Dr. Ian Carroll at the 2023 Cedars-Sinai Intracranial Hypotension Conference
Dr. Ian Carroll, associate professor of anesthesiology, perioperative and pain medicine, and chief of the Stanford Headache Program, presented this talk on iatrogenic leaks titled “Iatrogenic Leaks:PDPH, needle type data; can we reduce the incidence” 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] Thank you. It’s great to be here. And I appreciate the invitation of the Spinal CSF Leak Foundation and Dr. Schievink in having us here. The research I’m going to talk about will be presented tomorrow when people are talking about ongoing research, but I wanted to cover what we as a field have learned about, in terms of iatrogenic leaks today.
[00:00:33] So, for patients who are watching who want to read more about the stuff that I talk about in this talk, I share a Google Drive that has articles about it, including a subfolder on chronic post-puncture leaks.
[00:00:48] So first I want to tell you about a patient who came to see me after being seen at Mayo. In 2018, she had a venous sinus thrombosis and a venous infarct and had hemorrhagic conversion requiring a craniotomy to evacuate her hematoma, and at the time had a lumbar drain placed. She was in the neuro ICU for a couple of weeks, completely recovered. A year later, having recovered from her venous sinus thrombosis, she has orthostatic headache still that has persisted from the time that she left the hospital, and has a radionuclide cisternogram to look for a leak. It was read as negative for MRI of the brain and full spine, or normal, per the report. And she underwent, I think an appropriate and aggressive evaluation there looking for a CSF-venous fistula, and they did bilateral digital subtraction myelograms at Mayo showing no evidence of a CSF leak.
[00:01:45] She had several multi level thoracic epidural blood patches at Mayo, no relief. And I was suspicious of the lumbar drain site—maybe it was something that simple; decided to give her some lumbar patches. This is one of the images from when I patched her. So the tip of the needles here entering the epidural space, the dorsal dura is far forward. I’ve gotten nowhere near the dorsal dura. And on the first image, you see contrast has not spread up to the level above, but I give a nice big bolus of contrast, and in the next image, contrast has spread in the epidural space up here.
[00:02:29] And what you see where the blue arrow is something that we’re calling intravasation. This is contrast that has gone from the epidural space into the intrathecal space, the opposite of what we see with a positive myelogram. And so, Andy [Callen] and I also wrote up a paper that was recently published in headache showing a CT equivalent of this.
[00:02:51] This is something that I’m hoping other people in the audience will start to look for as they evaluate patients with potential post-puncture leaks. If you put a large amount of contrast, can you see it get into the thecal sac? And that’s it, a little blown up. You can see that it’s going in at a level other than where my needle is.
[00:03:13] At surgery, this is what was found. That’s a nice epithelialized hole where the lumbar drain had been. This is what it looks like after being surgically closed. She’s now two years out. This was done before COVID, the surgery. She remains symptom free. We went back to look at her MRI. This is what her MRI looks like.
[00:03:35] This was read as negative. You can see that, really, I don’t think anyone would call anything on the sagittal. I wonder how many people would call something on the axial. This is something that actually Dr. Callen has taught me to look for, this kind of little line in the epidural space that seems to go without some interruption to the intrathecal space.
[00:03:58] You only see it on that one slice of the MRI, you don’t see it anywhere else, but that’s where the hole was. So just how unusual is this? One of my colleagues, Pam Flood, while at Columbia, did research looking at patients who had an accidental dural puncture during the placement of a labor epidural. And she looked at 40 cases and 40 controls, but did something that most anesthesiologists never do: she contacted them a year to two years later to ask them about new chronic headache.
[00:04:30] And this is what she found is, one to two years after an uncomplicated epidural without an accidental dural puncture, the rate of having new chronic headaches is about 5%. But if you ask the patients who had an accidental dural puncture that was recognized and thought to have been treated appropriately at a major university, 30 percent of them are having new chronic headaches.
[00:04:54] If they had a blood patch as part of their treatment, they’re still at four times the baseline risk. It doesn’t reduce their risk 90%. The baseline control rate of having headaches was 5%. If they had a known puncture and were patched, their risk is still 20%. And conservative management did terribly at 40%.
[00:05:17] For a long time, this was a paper all by itself, but over the last several years, what you see is there’s now at least six papers out there showing essentially the same thing. What you see in blue is the rate with unintended dural puncture of chronic headache after unintended dural puncture. The red is the control rate for each study.
[00:05:37] This is clearly real. One of the things that’s troubling to me about this is the implications for the wards of obstetric anesthesiologists that are putting in epidurals for childbirth, which is a major part of every hospital in the country. This has such implications for that, you would think one of these papers would have been published in the journal anesthesiology, which is the main journal of anesthesiology.
[00:06:02] None of them have, and I think it’s because the implications, which are really quite scary.
[00:06:10] These are the papers that looked at imaging findings among such patients, including one by Dr. Schievink. And these were the rates of positive brain imaging among those patients. So on the right is my fantasy where we have a test that shows 100 percent that they’re leaking. But that’s also there to show the scale.
[00:06:32] The percent of actual positive imaging is tiny in these patients. And I put out to you that if I were to ask you—we all know that we have a problem in the field that we don’t have a gold standard to compare our imaging to. Like, how many patients with leaks are missed? What’s the gold standard? To me, the gold standard is the leak you caused.
[00:06:53] A new post-dural headache that was caused by a needle you put in that remains a post-dural headache. That’s as good as a gold standard as you’re ever going to get. This data suggests that our sensitivity is low. In Schievink’s data, there were 27 patients. Nine patients had positive spine findings on digital subtraction myelography.
[00:07:15] Nine out of the 27. Only three had positive brain MRIs. Three out of the 27. If you count the whole 27 as the true positives, that puts brain MRI sensitivity at 11%. If you count only the patients who had confirming DSMs showing the spinal leak, then your sensitivity’s 33%, three of the nine. We can’t walk around pretending that because the MRIs of the brains are negative, we should dismiss the likelihood that these patients have in fact a true leak.
[00:07:48] What about deliberate dural punctures? There’s much less information about this. Jeskins in 2001 published a study in which she compared patients who had undergone spinal anesthesia using the best kind of needles we have, the 25-gauge Whitacres, pencil point, small gauge, used for spinal anesthesia. She was able to collect 56 patients who had dural punctures followed by an acute early CSF headache, an acute post-puncture headache, and 61 who had dural punctures and did not have any acute post-dural puncture headache. When she followed them up after the years two to eight, and then assessed what percentage of them had new chronic headaches. If you had an early post-dural puncture headache, you had a much higher risk of going on to having chronic problems compared to people who’d have the same kind of dural puncture, but no acute post-dural puncture headache. The early acute post-dural puncture headache is a marker for long term risk.
[00:08:54] That was highly statistically significant. So what can we do to reduce the risk of early post-dural puncture headache. I want to show you that there’s two different kinds of needles that deliberate dural punctures are done with. One are called pencil point needles. Whitacre is one kind of pencil point needle.
[00:09:12] That’s on the right. There’s other kinds of pencil point needles. I know at Duke, they use Gertie Marx. There’s another kind called Sprotte. And then on the left, you have what are called conventional or cutting point needles, like the Quinke needle. That is the default needle that’s in lumbar puncture trays all over this country.
[00:09:30] There’s data suggesting that 98%—this is from a neurology paper—98% of neurologists exclusively use cutting needles as their default needle in the UK. There’s data that 80% of neurologists are using cutting needles. And the issue is, does the needle type in fact really make a difference in terms of causing problems?
[00:09:51] And in this case, the issue of best practice when deliberately puncturing the dura has been settled. Conventional cutting needles cause excess post-dural puncture headache and should be abandoned. We should not be using those needles anymore. This is a meta analysis in The Lancet from 2018. It included findings from 110 prospective randomized trials that included 31,000 patients in randomized prospective trials comparing conventional needles to those pencil point needles.
[00:10:28] In terms of the frequency of post-dural puncture headache—and this is out of 31,000 patients. This is as good as it gets—the cutting needle rate of acute post-dural puncture headache was 11%. With the pencil point needles, it’s 4.2%. That suggests that the lion’s share, the majority of early post-dural puncture headaches are attributable to a choice made by a physician, i.e. a preventable consequence of medical practice, as opposed to just something that happens.
[00:11:06] That’s highly statistically significant, and it means that 62% of post-dural puncture headache, early post-dural puncture headache, are preventable using a needle that costs 15 dollars more. It’s not a 1 percent complication.
[00:11:23] We’re talking 11% Versus4. 2%. There is no increase in the rate of procedure failure when you use these pencil point needles and, the cutting needles are not more successful. This is their forest plot. And what you see here is first, on the forest plot, the stuff on the left favors atraumatic needle. The stuff on the right favors conventional needles.
[00:11:48] What I want to point out is these things where it doesn’t appear to make a difference is success on a first attempt or overall failure rate. There is no evidence, despite thousands and thousands of patients, that you are less successful doing a lumbar puncture when you use a pencil point needle. In contrast, not only is the frequency of post-dural puncture headaches worse, as I just told you, but markers of severity of the post-dural puncture headache are clearly worse.
[00:12:17] The need for intravenous fluid or PCA or the need for a blood patch are much worse when you use a conventional needle. Conventional needles make the frequency of early post-dural puncture headache more frequent, and the severity worse as well. So PDPH is a preventable complication of choosing to use a conventional cutting needle, and the limited data suggests that acute post-dural puncture headache identifies the risk of long term complications.
[00:12:45] Since 2006, the American Academy of Neurology has recommended that atraumatic pencil point needles be used as the standard diagnostic lumbar puncture, and this has been adopted also by, you can see, a clinical practice guideline published in the British Medical Journal, and then the European Journal of Neurology also saying, change practice now.
[00:13:08] So for those of you who are working in this field, at the very minimum, when you find yourself in a hole, you gotta stop digging. The people in this room should not be using conventional needles anymore. We know better, and we’re dealing with patients with dura that is at risk.
[00:13:26] This is the acute post-dural puncture rate by needle type. The Tuohy needle on the left is what happens when you accidentally puncture the dura when you’re putting in an epidural catheter, but the 11 percent with the Quinke and the 4.2 with the Whitacre. So, dural punctures can cause two different headache syndromes, the acute post-dural puncture syndrome, and then this chronic syndrome that appears to be real in multiple studies that we have disagreement about.
[00:13:55] So, the post-dural puncture headache is broadly recognized. And the chronic post-puncture syndrome is broadly unknown. In the post-dural puncture headache, early on imaging is often positive. And then the chronic post-puncture syndrome, imaging is rarely positive. There is broad agreement that ongoing CSF leak is the cause of the acute post-dural puncture syndrome, but we’re starting to see the emergence of papers, saying, “Hey, if the imaging is always negative in these patients, maybe they don’t have a leak,” even though it started the day of a puncture, it’s positional, which we associate with orthostatic, with CSF loss, and remains positional in at least a third. We know that changing the needle type is a proven method for reducing post-dural puncture headache acutely, but right now, changing the needle type is not yet proven or tied with any clinical research to the long term sequelae.
[00:14:55] If we, in fact, find that the needle type is tied to the risk of the long term syndrome, that would be an important piece of evidence that they really are leaking. And again, that feeds into this broader narrative of what are the deficits in our imaging sensitivity. If we had good evidence that these long term headaches really are leaks, that tells us how far we have to go in improving our imaging sensitivity. Thank you very much guys.