Good afternoon, and welcome to the SCCM Neurocritical Care Board Review for subarachnoid hemorrhage. My name is Nikhil Patel, I am a neurointensivist at Atrium Health's Carolinas Medical Center in Charlotte, North Carolina. I have no disclosures. Here are our learning objectives for today. Subarachnoid hemorrhage is the least common type of stroke, comprising 1-6% of all strokes, but it has a disproportionate impact on death and disability given the younger population affected and the severity of the condition. The median age of onset is 53 years, with women affected more than men. Case fatality can be up to 50% and about 30-50% of survivors will experience long-term disability. Despite the severe nature of this condition, both incidence and case fatality have decreased worldwide over the past several decades. Subarachnoid hemorrhage can be categorized into primary or secondary causes. Primary subarachnoid hemorrhage is overwhelmingly caused by a ruptured intracranial aneurysm, with this cause making up 85% of cases. Another 10% of primary SAH patients will have non-aneurysmal perimensin-cephalic SAH, which we'll discuss later, and 5% will have another vascular malformation as the cause. Most of this discussion will focus on aneurysmal subarachnoid hemorrhage. Secondary subarachnoid hemorrhage typically has a different natural history than aneurysmal subarachnoid hemorrhage, although some aspects may be similar. The pattern of blood is typically more cortical in secondary SAH, and there are other clues that point away from a ruptured aneurysm being the cause. These are some conditions to remember or keep in mind when evaluating the cause of secondary subarachnoid hemorrhage, and this is an example of one. So here we have a convexity subarachnoid hemorrhage in the right frontal lobe highlighted there. In this particular case, it's secondary to RCVS. About 5% of people have an intracranial aneurysm. Most of these don't rupture, but there are certain risk factors that will increase the risk of rupture and are typically split up into modifiable and non-modifiable risk factors. Improved recognition and treatment of the modifiable risk factors listed here have been elicited as one of the reasons that the incidence of subarachnoid hemorrhage has been decreasing over the past several decades. And current guidelines recommend screening for aneurysms if the patient has two or more first-degree relatives with aneurysms or subarachnoid hemorrhage. The classic presentation of subarachnoid hemorrhage is a patient experiencing a severe headache, often the worst of their life, with associated nausea and vomiting. Loss of consciousness can occur if the aneurysm rupture is severe, and back pain can ensue later on in the course as the subarachnoid blood settles out in the lumbar cistern and irritates the nerve roots. On exam, patients are frequently somnolent. They can have focal cranial nerve palsies, either from mass effect from the aneurysm itself or increased ICP, meningitis is common, and focal neurological deficits can also occur, especially if there is associated ICH along with the subarachnoid hemorrhage. Seizures can also occur, but can be difficult to distinguish from posturing in the acute setting. We'll discuss this further when diving into the management of seizures in subarachnoid hemorrhage. Multiple grading scales are commonly used to standardize communication about the severity of SAH. The two most commonly used scales are the Hunt and Hess scale, shown on the left, and the WF and S scale, shown on the right. Higher numbers of each delineate increased severity. Both scores are indeed strongly correlated with mortality and functional outcome, but similar to the ICH score, they should not be used solely as a prognostication tool or as a basis of a goals of care discussion. Patients frequently can improve quite dramatically with basic neuroresuscitation, so using these scales as a singular predictor of mortality runs a high risk of feeding into a self-fulfilling prophecy. The diagnosis of subarachnoid hemorrhage is most often made on head CT. It is rapid, it's widely available, and it's pretty accurate when symptom onset is within several days. As time since a patient's initial symptoms progresses, the sensitivity of head CT decreases. This is when MRI can be helpful. The SWI and GRE sequences can be useful to detect subacute subarachnoid hemorrhage, and MRI can also help diagnose secondary causes of SAH. In cases of subacute presentation, lumbar puncture can be very helpful in clinching the diagnosis. Typically, we look for xanthochromia, which characterizes the yellowish appearance of CSF that results from the breakdown of blood products. It can be difficult to know when to order a lumbar puncture, but the Ottawa subarachnoid hemorrhage rule is one tool that could be used to determine when an LP can be high yield in the diagnostic workup. Previously, we touched on clinical grading scales that are used to communicate the severity of SAH. There are also radiographic scales that can be helpful in communicating the severity of SAH as well. These also have the benefit of risk stratifying patients by the likelihood of symptomatic vasospasm. One of the most commonly used scales is the modified Fisher scale shown here, which categorizes bleeds based on the thickness of the SAH clot as well as the presence of IVH. Higher grades predict higher likelihoods of symptomatic vasospasm. Here is an example of a modified Fisher 4 subarachnoid hemorrhage. There is only one cut of the head CT shown, but this is actually a thick SAH with greater than one millimeter in depth that has IVH present. After the diagnosis of SAH is made, the next step is to identify the aneurysm. This can be done with several imaging modalities, the most common being CTA and DSA, or digital subtraction angiography. In general, CTA is the initial imaging modality of choice because it's noninvasive, it's quick, and it's fairly sensitive. But DSA is often done as well as the CTA, either in cases when no aneurysm is seen on CTA or for better characterization of the aneurysm for proper surgical planning. MRI and MRA can be used as well, especially if no aneurysm is identified on CTA or DSA. Or if nonaneurysmal causes are suspected. On the right here is an example of an ICA terminus aneurysm seen on CTA and its reconstructions, as well as the same aneurysm seen on DSA and its corresponding reconstructions. We touched earlier on nonaneurysmal perimesencephalic SAH as another cause of primary subarachnoid hemorrhage. So let's dive a little deeper on this now. This type of SAH is characterized by localized bleed in the anterior pons and or midbrain, shown here on this CT scan, with blood sometimes extending into other cisterns and fissures. There's no overt IVH, although there may be sediment that settles into the occipital horns of the lateral ventricles. The clinical course of this type of SAH is more benign, although the cause of these types of bleeds is unknown. It's suggested that there is a venous origin to the bleed, but this is not fully described. This is typically still a diagnosis of exclusion. So many centers will perform some sort of vessel imaging to attempt to identify an aneurysm and repeat that vessel imaging seven days later, usually, to ensure that there's no aneurysm identified. When a patient presents with an SAH, there is a profound inflammatory response, both in the brain and systemically from the aneurysm rupture. This requires patients to be quickly stabilized and resuscitated in order to address multi-organ dysfunction that can ensue. We'll talk about the following hyperacute management considerations. Hemodynamic instability can manifest in several ways, the most common of which is acute hypertension. Acute hypotension can also occur in the setting of stress cardiomyopathy and cardiac arrhythmias. Acute left ventricular dysfunction is seen in up to 30% of patients, and it's thought to be due to the sudden catecholamine surge following the initial aneurysm rupture. Bedside ultrasound on admission can be helpful for a timely diagnosis. Cardiac arrest can also occur, and while patients that present in cardiac arrest from SAH often have high mortality, a good outcome may be possible with up to 25% of patients surviving to discharge in one multi-center study. Respiratory complications can occur via several mechanisms. First, a low level of alertness leads to an inability to protect the airway, leading to aspiration. Aspiration can also occur from vomiting as a symptom of elevated ICP, and hypoxia can result from neurogenic pulmonary edema, shown in this chest x-ray to the right, the precise etiology of which is poorly understood. Acute hydrocephalus occurs in about 20% of SAH patients, with the onset occurring within the first several hours or days after the aneurysm rupture. The mechanism is due to impaired CSF reabsorption. This can manifest as depressed level of consciousness, downward gaze, cranial nerve 6 palsy, and is treated with the placement of an external ventricular drain, or EVD. Aneurysm re-bleeding is one of the most feared hyperacute complications, as it's commonly lethal with up to 60% mortality. The highest risk is within the first 72 hours, with higher grade SAH, larger aneurysms, hypertension, and coagulopathy being the risk factors for re-rupture. With that in mind, the risk of re-bleeding is reduced by lowering the blood pressure, typically to less than 140 to 160, based on institutional norms, securing the aneurysm quickly, and correcting coagulopathy. Short-term use of antifibrinolytic medications can be used, such as TXA or Amicar, if there's a delay in treating the aneurysm, but is associated with thrombotic complications with longer duration of use. With the goal of treating the aneurysm quickly, there are two major options to do that, coiling shown on the left, and clipping shown on the right. In general, coiling confers a lower epilepsy risk, a lower risk of death and dependency as time goes on, and is the only option in certain aneurysm locations, such as posterior circulation aneurysms. On the other hand, clipping has a lower risk of re-bleeding, and gives the opportunity for hematoma evacuation in cases of associated ICH. It's also the better option in many cases of irregularly shaped aneurysms. Acute global cerebral edema occurs in 10% to 30% of patients on presentation, and is associated with higher grade bleeds and poor outcomes. There are no specific treatments for this, but hyperosmolar therapy can be considered if there is evidence of sustained elevations of ICP, typically appreciated when placing an EVD for hydrocephalus. Seizures can occur at any point during the hospital course, but up to a quarter of patients may present with seizure-like symptoms at onset. It's often hard to distinguish true epileptic seizures from posturing, and given that this is all happening emergently, confirming any activity with EEG can be difficult. There isn't much high-level evidence on AED management, but it's common practice to load AEDs on presentation. Theoretically, any seizure activity in the hyperacute period may increase ICP and risk the aneurysm re-bleeding, and given the profound harms that can result from that, many believe the risk-benefit analysis favors the use of AEDs. To balance that aggressive treatment of potential seizures up front, it's recommended that AEDs be discontinued after their aneurysm is secured, or after a short course of five to seven days, as there is an association between cognitive dysfunction and duration of AED therapy. Almost all seizures will be non-convulsive in nature, so the liberal use of continuous EEG is recommended. There will be more EEG examples during other talks in this course, but this EEG recording demonstrates a left seizure in the... Aside from the hyperacute systemic management considerations, the most prominent sequela of subarachnoid hemorrhage is the severe brain injury that can result. This has historically been attributed to large artery cerebral vasospasm causing ischemic injury to the brain. As we have discovered more about this condition, however, the conventional thinking has evolved away from solely large artery vasospasm being the major source of injury, and shifted toward a paradigm of complex multifactorial pathophysiology. One way to think about the brain injury is to recognize two phases of injury, early brain injury and delayed cerebral ischemia. Early brain injury is damage that ensues from the initial aneurysm rupture, which can lead to a sudden transient ICP elevation, transient global ischemia resulting from that ICP spike, and a series of further processes that leads to cell injury and death. There can also be two other mediators of early brain injury, one being the systemic complications that we just discussed, and two being any injury from the actual subarachnoid intracerebral or intraventricular blood itself. In this CT, we see that there is a subarachnoid hemorrhage along with hemorrhage in the brain parenchyma itself, both of which will lead to early brain injury. And this injury is separate from delayed cerebral ischemia, which is an SAH-associated brain injury process that typically develops three to 21 days after the initial aneurysm rupture. Traditionally, this has been thought to be due to large artery vasospasm, but the etiology of this has undergone a paradigm shift over the past several decades to involve multiple pathophysiological processes. This is the most feared complication of SAH, and most of the ICU management revolves around preventing, diagnosing, and treating this. And it's feared because it has a very strong association with functional outcome. There are many different terms and definitions used in this space, and they can be quite confusing and often overlapping. So let's dive into that a little bit further. One of the key features in aneurysmal subarachnoid hemorrhage is that up to 70% of patients will exhibit some degree of vessel narrowing on angiography, typically between 3 and 21 days after initial onset. However, only about 30% will develop clinical symptoms attributable to ischemia from this vasospasm. This underlines the importance of distinguishing angiographic vasospasm, which is a radiographic feature appreciated on imaging, from symptomatic vasospasm, which refers to clinical symptoms experienced by the patient that are attributable to ischemia from the visible vasospasm. Angiographic vasospasm does not correlate well with outcome, but symptomatic vasospasm is associated with DCI and outcome. Of note, symptomatic vasospasm and DCI often have overlapping definitions in the literature and may be referring to the same concept. Diagnosing DCI can be very difficult. In most patients, serial monitoring of the neurological exam can help detect when symptomatic vasospasm is occurring, and usually we suspect DCI if there is a new focal deficit or if there is a greater than two-point decrease in GCS that lasts for at least one hour and can't be explained by another cause. However, many high-grade patients have poor exams to begin with, and this makes it very hard to detect any changes. These are also the patients at highest risk of DCI, so this makes the process of diagnosis even more challenging. Particularly in these patients that have poor exam findings, diagnosis and treatment initiation of DCI may be somewhat subjective and based on neuroimaging findings rather than an exam finding. Transcranial Doppler, or TCD, is a commonly used diagnostic modality to augment the neurological exam when trying to diagnose DCI. Its advantages are that it's non-invasive and particularly well-suited for detecting DCI in the MCA and ICA. It's less reliable when examining the ACA and posterior circulation, and the overall sensitivity and specificity of the test are very operator and bone window dependent. The table to the right gives some commonly used thresholds for detecting MCA vasospasm based on the flow velocity detected. The Lindergaard ratio, which is the ratio of velocity of the MCA to the extracranial ICA, can be helpful in increasing the specificity of TCDs. Other diagnostic tools, such as angiography with DSA or CTA, can be helpful and are often used to either confirm symptomatic vasospasm based on an exam change or to screen for vasospasm in high-grade patients with unreliable exams. DSA is considered the most accurate, but CTA is commonly used as it's non-invasive and quicker, but it can overestimate vasospasm. CTE perfusion can be a helpful addition to CTA to increase the accuracy of diagnosis. Other diagnostic modalities are also used, the most common of which is EEG. It appears to have improved sensitivity compared to TCDs, but it's resource intensive. Brain oxygen monitoring and cerebral microdialysis are also used at some centers, but more data is needed before these technologies become more widely adopted. Treatment of suspected DCI is highly variable because limited, high-quality clinical trial data are available to help guide management. In general, management is split up into tiers. Most centers will use induced hypertension, either with norepinephrine or phenylephrine, as first-line therapy for suspected DCI. The details of this are much debated, including the threshold to initiate treatment, choice of vasopressor, and specific BP targets. Second-line options are generally different types of endovascular therapy, such as IA vasodilators or balloon angioplasty. Many other therapies are adjunctly used, such as intrathecal vasodilators, antiplatelet agents, and even invasive options such as intraaortic balloon pumps. In terms of prophylaxis, only nemotipine has been consistently demonstrated in RCTs as an agent to help reduce the incidence of DCI. This is despite its limited activity in improving angiographic vasospasm. This finding, along with other agents that showed significant activity in improving angiographic vasospasm, but limited activity in improving DCI, is what initially led to the paradigm shift away from large artery vasospasm being the sole cause of DCI. Dosing of nemotipine is 60 mg every 4 hours, but in the setting of large blood pressure variations, the dosing can be split up into 30 mg every 2 hours for a more consistent effect. The fact that many agents have been studied, but only nemotipine has demonstrated benefit in reducing this rate of DCI, underlines the importance of this medication in improving outcomes of patients with SAH. Intravascular volume status is another topic that's very important, but also variability managed across centers. There is a strong association between hypovolemia and DCI, and while this is the case, the historic practice of triple H therapy, hypovolemia, hypertension, and hemodilution as a whole is no longer recommended. Specifically, a hypovolemic strategy does not reduce the risk of DCI and leads to systemic complications such as pulmonary edema and acute kidney injury. Targeting uvolemia is recommended by AHA and NCS guidelines, although how this should be measured and achieved is not well studied, resulting in large practice variation. There are a few other management considerations to keep in mind when taking care of patients with SAH in the ICU. Fever is very common, found at some point in up to 70% of SAH patients, is more common in high grade patients, and often parallels the onset of DCI. The causes can be split up into infectious, inflammatory, or neurogenic. It can be very hard to distinguish between infectious and non-infectious causes, so having a systematic approach for guiding antibiotic decisions is key. Current guidelines recommend fever control in these patients, but RCTs are still needed to guide best practice. Hyponatremia is the most common electrolyte disturbance in patients with SAH and is typically caused by either SIADH or cerebral salt wasting. This diagnosis can be difficult because patients are commonly supplemented with sodium, treated with hypertonic saline for ICP issues, or given IV fluids for suspected DCI prior to or during the onset of these conditions. Labs can help, but volume status is the most important factor in separating these two conditions, with cerebral salt wasting being a hypovolemic state and SIADH being a uvolemic or hypervolemic state. Polyuria is also helpful and often seen in cerebral salt wasting. In cases of SIADH not associated with subarachnoid hemorrhage, fluid restriction is commonly used, but this risks precipitating DCI in subarachnoid hemorrhage, so that practice is not done in SAH-associated SIADH. There are several other considerations to remember when taking care of SAH patients. Patients frequently have venous thromboembolism, so it is important to start pharmacologic prophylaxis in a timely fashion. It's recommended to start this 24 hours after the aneurysm is secured. Similar to other critical illnesses, SAH patients can become hyperglycemic, and it's important to avoid both hypo- and hyperglycemia. Anemia is also of particular concern in SAH patients because of the potential compromised oxygen delivery to the brain as a contributor to DCI. The optimal target is unknown, but a large multi-center trial called Sahara is currently enrolling comparing a hemoglobin target of 8 versus 10 grams per deciliter, and this should help answer this question. I hope you enjoyed this presentation, and thank you for your attention. I'll be happy to answer any questions during our Q&A session.

subarachnoid hemorrhage

management

intracranial aneurysm

diagnosis

treatment

delayed cerebral ischemia