Delayed cerebral ischaemia after subarachnoid haemorrhage: pathophysiology, diagnosis and neurotherapeutic opportunities

Abstract

Delayed cerebral ischaemia is a serious complication following aneurysmal subarachnoid haemorrhage (aSAH), leading to poor outcomes. This review discusses the pathophysiology of DCI, focusing on the role of cerebral vasospasm, reduction in the cerebral blood flow and the factors that affect the vasospasm and DCI. Biomarkers play a crucial role in DCI’s recognition, risk stratification, and outcome prediction. Goal-directed management aims to ameliorate the reduced cerebral blood flow, restore circulation and substrate delivery, and protect neurons. These strategies include the maintenance of cerebral blood flow and restoring substrate delivery by optimising oxygenation and maintaining normoglycaemia. Neuroprotective agents, such as calcium channel blockers like nimodipine, play a crucial role by reducing neuronal calcium influx and improving collateral circulation. Multimodal monitoring, incorporating techniques like transcranial Doppler, continuous EEG, and intracranial pressure monitoring, is essential for guiding treatment decisions and evaluating the efficacy of interventions. In the future, research should focus on a comprehensive approach to DCI management with pharmacological interventions and individualised treatment strategies.

1. Introduction

Delayed cerebral ischaemia (DCI), a consequence of aneurysmal subarachnoid haemorrhage (aSAH), has significant implications on mortality and morbidity despite advancements in aSAH care. Delayed cerebral ischaemia refers to a new focal neurological deficit or a two-point reduction in the Glasgow Coma Score (GCS) for at least one hour without identifiable aetiology. A drop in GCS assessing consciousness can indicate worsening ischaemia, although it is not specific to DCI. A lower GCS score on admission is generally associated with a poorer prognosis and may reflect the severity of initial ischaemia, not DCI [1]. Radiographic evidence of vasospasm is observed in approximately 50–70% of individuals affected by aSAH, with roughly 30% displaying associated neurological impairments. This ultimately leads to sustained neurocognitive impairments in half of the survivors [2–7]. Angiographic vasospasm is recognised by the presence of narrow cerebral arteries on a cerebral angiogram, with the degree of narrowing quantified as a percentage reduction in vessel diameter. Using transcranial Doppler ultrasonography, sonographic vasospasm measures blood flow velocity in cerebral arteries, with an increased velocity suggesting vasospasm, with specific velocity thresholds defining vasospasm. While both methods assess vasospasm, they do not always perfectly correlate. Angiography is considered the gold standard but is invasive, whereas TCD is less precise and operator-dependent but is repeatable when studying temporal changes. In the last decade, numerous attempts have been made to understand the pathobiology and develop novel neurotherapeutic options to treat DCI. While cerebral vasospasm has long been considered the primary cause of DCI, recent research suggests that other mechanisms, such as microthrombosis, cortical spreading depolarisations, and inflammation, also play significant roles. These mechanisms can contribute to neuronal injury and worsen the neurological outcome. Clinically, there is an increasing recognition of the heterogeneity of the DCIs, ranging from subtle cognitive deficits to large vessel strokes. Advanced neuroimaging techniques may help us to identify these subtle changes to improve our understanding of DCI [8].

While DCI typically occurs between 4 and 14 days after the haemorrhage, large blood vessel narrowing, or vasospasm, can occur as early as 3–4 days after SAH, typically peaking around 7–10 days. However, it can sometimes occur later or even be delayed for several weeks. While traditional grading scales like Hunt & Hess, Fisher grade, and WFNS are associated with overall outcomes after SAH, they are not reliable predictors of DCI specifically [9]. Recent studies have highlighted the importance of coma as a strong predictor of mortality, reflecting the severity of the initial brain injury [10]. Other factors, such as the amount of blood seen on initial imaging and early brain injury, are also being investigated as potential predictors of DCI [8]. Cerebral ischaemia following aneurysmal SAH is categorised into three phases: early brain injury, occurring within 72 h and primarily caused by the initial haemorrhage and resulting mechanical and physiological insults; delayed cerebral ischaemia, typically occurring between 3 and 14 days after SAH and thought to be caused by a combination of vasospasm, microthrombosis, inflammation, and spreading depolarisations; and chronic ischaemia, which can occur weeks or months after SAH and may be related to persistent vasospasm, microvascular dysfunction, or other long-term consequences of the initial injury.

2. Pathophysiology

Delayed cerebral ischaemia after subarachnoid haemorrhage involves several intertwined pathophysiological mechanisms (Fig. 1). Early brain injury initiates a cascade of events, including ischaemia, inflammation, neuronal death, and blood–brain barrier disruption, setting the stage for DCI by increasing the brain’s vulnerability to subsequent insults like vasospasm and microthrombosis [11]. EBI also disrupts cerebral autoregulation, hindering the brain’s ability to maintain adequate blood flow, further contributing to DCI [8]. Large-vessel vasospasm, traditionally considered the primary culprit in DCI, occurs due to the narrowing of large cerebral arteries, typically 3–14 days post-SAH, reducing cerebral blood flow and leading to ischaemia. [12] This vasospasm is triggered by spasmogenic substances released from the blood clot surrounding the ruptured aneurysm [13]. Concurrently, microthrombosis and microcirculatory dysfunction exacerbate the ischaemic cascade [8]. Microthrombi formation within small brain vessels further restricts blood flow, particularly in microcirculation [14], compounding the effects of large-vessel vasospasm. Dysfunction of the microcirculation, including impaired autoregulation and endothelial dysfunction, further compromises blood flow and tissue oxygenation [8].

Fig. 1.

Pathophysiology of delayed cerebral ischaemia. This diagram summarises the cascade of events that results in DCI after a SAH. (aSAH aneurysmal subarachnoid haemorrhage; TCI transient cerebral ischaemia; Hb haemoglobin; RBCs red blood cells; BBB blood–brain barrier; NO nitric oxide; CBF cerebral blood flow; ICP intracranial pressure; NMDA N-methyl-D-aspartate; DCI delayed cerebral ischaemia; K + potassium ions). Created with Biorender.com

Beyond vasospasm and microcirculatory compromise, cortical spreading depolarisations play a critical role in DCI. These neuronal and glial depolarisation waves propagate across the cerebral cortex, causing significant metabolic stress and exacerbating neuronal injury in the ischaemic environment. [9]. The inflammatory response triggered by SAH further contributes to DCI by impairing microcirculation and promoting neuronal [8]. Inflammatory mediators increase blood–brain barrier permeability, leading to oedema and further compromising cerebral blood flow. This inflammatory cascade can also trigger neuronal apoptosis, contributing to neurological deficits [8]. The brain’s autoregulatory mechanism, responsible for maintaining stable blood flow despite systemic blood pressure fluctuations, is disrupted following SAH, rendering the brain vulnerable to blood pressure changes and increasing the risk of ischaemia [8, 15].

The combined effects of these mechanisms can lead to a critical reduction in cerebral blood flow, culminating in DCI [15]. This complex interplay necessitates a multi-pronged management approach, targeting large-vessel vasospasm and other contributing factors, including microcirculatory dysfunction, CSDs, inflammation, and impaired autoregulation. Further research is crucial to fully elucidate the interplay of these mechanisms and develop more effective therapeutic strategies [16]. This improved understanding will pave the way for neurotherapeutic opportunities to mitigate the devastating consequences of DCI after SAH. Non-invasive neuromonitoring techniques, such as near-infrared spectroscopy, may guide targeted interventions for early detection and management [9].

3. Diagnosis and monitoring of DCI

DCI is clinically relevant and associated with poor clinical outcomes, neuro-cognitive impairment, and reduced quality-of-life metrics. DCI primarily involves both large and small cerebral arteries, impacting both macrocirculation and micro-circulation [8]. Vasospasm and DCI are interlinked because arterial narrowing can result in decreased cerebral blood flow (CBF), leading to cerebral hypoperfusion and ischaemia. Differentiating vasospasm and DCI in clinical practice is essential, as not all patients with imaging evidence of arterial narrowing develop DCI, and up to 50% of patients with DCI have no imaging evidence of angiographic vasospasm [17, 18].

DCI is currently diagnosed with clinical examination, trans-cranial Doppler (TCD), computed tomography angiography (CTA) and computed tomography perfusion (CTP) [18, 19]. CT scans detect DCI-related cerebral infarcts as hypodense areas. While CT can show oedema (hypodense) and hyperaemia (subtle hyperdensity or loss of grey-white differentiation), distinguishing them can be difficult. Initial CT is done on admission; follow-up CTs are performed for clinical deterioration, suggesting DCI or other complications. Frequency depends on patient stability and suspicion of new deficits. MRI is more sensitive than CT for detecting early ischaemic changes, including diffusion-weighted imaging abnormalities before CT-visible infarction. While MRI is beneficial when clinical DCI suspicion is high despite negative or inconclusive CT, its availability is limited, and it may be contraindicated in some patients. MRI is typically performed when clinical or TCD findings suggest DCI, but CT is inconclusive, providing better visualisation of the subtle changes. Due to longer scan times and patient stability concerns, MRI is less frequent than CT, generally performed as clinically warranted, especially with significant clinical changes or ongoing ischaemia concerns. Integrating these neuroimaging modalities facilitates a more comprehensive understanding of the pathophysiological processes underlying DCI, enabling earlier detection and appropriate management. CT and digital subtraction angiography guide endovascular treatment decisions for vasospasm, such as angioplasty. Imaging also helps assess the extent of infarction and guide supportive care.

CSF biomarkers may help diagnose and treat patients with vasospasm and DCI. Several biomarkers are being investigated for DCI, including those related to neuronal injury (S100B, NSE), inflammation, and blood–brain barrier disruption. The optimal sampling frequency and their role in identifying high-risk individuals are still under investigation. A multimodal approach to monitoring using brain tissue oxygenation and continuous microdialysis may be helpful in the early detection of DCI [20–22], especially in the high-risk population. These patients, hence, should be managed in high-volume institutions experienced in managing SAH with defined imaging protocols in place to distinguish treatment-related infarction from DCI in the clinical course (Fig. 2).

Fig. 2.

Prevention and management of DCI in resource-poor environments and tertiary centres. DCI delayed cerebral ischaemia; PO per os (by mouth); CPP cerebral perfusion pressure; GCS Glasgow Coma Score; NIHSS National Institute of Health Stroke Score; TC(C)D transcranial (colour) doppler; DSA digital subtraction angiography; CTP/CTA computed tomography perfusion/angiography. Created with Biorender.com

TCD monitors cerebral blood flow velocities, which are elevated in vasospasm. TCD monitoring should typically be started on the first day after SAH and repeated regularly, with frequency depending on the institution’s protocol. Brain tissue oxygen tension (PbtO₂) monitoring detects cerebral ischaemia with invasive probes in the brain parenchyma. Low PbtO₂ values suggest inadequate oxygen delivery and may indicate DCI or vasospasm. Interpretation requires considering other factors like blood pressure and arterial oxygen saturation. Global cerebral ischaemia, affecting the entire brain due to reduced systemic blood flow (e.g. cardiac arrest), contrasts with territorial ischaemia, affecting a specific arterial territory due to focal occlusion (e.g. thromboembolism); DCI typically manifests as territorial ischaemia. Differentiating these requires integrating clinical, imaging (CT/MRI), and physiological (EEG, regional cerebral blood flow, brain tissue oxygen monitoring) data. Global ischaemia presents diffuse deficits and widespread changes, while territorial ischaemia manifests with focal deficits and localized abnormalities. Multimodality monitoring is particularly valuable, especially in DCI after SAH.

VASOGRADE is a simple risk stratification grading scale, which allows DCI risk stratification on presentation after subarachnoid haemorrhage. The VASOGRADE scale combines the World Federation of Neurosurgical Societies (WFNS) and the modified Fisher score [23], allowing tailoring of monitoring strategies, decisions on the length of ICU stay, and adjustment of treatment support.

In conclusion, diagnosing DCI involves a multifaceted approach encompassing clinical assessment, neuroimaging, and physiological monitoring. While CT remains the initial imaging modality for detecting cerebral infarcts, MRI offers greater sensitivity for early ischaemic changes and differentiating oedema from hyperaemia. TCD aids in identifying vasospasm, while PbtO₂ monitoring can detect critical reductions in brain tissue oxygenation. The use of biomarkers for DCI is an evolving field, with ongoing research exploring their potential for risk stratification and early detection. Integrating these diagnostic modalities allows for a comprehensive assessment of DCI, guiding timely and targeted interventions to improve patient outcomes. Further research is needed to refine diagnostic criteria and develop more sensitive and specific tools for early DCI detection.

4. Imaging techniques for detecting DCI

Transcranial Doppler (TCD)

TCD is a non-invasive, readily available, low-cost tool that can be performed at the bedside, allowing sensitive assessments and progression of the significant cerebral vessel vasospasm. It works on the principle that flow velocity increases in narrowed blood vessels compared to non-stenosed adjacent vessels. TCD is less sensitive to DSA, CT and MRI-based techniques and requires operator experience to acquire scan windows [24]. Typically, three acoustic windows are used: transtemporal (for middle, anterior, posterior cerebral and terminal internal carotid arteries), transforaminal (for the vertebrobasilar system), and transorbital (for the ophthalmic and cavernous segment of the internal carotid arteries). The vessels are evaluated using greyscale and colour Doppler flow imaging, and spectral waveforms are acquired from proximal, mid, and distal middle cerebral arteries. Diagnostic mean flow velocity (MFV) thresholds for evidence of vasospasm are mild (120–150 cm/sec), moderate (150–200 cm/sec), and severe (> 200 cm/sec). Diagnostic thresholds for the anterior cerebral artery and posterior cerebral arteries for the diagnosis of vasospasm are 80 cm/s and 85 cm/s, respectively [25]. If routine TCD follow-up is to be performed, it needs sequential serial assessments; a single value has a poor predictive value. A recent meta-analysis found TCD to have high sensitivity (90%) and negative predictive value (92%) for detecting vasospasm in patients with aSAH [26]. The American Heart Association (AHA) and American Stroke Association (ASA) recommend that TCD is useful in diagnosing and monitoring therapeutic interventions in arterial vasospasm (Class IIa; Level of Evidence B). Traditional transcranial Doppler (TCD) criteria for vasospasm severity, based on mean flow velocity increases, are often insufficient due to the influence of factors like age, blood pressure, and collateral circulation. Current research focuses on more nuanced approaches, including the Lindegaard Ratio to normalise for individual variations, serial TCD measurements to track velocity changes over time, and combining TCD with other modalities like neuroimaging and brain tissue oxygen monitoring for a more comprehensive assessment of cerebral haemodynamics. Furthermore, advanced TCD techniques such as transcranial color-coded duplex sonography offer additional information about flow patterns and vessel morphology, potentially enhancing vasospasm detection [27].

Computed tomography angiography (CTA) and computed tomography perfusion (CTP)

CTA/CTP are sensitive and specific for diagnosing DCI and planning targeted therapies for vasospasm and perfusion deficits (Fig. 3) [18, 28]. CTA is widely available without complex post-processing of the images. In a meta-analysis, CTA-based studies suggested a sensitivity of 80% and specificity of 93% with better diagnostic accuracy for proximal segmental vasospasm [29]. CTP is done by continuously acquiring images during the washing in and out phases following an intravenous bolus of iodinated contrast. It needs dedicated post-processing; commonly mapped parameters are mean transit time (MTT), CBF and cerebral blood volume (CBV) to discriminate the presence of DCI. A systematic review and meta-analysis for CTP suggested a pooled sensitivity and specificity for DCI to be 0.84 and 0.77, respectively. [30] AHA/ASA recommends perfusion imaging with CT or magnetic resonance, which can help identify regions of potential brain ischaemia (Class IIa; Level of Evidence B). CBF < 25 ml/100 gm/min in CTP is a concerning sign of developing clinically relevant DCI. Perfusion imaging is advantageous over other radiographic modalities as it can distinguish blood flow from velocity and vessel calibre.

Fig. 3.

A computed tomography angiogram showing right anterior cerebral artery vasospasm; B computed tomography perfusion showing reduced cerebral blood flow; C computed tomography perfusion showing reduced mean transit time

A patient with a decreased flow pattern on perfusion imaging but unremarkable large-vessel imaging and unremarkable TCD might still be at high risk for impending DCI. Conversely, patients with a reassuring blood flow perfusion scan but large vessel vasospasm in imaging and higher TCD velocities may not need escalation in their management. There are many challenges to implementing perfusion imaging in SAH patients, like the appropriate timing, modality, and interpretation of the scans. The frequency of CTA or CTP studies depends on the clinical scenario and the suspicion of DCI. These studies are typically performed when there is a change in the patient’s neurological status or concern for evolving ischaemia. While no strict guidelines exist, daily or every-other-day imaging might be considered in high-risk periods. Newer modalities like multimodality monitoring, including near-infrared spectroscopy [31, 32], offer continuous monitoring of cerebral oxygenation and may help detect early signs of ischaemia before they become apparent on CT or MRI. Intraparenchymal NIRS is another monitoring tool for early detection of DCI. A combined intraparenchymal near-infrared spectroscopy and intracranial pressure (NIRS-ICP) probe may be a valuable tool for the early detection of cerebral perfusion deficits and impending DCI. [33]

Digital subtraction angiography (DSA)

Digital subtraction angiography is the gold standard for visualising cerebral vasculature and confirming vasospasm. However, it is an invasive procedure typically reserved for cases where endovascular treatment is considered. DSA is not routinely used for the primary diagnosis of DCI itself but rather for identifying and treating the underlying vasospasm that may be contributing to DCI [34].

In the ICU, magnetic resonance imaging (MRI) can be challenging due to patient instability and the need for specialised MRI-compatible monitoring devices and equipment. Motion artefacts can degrade image quality, and some patients may have contraindications to MRI, such as implanted metallic devices. Furthermore, transporting critically ill patients to the MRI suite can be logistically complex and pose risks. Despite these challenges, MRI remains a valuable tool in selected ICU patients when the diagnostic information outweighs the risks and logistical hurdles.

5. Measuring local and global physiology for monitoring a DCI

Continuous electroencephalogram (cEEG)

cEEG is a promising tool for assessing electrical changes due to ischaemia of brain tissue. Continuous monitoring detects changes associated with cortical ischaemia, slowing intrinsic rhythm when an ischaemic threshold is reached with a shift from the alpha to delta frequency. This is related to corresponding ischaemic changes at the cellular level [35]. As angiography confirms, an alpha-delta ratio and relative alpha variability strongly correlate with DCI and vasospasm [36]. Furthermore, various cEEG patterns also help in prognostication, like nonconvulsive status epilepticus, generalised periodic epileptiform discharges, and periodic lateralised epileptiform discharges, correlate with poor prognosis, especially in high-grade aSAH [37]. Many high-volume centres use cEEG as an adjunct in monitoring for DCI, especially in poor grade aSAH. Newer advances include using cEEG to detect spreading depolarisations the transient waves of neuronal silencing that can precede and contribute to DCI. Dreier et al. and Meinert et al. discuss the role of SDs in ischaemia after subarachnoid haemorrhage and the use of less-invasive subdural electrocorticography for their detection [9, 38].

Near-infrared spectroscopy (NIRS)

A simple, practical, non-invasive, bedside technology for monitoring brain tissue oxygen saturation, change in haemoglobin volume, and an indirect tool for CBF. NIRS senses decreased cortical oxygen saturation and total haemoglobin, translating into predictable values. NIRS probes are attached to the scalp with adhesive pads, and continuous monitoring is obtained. NIRS technology allows the selective extraction of samples beyond a specific depth. NIRS uses two photodetectors for each light source, subtracting far-field detection from the near field and providing the selective tissue oxygenation reading. NIRS monitors continuous cortical blood oxygen saturation, total brain haemoglobin, oxyhaemoglobin, and deoxyhaemoglobin, which is helpful with other multimodal assessments [20, 21]. NIRS shows potential as a valuable instrument for early detection of DCI in individuals with poor-grade SAH at the bedside [39, 40]. NIRS has several limitations; the measurements of haemoglobin concentrations are averages and represent only a tiny area of brain tissue; the venous oximetry is associated with interference from nearby structures like the scalp, cerebral venous sinuses, skull [41, 42].

Brain tissue oxygenation (PbO₂)

Invasive brain tissue oxygenation monitoring involves placing an oxygen-sensitive probe (a Clark electrode) directly into the brain tissue. The electrode is inserted into a small opening in the skull. It is then positioned in the area of interest to continuously measure the partial pressure of oxygen in the tissue (PbO2). Invasive PbO2 monitoring is considered the gold standard for measuring brain tissue oxygenation because it provides direct and precise measurements of oxygen concentration in the brain tissue. [31] A level < 20 mmHg PbO₂ suggests focal cerebral hypoxia, and Ievels < 15 mmHg require an immediate intervention to optimise cerebral tissue oxygenation [32]. Disadvantages include invasiveness. While PbO2 monitoring does provide a localised measurement, its clinical utility stems from its ability to reflect the metabolic status of at-risk tissue, guide therapy, and correlate with global outcomes.

Microdialysis of the brain

Microdialysis monitoring is an invasive technique using microdialysis catheters which allows continuous bedside measurements of extracellular concentrations of lactate, pyruvate, lactate/pyruvate ratio, glutamate, glucose, and glycerol in brain tissue, thereby screening for anaerobic metabolism, excitotoxic cell injury, ischaemia, and cell death [31]. The characteristic cut-off values are lactate (≥ 4.0 μmol/L), lactate/pyruvate ratio (> 40) indicative of anaerobic metabolism), glutamate (> 10 μmol/L), glucose (< 0.2 μmol/L), and glycerol (> 90 μmol/L) [16]. The disadvantages include invasiveness and the fact that only a small area of tissue is sampled, but it can be used with cerebral oximetry to detect ischaemia in patients with SAH. The most reliable markers for predicting ischaemia are glutamate and lactate, as their values peak 24 h before clinical ischaemia, followed by glycerol concentrations, which peak 12 h before clinical ischaemia. Furthermore, raised lactate/glutamate and lactate/pyruvate ratios are good prognostic indicators in SAH [43]. High lactate alone does not indicate ischaemia but signifies a hypermetabolic state that may show an improved outcome. The microdialysis catheter should be placed in an area at risk for DCI. The values for up to 6 h after insertion are unreliable due to the insertion effect. Brain tissue oxygen monitoring and microdialysis are used to measure cerebral metabolism and detect ischaemia in real time. These techniques are typically employed in patients at high risk of DCI, and measurements are often taken continuously. When ischaemia is detected, therapeutic interventions, such as increasing blood pressure or administering vasopressors, are initiated to improve cerebral perfusion. New advances include combining microdialysis with other monitoring modalities, such as EEG and near-infrared spectroscopy, to provide a more comprehensive picture of cerebral function and metabolism.

CSF biomarkers

Ongoing research in molecular biology suggests other reliable CSF biomarkers, including miRNA, mRNA, fibrin degradation product (FDP), soluble receptors, calcium-binding protein (CBP) and high-mobility group box one protein (HMGB1) [44]. Markers of endothelial dysfunction include serum (MMP-9), VWF, and vascular endothelial growth factor (VEGF); their elevated levels were associated with an increased chance of development of DCI [45]. Neuron-specific enolase (NSE) and S100β levels measured in CSF and blood are sensitive biomarkers for predicting in-hospital mortality in patients with aSAH; the exact cut-off levels are still under investigation [46–48]. A strong pathophysiological association exists between cell-free haemoglobin found in CSF (CSF-Hb) and consequent secondary brain damage following a subarachnoid haemorrhage [49]. If one is in place, CSF biomarkers can be sampled via lumbar puncture or from an external ventricular drain. The frequency of sampling depends on the specific biomarker and the clinical scenario. Some biomarkers, such as S100B and neuron-specific enolase, may be measured daily or every other day, while others may be measured less frequently. Akeret et al. discuss a study protocol for validating cerebrospinal fluid haemoglobin as a monitoring biomarker for aneurysmal subarachnoid haemorrhage-related secondary brain injury [49].

6. Management

Early detection is the key to preventing DCI. Management goals include optimising CB, reducing cerebral metabolic demand (CMD) and preventing secondary brain injury (Fig. 4). The initial step is to optimise the rheology of blood flow by manipulating blood pressure and oxygen delivery by optimising haemoglobin targets and volume status. Extensive trials on popular”3H”therapy (hypertension, hypervolaemia, and haemodilution) revealed the benefit of targeted euvolemia (AHA/ASA recommendation class 1 Level B) while avoiding hypervolaemia and hypovolaemia. Fluid management in aSAH patients, particularly those at risk for DCI, aims to maintain adequate cerebral perfusion while avoiding complications like hyponatremia and cerebral oedema. Crystalloids, such as isotonic saline (0.9% NaCl), are commonly used for initial resuscitation and volume replacement. Hypertonic saline solutions (e.g., 3% NaCl) may be considered in specific situations, such as symptomatic hyponatremia or to reduce cerebral oedema, but require careful monitoring of serum sodium levels. Colloids, such as albumin, remain controversial due to concerns about increased risk of cerebral oedema and worse outcomes in some studies. The choice of fluid should be individualised based on the patient’s specific clinical status, electrolyte balance, and haemodynamic parameters [50–53]. Hyponatraemia, likely due to cerebral salt wasting (CSW) and a cause for volume contraction, is independently associated with poor outcomes. Optimal blood pressure targets during vasospasm treatment in DCI/SAH are a subject of ongoing debate and may vary based on individual patient characteristics. Generally, higher MAP targets are often employed to improve cerebral perfusion. A common target range for MAP during vasospasm is 80–100 mmHg or even higher in some cases. CVP targets typically aim to maintain adequate preload and may range from 8 to 12 mmHg, but this can vary. It is crucial to individualise blood pressure targets based on the patient’s clinical status, presence of other medical conditions, and response to therapy. Continuous monitoring of cerebral perfusion pressure (CPP), if available, can help guide blood pressure management [16, 53, 54]. Avoid hypotension at any cost: keep MAP > 65; other targets include keeping CPP ≥ 60 mmHg. Inotropic agents are often required to augment blood pressure and improve cerebral perfusion in patients with DCI/SAH, particularly those with vasospasm. Commonly used agents include norepinephrine, dopamine, and vasopressin. Norepinephrine is often considered a first-line agent due to its potent vasoconstrictive effects with relatively less impact on heart rate. Dopamine, while also increasing blood pressure, can have variable effects on heart rate and may increase the risk of arrhythmias. Vasopressin can be useful in patients with refractory hypotension or those requiring high doses of other vasopressors. Each agent has potential advantages and disadvantages, including effects on cardiac output, heart rate, and potential for arrhythmias. The choice of agent should be individualised based on the patient’s haemodynamic profile and clinical status [55, 56].

Fig. 4.

Guidance flow diagram for preventing and managing delayed cerebral ischaemia (DCI). EEG electroencephalograph, TCD transcranial doppler, TCCD transcranial colour doppler, PtiO2 brain tissue oxygen saturation, CTP computed tomography perfusion, CTA computed tomography angiography, GSC Glasgow Coma Score, NIHSS National Institutes of Health Stroke Scale, FVMCA flow velocity in the middle cerebral artery, LPR lactate-pyruvate ratio, CBF cerebral blood flow, DSA digital subtraction angiography, CPP cerebral perfusion pressure, ICP intracranial pressure, CSF cerebrospinal fluid, PO per os (by mouth). Created with Biorender.com

Oral nimodipine is the only therapy for aSAH shown to have a clear association with improved neurological outcomes; stroke guidelines recommend oral nimodipine, citing class I level A evidence. Its primary action extends beyond simply dilating large cerebral vessels; it acts as a neuroprotective agent through several mechanisms. These include improving collateral circulation by enhancing blood flow in smaller vessels and collateral pathways, thus maintaining perfusion in areas affected by vasospasm; reducing neuronal calcium influx by blocking calcium channels, which mitigates ischaemia-induced neuronal death; and exerting anti-inflammatory and antioxidant effects, further strengthening its neuroprotective properties [57]. Nimodipine (60 mg 4-hourly) should be started as early as possible. The administration of an intravenous calcium channel blocker is used in some neurointensive care, but these therapies are often associated with hypotensive episodes so should be used with caution.

Intrathecal administration of nimodipine and nicardipine, which cross the blood–brain barrier, is evaluated in trials using the sustained-release formulation, such as prolonged-release implants placed at the time of aneurysm clipping. Nicardipine, like nimodipine, is a calcium channel blocker. It primarily blocks L-type voltage-gated calcium channels in vascular smooth muscle, leading to vasodilation. This vasodilatory effect can help improve cerebral blood flow in patients with vasospasm. Early studies of intrathecal nicardipine have shown encouraging results [58]. Limitations include the need for surgical implantation and prolonged therapy. The growing use of interventional therapies could potentially explore options for implanting sustained-release formulations without surgical implantation [59].

Within a limited number of uncontrolled, non-randomised groups, intravenous milrinone seems to be a safe and viable treatment option for DCI. Intravenous milrinone was effective in treating clinical and angiographic vasospasms in more than two-thirds of cases, with hypotension and tachycardia being the most frequent haemodynamic side effects [60].

The volume of subarachnoid blood is a strong predictor for the development of DCI. Meta-analysis on enhanced clearance of bloody CSF by instilling tPA into the cisterns and ventricles suggests potential benefits [61, 62]. Large randomised controlled trials (RCTs) showed a risk reduction for significant DCI [63]. Alternatively, tPA instilled into the ventricular drain could help clear clots and blood from CSF. Similarly, in patients undergoing IR-guided coiling, instilling urokinase via microcatheters inserted in the lumbar region and navigated up to the cisterns reduced DCI. [64] Another approach for clearing blood in CSF is through lumbar CSF drainage. Further large RCTs are needed to recommend such therapies.

7. Angiographic treatment

Balloon and chemical angioplasty (infusion of intra-arterial vasodilator) are two approaches to angiographic treatment [65, 66].A recent study which compared distal balloon angioplasty with chemical angioplasty found distal balloon angioplasty to be associated with a lower rate of subsequent (recurrent) vasospasm [67]. In medically refractory cases, aggressive interventional therapy (balloon angioplasty, IA vasodilator infusion, or both) should be performed with the same concern and urgency as acute ischaemic stroke: AHA/ASA Class IIa; Level of Evidence B. Per studies by Rosenwasser and colleagues, early aggressive treatment in < 2 h seems to be the most effective way to prevent cerebral infarction and improve neurologic outcomes in medically refractory patients [68]. Chemical angioplasty involves the intra-arterial administration of vasodilators to treat vasospasm. Agents commonly used include intra-arterial verapamil and milrinone. Verapamil, a calcium channel blocker, is a commonly used agent. Milrinone, a phosphodiesterase-3 inhibitor, has also shown promise in some studies. The choice of agent and the specific protocol used can vary between institutions. These agents are typically administered under fluoroscopic guidance to target the affected vessels [69]. Refractory vasospasm, defined as vasospasm unresponsive to medical management, can be challenging to treat. Treatment modalities may include repeated or higher doses of intra-arterial vasodilators, balloon angioplasty, or even consideration of alternative diagnoses. Monitoring treatment response involves assessing clinical status, including neurological examination and imaging studies (e.g., CT angiography, digital subtraction angiography). Transcranial Doppler (TCD) ultrasound can be used to monitor cerebral blood flow velocities and assess the severity of vasospasm. Continuous EEG monitoring may also help detect changes in brain activity that could indicate worsening ischaemia. Multimodal monitoring, including ICP, PbtO2, and microdialysis, can provide additional insights into cerebral physiology and help guide treatment decisions.

8. Other intravascular and oral agents

Several intraarterial, intravenous and oral vasodilators are being trialled; they increase the diameter of blood vessels without any clear outcome benefits; the complex etiopathogenesis of DCI could explain this. Clazosentan, papaverine, fasudil, colforsin daropate hydrochloride, magnesium, nitric oxide analogues, erythropoietin, sildenafil citrate, and free radical scavengers like nicaraven are among some of the medications that have shown promise in small studies. Goal-directed management protocols for DCI centre around three key areas: improving cerebral blood flow through strategies like induced hypertension, hypervolemia, and vasopressor use; restoring substrate delivery by optimising oxygenation, ventilation, and maintaining normoglycemia; and protecting neurons with neuroprotective agents such as calcium channel blockers like nimodipine and nicardipine, which target specific pathways of neuronal injury, including reducing neuronal calcium influx, improving collateral circulation, and exerting anti-inflammatory and antioxidant effects. These strategies are typically combined and personalised based on the patient’s condition, with multimodality monitoring guiding interventions and evaluating their efficacy [16, 31, 32, 59].

9. Conclusions and future directions for DCI

Delayed cerebral ischaemia remains a significant challenge in managing aSAH. Despite advancements, DCI continues to cause substantial morbidity and mortality. Current management focuses on early detection and aggressive intervention, including optimising cerebral blood flow, ensuring adequate substrate delivery, and employing neuroprotective strategies. Induced hypertension, often combined with maintaining euvolemia and vasopressor support, remains a cornerstone of DCI management, aiming to maintain sufficient cerebral perfusion pressure. Simultaneously, optimising oxygenation, ventilation, and normoglycemia are crucial for supporting neuronal function. Nimodipine, a calcium channel blocker, is a mainstay of treatment, offering neuroprotection through various mechanisms.

Beyond these established approaches, ongoing research explores additional avenues for DCI management. Intrathecal administration of calcium channel blockers like nimodipine and nicardipine, which bypass the blood–brain barrier, shows promise. For refractory vasospasm, options include repeat chemical angioplasty, endovascular rescue therapies such as intra-arterial thrombolysis or mechanical thrombectomy [70] and, rarely, surgical interventions like bypass. Chemical angioplasty, commonly employing intra-arterial verapamil (a calcium channel blocker) for vasodilation, is a primary treatment for vasospasm. Intra-arterial milrinone is also being explored, particularly for refractory DCI [60].

Future research directions include refining diagnostic tools for earlier DCI detection, exploring novel therapeutic targets beyond calcium channel blockade, and personalizing treatment strategies based on individual patient characteristics. Advanced neuroimaging techniques and biomarkers hold the potential for improving diagnostic accuracy. Promising therapeutic agents requiring further investigation include clazosentan, papaverine, fasudil, and free radical scavengers. Optimising multimodal monitoring and integrating data from various sources can enhance individualised treatment decisions.

Ultimately, effective DCI management requires a multidisciplinary approach involving early detection, aggressive intervention, and personalised strategies. Continued research efforts are essential to refine existing therapies, identify novel targets, and improve outcomes for patients affected by this devastating complication of aSAH.

Data availability

No datasets were generated or analysed during the current study.