Aneurysmal subarachnoid haemorrhage

Introduction

Spontaneous subarachnoid haemorrhage accounts for only 3–4% of all stroke but the consequences can be devastating.^1,2 Approximately 85% of cases were due to ruptured intracranial aneurysms.^3–5 Advances in understanding of the disease, neurointensive care, endovascular treatment and refinement of microsurgical treatment have all led to improvement in management outcome in the recent decade. Nevertheless, its associated complications such as cerebral vasospasm remain a major cause of morbidity and mortality in this group of patients.

Epidemiology of spontaneous subarachnoid haemorrhage

The incidence of spontaneous subarachnoid haemorrhage was thought to be quite stable over the years. In a meta-analysis done in 1996, the overall incidence rate was 10.5 per 100 000 person years. Finland and Japan had a higher incidence of 22.0 and 23.0 per 100 000 person years, respectively.⁶ The incidence would be 7.8 per 100 000 person years if these two regions were excluded from analysis. Women also have a higher risk than men (1.6-fold, 95% confidence interval 1.5–2.3). In the neurosurgical unit of our university-affiliated hospital, we managed approximately 60 patients per year for spontaneous subarachnoid haemorrhage. Assuming 15% of this group of patients were dead on arrival to the emergency department and a population base of 1.2 million, we arrived at an estimated local incidence of 6 per 100 000 person per year.

A non-modifiable factor in the Caucasian population is familial predisposition to aneurysmal subarachnoid haemorrhage. In the Caucasian population, between 5 and 20% of patients with subarachnoid haemorrhage have a positive family history, the risk being in first-degree relatives.⁷ This, however, contradicts our finding locally, in a review of 200 consecutive Chinese patients admitted to our unit for spontaneous subarachnoid haemorrhage, none had a family history of spontaneous subarachnoid haemorrhage. Other important risk factors include smoking, hypertension and excessive alcohol intake. The proposed protective effects of white ethnicity compared to non-white ethnicity, hormonal replacement therapy, hypercholesterolaemia and diabetes in the aetiology of spontaneous subarachnoid haemorrhage remain uncertain.⁸

Only a minority of patients with aneurysmal subarachnoid haemorrhage were associated with specific hereditary disorders such as autosomal dominant polycystic kidney disease.⁹ The associations with other connective tissue diseases such as Marfan Syndrome are even weaker.¹⁰

Clinical features of spontaneous subarachnoid haemorrhage

Sudden severe headache is a cardinal symptom.¹¹ When patients were asked how long it took for their headache to reach its maximal severity, 50% of patients with subarachnoid haemorrhage described it as instantaneous, 20% of them described it as developing rapidly over 1–5 min, and the rest said it escalated over more than 5 min.¹² The headache usually persisted for days but definitely lasts more than an hour. The condition accounts for 3% of patients presenting to emergency departments with headache and approximately one in four people presenting with sudden severe headache will have had a subarachnoid haemorrhage.^13,14 Other symptoms of meningism, such as nausea, vomiting, photophobia and neck stiffness, are also commonly present. Loss of consciousness may also occur; it may be transient (50% at onset) or persist into coma. Epileptic seizures occurred at presentation in approximately 6–16% of cases.^15–17 Focal neurological deficits such as dysphasia or hemiplegia may occur, especially in cases of ruptured middle cerebral artery aneurysm associated with intracerebral haematoma. One to two per cent of patients with spontaneous subarachnoid haemorrhage may present with acute confusion and in most such patients a history of sudden headache is lacking.¹⁸ Sudden death may occur in 10–15% of cases of spontaneous subarachnoid haemorrhage.¹⁹ Trauma and spontaneous subarachnoid haemorrhage are sometimes difficult to differentiate in cases of syncope and significant head injury. Third nerve palsy can occur after rupture of an aneurysm of the internal carotid artery at the origin of the posterior communicating artery or aneurysm of the basilar bifurcation. Abducens nerve palsy may occur as a result of the raised intracranial pressure, either at the time of rupture or subsequent hydrocephalus. Cerebellar signs are common in patients with vertebral artery dissecting aneurysm. Clinical evaluation is usually communicated through Glasgow Coma Scale, and motor deficit/dysphasia excluding cranial neuropathies in terms of the World Federation of Neurological Surgeons Grading Scale (Table 1).

Investigations for suspected spontaneous subarachnoid haemorrhage

As no clinical symptoms and signs would be diagnostic of spontaneous subarachnoid haemorrhage, the diagnosis must be excluded if there is any clinical suspicion. Unenhanced computed tomography (CT) of the brain is the first line of investigation. Blood has the typical hyperdense appearance in the basal cisterns and the pattern of haemorrhage often suggests the location of any underlying aneurysm.²⁰ The time of haemorrhage to scanning is crucial to obtain a positive image. Modern CT scanners miss approximately 2% of subarachnoid haemorrhage within 12 h and approximately 7% by 24 h.^11,21 Subarachnoid blood is almost completely reabsorbed by day 10. Radiological changes can be quite subtle and require careful examination to identify. However, venous congestion in the subarachnoid space, as caused by generalized cerebral oedema, is well known to mimic basal cistern subarachnoid haemorrhage.^22,23

Lumbar puncture is still an essential investigation for exclusion of spontaneous subarachnoid haemorrhage with a history suggestive of the diagnosis though negative brain imaging. It should preferably be carried out 12 h after the onset of headache, feasible up to 2 weeks, unless there is a suspicion of an alternative life-threatening diagnosis, such as meningitis. The delay is essential for sufficient red cell lysis to form bilirubin from oxyhaemoglobulin.²³ These pigments will give the diagnostic yellow tinge or xanthchromia after centrifugation. If the supernatant appears crystal-clear, the specimen needs to be stored in darkness until the absence of blood pigments is confirmed by spectrophotometry.²⁴ Bilirubin equals subarachnoid haemorrhage as it is only synthesized in vivo.

Magnetic resonance imaging (MRI) using fluid attenuated inversion recovery (FLAIR) technique can also reliably show the subarachnoid haemorrhage.²⁵ It is especially useful for diagnosis in the subacute phase.^26,27 The main disadvantage is the time required for the patient to lie still and, thus, is not suitable for restless patients, which are common in this patient group.

Investigation for the cause of spontaneous subarachnoid haemorrhage

In 85% of patients with spontaneous subarachnoid haemorrhage, it is due to ruptured intracranial saccular aneurysm in the skull base. Ten per cent of cases may be due to non-aneurysmal perimesencephalic haemorrhage and 5% of cases may be due to other aetiologies such as arterial dissection, cerebralarteriovenous malformation, dural arteriovenous fistula, vascular lesion around the spinal cord and septic aneurysm.²⁴ In a recent review of the aetiological data of 101 patients with spontaneous subarachnoid haemorrhage within 48 h after ictus in our institute, there were 27 (25%) anterior cerebral artery aneurysms, 35 (32%) internal carotid artery aneurysms, 12 (11%) middle cerebral artery aneurysms, 20 (18%) posterior circulation aneurysms, four (3.6%) cerebral arteriovenous malformations, two (1.8%) Moya-moya disease and one (0.9%) dural arteriovenous fistula.

The vascular diagnosis needs to be established by angiography. The options include catheter angiography, computed tomographic angiography and magnetic resonance angiography. Catheter angiography remains the gold standard, although it carries a small risk of neurological morbidity and mortality of 1.8% in this acute setting, resulting from embolism or rebleeding.²⁸ It requires the presence of trained neuroradiologists and supporting staff. In cases where endovascular treatment is preferred, it can be done at the same setting for endovascular treatment. Modern MR angiography is also sensitive in making the vascular diagnosis, but is inconvenient for patients who are restless or who require close monitoring.²⁹

In clinical practice, we preferred multislice computed tomographic angiography as the initial investigation for diagnosis and planning of treatment. It can be done within a minute, is feasible in cases requiring intensive monitoring, intubation, cerebrospinal fluid drainage, and even in some restless patients. It also provides anatomical information for planning of microsurgery or endovascular treatment. In recent case series using multislice computed tomographic angiography as the initial vascular investigation, a high sensitvity of 96–100% and a high specificity of 97–100% were reported.^30–32

General medical management of aneurysmal subarachnoid haemorrhage

Neurointensive care management is one of the cornerstones for improvement in the clinical care of patients with aneurysmal subarachnoid haemorrhage. To realize the magnitude of importance, we reviewed a cohort study of 580 patients with subarachnoid haemorrhage in a neurological intensive care unit.³³ Fever, anaemia and hyperglycaemia affected 30–54% of patients with subarachnoid haemorrhage and were significantly associated with mortality and poor functional outcome, after adjustment for age, Hunt-Hess grade, aneurysm size, rebleeding and cerebral infarction due to vasospasm. The study emphasized the importance of critical care strategies directed at maintaining normothermia, normglycaemia and prevention of anaemia.

Oral nimodipine is the standard medical treatment for all patients with aneurysmal subarachnoid haemorrhage, presented within 96 h after haemorrhage. The ideal dosage is 60 mg every 4 h for 3 weeks. The neuroprotective effect was thought to be the mechanism of action. In the meta-analysis of all published randomized trials of prophylactic nimodipine used in patients who have experienced subarachnoid haemorrhage,³⁴ nimodipine improved outcome according to all measures examined. The odds of good and of good plus fair outcome were improved by ratios of 1.86:1 and 1.67:1, respectively, for nimodipine versus control (P < 0.005 for both measures). The odds of deficit and/or mortality attributed to vasospasm and CT-assessed infarction rate were reduced by a ratio of 0.46:1–0.58:1 in the nimodipine group (P < 0.008 for all measures). A meta-regression yielded findings indicating that the treatment effect of nimodipine in individual trials was positively correlated with the severity of subarachnoid haemorrhage in enrolled patients.

A daily fluid intake of at least 3 L, unless cardiac failure occurred, was suggested by a cohort study and recent guidelines to prevent sodium depletion and hypovolaemia.³⁵ In these patients, fluid balance and cardiac function should be carefully monitored. In this study, the outcome of patients (with delayed aneurysm surgery at day 12) was improved as compared to the historical cohort.³⁵

Analgesics, such as paracetamol 1 g every 6 h or dihydrocodeine 30–60 mg every 4 h may be prescribed for headache, thus anti-emetics and stool softeners are required. Graduated compression stockings are good measures to prevent deep vein thrombosis.

Prevention of rebleeding

Prevention of rebleeding is usually done through occlusion of the ruptured aneurysm by either microsurgical clipping or endovascular intervention. Microsurgical clipping, the extravascular surgery, has had a long established history in terms of techniques and efficacy, whereas endovascular embolization, the new treatment in clinical practice since 1995, is gaining popularity. For aneurysms suitable for either treatment, endovascular coil embolization confers an absolute risk reduction of 7% (25% relative risk reduction) for death and dependency at 1 year and up to at least 7 years.^36,37 Although the conclusion is mainly influenced by one large trial which included mostly young good neurological grade patients with small anterior circulation aneurysms, ruptured posterior circulationaneurysms had long been preferentially treated by endovascular coiling due to a relatively higher neurological morbidity and mortality associated with microsurgical clipping. Our centre, as in most centres in the UK, had two-thirds of the ruptured aneurysms occluded by endovascular coiling and one-third occluded by microsurgical clipping.

Aneurysm occlusion is usually attempted within the day after the diagnosis was made, although there was no evidence to support the best timing, at least, for microsurgical clipping. Aneurysm occlusion would also facilitate subsequent management of hydrocephalus and vasospasm, without the concern of causing a rebleed. Deteriorating neurological condition associated with intracerebral haematoma from ruptured middle cerebral artery aneurysm can be managed effectively with evacuation of the haematoma and microsurgical clipping. Prompt surgical evacuation of massive intracerebral hemorrhage (ICH) from ruptured middle cerebral artery aneurysm improves outcome.^38–40

Role of hypermagnesaemic treatment in aneurysmal subarachnoid haemorrhage

Delayed ischaemic neurological deficit or clinical vasospasm remains a major cause for delayed neurological morbidity and mortality for patients with aneurysmal subarachnoid haemorrhage. Magnesium is a cerebral vasodilator. In an experimental model of drug or subarachnoid haemorrhage-induced vasospasm, magnesium blocks voltage-dependent calcium channels and reverses cerebral vasoconstriction. Furthermore, its antagonistic action on N-methyl-d-aspartate receptor in the brain prevents glutamate stimulation and decreases calcium influx during ischaemic injury.⁴¹ From the meta-analysis of the three current pilot randomized controlled clinical trials in patients after aneurysmal subarachnoid haemorrhage, magnesium sulfate infusion was effective for patients with aneurysmal subarachnoid haemorrhage in terms of reduction of clinical vasospasm and improvement in neurological outcome. Favourable outcome was achieved in 71% of the magnesium-treated group and 61% of the placebo group, P = 0.041. Symptomatic vasospasm or delayed cerebral ischaemia was noted in 19% of the magnesium-treated group and 28% of the placebo group, P = 0.036. The number needed to treat (NNT) for achieving favourable outcome and reducing symptomatic vasospasm were 10 and 11, respectively.^42–46 Two multicentre trials are ongoing and should provide definitive information about the use of magnesium sulfate infusion in patients after aneurysmal subarachnoid haemorrhage in 2–3 years.