Slides
Transcript
Good morning. Thank you, Jill. I’m very honored and flattered to be in this beautiful place and to contribute to this marvelous meeting. Now we’ve heard two talks how important or how severely affected patients are, and how there’s still a big discrepancy between how patients and caregivers perceive this disease. And patients really do lose their lives. They do lose their jobs.
And now my job is to tell you we’re talking about spontaneous intracranial hypertension caused by a spontaneous leak at the level of the spine—a spinal CSF leak. How do these leaks look like now? And I will try to give you some perspective from the surgical—to the surgical camera, to the surgical microscope, and also with the help in the later part with combining imaging, modern imaging, high, ultra-high resolution cone beam CT and surgery.
There are many types of leaks, but basically over the last decade probably 95% or 99% of the patients can be stratified into one of a type 1, type 2, or type 3 leak. It’s basically a dural tear. It’s a dural tear either on the medial ventral aspect, so a ventral leak, or on the lateral aspect, so we call it a lateral leak. And roughly 10 years ago, Wouter discovered the third form of direct CSF-venous fistula.
All this is only possible, what we show at the beginning. I would like to thank my fantastic team back in Bern and now in Freiburg—without the help of these people nothing would be possible. It is the same journey in Freiburg, from probably one or two patients a year, now due to the work of my team, we have our own specialized ward only for CSF patients. We treat over 500 patients a year. I personally did over 500 surgeries. We do many, many blood patches.
Bottom line is you do need a new radiologist who loves the disease, who is really into it, focuses on it, and tells the surgeon where exactly the leak is. There is no way, there is no way, performing surgery without exactly knowing the height and the level and the aspect of the leak at the dura. There is no room for exploratory surgery.
Having two type 1 leaks, the ventral leaks caused probably most of them, if not all, by a tiny little microspur, discogenic, whatsoever—a tiny little, smaller than a corn of rice, piece of bone and cartilage cutting the dura like a knife. And we approached these patients from the posterior aspect.
This is a historic video. This is not minimally invasive, but you have a good overview. This was one of the first patients, a physician herself, operated on. She told me she’s functionally paraplegic, couldn’t stand up, and you see CSF oozing out. And it’s very exciting because I didn’t know what to expect at the time.
So I cut the dura to search for the leak in the dura dorsal aspect. Grab the dentate ligament, and then you see this little tiny microspur, iceberg knife, whatsoever, that really is into the dura, cutting it open, and most importantly holding it open. So blood patching doesn’t work in these patients. Removed the microspur at that time, sutured it. You can clearly see one 6-0 microsuture and the leak is closed and the thing is over.
This is because I was trained as a vascular and tumor surgeon, as Wouter, doing cavernomas out of the spine. So we were knowing how to be able to rotate the spinal cord a little bit to have access to the ventral side. Nowadays with my team in Freiburg we do this minimally invasively. Only two centimeters skin incision, tubular retractors, bony removal is minimal. It’s less than a coin. It’s like 15 mm or so, though no weight-bearing structures are affected. This is really minimally invasive. And then the steps are that we first cut the dura, then we cut the dentate ligament. This is the key step, to free, to liberate the spinal cord, that it is able to move to the side. Don’t push to the spinal cord. Always push on the lateral side at the dura to move the whole dural tear into your field of view, not otherwise.
This is a video how you can open the dura dorsolaterally, and then you have on the ventral aspect, you have always the two to four mm similar-looking ventral dural tear. The microspur, it’s smaller than a corn of rice, very tiny. Sometimes it’s even dissolved. And then you can close it. We use this fibrin mesh TachoSil from the outside, and you can see the yellow surface of the glue looking through the tear. Pinch it a little bit into the hole and cover it with a second layer. So we do a sandwich patch, and there’s no need for suturing it. It’s just an 8 mm opening of the dura. And again it’s not spine surgery, it’s spinal cord surgery. You have to cut the dentate. You have to know how to handle the arachnoid, how to handle the membranes, how to handle the ligaments, and not to touch the spinal cord.
And this now is, I think, one of the key slides. What do we see through the microscope? What you don’t see on imaging or don’t see directly on imaging. These are these membranes. These membranes that connect a kind of a vessel from the intradural to the extradural space. This was only the origin, the initial cut, but then somehow these membranes form or herniate and keep the ventral leaks open. So that’s the reason why blood patching doesn’t really work in these cases.
And same video once again. Now you saw it. Look again. Same video. You think it’s closed. There’s a membrane. It’s not closed because I sucked out the CSF. The membranes collapse, but the membranes are still there and hold the channel from the intra to the extrathecal space open. So it’s all on membranes. I would love to have it drawn even bigger, but this is the key step. And these membranes even prevent—why does a 2 mm channel dura not heal? Usually they all heal by themselves, because the membranes prevent the dura from healing.
And if you don’t believe me, my team has built a so-called funnel phantom box. So this is a model of the spine and these membranes, and we put it outside in the poster area. You can look at it. You can see how we think these membranes work and why it is so important to cut these tiny little membranes. The suturing is not the key issue. It’s the key issue to disconnect the intrathecal from the extrathecal space. And Sophia and Victoria, the op nurses, are here and can explain how we approach these membranes. And you don’t see them on imaging, but you see the sequelae of these membranes.
The SLECs have been mentioned very often right now. What are SLECs? Extrathecal fluid collections. If you look at the initial second of a myelogram, the contrast runs as fast extrathecally as intrathecally. There’s no barrier for fluid. But if you look at the MRI in the patient after a couple of days, it’s contained. And why is it contained? Because you see these SLECs, they’re all contained by these membranes that hold the fluid together.
So once again, this is a still image out of a surgical video. This is the dura. This hook is parallel to the dura, and here it is the border from the dura, the transition zone to this membrane. Now I cut this open and only then CSF is oozing out.
This now is an over clear case. It’s a case with superficial siderosis and, as Matthijs told you, it’s a longstanding disease. So the membranes are very, very prominent. Just to show you again how important these structures are. Typical case microspur SLEC collection, then the 2 cm skin incision, microtubular retractor, and get access to the dura and to the SLEC space. Then I dissect a little bit the epidural space and you see there is stuff that doesn’t normally belong there. If you do disc surgery, there are no membranes at all. So dissect it a little bit, cut this vein, and then you see now this is the membrane, this is the dura, and this is the membrane. It’s a pathologic structure.
To further demonstrate it, we injected sodium fluorescein at another level in the spine. It’s a green substance, it’s fluorescent, and you can clearly see it’s watertight. This is the dura, looks a little bit different, and this is a neo-membrane. And only when I open it with a sharp hook suddenly all the CSF is oozing out.
And if you are not used to this surgery, surgeons misdiagnose the membrane for the dura because if you look into it, it’s beautiful. It’s a beautiful neospace. It looks like the real dura, and so they try to suture the neomembrane. You’ll fail. You cannot suture the neomembrane, and you haven’t even attacked the dural tear yet. So don’t be misguided by membranes. And only then, only then, I open the dura.
This is now the crucial step to open the dura. Cut the dentate ligament and now lateralize the dura. Bring the dural tear into your field of view. And now you see yellow instead of green. Why yellow? Because it’s superficial siderosis. It’s a longstanding disease. And you see this is the tear. These are these membranes going through. You need to be aware of the anatomy. The microanatomy can only be seen under the microscope.
So we believe that it is all about these membranes and that they are paramount for getting our patients healthy again. And this is the reason why blood patching, and Eike has shown this nicely, only helps I think in 10% or so of the patients.
So now if you know the anatomy, that there are barriers for the blood that cannot reach the actual site of injury. It doesn’t work.
This was a wonderful paper found by a colleague in 1918 in New York, and he already did surgery at the time and post-lumbar puncture syndrome. And he also pinpointed the very detailed anatomy—where is the hole, where’s the hole in the dura, and where’s the hole in the arachnoid? And only if it matches and if the membranes pinch through the hole, then it does stop the healing process, and you get the disease. This is at least currently our understanding.
So microsurgical anatomy reveals these microsurgical anatomy that you now can fuse with what we know from imaging to get a bigger picture of this disease. And it holds the channel open like a vessel. And there’s even a high-flow CSF from the inside to the outside. And there is this flow-void sign, which we all call Lalani-sign. And Niklas has shown recently that it is indeed a high-flow system. This is now important for you how to diagnose it. And for ventral leaks, you need a fast system. Since Niklas, correct me, since it’s a high-flow system and the thing is happening in the first seconds. So it’s a really high-flow system.
What also gives us insight into the anatomy, how does this disease really look like—idiopathic spinal cord herniations. This is also what was used to set there—double dura with genetic disease. It is not, it is not. It’s like superficial siderosis, a long-term sequela of spinal CSF leaks. And that is not only the arachnoid herniating out of the [dural] rent, trying to seal the rent. It’s also the spinal cord itself herniating out, enlarging the tear. And if you look closely to your spinal cord herniation cases, you almost always see SLECs, because this is the origin of the disease.
Next, type 2 – lateral leaks, or cysts. This is a cyst when you operate on it a couple of days or weeks after the patient was symptomatic. It is very thin, translucent. This is the dura. This is a cyst. It looks like arachnoid. Why? Because it is arachnoid. It is the arachnoid. It herniates through the dural tear. So the origin, the agent itself is the tear in the dura. And this is just a pressure thing. It’s like a herniation in belly surgery. This is a patient, how it looks like after a couple of months. Same principle, just more scarring and not translucent anymore.
And again, please another video. This now is a lateral leak where you do again always pinpoint the exact site of the lesion. There’s no way of exploratory surgery. Then you need to dissect it. And then you see that the dura, the normal dura, lateralizes a little bit to expose the root where the site of the cyst is. And in this case, the cyst is on the outgoing root. This is the dura. This is the outgoing root. And you see its outpouch, and you can just simply reduce it, put it back into the dura, and that’s it. And the origin of the disease is the tear of the dura, not the cyst itself.
So we put this on the cover—we were able to put this on the cover of Neurology 10 years ago. This is a cyst, but the CSF is oozing out next to the cyst.
So exactly, yeah, it is probably important to know that the dural tear is the origin of the leak and it is mainly at the axilla. Sometimes it’s a little bit distant from the axilla. And happily, 10 years later, people from Mayo Clinic did the same observation. So it’s not the dural—it’s the diverticulum as a secondary manifestation of the dural tear, and not the source of the leak. The actual CSF leak was from the edges of the dural tear itself. So the exact same mechanism.
Wouter put some science on it and gathered his experience. Sorry for that. And this is probably not so important whether it’s two or three millimeters apart. I think this is all a locus minoris resistentiae here. There’s a thin area whether it’s in the axilla or in the shoulder. Sometimes it’s a little bit distant. But most of the leaks, as shown in the video, the leak was on the root itself – it’s in this area. So there seems to be a weak predisposition site for spinal lateral leaks.
Then imaging, high resolution, ultra-high resolution cone beam myelography, depicts very, very nicely the anatomy. And you can really one-to-one transpose it to the surgical theatre. And if you look at the images Niklas and Horst are producing, you can clearly see the roots, the CSF, the tear. And this is, for instance, one of the leaks that normal CT, ultra-high resolution CT—you can see here, exit the roots. Here is the dural tear with a diverticulum, and this is the surgical situs, and exactly 2 millimeters below the root there is the dural tear. And this is the anatomy. So this is fantastic and it really helps us to pinpoint the lesion, to only attack this 2 mm where the disease is happening, and then it’s surgically an easy procedure just to cut it, to reduce it, and to suture the dura.
This is also nicely shown by Niklas together with Lalani. There’s a bottom branch sign that you can focus your diagnostic attack more to decide which cyst is the leaking cyst. So you have hints, but still do probably dynamic studies. And there’s another dynamic study, and there’s I think a poster out, that you can see that also in lateral leaks there might be high flow leaks.
Summarizing this, there is a clear notion now from Freiburg, from Cedars, and from the Mayo Clinic that probably these lateral leaks are just out-pouched arachnoid, and the cause is a lateral tear in the dura – as simple as that. Even the larger cysts, this was nicely shown by Mark Mamlouk and his group, that the larger arachnoid cysts that we used to call them are probably in many if not most cases also a long-term sequelae of spinal CSF leaks. So same procedure, same mechanism. You have a dural tear, and the arachnoid is herniating out and getting even big and bizarre, impinging the spinal cord causing symptoms. And this is important for the surgeon, how you can attack these lesions that you know that’s only one tiny 1 mm big tear in the dura and the large extradural cyst is only a sequela of this tear.
Then finally, as Wouter has described it 10 years ago, there’s a new form of CSF leaks, and we had a hard time in Europe first believing you, Wouter, that this does exist because we didn’t find it on imaging. And when we gathered experience and applied some tricks—lateral decubitus position, pre-pressurization, forced inspiration, resisted inspiration—it’s easy nowadays. And Horst and Niklas finds, I think, from 0 to 5% a year. Currently, I think we have 50% of our patients do have CSF-venous fistulas.
And this is a photograph from Wouter – how these leaks look like during surgery. And there are veins, and probably here filled with CSF, and more translucent probably here filled with blood.
So to show you that this really does exist, another surgical video. First diagnostic—always mandatory—good images up front that you really know where to put your tubular retractor in, and then you open the small tiny bony opening. I only dissect the epidural space. No coagulation because I want to find the offending vessel. And so then there it is. You see here a vein which is halfway filled with CSF and here with blood. That’s the root. That’s the dura. That’s the thecal sac. This is the finding. Now you can tell, very subtle finding. Is it really true? Again, we injected sodium fluorescein to visualize it. And you see the green stuff is normal, and slightly the vein is filling after a couple of seconds. And if I push with the suction device, you see how the CSF is oozing out through this vein. And that’s the only single offending disease and vessel. The other vessels are not yellow. So it’s only one tiny singular channel. This is very enigmatic and fantastic, and I still do not understand why there is no backflow. Maybe there is, but even very, very subtle and even very tiny. Direct CSF-fistulas nowadays, at least in the centers, it’s a standard of care.
So 95% or 99% of the patients do have one of these type 1, type 2, type 3 leaks. It’s good for us, for the neurologists, radiologists, and neurosurgeons to have now a group formation. We can stratify the patients, but there’s more. There are rare diseases. Probably you also found these sacral dural tears where there is no clear SLEC but more diffuse contrast egress in sacral patients. I cannot tell you how the anatomy looks like during surgery because we refrain from doing surgery in these patients, because a blood patch is usually very, very helpful in these mainly young women that present, at least to Freiburg, with these sacral dural tears.
There’s one another—probably it’s a singular finding—but this is not a rendering, this is an original high-resolution cone beam CT again done by Niklas. It’s beautiful, and it’s really beautiful, and you can see it’s not connected to a vein. It’s connected to the lymphatic system, and we even referred this patient to other centers in Germany. They proved it’s the lymphatic system. We did surgery. Surgery was easy because we exactly knew where the spot is. And it was indeed a direct CSF lymphatic fistula. Probably there are more. I don’t know yet. But we clearly need to refine our diagnostic tools to find such things.
And there’s even more. I think our understanding is now way better of CSF leaks, but there are still so many riddles and still enigmatic. What is this area around the exiting roof? What if there’s a tear, the patient becomes suddenly symptomatic, and we can diagnose them? But what if there is just locus of less resistance, and there’s a compliance change from standing to sitting, going back to probably a genetic disease, probably soft tissue weakness? So to say, if—as drawn here—if the thecal sac is wider during standing than sitting and we are symptomatic. I think Dr. Bahram Mokri introduced this concept first, so the closed SIH, so to say. There might be a disease where there is no need at all of a CSF leak, but you still have orthostasis. We don’t know yet. Very interesting, very intriguing.
I’m so happy to be part of this journey with my fantastic team. And I think from my perspective, it’s a duropathy. It’s a dural disease. And do not forget the surrounding structures—for instance, the microspur that cause some of these tears, the microanatomy of the dura itself, and of course the other layer, the arachnoid, that is probably key whether the body is able to heal itself or not.
If you want to see all these surgeries, you can always come to our visitor centers, to Freiburg and to Bern. We do a common course. We join up. We are all family and do surgeries in these days twice a year. Then you can come and see us working in Freiburg and trying to further CSF leaks. Thank you so much.