Keywords
Subarachnoid Hemorrhage - Brain Ischemia - Critical Care - Cerebral Hemorrhage - Intracranial
Aneurysm
INTRODUCTION
Aneurysmal subarachnoid hemorrhage (aSAH) accounts for 5 to 10% of all strokes, and
it is a condition with high morbimortality rates, predominantly affecting a younger
population, which leads to significant economic and social impact.[1] Long-term disability affects between a third and half of the survivors globally.[2]
[3]
Delayed cerebral ischemia (DCI) is a complication that occurs in approximately 30%
of aSAH patients, most commonly between 3 to 14 days after symptom onset,[4]
[5] and it represents a leading cause of poor outcomes and increased costs for those
who survive the initial bleeding.[6]
[7] This is a potentially reversible condition that may progress to permanent infarction
if not properly treated. Death and permanent neurological deficits occur in 30 to
40% of DCI patients, and these permanent neurological deficits have been an independent
predictor of mortality at 6 and 12 months.[8]
We herein aim to provide a synthesis and critical review of DCI's pathophysiology,
clinical presentation, and management strategies, highlighting which aspects have
more impact on patient outcomes. This includes a comprehensive examination of early
diagnostic methods, predictors, and the effectiveness of various therapeutic interventions.
Additionally, we will identify potential research gaps and areas that require further
investigation, ultimately offering clinicians an updated, evidence-based narrative
review.
DEFINITIONS
Historically, various terms have been used to describe DCI. Therefore, in 2010, Vergouwen
et al. proposed more precise definitions.[9] Later, in 2019, a multidisciplinary, international panel established additional
terminologies associated with it ([Table 1]).[10]
Table 1
Terminologies associated with DCI and aSAH[4]
[9]
|
Terminology
|
Definition
|
|
Clinical deterioration due to DCI
|
The occurrence of focal neurological impairment (such as hemiparesis, aphasia, apraxia,
hemianopsia, or neglect), or a decrease of at least 2 points on the Glasgow coma scale,
either on the total score or on one of its individual components (eye, motor, verbal).
Should last for at least 1 hour, is not apparent immediately after aneurysm occlusion,
and cannot be attributed to other causes by means of clinical assessment, brain CT
or MRI, and appropriate laboratory studies.
|
|
Cerebral infarction due to DCI
|
The presence of cerebral infarction on a CT or MRI scan of the brain within 6 weeks
after aSAH, on the latest CT or MRI scan made before death within 6 weeks, or proven
at autopsy. Should not be present on the CT or MRI scan between 24 and 48 hours after
early aneurysm occlusion and should not be attributable to other causes such as surgical
clipping or endovascular treatment. Hypodensities on CT imaging resulting from a ventricular
catheter or intraparenchymal hematoma should not be regarded as cerebral infarctions
from DCI.
|
|
Angiographic cerebral vasospasm
|
The arterial narrowing of large cerebral vessels is observed on radiological tests
such as CT angiography, MRA, or digital subtraction angiography.
|
|
Symptomatic cerebral vasospasm
|
Patients with aSAH develop clinical symptoms attributable to ischemia from visible
vasospasm on angiography.
|
Abbreviations: aSAH, aneurismatic subarachnoid hemorrhage; CT, computed tomography;
DCI, delayed cerebral ischemia; MRI, magnetic resonance image; MRA, magnetic resonance
angiography.
PATHOPHYSIOLOGY
Intracranial vasospasm was once considered the sole cause of DCI. However, it involves
various underlying pathophysiological mechanisms such as cerebral macro- and microvascular
dysfunction, microthrombosis, cortical spreading depolarization (CSD), and neuroinflammation
([Figure 1]).[11]
[12]
[13]
Abbreviations: BBB, brain-blood barrier; NO, nitric oxid; ADAMTS13, a disintegrin and metalloproteinase
with a thrombospondin type 1 motif, member 13; CBF, cerebral blood flow; RBC, red
blood cells; TNF, tumor necrosis factor type α; MyD88, Myeloid differentiation primary
response 88; MAPK, mitogen-activated protein kinase; NFKB, nuclear factor kappa B;
ICAM1, Intercellular Adhesion Molecule 1; MMP9, Matrix metalloproteinase 9; TRL4,
toll-like receptor 4; IL6, interleukin 6; IL 1 β, interleukin 1 β. Figure 1 Pathophysiology of DCI.
Vascular dysfunction
In macrovascular dysfunction cases, cerebral vasospasm is defined as “the arterial
narrowing of large cerebral vessels observed on radiological tests such as CT angiography
(CTA), magnetic resonance angiography (MRA), or digital subtraction angiography (DSA)”
([Table 1]).[9] The release of hemoglobin and hemolysis-mediated degradation products in subarachnoid
space triggers reactions with oxygen free radicals, inflammation, endothelial injury,
an increase in endothelin-1 (a vasoconstrictor), and a decrease in the vasodilator
nitric oxide (NO),[12] resulting in cerebral vasospasm. This phenomenon is detectable only by specialized
diagnostic methods ([Figure 2]) and occurs in up to 70% of patients after aSAH. It typically begins 3 to 4 days
after aneurysm rupture, peaks at 7 to 10 days, and resolves within 14 to 21 days.[14] Conversely, DCI is observed in around 30% of aSAH patients and does not always correspond
to areas of arterial narrowing.[14] In patients with aneurysmal aSAH who underwent DSA and serial CT scanning, DCI occurred
in 3, 10, and 46% of those with no/mild, moderate, or severe angiographic vasospasm,
respectively.[11]
Figure 2 Cerebral vasospasm on digital subtraction angiography (DSA). (A) DSA showing aneurism in the left anterior cerebral artery (ACA) segment A1. (B) ACA aneurism after 3D reconstruction. (C) ACA focal narrowing suggestive of cerebral vasospasm in the same patient.
Furthermore, in microvascular dysfunctions, arterioles and capillaries undergo impaired
cerebral autoregulation, neurovascular uncoupling, and breakdown of the blood-brain
barrier, which significantly trigger microvascular spasm.[11] Alterations in the meningeal lymphatic and central nervous systems' perivascular
glymphatics contribute to cerebral ischemia. The meningeal lymphatic system is responsible
for clearing extravasated erythrocytes, so its disruption allows blood products to
quickly penetrate brain parenchyma.[14]
Microthrombosis
Aneurysmal subarachnoid hemorrhage affects multiple stages of the blood coagulation
cascade, including stasis, blood turbulence, and endothelial dysfunction (the Virchow's
triad), low NO, increased P-selectin, higher levels of platelet activation, arteriolar
constriction, neuronal apoptosis, reduced fibrinolysis, and alterations in von Willebrand
factor, among others. Evidence of microthrombosis has been consistently linked to
cerebral ischemia after aSAH in animal models and clinical studies.[11]
Cortical spreading depolarizations
Cortical spreading depolarization is a slowly propagating wave of depolarization that
moves outward from its point of origin at a speed of 2 to 5 mm per minute, leading
to both a depression of electrical activity (spreading depression) and cortical hypoperfusion
(cortical spreading ischemia).[11]
Initially, CSD leads to an increase in CBF, resulting in hyperemia, which is followed
by a transient phase of oligemia and reduced cerebral perfusion. In pathological conditions,
CSD induces arteriolar vasoconstriction and an inversion of neurovascular coupling,
generating prolonged and severe hypoperfusion. Moreover, the release of glutamate
during CSD can cause neurotoxicity through excessive stimulation and cell death.[11]
Neuroinflammation
The blood in the subarachnoid space triggers an immediate inflammatory response. The
release of red blood cell degradation products and high mobility group box-1 (HMGB1)
proteins both activate the innate immune cells (particularly microglia), contributing
to DCI.[11] Recent reviews highlighted the many factors involved, summarized in [Figure 1].[11]
[12]
It should be noted that all these factors work together, even though they are described
separately. For instance, neuroinflammation promotes vasospasm and the occurrence
of microthrombosis, both of which contribute to microvascular dysfunction. The pathophysiology
of DCI is complex, as it is further influenced by systemic and neurological complications,
such as epileptic seizures and hydrocephalus.[13]
PREDICTORS OF DCI
Estimating the risk of DCI is crucial for tailoring the vigilance required for each
patient and reducing costs associated with extended hospital stays. The historical
predictors of DCI are the volume and extension of bleeding on initial CT, evaluated
by the original and modified Fisher (mFisher) scales,[15] and the patient's neurological status upon admission, assessed through the World
Federation of Neurological Surgeons (WFNS) and Hunt-Hess scales.[1]
[4]
The first radiological scale was introduced by Fisher et al. in 1980, in which patients
classified as grade 3 had a higher risk of vasospasm/DCI compared with those with
grade 4. Later, mFisher showed that thick blood and intraventricular hemorrhage (blood
in both lateral ventricles) were the strongest predictors of vasospasm and DCI (grade
4).[4]
The VASOGRADE scale combines the WFNS and mFisher scores to classify patients with
aSAH into green, yellow, and red. Compared with green, the yellow score has an OR
of 1.31, and red has an OR of 3.19 for developing DCI/vasospasm.[16] The VASOGRADE scale was also associated with functional outcomes of aSAH patients
with high specificity. This could serve as an early indicator for transitioning patients
from the intensive care unit (ICU) to the ward, helping to reduce costs and potential
medical complications associated with extended internations.[17]
Other described predictors of DCI include female gender, diabetes mellitus, early
rise in serum C-reactive protein, preexisting hypertension, a high white blood cell
count on admission, intracranial infection, smoking, hyperglycemia, hydrocephalus,
poor collateral status on DSA, and early systemic inflammatory response syndrome.[18]
[19]
[20] Notably, the aneurysm's location and size are not associated with this condition.[18]
High levels of cerebrospinal fluid (CSF) lactate and glucose in the first 3 days following
aSAH were independent predictors.[21] Dynamic changes in systemic inflammation response index have also been associated
with DCI,[22] as have higher levels of admission N-terminal probrain natriuretic peptide (NT-pro
BNP), which were also associated with neurogenic cardiac injury.[23]
An automated electroencephalography (EEG)-based approach for DCI prediction showed
improved accuracy when combined features were analyzed (α-delta ratio and percent
α variability, Shannon entropy, and epileptiform discharge burden).[24] Furthermore, transcranial Doppler (TCD) measurements of peak flow velocity in the
middle cerebral artery (MCA) above 200 cm/s, combined with EEG epileptiform abnormalities,
can predict it better than EEG alone on day 3.[25] In the “International Subarachnoid Aneurysm Trial” (ISAT), DCI incidence was higher
after surgical clipping of the aneurysm than after endovascular coiling, probably
related to blood vessel manipulation during neurosurgical procedures.[26]
The early clot clean rate was associated with reduced incidence of DCI and outcomes
at 30 days.[27] Poor arterial collateral status on cranial CT at admission also increased the risk
for this condition in a retrospective study.[28] Regardless of absolute values, greater hemoglobin decrement from admission to discharge
was independently associated with a higher incidence of DCI and poor functional outcomes.[29]
A systematic review identified seven potential biomarkers of clinical DCI: haptoglobin
polymorphisms 2–1 and 2–2, ADAM metallopeptidase with thrombospondin type 1 motif
13 (ADAMTS13), neutrophil/lymphocyte ratio, P-selectin, von Willebrand Factor (vWF),
and F2-isoprostane in urine. Further studies are needed to standardize cutoff values
and to explore if those biomarkers could be used as preventive or therapeutic targets.[30]
Neurofilament proteins (NFs) are consistently elevated in serum and CSF upon admission
in aSAH patients and may have a relationship with disease severity, prognosis, and
mortality. The C-reactive protein (CRP), lactate, microRNAs, estrogen, galectins,
D-dimer, neuroglobin, high-mobility group box1 protein (HMGB1), periostin, and glial
fibrillary acidic protein (GFAP) also have a potential role regarding DCI and poor
prognosis.[31]
Big data and artificial intelligence
The adoption of artificial intelligence and big data analytics may improve the prediction
of DCI, which can enhance patient management and treatment strategies. Various machine
learning (ML) models, such as the Random Forest, XGBoost, Support Vector Machines,
Multilayer Perceptron, Gradient Boosting Decision Trees, and Decision Trees, have
been developed and validated across multiple centers.[32] They demonstrate higher accuracy, outperforming standard models and logistic regression
scoring systems like the VASOGRADE and the Subarachnoid Hemorrhage International Trialists
(SAHIT) prediction models. These models leverage large datasets from clinical records
to identify complex patterns and risk factors that may not be apparent through conventional
analysis. Moreover, ML models can easily combine clinical variables with image features
or biomarkers (such as heart rate variability, matricellular proteins), which increases
reliability.
Clinical variables such as age, mFisher, Hunt and Hess score, and external ventricular
drain (EVD) placement are significant predictors in ML models.[32] Multicenter collaborations and standardized data sharing are essential for further
refining these predictive models, ensuring they can be generalized across diverse
patient populations and clinical settings. As these technologies advance, they hold
the potential to significantly improve patient outcomes by enabling timely and accurate
diagnosis and intervention for DCI following aSAH.[33]
PREVENTION
Pharmacological therapy
The only pharmacological therapy with confirmed effectiveness and safety in DCI prevention
is nimodipine, a dihydropyridine L-type calcium channel antagonist.[5]
[6] This drug's exact mechanisms of action have been investigated and other effects
were proposed beyond vasodilation, including reduction of CSD, increased endogenous
fibrinolysis, and reduced microthrombosis.[34] Nimodipine reduces DCI even without improvement of vasospasm,[34] and a meta-analysis of randomized control trials (RCTs), including a total of 1,202
patients, found that nimodipine improved all eight outcome measures: good clinical
outcome, mortality, morbidity, death attributed to vasospasm, DCI, cerebral infarction,
and rebleeding.[35]
Current guidelines recommend starting nimodipine within the first 96 hours of subarachnoid
bleeding at a dose of 60 mg, every 4 hours, for 21 days (class I, level of evidence
A).[5] If arterial hypotension occurs, the dose may be reduced to 30 mg every 2 to 4 hours.[36] Regarding the route of administration, the randomized “Nimodipine Microparticles
to Enhance Recovery While Reducing Toxicity After Subarachnoid Hemorrhage” (NEWTON)
trial found no improvements in vasospasm when nimodipine sustained-release microparticles
directly into the ventricles were compared with standard oral administration.[37]
Recently, it was demonstrated that localized nicardipine release implants (NPRIs),
placed around the basal cerebral vasculature, are effective in preventing vasospasm
and DCI.[38] A recent RCT including 41 patients showed a lower incidence of cerebral vasospasm,
reduced clinical need for rescue therapy, lower rates of new cerebral infarcts, and
a higher proportion of favorable functional outcomes in the NPRIs group.[39]
One study analyzed the impact of carotid siphon calcification (CSC) on vasospasm and
outcomes, identifying a negative impact only in patients not using aspirin. This suggests
aspirin may mitigate microcirculatory impairment, indicating potential benefits of
aspirin in DCI prevention in patients with aSAH and concomitant carotid atherosclerosis.[40]
The administration of clazosentan, magnesium, statins, aspirin, enoxaparin, erythropoietin,
fludrocortisone, methylprednisolone, and prophylactic balloon angioplasty have been
tested in DCI prevention with negative results.[36] Guidelines from different countries recommend other drugs. For example, tirilazad,
a nonglucocorticoid 21-aminosteroid that inhibits lipid peroxidation and free radical
production, is approved for treating aSAH in 21 countries. Fasudil has been used in
China and Japan since 1995. Furthermore, cilostazol and clazosentan are recommended
in Japan to reduce DCI.[34]
Ongoing studies are evaluating anakinra, cilostazol, clazosentan, deferoxamine, dexamethasone,
IV heparin, isoflurane sedation, ketamine sedation, magnesium-rich artificial CSF,
IV milrinone, nadroparine, NPRIs, and intraventricular fibrinolysis with tissue plasminogen
activator.
Noteworthy, one systematic review found that pharmaceutical treatments decreased the
incidence of both cerebral infarction (relative risk [RR]: 0.83; 95%CI: 0.74–0.93)
and poor functional outcome (RR: 0.92; 95%CI: 0.86–0.98).[6] This indicates that the clinical features of DCI may evolve. Clinical diagnosis
shows a lower rate of interobserver agreement. In contrast, cerebral infarction observed
on neuroimaging is a result of DCI that may not be present in all patients but is
strongly correlated with functional outcomes 3 months after aSAH. It also exhibits
a high interobserver agreement rate, facilitating diagnosis in sedated and comatose
patients. Currently, there is a consensus that cerebral infarction is a preferable
primary outcome for observational prevention trials, with the clinical definition
of DCI reserved as a secondary outcome.[41]
Euvolemia
Maintaining euvolemia and avoidance of hypervolemia have proven benefits in preventing
DCI and improving functional outcomes. It is recommended in current guidelines as
patients with documented volume depletion have a higher chance of developing DCI.[2]
[5] Early goal-directed fluid therapy (EGDT) guided by transpulmonary thermodilution
may be an interesting option and was shown to reduce DCI.
Regarding clinical outcomes, in 2014, Mutoh et al.[42] conducted a prospective trial with 160 patients randomized to EGDT guided by preload
and cardiac output measurements or conventional therapy guided by central venous pressure
and fluid balance. Among those with poor-grade hemorrhages, DCI and the ICU length
of stay were significantly reduced with EGDT. Good outcomes (mRs 0–3 in 3 months)
were significantly higher in the EGDT group (52 versus 36%; p = 0.026).[42]
Another prospective, randomized trial with 108 patients achieved similar results:
reduced DCI and disability with EGDT.[43] However, liberal fluid administration without goal-directed targets and consequent
hypervolemia is not recommended due to the increased risk of pulmonary edema and cardiac
complications.[2]
[5]
Cerebrospinal fluid diversion
Regarding common neurological complications, CSF diversion is used to treat intracranial
hypertension and hydrocephalus[44] While external ventricular drains (EVDs) are an important measure to reduce intracranial
hypertension after aSAH, they have questionable benefits. Theoretically, the removal
of blood clots may reduce DCI. However, aggressive drainage has not been proven to
benefit.[45]
Lumbar drains have consistently been shown to reduce DCI and improve functional outcomes
after aSAH. It also reduces oxidative stress and looks more successful than EVDs for
blood clot removal.[46] An ongoing trial in the United States (NCT03065231), now recruiting, aims to answer
whether these drains are superior to EVDs for DCI prevention.
More recently, CSF filtration with neurapheresis has shown the potential to reduce
clot burden and potentially DCI after aSAH. The fluid is removed, cleaned, and returned
to the lumbar spine. However, further studies with larger populations are needed for
clarification.
DIAGNOSIS AND TRIGGERS FOR INTERVENTION
DIAGNOSIS AND TRIGGERS FOR INTERVENTION
Serial neurological examination is the gold standard for diagnosing DCI in awake patients,
which must performed every 2 to 4 hours.[47] For high-grade aSAH patients who are in a coma or sedated, additional methods to
detect DCI are recommended, including TCD, DSA, CTA, CT with perfusion (CTP), continuous
electroencephalography (cEEG), partial brain tissue oxygenation monitoring (PbTiO2),
and cerebral microdialysis (CMD).[48] In 20% of patients diagnosed with DCI-related infarctions, no clinical deterioration
was observed before the neuroimaging findings.[36]
Digital subtraction angiography and computed tomography angiography
The use of DSA is the gold standard for detecting vasospasm; however, its availability
is limited, particularly in low- and middle-income countries. Vasospasm on DSA can
be classified as mild to moderate (grade I: 0–25%; and II: 26–50% narrowing) and severe
(grade III [50–75%] and grade IV [> 75%]).[47]
[48] In contrast, CTA is a less invasive method and has frequently been used instead
of DSA. In a meta-analysis of 7 studies, comprising 1,646 arterial segments, CTA had
a pooled sensitivity of 82% (95%CI: 68–91%) and a specificity of 97% (95%CI: 93–98%),
compared with the gold standard DSA.[49] The disadvantages of CTA include lower accuracy for medium and small vessels, susceptibility
to artifacts such as metal clips and coils, and overestimation of arterial narrowing.[47]
Transcranial Doppler ultrasonography
As a non-invasive, safe, and quick method, TCD allows real-time bedside assessment
of CBF of the intracranial vessels. It is recommended in the period of highest risk
for vasospasm and DCI (3–14 days), but this time may be extended in higher risk patients.[5] Furthermore, TCD is most reliable for detecting MCA vasospasms, with less accuracy
for other arteries ([Table 2]
[3]).[50] A mean flow velocity (MFV) in the MCA > 120 cm/s or an increase ≥ 50 cm/s over 24 hours
indicates vasospasm, while an MFV greater than 200 cm/s suggests severe vasospasm
([Figure 3]).
Figure 3 Cerebral vasospasm on transcranial Doppler. (A) Transcranial Doppler showing severe vasospasm on left middle cerebral artery (MCA)
in a patient with aneurismal subarachnoid hemorrhage at 7-days postbleeding: mean
flow velocity (MVF) on MCA of 221 cm/s, MFV on left internal carotid artery (ICA)
of 30 cm/s, Lindegaard ratio (LR) of 7.3. (B) Moderate vasospasm on right MCA, with MFV of 131 cm/s, and right ICA MFV of 32 cm/s,
LR of 4.09.
Table 2
Transcranial Doppler criteria for middle cerebral artery vasospasm
|
Mean flow velocity (cm/s)
|
Lindegaard index
|
Interpretation
|
|
> 120
|
≤ 3
|
Hyperemia
|
|
≥ 120
|
3–4
|
Light spasm + hyperemia
|
|
≥ 120
|
4–5
|
Moderate spasm + hyperemia
|
|
≥ 120
|
5–6
|
Moderate spasm
|
|
≥ 180
|
> 6
|
Moderate to severe spasm
|
|
≥ 200
|
≥ 6
|
Severe spasm
|
|
> 200
|
4–6
|
Moderate spasm + hyperemia
|
|
> 200
|
3–4
|
Hyperemia + light spasm (often residual)
|
|
> 200
|
< 3
|
Hyperemia
|
Table 3
Transcranial Doppler criteria for other cerebral arteries (except MCA) vasospasm[50]
|
Artery
|
Spasms (MFV, cm/s)
|
|
Possible
|
Probable
|
Definitive
|
|
ICA
|
> 80
|
> 110
|
> 130
|
|
ACA
|
> 90
|
> 110
|
> 120
|
|
PCA
|
> 60
|
> 80
|
> 90
|
|
BA
|
> 70
|
> 90
|
> 100
|
|
VA
|
> 60
|
> 80
|
> 90
|
Abbreviations: ACA, anterior cerebral artery; BA, basilar artery; ICA, internal carotid
artery; MCA, middle cerebral artery; MFV, mean flow velocity; PCA, posterior cerebral
artery; VA, vertebral artery.
To prevent misinterpretation due to conditions like anemia and fever, which can cause
a hyperdynamic state, evaluation with the Lindegaard ratio (LR) is recommended, calculated
by dividing the MFV of the MCA by the MFV of the extracranial ipsilateral internal
carotid artery. An LR greater than 3 indicates vasospasm, and greater than 6 suggests
severe vasospasm.[51] Several publications have shown high specificity of TCD (94–100%), though it has
variable sensitivity (39–96%) compared with DSA in the MCA.[51]
Additionally, the absence of a transtemporal window in approximately 20% of individuals
may limit the effectiveness of the method in some patients.[52] As such, TCD should ideally be used in conjunction with neurological examinations
and other monitoring modalities to improve the accuracy of detecting DCI.
Limitations of the IL include ICA plaques, stenosis, and occlusions, which may influence
the blood flow velocities.[53] Furthermore, monitoring blood flow in the ICA in the ICU setting may be hampered
by the presence of invasive devices and cervical edema. These factors motivated the
study of cerebral vein ultrasound as an alternative for detecting vasospasm.
Mursch et al.[53] prospectively investigated the velocities in the Rosenthal basal vein (RBV), MCA,
and extracranial ICA in 66 patients after spontaneous aSAH, and concluded that in
patients with increased MCA MFV (above 120 cm/s), those who also had increased velocities
in the RBV had a better prognosis. This finding suggests that an increase in the RBV
velocities concomitant with an increase in the MCA reflects a state of hyperemia.
In turn, patients with normal RBV and increased MCA velocities probably have vasospasm.
Another study prospectively evaluated MCA MFV, MCA peak systolic velocity (PSV), IL
using MFV and PSV, and an original arteriovenous index (AVI) between the MCA and the
RBV, using MFV and PSV.[54] They compared both IL and AVI with the gold standard DSA in the diagnosis of vasospasm.
Interestingly, the AVI showed higher accuracy in the diagnosis of arterial vasospasm
compared with the IL. An AVI > 10 (considering MFV) and an AVI > 12 (considering PSV)
provided the highest accuracies of 87 and 86%, respectively. Regarding the IL, the
accuracy was higher using a threshold of > 3 for MFV (84%) and for PSV (80%).[54]
Computed tomography perfusion
As a relatively fast-performing, minimally invasive, and reasonably cost-effective
method, CTP is used to evaluate both macro and microvascularization, as well as cerebral
tissue perfusion ([Figure 4]). Therefore, it plays an important role in predicting and diagnosing DCI, especially
for patients with poor grade aSAH who are sedated or comatose, where it is difficult
to rely only on serial clinical examinations. Several studies have demonstrated that
an increase in the mean transit time (MTT) above 6.5 seconds, and reduction of CBF
below 25 mL/100 g/minute, in the appropriate context, has high accuracy and negative
predictive value.[47]
[55] A systematic review of 882 patients confirmed these results.[56]
Figure 4 Delayed cerebral ischemia on computed tomography with perfusion images. (A) Computed tomography with perfusion (CTP) showing prolonged mean transit time (MTT),
time to detection (TTD) and reduced cerebral blood flow (CBF) in right parietal lobe,
with preserved cerebral blood volume (CBV) in a patient with DCI after aSAH. (B) Brain computed tomography showing definitive cerebral infarction due to DCI at left
frontal and parietal lobes in another patient with aSAH (anterior cerebral artery
aneurysm). (C) Cerebral infarction associated with DCI at anterior cerebral arteries territories
bilaterally (anterior cerebral artery aneurysm).
Continuous electroencephalography
The use of cEEG provides noninvasive, continuous, real-time data on cortical activity,
with well-established sensitivity to ischemia. Claassen et al. first demonstrated
that a reduction in the α/delta ratio of more than 10% from baseline, and decreased
α variability, were the parameters best correlated with early, reversible stages of
DCI.[57]
In a retrospective study, cEEG monitoring detected DCI and cerebral vasospasm with
specificities of 82.9% (95%CI: 66.4–93.4%) and 94.4% (95%CI: 72.7–99.9%), respectively.[58] Additionally, enhanced delta patterns, epileptiform activity, and nonconvulsive
status epilepticus are all associated with poor outcomes.[59]
To improve the exam's quality, researchers have utilized relative rather than absolute
parameters, exclusively artifact-free monitoring, as well as longer monitoring times.
Near-infrared spectroscopy
Near-infrared spectroscopy (NIRS) is another continuous, noninvasive examination technique
that provides information about CBF by estimating intracerebral oxygen saturation.
A decrease in regional cerebral oxygen saturation is correlated with ischemia, being
a promising tool for DCI detection. However, more studies are needed to confirm the
adequate thresholds for diagnosis and triggers for intervention.[60]
Invasive monitoring
Intracranial hypertension is a complication of aSAH associated with alterations in
cerebral metabolism leading to ischemia; thus, it requires aggressive treatment. In
several neurocritical care patients, maintaining cerebral perfusion pressure (CPP)
above 70 mmHg, using invasive intracranial pressure (ICP) and arterial pressure catheters,
has been associated with a reduced risk of tissue hypoxia and brain injury.[36]
Furthermore, CMD offers insight into the composition of cerebral interstitial fluid
and metabolism. A lactate/pyruvate ratio (LPR) > 40 or glucose < 0.5 mM indicates
an energy crisis and subsequent hypoxemia/ischemia.[61] Therefore, LPR is considered a warning sign for increased surveillance. This metabolic
deterioration can precede cerebral infarction by several hours and is specific for
DCI, with lactate > 4 mmol being highly specific.[36]
[61]
The PbTiO2 values can be interpreted as a marker for regional cerebral blood flow,
reflecting the balance between oxygen supply, diffusion, and consumption. Cerebral
hypoxia (PtiO2 < 20 mm Hg) indicates regional ischemia in SAH patients and suggests
intervention together with other parameters.[47]
[59]
Recent guidelines emphasize the lack of evidence to determine ideal triggers for intervention.[5]
[6] It is important to highlight that TCD, CTA, and DSA identify vasospasm but not DCI.
Therefore, ideally, multimodal monitoring methods should be integrated with clinical
practice to enhance diagnosis and management ([Table 4]).
Table 4
Diagnosis modalities for delayed cerebral ischemia and cerebral vasospasm after poor-grade
SAH[47]
|
Complementary Exam
|
Pathological alteration
|
|
Transcranial Doppler ultrasound
|
MCA MFV > 120cm/s, LR > 3: vasospasm
MCA MFV > 200cm/s or LR > 6: severe vasospasm
|
|
Digital subtraction angiography
|
> 70% narrowing: severe vasospasm
|
|
CTA
|
Arterial narrowing
|
|
CTP imaging
|
MTT > 6.5 seconds; reduction of CBF < 25mL/100 g/minute; or 1.5-fold prolongation
compared with baseline indicative for DCI
|
|
Continuous electroencephalography
|
Alpha/delta ration < 50%, reduction in α variability, epileptiform discharges, no
reactivity
|
|
PbtiO2
|
< 20 mmHg
|
|
Cerebral microdialysis
|
LPR > 40
glucose < 0.5 mM
|
|
Cerebral blood flow monitor
|
< 20 mL/100 g/minute
|
Abbreviations: CBF, cerebral blood flow; CTA, computed tomography angiography; CTP,
computed tomography with perfusion; DCI, delayed cerebral ischemia; LPR, lactate pyruvate
ratio; MCA, middle cerebral artery; MFV, mean flow velocity; MTT, mean transit time;
PbTiO2, partial brain tissue oxygenation; SAH, subarachnoid hemorrhage.
MANAGEMENT OF DCI
Once the diagnosis of DCI is established, prompt interventions are warranted to prevent
cerebral infarction and permanent neurological deficits. Guidelines and observational
studies from high-income countries show that treating these patients in high-volume
centers, with at least 35 aSAH cases per year—ideally 60 or more—leads to improved
outcomes.[5] Initially, volume expansion with crystalloids (bolus of 15 mL/kg of isotonic saline)
should be administered targeting euvolemia, with potential increase in cardiac output,
cerebral perfusion, and cerebral oxygenation analyzed by PbtiO2.[62]
[63] The so-called “Triple H Therapy”, composed of hypervolemia, arterial hypertension,
and hemodilution, is contraindicated.[63]
Hemoglobin optimization
Current guidelines recommend a hemoglobin (Hb) goal above 7 g/dL for patients with
aSAH.[6] However, the appropriate threshold for patients with active DCI is unclear. Transfusion
of red blood cells increased the partial pressure of PbTiO2 in low-grade aSAH patients
with a baseline Hb of 8 g/dL.[64] Conversely, nonguided blood transfusions can lead to complications, poor outcomes,
and increased mortality in aSAH patients.
A recent meta-analysis indicates that even mild anemia correlates with cerebral infarction
and poor outcomes, particularly with Hb levels between 9 and 10 g/dL.[65] The “Transfusion Strategies in Acute Brain Injured Patients (TRAIN)” trial compared
liberal versus restrictive transfusion approaches (maintaining Hb > 9 or > 7 g/dL,
respectively) in SAH and traumatic brain injury patients, showing better neurological
outcomes in the liberal group.[66] On the other hand, the “Aneurysmal Subarachnoid Hemorrhage – A Red Blood Cell Transfusion
and Outcome (SAHaRA)” study showed that a liberal transfusion strategy (mandatory
at Hb ≤ 10 g/dL) did not result in a lower risk of an unfavorable neurologic outcome
at 12 months compared with a restrictive strategy (optional at Hb ≤ 8 g/dL).[67]
Hemodynamic augmentation
The rationale of hemodynamic augmentation using induced hypertension (IH) or inotropic
drugs is to improve CBF and CPP, targeting improvement of cerebral perfusion.
Induced hypertension
Regarding IH, norepinephrine is widely preferred due to its ability to stimulate α
and β receptors, resulting in lower rates of tachycardia and a more predictable hemodynamic
response, comparable to phenylephrine.[68]
[69] Vasopressin is reserved for refractory cases requiring multiple vasoactive drugs.
Also, IH is safe in unruptured aneurysms.[70]
Nonetheless, there is still a lack of robust evidence regarding IH in preventing and
treating DCI. In an observational retrospective trial, Haegens et al.[71] demonstrated the effectiveness of IH in preventing DCI-related cerebral infarction
in patients with clinical DCI. The authors[71] included479 patients with clinical signs, and 20% of patients treated with IH developed
a DCI-related cerebral infarct, compared with 33% in the no-IH cohort, with statistical
significance. Also, IH prevented poor outcome. The RCT “Hypertension Induction in
the Management of Aneurysmal Subarachnoid Haemorrhage with Secondary Ischemia (HIMALAIA)”
compared the efficacy and safety between IH and no-IH in patients with clinical symptoms
of DCI, with a primary outcome of 90-day mRS. However, it was prematurely stopped
due to slow recruitment (only 41 patients included) and concerns about the treatment's
adverse events.[72]
A suggested approach is starting norepinephrine in those with DCI who do not have
a basal elevated SBP and/or contraindications (such as a recent myocardial infarction,
decompensated congestive heart failure, and pulmonary edema), with an initial target
of 160 to 180 mmHg. Neurological reevaluations should be done in 30 minutes intervals
and, in cases without improvement, SBP should be augmented up to 220 mmHg, or until
there are collateral effects. If there is improvement, that target should be maintained
for 48 hours and slowly tapered, based on neurological exams or ancillary tests. If
there is recrudescence of symptoms, therapy should resume ([Figure 5]).[36]
Abbreviations: BP, blood pressure; DCI, delayed cerebral ischemia; HI, Hypertension induction; MAP,
mean arterial pressure; SBP, systolic blood pressure. Figure 5 Proposed hemodynamic augmentation strategies for treatment of DCI.[74]
Inotropic therapy
Milrinone, a phosphodiesterase III inhibitor, exhibits positive inotropic and direct
vasodilatory effects. It is administered intraarterially, followed by continuous IV
infusion.[73] Optimal administration should ideally be paired with hemodynamic monitoring, either
minimally invasive or invasive (e.g., transpulmonary thermodilution or pulmonary artery
catheter), targeting a cardiac output above 4 L/min/m[2].[36] In patients with high-grade aSAH, goal-directed hemodynamic is more effective than
nontargeted management.[62]
In a retrospective case series, Lannes et al.[73] tested the effectiveness of milrinone in reversing DCI and found aSAH-associated
vasospasm resolution following the treatment. The study[73] included 88 patients diagnosed from 1999 to 2006, and favorable functional outcomes
were achieved in 75% (mRS 0–2). Notably, most participants had low-grade aSAH and
needed norepinephrine to maintain SBP and CCP. The Montreal Neurological Hospital's
protocol was followed, administering a milrinone bolus of 0.1 to 0.2 mg/kg and continuous
infusion of 0.75 to 1.25 mcg/kg/min. When BP decreased below baseline, norepinephrine
was used to restore it to previous levels, maintaining mean arterial pressure (MAP)
at 90 mmHg. The dosage was progressively increased if symptoms did not improve, with
reassessments every 30 minutes ([Figure 5]).[73]
The predictors of refractory vasospasm and DCI despite IV milrinone treatment were
investigated retrospectively in a cohort predominantly with high-grade aSAH. Intravenous
milrinone was initiated at a dose of 0.25 to 2.5 mcg/kg/min, following diagnosis.[75] Only 21% developed myocardial infarction, and 19% required endovascular rescue therapy
(ERT). Moreover, 65% of patients achieved favorable outcomes (mRS 0–2), highlighting
milrinone's potential benefits.
The MILRISPASM, a controlled observational study, compared patients with cerebral
vasospasm after aSAH who were treated with intravenous milrinone (0.5 µg/kg/min as
part of a strict protocol) and IH, to a historical control group that received only
IH. Patients treated with milrinone exhibited lower rates of ERT and cerebral infarctions,
with better functional outcomes (mRS 0–1).[76] Another retrospective study showed improvement in DCI assessed by TCP after milrinone
therapy.[77]
A recent publication examined sonographic and clinical outcomes during DCI treatment
with either IH or milrinone. The TCD was performed immediately before (t0), and at
45 (t1) and 90 minutes (t2) after therapy initiation, measuring and comparing mean
BFV and their kinetics over time. The National Institutes of Health Stroke Scale (NIHSS)
and Glasgow coma scale were also evaluated at these prespecified intervals. The analysis
of 27 DCI events and 63 spastic vessels demonstrated a significant time-dependent
decrease in BFV with these treatments. When comparing therapies, milrinone effectively
reduced cerebral BFV, and norepinephrine did not. Clinical improvement was observed
with both strategies. This study reinforces milrinone's potential to improve vasospasm,
while IH may enhance CBF through collateral vessels. It also confirms that, depending
on the therapeutic strategy, clinical and sonographic improvements may not necessarily
coincide.[51]
Endovascular rescue therapy
For ERT, both intra-arterial (IA) vasodilating agents and balloon angioplasty (BA)
are used.[48] Specific criteria for intervention triggers for ERT are not well established, but
they are most used in cases of refractoriness in optimized clinical therapy. In a
large observational cohort study, involving over 100,000 patients from an aSAH database,
between 2009 and 2018, ERT was linked to lower rates of in-hospital mortality and
poor discharge outcomes.[78]
Intra-arterial vasodilators
Cerebral IA vasodilators include fasudil, verapamil, nicardipine, nimodipine, and
milrinone.[48] An international online survey with 201 physicians showed that IA nimodipine was
the preferred drug.[79] The advantages of IA vasodilators include their diffuse effect and favorable safety
profile. Therefore, its pharmacotherapy is recommended for mild to moderate vasospasm
(grade I: 0–25% narrowing; and II: 26–50%), particularly in more distal vessels. The
disadvantages of IA vasodilators are the shorter duration of effect, potential to
increase intracranial pressure due to cerebral vasodilation, and the risk of arterial
hypotension, which demands active clinical monitoring during and after the procedure.[48]
Balloon angioplasty
This technique promotes vasodilation through mechanical stretching of the vascular
endothelium and disrupting the smooth muscle cells and extracellular matrix of arteries
to facilitate increased CBF.[80] It is preferred for treating severe focal vasospasm (grade III: 50–75%; and IV: > 75%)
involving proximal intracranial vessels, requiring a minimum vessel diameter of 2
to 2.5 mm.[48]
“The Balloon Prophylaxis of Aneurysmal Vasospasm” randomized controlled trial compared
prophylactic balloon angioplasty in patients with low-grade aSAH within the first
96 hours and indicated a lower incidence of infarction associated with DCI and neurological
deficits in the treated cohort. However, four patients experienced catastrophic vessel
rupture, making the study negative and the intervention contraindicated as a preventive
strategy.[80] Although it is considered the most effective vasospasm therapy in neurology, caution
and experience with balloons are necessary.
In summary, guidelines recommend:[2]
-
In patients with aSAH and symptomatic vasospasm, elevating systolic BP values may
reduce the progression and severity of DCI;
-
Use of IA vasodilators and/or cerebral angioplasty may be reasonable in patients with
severe vasospasm;
-
In patients with aSAH at risk of DCI, prophylactic hemodynamic augmentation should
not be performed to reduce iatrogenic patient harm (HOH).
Notably, there is no robust evidence to create recommendations about how to initiate
and withdraw IH, and asymptomatic vasospasm in awake patients should not be treated
due to potential complications of hemodynamic augmentation. Rather, its detection
should lead to increased vigilance and closer neuromonitoring. Milrinone may be useful
in refractory cases, and ERT may be employed at any stage of treatment, depending
on patients' characteristics and vasospasm location.[36]
[48]
In conclusion, DCI is a critical determinant of long-term outcomes for patients following
aSAH. It demands the implementation of close, multimodal neuromonitoring to ensure
accurate diagnosis. This condition must be treated as a neurological emergency, requiring
the immediate initiation of interventions such as maintaining euvolemia, managing
intracranial hypertension, and employing endovascular therapy for refractory cases.
We must address several questions and evidence gaps including intervention triggers,
blood pressure targets, the efficacy of hemodynamic augmentation, intrathecal vasodilators
antiplatelet therapy or low-dose heparin to prevent microthrombosis, oxygen and microdialysis
targets to improve brain perfusion, inhalational anesthetics as neuroprotective agents,
stellate ganglion block to prevent vasospasm, optimal hemoglobin threshold, CSF drainage
efficacy, and the most effective prevention strategies, among others. These efforts
can significantly reduce the incidence of DCI and enhance long-term patient outcomes.
Bibliographical Record
Ingrid Pereira Marques, Carolina Rouanet Cavalcanti de Albuquerque, Natalia Vasconcellos
de Oliveira Souza, João Brainer Clares de Andrade, Gisele Sampaio Silva, Pedro Kurtz.
Delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage: a narrative review.
Arq Neuropsiquiatr 2025; 83: s00451809885.
DOI: 10.1055/s-0045-1809885
