Slides
Transcript
Okay, good morning everyone, and thank you to the Spinal CSF Foundation and to Dr. Callen for inviting me. I’m very pleased to be talking to you today about CSF-venous fistulas – known and unknown. This talk is going to start exactly the way that you thought it would start – with the discussion of epistemology.
So, one of the things that epistemology is sort of the science of understanding knowledge. And since we’re talking about knowns and unknowns, it seems relevant. One of the concepts in epistemology is something called epistemic humility, and that is basically the intellectual virtue recognizing that knowledge is always provisional. It’s incomplete, and it involves being open to learning and new knowledge acquisition and willing to change your beliefs and assumptions.
Immanuel Kant was a German philosopher who wrote a lot about this. But perhaps a little more contemporaneously, there’s this master of philosophy from this century – Donald Rumsfeld – who was actually both the youngest and the oldest Secretary of Defense, if you didn’t know that. But he said this quote:
“Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tends to be the difficult ones.”
And so, since we’re talking about the knowns and the unknowns, I’m going to use Donald Rumsfeld’s pattern of epistemology to talk about these things. So, we’re going to start with the known knowns.
So, we know that CSF-venous fistulas exist. In 2014, which really wasn’t that long ago, Dr. Schievink and Dr. Maya and Dr. Moser at Cedars-Sinai published a case report of three patients who had CSF-venous fistulas. And even though at the time we thought this was a rare phenomenon, it turns out to be quite common.
We also know that these fistulas go into the veins around the spine, into what we call the epidural venous plexus. And there are different parts of the epidural venous plexus – there’s what we call the EVVP, or the external epidural venous plexus, which surrounds the spinal column and the vertebral bodies. This is an example of a venous fistula into that EVVP. There are fistulas into the IVVP, or the internal epidural venous plexus, which are the veins that surround the thecal sac within the spinal canal. And here’s an example of that. And then there’s a network of veins that go through the vertebral bodies called the basivertebral venous plexus, and we can also see fistulas into those as well.
One thing I think that everybody pretty much agrees on now is that decubitus imaging is the way to go for detecting CSF-venous fistulas. This was described both on CT and on digital subtraction myelography in 2019, and this has really revolutionized and greatly enhanced our ability to detect venous fistulas. So, if you want to detect these, you really need to be laying on your side.
We know that venous fistulas can be detected with a variety of different imaging modalities, including digital subtraction myelography or fluoroscopic-based techniques, and with CT as well. These are some nice illustrations from the group in Switzerland about how one is positioned for a digital subtraction myelogram – laying on your side on a tilting table. Contrast is injected into the lower part of the spine and runs up the spine, and images are acquired as that is occurring.
Now, unfortunately, CT scanners don’t really tilt, so you need some sort of mechanism to tilt the patient. Engineers look at me with a look of horror whenever I try to describe a tilting table on CT. One said to me, “This is literally the last thing that we want to happen under CT.” So, you have to make it yourself. There’s a variety of different ways of doing this, and some people have come up with a lot of good ideas. We use just a simple wedge that looks like a big doorstop that people will lay on their side during imaging.
This is an example of how a patient is often positioned, and depending on what technique exactly you want to use will determine how the patient’s head is positioned. We do what some may consider to be a static decubitus CT myelogram where the head is elevated in order to pool contrast in the thoracic spine. Other folks will do what’s called a dynamic, where they want the head down because they want contrast to be moving during the process of imaging. Both of those can be effective ways of detecting CSF-venous fistulas.
So, we can use fluoroscopy or CT myelography. Patients are in the decubitus position. And we know that these mostly – although not exclusively – but the large majority of them occur in the thoracic spine. So, that’s where we tend to focus our evaluation, at least in the beginning.
This is an example of a myelogram where a patient wasn’t imaged in the decubitus position and they were placed in the prone position. You can see that there is a fistula here, although you may wonder where that is because it’s actually quite difficult to see. But, when the patient is turned into the decubitus position – same patient, same scan, just turned on their side – now you’re starting to see a little bit more of this fistula. On the last slide, or sorry, this is where you’re starting to see a little bit of that fistula.
And on the last image, this is a separate scan where contrast was injected, and the patient was kept on their side. Now the fistula is very obvious. So, this is just an example of how that decubitus position is really important.
Another known known is that there are multiple ways of treating CSF-venous fistulas. And in this session, the speakers that follow are going to talk about surgery, venous embolization, and epidural blood patch. We know that all three of these methods, to some extent, are capable of treating venous fistulas. We don’t know exactly how they compare with each other. The literature is not totally complete, and we all have our own preferred algorithms. But we know that all three of these treatments work to some degree.
Okay, now we’re going to talk about the known unknown – so what are things that we know we don’t know? The first is we don’t really know why CSF-venous fistulas form. Despite a lot of smart people thinking about this, this is very difficult to prove directly. Now we do know that the CSF and the veins around the spine occur in close proximity to each other, and there are normal places where spinal fluid is resorbed into the veins around the spine. So, we know that anatomically, the veins and the spinal fluid are in close contact.
If you look at this paper from 1976 and you look at where these connections, called arachnoid granulations occur, they occur predominantly in the thoracic spine. And so, they occur in the same areas where CSF-venous fistulas tend to occur. This is a schematic illustration from another paper just illustrating the fact that they oftentimes occur along the nerve roots. Occasionally, they occur directly off of the thecal sac, and they sort of push their way into blood vessels under normal conditions.
This is a nice illustration from Lydia Gregg at Johns Hopkins University. She is a talented illustrator – just illustrating that same relationship. And so, the one hypothesis is that when these rupture, it allows a venous fistula to form.
One thing that we also don’t know about is the relative factors that make these venous fistulas more conspicuous. So, people have talked about timing, and this is from a paper from Dr. Mark, who found that under digital subtraction myelography they typically will see these fistulas at about 9 seconds, and they’ll persist for about 45 seconds. In digital subtraction myelography, the contrast is moving. So not only is time changing but the density of the contrast is also changing, and it’s hard to know exactly what the contributions of each of those are.
Here’s a paper from Dr. Callen talking about timing as well, using a slightly different technique under CT. Here they found that in about half of cases, the first pass imaging showed the fistula at its best. This is a more recent paper from Diogo Edelmuth, who’s in São Paulo, Brazil, and he basically looked at factors that promoted visualization of fistulas. What he found is that when the contrast concentration in the spinal thecal sac was the highest, then those were the fistulas that were correlated with visualization of fistulas. You see all the green triangles – those are fistulas that are visible – and you can see they tend to cluster when the density is highest. At least in this study, this was much more important than timing, but I think we still have some disagreements about whether that’s the case. Both methods are capable of detecting fistulas at the end.
We know that breathing can affect the visibility of fistulas. This is a case I did under fluoroscopy. Here when the patient is taking a deep breath in, you can see a venous fistula. As I ask the patient to hold their breath and do a Valsalva maneuver, you can see that fistula becomes progressively less extensive and then disappears altogether. Then when the patient takes a deep breath in again, it comes back. So, we know that respiration has some effect. We don’t know exactly what the effect is. We know that in some cases, doing a technique called resisted inspiration – where you suck in through a straw – may promote visualization of some fistulas, but this is not 100%.
This was a paper that came out of Duke. We had eight patients. This is a very small number. And among those eight patients, the resisted inspiration was helpful in a little bit over half of cases. So, this is not a cure-all. This is not the case that all fistulas are going to be most visible this way. I think people are still interested in this topic. We don’t really know whether the different modalities we talked about, whether one is better than the other, and it’s very difficult to compare those.
This was a paper of Dr. Lützen and Dr. Beck, which came out of Freiburg, Germany, where patients got both DSM and CT myelography. They found more fistulas with CT, but I think this is a topic that is not fully resolved yet. I think we don’t – no single paper from any institution should be considered the final word. So, I think this is something that is interesting, but we don’t know for sure yet whether one of these modalities is better. Both are capable of identifying fistulas.
You may hear talk about photon-counting CT. This is a relatively new technology, and I’m sorry this video won’t play of the standard CT, but for the photon-counting CT, it basically provides us with higher spatial resolution and allows us to see some very small fistulas – like in this particular case. The fistula, you’ll see, coming out here – that can be difficult to identify in some cases, but we’ve never really compared randomized patients to either standard or photon-counting CT, so we don’t know for sure what the contribution of photon-counting CT is.
These are a couple of papers. Dr. Madhavan was the first author on both of these. One of them also came from Duke. Here, we’re talking about six cases in the first paper and then sort of a case series of a larger group of patients – 57 patients, but uncontrolled, so no comparator group. So I think this is something we’re all interested in. Perhaps there is some utility to photon-counting CT. I personally think there will, be but recognize that that’s not something that we’re completely sure about yet.
So, we’re in the section on known unknowns, and I would say the literature is still inconclusive on a lot of these pieces. I think if you’re a patient, if you’re getting a good high-quality decubitus CT myelogram regardless of which of these techniques, you should feel reassured that you’re getting a high-quality examination if it’s being done in the hands of someone who’s experienced in that technique. That operator experience is probably more important than any of these other technical factors.
It has to work at the institution where you’re at. Equipment availability, referral patterns, experience patterns definitely matter. And again, I think operator experience is probably more important than technology. We also don’t know about recurrence after surgery. This is a sort of a systematic review – not a ton of papers that were reviewed here, and you can see the sample sizes for all of these papers are quite small. They found a recurrence rate of 6%. Now, this was in 2022, which seems like a long time ago now. I think most people probably believe that the recurrence rate is likely a little bit higher than that depending on what technique you use.
Then rebound high pressure is something that’s very important and really has very little literature at all when it comes to the treatment of CSF-venous fistulas. They found a 16% rate of rebound high pressure, which seems grossly an underestimation based on my experience. I would say it’s more like 50% or 75% of patients will have rebound high pressure.
So finally, I’m going to end up with the unknown unknowns, and by definition, these are things that we don’t know that we don’t know. So it’s very hard to actually talk about that. But I love this quote from Hamlet: “There are more things in Heaven and Earth than are dreamt of in your philosophy.” Even CSF-venous fistulas, when we first learned about them, were one of our unknown unknowns – we didn’t know that they existed until we happened to find that they did.
There’s this nice book by Thomas Kuhn who talks about the structure of scientific revolutions. In the book, he says that normal science doesn’t aim at novelty, but it aims at clearing up the status quo, and it tends to discover what it expects to discover. So, we’re essentially validating things that we already think we know, and we have a tendency to see what we expect even when it’s not there. It takes a long time for an anomaly to be seen for what it is, which is something contrary to the established order. So, it’s very hard for us to get to these revolutions because our brain is trained to look for things that we already know, and it’s hard to find things that we don’t know.
This is my last slide. There was a study done by Bruner and Postman, and they asked subjects to identify playing cards after a short exposure. Most of them – the playing cards were normal, but some were anomalous, like a red six of spades or a black four of hearts. The anomalous cards were almost always identified without any hesitation at all as being completely normal. So you would see the anomalous card, and the person looking at it would say, “That’s a normal card.” The black four of hearts, for instance, might be identified as a spade or a heart but without any hesitation. As people were increasingly exposed to these anomalous cards, they began to recognize them more frequently. But it was often difficult for them to identify what exactly was wrong, and some subjects still made incorrect identifications even at 40 times the average exposure were needed to identify normal cards for what they are. So, our brains don’t work to find these anomalies, and we really have to be actively searching for them.
And so, you know, I just want to end there by saying that our knowledge is incomplete on CSF-venous fistulas, and we know that. We know there are gaps in our knowledge that prevent us from making the correct diagnosis or instituting the correct treatment. But I’m really optimistic because we have more people involved in this now than ever before, and it’s much more likely, with all those people involved, that we’re going to find these anomalies. We’re going to find things that we weren’t expecting to find – like we found the CSF-venous fistulas – and it probably won’t be what we expect them to be. If you’re a patient, it won’t be what you expect it to be, because that’s the definition of anomalous. So, I think we need to keep an open mind to what we find. Whether you’re a physician or a patient, we can’t afford to be overly anchored in what we think we know – what we think our known knowns or known unknowns are. We have to keep an open mind if we want to move forward.
Thank you very much.