Aneurysmal Subarachnoid Hemorrhage

Summary

Aneurysmal subarachnoid hemorrhage is a neurological emergency that requires prompt patient stabilization, diagnosis, and treatment. While the classic presentation of a severe headache is common, there are a wide range of presentations including mild pain and coma. Treatment advances have moved towards early endovascular coiling of the ruptured aneurysm in most cases, based on two randomized clinical trials. Following securing of the aneurysm, there is evidence that patients are best served at centers employing multidisciplinary teams with specialized training in neurocritical care. The critical care management of the SAH patient is essential for treatment and mitigation of complications, in particular a syndrome of delayed neurological decline. Further study is essential for optimizing the acute care and improving outcomes in patients with aneurysmal SAH.

Introduction

Aneurysmal subarachnoid hemorrhage (SAH) is a neurological emergency that requires prompt diagnosis and management to prevent life-threatening rebleeding and optimize patient outcomes. A major challenge lies in the broad range of severity and in continuously tailoring treatments during the complex evolution of the disease. Care of this patient population continues to improve with advances in endovascular therapy and the introduction of dedicated neurocritical care teams. Here, we cover the essential approach to the patient with aneurysmal SAH, review current controversies, and discuss ongoing work aimed at improving outcomes in survivors.

Evaluation

The classic chief complaint of the patient presenting with a ruptured aneurysm is the sudden onset of the worse headache of their life or a thunderclap headache. While a severe headache is a common symptom, presentations can range from no headache to coma¹. Furthermore, most patients with a headache present to the emergency room, but a subset present to Primary Care. A heightened index of suspicion is important when triaging such patients. The emergency management of patients with acute headache, which has a broad differential diagnosis, is covered in a separate article, “Headache Emergencies,” in this issue of Neurologic Clinics. Other common symptoms of SAH include neck stiffness, photophobia, and vomiting. Transient loss of consciousness has also been described in patients with SAH related to a transient intracranial circulatory arrest². The range of presenting clinical severity is summarized in the Hunt and Hess (HH)³ and World Federation of Neurological Surgeons (WFNS) scales⁴, which are commonly used to communicate severity of a patient’s initial presentation (Table 1). The more general terms high- and low-grade, or poor- and good-grade, tend to assume cutoffs of 3 or higher in either severity score.

Table 1:

Common aneurysmal subarachnoid hemorrhage scales

Hunt and Hess
1	Asymptomatic or minimal headache and slight nuchal rigidity
2	Moderate-to-severe headache nuchal rigidity, no neurological deficit other than cranial nerve palsy
3	Drowsy, confusion, or mild focal deficit
4	Stupor, moderate-to-severe hemiparesis, possibly early decerebrate rigidity and vegetative disturbances
5	Deep coma, decerebrate rigidity, moribund appearance
WFNS
1	GCS 15, absent motor deficits
2	GCS 13/14, absent motor deficits
3	GCS 13/14, motor deficit present
4	GCS 7 to 12, present or absent motor deficits
5	GCS 3 to 6, present or absent motor deficits
Fisher
1	No detectable SAH
2	Diffuse SAH, no localized clot > 3mm thick or vertical layers > 1 mm thick
3	Presence of localized clots and/or vertical layers of blood 1 mm or greater in thickness
4	Intra parenchymal or intra ventricular hemorrhage with either absent or diffuse SAH
Claassen
0	No SAH or IVH
1	Minimal/thin SAH, no IVH in both lateral ventricles
2	Minimal/thin SAH, with IVH in both lateral ventricles
3	Thick SAH, no IVH in both lateral ventricles
4	Thick SAH, with IVH in both lateral ventricles
Modified Fisher
0	No SAH or IVH
1	Minimal/thin SAH, no IVH
2	Minimal/thin SAH, with IVH
3	Thick SAH, no IVH
4	Thick SAH, with IVH

Abbrev. GCS Glascow coma scale, SAH subarachnoid hemorrhage, IVH intraventricular hemorrhage

All patients with a suspected ruptured aneurysm should have a non-contrast head CT. Subarachnoid blood on head imaging should prompt vessel imaging, usually a CT angiogram. A ruptured aneurysm is more commonly is associated with blood within the basal cisterns (Figure 1A), as opposed to convexal sulcal subarachnoid blood. When the initial non-contrast head CT is negative for blood but there is still a clinical suspicion for a ruptured aneurysm, then the traditional approach has been to perform a lumbar puncture (LP)⁵. Frank blood and red blood cells (RBC) are commonly seen and are distinguished from a traumatic LP as RBC counts that do not diminish in sequential collecting tubes (Figure 1D–E). An additional classic finding is xanthochromia—due to RBC breakdown products—which is apparent in the cerebrospinal fluid (CSF) on sample collection or after the sample has been spun down in the laboratory.

Figure 1:

(A) Axial CT scan in a patient with ruptured anterior communicating aneurysm showing diffuse and thick subarachnoid hemorrhage seen in the basal cisterns (quadrigeminal cisterns, peri-mesencephalic cisterns and in the Sylvian fissures). Minimal intraventricular hemorrhage (white arrow) is also noted layering in the occipital horns of the lateral ventricles which are significantly dilated. (B) Axial CT scan of the same patient showing significant dilatation of the ventricular system consistent with communicating hydrocephalus. A ventricular drain was recently placed with its tip noted in the right lateral ventricle. (C) Follow-up CT scan after endovascular coiling of the ruptured anterior communicating aneurysm with resolution of the subarachnoid blood and significant improvement of the ventricular dilatation. Streak artifacts from the coils is noted. (D) Cerebrospinal fluid (CSF) in a patient with subarachnoid hemorrhage (SAH) showing the red blood cells in the final tube. (E) CSF in a patient without SAH with blood clearing in sequential tubes consistent with a traumatic lumbar puncture.

There have been recent proposals to change practice by foregoing an LP if there is a negative early head CT and certain criteria are met^6–9. A meta-analysis of a series of studies concluded that a modern head CT 6 hours after ictus can rule out SAH with 98.7% sensitivity in the setting of severe headache if there are no other neurological symptoms present and the scan is interpreted by an experienced radiologist¹⁰. In general, when the suspicion for aneurysm rupture remains high, despite an early negative head CT, most clinicians would still carry out an LP, assuming no contraindications. Under specific circumstances when the suspicion for an underlying aneurysm is low, a head CT within 6 hours to rule out a ruptured aneurysm may be reasonable.

When there is documented SAH—either by CT or lumbar puncture—initial vessel imaging should be followed by digital subtraction cerebral angiography (DSA) to increase the sensitivity of detecting an aneurysm that was not initially seen or to better characterize an aneurysm for endovascular coiling or surgical clipping^11–13. Other rarer pathologies such as dural arteriovenous fistula, arteriovenous malformation, and distal vasculopathy may also be better characterized on DSA compared to CT imaging.

Initial Management

As with most other neurological emergencies, application of advanced cardiovascular life support (ACLS) in the unstable patient with suspected SAH is the priority, including attention to the airway, breathing, and circulatory status. If the patient is progressing towards coma or there is impending respiratory failure, the patient should be intubated. If the patient is hypotensive, then measures to maintain an adequate blood pressure (BP) need to be instituted, including the use of vasopressors for a mean arterial pressure (MAP) ≥ 65 mm Hg. There are specialized disease-specific considerations (e.g. stress cardiomyopathy, discussed later), but work up of these potential complications should not delay the initial stabilization of the patient while the differential diagnosis remains broad.

Once the patient is stabilized, initial medical management and evaluation runs in parallel. If the pattern of blood on head CT is consistent with SAH (Figure 1A) then an upper systolic blood pressure (SBP) goal of <160 mm Hg is reasonable (Table 2). The SBP goal in the setting of SAH should be distinguished from 2 recent RCTs that suggested a potential benefit of <140 mm Hg in patients with confirmed primary ICH^14,15. SAH and ICH are not equivalent and, therefore, a SBP goal of <140 should not be assumed to be beneficial in SAH. There are no high-quality randomized studies of BP goals in SAH patients to guide recommendations. However, the potential for harm in over-correcting high SBP in this distinct population is not insignificant, especially in comatose patients when intracranial pressure (ICP) may be high and cerebral autoregulation dysregulated. In such patients, overtreatment of the blood pressure can lead to diffuse brain hypoperfusion and have significant negative consequences.

Table 2:

Recommended treatment parameters and orders

Prior to Aneurysm Securing
MAP	≥65 mm Hg
SBP	<160 mm Hg
ICP	<20 mm Hg
CPP	≥60 mm Hg
Oxygen saturation	>93%
EVD status	raised (e.g. 20 cm H2O) or closed, as tolerated
After Aneurysm Securing
MAP	≥65 mm Hg
SBP	up 220 mm Hg depending on clinical status
ICP	<20 mm Hg
CPP	≥60 mm Hg
Oxygen saturation	>93%
EVD status	controversial
TTE	On admission to establish baseline
TCD	Daily when available
Nimodipine	60 mg q4hr PO × 21 days
Fludrocortisone	0.1 mg one to three times per day for hyponatremia
Head CT	24–48 hours following aneurysm securing
Vessel or perfusion imaging	At 4–8 days post-SAH for high risk patients
Hemoglobin	> 7.0 g/dL, consider higher goal for DCI
Temperature management	≤ 37.5°C
Volume status	Euvolemia with isotonic fluids

Abbrev. MAP mean arterial pressure, SBP systolic blood pressure, ICP intracranial pressure, CPP cerebral perfusion pressure, EVD external ventricular drain, TCD transcranial doppler, CT computed tomography, CTA computed tomography angiogram, DSA digital subtraction angiography, TTE transthoracic echocardiography

The patient should be monitored for signs and symptoms of hydrocephalus or raised ICP that would warrant emergent placement of an external ventricular drain (EVD) (Figure 1B)¹⁶. There should be suspicion for symptomatic hydrocephalus in any high grade (HH grade 3 or above) patient, including those with progressive lethargy, limited vertical extra-ocular movements, stupor, hemiparesis, or decerebrate posturing without an alternative cause. If no one is available to emergently place an EVD, then the patient should be treated empirically with bridging osmotherapy and transferred to a higher level of care—ideally a Comprehensive Stroke Center¹⁷—where a qualified clinician can place an EVD. It is worth noting that some centers place an EVD in all patients with SAH for measurement of ICP, prophylaxis for possible hydrocephalus, and to remove blood from the CSF. In such cases, placement of an EVD would be non-emergent. Our recommendation is to place an EVD emergently if the patient is symptomatic or if there are clear signs of hydrocephalus on imaging.

A limited course of seizure prophylaxis is a reasonable approach in the setting of a potentially untreated, ruptured aneurysm since the sequela of seizures is of more potential harm than anticonvulsant side effects in the acute phase. Additionally, patients may present with a seizure-like episode at admission, and it would be reasonable to start an anticonvulsant for empiric seizure treatment until further history and work up is obtained. Levetiracetam has become a common first line agent given its better side effect profile and pharmacodynamics, compared to phenytoin, which has been associated with worse cognitive and neurologic outcome after SAH¹⁸. However, there are no high quality data regarding the use of anticonvulsants—either overall benefit or harm—or for the choice of any specific anticonvulsant over the other in the setting of SAH.

Pain management should take a step-wise approach, starting with non-sedating medications such as acetaminophen. NSAIDS are to be avoided given an increased risk of rebleeding prior to securing of the aneurysm. If pain is persistently severe then modest doses of opioids such as oxycodone or hydromorphone can be added, but given the importance of the clinical exam in the early phase post aneurysm rupture expectations should be set with the patient that complete pain freedom may not be the goal.

The patient with aneurysmal SAH should ideally be admitted to an accredited Comprehensive Stroke Center¹⁷ with a neuroendovascular team, neurosurgery, and an intensive care unit staffed with clinicians and nurses with specialized training in neurocritical care. This multidisciplinary approach allows for optimal aneurysm treatment and management of potential complications specific to aneurysmal SAH.

Aneurysm treatment

Digital subtraction angiography (DSA) should be performed in all patients who have a suspected aneurysm seen using other modalities (e.g. CT angiography or magnetic resonance angiography) or in patients in whom non-invasive vessel imaging is negative but there remains a clinical suspicion of an underlying vascular lesion¹⁹. DSA is important both for aneurysm treatment (Figure 2) and for open surgical planning. The diagnostic approach in patients with non-aneurysmal SAH or patients with an indeterminate diagnosis is variable and has been covered elsewhere¹³.

Figure 2:

Anteroposterior angiogram of the left internal carotid artery in a patient with subarachnoid hemorrhage showing a carotid terminus aneurysm (white arrow) (A). Balloon-assisted coiling of the carotid terminus aneurysm was performed (B) with complete occlusion of the aneurysm (C, black arrow).

The International Subarachnoid Aneurysm Trial was a landmark randomized trial that demonstrated better clinical outcomes at one year for patients with ruptured aneurysms treated with endovascular coiling compared to surgical clipping (190/801, 24% dependent or dead at 1 year in coiling arm vs 243/793, 31% in clipping arm, p=0.0019)^20,21. The Barrow Ruptured Aneurysm Trial was another randomized trial of SAH patients treated with alternating clip vs coil strategy. At 1 year, poor outcome was higher in the clip versus the coil group (33.7% s 23%; OR 1.68, p=0.02)²². Furthermore, there is evidence that suggests that coiling is associated with a lower rate of DCI²³. With advances in endovascular techniques and these recent randomized clinical trial data, most ruptured aneurysms are being treated by coiling. There are instances where open surgical clipping may be favored, such as a patient with a ruptured MCA aneurysm with large hematoma and mass effect requiring evacuation. Still, some surgeons prefer to secure the aneurysm by coiling in such instances followed by immediate craniectomy decompression²⁴. There are new endovascular techniques for aneurysm treatment involving a flow diverter device (resembles a stent with more metal surface area) and expandable intrasaccular flow disruption devices such as the WEB^25,26. However, flow diverter and stent assisted techniques are usually considered a last resort due to obligate dual antiplatelet therapy to prevent stent or flow diverter device thrombosis and risk of hemorrhage^25,26. Therefore, the patient should be evaluated by a multidisciplinary group experienced in endovascular approaches and/or open neurosurgical clipping to determine the optimal approach for each individual patient²⁴.

The timing of aneurysm treatment should be early following aneurysm diagnosis to reduce the risk of aneurysm re-rupture. The data on risk of aneurysm rebleeding after subarachnoid hemorrhage traces to studies from the 1970s and 1980s as most aneurysms are rapidly secured in the modern era^19,27,28. After SAH, the risk of re-rupture is highest in the first day (4%) and then is estimated at 1% to 2% each day in the first month²⁹. With conservative management, the risk of aneurysm rebleeding is 20% to 30% in the first month and then ≈3% per year. The mortality associated with aneurysm re-rupture is estimated at 67%³⁰. Ruptured aneurysms therefore need to be treated early to prevent re-rupture.

Ultra-early treatment is considered to be within the first 24 hours after presentation. The question of whether or not immediate emergent treatment—overnight when needed—is superior to waiting for daytime teams to arrive is controversial but could be considered in cases where there is already clinical suspicion or imaging evidence of aneurysm re-rupture^31,32. The current evidence supporting emergent versus within-24-hour treatment is inconclusive, but if there is an effect on rebleed rates, it is likely a very small effect³². Therefore, an ultra-early <24 hour treatment protocol is a reasonably rapid protocol when taking into account systems of care, resource utilization, and the benefit of proper surgical-endovascular planning.

If there is an unavoidable delay in aneurysm treatment beyond 24 hours, clinicians could consider a limited course of antifibrinolytics such as aminocaproic acid (Amicar) or tranexamic acid (TXA) to mitigate the risk for aneurysm rebleeding. In the era prior to early aneurysm treatment several decades ago, these agents were associated with myocardial infarction, pulmonary embolism, and other thrombotic complications. However, in the current period of early ruptured aneurysm treatment, the risk of complications up to 48–72 hours remain low to inexistent^33–35. One caveat that remains, however, is that antifibrinolytics can cause complications if the agent remains at a therapeutic level during endovascular treatment. Furthermore, a recent RCT found that ultra-early <24 hour use of TXA had no detectable effect on long term functional outcome³⁶. Therefore, we would reserve the use of antifibrinolytics only for patients in whom a delay of >24 hours is unavoidable and, when used, would hold the medication for at least 2 hours prior to an aneurysm securing procedure. It is critical to discuss utilization of antifibrinolytics with the physicians involved in securing the aneurysm (i.e. neurointerventionalist or neurosurgeon) as its effect on endovascular or open surgical procedural thrombotic complications is not well established and current clinical practice remains variable.

Finally, there are patients who present with aneurysmal pattern SAH yet the initial angiogram studies are negative. In these instances, the patient often undergoes repeat angiography in 1 week to look for an underlying lesion that may have been masked by hemorrhage at the initial study. When the repeat study is negative, the etiology is often referred to as a venous bleed, as in peri-mesencephalic SAH¹³. In instances where the clinical history points to the onset of pain in the neck, back, or there is a high burden of hemorrhage in the posterior fossa or foramen magnum, a search for spinal aneurysm may yield the etiology of the patient’s hemorrhage. Such findings are rare, thought to relate to dissection, and may resolve on their own prior to securing³⁷.

Critical Care Management

All patients with a suspected or confirmed ruptured aneurysm should be admitted to an ICU, preferably staffed with a clinical and nursing team with specialized training in neurocritical care. The only exception is at some experienced centers where good-grade SAH patients are admitted to an intermediate step down unit for lack of ICU beds. Studies have demonstrated that patient outcomes are improved when a multidisciplinary neurocritical care service is involved^38,39. The value of specialized care is underscored by the absence of firm evidence-based recommendations and the complexity of patient management based on time from aneurysm rupture, vasospasm management, and severity of disease. Therefore, there is no substitute for an experienced multidisciplinary neurocritical care team. What follows is an overview of general neurocritical care principles for this patient population to help guide management decisions.

Initial critical care considerations

Prior to securing of the aneurysm, initial management measures described above should be continued to prevent aneurysm re-rupture, including adequate BP control and anticonvulsants for seizure prophylaxis. If there is an EVD in place, the drain should be kept at a high level (e.g. 20 cm H₂O) or closed, as tolerated, to minimize the transmural pressure gradient across the aneurysm wall^40,41 (Table 2).

Once the aneurysm is secured, the focus of care shifts to preventing secondary brain injury, minimizing and treating complications. BP parameters can be liberalized based on medical co-morbidities (e.g. a lower upper BP limit in the setting of known heart failure). Anticonvulsant medications can be stopped if the clinical suspicion for seizures is low. Important general critical care practice applies to the patient with aneurysmal SAH such as ventilator liberation strategies, deep venous thromboembolism prophylaxis, infection control, and removing unnecessary lines and catheters on a daily basis. In the era of COVID-19, modified protocols tailored to the management of SAH have been reported to optimize SAH care and protect health care workers⁴². Additional COVID-19-specific recommendations are discussed in the article on Neurologic Emergencies during the COVID-19 pandemic.

The management of the SAH patient evolves with the time from aneurysm rupture. As such, we will address potential complications encountered in the ICU and their management strategies as they appear in an approximate chronological order.

Stress cardiomyopathy and neurogenic pulmonary edema

An adrenergic surge in the setting of severe SAH can lead to stress cardiomyopathy (SCM) which is an acute and reversible form of heart failure⁴³. The syndrome is also known as takotsubo cardiomyopathy for the characteristic shape of the left ventricle on catheter angiography or echocardiography which resembles a traditional Japanese octopus trap (tako たこ trans. octopus and tsubo つぼ trans. pot)⁴⁴. The treatment depends on whether or not there is an associated left ventricular outflow tract obstruction (LVOT)⁴⁵. In the majority of cases, there is no LVOT obstruction and the patient can be treated with vasopressors, inotropes, or in severe cases mechanical circulatory support. In the less common case of LVOT obstruction, a beta-agonist can cause a paradoxical worsening of shock and the approach shifts towards beta-blockers, pre-load optimization, and, when necessary, careful titration of alpha-agonists such as phenylephrine. The occurrence of SCM is important to consider as it influences the differential diagnosis for hypotension and shock in SAH patients. However, available diagnostics are non-specific and definitive diagnostics tend to be contraindicated. Therefore, clinical context (i.e. pretest probability) is key to the possible diagnosis of SCM. ECG and elevation of cardiac enzymes can be indistinguishable from ST-segment elevation myocardial infarction (STEMI). The diagnosis can be supported by the pattern of regional wall motion abnormalities based on TTE in which apical dilation and hypokinesis is more consistent with SCM and abnormalities in the distribution of a coronary artery could raise additional concern for a STEMI. The gold standard diagnostic procedure is catheter angiography and the exclusion of coronary artery disease. The incidence of SCM is high enough in SAH patients compared to the general population such that catheter angiography—and the accompanying use of periprocedural heparin—is typically not recommended for most patients, especially in the setting of a ruptured aneurysm. However, there can be concomitant coronary syndromes with SAH so a primary or concomitant coronary syndrome should not be completely excluded from the differential diagnosis.

Neurogenic pulmonary edema is a related hyperadrenergic syndrome which affects the lungs and should be on the differential diagnosis if there is unexplained early hypoxemia in a patient with severe, high-grade SAH. It can be thought of as a form of acute respiratory distress syndrome (ARDS) which usually resolves within 24–48 hours. There can be pulmonary edema exacerbated by concomitant heart failure due to SCM but neurogenic pulmonary edema is also seen in the absence of SCM. Treatment is supportive care with lung protective therapy according to current ARDS management strategies^46–48.

Hyponatremia

Hyponatremia (Na < 135 mmol/L) is common early in the hospital course of patients with SAH. The etiology has historically been controversial but the two leading causes are attributed to cerebral salt wasting (CSW)⁴⁹ and syndrome of inappropriate antidiuretic hormone secretion (SIADH)⁵⁰. The typical treatment of SIADH involves fluid restriction, which can be harmful in the patient with aneurysmal SAH as it may increase risk for DCI. The treatment for CSW involves fluid administration, which can conflict with SIADH. There are classic urine and blood study values that have been reported to favor one diagnosis over the other⁴⁹ but we have found that these studies are rarely revealing. As SIADH tends to occur in the setting of euvolemia or hypervolemia and CSW leads to hypovolemia, volume status may provide insight into etiology; however, volume status itself is notoriously challenging to determine in critically ill patients^51,52. It is also possible that patients have both SIADH and CSW although it is unclear how such a situation could be proven. Administration of hypertonic saline solutions (typically 1.5–3% sodium chloride IV) is the most effective treatment in maintaining eunatremia. Some clinicians will administer salt tablets (1–3 g 3 times a day), but this is not always effective and may not be well tolerated in awake patients. Fludrocortisone may have a role prophylactically or in response to hyponatremia⁵³ to minimize the use of continuous IV infusions (Table 2).

Fever

There are 2 common questions regarding fever following SAH: (1) what is the source of fever and (2) should fever be controlled? Following SAH both infectious and non-infectious fever is common. The etiology of non-infectious central or neurogenic fever in the setting of SAH is thought to involve hypothalamic irritation from blood products. Within the first 72 hours of admission the rate of non-infectious fever is higher in SAH⁵⁴, but in practice empiric antibiotics for possible infection is the usual course of action. Therefore, non-infectious fever remains a diagnosis of exclusion but should be higher on the differential in the SAH population to decrease the rate of unnecessary antibiotic use if an infectious source is ruled out.

Most neurointensivists would treat fever itself (>37.5°C) in SAH patients. Although there are no high quality data to support the approach, there is a strong rationale that reducing fever might minimize secondary brain injury. Association studies have found a relationship between fever and worse outcomes in SAH^55,56 but there is an absence of supportive randomized evidence⁵⁷. The reason for debate is that there is a hypothesized role for beneficial effects of fever in the setting of infection⁵⁸. There are also risks associated with implementing fever control measures such as intravascular or external cooling devices and their accompanying anti-shivering medications⁵⁹. What is likely is that the benefit of fever control probably depends on the specific patient population and circumstances that best balance risk and benefit⁶⁰. Therefore, empiric randomized trials are urgently needed. The ongoing INTREPID trial is a randomized study that takes the unique approach of prophylactically preventing fever in brain injured patients, including SAH patients (INTREPID, ClinicalTrials.gov Identifier NCT02996266). It is a step towards determining the impact of fever in SAH patients and whether treatment is beneficial. Additional adequately powered randomized trials to treat incident fever are also needed⁶¹. In the meantime, it is reasonable to treat fever empirically with antibiotics and antipyretic medications (such as acetaminophen) and then escalate to cooling devices for persistent fevers on a case-by-case basis.

Delayed cerebral ischemia

The mainstay of critical care management is the prevention and treatment of a syndrome of delayed neurological deficits that typically occurs 3 to 14 days after aneurysm rupture⁶² with the peak on days 7 to 9 post bleed. The syndrome is variably termed delayed cerebral ischemia (DCI)⁶³, delayed ischemic neurological deficits⁶⁴, delayed cerebral infarction⁶⁵, symptomatic vasospasm⁶⁶, or vasospasm⁶⁷. Historically, the cause of delayed neurological deficits was thought to be related to large vessel arterial vasospasm (Figure 3). This led to the practice of fluid administration, permissive or induced hypertension, and administration of vasodilators to increase blood flow. One vasodilator, the calcium channel antagonist, nimodipine, was designed to dilate cerebral blood vessels. The use of nimodipine has the strongest supportive evidence of any DCI therapy. A pair of randomized trials in the 1980s showed that it reduced the rate of poor outcome or delayed infarcts by as much as 50% when used prophylactically^68,69. However, nimodipine does not appear to exert its effect solely through dilation of large cerebral arteries given early observations that outcomes could be improved without improvement in angiographic vasospasm and, conversely, that there could be poor outcome even with reversal of vasospasm. Regardless, nimodipine 60 mg every 4 hours used as prophylaxis remains standard of care for most patients with aneurysmal SAH (Table 2). A notable exception exists in Japan where standard of care is use of the rho kinase inhibitor fasudil, rather than nimodipine, to prevent DCI. Finally, patients with symptomatic vasospasm refractory to medical management are commonly referred for endovascular therapy with balloon angioplasty or intra-arterial vasodilators.^70–73.

Figure 3:

Anteroposterior angiogram of the left internal carotid artery in a patient with aneurysmal subarachnoid hemorrhage at presentation (A) and 8 days later (B) showing interval development of severe vasospasm of the left supra-clinoid carotid artery, left anterior and middle cerebral arteries (white arrows).

Monitoring and DCI Detection.

It is common practice to obtain a head CT 24–48 hours following aneurysm securing to distinguish treatment-related infarction from DCI later in the patient’s hospital course⁷⁴. Furthermore, many centers will use regular monitoring with transcranial Dopplers (TCD) and interval vessel or perfusion imaging 4 to 8 days post-SAH to detect early signs of vasospasm (see Table 2). In high grade patients who have a poor exam and are not following commands, these radiographic modalities may be the only triggers to raise a clinical suspicion for DCI. In contrast, for low-grade patients or those with a good examination some clinicians would argue that such monitoring is not necessary. Perfusion imaging may have an advantage over other radiographic modalities in that it can distinguish blood flow (the metric of interest) from velocity and vessel caliber⁷⁵. Therefore, a patient with decreased flow on perfusion imaging but unremarkable large vessel imaging and TCDs might still be at high risk for impending DCI. Conversely, a patient with a reassuring blood flow perfusion scan but large vessel vasospasm and elevated TCD velocities may not warrant escalation in their management. However, in practice there are challenges to implementing perfusion imaging in the SAH patient population and persistent questions about the appropriate timing, modality, and interpretation of the scans⁷⁶. Non-invasive CBF monitors may address some of these issues but the technology remains in development⁷⁷. Continuous EEG (cEEG) is another potential modality that may find use in detecting impending DCI. While promising, the cEEG approach is resource intensive and requires a reading epileptologist who is specially trained in grading records for DCI risk. Finally, invasive multimodality monitoring using brain tissue oxygenation, microdialysis, and thermal diffusion flow measurement is a promising but complex approach that requires additional development for more widespread use⁷⁸. Regardless of the diagnostics used at any specific institution, the suspicion and triggers for care escalation for DCI should be driven by the patient’s clinical exam. Once DCI is suspected, the following measures can be attempted, all with limited supporting evidence.

Hypertension.

Strategies to prevent and treat DCI include BP, volume, and hemoglobin optimization. The historical “triple-H” treatment (hypertension, hypervolemia, and hemodilution)⁷⁹ approach has been outdated for some time but is worth discussing as a conceptual framework. Of these only permissive or induced hypertension remains in practice, and even this practice has recently been questioned⁸⁰. The rationale behind elevating blood pressure in the SAH patient is to improve blood flow in the setting of altered cerebrovascular autoregulation that can occur in the setting of brain injury. In normal, uninjured brain, cerebral blood flow (CBF) should not vary with changes in mean arterial pressure (MAP) within a range of approximately 60–150 mm Hg⁸¹. However, this autoregulation is thought to break down in the setting of acute brain injury, including SAH. Therefore, it may be necessary to modulate CBF through changes in BP. A common practice is to attempt a trial of induced hypertension with vasopressors if there is a clinical neurological change. A more controversial practice is inducing hypertension for TCD or imaging findings that demonstrate vasospasm as a prophylactic measure in the absence of symptoms. Typically, BP is induced to approximately 20% above baseline BP with MAP or systolic blood pressure (SBP) goals, with an upper SBP limit of 220 mm Hg (Table 2). The choice of MAP or SBP is not standardized and there is limited high quality evidence supporting specific targets or protocols. A recent international multicenter trial attempted to determine the benefit of a protocolized induced hypertension approach in patients with clinical symptoms of DCI but was not able to reach their enrollment goal of 240 patients, did not find a benefit to prophylactic hypertension in their analyzed patients, and found a higher rate of serious adverse effects in the induced hypertension group which was validated in a systematic literature review⁸⁰. Therefore, there is a disconnect between the practice of induced hypertension and the current level of evidence.

Cardiac Output.

Cardiac output enhancement is related to induced hypertension, but is a distinct and independent strategy to increase CBF to prevent or treat DCI. Titrating supranormal cardiac output with inotropes, such as dobutamine, and a pulmonary artery catheter used to be practiced but has now largely fallen out of favor. Trials have thus far failed to show benefit of such therapy and many clinicians are wary of adverse effects such as cardiac arrhythmia and central venous line complications related to dobutamine titration with a PA catheter. Various device-based strategies have been designed to enhance CBF through increasing cardiac output (e.g. aortic balloon pumps) or diverting flow to the brain by occluding flow within the descending aorta⁸². However, these devices have a large potential for morbidity with no clear demonstration of benefit. Therefore, we would not recommend routine guided pharmacologic or device-based cardiac output therapy given an unfavorable risk-benefit ratio in most patients. Such therapy may be considered for a poor-grade patient who has failed more conservative approaches.

Vasodilators.

One promising, less-invasive approach that may act through a combination of enhanced cardiac output and cerebral vasodilation is the administration of IV milrinone⁸³. As a phosphodiesterase inhibitor, IV milrinone has inotropic properties and causes peripheral and cerebral vasodilation. It is also used by angiographers intraarterially as a cerebral vasodilator. A protocol developed by clinicians at the Montreal Neurological Institute has proven safe and suggests favorable outcomes in a single center retrospective cohort without the need for invasive hemodynamic monitoring^83,84. This approach has been adapted by other centers, again with a good safety profile, but further study is likely required to meet the evidence threshold to incorporate or reject widespread milrinone therapy in clinical practice⁸⁵.

Volume.

A longstanding cornerstone in prevention and management of DCI is volume optimization⁸⁶. The current practice is for goal euvolemia. Hypervolemia is no longer favored and likely leads to increased complications and increased length of stay. A single center randomized study comparing euvolemia versus hypervolemia demonstrated an effect on cardiac filling pressures but did not find a difference in CBF or blood volumes⁸⁷. Hypovolemia could lead to low cardiac output, hypotension, and potential precipitation or exacerbation of DCI. Therefore, current practice focuses on avoiding hypovolemia and maintaining euvolemia through clinical assessment of volume status and administration of PO and isotonic IV fluids as needed.

Hemoglobin.

The optimal hemoglobin concentration in SAH patients remains unclear although there is agreement that the hemoglobin concentration should be at least 7.0 g/dL. Patients with SAH tend to develop a relative anemia which is thought to be due to hospitalization-associated anemia and fluid administration. One prospective study transfused 1 unit of RBCs in patients with a hemoglobin concentration <11 g/dL and found improvements in oxygen delivery—as measured with PET imaging—that were superior to induced hypertension or an equivalent fluid bolus⁸⁸. On the other hand, blood transfusions have known complications in general (e.g. greater infection rate, acute lung injury, transfusion reactions) and in SAH patients have been found to have a greater risk of thrombotic complications⁸⁹. The best hemoglobin goal is unknown and requires further study. In the meantime, when there is a concern for active cerebral ischemia it is reasonable to aim for a transfusion hemoglobin goal of greater than 8.0 g/dL or even higher depending on the clinical scenario⁹⁰.

Hemodilution and Viscosity.

Hemodilution (from classic triple-H therapy) should not be a goal. A related strategy is attempting to improve blood viscosity through a continuous low dose infusion of mannitol. The rationale is to improve CBF by making it easier for RBCs to transit the capillary bed^91,92. There is evidence that a continuous low dose mannitol infusion can improve CBF in SAH patients independent of its effect on ICP and without causing a decrease in blood pressure⁹³. Mannitol in SAH patients was more common in the past but has largely fallen out of clinical practice, perhaps out of a concern for its diuretic effects. The historical papers demonstrate safety^91–93, however, and the effect of continuous mannitol after SAH may warrant further study.

Future Work.

DCI remains an important problem with a sizeable proportion of patients that could benefit from new therapies to prevent and fully treat the disease. The focus of these therapies should focus on long term functional outcome, minimizing procedures and complications, and decreasing length of stay so that all patients with ruptured aneurysms might benefit from new approaches.

There have been several additional investigational therapies that target the prevention and treatment of DCI^82,94,95. One of these promising therapies involves use of the phosphodiesterase inhibitor cilostazol, which exerts its effects through cerebral vasodilation and inhibition of platelet aggregation⁹⁶. In a single center randomized trial, cilostazol administration was associated with less symptomatic vasospasm compared to placebo⁹⁷. A subsequent single center randomized trial from an independent group found that cilostazol use was associated with fewer sequelae of spreading depolarizations⁷³, a phenomenon thought to be part of the pathophysiological pathway leading to secondary brain injury in SAH^98–101. These studies were conducted in Japan where fasudil, rather than nimodipine, is used for DCI prophylaxis. Therefore, they are not generalizable and additional studies are required.

Another promising approach involves targeting arterial vasospasm. SAH leads to elevated levels of the potent endogenous vasoconstrictor endothelin 1 (ET-1). A phase III randomized trial of the ET-1 receptor antagonist clazosentan demonstrated a reduced rate of rescue therapy, but did not meet the threshold for statistically significant improvement in the primary composite outcome of all-cause mortality, rescue therapy, DCI-related cerebral infarcts, and neurological deficits in patients undergoing surgical clipping¹⁰². As a result, a related phase III randomized trial in patients undergoing endovascular coiling of their aneurysm was stopped early¹⁰³. Regardless, the study in coiled patients demonstrated that at higher doses, clazosentan led to a decreased rate in the primary composite outcome. Neither study met their predefined secondary outcome of functional improvement, possibly due to underpowering due to better-than-expected outcomes in placebo groups. A major caveat is that clazosentan led to an increased rate of pulmonary edema. An ongoing follow up study called REACT (ClinicalTrials.gov Identifier NCT03585270) will determine if clazosentan can decrease the incidence of clinical deterioration due to DCI in high risk patients and to address remaining concerns.

Finally, work towards predicting which patients develop DCI is ongoing. Publication of the original Fisher scale was an important step towards using an early head CT to predict the risk of vasospasm and poor outcome¹⁰⁴. However, it was designed as preliminary scale and does not take into account the additive effect that intraventricular hemorrhage may have on risk for DCI (Table 1). The modified Fisher (mF) scale was developed as an ordinal radiological scale that built on Fisher’s work and incorporates the amount of SAH and presence of intraventricular hemorrhage (IVH) in any ventricle. An increase in the scale from 0–4 is associated with an approximately step-wise increased risk of developing DCI¹⁰⁵. The Claassen scale is very similar to the mF scale but requires IVH in both lateral ventricles rather that blood in any ventricle¹⁰⁶; given this specificity for blood in both lateral ventricles it is not as widely used as the mF scale. There has been increasing acknowledgement of issues with interrater reliability with the different Fisher scales¹⁰⁷. For research purposes, it is important that raters are using the same scoring criteria and that the different versions of the scales are not conflated with each other. In clinical practice, the modified Fisher (mF) and Claassen scales are the most helpful and commonly used scales to communicate the amount and distribution of intracranial blood, when used consistently within an institution. Aside from the Fisher scales, there are other potential predictors of DCI which are being explored. The Hijdra scale is another CT-based scale that quantifies the amount of blood in several basal cisterns in greater detail that the Fisher scales¹⁰⁸. Investigators have reported increased sensitivity in early prediction and detection of DCI or secondary brain injury by supplementing clinical and CT data with TCD¹⁰⁹, continuous EEG^110,111, inflammatory biomarkers^112,113, and common lab values¹¹⁴. Future work will determine how such approaches can influence interventions and impact outcomes.

Seizures

Seizures should remain on the differential diagnosis for hospitalized SAH patients in a coma or more subtle changes in mental status. The only way to confirm a suspicion of non-convulsive seizures is through continuous EEG (cEEG). Convulsive seizures can be treated empirically but should also be followed up with cEEG to capture subsequent convulsive or non-convulsive events. The duration of monitoring remains controversial, but a good rule of thumb is 24 hours in awake patients and at least 48 hours in comatose patients¹¹⁵. The exception may be in patients in whom there is not a high clinical suspicion of seizures and who have an absence of epileptiform abnormalities on cEEG in the first 4 hours of recording¹¹⁶.

If seizures are detected, clinically or electrographically, then they should be treated according to local practice, which varies considerably. There is no evidence that seizure control in SAH leads to improved outcomes; however, given the likelihood that seizures on top of SAH could lead to even worse neurological injury, we advocate for seizure detection and control measures in the ICU. Following control of confirmed seizures, they recommend maintenance or careful slow down-titration of anticonvulsants, to reduce the risk of seizure recurrence in the resolution phase of the disease.

It is worth commenting on whether anticonvulsants should be continued in the common instance of isolated seizure-like episodes on presentation. Non-epileptic convulsions in the setting of syncope are very common. One study found no association between seizure-like episodes and the risk of developing subsequent seizures¹¹⁷. Therefore, if there are no concerning features of presenting abnormal movements (e.g. development of status epilepticus), it is reasonable to defer anticonvulsants in patients where there is only an isolated seizure-like episodes on presentation.

Hydrocephalus and EVD management

Hydrocephalus is a clear indication for EVD placement and CSF diversion. A controversial area of post-aneurysm treatment involves how to manage the EVD once it is in place⁴¹. The issue is important as it can impact ICU length of stay and is thought to influence patient outcomes. There are prospective randomized and retrospective studies that have suggested that keeping the EVD closed and only opening intermittently when needed can prevent EVD-related complications, specifically reducing the rate of drain malfunction and ventriculostomy-associated infection^118,119. However, surveys of actual practice patterns have revealed that most institutions take the opposite approach of keeping the EVD open and draining continuously by default once the aneurysm is treated^120,121. The rationale is that draining CSF might help to clear red blood cells and breakdown products that are thought to contribute to neuroinflammation. However, if the goal is to drain blood products from the CSF, then lumbar drainage is likely more effective and leads to fewer complications than EVD drainage¹²². Regardless, there are currently no high quality data to suggest that CSF drainage either improves or worsens functional outcomes^41,123.

Another controversial aspect of EVD management is how to discontinue the drain once the clinical decision has been made that it is no longer needed. The choice in discontinuation strategies is typically between what is referred to as a gradual wean vs. rapid wean. The most common approach is a gradual wean whereby the drain is raised in steps every 24 hours and then clamped (closed), usually over the course of 3–4 days⁸². A typical gradual wean scenario is a patient with an EVD open and continuously draining at 10 cm H₂O above the tragus of the ear. When the decision is made to attempt drain discontinuation, the drain is raised to 15 cm H₂O for 24 hours, then 20 cm H₂O, and then closed. A wean failure is typically defined as the patient developing symptoms (e.g. headache or drowsiness), an elevated ICP, CSF leakage from around the EVD site, or signs of radiographic hydrocephalus at any point during the process. Once the EVD has been closed for 24–48 hours and repeat imaging does not show radiographic hydrocephalus then the EVD is discontinued. In contrast, a typical rapid wean consists of the EVD being closed for 24–48 hours regardless of the starting height of the drain. A single center randomized trial of rapid vs. gradual EVD weaning found that a rapid wean led to shorter ICU length of stay without a detectable effect on the rate of vasospasm¹²⁴. A more recent single center before-and-after retrospective study also found that a rapid wean was associated with shorter ICU length of stay and additionally found that patients undergoing a rapid EVD wean had a lower rate of permanent CSF diversion in the form of a ventriculoperitoneal shunt¹¹⁹. A prospective multicenter study would be helpful to determine the generalizability of the rapid vs. gradual wean findings.

Key Points.

• Aneurysmal SAH is a neurological emergency ideally treated in a multidisciplinary Comprehensive Stroke Center.

• Initial management should focus on principles of advanced cardiovascular life support and prompt diagnosis.

• Early complications are aneurysm rebleeding and acute hydrocephalus.

• Most patients have better outcomes after endovascular coiling of the aneurysm although there are instances when open surgical clipping may be indicated.

• After the aneurysm is treated, survivors require additional monitoring to prevent and manage complications, including a syndrome of delayed neurological decline known as delayed cerebral ischemia or symptomatic vasospasm.

Clinics Care Points.

Pre-aneurysm securing

After the initial airway, breathing, and circulation assessments, the patient should be evaluated for symptomatic hydrocephalus or a clinical presentation concerning for elevated intracranial pressure (i.e. high grade SAH) and considered for an external ventricular drain.

The patient should be evaluated by cerebral angiography for consideration for endovascular coiling vs. open surgical clipping to secure the culprit aneurysm.

When there is equipoise between securing the aneurysm with coiling vs clipping, more patients have better outcomes at one year after coiling compared to clipping.

Post-aneurysm securing

The patient should be monitored and treated for potential medical complications of subarachnoid hemorrhage.

Delayed neurological decline is a common occurrence after aneurysmal subarachnoid hemorrhage, typically arising day 5 to 12 after SAH. If clinically evident, this is termed delayed cerebral ischemia or symptomatic vasospasm.

The tenets of medical therapy for the prevention and treatment of symptomatic cerebral vasospasm, or delayed cerebral ischemia, are nimodipine administration and euvolemia.

Patients with persistent symptomatic vasospasm refractory to medical management should be considered for endovascular therapy with intra-arterial vasodilators or balloon angioplasty. This can be bridged with intravenous vasodilator infusion in the neurointensive care unit.

Synopsis.

Aneurysmal SAH is a neurological emergency which requires immediate patient stabilization and prompt diagnosis and treatment. Early measures should focus on principles of advanced cardiovascular life support. The aneurysm should ideally be evaluated and treated in a Comprehensive Stroke Center by a multidisciplinary team capable of endovascular and operative approaches. Once the aneurysm is secured, the patient is best managed by a dedicated neurocritical care service to prevent and manage complications, including a syndrome of delayed neurological decline. The goal of such specialized care, and ongoing research, is to prevent secondary injury, reduce length of stay, and ultimately to improve outcomes for survivors of the disease.