I'd like to thank Dr. Bola, Rob Bola, for coming to the stage and talking about pediatric blunt cerebrovascular injuries, imaging, and management update. Good afternoon. Thanks for the opportunity to talk about this and present today. I have no disclosures. So this is just a brief outline of my talk, but we're going to just sort of go over some of the adult BCVI, or blunt cerebrovascular injury, literature. We're going to talk a little bit about radiation exposure in children, and then I'm going to try to convince you that selective screening in children is the most appropriate thing for this fairly rare, but potentially devastating, type of injury. And then we're going to talk about treatment, and then we're going to talk about future directions for clinical research on this topic. So in adults, traumatic blunt cerebrovascular injury complicates about 1 to 4% of blunt trauma, depending on the series you look at. It's an established risk factor for delayed stroke, and post-traumatic stroke has an extremely high morbidity and mortality. And there's pretty clear evidence in the adult trauma population that early recognition and treatment reduces that risk. In terms of how to screen trauma patients with blunt injury, CTA is really the screening method of choice. This hasn't always been the case, but now with 64-slice and better scanners, it has very high sensitivity. Digital subtraction angiography is invasive and challenging to perform in the acute trauma setting. Currently, MRA is time-consuming. It's poorly studied in terms of its sensitivity for this type of injury for screening, and ultrasound is operator-dependent. So CTA is really the method of choice for screening. This is the modified Denver criteria, looking at signs and symptoms of blunt cerebrovascular injury, as well as risk factors. So this was developed by somebody named Walter Biffle in Denver, and it's pretty well used in the adult population. So if you have four risk factors, there's over a 90% risk of a blunt cerebrovascular injury. If you have no risk factors but blunt trauma, there's a 20% risk in the initial publication. So the risk factors are a high-energy transfer mechanism, a Laforte 2 or 3 fracture, certain patterns of cervical spine fracture, including fractures extending through the transverse foramina at C1 to C3, basilar skull fractures, especially those with carotid canal involvement and petrous bone fractures, and then diffuse axonal injury with a low GCS score. The Eastern Association for the Surgery of Trauma developed screening guidelines largely based on the Denver criteria, including neurologic abnormality not explained by a diagnosed injury, a blunt trauma with epistaxis, a low GCS score, a petrous bone fracture, DAI, and similar patterns of facial, skull base, and cervical spine fractures. This is a quite useful grading scale also developed by Walter Biffle, which is 1 through 5 in terms of blunt cerebrovascular injury grading. So grade 1 injury is an irregularity of the vessel wall or a dissection with an intramural hematoma with less than 25% luminal stenosis. Grade 2 is a visualized raised flap or dissection with greater than 25% luminal narrowing. Grade 3 is a pseudoaneurysm. Grade 4 is a vessel occlusion. And grade 5 is a vessel transection. And this is very useful in terms of trying to determine what the appropriate treatment is. So, you know, in children we have other constraints. Diagnostic radiation is more harmful. Children are more sensitive to radiation. And it's estimated that almost 5,000 new cancers are attributable to diagnostic radiation exposure. A high-dose head CT, like a CTA in a child, triples the risk of brain cancer. And it's estimated about a third of pediatric CT scans are unnecessary. And so there's concepts that you're all familiar with introduced in the pediatric literature, including ALARA or as low as reasonably achievable, in order to try to minimize the rate diagnostic radiation exposure in children. This just looks at trends in CT in the United States. And the top part of that line graph is head CT. So by far the most common type of CT performed is a head CT. Greater than 4 million CTs are performed in children annually in the United States. This is an editorial that was in the New York Times in 2014, which is right around the time I first gave any talk on this topic at the ASPN in Costa Rica. And this was published by two radiologists just saying that we're overdoing diagnostic imaging in the United States. So this is in the lay press. And that was several years ago, but it's really not going away. The FDA produced new guidelines for diagnostic imaging in children in January of this year, 2018, just identifying pediatric patients as more radiosensitive than adults, with a higher cancer risk per unit dose of ionizing radiation. And emphasizing the need for diagnostic radiation to use pediatric-specific equipment and to minimize their exposure. So this is a special constraint that means in pediatric blood supravascular injury we would really have to take a different tact and perform selective screening rather than broad screening. And until recently the pediatric literature was pretty sparse. It's thought that there's a lower incidence of blood supravascular injury in children. So 1 to 4% in the adult literature, maybe about 0.4 to 1% in children. But that may be because we're not screening them. And most studies suggest that we're really not, most children's hospitals and large children, large pediatric trauma centers don't do a good job of screening. So it may be that we're just not looking. And the East Society suggests that we should apply their adult criteria to children, which is really unselective in my opinion. This is just at Primary Children's Hospital where I practice looking at the rates of CTA over a fairly constant traumatic brain injury and head CT for trauma data, but increasing use of CTA over time. And so I'd like to make the case that selective screening is appropriate, as I said, that radiographic patterns of injury are different in children, that in the trauma setting all of these patients have already had a non-contrasted head CT. It's expensive. And then it's a binary decision in acute trauma. So, you know, either you get a CTA in the trauma bay or you don't. And it would be nice to have a prediction rule to say this child's at high risk versus low risk and we should proceed with a CTA. They've successfully decreased head CT use. In other cases, the PCAR network, which I'm sure you're all familiar with, has decreased the use of head CT in patients at very low risk of clinically important head injuries after trauma. This is a cost-effectiveness analysis that came out of the neuroradiology group at Yale, which basically demonstrates that as injuries become less common, that selective screening makes more sense just from a cost-effectiveness standpoint. So this led us to develop the Utah score, which we initially published in 2015, which looked at variables to predict the risk of blunt cerebrovascular injury in children. So a GCS less than 8 or 8 or less, a focal neurologic deficit, a carotid canal fracture, a petrous temporal bone fracture, or a stroke on CT. And if you had a score of 2 or less, you had about an 8% risk of injury. And if you had a score of 3 or higher, you had almost a 40% risk. So we dichotomized our score that way. And recall that in the Denver criteria in the initial publication, zero risk factors was associated with a 20% risk. So we thought this was pretty good, but it needed to be validated. So we did a multi-center validation, including Vanderbilt, St. Louis Children's, Texas Children's, and then our center was the training cohort. And while we validated it, so a score of less than or equal to 2 was a 2.7% risk of injury, a score of 3 or greater was an 18% risk. So we felt like that was a reasonable dichotomization. But if you look at the left-hand column, 9 of 22 injuries were misclassified as low risk. So I think one of the reasons that this looks good is that it's a fairly low risk to begin with. So this is certainly something we can and should improve upon. We also applied the Denver criteria to our total cohort, so both the training and validation groups. And the Denver criteria did not predict blunt cerebrovascular injury in children. I think, again, emphasizing the point that patterns of injury are different, that the risk factors in children are different. So since then, a few other centers have looked at their population. So CHOP looked at their population, and they had a .1% risk of BCVI, and they identified 11 in over 11,000 patients with blunt trauma. But they only screened patients, only 6% of patients who met the Denver criteria. They concluded that the Denver criteria made sense and predicted BCVI in children. But again, very, very low screening rate. The other group that did a large evaluation was Seattle. So the Harborview group, they identified 96 blunt cerebrovascular injuries in over 7,000 patients, so 1.3%. And they did a head-to-head comparison of the East trauma guidelines, the Denver criteria, and the Utah score. And they demonstrated that there was a 17% failure rate with the Utah score in their population, and one of those patients went on to develop a stroke. So it did better than the East guidelines, but the Denver criteria, which is more broad and less selective, did not miss any injuries. They had a much higher median age than our group, either the validation or the training cohort. So I think this is probably one of the reasons that our score was less appropriate in their setting. So the mean age was 16. These were also very sick patients, so the median admitting GCS score was 3. So recently, earlier this year, the Hermann Hospital, University of Texas group produced the McGovern score, which basically added to our classification score, the Utah score, mechanism of injury. And they identified a high-risk mechanism of injury as a motor vehicle collision or a pedestrian struck by a motor vehicle. And this significantly improved the misclassification rate, so from 40% down to 19%. And they, again, stated the high-risk score of three or greater. And this just demonstrates the ROC curve for the blue is the McGovern score and the red is Utah. So I think that's a good improvement upon our score and an important contribution to the literature for screening. So I'm just going to pivot really quickly here to treatment. We already presented this slide, but we grade these from one to five. So treatment itself in grade one to two injuries is somewhat controversial. So some of these are indeterminate. And so this was the Parkland group looked at their treatment patterns for grade one to two injuries. So this is in adults. So 31% of patients weren't treated, either because they felt it was not indicated or contraindicated. And in general, most injuries resolved. This is 117 grade one to two blunt cerebrovascular injuries in 100 patients. The stroke rate was low, and they did not identify a clear correlation between treatment and outcome. So I think in low-grade injuries, whether to treat remains somewhat controversial, but there's very sparse literature, and it's really only in adults for that. Grade four injuries are a different story. So this is another study looking at grade four injuries. And essentially, they're very lethal. So this study looked at 82 grade four injuries, 13 carotid and 69 vertebral artery injuries. There was a very high stroke rate, so 70% of the carotids went on to stroke. So this is vessel occlusion that then recanalizes and embolizes. And so 67% of the strokes occurred less than 10 hours after admission. So, you know, just another example that these higher-grade injuries can be quite lethal. There were many contraindications to medical treatment in their study. So these patients generally were sick. Our, the initial data set that we used to develop the Utah's core, we looked at treatment, and this was Rob Naftal in the group from Vanderbilt. And so in children in that cohort, for grade, low-grade injuries, many were not treated. So you can see for grade one, 19 were not treated. Eight got antiplatelet agents, aspirin. Three got anticoagulation. And then higher-grade treatments, higher-grade injuries were generally uniformly treated, but there were very few. There were no adverse events in this pediatric treatment study. And there were two delayed strokes, and one of those patients died. So general reasons for not treatment were it was contraindicated in 38, or the physicians felt it was not indicated, or the patient proceeded to death in 58% of the patients who were not treated. So again, I think treatment is somewhat controversial for lower-grade injuries, but extremely important for higher-grade injuries. In terms of how to treat, the CATUS trial was published in 2015 in Lancet Neurology, which is a randomized trial for essentially cervical artery dissection, looking at anticoagulation versus antithrombotic therapy with aspirin, and there was no difference. So I think if you're making a decision to treat medically, there's no difference that's known between aspirin and anticoagulation. So I think many of us are more comfortable with aspirin, and I think that that's fine. In terms of endovascular treatment, the Memphis Group, who initially, prior to the Denver criteria, published initial risk factors for blunt cerebrovascular injury, looked at 365 patients with digital subtraction angiography, confirmed BCVI, and they separated it into two groups. Their cohort historically, 2011 to 2012, and then 2013 to 2016, and they had a significant decrease in the rate of endovascular stenting, from 34% early to about 9% late. And they concluded that most patients are adequately managed medically. They advocated reserving endovascular treatment for enlarging pseudoaneurysms and what they called dissections with significant narrowing. So I think grade three. So I think future directions are to merge retrospective data sets. So we've talked with Manisha, who led the group at University of Texas, and I think we're going to try to get all of our data sets together to improve upon the Utah and the McGovern score to make a better score for selective screening for children. And that will improve clinical prediction and make screening even better. I think prospectively evaluating that score will be very important. Right now, all the literature is only on patients who were screened. So we don't know patients, you know, a child who comes into the emergency room with a skull base fracture who didn't get screened and what their risk is. And we don't want to increase CT use by putting out a score that might do that and do more CTAs in children because that wasn't the goal. So I think we need to prospectively evaluate these scores. So just in conclusion, screening, CTA is really the method of choice. Risk factors in children are unique, and selective screening is appropriate. Early identification, especially of high-grade lesions is critical, but we still don't have a great clinical prediction rule for who to screen, but we're getting there. In terms of treatment, it clearly prevents stroke, but it's more controversial in low-grade injury where there's a low stroke rate. We would recommend considering treatment of low-grade BCVIs if it is generally fairly low risk to put somebody on aspirin. And again, in that pediatric study of treatment of our cohort, there were no adverse events. Aspirin and anticoagulation are equivalent according to the CATUS trial. Endovascular treatment is controversial, but it's rarely required. So potentially in high-grade injuries where there's a contradiction to medical management or in a high-grade lesion that's progressive despite medical management. So that's it. Thank you very much. Thank you.

pediatric blunt cerebrovascular injuries

selective screening

computed tomographic angiography (CTA)

Utah score

treatment