Aneurysmal Subarachnoid Hemorrhage and Early Brain Injury: A New Pathophysiological Perspective

Abstract

Non-traumatic subarachnoid hemorrhage is a devastating neurological emergency, the main cause of which is aneurysmal rupture. The treatment of the aneurysm, whether microsurgical or endovascular, is essential for the recovery of these patients, however, a series of pathophysiological events in the days following the bleeding cause great damage to the brain tissue. For many years efforts have been focused on the prevention and treatment of cerebral vasospasm, which is believed to be the cause of late cerebral ischemia. However, new pathophysiological perspectives point to a series of events that begin immediately after bleeding, known as early brain injury, mainly involving brain microvascular dysfunction, cortical spreading depolarizations and neuroinflammation, which we discuss below.

1. Introduction

Non-traumatic subarachnoid hemorrhage is a devastating neurological emergency, resulting from the rupture of cerebral aneurysms in most cases. It accounts for about 5–7% of strokes each year, with high morbidity and mortality [1]. Approximately 10% of patients die before receiving medical care [2]. This occurs due to a metabolic collapse caused by the sudden and sustained increase in intracranial pressure (ICP), reducing cerebral perfusion and causing global cerebral ischemia. For those who do receive medical care, about 25% die in the first 24 to 72 hours, with the level of consciousness, hemorrhage volume and neurological deficits on admission being determinants for the prognosis. In this time window, a series of pathophysiological events are known as early brain injury (EBI) [3], including neuroinflammation, microvascular dysfunction and cortical spreading depolarizations (SD), which recent studies point out as main events related to worse prognosis in aneurysmal subarachnoid hemorrhage (SAH). Survivors of this period are still subject to secondary brain injuries resulting from vasospasm of medium and large caliber intracranial arteries, in up to 70% of patients, and even delayed cerebral ischemia (DCI), in 30% of cases [1]. Typically, vasospasm has its highest incidence between the third and fourteenth day after the ictus, occurring due to the imbalance between vasodilator and vasoconstrictor factors in the subarachnoid space, including an increase in the concentration of bilirubin oxidation products, formed from hemoglobin [4]. New focal neurological deficits or a two-point drop on the Glasgow Coma Scale (GCS), within an hour, in turn characterizes DCI [5].

For many years, it was believed that the worst neurocognitive outcomes of SAH occurred as a result of distal hypoperfusion caused by vasospasm of the cerebral arteries, therefore, efforts were focused on the treatment of this entity [6]. Although not yet fully elucidated, the pathogenesis of vasospasm is correlated with high concentrations of Endothelin-1 (ET-1) in the cerebrospinal fluid (CSF). This peptide, made up of 21 amino acids, synthesized by the endothelium, has a potent vasoconstrictor action, also present in other non-neurological pathologies, such as infectious diseases, pulmonary arterial hypertension and neoplasms. In the brain, ET-1 activates ET-_A receptors, leading to increased cellular calcium influx into smooth muscle, resulting in vasoconstriction [7]. Clazosentan, a highly selective ET-A receptor antagonist, has studied as a treatment for vasospasm and DCI. A double-blind, randomized, phase 2 study of 413 patients showed a reduction in the incidence of angiographic vasospasm from 66% in the placebo group to 23% in the control group (65% risk reduction - 95% CI, P < 0.0001) [8], however, the benefit in morbidity and mortality related to vasospasm, evaluating the functional outcome in patients was not confirmed by the subsequent phase 3 study [9]. These results served as a turning point in the field of post-SAH brain injuries, highlighting the importance of researching other pathophysiological mechanisms [10]. From then on, a series of studies were concerned with investigating the mechanisms in the early stage of the disease, with great advances in EBI, which is the focus of the study in this chapter.

2. EBI mechanisms

Brain damage in SAH begins immediately after bleeding. A sudden increase in ICP, caused by intracranial bleeding, causes a reduction in cerebral perfusion and impairment of brain hemodynamics, impairing the mechanisms of self-regulation of brain blood flow, which results in ischemia of the brain tissue, triggering a series of events called “ischemic cascade” [11]. During oxygen and glucose deprivation, there is a reduction in the production of adenosine triphosphate (ATP) and consequent reduction in the activity of Na⁺/K⁺-ATPase, with loss of cellular ionic homeostasis, with an increase in the concentration of intracellular Na⁺ and cytotoxic edema. The exit of K⁺ from neurons and the entry of Na⁺ into the cytoplasm cause cell depolarization. Depolarized neurons are unable to restore their action potentials and die within minutes as a result of energy collapse and a cycle of loss of ionic homeostasis, cytotoxic edema, proteolysis and disintegration of cell membranes [12]. In addition, anaerobic metabolism in areas of greater energy susceptibility generates tissue acidosis, with breakdown of the blood-brain barrier (BBB) and dilation of pre-capillary arterioles, generating vasogenic edema, increased brain volume and ICP, reduced brain perfusion pressure and encephalic hemodynamic collapse.

Concomitantly to this, after bleeding, platelets activated, leading to platelet aggregation and formation of microthrombosis in the microcirculation (which may contain red blood cells and white blood cells in their composition). This causes increased cerebrovascular resistance and perfusion impairment. The reduction in cerebral flow generates blood stasis and the propagation of intravascular microthrombosis, perpetuating the process of tissue ischemia [13].

In addition to perfusion impairment, “non-ischemic” mechanisms are related to EBI, such as energy/mitochondrial dysfunction caused by SD and, finally, a complex and wide range of reactions caused by microglial activation, culminating in neurinflammation (Figure 1) [11].

Figure 1.

Complex pathophysiological mechanisms that contribute to EBI after SAH. Reprinted from Rass et al., current neurology and neuroscience, 2019 [11].

2.1 Microvascular dysfunction

The brain is a high compliance organ, and it is estimated that cerebral vascular resistance is largely regulated by precapillary arterioles, with a central role in the control of brain hemodynamics. Angiographic vasospasm, observed in up to 70% of patients with SAH, refers to a reduction in the caliber of large vessels, not being correlated in the same proportion to brain tissue ischemia, since the incidence of DCI is observed in half of patients with vasospasm angiographic and occurs even in patients without vasospasm [14]. Such evidence suggests the increasing participation of microvascular dysfunction in brain damage.

At this moment, the only drug that has shown to be beneficial in reducing DCI has been the calcium channel blocker nimodipine. It has no significant effect on vasospasm; however, it inhibits vasoconstriction at the level of precapillary arterioles, suggesting that targeting microvascular dysfunction can improve results in SAH [15]. Some experimental studies with animal models of SAH have investigated microvascular reactivity with direct observation of vessels in vivo, as well as the behavior of pial arterioles against vasoactive agents, including adenosine, acetylcholine, carbon dioxide and nitric oxide (NO). In rodents, great damage to microvascular reactivity was observed in the face of these interventions [16, 17].

Another aspect of the microvasculature dysfunction involved in the pathogenesis of cerebral ischemia in SAH is microthrombosis. It is commonly found throughout the brain after SAH, with a greater presence in areas of arteriolar constriction [18] in a pattern called “string of pearls.” The main findings of the study by Friedrich et al., with a model of SAH in rodents, are that more than 70% of the arterioles derived from the middle cerebral artery (MCA) presented “string of pearls” constrictions in the first 3 hours after SAH and that these constrictions persisted for at least 3 days after the hemorrhage, suggesting that the vessels become spastic. The arteriolar diameter was reduced by up to 50% and the more constricted the arterioles were, the more often the vessel lumen was occluded by microthrombosis (30% of all spastic vessels). Small reductions in arteriolar diameter can significantly reduce blood flow with increasing cerebral vascular resistance (Poiseuille’s law) and microthrombosis can completely disrupt microperfusion, explaining the severe early and cerebral perfusion pressure-independent reduction in cerebral blood flow after SAH [17].

In addition to the mechanical impairment of obstructing blood flow, activation of the coagulation cascade and neuroinflammation are closely linked to each other through a process known as thromboinflammation. A sequence of events that include endothelial adhesion, platelet activation and inflammation, culminate in microthrombosis, ischemia, vasogenic edema and EBI [19]. Platelets amplify the inflammatory cascade. Once activated, platelets display P-selectin and release cytokines that promote leukocyte adhesion and transmigration at sites of platelet deposition. Neutrophils also release factors that promote platelet activation. Activated platelets and inflammatory cells contribute to endothelial rupture, perpetuating the cycle of microthrombosis and inflammation, even far from the site of aneurysm rupture [20]. A study involving 127 patients evaluated platelet activation through the maximum amplitude of thromboelastography (TEG) and inflammation through serum levels of C-reactive protein (CRP) and percentage of neutrophils in venous blood, collected at 0–24, 24–48 and 48–72 hours after aneurysm rupture. They found that patients with high-grade SAH (by Hunt-Hess, NIHSS and GCS scales) had significant elevations in platelet activation and inflammation compared to patients with less severely affected SAH (Hunt-Hess 1–3). Furthermore, there was a “dose-response” effect with incremental increase in platelet activation and inflammation as the Hunt-Hess grade and EBI worsened. Even after controlling for other factors, platelet activation was independently associated with EBI [20].

SAH also generates an imbalance of vasoactive substances, both endogenous vasoconstrictors, such as ET-1, and vasodilators, such as NO [21]. Activated leukocytes in the CSF of patients with SAH synthesize and release ET-1, causing arterial vasoconstriction. As mentioned above, clazosentan, a selective endothelin ET-_A receptor antagonist was able to reduce angiographic vasospasm [9]. The benefits of this intervention, although uncertain so far with the publication of CONSCIUS-2, are still extensively studied, and another phase 3 study (REACT) is underway, which is intended to assess its efficacy and safety [22]. Japan is the first country where its use approved for the prevention of vasospasm, cerebral infarction related to vasospasm and symptoms of cerebral ischemia after SAH [23].

On the other hand, NO is capable of inducing cyclic guanosine monophosphate (cGMP) mediated vasodilation, in addition to being involved in the post-SAH inflammatory response. In vascular endothelial cells, in the presence of oxygen, the nitrogenous guanidino terminal of L-arginine produces the gaseous free radical, NO and L-citrulline in a process catalyzed by the enzyme nitric oxide synthase (NOS). NO crosses the space of the endothelium to the vascular smooth muscle and directly stimulates the soluble guanylate cyclase enzyme and the consequent formation of intracellular cGMP, resulting in the relaxation of vascular smooth muscle cells. Constant levels of NO keep the arteriolar diameter under normal conditions and prevent platelet and leukocyte activation [24, 25]. In the brain, NO production is mainly provided by neural NOS (nNOS) and endothelial NOS (eNOS). In systemic inflammatory processes its synthesis is mediated by inducible NOS (iNOS). Immediately after bleeding, different mechanisms result in reduced NO bioavailability, such as reduced synthesis, uncoupling of eNOS, endothelial damage, increase in NOS inhibitors (dimethylarginine) and NO sequestration by reactive oxygen species [24]. In the inflammatory response, the upregulation of iNOS by microglia and astrocytes can generate increased levels of NO, inflammation and cytotoxicity by distant microvascular uncoupling. This supports the hypothesis that NO can regulate both microvascular function and neuroinflammation [21].

Attempts to restore NO production balance have been effective in experimental models. Drugs such as L-arginine and S-nitrosoglutathione have shown efficacy in improving outcome after SAH in animal models, but have been associated with drops in systemic blood pressure. However, inhaled NO has limited effects on systemic blood pressure and has been shown, in rodents, to reduce the number and severity of microvascular constrictions with subsequent reduction in mortality and improvement in outcomes. In patients with SAH, NO donors, including sodium nitroprusside and transdermal nitroglycerin, have been used. Some studies have shown promise, however, they are underpowered and side effects of systemic hypotension, headache and rebound hypertension limit routine use [26].

Other therapeutic strategies sought to modulate NO production by interfering with microvascular relaxation in other ways. Phosphodiesterase V (PDE-V) is a regulator that inhibits vascular smooth muscle cell relaxation and subsequent vasodilation. PDE-V inhibition using sildenafil has shown promising results in experimental SAH [27]. Another medication with similar action, the phosphodiesterase III (PDE-III) inhibitor milrinone, showed some effectiveness in reducing vasospasm, with a reduction in the need for endovascular angioplasty and improved results [28]. In addition to PDE inhibitors, magnesium sulfate has shown promise in experimental SAH, with reduction of infarct areas, reversal of vasospasm and improvement of cerebral perfusion based on its ability to promote relaxation of vascular smooth muscle cells [29].

2.2 Cortical spreading depolarizations (SD)

SD was discovered in animals by the Brazilian neurophysiologist Aristides Leão. He suggested that SD is involved in both migraine aura and cerebral ischemia in humans [30]. The latter was recently confirmed by the Cooperative Studies on Brain Injury Depolarizations (COSBID) [31]. SD is characterized by a massive depolarization of all types of nerve cells and spreads through the cortex by contiguity with the surrounding brain tissue and is the pathophysiological component of EBI in SAH [32, 33]. SD is present in several diseases, such as migraine with aura, subdural hematoma, intraparenchymal hemorrhage, traumatic brain injury (TBI) and SAH, with a wide spectrum of clinical symptoms, which may or may not generate secondary brain injury, due to the influx of intracellular water and cytotoxic edema [32].

Studies in animal models show that SD is triggered when a sufficiently strong stimulus simultaneously depolarizes a critical minimum volume of brain tissue. The depolarizing stimulus overloads extracellular K⁺ clearance mechanisms, causing extracellular K⁺ to exceed a critical threshold concentration. These thresholds may vary in different species and brain regions, depending on neuronal and excitatory synaptic density, age and other factors. The inciting event causes a sudden drop in membrane resistance through the opening of ion channels. As a result, intracellular and extracellular ions move along their transmembrane concentration gradients. The massive efflux of K⁺ increases the extracellular concentration. To ensure transmembrane ionic balance, excess K⁺ is reciprocated by cellular influx of Na⁺ and Cl⁻ which pulls water, causing cell swelling [32]. Depolarization also triggers Ca²⁺ influx and a more than 10-fold drop in extracellular concentration, which, along with Na⁺ and water influx, leads to the release of many neurotransmitters and neuromodulators within depolarized tissue. Extracellular concentrations of glutamate, aspartate, glycine, gamma-aminobutyric acid (GABA) and taurine increase during SD. The massive increase of these substances is capable of depolarizing neighboring cells and is the critical factor for the contiguous propagation of the depolarizing wave [33]. Elevated extracellular levels of glutamate, a strongly depolarizing excitatory amino acid, further fuel DS and facilitate its propagation by activating N-methyl D-aspartate (NMDA) receptors [34, 35]. In healthy brain tissue, this ionic redistribution is self-limiting. A number of mechanisms, including Na⁺ -K⁺-ATPase, intracellular Ca²⁺ buffering and vascular clearance, help restore homeostasis within minutes. In humans without acute injury, the brain hemodynamic response results only in cerebral hyperemia. However, since these processes are highly dependent on glucose and oxygen, that is, high cerebral metabolic rate [35], in the scenario of acute brain injury, as in SAH, areas that are more susceptible evolve with greater tissue damage and ischemia [36].

The gold standard for SD detection is intracranial electrocorticography (ECoG), with records captured through implanted subdural electrodes [37]. The development of less invasive methods would help expand the number of centers capable of detecting SD. When intracranial ECoG is recorded simultaneously with scalp electroencephalogram (EEG), correlations were found between the two modalities. However, it has not yet been possible to detect SDs with good enough reliability with scalp EEG alone [37].

There is a high association between SD, tissue damage and functional outcomes in SAH. Currently, one of the most discussed therapies has been the use of the NMDA receptor antagonist ketamine [38]. Small retrospective studies support its use in SAH, however, there are no randomized clinical trials that provide robust evidence of its benefit. An interesting retrospective cohort study of 66 SAH patients strongly suggests that its use is beneficial. Thirty-three of 66 patients received ketamine during electrocorticographic neuromonitoring of SD in neurointensive therapy. The decision to administer ketamine depended on the need for deeper sedation, therefore, patients receiving ketamine were expected to have a worse clinical outcome. However, in patients who received ketamine, a significant decrease in the incidence of SD was seen when the infusion was started (p < 0.001). Even if functional outcomes are not evaluated, such work provides the basis for the production of a randomized multicenter study for this intervention [39].

Another promising therapeutic option is the PDE-III inhibitor cilostazol. The proposed mechanism of this medication is to reduce SD, due to better neurovascular coupling and improved cerebral blood flow (CBF). A randomized trial involving 50 patients with SAH on ECoG monitoring found a significant reduction in SD with cilostazol [40]. Such a study also provides support for larger studies.

2.3 Neuroinflammation

Neuroinflammation plays an important role in cell damage after aneurysmal rupture. Several inflammatory mediators, for example, interleukin-1β (IL-1β), interleukin-6 (IL-6) and tumor necrosis factor α (TNFα) are released in both serum and CSF in the first hours after SAH. The products of erythrocyte degradation in the subarachnoid space lead to the accumulation of hemoglobin and its products, which activate the toll-like receptor 4 (TLR4), initiating the inflammatory cascade (Figure 2) [41]. At that moment, one of the most important immune cells of the central nervous system (CNS) is activated: the microglia. They play a fundamental role in the pathophysiological process of EBI, as precursors of neuroinflammation. Activated microglia promote the process of endothelial adhesion, recruiting inflammatory cells to the subarachnoid space (neutrophils and macrophages), leading to BBB injury, apoptosis and neuronal edema, in addition to the production of inflammatory cytokines [42]. In SAH, TLR4 expressed by microglia is activated by ligands such as high-mobility group box protein 1 (HMGB1), heme and methemoglobin and mediates a series of intracellular pathways involving nuclear factor-κB (NF-κB) activation. After stimulation by these molecules, microglia initiate the production of pro-inflammatory cytokines, such as TNF-α, interleukins IL-1β, IL-6, interleukin-8 (IL-8) and interleukin-12 (IL-12), that result in tissue necrosis [43].

Figure 2.

The breakdown of red blood cells causes the release of heme, hemin and methemoglobin. Through interactions with TRL4 in microglia, HMGB1 is increased. This increase leads to downstream activation of NF-κB and the release of pro-inflammatory cytokines. Reprinted from Luck-Wold et al., International Journal of Molecular Sciences, 2016 [41].

IL-1, in particular, increases BBB permeability, glial-mediated neurotoxicity and promotes ischemic changes after SAH in preclinical models [44]. IL-6, on the other hand, has a strong correlation with worse outcomes in SAH. One study was able to correlate elevated serum levels of IL-6 with measures of baseline parameters of brain injury, that is, cerebral edema scores based on imaging studies (Computerized Tomography) and the patient’s respective Hunt and Hess grade [45]. In another study, there was also a correlation between higher serum levels of IL-6 in the CSF in cases with grade 4 bleeding vs. 3 by Fisher and in patients with World Federation of Neurological Societies (WFNS) grade 5 SAH, suggesting that the role of IL-6 signaling in the early inflammatory response is a sensitive biomarker of early brain injury. In addition, high levels of IL-6 for more than 72 hours after the injury are correlated with worse outcomes [11]. Published in 2015, an interesting study based on neuromonitoring with cerebral microdialysis identified higher concentrations of IL-6 in the brain compared to systemic concentrations, corroborating the hypothesis of cerebral compartmentalization. In it, there was still an association of high levels of IL-6 with GCS, metabolic disorder and cerebral perfusion pressure (CPP) lower than 70 mmhg [46].

Other substances involved in the neuroinflammation process are matrix metalloproteinases (MMPs), especially MMP-9. The loss of BBB integrity is related to the upregulation of these proteinases in aHSA and is related to the pathophysiology of EBI [41]. In animal models with rats, increased expression was associated with apoptosis of hippocampal neurons [47]. Increased expression and subsequent activation of MMP-9 may occur in response to reactive oxygen species and pro-inflammatory cytokines such as TNF-a and interleukin-17 (IL-17). The source of MMP-9 in SAH is not well described, but evidence obtained from ischemia-reperfusion models indicates that neutrophils may be the main source of MMP-9 acting on the BBB [48]. MMP-9 can also drive neuroinflammation through activation of pro-inflammatory signals and clotting factors, triggering a positive feedback loop that promotes thromboinflammation and neurotoxicity [49]. The increase of MMP-9 in the CSF by cerebral microdialysis techniques correlates with the extent of EBI, vasospasm and DCI [46].

Faced with this pathophysiological hypothesis, drugs with cerebral anti-inflammatory action were studied to improve the outcome of SAH in the acute phase. A study carried out at the University of Zurich, involving 138 patients, showed that the use of non-steroidal anti-inflammatory drugs (NSAIDs) (acetaminophen, dipyrone, diclofenac and ibuprofen) independently reduced serum levels of IL-6 and CRP, as well as obtained better functional outcomes [50]. Another drug studied in this context was heparin. Despite its anticoagulant effect, this polysaccharide from the glycosaminoglycan family has a pleiotropic effect, showing broad anti-inflammatory and immunomodulatory activity, even at doses that do not cause anticoagulation. An animal study showed that low-dose heparin is able to attenuate neuroinflammation in SAH [51]. Table 1 summarizes the main drugs studied for the treatment of EBI.

Medication	Mechanism	Potential benefit
Clazosentan	Selective antagonist of Endothelin receptors	Vasospasm reduction
Nitric oxide (NO)	Relaxation of endothelial smooth muscle cells via cGMP	Vasodilation
Sildenafil	PDE-V inhibitor	Vasodilation
Cilostazol	PDE-III inhibitor	Neurovascular coupling
Milrinone	PDE-III inhibitor	Vasodilation
Magnesium Sulfate	Blocks intracellular calcium influx and improves blood rheological function	Vasodilation, neuroprotection and increased cerebral blood flow
Ketamine	Blocking NMDA receptors	Reduction of propagation of excitatory stimuli
NSAIDs	Reduction of circulating levels of IL 6 and CRP	Decreased brain inflammatory response
Heparin	Anti-inflammatory and immunomodulatory activity	Decreased brain inflammatory response

Table 1.

Summarizes the main study medications and potential beneficial effects in the treatment of EBI.

3. Conclusion

The pathophysiological knowledge of SAH is of fundamental importance to promote paradigm shifts and new therapeutic proposals for this devastating disease. In the last 20 years, the paradigm that ischemic lesions secondary to aneurysmal rupture started only after the third day fell apart with the emergence of the new concept of EBI. A pathology with enormous morbidity and mortality for years, with great financial and social impact, requires new therapeutic weapons. Clazosentan, ketamine, cilostazol, non-steroidal anti-inflammatory drugs and heparin have shown to be possible options, but they require larger studies to consolidate their applications.

Acknowledgments

I thank everyone who participated in this work, especially my wife, MD Karylla Marques for her unconditional support and Dr. Salomón Rojas, my master and great friend, for teaching me to walk in the paths of neurointensivism.

Conflicts of interest

The authors declare that there are no conflicts of interest of any kind.

Acronyms

ATP

adenosine triphosphate

BBB

blood-brain barrier

CBF

cerebral blood flow

cGMP

cyclic guanosine monophosphate

CNS

central nervous system

CPP

cerebral perfusion pressure

CPR

C-reactive protein

CSF

Cerebrospinal fluid

DCI

delayed cerebral ischemia

EBI

early brain injury

ECoG

electrocorticography

EEG

electroencephalogram

eNOS

endothelial NOS

ET 1

endothelin-1

GABA

gamma-aminobutyric acid

GCS

Glasgow coma scale

HMGB1

high-mobility group box protein 1

ICP

intracranial pressure

IL-1 β

interleukin-1β

IL-12

interleukin-12

IL-17

interleukin-17

IL-6

interleukin-6

IL-8

interleukin-8

iNOS

inducible NOS

MCA

middle cerebral artery

MMP-9

metalloproteinases-9

MMPs

metalloproteinases

NF-κB

nuclear factor-κB

NIHSS

National institutes of health stroke scale

NMDA

N-methyl D-aspartate

nNOS

neural NOS

NO

Nitric oxide

NOS

Nitric oxide synthase

NSAIDs

non-steroidal anti-inflammatory drugs

PDE-III

phosphodiesterase III

PDE-V

phosphodiesterase V

SAH

subarachnoid hemorrhage

SD

spreading depolarizations

TBI

traumatic brain injury

TEG

thromboelastography

TNFα

tumor necrosis factor α

TRL4

toll-like receptor 4

WFNS

world federation of neurological societies