Dr. Andrew Callen at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Andrew Callen, assistant professor of neuroradiology and director of the CSF leak program at the University of Colorado Anschutz, presented this talk, “Iatrogenic Leaks: Imaging of Post Dural Puncture Headache,” 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
[00:00:12] Hi, everyone. I’m Andy Callen. It’s great to be here. Thank you very much for inviting me to speak. Thank you to Dr. Carroll for giving us the great context for this part of the talk, which is going to be the imaging of post-dural puncture headache.
[00:00:24] All day we’ve been hearing about how this is such a challenge, both in the brain and the spine. And so as a radiologist getting up here and talking about imaging and post-dural puncture headache, I could actually just end the talk now and say it’s always negative and walk away, but I’m not going to do that.
[00:00:39] I’m going to be focusing on the spine in particular, because I think that our concept of negative is very important. When we hear studies that say, you know, there’s this many negative brains with this many negative spines, we have to ask ourselves, were they truly negative? Were they read as negative and misinterpreted? Was there something subtle that was missed? Or what have you.
[00:00:57] I want to focus on a couple concepts here. Number one, the acute post-dural puncture headache, things that we may more commonly associate with a classic CSF leak, so to speak, epidural fluid, etc. And then the more challenging chronic CSF leak and post-dural puncture headache, where we may not see an epidural fluid collection, but we may see epidural edema, and then visit this concept that has been sort of brought up a couple times so far today of the bleb or the pseudomeningocele with a couple very illustrative cases.
[00:01:26] Let’s look at a true positive that I think we can all agree on. So here’s a 22-year-old woman who is two days postpartum after a labor epidural. Now she has a severe orthostatic headache. Not all these patients get neuroimaging. A lot of them just get a blood patch and go home. This one did, and she has diffuse, smooth pachymeningeal enhancement and sagging of the posterior fossil structures, pituitary venous engorgement. We’re all experts in this now, we can see it from across the room. There’s also imaging on her spine MRI. She has an epidural fluid collection. In this case, predominantly ventral, indicated by the arrows here, that posterior displacement of the dura that we’re getting used to seeing.
[00:01:59] But I want to draw your attention to something else, and that is what’s going on in the posterior epidural space, this lumbar epidural edema, this intercalation of fluid signal along the lumbar epidural fat in an undulating fashion back here. We’ll come back to that soon. Anybody who performs myelography is familiar with the iatrogenic puncture, the extradural contrast on your myelogram that we either don’t comment on at all, or just bury in the findings and we say, yeah, there’s some extradural contrast here, but I did that. That’s not the leak that we’re looking for.
[00:02:30] We see it here on screen left, the CT myelogram, the lateral arrows showing us the contrast sort of leaving the thecal sac there. We also see a little bit of disruption of the posterior dura and displacement of that epidural fat posteriorly. When we perform MR myelography, which has exquisite contrast sensitivity, we can see that extra thecal gadolinium as well.
[00:02:50] We have to be very careful to not misinterpret that as a spontaneous leak. And I think that intuitively we know that most of the patients we see, that we see iatrogenic contrast, extradural contrast on, aren’t all suffering from post-dural puncture headache. And that’s been borne out in the literature to some degree.
[00:03:05] So this very interesting study out of the Japanese group looked at patients who had lumbar puncture related extradural fluid. They took 53 patients who had an MRI obtained within two weeks of a lumbar puncture, and they saw that really, when you looked at those who had extradural fluid on their MRI compared with those who did not, there was really no difference in the presence of post-dural puncture headache among those two populations.
[00:03:29] So just extradural fluid doesn’t correlate with the presence of a post-dural puncture headache. This seems to align with our experience that we see when performing CT myelography. In my patients, I offer in a subset of them intrathecal gadolinium myelography. And some of them aren’t really keen on using that, which I think is totally reasonable.
[00:03:46] And so in those patients I’ve been just experimenting with just water. And rather than putting in GAD and then doing a T1 fat saturated sequence, just simply repeating a 3D T2 fat saturated sequence and sagittal STIR stations throughout the spine after water infusion, concurrent with their dynamic CT myelogram, to see if we can pick up slow leaks that way in a non-inferior manner.
[00:04:06] Here’s one patient who had pre saline augmented myelogram pre-procedure on screen left, and then post saline immediately after taken to the MR suite, here on screen right. Now, this patient did not suffer from a post-dural puncture headache, but nonetheless, we see these sort of classic hallmark features of that iatrogenic leak, the leak that I gave him, with a non cutting needle, Ian, don’t worry, but, um, it was, we see the disruption of the posterior dura there, and we see this edema along the posterior epidural fat.
[00:04:31] That edema along the posterior epidural fat has a very unfortunate name. It’s called the dinosaur tail sign. And I say it’s unfortunate because every time I say this to a patient, I feel myself cringing. But nonetheless, I think it’s important to a certain degree. And let’s talk about why. So in this paper, also published by that same group out of Japan, they looked at people with confirmed SIH as well as people with other signs of post-dural puncture leak on imaging. They saw that six out of the seven patients with SIH had this lumbar epidural edema, which I’ll refer to it as from from here on out, rather than the dinosaur tail sign. And then 19 with other signs of post-dural puncture leak also had this lumbar epidural edema.
[00:05:06] Now. Six out of the 23 that had that lumbar epidural edema had a post-dural puncture headache. So, not the majority of them, and none without the sign had a headache. Compare that to controls who had no prior puncture, no SIH, two out of the 35 had this finding. However, it was limited only to a single inner space. There was no people with this finding and more than one inner space in their study. So perhaps its absence is more valuable than its presence when we’re looking at it on imaging.
[00:05:33] Here’s a patient I took care of who one day, postpartum, had classic signs of perhaps dural violation from her epidural anesthesia. The yellow arrows are pointing to our epidural fluid collection. The red arrows are again pointing to our lumbar epidural edema. She had a lumbar blood patch with anesthesia in the hospital. Her headache was about 70 percent improved, but she came back to see me five months later, and she just had a persistent orthostatic headache.
[00:05:53] It was better, but it was still there, and interfering with her quality of life, preventing her from being able to take care of her baby in the way that she would like to. And this MRI was read as negative. And maybe it would be thrown into the pile of MRIs that say, yep, this is another case of negative MRI in the case of post-dural puncture headache, but I want to draw your attention now to that lumbar epidural fat.
[00:06:12] And I would argue that three consecutive inner spaces here show that lumbar epidural edema. So, to me, I would actually write this in my report as a radiologist, I wouldn’t say that I can localize where the leak is, but I would say that there’s some edema there which has been associated or described in the context of a post-dural puncture headache.
[00:06:29] And there’s our little stegosaurus at the top, just reminding us about the name.
[00:06:34] I want to draw you to another case which was very, very important to me and I learned a lot from. So this was a 42 year old woman who had been suffering from a disabling orthostatic headache for six years. Here’s her brain MRI, non contrast, and a post contrast axial T1 image on screen left, showing no signs of intracranial hypotension.
[00:06:52] But I also want to draw your attention to her 3D T2 fat saturated imaging on the right. Which shows a number of these very enlarged, irregular perineural cysts. I think that any of us, when we’re in this field, we see these sort of these diverticula, we get very excited. Oh, we must find a fistula or leak at one of these levels, where we can’t wait to perform our imaging.
[00:07:08] And she actually had both DSMs as well as dynamic CTMs, which showed nothing. So when we were looking at her and her imaging, we went back to the very first MRI she had, the MRI that she had before any of her myelography. And at the L3, L4 level along the posterior dural margin, there was this very, very small outpouching.
[00:07:25] And we asked ourselves, What is this? What could this be? And at the time, we didn’t really think that there was any sort of relationship with her clinical history, but talking to her further, she did have a labor epidural 18 years prior to presentation. And, briefly after that epidural, she had a brief self-limiting headache, which went away on its own. She did not receive a blood patch at the time, and then was headache free for 12 years prior to the current presentation of six years of persistent orthostatic headache.
[00:07:51] I went to patch that finding, and during my test injection with a little bit of contrast dye, I saw a finding similar to what Dr. Carroll showed before. Here’s my needle with a tip in the epidural space. Here’s some epidural contrast. I hit my pedal on my fluoro CT machine right as I was injecting, and saw this little bit of intravisation of contrast here into the subarachnoid space and layering along the dependent thecal sac.
[00:08:11] The response, the clinical response that she had from this epidural patch with fibrin was much greater than several large volume blood patches that she had had previously. And therefore she was taken to surgery for definitive repair. Intraoperatively, my surgeon, Dr. Lennerson, who provided these photos for me, saw some very interesting things.
[00:08:28] Number one, there was this membrane covering this finding, which I think has been described by some of our colleagues in the audience today. And once that membrane was sort of removed, he noticed leaking from this structure, but it was not from the dome, but rather from the base. And particularly as he moved it around and begin to reduce it, leaking was coming from the base of this pseudo meningocele, this bleb.
[00:08:47] And, you know, I wonder if that sort of gives us a hint at what’s possibly going on here, a needle comes in, violates the dura and arachnoid. It’s removed. There’s herniation of this arachnoid through that defect, which may act as sort of a ball valve mechanism, so to speak, and there’s leakage from the base, which is perhaps intermittent, not well captured, particularly when the patient’s in the supine position during our examinations. And when we try to patch these things, perhaps that ball valve mechanism is slightly getting in the way of our patch, which may account for the sort of intermittent or partial, incomplete nature of the clinical response to our patching. Nonetheless, with operative repair, with reduction of this bleb, with dural closure and patching intraoperatively, she did have durable symptom resolution.
[00:09:25] There’s a couple points of significance here, but one that I like to draw everyone’s attention to, which I think has been sort of hinted at several times so far, is that the classic conception of a post-dural puncture headache is something that occurs in the acute or subacute period.
[00:09:36] So, you know, we weren’t thinking at all that she could have possibly had a chronic post-dural puncture headache from years and years prior, but that was the case in this patient. And so I think it’s very important to remember that post-dural puncture headache can occur in the chronic period, even though it’s not classically conceptualized as such.
[00:09:52] Here’s a couple more examples of some blebs, so to speak. Here’s a 59-year-old gentleman with an intrathecal catheter who developed a new orthostatic headache. A couple of months after that catheter was placed, we see that outpouching of dura around where the catheter entry site is.
[00:10:06] Here’s a 48-year-old woman who has no prior instrumentation at all, who had a severe orthostatic headache with temporary relief to non-targeted patches. However, when we scrutinized her MRI here before any myelography was performed, we noticed this outpatching here in her thoracic spine. I’m not showing it, but there was a sharp osteophyte on her CT that we thought may have been contributing, and a fibrin patch at that level did lead to a more durable clinical response for her.
[00:10:28] So in conclusion, I think that we should think about the word negative. What does it mean? Does it mean that these patients actually have completely normal imaging? Or perhaps there’s subtle imaging findings, which we may not be classically associating with a leak that we should be paying more attention to.
[00:10:41] We should think about what kind of sequences we’re using when we image our patients. We should use thin slice, t2 fat saturated imaging because we want to avoid a repeat puncture if at all possible. I always tell my patients that the worst part of my job is that I have to put a hole in their dura in order to find a new leak.
[00:10:55] So if I can avoid that by just doing an MRI and trying patching based on this, that is the best possible outcome. I want everyone looking at MRIs of their patients to scrutinize the dural contours of their patient’s MRI to look at the lumbar epidural space. Just take note: do we see that sort of lumbar epidural edema? And also take a very detailed history, understand where the spine may have been instrumented in the past in order to scrutinize those areas of dura even more carefully.
[00:11:18] Thank you very much.