Hi, everyone. I'm Dr. Deepa Malliandy, and welcome to the hemorrhagic stroke lecture in this series. I have no disclosures for this talk. So the goal for this talk is to provide a high-yield review of hemorrhagic stroke and to review key principles and updates in management. The objectives are for you to be able to determine early blood pressure control targets, identify imaging patterns associated with increased risk of hematoma expansion, weigh the risks and benefits of antithrombotic reversal and hemostatic treatments, and to identify patients for urgent surgical treatment. We're going to use the following case to review all four objectives. So we'll be talking about a 67-year-old man with a history of CAD and prior acute ischemic stroke on daily aspirin and clopidogrel who was taken to the ED with acute onset of being poorly responsive with a GCS of 9 and noted to have right hemiplegia of 60 minutes duration. His vitals are BP's 230 over 115, and he's got a heart rate of 80 beats per minute. He's intubated for airway protection, and his stat CT had is shown below. So the questions we're going to be looking at, again, is what should his initial blood pressure goal be? What imaging features, if any, are associated with hematoma expansion here? Is there a role for antithrombotic reversal with platelet transfusion, DDAVP, or both, or other hemostatic treatments? And is there a need for emergent surgical intervention? But first, we need to go over some basic terminology and review the etiologies of hemorrhagic stroke as well. Often in our daily clinical lives, terms hemorrhagic stroke, ICH, and IPH are used interchangeably. But for the purpose of clarity, we'll quickly review how they differ. Hemorrhagic stroke or infarction is any vascular cause for hemorrhage within the brain parenchyma, ventricles, or subarachnoid space. Intracerebral hemorrhage, often abbreviated ICH, includes hemorrhagic stroke as well as hemorrhage from tumors, trauma, infection, and hemorrhagic transformation of ischemic stroke, et cetera. Intracranial hemorrhage, also often abbreviated ICH, broadly includes bleeding of any cause within any compartment. So intraparenchymal, ventricular, epidural, subdural, and subarachnoid bleeds. And lastly, intraparenchymal hemorrhage, abbreviated as IPH, refers specifically to bleeding within the brain parenchyma. For this review, we're going to focus on the first bullet. If I use one of the other terms, it's because the recommendations or guidelines or studies did not specifically refer to hemorrhagic stroke only, or the authors used that term specifically. Within hemorrhagic stroke, there are primary and secondary etiologies. Primary includes both hypertension and cerebral amyloid angiopathy. Secondary etiologies include coagulopathy, any of the vascular malformations, aneurysms that are distal enough to cause parenchymal hemorrhage instead of subarachnoid predominant hemorrhages, cerebrovenous sinus thrombosis, and vasculopathies, including the vasculitides. Now we can move on to management. So just to review the four main categories of management per our objectives. So we obviously always start with the ABCs, and that includes blood pressure management. Then we want to interpret the images to get a sense of the severity and help assess the risk of hematoma expansion, in addition to other acute neurological complications that we may need to anticipate, determine the need to reverse anticoagulation, antiplatelet therapy, or treat coagulopathies, and what, if any, surgical management is needed. So there are two main RCTs that inform our decisions on blood pressure control targets in hemorrhagic stroke, and those are InterAct2 and Attach2. So InterAct2 was a nearly 3,000-subject study that enrolled those with a systolic blood pressure of 150 to 220 and randomized them to seven days of either a systolic blood pressure less than 140 or less than 180. The medications used were left to the discretion of the doctor, so whether that was with oral agents as needed, IV injectables, or continuous infusions, that was really up to the discretion of the treating team. They did see better 90-day modified Rankin scores, and many of the quality-of-life factors were improved, and they did not see an increase in adverse events. Something to note about the patients enrolled in this trial was that only 50% of them had a systolic blood pressure above 180, and most of them had a bleed over 15 millimeters. In the Attach2 trial, there was 1,000 patients, and they only enrolled patients with greater than a blood pressure of 180 presenting, and these patients were randomized then to a systolic blood pressure target of 110 to 139 versus 140 to 179, but this was only maintained for 24 hours. This was a trial around the time that nicartapine became available, so the trial required everyone to use nicartapine to achieve their blood pressure targets. They did not find a difference in functional outcomes specifically. They did not see any improvement in functional outcome as was seen in Interact2, and they did see an increase in renal adverse events. In their patient population, obviously 100% of them had a blood pressure over 180 because that was an enrollment criteria. 10% of them only had hemorrhages above 30 milliliters, and because the trial only required that strict blood pressure parameter for 24 hours, there was a significant amount of blood pressure fluctuation and rebound hypertension after that 24 hours was over. There was a pre-planned pooled analysis of these two trials, so when they looked at all the data combined, what they noticed was that there's a 10% better functional outcome for every 10 millimeters mercury smooth blood pressure drop, so not sudden drops, and actually in terms of sudden drops in blood pressure fluctuations, what they found was dropping the blood pressure by 60 millimeters of mercury or more within 60 minutes was actually associated with lower odds ratios of good outcomes. A smooth blood pressure lowering to a safe target seems to be the best approach, and we'll talk about how we apply this data to clinical practice today. So this table is adapted from the recent AHA-ASA 2022 guidelines on the management of acute intracerebral hemorrhage, and their blood pressure recommendations are as follows. So continuous smooth and sustained acute blood pressure control that avoids large variability in systolic blood pressure can be beneficial for improving functional outcomes, and we talked about that. Also, initial treatment within two hours of ictus and reaching target within one hour can be beneficial to the risk of hematoma expansion and improve functional outcomes. And then in mild to moderate severity cases presenting with a systolic blood pressure of 150 to 220, acute lowering of blood pressure to 140 is safe and may be reasonable for improving functional outcomes. Just note that the goal range here is 130 to 150, so they do note that even for moderate severity bleeds with a presenting blood pressure of over 150, acute lowering to less than 130 may be harmful. So it's really that lower blood pressure target that is important to avoid, anything less than 130. And then for severe ICHs or large ones or those that require surgery, the safety and efficacy of intensive blood pressure lowering is really not well established. And oftentimes in these cases, we'll have an intracerebral monitor or an intracranial monitor or some type of intracranial pressure monitor. And so what would really be targeted here ideally in these cases would be cerebral perfusion pressure at either a target of 50 to 60 or 60 to 70, depending on whether you level at your institution, the arterial line at the tragus or the phlebostatic axis. So if we apply that information to our patient case, we see that our patients presenting blood pressure was outside of what was studied in the trials for the most part. And we see that this is likely going to be a very severe hemorrhage for this person given, you know, particularly the load of intraventricular hemorrhage. So how do we target a blood pressure goal for them? In this case, we could argue that with the acute hydrocephalus, there would be harm with rapid reduction in blood pressure that's too severe. And there are going to be cerebral perfusion concerns with acute hydrocephalus. So what we'll want to do is smoothly lower like hypertensive emergency while keeping a target of CPP of at least 60 to 70 in mind for this patient. So that's how we would go about that in this case. Here is a summary list of some common signs on imaging that have been associated with hematoma expansion. Before we get into them individually, just know that there are others as well. And also various studies have compared them to each other and to other signs, and therefore their sensitivities vary greatly depending on the study, the methods, and the sample size. But the specificity is consistently high, usually greater than 80% and up to 100. There are some mimics for some of these that could explain some of the differences in specificity as well. So we'll look at the spot sign first. So here, this is a CT angiogram of the head. You see the outline of the hematoma itself. And then here at the periphery, you see this bright white spot. And so a spot sign is defined as a 1 to 2 millimeter focus of enhancement typically found at the periphery of the hematoma. It's thought to represent a microaneurysm, active extravasation, or secondary vessel disruption, all of which are hypothesized to lead to hematoma expansion. And notice I said focus of enhancement. So again, this is a CT angiogram, and it is a study that does require contrast. So now looking at the black hole sign, you can think of it sort of as a poor man's spot sign. But essentially, it's seen on a native CT head, so no contrast. And here, you can see it right here, this irregular shaped focus of hypodensity. And it can be irregular, it can be round, or it can be tubular in shape, but it's thought to signify the same as a spot sign on CTA. So then the swirl sign takes it one step further. So this is similar to a black hole sign, but slightly different, almost a more severe form of a black hole sign if you want to think about it that way. But swirl sign is defined as a hypo or isodensity within the hyperdensity of the hematoma. So here we see the same hematoma, and here we see, in our case, a hypodensity within it. But it has to extend across at least two contiguous five millimeter axial CT head slices, which we see it does here on two contiguous slices. And it's thought to represent turbulent flow into the hematoma and has been correlated with areas of ongoing hemorrhage in the OR during hematoma evacuation. It's also been found to be highly reproducible and independently correlates with hematoma expansion and poor functional outcomes. Another sign is the island sign. So island sign requires three independent small focuses of hemorrhage adjacent to the parent hemorrhage or four foci of hemorrhage or more that can be attached to the parent hemorrhage. But both of these are risk factors for hematoma expansion. Lastly, blend sign also correlates with hematoma expansion. And it's defined by a mixed density bleed with clear demarcation between an area of hyper and relative hypodensity. So here, the small component of this hematoma is far more hyperdense and clearly delineated from this relative area of hypodensity in the hemorrhage. Same thing over here, you see this well demarcated area of hyperdensity and a relative hypodense area of the hemorrhage adjacent to it. So that's a lot of signs. And as I mentioned, there are more. So the BAT score is one example of a score created from combining some of the most predictive features. So if CT shows blend sign, that would be one point. If they have any intra-clot hypodensity, such as swirl or black hole sign, that's two points. And if their non-contrast CT was done within two hours of their symptom onset, that's another two points. So a BAT score of three or more predicts hematoma expansion with 89% specificity. The main reason for considering these signs and scores is to identify patients who may experience a more severe disease course, worse outcomes, or may benefit from more aggressive interventions early on. Looking back at our case again, we see that in the first two images highlighted in red, there's a swirl sign, and in the third image, a black hole sign. So knowing that the onset was 60 minutes ago yields a BAT score of four. This alerts us to the potential that our patient may be at increased risk for hematoma expansion. So next, we're going to talk about reversal of coagulopathic states in order to answer question number three. So while our patient is not on an anticoagulant, I did want to review a few considerations for reversal. So in general, we would reverse an INR of greater than 1.3 or a prolonged APTT. There are a few instances that you would want to reconsider this approach, though. So if there is a small bleed and the INR is less than two, you may need to discuss the risk versus benefit with the patient and their families. So risks that you would actually consider that would give you pause for reversal would include things like life-threatening thrombosis, so DIC, HIT, active ischemia, and of course, the situation of a hemorrhage associated with cerebral venous sinus thrombosis, in which case the treatment would be anticoagulation and reversal would be contraindicated. Again, we always would want to correct thrombocytopenia. So these are a couple of considerations to keep in mind. So this follow-up figure is adapted from the 2022 Spontaneous ICH Guidelines as well. And while I don't want to review the whole chart, there's just a couple of things I wanted to draw your attention to. One is in the INR range of 1.3 to 1.9, the dose for four-factor PCC is lower than what you're typically used to dosing for therapeutic INR from the package inserts. So I wanted to draw attention to that. The other is a recommendation for activated charcoal. So if a DOAC is consumed in less than two hours from presentation, it is recommended to administer activated charcoal, though there may potentially be efficacy up to eight hours. And then lastly, with respect to iteracizumab and andexanet-alpha availability. So both iteracizumab and andexanet-alpha are two-way recommendations, which is a moderate recommendation, whereas if unavailable, then PCCs or four-factor activated PCCs are considered a 2B or weak quality evidence, but are still recommended. So this brings us to antiplatelet reversal. Recall our patient was on dual antiplatelets. So should we transfuse platelets, give DDAVP, give both or neither, and how should we decide what to do? The next four studies we're going to cover are not intended to be a comprehensive review of the literature, but more so represent current theories and ongoing areas of research in this aspect of ICH management. One of the largest and only multi-center RCTs for platelet transfusion and ICH is the PATCH trial. It was published in The Lancet in 2016. The trial randomized a total of 190 patients across Europe with supratentorial non-surgical, so smaller bleeds, with minimal IVH, presenting GCS greater than 8, and who were taking either mono or dual antiplatelet therapy for at least 7 days to either receive platelet transfusions or not. If you were randomized to the treatment arm and were on a COX inhibitor, aspirin, with or without an ADP or reuptake inhibitor, dipyrimidol, you received 1 unit of platelets. And if you were on an ADP inhibitor, clopidogrel, alone or in combination with aspirin, you received 2 units. What the PATCH trial demonstrated was that platelet transfusion was associated with an odds ratio of 2.5 for increased odds of death dependence, including in-hospital mortality, and increased serious adverse effects. That's a big deal, and it's been well accepted since the publication of the trial that for bleeds that fit these specific inclusion criteria, platelet transfusion is not recommended. However, what about the platelets that don't fit this criteria, and are there other aspects of the case that would make it so these results may not apply or be as applicable? One thing to consider is that in Europe at the time the trial was done, the buffy coat method was used of extracting platelets, which tends to leave more small inflammatory particles behind in the platelet packs. So this actually might increase your risk of transfusion reaction or an immunogenic reaction to the platelets. Also, these were smaller bleeds with no anticipated surgical needs for at least 24 hours. And lastly, they did not look at any imaging features, so we don't know if any of these patients had, you know, some of the high-risk features we just reviewed. And so we don't know if there are subgroups that would still benefit from platelets at this point. What's interesting is that in 2018, a study was published in World Neurosurgery that looked at the effects of platelet transfusion on outcomes when matching for case severity using specifically the ICH score. This was a single-center retrospective study. The center had predetermined criteria that would qualify patients based on subgroups that they felt were high-risk for platelet transfusion, which included if the platelet function assay was positive, if the patient needed any neurosurgical intervention such as an EVD or a crani, if there was a hematoma expansion on a repeat scan, or if there was either a big bleed or a bleed in a bad spot, essentially. So interestingly, prior to matching by ICH score, their results showed platelet transfusion was associated with increased odds of neurological decline of 4.7, a need for surgical intervention of 7.2, a higher MRS score of 3.6, and death of 6.1. These are definitely not good, right? But once they matched their pairs by ICH score, none of the odds ratios were significant, and all of the effects or the negative effects that we thought we were seeing from platelet transfusion was equalized. Notably, this study used apherese platelets for the initial years, then changed to leukoreduced platelets from whole blood. But given they saw the significant odds ratios for negative outcomes prior to ICH score matching, the method of platelet preparation is less likely a determining cause here. And this was a small single-center retrospective study, and so to be able to see this pattern is actually impressive. And obviously, there's some bias in the selection of who got platelets, and not all the biases accounted for by the ICH score alone. It does identify subgroups for who the risk of platelet transfusion is at least balanced against the benefit. But what is the benefit? Is it all just theoretical? Probably not, right? So the third study published in 2019 here is in Experimental and Therapeutic Medicine, and it looks specifically at intraoperative and perioperative platelet transfusions guided by TAG. So this was also a single-center case control study in which platelet transfusion was given to those with a TAG showing greater than 89% inhibition of platelet function. They found this decreased their estimated blood loss, blood transfusion, residual clot volume, drain output post-op, need for reduced surgery, and allowed for earlier drain removal. These results were also replicated in non-neurosurgical patients, such as cardiac surgery patients and others as well. So there does appear to be benefit to operative patients when treatment is guided by TAG or some other means of assessing platelet function. The last of the four studies was published in Crick Care Medicine in 2020. It was a single-center retrospective study that looked at a similar population to PATCH, but included IVH. They also determined who to treat with the DDAVP in two units of platelets based on their local protocols. So they hypothesized that since intraoperative DDAVP had been shown to confer similar benefits to what was seen in the TAG-guided platelet transfusion trial, that combining the two therapies would provide added benefit. Their study showed that increased risk of higher MRS, need for EVD, need for operative intervention, and mechanical ventilation were all associated with the treatment arm, which was not what they were expecting to see. However, their treatment group was defined to be a higher risk group based on their decision tree and had older patients with higher NIH stroke scales and more IVH. So it's hard to interpret their results, which likely reflect that selection bias. Regardless, for most small, supratentorial, non-operative, and non-operative includes don't need an EVD bleeds, empiric platelet transfusion is not warranted and most likely will result in harm. However, there may be subgroups of patients who may still benefit from either platelets, DDAVP, or both. The 2022 ICH guideline interpretation of these studies, and many more like them, are as follows. So for patients with spontaneous ICH being treated with aspirin who require emergency neurosurgery, platelet transfusion might be considered to reduce post-operative bleeding and mortality. For patients with spontaneous ICH being treated with antiplatelet agents, the effectiveness of desmopressin with or without platelet transfusion to reduce the expansion of the hematomas uncertain. And as we mentioned, for those who are treated with aspirin and not scheduled for emergency surgery, platelet transfusions are potentially harmful and should not be administered. It's important to note here that in the PATCH trial, although they enrolled patients on dual antiplatelet, the large majority of patients were on monotherapy with aspirin, and there was only four patients on dual antiplatelet in the entire study. And if I recall correctly, only one of those was on clopidogrel. So as far as RCT data goes, we don't have much for the other antiplatelet agents. And that's one of the reasons why the guidelines are specific to aspirin here. Another thing to note is that it's unclear when it comes to EVD placement if this is considered emergency neurosurgery and whether platelet transfusion carries the same risk. So let's say you have a small ICH, but a decent amount of IVH, and that patient is going to get an EVD but does not need a craniotomy or anything like that. Is that patient truly considered the same as a patient who's going to OR for hematoma evacuation in terms of platelet transfusion and risks of intraoperative or procedural hemorrhaging on an antiplatelet therapy? So these are some considerations and answers we don't have as yet. An important quick note on ticagrelor is that it is not reversible with platelets, as you may recall, and that there is an antibody antidote that's under phase III clinical trial testing currently. Lastly, with respect to other hemostatic treatment options, the recommendations from the guidelines are that the effectiveness of recombinant factor VIIa for improving functional outcome is still unclear in spontaneous ICH. And in patients at high risk for hematoma expansion by imaging criteria, it's still felt that the effectiveness of tranexemic acid has not been well established in terms of its ability to improve functional outcome, though there is an ongoing large RCT for this currently. And so hopefully we'll have more information regarding this question before too long. So then how does our review of that literature and guidelines inform how we should manage this patient? So should we give platelets, DDAVP, both, none, something else? How do we decide? I think what's most important is to recognize that there's a paucity of data addressing patients on P2Y12 inhibitors, as well as those with high-risk imaging features. And spoiler alert for the next question on whether or not emergent neurosurgical intervention is required, obviously they will need an EVD for this completely obstructed fourth ventricle. But having said that, does an EVD qualify as emergent neurosurgical intervention? There's no specific data on EVDs. And if you think about it, many EVDs get placed in the O.R. emergently, immediately prior to the O.R. And so teasing out that EVD-specific data is not always possible. Now, in cases like this, it's always best to have a multidisciplinary conversation with your colleagues in stroke and neurosurgery. My personal practice, preference, or pattern, and the way I advise my colleagues is to obtain a tag with platelet mapping. And if the inhibition is greater than 89% and the patient is on a P2Y12 inhibitor, my practice preference is to give DDAVP as a dose, as well as one unit of platelets, and reassess for the platelet inhibition percentage. And I target a platelet inhibition percentage under 90%, if and until they need to go to the O.R., and in which case, you can always hang platelets in DDAVP again. But that tends to be my practice pattern, and the reason to give both is just the quantity of platelets required to reverse Plavix is large. And so instead of giving multiple units of platelets to try to get under 90%, my preference is to use a combination of platelets and DDAVP. But there's no data to really say that any one management style in this case would be right or wrong. Would our patient potentially benefit from intrathecal TPA via the EVD? Do they require craniotomy or ectomy for hematoma evacuation, or do they qualify for minimally invasive image-guided resection with or without post-evacuation TPA to the hematoma cavity? So as I mentioned before, our patient requires an EVD for acute hydrocephalus that's obstructive. I do want to mention, though, there is some really important things to keep aware of for infratentorial hemorrhages. So in an infratentorial hemorrhage, evidence of hydro or partial effacement of the fourth ventricle, any decline in exam, and any bleed greater than 15 milliliters requires emergent suboccipital decompressive craniotomy. An EVD alone could result in upward herniation. And while EVDs can be placed prior to the OR, they should be done essentially en route and should be clamped as soon as they're placed in most cases until the decompression can occur. So what about intrathecal TPA via the EVD? And to understand what our options are and what we can expect from this therapy, we're going to take a look at the CLEAR-3 trial. So the CLEAR-3 trial is a multicenter, double-blinded, placebo-controlled, randomized trial that looked at placebo versus TPA being instilled into extraventricular drains for clearing IVH. And the inclusion and exclusion criteria are not listed here for you to learn or memorize. It's just to make you aware that there are several factors for safety that were considered and adhered to prior to instilling TPA into these EVDs. If these criteria were met, however, the findings show overall a decreased rate of mortality, but no real improvement in three-month outcomes. But it was found to be safe, and a subgroup of patients actually did favor TPA. And that included those with at least 20 to 50 mLs of IVH to begin with and who ended up having a larger amount of their IVH cleared by the TPA. So still a subgroup of patients that seem to have benefited not just in mortality, but in outcomes from treatment. So what about the indication for open surgical evacuation and decompression via craniotomy? So the first trial, the STITCH trial, was done in 2005. And it found no benefits to early surgery versus initial medical management. There was, however, a subgroup of superficial, so less than 1 centimeter deep, lobar hemorrhages of 10 to 50 mLs that suggested favoring early surgery. So in order to study the subgroup, STITCH2 was conducted in 2013. It was published specifically looking at less than 1 centimeter deep lobar hemorrhages. However, instead of the 50 mL cutoff, they increased it to 100 mLs and specified that there should not be any IVH. And they looked at, again, early surgical management versus initial medical management and found a small survival benefit, but no improvement in outcomes. And again, this is for open craniotomy and unclear how much the increase in cutoff for hematoma volume impacted these results. Lastly, the MISTI3 trial that was published in 2019 was an international double-blinded RCT that looked at minimally invasive image-guided evacuation. And they would leave a drain in place via which to instill TPA into the hematoma cavity intermittently to remove additional clot. In addition to the benefit in all-cause mortality, they were also able to show that if they could get the residual hematoma volume to less than 15 mLs, they had an increased chance of good outcome at one year. So back to our case for a summary of our management. So what should the blood pressure goal be for this patient? We said a CPP-based target once the EVD is placed would be ideal. It may be that the patient does not have elevated intracranial pressure once the EVD is placed. And if the hematoma does not expand significantly and they don't develop severe cerebral edema or hematoma expansion as is predicted based on their high-risk features, then that can be re-evaluated to a systolic blood pressure goal as opposed to a CPP target. Are there any high-risk features by imaging? We said yes. We mentioned that there was both a swirl sign and a time-to-presentation from ictus of less than 2 and 1 1⁄2 hours. So their BAT score is 4, which we said predicted hematoma expansion with 89% specificity. And then as far as their antiplatelet therapy goes, should we transfuse with platelets or DDAVP or both? We said in our situation, there was very little data to guide management, and TAG-guided therapy was recommended. In most cases of non-surgical ICH in aspirin therapy alone, the recommendation is to not give platelets or DDAVP empirically. And in surgical cases on antiplatelets, the recommendation is at least for platelet transfusion. And there's insufficient data for DDAVP with or without platelets at this time. Does this patient need emergent surgical intervention? We just reviewed that, at minimum, they would need an EVD. And likely, if there's no EVD-tracked hemorrhage and if the hematoma doesn't show significant expansion on the stability scan, this patient would be a good candidate for intrathecal TPA. So that is a review. And thank you all for your time and attention today.

hemorrhagic stroke

management

blood pressure control targets

hematoma expansion

imaging patterns

reversal of coagulopathic states

surgical treatment