those three compartments, and they’re kind of easy to visualize, not just as a surgeon, but just in general.
Then we’ve talked about these cysts. Usually these cysts don’t really rupture, but occasionally they can. Or people can have these very complex types of cysts or just enormous outpouching and widening of that dural sac.
And then there are the fistulas. The vast majority of those are CSF-venous, but very, very rarely CSF lymphatic fistulas. And then Dr. Farb in Toronto was really the first one who described what he calls distal nerve root tears, and I think that’s a really good description this stuff is really rare.
For these ventral leaks, as Dr. Beck mentioned, those are almost always caused by some type of calcification. Here you can see that really well on the post-myelogram CT scan. There’s a piece of bone that has punctured the dura. This is what it looks like during surgery. We use intraoperative ultrasonography a lot. You can find the piece of bone. Sometimes you can find that dural hole. And this is what it looks like at surgery. This is the normal dura in front of the spinal cord. This is the spinal cord. This is a sharp little piece of bone that has caused that hole in the dura.
These little holes in the dura on average are about four or five millimeters, but they can be really tiny, and that’s actually more difficult to find. As you can imagine, this one was about a millimeter and a half. This one was much larger. This was about a centimeter and a half. The largest we’ve seen was just under an inch or so. And as I mentioned, it’s caused by some type of calcification, and that can be really minute. It’s just a tiny little speck of calcium on the CT. This is more like a typical type of microspur, like a little scalpel that has cut the dural sac. Sometimes it’s more of a typical calcified disc. That can look kind of scary, right, when it’s protruding really into the spinal cord.
It’s kind of fun to remove those discs at surgery and also I think for the residents, because else, you know, oh yeah, there’s another tear in the dura. We’ve seen that a lot. But then when there’s kind of what we call a juicy type of cyst—here’s the hole in front of the spinal cord, that’s the hole in front of it. Here you can see the disc kind of poking out, and then you can easily remove that. It’s kind of like a calcified disc, a little bit like lobster meat with a little sand in it. And then you can remove that, and that should be a really safe and not a fun procedure to do.
So there are different ways to repair these leaks in front of the spinal cord. When I started doing those in 2009, we started doing digital myelography in March of 2009. So then we finally knew where those leaks were located. I would do kind of a destructive type of procedure. We would remove facets of the spine that would destabilize the spine, and then we would fuse it over many levels. And I would look inside of that dural sac. I would look outside of the dural sac. It’s kind of difficult to find that tear. You kind of knew where it was, but not really. So the results of that were not that good. Early in 2011, I looked at the results over the past two years. Those were kind of pitiful, really bad.
So then I started doing intradural surgery, like what Dr. Beck showed you. That involves a laminectomy. You remove some bone from the back. You open that dural sac. You go around the spinal cord, find a little piece of bone, remove that if necessary or possible, and then seal that hole.
And you can do that with a traditional laminectomy, and you can add a laminoplasty to that to cover that bony defect. You can do just a laminectomy on one side, right? Or you can do the laminectomy through a little tube as Dr. Beck showed you. That’s just less and less invasive, this type of surgery.
So this extradural surgery with the multi-level fusion, the destructive surgery, I don’t do that anymore. We either do a regular full laminectomy. I think that’s really the safest way and the most effective way. If you don’t really see it coming off to the left side of the spinal cord, it’s really easy and straightforward to go to the right side. But it is more bone removal, it’s a longer recovery. You can either then do a hemi-laminectomy that’s almost as safe and good, or you can try to do it through a tube, but I find that difficult because it’s a bit of a challenge to operate through this little tube to close the dural sac.
These are some publications from Europe. This is from Dr. Beck’s publication, and he very proudly showed the small amount of bone that he removed with the hemi-laminectomy. That’s from Switzerland or Germany. So then of course the French wanted to do a better job. They came out with this paper a few years later where they showed, oh, we do it through a much, much smaller type of bony opening. Then Dr. Beck didn’t really like that either, so he came back with another paper where he showed that he can do it through just as small of a bony opening as the French could do. Not much has changed in the last 70 years.
Between 2009 and 2011, we would do these very extensive destructive procedures, really long incision. Then in the summer of 2011, I did it just intradural, no need for a fusion, anything like that. It’s a bit of a smaller incision. And then we’ve been doing some surgeries through the little tube, just like an incision that’s about an inch.
But basically those are all very similar procedures. These intradural procedures, you remove some bone right from the back, a little bit of bone, a lot of bone, and do a laminoplasty, but it’s basically the same. You open the dural sac, you find the spinal cord, you go around the spinal cord, and then fix it. But that’s kind of, I think Dr. Maya said that embolization is the new kid in town, and for these ventral leaks, kind of the new kid in town is endoscopy. We’ve only just started doing that a couple of months ago.
Here’s what it looks like when we do endoscopic surgery for ventral leaks. It’s just through a single portal. This is a little tube, it’s an 8 mm tube, so really small. That’s the outer diameter. The inner diameter is only 7 millimeters. But it’s only one single port, right? So it’s not like most laparoscopies where you have two or three ports, just a single port.
And I don’t really know how good the results are going to be. We’ve just started doing it earlier this year. This is Dr. Walker, who’s one of our spine surgeons in our group, and is like the virtuoso of using the endoscope. So we do all of these surgeries together.
This was actually the first one we did, a patient I had operated on a few years ago. This is a guy who lives up in the mountains near Los Angeles. This is what his digital subtraction myelogram showed. He had a ventral tear. This is what it looked like at surgery. Surgery went great of course, right? Like what all neurosurgeons would say, but he still had a persistent leak. For a couple of years, symptoms were not that bad, but then his symptoms came back. They were just as bad as they were before. Then we did this endoscopic approach, and you can see here this is the endoscope, one instrument. Here you can see the dural tear, there’s some scarring that’s connecting the dural tear to the remaining disc. When you go intradural, this is the disc that this patient had on this MRI scan, and this is what it looked like on the post-myelogram CT. Right – there’s his disc. It was kind of difficult to remove that completely safely, so I just removed the part that I thought was responsible for the leak, but you can see it was still there. Then it was easy to remove that remaining disc through the endoscope.
This is a little video of what that looks like. This is somebody else, somebody we operated on a week or two ago. This is the digital myelogram. This is a high-speed drill that you introduce through that single port. You kind of expand the opening. You’re going to go through the foramen. You can remove the disc.
Here’s the dural defect. You could see that there, right there. You can see that right there. And then we also put this TechoSil inside and outside of the dura, but maybe that’s not the right type of material. I don’t know. I think we’re going to hear about that tomorrow. I’m looking forward to that. Although it’s kind of strange that a neurologist is going to talk about surgery for ventral leaks, but we’ll see. We’ll give her some really difficult questions.
This is the leak before surgery. You can see that ventral collection, and then this is 48 hours post-op, it’s all gone. And then this is the little entry point of the scope that we used.
But I think really the gold standard is posterior intradural approach. We’ll have to give it a couple of years to see how good endoscopy is. A whole bunch of papers have come out in the last year or two about the success of ventral tear surgery, and the results are pretty good, so about 90% success rate, radiographic success rate of fixing the leak.
This is some results of the first 400 patients that we’ve operated on at Cedars. The radiographic result is really good, 96% cure rate. But as you’ve heard earlier this morning, that doesn’t necessarily mean that all the symptoms of the leak are gone. Complications are low, but it’s not zero. Once every few years, you’ll have a patient who’s totally fine, and then after surgery they cannot move one of their legs. And you can only imagine, that’s a horrible thing to happen for somebody who might be mildly disabled by an occasional headache.
In the last few months we’ve been trying to contact patients who we’ve operated on more than 10 years ago to see the durability of fixing the dura for these ventral leaks. As you can see, over more than 400 person-years of follow-up, we haven’t had a single patient who had a recurrent CSF leak.
The lateral leaks, Dr. Beck has talked about that. He has shown this picture. Those are the different types, right? It can be at the shoulder of the nerve root, it can be at the axilla at the armpit of the nerve root, or it can be far removed from the nerve root at the level of the pedicle. And that approach is all a little bit different. There are some technical nuances.
These lateral leaks are not caused by any type of bony protuberance or a microspur.
We’ve done some research with the genetics institute at Johns Hopkins, sponsored by NIH, and about one out of four patients who have these spontaneous lateral leaks—those are usually younger patients, like in their 20s or 30s—we have found mutations in the fibrillin-2 gene. Fibrillin-1 gene mutations cause Marfan syndrome, and it looks like fibrillin-2 gene mutations cause these spontaneous leaks on the side of the spinal cord.
There were two papers I found recently, and the success rate, especially Dr. Beck’s group, 100% success rate. I think the success rate of ventral leak repair is a little bit higher than the repair of lateral leaks, and I think that’s just because these lateral leaks have an underlying problem with the dura. So I think the success rate is a little bit less with those.
And then there are dorsal leaks. There’s not that much information on that, not much data have been published. There can be different reasons for those. I think that’s kind of a heterogeneous group.
This was somebody where there was a sharp piece of bone that had caused the dural defect. Then this was somebody who had four little what looked like little pseudomeningoceles in the thoracic spine that kind of looked like what you would see after a dural puncture, but this patient never had any type of needle procedure to her spine. And then this just looked like a little dorsal tear just like you can see it ventral to the cord. This was just dorsal to the cord and much, much easier and safer to repair those.
Then all the patients have these simple cysts along the spine, as Dr. Kranz has shown more than a decade ago now. That’s actually normal, right, so a lot of radiology reports won’t even mention that because it’s a variant of normal. Two out of five people happen to have one or more of these simple meningeal diverticula. Occasionally these can rupture.
This was somebody that presented with fluid in the lungs. They tested it for beta-2 transferrin, that’s specific for spinal fluid, and that was positive for beta-2 transferrin, and it was caused by one of these cysts having ruptured and causing fluid around the lungs itself.
Then there can be these cases of what’s called dural ectasia or complex cysts. That’s what that looks like. We try not to operate on that because the dura is so incredibly attenuated and fragile. These are some examples of that on MR myelography. That’s a non-invasive way of looking at spinal fluid. They all look similar but not identical. Everybody is unique. This is why I generally try not to operate on those, but sometimes you have to of course. Like this is one of our residents who is making the incision, and then you see all the CSF pouring out, and it’s not the original leak, that’s the leak that was caused by my surgery. And then this is in normal speed there. Also, you can see CSF really is the most attractive type of body fluid that we have.
So these CSF-venous fistulas, we’ve all been talking about that. I think that’s actually the same pictures that Dr. Maya showed. Most of those are in the thoracic spine, almost always with one of those cysts, but not always. You can also see it anywhere else. You can have it in your neck, lumbar, sacrum as well.
This is kind of a surgical classification. Most of them are pretty far along the nerve root, and you can just put a little clip on the nerve root, that’ll take care of the fistula. Sometimes if you’re lucky, or the patient is lucky, it’s more proximal, easy to identify at the time of surgery, and you don’t really need to even touch the nerve root. This is what this looks like at the time of surgery. The D stands for diverticulum. There’s a little diverticulum, and then the V, that stands for vein. So that’s the vein that’s kind of sucking out the spinal fluid.
This is just an example of where it comes right off the nerve root, very proximal, and it’s easy just to put a little clip on it, and you can cauterize it, of course. Here it was somewhat less obvious, and we have to dissect those veins away from the nerve root, and then you can put a little clip on the vein without disturbing the nerve root itself. Dr. Maya mentioned that some of these are associated with vascular either venous, venolymphatic malformations. Try to stay away from that.
It’s not always as obvious. This is somebody with Blue Rubber Bleb Nevus Syndrome. He had these vascular malformations in his GI tract and along his tongue, the bottom of his feet. This is just an example of when this is the vascular malformation right on the myelogram. This is a cyst. This is really easy to fix. You just put a little clip here, that’s done, you don’t even see the vascular malformation. But like what Dr. Maya showed you earlier, when I operated on somebody at lumbar one level, that was a fistula that was surrounded by a vascular malformation, and that tends to bleed.
For example, this individual young woman from Manhattan has this very complex vascular malformation intraspinal and extraspinal. She has a couple of leaks, fistulas, and then Dr. Maya was able to cure that with embolization. We think that fistulas don’t react as well to blood patches as other types of leaks.
So when we see somebody, not if it’s right away, if we see somebody through the ER within a few weeks of a fistula, we do blood patches. But if it’s after more than a month or so, we just go directly to myelography, and then we do specific treatment. That can be fibrin glue injections. Dr. Carlton Jones showed you that. I think on average it’s about 70% to 80% cure rate. When we started doing it, I think we cured 1 out of 10.
The other, embolization, really, really interesting, I think kind of a similar cure rate, also 70% to 80%. And then surgery. I was surprised that Dr. Carlton Jones, who is almost paying attention, said that that’s high complexity treatment. I think surgery is really low complexity treatment. I think 99% of neurosurgeons will be very easily able to fix CSF-venous fistulas, unlike fibrin glue. I think most neuroradiologists at this time would not really be able to fix those. I think it’s actually a really straightforward procedure. You kind of have to scratch your head, right? You put a clip on the nerve root sleeve and there’s still a fistula. That should be an extremely rare occasion.
I won’t really go through this because we’re running out of time, but there’s very little you have to fear of clipping the nerve root, except if it’s in the cervical or lumbar spine. If it’s just in the thoracic spine some people are worried about the artery of Adamkiewicz, you know, that can cause a stroke of the spinal cord. I know I’ve clipped dozens of arteries of Adamkiewicz, nothing has ever happened. I did a study when I lived here in Amsterdam in the 80s, it’s called “Does the artery of Adamkiewicz exist in the albino rat?” We wanted to come up with an animal model to cause spinal cord infarction, and the rat doesn’t have an artery of Adamkiewicz. There are lots of different arteries supplying the spinal cord. So I think actually people are very similar to albino rats, at least as far as that’s concerned.
Okay. I’ll just talk briefly about this. The first 10 years that we did that, never saw any complication, but then over the last few years there have been some patients, usually white middle-aged men, I don’t know why, and they complain of this bulge. We clip the lower thoracic nerve root and then days or weeks, months later they come back and they say, “Oh, I developed a hernia, I’m going to have hernia surgery.” It’s just because you interrupt the motor supply through that lower thoracic nerve root, and that can be kind of disfiguring. Most people don’t like it.
This, you also can get that with embolization. This was somebody who had two levels embolized, T11, T12. Had a little bit of pain, had this bulge in his abdomen. Unfortunately, his general surgeon thought it was a hernia, did a laparotomy, accidentally caused a little nick in his small bowel. He was in the hospital for about a month trying to recover from that. So that’s really something you have to alert your patients to, not just with surgery but also with embolization.
Dr. Maya talked about this before. We did a little poll a few months ago and I asked patients who I had operated on, “Why did you choose surgery over embolization?” And just like what Dr. Maya mentioned, most of them said, one and done, I’m from far away, just want to get it over and done with, go back home. Two very sweet patients just didn’t want to disappoint me. Two other patients were really worried about the long-term effects of Onyx. I think there was something on the internet. One patient had had embolization of the middle meningeal artery. That’s a treatment for subdural hematomas. This was not done by anybody here, but it was a pretty major medical center somewhere else in the country, and they had caused permanent blindness in one eye. So she said, “Oh, don’t talk to me about embolization.” And then one patient said, “What is embolization?” And that’s kind of the right answer. I really like that.
You already saw this beautiful, beautiful imaging of Dr. Lützen and Dr. Beck. These are just, and I’ll end with this, these distal nerve root tears. We’ve only seen less than one-tenth of a percent of our patients have that, but I think it does occur.
So again, imaging is key to success. We try to become less and less invasive. I think endoscopy is going to play a really large role in fixing ventral leaks in the future. Dural repair surgery is durable, and I think it is time for a randomized control trial. I think it’s easy to do, right? You do embolization versus clipping for fistulas. All right. Thank you.