Management Principles for Spontaneous Intracerebral Hemorrhage

My name is John Terry. I am a neurologist and  Associate Professor inat the Department of Neurology. I am a neuro-interventionist, neuro-intensivist and stroke neurologist so a lot of my day consists of dealing with stroke patients. Today, I’d like to speak about the management principles for spontaneous intracerebral hemorrhage. 

There have been multiple iterations of guidelines for management of this disease, the most recent of which was in 2010 when the American Heart and Stroke Association published their guidelines for the management of this disease. I will be referring back to this particular document as we go along to point out some of the major recommendations and also the updated recommendations. 

We really know probably the least amount about spontaneous intracerebral hemorrhage which is a type of stroke. If you look back historically in 1999, there was an iteration of intracerebral hemorrhage guidelines and at that time there were only five small randomized medical trials and four randomized surgical trials dealing with acute intracerebral hemorrhage. 

What I would like to do is review the epidemiology of this disease, talk about the pathophysiology and the treatment options. The blood supply to the brain varies by location. There are several areas that are fed predominantly by very small blood vessels that branch directly off of the larger blood vessels. These areas include the basal ganglia, the thalamus, the pons and the cerebellum. The way that the circulation differs in these areas is that, unlike the normal pattern where the arteries as they get farther away from the heart branch like trees and become progressively smaller, these areas have very small branches that come off main arterial trunks. If one looks at the blood pressure throughout the circulatory system, one can see that initially, for instance, in the aorta the blood pressure is very high and it is pulsatile meaning it is at its highest when the heart contracts and lowest when it relaxes. As you move farther away from the heart, that pulsatility decreases as does the pressure. And so, in effect, the small arteries are protected from the systemic blood pressure by the intervening branches of the intermediate sized arteries. In the areas that we have been speaking about in the brain where you have small branches that come directly off large branches you don’t have the benefit of the pressure decrease over the length of the arterial system so that these arteries are exposed to higher pressures than other arteries of similar size in a more normal configuration. 

The walls of these small arteries are subject to degeneration over time, however, people that have high blood pressure have much more rapid degeneration. I have a slide here that shows a cross-section of several of these small arteries. Some of these arteries are normal in which case you see a round wall at the circumference of the artery. The wall is of uniform thickness and within the walls you can see red blood cells. In others you see that the wall looks irregular. It is thickened and has various degrees of thickening around the outside, and it also is filled with red blood cells. The thickening and irregularity represent degeneration that occurs over time and especially in patients with high blood pressure. It involves deposition of abnormal protein and the result of this thickening is that the artery wall becomes more brittle and more prone to rupture. 

Additionally, small out pouchings of the arterial wall can occur. These are termed Ccharcot-bouchard aneurysms, and these are felt to be a large factor in patients who have intracerebral hemorrhage where the out pouching becomes very narrow and may rupture and bleed. 

If you look at the epidemiology of intracerebral hemorrhage, it causes about 10 to 15% of strokes, so it is a small minority of patients that present with stroke symptoms who have an intracerebral hemorrhage. If you look, roughly 70% of patients that come in with stroke symptoms have what we term an ischemic stroke, which is a situation where a blood clot forms within an artery and cuts off blood flow to part of the brain. In intracerebral hemorrhage, a blood vessel ruptures and blood leaks out. 

Only about 37,000 to 52,000 cases of intracerebral hemorrhage occur each year which is smaller than the number of cases of ischemic stroke. If one looks at the location of these hemorrhages about 50% were are what we call “deep” in location,” meaning they occur not on the surface of the brain but more towards the center. The most common area that is involved is called the basal ganglia. About 16% occur in the brain stem and cerebellum in the lower and more posterior parts of the brain. And based on other demographic characteristics, the incidencets of intracerebral hemorrhage varies from about 10 to 20 cases per 100,000 people per year. 

Although intracerebral hemorrhage is more rare than ischemic hemorrhage, unfortunately it has the highest mortality of all stroke types with only about 38% of victims surviving the first year. The 30 day mortality has been reported anywhere from 35% to 52% and in some ways depends on the location of the hemorrhage so that for patients that have deep hemorrhages about half of the patients will die within the first year. For those that have brain stem hemorrhages, 65% will die in the first year. And if you look at the deaths about half of them occur in the first two days. So this type of stroke really is very serious and has a high mortality. 

Unfortunately, it is the type of stroke that we have the least knowledge of what the proper treatments are, at least in terms of treatment options. As I said earlier, about half of the people die in the first couple of days. This has been a fact has been studied in order to try to find a way to mitigate these deaths. A physician by the name of Dr. Brott performed a study in 2007 which looked at the first three hours after the onset of intracerebral hemorrhage and in patients that came into the hospital after this had occurred. They performed a neurologic evaluation and a CT scan at presentation, at one hour, and at 20 hours after the patient arrived. What they found is that 26% of the subjects showed a substantial growth in the volume of the hemorrhage between the baseline and the one hour scan. So in the first hour after being in the hospital continued bleeding occurred which resulted in an enlargement of the blood collection noted in the brain. 

Additionally, about 12% of subjects had substantial growth between the one and 20 hour CT scans. They also found that patients that did have hematoma enlargement or continued bleeding had worse outcomes over time, and the 30 day mortality was significantly higher in patients with enlargement compared to those without enlargement. It was felt that possibly this could be a point where intervention could be applied to help improve outcomes, since this is a phenomenon that occurs in the hospital after the patients had presented and occursred in a time frame which could be addressed. 

In terms of looking at factors that influence the outcome of patients, the volume of the hematoma or intracranial blood collection has a big impact on outcome and is a powerful predictor of death by 30 days. Also, the extent of disability as measured by the Glasgow Coma Scale at presentation is a good predictor of how people will do. Hydrocephalus is another indicator of increased mortality. Hydrocephalus occurs when blood breaks into fluid filled spaces in the brain and can block outflow of that fluid so that it is trapped and the fluid can build up and increase pressure in the head. 

As we just said, we know that the early risk of neurologic deterioration in intracerebral hemorrhage is high and is usually associated with hematoma enlargement. EIn the expansion of, the hematoma is associated with a nearly five fold increase in clinical deterioration, poor outcome and death. So the prevention of injury from hematoma development enlargement really has been the focus of our therapeutic intervention trials. 

Looking back at the American Heart Association recommendations for treatment of intracerebral hemorrhage, they point out that although there are some clues that a clinician can glean from the patient on presentation, it is impossible to know that whether a person that presents with stroke symptoms has had hemorrhage or has a blocked cerebral artery so that imaging, in particular CT scanning, is important to perform early on. 

Hematoma expansion has been shown to be predictive of clinical deterioration and increased morbidity and mortality. CT angiod geography or contrast CT may identify patients at high risk of hematoma expansion based on the presence of contrast extravasation within the hematoma. Other forms of imaging are reasonable to perform to try to more specifically identify causes of the hemorrhage. Specifically , entities like arteriovenous malformations, tumors, cerebral vein thrombosis or aneurysms as these entities have other treatment considerations that need to be assessed. 

In terms of medical treatments for these patients, as I had said before, in 1999 there were only four small randomized trials looking at patients being treated for intracerebral hemorrhage. One of the trials looked at the use of steroids as a possible way to decrease swelling around the hematoma in the brain. One of them looked at hemodilution, or giving fluids to make the blood thinner, to help improve flow. And one of the trials looked at glycerol which is a medication that may also help decrease the swelling. None of the four studies showed any benefit of the three therapies discussed here. And, in particular, one of the trials showed that patients treated with steroids were more likely to develop infectious complications than those who received placebo. In this particular intervention not only did it not help the patient, it was actually harmful. 

Another treatment strategy that has been discussed or thought of has been to use medications that favor clot formation in order to try to stop bleeding from occurring. Clot formation is the natural way that the body responds to bleeding and the natural way that the body can stop bleeding. Unfortunately, several trials using different mediations have been attempted and have not been shown to benefit patients overall in outcome and, in particular, harm may occur due to unwanted clotting which may result in things like deep venous thrombosis, heart attack and other problematic clotting. 

So the recommendations from the American Heart Association are not to use pro-thrombotic medications or medications that favor clot formation. 

Another strategy that has been discussed is to decrease the arterial pressure which may help decrease the amount of blood that is forced out of the area of rupture. An initial concern about this potential strategy is that if the blood pressure were lowered, the concern was that that would lower the circulation of blood around where the hematoma is because the tissue is compressed around the hematoma. A physician by the name of Dr. Powers performed a study looking at cerebral blood flow during reductions and in arterial pressure in patients with acute hemorrhage u. Using a PET scanner.  I it was a small subject trial including 14 patients with intracerebral hemorrhages who were studied 6 to 22 hours after the onset of hemorrhage. The regional blood flow was measured with a PET scanner and then patients were randomized to receive either a blood pressure lowering medicine or placebo. The goal was to lower the blood pressure by about 15% in those receiving blood pressure reducing medications. After the blood pressure was reduced to the target goal, cerebral blood flow was measured again. They used nicardipine or labetalol as the agents and were able to lower the mean arterial blood pressure from 143 to 119 which was about a 16% change. 

What they found is there was no significant change in blood flow either globally in the brain or around the intracerebral hemorrhage. They concluded that there was a less than 5% chance that cerebral blood flow would fall by more than 2.7 mils per 100 grams of tissue per minute. 

This trial really was one of the first that supported the idea that lowering blood pressure in patients with intracerebral hemorrhages is actually safe. Other trials were published that supported this idea. T– two Japanese trials – one in 1998 and one in 2004- were published that showed a rough dose response relationship between blood pressure and hematoma enlargement. 

So we know that acute high blood pressure is very common in patients presenting with intracerebral hemorrhages. In fact, it is almost unheard of that patients are not hypertensive when they come in. And we now have evidence that there is not a problem with lowering the blood pressure around the clot or in the brain in general. It is felt that elevated blood pressure may contribute to hematoma expansion by forcing more blood out through the hole that has occurred in the arterial system. 

It has been an area of great interest in terms of potential treatment interventions, t. The thought being that if it is safe to lower blood pressure when a patient comes in, doing so may result in less hematoma growth. An NINDS sponsored trial named ATACH for Antihypertensive Treatment of Acute Cerebral Hemorrhage was conducted in the early 2000s. It was a three year multi centered, open labeled, non randomized phase one pilot trial which was aimed at trying to establish the safety of decreasing blood pressure in patients that present with intracerebral hematomas. 

They used a 4 Tier dose escalation design so that initially the target range of blood pressure is 170 to 200. After the requisite number of patients were enrolled, the data was reviewed for evidence of harm. No harm was found and therefore the target range was lowered to 140 over 170 and that process was repeated and the final target goal was 110 to 140 millimeters of mercury. The results of this trial really showed that it is not only safe, but feasible, to rapidly lower blood pressure early in the course of patients presenting with this kind of stroke. 

A more recent trial, which is published in the New England Journal of Medicine in 2013, is the INTERACT II trial that stands for Intensive Blood Pressure Reduction and in Acute Cerebral Hemorrhage Trial II. This was a much larger trial. It included 2,839 patients that were treated within the first six hours after the onset of their hemorrhage. 719 were treated to a low systolic blood pressure goal of less than 140 millimeters of mercury. The remaining 786 were treated to a systolic blood pressure goal of less than 180 millimeters of mercury. The primary outcome was a modified Rankin score of 0 to 2 versus 3 to 6 at 90 days. They also included a prespecified ordinal analysis of the outcomes. Secondary outcomes included safety and quality of life. The results of this trial show that there was no difference in the primary outcome or adverse events. And the P value there was 0.06 which was almost significant but not quite. However, in the ordinal analysis, a benefit of intensive treatment was shown to a level of significance of P = 0.04 and they also found improved quality of life in the intensive group. So this is another major trial that shows that lowering the blood pressure is safe and actually may confer some benefit as demonstrated by the end points that were chosen for the trial. 

The current recommended guidelines for treating elevated blood pressure and spontaneous intracerebral hemorrhage are more liberal than those that we have been speaking of in the trials. Typically, we want to keep the systolic blood pressure less than 200 millimeters of mercury and the target goal really is less than 160. So those were the medical treatment options. The trial looking at trying to form a blood clot to plug the hole, did not show benefit. However the trials showing decreasing the blood pressure showed that it was safe and probably does have some beneficial impact. 

What about surgical trials? In the past, surgery has been a major consideration for patients with intracerebral hemorrhage and that treatment pattern actually was present for some time before clinical trials were undertaken to try to establish whether or not it was the best way to treat these patients. A Seminole seminal trial called the STICH trial which stands for International Surgical Trial and Intracerebral Hemorrhage randomized 1033 patients from 107 centers over an eight- year period beginning in 1995. So this was a large multinational, multicenter trial that gave us some information about surgery in these patients. Patients were randomized within 72 hours and then operated on within 96 hours after the onset of their stroke if they had a clot that was greater to or equal to 2 centimeters in diameter. 

The patients were randomized only if the neurosurgeon was uncertain as to the benefit of the surgery. The patients with Glasgow coma scores less than five were excluded because of the poor outcome that those patients would likely have. The primary outcomes that were looked at were death and, disability as measured by the extent of Glasgow outcome scale at six months. Secondary outcomes were death, Barthel index score and the modified Rankin score at six months. 

506 patients randomized in surgery and 530 were randomized to medical therapy. The groups were well-matched in terms of their characteristics. There was about a 26% crossover rate so that 26% of patients that initially were randomized to the medical arm were operated on, presumably because they had worsening in their exam. Craniotomy was the type of surgical intervention used in 85% of the subjects who crossed over. Whereas if you look at the patients that were operated on as a whole, 75% of the patients underwent craniotomy with the remainder undergoing less invasive surgical techniques. And 93% of the patients were available for analysis at follow-up at six months. 

In an intention to treat analysis, surgery within 96 hours was associated with a non-significant absolute benefit of 2.3% at six months. And o Overall when you review the results of the study there was no clear benefit of surgical intervention in these patients, which was somewhat of a disappointing outcome. However, it is important information to know that the thought currently is that open craniotomy in general does not benefit patients who have had an intracerebral hemorrhage. Some subgroup analysis showed that patients with a Glasgow coma score of 9 to 12, those with low bar clots, and those with clots 1 centimeter from the surface may have been helped by early surgery. However, this wasn’t statistically significant.   

It is interesting though that patients in the Glasgow coma scale score a range of 9 to 12 normally would be only moderately symptomatic from their hemorrhage.  , Wwhereas patients in coma were actually found to do better with medical management. So surgery hurt the worst affected patients. Craniotomy has not been shown to be helpful in this disease process. One of the potential thoughts on why that might be the case is that many of the blood clots that accumulate in intracerebral hemorrhage do so deep in the brain, so that in order to reach them to remove them the surgeon must go through normal brain which causes additional injury. This additional injury may offset any benefit from the surgery itself. 

So consideration to less invasive ways of removing clot has been a more recent area of interest in treating these patients. The potential advantages of minimally inevasive evacuation include possibly reduced operative time, the possibility of avoidance of general anesthesia (because these cases may potentially be performed under local anesthesia), reduction of tissue trauma, especially for the deep lesions, and facilitation of earlier evacuation thaen it is possible or practical with conventional craniotomy, as craniotomy typically takes a fair amount of time to mobilize the team and prepare the operative room to do the procedure. 

The limitations of this approach may include reduced surgical exposure so that targeting the lesion may be more difficult, an inability to treat structural lesions such as arterial venous malformations or aneurysms which can be the cause of intracerebral bleeding, the potential for rebleeding if clot dissolving medications are used to aid in removal of the clot, and the possibility of increased of risk of infection related to prolonged indwelling equipment that may be used to help drain the clot over time. 

Looking back at the STICH trial results it suggested that subjects treated with a non-craniotomy surgical approach had a worse outcome than those treated with conservative management. However, the confidence interval in this estimation included one which brings into question its validity. It was unclear in that particular trial how a given patient was chosen for craniotomy versus less invasive procedures so that the information that comes out of this trial is not very helpful in helping to answer the question about whether less invasive methods may help. 

As I said earlier, clot dissolving medicine, specifically TPA, hasve been considered in treating patients with intracerebral hemorrhage. In general, the method of treatment would be gaining access to the area of the brain where the hematoma is and instilling TPA to help break the clot down and then allowing that to drain out of the head. 

There have been many small trials looking at various protocols and methods for carrying out this type of treatment. In general, these trials have shown that it is relatively safe. Given these results the NINDS funded a multi-center randomized controled trial called MISTIE which stands for minimally invasive stereotactic surgery plus TPA for intracerebral hemorrhage evacuation. This trial was designed to test the hypothesis that early use of minimally invasive surgery plus TPA for three days is safe for the treatment of intracerebral hemorrhage, and it will produce a clot size reduction compared with medically treated patients. 

Again, going back to the American Heart Association recommendations for intracerebral hemorrhage regarding surgery, the current recommendations state that for most patients with intracerebral hemorrhage the usefulness of surgery is uncertain and not recommended. A sSpecial case in point that we won’t get into, but is worth mentioning, is that patients with cerebellar hemorrhage are usually felt to be good candidates for surgery. But hematomas in other locations in general are felt not to be good candidates. The effectiveness of minimally invasive clot evacuation is still being evaluated, but is not currently recommended outside of the clinical trial.

Let’s talk a little bit about the MISTIE trial. We know that there is a strong correlation or relationship between the volume of the hematoma in the brain and the patient’s outcome at 30 days. I have a slide here that shows a table that describes outcome at 30 days related to the volume of the intracerebral hematoma, and in general what one sees is that patients that have very small hematomas tend to do well and as the hematoma increases in size their outcomes become progressively worse.

The inclusion criteria for the MISTIE trial were patients that were older than 18, those that had a Glasgow coma score of less than 13 or an NIH stroke scale of greater than six so that these patients had significant effects from their hemorrhage, but were not comatose and devastated by their hemorrhage. Patients needed to have a spontaneous supratentorial hemorrhage of greater than or equal to 30ccs in volume and over time the clot needed to be stable. So that the patients come in and are evaluated with a CT scan. Another scan is performed at six hours and they are not able to participate in the trial until their two scans six hours apart showed no significant change in the hematoma volume. 

I have a slide here of one of the patients who was enrolled in this trial here at our hospital who had a catheter placed into their hematoma. What the slide shows is that over the course of about one week the hematoma went from a fairly significant size to almost completely gone. In contrast, this other slide shows a patient that was randomized through the trial to the medical arm and over the same amount of time there was not much change in the hematoma over the first week. And actually at day 22 the hematoma had only come down in size very slightly. So this shows a fairly dramatic comparison that with the catheter method the hematoma can be drained fairly rapidly. 

The MISTIE trial has had several iterations. In the MISTIE II trial when the data were looked at of outcomes of patients that were enrolled, what they found was when comparing a similar plot of hematoma volume versus outcome to those that were not treated there were more patients that fell in the good outcome ranges that had actually had larger hematomas, which was support of the fact that patients who had their hematomas drained had better outcomes. And this is also in concert with the observation that has been made in the past that lower volumes tended to have better outcomes and higher volumes tended to have worse outcomes. 

If one looks at the day 365 modified Rankin scale score which is a score of functional abilities, you can see there is roughly 14% reduction in the amount of disability of patients among those who received the catheter drainage compared to those who received only medical therapy. 

Also some interesting observations from the data is that the patients that underwent the catheter drainage part of the study had a shorter length of stay in the hospital and even given the cost of the surgical treatment had a lower cost of care for their hospital stay. 

We did learn some interesting aspects about recovery after hemorrhage in the study as well. The study is unique in the fact that it followed patients for up to a year after their hemorrhage, which is longer than most other study’ies used to have a follow-up time. In those patients that were treated medically what was found was that there was a period of recovery that tended to span the first 180 days where gradual improvement in ability occurred, but after 180 days that improvement plateaued and there was not much significant improvement after that. Whereas those that had the drainage procedure had steady improvement throughout the entire 365 days over which they were followed during the trial. In fact, the rate of recovery was equal or f even greater in the second 180 days compared to the first. 

Again, back to the American Heart Association Guidelines, the recommendations are aggressive full care early after the onset of intracerebral hemorrhage. It recommends against routine open craniotomy for surgical treatments, and it recommends blood pressure control according to the levels in the guidelines. So in summary, what we know about intracerebral hemorrhage treatment thus far is that blood pressure control is a cornerstone of the treatment. It does not cause harm and probably is beneficial. Routine use of open craniotomy for surgical treatment of the disease does not confer benefit, however, less invasive methods particularly drainage of the hematoma cavity with an inserted catheter is showing great promise for helping these patients in the future