Aneurysmal SAH is a severe disease, and the post-haemorrhage period fraught with potential complications that must be recognized and treated early for favourable outcome. While diagnosis of SAH is often straightforward from clinical history and initial CT, some patients will require cerebrospinal fluid evaluation. The aneurysm must be secured urgently to reduce rerupture and clinical worsening. Endovascular coiling is preferable when feasible, but surgical clipping is sometimes needed based on patient or aneurysmal characteristics, or presence of intraparenchymal haemorrhage requiring evacuation. Treatment of symptomatic hydrocephalus with CSF diversion is also crucial. Patients with aneurysmal SAH should be managed by a team of nurses and physicians with neurocritical care, neuroendovascular, and neurosurgical expertise, preferably in a dedicated neurosciences intensive care unit. Early complications include aneurysmal rebleeding, hydrocephalus, and neurogenic cardiopulmonary injury. In the subacute phase, delayed cerebral ischaemia and hyponatremia are more commonly seen. With optimal multidisciplinary management, many patients can return to their previous level of function only weeks after the aneurysm rupture. Still, most treatments in SAH are based on insufficient evidence, and more collaborative research from the bench to the bedside is necessary to continue improving patient outcomes.
In patients with subarachnoid haemorrhage (SAH), the initial presentation to the Emergency Department can be very dramatic, with a thunderclap headache (i.e. maximal at onset) followed by sudden unresponsiveness or seizure. In other patients, the presentation may be milder, and clinicians must remain vigilant to recognize subtle clinical presentations. With advances in the quality of computed tomography (CT) scanning, the great majority of cases of aneurysmal SAH are easily identified with this modality. In cases with a clinical history that suggests aneurysmal SAH but negative CT scans, further diagnostic studies may be necessary to determine whether a haemorrhage has occurred.
Following the diagnosis of aneurysmal SAH, the patient must be carefully monitored for multiple early and late complications, and the aneurysm must be definitively secured to prevent re-rupture and clinical worsening.
Over time, the outcomes of patients with SAH have improved, with an overall decrease in mortality (Nieuwkamp et al., 2009). Predictors of good functional outcome in SAH patients include low World Federation of Neurological Surgeons (WFNS) grade, absence of intraparenchymal haemorrhage, absence of delayed infarction, and no requirement for blood transfusion (Pegoli et al., 2015). High-grade initial WFNS score, increased circulating neutrophils, poor haemodynamic status, significant radiographic neurological injury, systemic inflammatory response syndrome, and cardiac dysfunction have been shown to be predictive of a poor clinical outcome (van der Bilt et al., 2009; Tam et al., 2010; Ibrahim et al., 2014). In patients who survive the protracted clinical course and complications associated with aneurysmal SAH, many have good functional outcomes, but cognitive impairment may persist beyond the subacute clinical period (Raya and Diringer, 2014).
A multicentre cohort study evaluating all patients who presented to an Emergency Department with a thunderclap headache (defined as headache peaking within 1 hour of onset) found that 7.7% (240/3132) had a final diagnosis of SAH. CT of the head performed within 6 hours of symptom onset was 100% sensitive and specific for SAH. This required image interpretation by an experienced radiologist, as four false negative readings by physicians in other specialties and trainees were later corrected. Beyond 6 hours, CT performance decreased, with sensitivity 86%, specificity 100% (Perry et al., 2011).
Magnetic resonance imaging (MRI) is more sensitive than CT for the detection of SAH. In one study evaluating detection of SAH present in 146 subarachnoid regions in 25 patients, CT detected 75% of involved regions. In contrast, fluid-attenuated inversion recovery (FLAIR) MRI detected 87%, susceptibility-weighted MRI detected 88%, and combined FLAIR and susceptibility-weighted imaging (SWI) detected 100% (Verma et al., 2013).
When a lumbar puncture is performed and blood is found in the cerebrospinal fluid (CSF), careful analysis of the CSF is needed to differentiate between traumatic tap and aneurysmal haemorrhage as the source. In a multicentre, observational study evaluating CSF findings in alert patients presenting with headache and in whom lumbar puncture was performed after negative CT scans, or in advance of CT imaging at the treating physician’s discretion, 36.9% (641/1739) had a bloody tap, but only 0.9% (15/1739) had a final diagnosis of SAH. The combination of <2000 × 106 red blood cells/L and absence of xanthochromia excluded SAH (negative predictive value 100%: 95% CI: 99.2–100%), while presence of either or both of these features increased the likelihood of SAH (positive predictive value 21.4%, 95% CI: 12.9–33.2%) (Perry et al., 2015).
The usefulness of xanthochromia alone as a CSF finding in CT-negative patients has varied in different studies. In one single-centre study, patients presenting within 14 days of onset of a thunderclap headache with negative CT scan underwent subsequent CSF analysis. All patients who had evidence of xanthochromia underwent 4-vessel catheter angiography to evaluate for evidence of aneurysm. Among 152 patients, 12% had xanthochromia on visual CSF inspection; and of those, 72% were diagnosed with aneurysmal SAH. Accordingly, for presence of aneurysmal SAH, xanthochromia in CT-negative patients had sensitivity of 93% and specificity of 95% (Dupont et al., 2008). However, in the Ottawa multicentre study of 1739 patients presenting within 14 days of headache onset, xanthochromia had lower sensitivity (47%), though higher specificity (99%) (Perry et al., 2015).
Studies have sought to identify prediction rules that would indicate patients presenting with headache who do not require CT imaging in order to reduce radiation exposure and care costs. However, given the importance of not missing an underlying aneurysm, imaging selection scales are weighted to ensure 100% sensitivity, resulting in relatively low specificity. For example, in a multicentre cohort of 2131 patients, a clinical decision rule, the Ottawa SAH Rule, was derived, which indicated imaging for any patients with age ≥40 years, witnessed loss of consciousness or onset during exertion, neck pain or stiffness, thunderclap onset (pain peaking at onset of headache), or limited neck flexion to predict the presence of SAH. This rule had 100% sensitivity but only 15.3% specificity for the recognition of SAH (Perry et al., 2013). In a validation cohort, this rule had retrospectively retained sensitivity of 100%, and specificity dropped further to 7.6%.
The data suggest that in patients presenting with acute onset, rapidly peaking headache where clinical suspicion for SAH is high, the initial evaluation should include a non-contrast CT scan of the head. If the CT scan is unrevealing, MRI of the head with gradient echo or SWI could be considered; however, a negative result still requires spinal fluid evaluation to investigate for SAH. Lumbar puncture should evaluate for red blood cells and xanthochromia. As xanthochromia can take hours to develop, ensuring several hours (e.g. 4–6 h) have passed since headache onset before evaluating the spinal fluid may be reasonable. Once an SAH is identified, noninvasive and catheter digital subtraction angiography is necessary for identification and possible management of an identified symptomatic aneurysm.
There is very limited research on best analgesic regimens for headache in patients with acute SAH. A small retrospective analysis of 53 patients who received gabapentin after SAH found few adverse effects; however, the efficacy was uncertain, as these patients were also receiving other analgesic medications (Dhakal et al., 2015).
In patients with SAH, sedating medications that may confound serial neurological examinations should be avoided. Opioid medications have adverse effects, including sedation, constipation, and decrease in central respiratory drive, and, in SAH patients, the latter can cause a rise in carbon dioxide with resulting cerebral vasodilation and increased intracranial pressure. Therefore, opioid medications should be avoided if possible. As bleeding risks are an especial concern in patients with SAH, non-steroidal anti-inflammatory agents, which adversely affect platelet aggregation, are typically avoided. For those who have suffered a seizure, analgesic medications such as tramadol, which can further lower the seizure threshold, should be used with caution.
In general, for SAH patients, acetaminophen and judicious use of tramadol and low-dose codeine may be helpful in managing pain without confounding a neurological examination or causing significant risk of increasing intracranial pressure.
Valsalva manoeuvres, as during coughing or straining at urinary or bowel movements, produce a brief surge in arterial blood pressure that may increase risk of re-rupture in unsecured aneurysms. In a study of 456 patients whose activities at onset of aneurysmal rupture were known, 13% were engaged in micturition or defaecation (Matsuda et al., 2007).
It is reasonable to take measures to prevent Valsalva manoeuvres in patients in whom a ruptured aneurysm is not yet secured. Stool softeners should be prescribed routinely. Patients are placed at bed rest and instructed to avoid lifting objects. Antiemetics should be used promptly if nausea develops.
Elevated blood pressure is a risk factor for early rebleeding in the first 24 hours after initial aneurysmal rupture. In a study of 273 SAH patients, 14% had rebleeding in the ambulance or at the referring hospital before admission arrival to a comprehensive centre. Elevated systolic blood pressure (SBP) (>160 mm Hg) was associated with rebleeding, odds ratio (OR) 3.1 (95% confidence interval [CI]: 1.5–6.8) (Ohkuma et al., 2001).
Whether active blood pressure control reduces rebleeding has not been studied in a randomized trial in the modern management era. In the American Cooperative Study, conducted between 1963 and 1970, 1005 patients with ruptured aneurysms were randomized between four treatment modalities: one arm consisted of drug-induced lowering of the blood pressure, another of bed rest alone, and the other two arms were surgical: carotid ligation and intracranial surgery. In the intention-to-treat analysis, antihypertensive drugs compared with bed rest alone failed to reduce the rate of rebleeding (Torner et al., 1981). However, this study was performed in the pre-CT era, so diagnosis of rebleeding may not have been fully accurate, and the duration of delay before securing the aneurysm was prolonged.
Some nonrandomized studies have tested how well particular continuous intravenous antihypertensive regimens can attain target blood pressures in acute SAH patients awaiting early surgical or endovascular aneurysm treatment. For example, a single arm study analysed 28 acute SAH patients prior to aneurysm securement who had not achieved goal blood pressures with intravenous labetalol or hydralazine. Addition of intravenous nicardipine was effective at lowering blood pressure, decreasing mean SBP from 177 mm Hg at admission and 156 mm Hg with initial agents to 143 mm Hg. Failure to achieve target blood pressure occurred in 4%, hypotension occurred in 11%, and there were no episodes of rebleeding (Varelas et al., 2010).
In the acute setting, prior to intervention on a ruptured intracranial aneurysm, it is reasonable to control blood pressure to <160 mm Hg to deter early re-rupture. This can be accomplished with small doses of labetalol or hydralazine. In patients with refractory hypertension, continuous infusion of calcium channel blocking agents such as nicardipine or clevidipine can be a reasonable option. Despite theoretical concerns that sodium nitroprusside could increase intracranial pressure due to its potent vasodilatory effect, this effect is rare in practice. Sodium nitroprusside can be useful for the most recalcitrant cases of acute hypertension, though it is prudent to reserve its use for such cases.
After the aneurysm is secured, hypertension should generally not be treated in the absence of signs of heart failure or extremely high pressures reaching the range that could produce hypertensive encephalopathy. Permissive hypertension is especially advisable in patients deemed to be at high risk of delayed cerebral ischaemia (DCI).
Hypotonic solutions (including lactated Ringer’s solution) should be avoided in patients with SAH to avoid worsening of brain oedema. Electrolyte imbalances are not uncommon upon presentation. Particular attention should be paid to hypomagnesaemia, which is prevalent and can increase the risk of cardiac arrhythmias. Avoiding intravascular volume depletion should be a priority, especially within the period of maximal risk of DCI (days 3–10), but inducing prophylactic hypervolaemia is not beneficial and can in fact be detrimental (Lennihan et al., 2000; Martini et al., 2012; Kissoon et al., 2015).
For all patients hospitalized with an aneurysmal SAH, careful fluid monitoring requires placement of a urinary catheter. Patients should be hydrated with isotonic or hypertonic solutions. Intravascular volume depletion should be avoided, but without trying to induce hypervolaemia. Additionally, regular monitoring of electrolytes, particularly sodium, allows for early recognition of delayed complications, namely, cerebral salt wasting (CSW) and the syndrome of inappropriate antidiuretic hormone (SIADH).
Patients who present with SAH are at risk for multiple medical complications, and hyponatraemia is one of the most common. Prompt recognition and treatment are necessary to prevent additional complications.
The cause of hyponatraemia in SAH patients had previously been attributed to SIADH (Doczi et al., 1981). However, a prospective evaluation of 21 consecutive patients with SAH found that hyponatraemia developed in 9 of 21 patients between post-bleed days 3 and 9, and evaluation of fluid status in those patients identified a decreased plasma volume with a preceding negative sodium balance, indicative of CSW (Wijdicks et al., 1985).
Subsequently, a prospective study of 208 SAH patients identified an incidence of hyponatraemia of 34%, and of these patients, 24% developed DCI compared with 12% of SAH patients who did not have hyponatraemia (Hasan et al., 1990).
A systematic review completed in 2011 identified seven articles evaluating management options for patients with hyponatraemia and SAH, due to either SIADH or CSW. The authors concluded that maintaining a normal volume status and correcting sodium to normal levels was beneficial. Early treatment with mineralocorticoids improved excess sodium loss and serum sodium levels. Most studies evaluated either fludrocortisone or hydrocortisone, with a better side effect profile being seen in patients receiving fludrocortisone. While the use of hypertonic saline improved laboratory values, the clinical benefit was uncertain and, likewise, use of albumin had an unclear clinical benefit. The use of vasopressin antagonists could not be recommended in SAH patients, as these medications had not been studied in this population, and the benefit of the medication compared with the potential harm of inducing hypovolaemia is unknown (Rabinstein and Bruder, 2011).
Hyponatraemia in SAH patients is common and primarily due to CSW, although SIADH can occur as well. Maintenance of normovolaemia and correction of sodium to normal levels are recommended. For patients with CSW, mineralocorticoid agents such as fludrocortisone can improve the degree of natriuresis. Maintenance of appropriate volume can be effectively achieved with either isotonic or hypertonic crystalloid fluids. The use of albumin is safe, but the clinical benefit is unclear.
It is common for patients with SAH to present with or develop anaemia during their hospitalization. Anaemia is typically multifactorial.
Lower haemoglobin levels have also been associated with an increased risk of an unfavourable functional outcome, especially in patients who have vasospasm (Smith et al., 2004; Wartenberg, et al., 2006; Kramer et al., 2008; Stein et al., 2015). However, red blood cell transfusion (RBCT) has also been consistently associated with worse outcomes and more systemic complications in several retrospective studies (Kramer et al., 2008; Levine et al., 2010; Festic et al., 2013; Pegoli et al., 2015). An extensive review in 2011 included 27 studies evaluating the safety and clinical benefit of RBCT in patients with aneurysmal SAH. Overall, the authors concluded that RBCT might improve some aspects of brain physiology, but is associated with increased risk of medical complications, vasospasm, and poor functional outcomes. Yet, it is unclear if RBCT is responsible for worse clinical outcomes or simply a marker of disease severity (Le Roux, 2011).
Prospective data are very scarce. A small pilot study of 44 patients randomized to a haemoglobin goal of 10 g/dL or 11.5 g/dL found that severity of physical deficits and levels of independence at 14 days, 28 days, or 3 months were not significantly different between the groups (Naidech et al., 2010). Of note, the two haemoglobin targets evaluated in this pilot study would be considered too high by many clinicians.
The presence of fever has been shown to negatively impact functional outcomes in SAH and has also been associated with cerebral vasospasm (Oliveira-Filho et al., 2001; Rabinstein and Sandhu, 2007). In a large observational study of 584 consecutive patients with aneurysmal SAH, fever occurred in 48% (Kramer et al., 2017). Fever burden was an independent predictor of poor outcome, OR 1.14 per day of fever, 95% CI: 1.06–1.22; p = 0.0006. In aneurysmal SAH, it is not uncommon for an underlying infectious or systemic inflammatory source to be absent, indicating a likely central cause of fever secondary to neurological injury. In the large observational study, fever without an identified infectious source occurred in 55% of patients (Kramer et al., 2017).
No randomized trial has evaluated induced normothermia in patients with SAH. One case–control study matched 40 consecutive febrile SAH patients treated with a surface-cooling device during the first 14 days after SAH to 80 prior febrile SAH patients managed with conventional fever control, matched by age, Hunt and Hess grade, and SAH severity (Badjatia et al., 2010). On average, over the 2-week treatment period, temperatures were 0.9°C above normal in the conventional treatment group versus 0.2°C above normal in the induced normothermia group, p < 0.001. Although patients with induced normothermia had increased rates of cardiac dysrhythmia and hyperglycaemia, induced normothermia was associated with a reduction in poor outcomes at 1 year, OR 0.2, 95% CI: 0.1–0.6, p = 0.004.
One small randomized trial evaluated mild therapeutic hypothermia (targeted temperature 34.5°C) or conventional management in 22 poor-grade SAH patients (Choi et al., 2017). Non-significantly fewer patients in the hypothermia group had symptomatic vasospasm (18% vs 36%), delayed cerebral infarction (36% vs 46%), and mortality at 1 month (0% vs 36%).
Temperature should be closely monitored and fever should be treated. Central fever is common in neurological patients, but must be a diagnosis of exclusion. Careful evaluation for infection (including ventriculitis/meningitis in patients with a ventricular or lumbar drain), deep vein thrombosis (DVT) or pulmonary embolism, and drug fever must be completed prior to making this diagnosis. For fever occurring between days 3 and 14 after aneurysmal rupture, cerebral vasospasm should also be considered. Management of fever begins with proper identification and treatment of the underlying cause, but also the fever itself should be treated, even if deemed central. Given the known association of fever with poor outcome, a goal of normothermia is reasonable. Various methods, including antipyretic medications, surface or intravascular temperature-modulating devices, and/or cold saline infusions are commonly utilized. There is insufficient evidence at present to support the routine implementation of targeted temperature management to normothermic or hypothermic targets, but further studies are warranted (Madden et al., 2017).
Deep vein thrombosis is not as common after SAH as is ischaemic stroke, likely because the patients are restless, less often have leg paralysis, are more aggressively hydrated, and are young. Among 15,968 SAH admissions for aneurysmal SAH in the US Nationwide Inpatient Sample, symptomatic DVTs were documented in 3.5% and pulmonary embolism in 1.2% (Kshettry et al., 2014). Asymptomatic DVTs, found on screening lower extremity Doppler ultrasound, occur more frequently and are reported in 3–24% of patients (Mack et al., 2008; Ray et al., 2009).
In patients with acute SAH, there are limited data regarding the timing of initiation of venous thromboembolism (VTE) chemoprophylaxis. A meta-analysis analysed 18 randomized trials and 12 cohort studies of mechanical and pharmacological prophylaxis in patients undergoing a wide variety of neurosurgical procedures, but few of the patients had SAH and the preponderance of patients were treated with open craniotomy procedures rather than minimally invasive endovascular procedures (Collen et al., 2008). Both low-molecular-weight heparin (LMWH) and intermittent compression devices (ICDs) were effective in reducing DVT (LMWH: risk ratio [RR] 0.60, 95% CI: 0.44–0.81; ICD: RR 0.41, 95% CI: 0.21–0.78). There was no statistical difference in intracranial haemorrhage rates between LMWH and nonpharmacological methods (RR 1.97, 95% CI: 0.64–6.09), though combined intracranial haemorrhage and minor bleeding were higher with pharmacological therapy.
For patients admitted with SAH, given the likelihood of limited mobility during the hospital course, VTE prophylaxis should be prescribed. Prior to securing the ruptured aneurysm, ICDs should be used. After the aneurysm is secured (at least 24 hours after the intervention), and if the clinical situation allows, pharmacological prophylaxis may be added (Nyquist et al., 2016). Special consideration and discussion with the neurosurgery team are needed prior to initiation of pharmacological prophylaxis in patients with CSF diversion using an external ventricular drain (EVD) or lumbar drain (LD), and in those where a major surgical procedure (e.g. clipping, haematoma evacuation) was required.
There are no prospective data demonstrating that screening with serial Doppler ultrasounds is cost-effective or beneficial in patients with SAH. Because some of these patients may be unable to receive pharmacological prophylaxis for a prolonged time and have limited mobility, screening ultrasound examinations may be reasonable in certain cases, particularly in patients with poor-grade SAH who may have a higher likelihood of developing a DVT.
The prophylactic use of antiepileptic medications is controversial. Seizures in the first few weeks after aneurysmal rupture occur in about 10% of patients, and most occur soon after the initial haemorrhage (O’Connor et al., 2014; Panczykowski et al., 2016). Intracranial haematoma and intracranial surgery likely increase the risk. A recent survey of several US medical centres regarding the use of prophylactic antiepileptic drugs (AEDs) in SAH found that 68% of respondents prescribe prophylactic AEDs (Dewan and Mocco, 2015). This practice may not be safe, at least if phenytoin is the drug chosen. Phenytoin use has been associated with worse functional and cognitive outcomes in aneurysmal SAH (Naidech et al., 2005) and it can accelerate the liver metabolism of nimodipine, likely making it less effective (Tartara et al., 1991).
While no randomized trials have included a treatment arm in which use of prophylactic AEDs was completely avoided, one small, single-centre trial randomized 84 patients to a brief (3-day) course of levetiracetam versus an extended (until hospital discharge) course of levetiracetam (Human et al., 2018). Although in-hospital seizures occurred non-significantly more often in the brief than extended prophylaxis group (9% vs 2%, p = 0.2), good functional outcome (modified Rankin Scale [mRS] 0–2) was more frequent in the brief treatment group (83% vs 61%, p = 0.04). Seizures were more likely to occur among patients with entry CT evidence of early brain injury (adjusted OR 12.5, 95% CI: 1.2–122; p = 0.03).
There is no clear evidence that routine use of prophylactic AEDs is beneficial in SAH patients without a history of seizures. If prophylaxis is started, a brief, 3-day period of treatment may be preferred for most patients, avoiding phenytoin and using a medication with few side effects (including less sedation) and limited drug-to-drug interactions, such as levetiracetam. A longer course of prophylaxis may be considered in patients at increased seizure risk due to early brain injury on imaging.
Patients who experience an SAH are at risk for multiple complications, both in the acute and subacute settings. Clinicians must be mindful of the likely causes of decompensation, and react urgently to stabilize the patient and manage each potential complication.
In the acute post-bleed phase, patients may experience re-rupture and rebleeding, hydrocephalus due to occluded arachnoid granulations or obstruction of foramina, and cardiopulmonary failure.
Rebleeding is a known early complication of aneurysmal SAH and typically occurs in the first 24–72 hours after the initial haemorrhage. The clinical presentation of rebleeding can be dramatic, with acute coma or sudden death occurring in some patients. The incidence of rebleeding after SAH has been estimated at 6% within 24 hours (van Donkelaar et al., 2015). The mainstay of treatment after rebleeding is urgent stabilization and definitive management of the aneurysm, with either endovascular or open surgical approaches. Hypertension may contribute to aneurysmal rupture; early blood pressure management has been discussed in a previous section.
Rebleeding after SAH is thought to originate from dissolution of the clot at the site of the ruptured aneurysm by natural fibrinolytic activity in the CSF after the SAH. Antifibrinolytic drugs cross the blood–brain barrier rapidly after SAH and reduce fibrinolytic activity. The two most commonly tested drugs have been tranexamic acid and ε-aminocaproic acid. Both agents are structurally similar to lysine and block the lysine sites by which plasminogen molecules bind to fibrin, thereby inhibiting fibrinolysis. Antifibrinolytic drugs potentially have a role in aneurysmal SAH in preventing rebleeding during the period between the initial aneurysm rupture and definitive structural treatment of the aneurysm by clipping or coiling. Most studies of antifibrinolytic agents were performed in an era when there was often a long delay between initial rupture and surgical or endovascular aneurysm treatment, rather than the current practice of proceeding rapidly to definitive aneurysm management.
A systematic review for the Cochrane Library identified 10 randomized trials evaluating the effect of antifibrinolytic therapy in a total of 1904 patients with aneurysmal SAH (Baharoglu et al., 2013). For death, data were available from all 10 trials and 1904 patients, and random allocation to antifibrinolytic therapy was not associated with altered mortality, OR 1.00, 95% CI: 0.85–1.18; p = 0.98 (Figure 14.1). For poor outcome (severe disability, vegetative state, or death), data were available from 3 trials and 1546 patients. Random assignment to antifibrinolytic treatment did not change the poor outcome rate, OR 1.02, 95% CI: 0.91–1.15 (Figure 14.2). With regard to rebleeding, all 10 trials and 1904 patients provided data. Treatment with antifibrinolytic therapy did reduce rebleeding rates, 12.7% versus 22.3%, OR 0.65, 95% CI: 0.44–0.97; p = 0.035 (Figure 14.3). However, a countervailing effect was seen for occurrence of cerebral ischaemia, for which 6 trials provided data on 1671 patients. Antifibrinolytic therapy increased cerebral ischaemia, 25.6% versus 20.3%, OR 1.41, 95% CI: 1.04–1.91; p = 0.03 (Figure 14.4). As prolonged administration of an antifibrinolytic drug increases the incidence of delayed ischaemic infarctions, thus offsetting the early benefit from prevention of rebleeding, a brief duration of antifibrinolytic administration is of interest.
Figure 14.1 Forest plot showing the effects of antifibrinolytic therapy vs control in acute SAH on death at end of follow-up. Reproduced from Baharoglu et al. (2013), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Figure 14.2 Forest plot showing the effects of antifibrinolytic therapy vs control in acute SAH on poor outcome at end of follow-up. Reproduced from Baharoglu et al. (2013), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Figure 14.3 Forest plot showing the effects of antifibrinolytic therapy vs control in acute SAH on rebleeding at end of follow-up. Reproduced from Baharoglu et al. (2013), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Figure 14.4 Forest plot showing the effects of antifibrinolytic therapy vs control in acute SAH on cerebral ischaemia at end of follow-up. Reproduced from Baharoglu et al. (2013), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Only 1 of the randomized trials evaluated brief duration therapy, assessing tranexamic acid 1 g IV every 6 hours until the aneurysm was secured or for not more than 72 hours (Hillman et al., 2002). Among 505 randomized patients, antifibrinolytic therapy reduced rebleeding (OR 0.22, 95% CI: 0.09–0.52; p = 0.0006) with a statistically non-significant offsetting increase in cerebral ischaemia (OR 1.35, 95% CI: 0.89–2.04; p = 0.16). Net effects did not reach statistical significance for final poor outcome (OR 0.85, 95% CI: 0.64–1.14, p = 0.28) or mortality (OR 0.83, 95% CI: 0.52–1.35; p = 0.46).
The use of antifibrinolytic agents may reduce the incidence of aneurysmal rebleeding in the acute setting. It is reasonable, in the absence of contraindications, to treat patients with an antifibrinolytic medication (we use tranexamic acid 1 g IV every 6 hours) for the first 72 hours after the haemorrhage or until the aneurysm is secured, whichever occurs first (Connolly et al., 2012). Initiation of antifibrinolytic may increase cerebral ischaemia complications associated with cerebral vasospasm and increases the risk of systemic thrombosis, so these medications should be avoided in patients who have entered the high-risk period of SAH-associated vasospasm (post-bleed days 4–14) and in those who have an increased thrombotic risk.
Hydrocephalus is a known complication of SAH, and can present shortly after the initial haemorrhage. It can be communicating (due to blood products blocking CSF absorption at the arachnoid granulations) or non-communicating (from foraminal obstruction). The incidence of hydrocephalus within 72 hours of haemorrhage is 15–20%, and common symptoms include depressed level of consciousness, downward eye deviation, and bradycardia (Yamada et al., 2015; Tso et al., 2016; Chen et al., 2017). The major modern methods of initial hydrocephalus management are EVD and LD, and the leading long-term management technique is ventriculoperitoneal shunting (VPS), with lamina terminalis fenestration used less often. Hydrocephalus is associated with worse functional outcome, and CSF diversion may not be effective in improving prognosis if not started early (Dupont and Rabinstein, 2013).
For patients with SAH-associated hydrocephalus, no large, randomized controlled trials have been completed comparing initial treatment strategies. Based on observational series, and the devastating course of untreated hydrocephalus, placement of an EVD is the mainstay of initial therapy. Of note, placement of an EVD is associated with a higher risk of rebleeding, but whether this relationship is causal or reflects bias by indication is uncertain. A systematic review of 16 observational studies with a total of 6804 patients found rebleeding occurred more frequently in patients undergoing EVD, 18.8% versus 6.4%, OR 3.92, p < 0.0001 (Cagnazzo et al., 2017). Pathophysiologically, placement of an EVD might increase rebleeding risk by reducing intracranial pressure, resulting in greater transmural pressure at the aneurysm site. However, the association may reflect confounding by indication, with underlying sicker patient and riskier aneurysm status separately causing more frequent treatment with EVDs and more frequent rebleeding. One small randomized trial enrolling 60 aneurysmal subarachnoid patients with EVDs placed compared continuous CSF drainage with intermittent CSF drainage and found no difference in frequency of vasospasm (65% vs 80%, p = 0.18), independent or better (mRS 0–2) outcomes (32% vs 35%, p = 0.85), or mortality (24% vs 12%, p = 0.24), though nonpatency of EVD occurred more often with continuous drainage (44% vs 12%, p = 0.03) (Olson et al., 2013).
Many patients treated with an EVD will be able to be weaned from ventricular drainage as the subarachnoid blood is reabsorbed. However, an important minority will require placement of a long-term ventriculoperitoneal shunt. Overall, in an analysis of 66 observational studies analysing 41,789 aneurysmal SAH patients, the eventual VPS insertion rate was 12.7% (Tso et al., 2016). Among patients who are initially treated with an EVD, ventriculoperitoneal shunting is pursued in over 30% (Dorai et al., 2003; O’Kelly et al., 2009). One small randomized trial in 81 patients compared rapid (within 24 hours) versus gradual (over 96 hours) weaning of an EVD and found no difference in subsequent requirement for VPS placement, 63% versus 63% (Klopfenstein et al., 2004). Nonetheless, EVD weaning over at least 2–3 days remains common across centres, and the value of this strategy deserves further investigation.
CSF diversion in patients with acute symptomatic hydrocephalus after SAH can improve the clinical status rapidly, and this treatment is indispensable in many cases. Removal of CSF can be achieved using lumbar puncture, LD, or an EVD. EVD is the only option when hydrocephalus is non-communicating (typically from obstruction at the aqueduct, thus causing dilation of the third ventricle but not the fourth ventricle). Deferring CSF diversion until after the ruptured aneurysm has been secured is advocated by some, but should only be acceptable if the delay will be short and the patient is not frankly symptomatic from the hydrocephalus.
Ventriculoperitoneal shunting (or, less often, lumboperitoneal shunting) is necessary when attempts at clamping the external drain fail because of recurrent symptoms or intracranial hypertension. Further randomized trials are needed to guide use of intermittent versus continuous EVD drainage and the rate of EVD weaning.
Patients with acute neurological injury, including SAH, are at risk of developing secondary cardiac injury in the absence of coronary disease, typically referred to as ‘stress cardiomyopathy’ or Takotsubo cardiomyopathy, mediated by a surge in adrenal circulating catecholamines or direct myocardial sympathetic stimulation by dysfunctional insular cortex. Neurogenic pulmonary oedema can also occur acutely after aneurysmal SAH, though most often combined with a cardiogenic component.
Approximately one-third of patients with SAH have elevation of troponin, and this, as well as elevated B-type natriuretic peptide, and Q wave, T wave, and ST segment abnormalities, is associated with death, poor overall outcome, and development of delayed cerebral infarction. Cardiac arrhythmias are also common, occurring in approximately one-third of patients, although life-threatening arrhythmias only occur in a small minority. Pulmonary oedema is associated with poor-grade haemorrhages, elevated serum troponin, and increased mortality (van der Bilt et al., 2009; Bruder and Rabinstein, 2011).
In the modern era of early aneurysm coiling or clipping, no randomized trial has assessed different therapies for prevention or treatment of neurogenic cardiopulmonary injury. In the preceding treatment era, a trial randomizing 224 aneurysmal SAH patients to prophylactic beta-blockade therapy with propranolol or control for 3 weeks found beta-blocker treatment associated with more frequent non-disabled outcome (72.1% vs 51.6%, p = 0.003) and fewer deaths (11.7% vs 22.6%, p = 0.02) (Neil-Dwyer et al., 1978; Neil-Dwyer et al., 1985). However, beta-blockers potentially adversely affect cerebral perfusion in the setting of vasospasm and may raise intracranial pressure, and have not been tested in a randomized study in the modern treatment era (Chang et al., 2016).
Cardiopulmonary complications are common in SAH patients. Electrolytes should be carefully monitored and adequately replaced to prevent cardiac arrhythmias. Treatment of stress-induced cardiomyopathy is supportive, with anti-hypertensives or diuretics as needed, and occasionally ionotropic agents. However, these treatments should be carefully monitored in patients with SAH, particularly those who are developing or are at high risk of vasospasm, as hypotension or hypovolaemia may precipitate cerebral ischaemia.
In the first day after hospital admission for the initial SAH, up to 15% of patients experience rebleeding. In patients who survive the first day, the risk of rebleeding without medical or surgical therapy directed at preventing recurrent haemorrhage is 35–40% over the next 4 weeks, and is more or less evenly distributed over those 4 weeks (Hijdra et al., 1987). Between 4 weeks and 6 months after the initial haemorrhage, the risk of rebleeding gradually decreases, from the initial 1–2% per day to a long-term risk of about 3% per year, if the ruptured aneurysm is not therapeutically occluded. Given its high rates of occurrence and morbidity and mortality, averting rebleeding is a dominant therapeutic aim in the management of aneurysmal SAH. The leading techniques for definitive aneurysm occlusion are: (1) endovascular coiling, in which catheter-delivered platinum coils are released to pack the aneurysm and induce thrombosis to completely fill the aneurysm sac; and (2) open surgical clipping, in which a small metallic clip or clips are placed across the neck of the aneurysm, excluding it from the circulation. The endovascular technique allows for the diagnostic angiogram and the therapeutic coiling to be completed in the same intervention. The differential efficacy and safety of each approach for different aneurysm targets must be considered when determining the ideal treatment for each individual patient.
In addition to whether and how to occlude a ruptured aneurysm, the best time period to do so has been an important focus of study. Performing the occlusion procedure early after aneurysm rupture prevents rebleeding and, with the aneurysm secured, permits more aggressive blood pressure and fluid support to treat vasospasm and avert delayed cerebral infarction. However, it requires the obliterative procedure to take place when the patient is under acute physiological stress.
The practice of performing aneurysm occlusion procedures early after initial SAH first became widespread in the era when microsurgical clipping was the main treatment procedure, after the publication of a randomized trial from Finland and a large international observational study. The Finnish study, the only randomized controlled trial (RCT) addressing this topic identified in systematic review, randomized 216 patients with aneurysmal SAH to planned surgery at 3 different time points after SAH: early (0–3 days), intermediate (3–7 days), or late (>7 days) (Ohman and Heiskanen, 1989; Whitfield and Kirkpatrick, 2001). The outcome of death or dependency at 3 months was lower in patients undergoing early- compared with intermediate-period surgery (OR 0.34, 95% CI: 0.12–0.93), and also tended to be lower compared with late-period surgery (OR 0.40, 95% CI: 0.13–1.02).
A similar advantage of early- over intermediate-period surgery was found in the large, observational International Cooperative Study on the Timing of Aneurysm Surgery, a prospective, nonrandomized study of 3521 patients (Kassell et al., 1990). Good recovery rates were higher when surgery was planned to occur in early (0–3 days) or late (15–32 days), compared with intermediate (e.g. 7–10 days) time periods: early 63%, intermediate 56%, late 63%, p = 0.046. Among the 722 patients enrolled in the United States, where patients may have had access to more advanced intensive care postoperatively, good recovery rates were highest with early period surgery: early 71%, intermediate 56%, late 63%, p < 0.05 (Haley et al., 1992).
The development of endovascular coiling further entrenched early aneurysm obliteration as a widely employed treatment strategy, as coiling, compared with clipping, potentially has less differentially higher intra-procedural risk when performed early after SAH. In the era of coiling, observational studies, but no randomized trials, have investigated whether performing endovascular aneurysm occlusion procedures ultra-early, in the first 24 hours after SAH, is advantageous compared with later time periods. A systematic review identified 8 observational series with a total of 1594 SAH patients (Rawal et al., 2017). Endovascular coiling in the ultra-early period at <1 day was associated with a reduced frequency of poor outcome compared with treatment at >1 day (RR 0.56, 95% CI: 0.45–0.70; I2 = 0%) but not when compared with treatment at 1–3 days (RR 1.12, 95% CI: 0.58–2.15; I2 = 80%) (Figure 14.5).
Figure 14.5 Forest plot showing the effects of ultra-early vs later endovascular coiling for ruptured cerebral aneurysms on poor functional outcome at end of follow-up in non-randomized studies. Figure courtesy of JL Saver, under a Creative Commons 4.0 CC-BY license (JL Saver).
A treatment policy of procedural treatment of the ruptured aneurysm in the early, initial 0- to 3-day period following SAH is highly reasonable, though supported preponderantly by observational series rather than randomized trials. Securing the aneurysm early after rupture reduces rebleeding frequency and permits more aggressive vasoactive management to prevent delayed cerebral infarction from vasospasm. Further compressing the target procedural period to just the first 24 hours after aneurysm rupture, rather than days 1–3, does not consistently provide additional improved outcomes.
A systematic review identified 4 randomized trials enrolling 2458 participants comparing endovascular coiling and surgical clipping of ruptured aneurysms (Lindgren et al., 2018). Only patients deemed treatable by either modality were enrolled. A preponderance of the patients, 87%, were contributed by one large trial, the International Subarachnoid Aneurysm Trial (ISAT) (Molyneux et al., 2002; Molyneux et al., 2015). The majority of enrolled patients had mild-to-moderate rather than severe-grade SAH, and most aneurysms were located within the anterior circulation. All 4 trials contributed data on outcomes through 1 year. Only ISAT contributed data on long-term outcomes, with follow-up extending to 18 years.
With regard to functional outcomes, allocation to endovascular coiling compared with surgical clipping and reduced death and dependency at 1 year, 24% versus 32% (RR 0.77, 95% CI: 0.67–0.87; p = 0.00005) (Figure 14.6). At 10 years, the reduction in death or dependency was sustained (RR 0.81, 95% CI: 0.70–0.92; p = 0.002).
Figure 14.6 Forest plot showing the effects of endovascular coiling vs surgical clipping for ruptured cerebral aneurysms on poor outcome at 1 year. Reproduced from Lindgren et al. (2018), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
For mortality, assignment to endovascular coiling was associated with a non-significant reduction in death at 1 year, 8.5% versus 10.7% (RR 0.80, 95% CI: 0.63–1.02; p = 0.07) (Figure 14.7). At 10 years, the reduction in mortality reached nominal statistical significances (RR 0.78, 95% CI: 0.64–0.96; p = 0.02).
Figure 14.7 Forest plot showing the effects of endovascular coiling vs surgical clipping for ruptured cerebral aneurysms on death at 1 year. Reproduced from Lindgren et al. (2018), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Data regarding DCI by 2–3 months were available for 2450 patients from the 4 trials. Allocation to endovascular coiling compared with surgical clipping reduced DCI, 23.8% versus 28.5% (RR 0.84, 95% CI: 0.74–0.96; p = 0.01) (Figure 14.8).
Figure 14.8 Forest plot showing the effects of endovascular coiling vs surgical clipping for ruptured cerebral aneurysms on delayed cerebral infarction. Reproduced from Lindgren et al. (2018), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.
Data regarding the degree of sustained aneurysm obliteration were available from 3 trials for 1626 patients. At 1 year, non-complete obliteration of the target aneurysm was more frequent with endovascular coiling than with neurosurgical clipping, 33.3% versus 16.4% (RR 2.02, 95% CI: 1.65–2.47; p < 0.00001) (Figure 14.9). However, substantial non-complete obliteration (<90% occlusion) was less frequent, and not statistically different, 7.6% versus 5.1% (RR 1.43, 95% CI: 0.93–2.21; p = 0.11) (2 trials, 1440 patients).
Figure 14.9 Forest plot showing the effects of endovascular coiling vs surgical clipping for ruptured cerebral aneurysms on incomplete aneurysm obliteration. Reproduced from Lindgren et al. (2018), with permission from the authors and John Wiley & Sons Limited. Copyright Cochrane Library, with permission.