Dr. Jürgen Beck at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Jürgen Beck, Director, Department of Neurosurgery, University of Freiburg, Germany, presented this talk titled “New Diagnostics: Cine MRI and CSF-PET” at the 2023 Cedars-Sinai Intracranial Hypotension Conference on July 8, 2023. The conference was hosted by Cedars-Sinai with generous support from the Spinal CSF Leak Foundation in Kohala Coast, Hawaii.
Transcript
[00:00:09] I would like to thank Wouter Schievink, Marcus Stoodley, and Connie Deline for the possibility to present our approaches, or maybe new approaches, to new diagnostics called Cine MRI and CSF-PET from my team in Freiburg. First of all, thank you very much to my excellent team in Freiburg, and these are just the people taking special interest in CSF leaks, including, of course, neurobiology, neurology, and even nuclear medicine.
[00:00:45] So, I had a long interest in finding additional diagnostic ways and CSF leaks and did a lot of work at the time when I worked in Bern in Switzerland. And, uh, it’s ultrasound of the optic nerve sheath and it was a big difference when the patients were lying or standing, but unfortunately nobody’s using it.
[00:01:06] And even if we don’t use it anymore because it’s so user dependent, and probably we use it from time to time after surgery or after embolization or after patching, whether to discriminate still having low pressure headache or whether there’s a high pressure rebound intercranial hypertension. And there was another hobby developed in, in Bern at the time, a lumbar infusion testing.
[00:01:32] And I think it’s very natural, very obvious if you do a lumbar puncture any way to challenge the system with fluid. And of course you should have a completely different response of the system when there is a hole. So this, we thought, okay, that’s the solution to this disease. And in the beginning, you have clearly different pressure curves and you have two different resistance to CSF outflow. But unfortunately, only the beginning of the season after three months, it levels out, and even if there’s a reverse response of the CSF system, to which we don’t really have an answer yet. So do we need, after seeing this beautiful presentations of Marcel and Peter, do we still need other diagnostics stuff? I think we do need.
[00:02:20] I’ve never heard anyone presenting the radiation dose that we are doing to our patients with these studies, and they are tremendous. So this is just, we are a referral center in Germany, we get a lot of patients from, from other centers and from all over Europe. And this is just one example of a dose a patient received at one day during diagnostic study for CSF. This is kind of radiotherapy. And we, we try to, to, to deliver the dosing with each study we do, and they are really quite significant. So there might be a way to improve our diagnostic workup. And then we get back to radionuclide cisternography, which was, I think, very, a breakthrough study in, in 50 years ago, published in Nature, showing first time in man, the CSF motion. The CSF is running from the ventricles to the subarachnoid space and vice versa, which was very interesting. But nowadays, cisternography isn’t done by anyone for the reason it has a very low spatial resolution or temporal resolution, and it takes very long several days to perform the study.
[00:03:32] And this is a slide by my nuclear medicine colleague Professor Maya, he says, even nuclear medicine practitioners are reluctant to perform a radionuclide cisternography to the stone-age image quality. But there is a progress, and we can do cisternography with PET, position emission tomography, and there are clear advantages of PET, so we have a way better spatial resolution.
[00:03:58] Yeah, it’s it’s faster, you can scan the whole body in 10 minutes. It’s a hybrid imaging system, and I think there’s the beauty of it. You can quantify, you can now really quantify what’s going on in the CSF system. You have a certain amount of a dose that is in the system and then you can measure the decline and a washout.
[00:04:19] This is a clear advantage. So our question was, are there any direct or indirect signs and quantitative signs of the halftime of the CSF washout in patients with spinal CSF leaks? And is CSF-PET useful for A, answering the question, is there a CSF leak? And B, where is the leak? Of course, the questions we all have to ask ourselves.
[00:04:45] It’s done in five hours, so we inject 40 to 50 millibecquerels of gallium dota and do a scan one hour, three hours and five hours after injection. And very interestingly, we had a very high sensitivity of the lack of trace accumulation over the convexities after five hours. So on the upper row there is normal physiologic accumulation of the tracer over the several hemispheres.
[00:05:14] And in the presence of a spine CSF leak, you don’t see that accumulation over the several hemispheres. So this is an indirect sign. Another indirect sign is extracellular tracer accumulation. So, in the presence of a CSF leak now you have a very high specificity. You have next to the thecal sac, you have this trace accumulation, and it is very important now to recognize this is not a localizing sign.
[00:05:41] This is a sign for the presence for the accumulation of the tracer next to the thecal sac, not where it egresses out of the thecal sac. So just to give you some more examples, um, when you have no leak, there is a clear accumulation over the several hemispheres. When you have an artificial leak here at the injection side, it’s still accumulation over the hemispheres, but with a spinal CSF leak you have no accumulation, and you have this para spinal, not localizing accumulation of, of the nuclide.
[00:06:16] So to summarize, gallium dota positron emission tomography for the diagnosis of spinal CSF leaks: we have a quantitative sign, the T halftime, if it is below 7.9, there is, uh, a washout, there is an egress through the spinal CSF leak, goes through the spinal CSF fistula. And, uh, you have this indirect science, not quantitative, whether it is accumulation over the hemispheres or not, or whether they have this paraspinal accumulation.
[00:06:46] So, it’s very good for saying yes, there is a leak, it’s not useful for localizing. But still, if you look at the millisieverts, the whole gadolinium CSF PET is only around 5 millisieverts as compared to the over 200 we’ve seen with repetitive myelograms. And probably—if I think of the patients Wouter Shievink has introduced to us with the behavioral variant, frontotemporal dementia, elucidating pathophysiology—and so probably this could be a study used as a gatekeeper to include patients and further end repetitive myelograms and imaging.
[00:07:28] So this has been done in nuclear medicine department together with our neurosurgical and neuroradiological department in Freiburg. And I would like to introduce another additional diagnostic tool: phase-contrast MRI. It’s very fast, it takes us, the whole sequence takes two minutes, and you don’t have to use intrathecal or intravenous, no contrast at all.
[00:07:51] This is how it looks like. And you’re looking at the motion of the spinal cord and the backflow and outflow and inflow of the CSF around the level of C2. The question is how much support is there in SIH, where there’s low volume, low pressure, loss of CSF, probably the motion of the spinal cord at this level is increased as compared to the normal state.
[00:08:18] In IIH, it might even be decreased, the motion there. And, um, these are the results. So, in SIH patients, indeed, there was a higher motion as compared to healthy patients, and even way higher than, uh, as compared to IIH patients. And interestingly, um, total displacement and also the velocity of the speed of the spinal cord motion was significantly, highly significantly different between patients and controls.
[00:08:48] And for the proof of concept, we included only so-called SLEC-positive patients with epidural fluid, and, of course, in these patients, you already know that they have a CSF leak, there is epidural fluid. But still, you could use this very fast, non-contrast diagnostic test to elucidate the pathophysiology in these patients.
[00:09:10] Again, think about BVFTD patients, for instance, or in patients where you are insecure. And, but the next step is, now let’s include patients with, um, fistulas. And we did that in the meantime. And now we over doubled our patient number and we have 47 SIH patients included. Also half of them now have type 2 and have type 3 lesions, so CSF venous fistulas, and the results are very robust. And even in type 3 lesions, there is a higher speed of spinal cord motion. And probably this is even also related to pathophysiology and all the signs and symptoms in these patients, because there’s an abnormal motion of the brainstem and the spinal cord.
[00:09:51] And so luckily, the results are corroborated and we have a high sensitivity and specificity as compared to healthy controls or healthy participants. This has been done knowing 57 patients and 68 healthy controls. So it might very well be useful—again, two minutes, fast scan, no gadolinium— for triggering for the diagnostics and also for follow up purposes, whether a fistula reopens or whether there’s a new CSF fistula forming. And probably also for the leftovers, so to say.
[00:10:27] So in the unknown group, and if you looked more specifically or more closely into this unknown group, there were 1, 2, 3, 4, 5 patients with a high Bern score. And probably we just missed finding the leak in these patients, and these are the candidates for redoing myelograms over and over again.
[00:10:48] And we can also use it for monitoring. This is a case of a CSF venous fistula, and you see before a targeted fibrin glue patch, there was this pattern of motion. They had relapse and there was basically no difference. And after surgery, you know, okay, now the fistula is closed. And again, two minute, no gadolinium. You can also use it—and a very interesting data I haven’t presented yet—you can use it also in IIH. So there seems to be a spectrum of the spinal cord motion from SIH to normal patients or to normal controls and to IIH patients. And this is a very difficult patient group, IIH patients. And after standing, it clearly normalized, but still having headache.
[00:11:29] And we showed that the stand was open. And it was a medication overuse differential in this patient. So it’s fast, two minutes, no gad. It’s diagnostic also for the CSF venous fistulas and probably for the unknown patients and probably to elucidate further on the pathophysiology. And also for monitoring patients with this disease.
[00:11:53] Special thanks here to Katarina Wolf, who is a bright neurologist who is doing her neurosurgery residency at the moment, and of course, also to Niklas Lützen and Horst Ulbach, and also to Professor Meier from nuclear medicine. And to give you some more news in diagnostics—so I think, Peter has mentioned it already—there is this ultrahigh-resolution cone-beam computed tomography just published a couple of weeks ago, where you can really clearly depict the CSF venous fistula with a very high image quality. And I’ve already shown these images in Freiburg at our meeting. This is a piece of art, I think, done by Niklas Lützen and Horst Ulbach, where you can clearly see the fluid level here in the cyst and then the CSF, the venous fistula is filling here, very high image resolution.
[00:12:45] And last, probably, we can also use just simply another location of looking, looking for the leak. This was also already mentioned by the speakers about imaging. Probably you missed the sacrum or the lower parts of the spinal cord. And with heavily T2 weighted imaging, we were able to find probably the, it’s just another location probably, it’s a fourth type of CSF leak, which we call the sacral dual tears, and which we found and published this year that there’s probably less than 10 percent in our cohort, it is indeed worthwhile including the lower part and the sacrum in high T2 weighted MRI imaging and you can find and localize these sacral CSF lesions, and usually they respond very well to blood patching.
[00:13:42] And probably the last slide for new diagnostics is what we are looking for is for a blood test for SIH. And there’s just very, very, very preliminary data from my group in Freiburg. This is a prospective proteomic study we did only in, I think, 10 or 20 patients, and there is a clearly distinct pattern of proteomics and metabolomics upregulation in the CSF. So this would be probably a target for further investigation. And so that’s it from Freiburg. Thank you very much for your attention.