Hello. My name is Patrick Brown. I'm one of the neurointerventionalists at Wake Forest Baptist Medical Center. I'm coming today to talk about our descriptive study looking at post-mechanical thrombectomy subarachnoid hemorrhage, looking at procedural characteristics and outcomes. This is our team. We have no relevant disclosures. Since 2015, as we all know, mechanical thrombectomy has developed into mainstay therapy for intracranial emergent large vessel occlusion after publication of several positive studies at that time. Then after publication of DAWN and DEFUSE 3 in January 2018, indications for mechanical thrombectomy expanded to include delayed presentations. And then also, we're doing more thrombectomy as operator comfort has increased. Expanded indications, including distal occlusions, have also been considered. So it's in that background that we wanted to look at procedural safety. Now, most of the time we think about procedural safety in terms of symptomatic intracranial hemorrhage, which is defined as generally as an increase in NIH stroke scale of four or more points compared to baseline in the setting of development of new intracranial hemorrhage. Discussions of safety are often limited to parenchymal hematoma and hemorrhagic transformation. And these have suggested relatively stable rates in the era of expanded thrombectomy indications. The discussion of subarachnoid hemorrhage generally has been limited to small series, and they often all predate the 2015 studies for the most part. And they suggest little to no substantial clinical effect in subarachnoid hemorrhage in the post-thrombectomy setting. So the goals of our study were to corroborate these findings in our current era to see if they hold true, and then report any trends in procedural characteristics that might suggest future development of subarachnoid hemorrhage, or be correlated with the future development of subarachnoid hemorrhage. So we did a single center retrospective review of our prospectively acquired database of cases at Wake Forest from January 2015 to June 2019. We did a clinical and imaging review to look at procedural characteristics associated with development of subarachnoid hemorrhage. And then we also looked at outcomes in the form of 90-day Rankin scores. We also did a logistic regression analysis of demographics and comorbidity data to see if they also had any predictive value. So quantification of subarachnoid hemorrhage turned out to be pretty difficult. As opposed to parenchymal hematoma, which is usually very localized and measurable, subarachnoid hemorrhage is quite difficult to measure. So as a result, we came up with a more qualitative look at subarachnoid hemorrhage. And we categorized it in terms of trace, which is barely perceptible and really not measurable, versus focal, which is a measurable amount of subarachnoid hemorrhage in one or a couple sulci, and then diffuse subarachnoid hemorrhage, which is measurable subarachnoid hemorrhage in multiple sulci. So what did we find? Well, during our evaluation period, we had 212 thrombectomy cases. 26 of those cases subsequently developed subarachnoid hemorrhage, with an overall subarachnoid hemorrhage incidence rate of 12.2%, which is on par with the previously published data. Then we compared post-thrombectomy subarachnoid hemorrhage cases with non-subarachnoid hemorrhage cases, looking at thrombectomy technique, particularly stent retriever versus aspiration versus combined techniques. We looked at procedure time, number of device passes, and long-term clinical outcome. So overall, what we found was that stent retriever use was strongly correlated with subsequent development of subarachnoid hemorrhage. 92% of our subarachnoid hemorrhage cases used a stent retriever versus 67% to 68% of cases that subsequently did not develop a subarachnoid hemorrhage. We also found that longer procedure times, more device passes were also associated with subsequent development of subarachnoid hemorrhage. And subarachnoid hemorrhage was associated with longer length of stay post-thrombectomy. These are all pretty strong correlations. What we didn't see was a statistically significant trend toward a poorer clinical outcome in the subarachnoid hemorrhage population. We saw 21% of the post-thrombectomy subarachnoid, or 21 of the 26 post-thrombectomy subarachnoid hemorrhage patients had a poor clinical outcome as defined by a Rankin score of 3 or more at 90 days, resulting in almost 81%. But this did not achieve statistical significance. When we look at volume and distribution using our qualitative and visual analysis of the image data, we saw that there was no significant correlation with any of the procedural characteristics or clinical outcome when looking at trace versus focal versus diffuse subarachnoid hemorrhage. When we look at subarachnoid hemorrhage only versus subarachnoid hemorrhage with a parenchymal hematoma, again, we see a strong correlation with use of stent retrievers, longer procedure time, more device passes, and a longer length of stay in the subarachnoid hemorrhage and or parenchymal hematoma population. But again, we see a trend toward poor clinical outcome or more poor clinical outcomes, but we don't see a statistically significant trend. And then finally, with our logistic regression analysis, we looked at demographics, including age, sex, race, and then comorbidity data, including hypertension, diabetes, coronary artery disease, AFib, prior stroke, and smoking. And the only thing that we saw that was significantly correlated with subarachnoid hemorrhage was age. And so, the subarachnoid hemorrhage population tended to have older age patients as opposed to the non-subarachnoid hemorrhage population. So, the average age for a subarachnoid hemorrhage patient was 73. With a non-subarachnoid hemorrhage patient, it was 66.5, and that was statistically significant. This bottom graph shows that we had a total of 73 patients over the age of 75 in our study, with 14 of those developing subarachnoid hemorrhage, and then 25 patients over the age of 85, with eight of them developing subarachnoid hemorrhage. So, 22 of the 26 subarachnoid hemorrhage patients that we saw were over the age of 75. So, we found that post-thrombectomy subarachnoid is more associated with use of a stent retriever, longer procedure times, more device passes, and a longer length of stay, as well as older age. Now, we saw that long-term clinical outcomes trend toward a worsened modified Rankin score, but do not reach statistical significance. And that was even when we included parenchymal hematomas into the mix. And finally, we saw no significant correlation with our qualitative subarachnoid hemorrhage volume and distribution analysis. So, future directions. We have the largest single-center series that we could find. But yet, I think that because of the rarity of subarachnoid hemorrhage, we do need to combine forces with multi-center data. You know, we trended towards statistical significance, but we didn't quite achieve it. Maybe with a larger patient cohort, we might get there. We saw poor correlation with a qualitative subarachnoid hemorrhage volume and distribution. So, we're working on a volumetric analysis of imaging data to see if there's a quantitative correlation. And then finally, future direction, we could look at what is the mechanism for development of subarachnoid hemorrhage. Thank you for your attention. These are our references. Thanks so much.

subarachnoid hemorrhage

procedural characteristics

SAH incidence rate

stent retriever use

age factor