Introduction

Spontaneous non-traumatic Subarachnoid Hemorrhage (SAH) is a rare form of stroke with an incidence of approximately 9:100,000 in western industrialized nations associated with high morbidity and mortality of up to 40%. This form of SAH mainly affects patients of younger age around the fifth decade of life and nearly 80% of cases are due to cerebral aneurysms (aneurysmal SAH, aSAH) [1]. The primary treatment goal of aSAH is to eliminate the source of bleeding by open surgical or endovascular treatment options [2]. As aSAH is often accompanied by intraventricular and intracerebral blood components, patients are at risk of consecutive occluding hydrocephalus and the development of cerebral edema [3]. Furthermore, a distinctive aSAH feature is the biphasic disease course: after initial neuronal damage within the first 72 hours due to toxic effects of subarachnoid blood, transient global ischemia, and direct damage to the brain parenchyma, delayed onset of neurological deficits, i.e. Delayed Cerebral Ischemias (DCI) occur, especially on days 3 to 14 [4]. Although hospital inpatient mortality rates from aSAH have decreased over recent years, 2 persistently high incidences of medical complications necessitate improved therapies and standardized practice in the post interventional care of patients in order to maintain a comprehensive approach [5]. Out of numerous, sometimes controversially discussed treatment options, we here discuss five main aspects of therapeutic and prophylactic strategies, including the key interventions of CSF drainage, vasospasm therapy and DCI prevention, BP and ICP management, as well as seizure treatment.

Immediate CSF diversion, hematoma clearance and antifibrinolytic therapy

Symptomatic acute or chronic hydrocephalus is common after aSAH (15-87%) and associated with worse outcome [2]. Pathophysiologically, Intraventricular Hemorrhage (IVH), but also ependymal and plexus choroideus damage, CSF hypersecretion from the choroid plexus, impaired CSF drainage via arachnoid granulations or inflammatory responses may cause a disruption of CSF circulation and consequently induce occluding or malresorptive hydrocephalus [2]. Therapeutic targets of CSF dynamic modulation after aSAH include surgical (CSF diversion via External-Ventricular-Drain (EVD) and/or Lumbar Drain (LD), ventriculoperitoneal drain, ventriculostomy) and pharmacological interventions (intracisternal or intraventricular fibrinolytics) [6,7]. Yet, after surviving the initial hemorrhage, still 9-48% of aSAH patients develop chronic, shunt-dependent hydrocephalus [2].

Surgical interventions are recommended for CSF diversion and to reduce the risk of shunt-dependent hydrocephalus

Recommendations and options

In patients with decreased level of consciousness and/or detected acute hydrocephalus, CSF diversion (and ICP measurement) should be applied immediately, 2 and in case of occluding hydrocephalus, primarily EVD placement should be initiated [2]. So far, no clear protocols for EVD management (i.e. optimal drainage and weaning strategies) have been developed. Intermittent drainage implies to maintain the EVD closed by default, and to drain CSF when ICP reaches a set threshold, or when patients become symptomatic. Recently, a systematic review and meta-analysis found lower incidences of EVD-related infections (RR=0.20, 95% CI 0.05–0.72) and EVD-blockages (RR=0.45, 95% CI 0.27–0.74) with intermittent, compared to continuous CSF drainage [8]. Rapid wean is defined as EVD closure within 48h of starting a weaning trial without reopening, unless evidence of increased ICP, clinical worsening, CSF leakage or progressive hydrocephalus [9]. In contrast, gradual EVD wean involves to leave the drain open and to raise it gradually over several days [9]. In a prospective multicenter study, Chung et al. suggested lower rates of permanent shunt-dependency and drain complications, as well as decreased lengths of ICU stay in patients with rapid, compared to gradual wean [9]. Although no clear recommendations have been established so far, guidelines emphasise the importance of implementing standardized care bundles in patients who require EVD treatment [2].

After restoration of ventricular communication, lumbar CSF drainage (LD) should be considered [2]. In 210 aSAH patients (WFNS Grade 1-3), LUMAS, a single-center prospective RCT first suggested that LD improves early clinical outcome, yet failed to show long-term outcome benefits [10]. Recently, the EARLYDRAIN trial demonstrated that prophylactic LD decreases rates of secondary infarction (absolute risk difference, −0.11; 95%CI, −0.22-0) and unfavorable outcome at 6 months (absolute risk difference, −0.12; 95%CI, −0.23-−0.01) in all grades of aSAH [11]. Furthermore, with mean CSF drainage rates of 5ml/h during the first 8 days, LD (with/without EVD) was associated with lower ICPs compared to EVD alone [11]. As unfavorable outcome after aSAH has clearly been linked to duration and magnitude of ICP elevation, these findings support the use of LD in addition to standard of care [11].

Considerations

Meningitis and ventriculitis are frequent and severe complications of CSF drainage (4-17 infection days per 1,000 EVD days),12 and a strong independent predictor of unfavorable outcome and longterm neurological dysfunction after aSAH [12]. A comprehensive systematic review and meta-analysis found that bolted and antibiotic-coated catheters were superior to tunneled/uncoated catheters (P<0.001), and antibiotic- vs. silver-impregnated catheters (P<0.001) in preventing infections [13]. Furthermore, best-practice recommendations include to remove catheters as soon as clinically possible, preferably in less than 5 days [13].

Although microsurgical fenestration of the lamina terminalis had been investigated in terms of facilitating CSF dynamics, a metaanalysis of almost 2,000 aSAH patients revealed no differences in rates of shunt-dependent hydrocephalus (10% in the fenestrated cohort vs 14% in the non-fenestrated cohort) [14]. Therefore, routine fenestration of the lamina terminalis is not recommended by international guidelines [2].

Intraventricular fibrinolytics for hematoma clearance may be considered while intravenous anti-fibrinolytic therapy has no proven effect to prevent re-bleeding.

Recommendations and options

Especially in patients with occluding hydrocephalus due to IVH components, installation of fibrinolytic agents (IVF) may be considered. Although the study was primarily negative, a prespecified secondary analysis of the CLEAR-III trial demonstrated benefits of IVF in terms of mortality [15]. Moreover, exploratory analyses of CLEAR-III suggest positive effects on functional outcome, however, effects on long-term functional outcome have not been found [15]. Recently, a systematic review and meta-analysis further identified IVF to promote hematoma clearance, reduce rates of catheter occlusion, decrease mortality and improve functional outcome [13]. Overall, IVF complication rates appear low and high dosage of alteplase (4mg/12h) had favourable effects compared to low-dosage of alteplase (1mg/8h), despite absence of dose-related adverse events [13].

Considerations

Current evidence indicates that routine use of antifibrinolytic therapy does not reduce rates of re-bleeding, or improve functional outcome [2]. After conflicting results from older studies, recommendations derive from the ULTRA trial-the largest RCT that evaluated ultraearly and short-term therapy with tranexamicacid after aSAH [16] Between 2013 and 2019, the authors found similar rates of good clinical outcome (mRS 0-3; 60 vs 64%), and of re-bleeding (10 vs 14%) between patients treated with antifibrinolytic medication (within a median of 185 minutes after disease onset, n=480) compared to patients who did not receive tranexamic-acid (n=475) [16].

Strict vascular monitoring, vasospasm treatment and DCI prevention

Neurological deterioration following the acute phase commonly occurs 3 to 14 days after aSAH onset and cerebral infarction due to DCI increases morbidity in patients surviving the initial hemorrhage [17]. With an OR of 3.7 (95% CI 1.9–6.9), cerebral infarction due to DCI represents one of the strongest predictors of in-hospital mortality [17,18]. Pathophysiologically, DCIs are attributed to cerebral vasospasm among other causes (e.g. impaired autoregulation, microthrombosis, cortical spreading ischemia) [19]. In total, 70% of aSAH patients have detectable vasospasm and approximately 30% develop delayed neurological deterioration and focal deficits that may progress to infarction [2,7].

Various diagnostic procedures are available to detect cerebral vasospasms.

Recommendations and options

Especially in high-grade aSAH patients, different methods for monitoring and detection of arterial narrowing may be applied in order to predict DCI [2].With a sensitivity of 90%, non-invasive transcranial Doppler ultrasonography (TCD) monitoring is considered reasonable, and TCD evidence of vasospasm was found to be highly predictive of DCI [2]. Yet, clinical value may be limited by patient anatomy (temporal bone window) and physiological measures such as heart rate and BP [2]. With a similarly high sensitivity, CT-Angiography (macroscopic vasospasm) and CTPerfusion (alterations to the microcirculation) can be useful in patients with suspected vasospasm [2]. Continuous measure of cerebral function using cEEG may complement other techniques. Although further validation and automatization is needed, detection of a decrease in alpha-power variability, a decrease in alpha-to-delta power ratio, and other patterns have been suggested to predict DCI [2]. So far, there is insufficient evidence that invasive multimodal neuromonitoring relevantly affects long-term clinical outcome [2]. Yet, monitoring of brain tissue oxygenation, lactate-to-pyruvate ratio, and glutamate may also be considered to predict DCI [2].

Considerations

Despite advantages of TCD as a noninvasive, safe and easy to perform bedside neuromonitoring technique, the predictive value for vasospasms and DCI detection is rather low (57%) limiting accuracy and applicability of this technique [20]. Therefore, the combination with other neuromonitoring modalities such as EEG and CTP may be considered to improve DCI prediction, especially in the setting when neurological examination is limited due to clinical severity and/or decreased level of consciousness [21,22].

The continuous administration of nimodipine and other neuroprotective agents is generally recommended.

Recommendations and options

Already 40 years ago, Allen, et al. demonstrated that administration of the dihydropyridine calcium channel blocker nimodipine is beneficial in preventing DCI and improving functional outcome after aSAH [23]. Since then, several trials and metaanalyses confirmed this finding, and early enteral administration of nimodipine (6 x 60mg per day) still represents the only recommended on-label prophylactic therapy [2]. Even in the setting of nimodipine-induced hypotension, consistent administration is recommended, as a therapy disruption has been associated with higher DCI incidences [24] Although enteral administration should be preferred, intravenous therapy seems similarly effective [25].

Considerations

Several RCTs assessed the relevance of statin therapy in aSAH patients [2]. Although treatment with statins significantly decreased the occurrence of vasospasm, no benefits on DCI, outcome and mortality have been observed, 26 consequently, routine use is not recommended [2].

Despite suggested neuroprotective effects, intravenous magnesium sulphate is not recommended routinely after aSAH. Among other studies, the MASH-2 RCT found no benefit in terms of cerebral infarction, outcome, or mortality [27].

Rescue strategies should be considered when vasospasms occur.

Recommendations and options

If clinical deterioration occurs, which can be determined using the NIHSS or GCS (defined as difference of at least 2 points), or with the help of specific prediction tools such as the VASOGRADE, sever vasospasms may be present [28,29]. Especially if neurological examination is unreliable, i.e. in patients with decreased level of consciousness, neuromonitoring techniques, e.g. TCD (most studies focus on the middle cerebral artery definition of vasospasms: velocity ≥ 120 cm/s or Lindegaard ratio ≥ 3) or continuous EEG monitoring (cEEG, with abnormalities decreased alpha power variability or late-appearing epileptiform abnormalities) should be applied.20,30 If clinical deterioration does not resolve by induction of euvolemia or BP elevation, additional neuroimaging (CTA and CTP) and rescue strategies should be considered [2,31]. Endovascular options like intra-arterial application of nimodipine and milrinone have been suggested to be beneficial in order to target the proximal and distal vasculature [32]. Intermittent infusion via a cervical catheter should be preferred over continuous infusion regarding efficacy and complications (e.g. hypotension and elevated ICP) [33]. Both intra-arterial applicable agents have been suggested to improve angiographic and long-term functional outcomes [34,35]. Yet, besides mostly retrospective case series and single-center cohorts, no high-quality randomized trials have been undertaken to compare different agents, and future studies are needed to evaluate functional outcome und risk stratification [2,34]. Besides that, angioplasty with compliant and noncompliant balloons has been used as a mechanical option to treat vasospasm in the proximal vessels and existing data suggests enhanced angiographic responses compared to intra-arterial application of vasodilators [36]. Furthermore, a combination of angioplasty and i.e. medication may result in longer durability of angiographic responses [2]. Poor angiographic responses during angioplasty may be due to ongoing and recurrent vasospasms and DCI. Furthermore, the risk of vessel rupture during angioplasty with consecutive high mortality should be considered.37 Overall, only limited data exists comparing mechanical and medicamentous therapeutic options [36].

Considerations

Due to the increased risk of organ failure (especially pulmonary edema and congestive heart failure), “triple-H therapy” including hemodynamic augmentation methods in form of hypervolemia, hemodilution, and hypertension is no longer recommended [38]. Not only hyper-, but also hypovolemia have been associated with worse outcome after aSAH, so that several studies recommended maintenance of euvolemia. Yet, no clear protocols and approaches (crystalloid infusions, vasopressor therapy) have been tested in RCTs [38]. Finally, because of the risk of gray matter changes and neurotoxicity intra-arterial papaverine should be avoided as a rescue strategy in refractory vasospasms [39].

Continuous BP monitoring, and normalization of BP

Compared to other forms of hemorrhagic stroke and intracranial bleeding, BP management seems to be more complex in SAH patients [7,40]. The acute phase of aSAH is often complicated by elevated BP, in terms of ongoing and recurrent hemorrhage, death and disability [40]. On the other hand, lowering BP reduces Cerebral Perfusion Pressure (CPP) and may increase the risk of DCI due to elevated ICP [40].

Continuous blood pressure monitoring is recommended.

Recommendations and options

Similar to other cerebrovascular diseases, autoregulation has been demonstrated to be impaired in aSAH patients [2,41]. Furthermore, elevated BP variability was found to be associated with DCI and worse outcome [2,41]. Therefore, continuous or highly frequent BP measurement is recommended, in order to minimize BP variability, 2 and closely respond to excessive changes in BP [2,41].

Considerations

Blood pressure monitoring should be part of standard intensive care unit bundles to be able to detect and react to corresponding BP fluctuations [2].

Hypertension should be treated immediately.

Recommendations and options

Approximate normalization of BP, i.e. treatment of hypertension within the acute phase of the disease seems justified to reduce the risk of ongoing and recurrent hemorrhage, i.e. hematoma progression [40]. Although available evidence in aSAH is insufficient to recommend specific BP goals, international guidelines advice that high BP should be treated [40]. Current recommendations for acute BP reduction derive from the two ICH RCTs INTERACT2 and ATACH2 [42,43]. In acute aSAH, systolic-BP<160mmHg (with CPP >70mmHg and MAP 60-100mmHg) may be a reasonable therapeutic goal [2,41] treatment may be started with nimodipine and analgesia, 40 but more importantly, agents to lower BP should be well tolerated, easy to titrate and effective to achieve and maintain BP targets without excessive variability [40]. The recently published INTERACT3 trial also emphasises the benefits of prompt blood pressure reduction in patients with ICH in order to improve chances of survival without major disability [44]. Yet, it is important to mention that not only early intensive BP lowering but also simultaneous administration of specific care bundles including treatment of hyperglycemia, fever and rapid reversal of abnormal anticoagulation are needed to improve outcome [44].

Considerations

So far, no controlled trials were able to demonstrate advantages of induced hypertension to effectively target DCI [40,45]. Although the trial was terminated after recruitment of only 41 patients, the HIMALAIA RCT suggested an adjusted risk-ratio for poor outcome of 1.0 (95% CI, 0.6-1.8), and a risk-ratio for serious adverse events of 2.1 (95% CI, 0.9-5.0) with induced hypertension [45]. Yet, since most aSAH patients develop some degree of vasospasm, 41 permissive arterial hypertension may be beneficial after the source of bleeding is secured [40]. Therefore, aggressive BP management should be avoided in the post interventional phase and loosened BP targets in patients with vasospasm seem reasonable [40].

Deep sedation, osmotic therapy, and hypothermia for ICP control

In aSAH patients, ICP elevation may be due to hydrocephalus, ischemia, intracerebral or subdural hemorrhage components, electrolyte imbalance, or a generalized cerebral edema. It is well known that elevation of ICP may induce local but also global cerebral ischemia and consequently increases the risk of poor outcome.2,7 Therefore, international guidelines recommend ICP target values <23mmHg, especially in patients with untreated aneurysm, and maintenance of CPP (60-90mmHg) [46]. Basic measures of ICP management include elevated head positioning as well as adequate sedation with regard to treatment of pain and agitation. Irrespective of the underlying disease, an increase in body-core temperature enhances cerebral metabolism, with augmented cerebral blood-flow and -volume. If compensatory mechanisms are exhausted, ICP may raise [47]. As fever (infectious and noninfectious) is the most common complication during the clinical course of aSAH, an increased risk of poor outcome for each day that a patient has fever (OR 1.4) has been described.48 Therefore, temperature management to control fever after aSAH is recommended to lower ICP [2]. Furthermore, several studies demonstrated that hyperosmotic agents may be used to manage ICP and brain relaxation [49,50].

Analgosedation may be used to support ICP control.

Recommendations and options

Although general anesthesia limits clinical evaluability of patients, analgesia and mechanical ventilation are recommended for ICP control in comatose patients and/or patients with depleted effect of CSF drainage [2]. Usual sedative medication comprises propofol and benzodiazepines. The analgesic component could be covered by opioids or ketamine. Specific pharmacological approaches have so far not been evaluated in RCTs, yet data from literature reviews suggest that bolus administration should be avoided [51]. Furthermore, the implementation of standardized ICU care bundle in patients who require anesthesia for >24h is recommended [2].

Considerations

Despite effects on ICP-control, no benefits of barbiturate therapy have been described in terms of death or disability [52]. In contrast, especially long-term barbiturate therapy has been associated with significant side effects (e.g. pneumonia) and poor outcome [52]. Therefore, long-term barbiturate therapy should be avoided. Propofol as well as inhalative anesthesia agents could have both beneficial and adverse effects [53]. The propofol infusion syndrome, a rather rare complication, is associated with increased morbidity and high mortality, limiting applicability of this drug. Therefore, the use of propofol is not recommended for a duration >7 days and doses >4mg/kg [54].

Osmotic therapy may be used to support ICP control while hypothermia is not generally recommended.

Recommendation and options

It has been demonstrated that administration of osmotic agents (i.e. mannitol and hypertonic saline) potently lowers ICP [50,55,56]. Therefore, short-term administration (over a few minutes) for the treatment of ICP crisis can be recommended; effects develop after a few minutes and last up to two hours [49]. Hypertonic saline increases blood sodium (target values: 155-160mmol/l), and bolus applications as well as continuous infusion may be considered. While most studies, e.g. a prospective study by Huang, et al. found no evidence to recommend one therapy over the other, 55 hypertonic saline (3%) was slightly more effective than mannitol (20%) in terms of lowering ICP, and was still partially effective in “mannitol failures” in a meta-analysis [50].

Considerations

While the use of osmotic agents is justified for prophylactic or therapeutic use in brain swelling, important side effects have to be taken into account. Due to its diuretic effects, mannitol (target osmolarity 330 mOsmol/l) can cause hypovolemia and hypotension. Furthermore, renal elimination may induce metabolic acidosis and renal damage, why daily dosages of 6 x 250mg should not be exceeded.2,57 Although mild hypothermia (32-36°C) has been suggested to lower ICP in different diseases with cerebral injury, there is only limited data on hypothermia application for ICP-control after aSAH [58,59]. As most studies reported high complication rates (e.g. infections, electrolyte disturbances, heart failure, arterial hypotension, arrhythmias, and coagulation disorders), hypothermia should only be considered in individual cases with pharmacologically refractory ICP elevation [58,59].

EEG monitoring, and early and/or prophylactic seizure treatment

Epileptic seizures are common complications of aSAH, affecting up to 26% of patients [2]. Based on seizure occurrence during the disease course, guidelines distinguish between onset-seizures (at the time of hemorrhage), early (during the first week), and late seizures (after week 1) [2].

Continuous EEG monitoring is recommended in every SAH patient for seizure monitoring.

Recommendations and options

Recent technical advances have made cEEG monitoring feasible, allowing for real-time detection of adverse events [60]. Although available studies on cEEG are very heterogeneous, a systematic literature review found improved detection and increased numbers of subclinical seizures by cEEG [60]. Therefore, current guidelines consider cEEG monitoring reasonable, especially in patients with decreased level of consciousness, fluctuating neurological deficits, or characteristics that have been associated with elevated seizure incidence, i.e. high-grade SAH, middle cerebral artery aneurysm and ICH [2].

Considerations

Although benefits of cEEG monitoring have been suggested, 2,60 and the use of quantitative software programs allow for timely interpretation of compressed raw data, 61 insufficient evidence exists whether more aggressive treatment based on these recordings leads to improved clinical outcome [60]. Overall, a full EEG montage should be preferred over a EEG bispectral index monitoring (BIS EEG) and data should be interpreted by experienced neuro-critical care physicians [60].

Seizures should be treated while prophylactic ASM is not recommended.

Recommendations and options

Onset seizures have clearly been associated with poor outcome after aSAH. Especially to prevent non-convulsive status or re-bleeding of unsecured aneurysms, Anti-Seizure Medication (ASM) should be administered upon presentation [2]. As there are no RCTs on duration of ASM use in patients with onset seizures, guidelines recommend treatment for ≤7 days to decrease long-term medication side effects [2]. Although administration of prophylactic ASM has in the past been a standard protocol, recent guidelines recommend against routine peri-interventional ASM use in ruptured aneurysms [2]. Yet, specific patient groups are at high seizure risk after aSAH, i.e. patients with higher aneurysm grade, lower Glasgow Coma Scale score at presentation, higher extent of subarachnoid blood on CT, younger age, those with middle cerebral artery aneurysms, and coexisting ICH [62]. In order to minimize medical complications, current guidelines therefore consider short term prophylactic ASM use appropriate within this specific patient group [2].

Considerations

Besides clear, RCT-proven benefits, ASM after aSAH may induce serious adverse effects such as arterial hypotension, organ toxicity, prolonged ventilator support, and overtreatment may potentially harm patients [60]. Phenytoin for seizure prevention has been associated with increased morbidity and mortality [2]. Amongst other studies, a systematic review of literature found phenytoininduced fever, deranged liver function, thrombocytopenia, rash, and Stevens-Johnson syndrome in over 20% of treated patients, 62 so that current guidelines state that the side-effect profile of phenytoin outweighs seizure prophylaxis and therapy benefits of this drug [2]. When comparing different durations of prophylactic antiepileptic medication after aSAH, a systematic review and metaanalysis with a total of 959 patients found that long-term exposure to prophylactic ASM (>3 days) was associated with poor clinical outcomes (OR 1.55; 95% CI 1.01–2.39; p=0.045) compared to short-term use [63]. Therefore, prophylactic treatment for >7 days in the management of delayed seizure risk is not recommended [2] (Table 1).

Figure 1: Schematic illustration of recommendations and options in the post-interventional care of aSAH patients. Colors distinguish between recommendations (green) and options (blue) in the post-interventional care of aSAH patients. DCI: delayed cerebral ischemia. CT: Computed Tomography. ICP: Intracranial Pressure. EEG: Electroencephalography. BP: Blood Pressure. CSF: Cerebrospinal Fluid.